{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 9495,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.5241050124168396,
      "epoch": 0.000526592943654555,
      "grad_norm": 111.6484146118164,
      "learning_rate": 0.0,
      "loss": 7.630620002746582,
      "mean_token_accuracy": 0.16127368062734604,
      "num_tokens": 12336.0,
      "step": 1
    },
    {
      "entropy": 1.5055236518383026,
      "epoch": 0.00105318588730911,
      "grad_norm": 111.12712097167969,
      "learning_rate": 5.012531328320802e-07,
      "loss": 7.332417964935303,
      "mean_token_accuracy": 0.17185717076063156,
      "num_tokens": 24672.0,
      "step": 2
    },
    {
      "entropy": 1.5026952624320984,
      "epoch": 0.001579778830963665,
      "grad_norm": 112.11380767822266,
      "learning_rate": 1.0025062656641603e-06,
      "loss": 7.473679542541504,
      "mean_token_accuracy": 0.1643187664449215,
      "num_tokens": 37008.0,
      "step": 3
    },
    {
      "entropy": 1.5194158256053925,
      "epoch": 0.00210637177461822,
      "grad_norm": 114.23673248291016,
      "learning_rate": 1.5037593984962406e-06,
      "loss": 7.426502704620361,
      "mean_token_accuracy": 0.1571561936289072,
      "num_tokens": 49344.0,
      "step": 4
    },
    {
      "entropy": 1.496931940317154,
      "epoch": 0.0026329647182727752,
      "grad_norm": 118.40357208251953,
      "learning_rate": 2.0050125313283207e-06,
      "loss": 7.894323348999023,
      "mean_token_accuracy": 0.1548401154577732,
      "num_tokens": 61680.0,
      "step": 5
    },
    {
      "entropy": 1.5049863159656525,
      "epoch": 0.00315955766192733,
      "grad_norm": 104.21913146972656,
      "learning_rate": 2.506265664160401e-06,
      "loss": 6.593452453613281,
      "mean_token_accuracy": 0.20564104989171028,
      "num_tokens": 74016.0,
      "step": 6
    },
    {
      "entropy": 1.4925725162029266,
      "epoch": 0.003686150605581885,
      "grad_norm": 114.10234069824219,
      "learning_rate": 3.007518796992481e-06,
      "loss": 7.3693318367004395,
      "mean_token_accuracy": 0.16597539372742176,
      "num_tokens": 86352.0,
      "step": 7
    },
    {
      "entropy": 1.5191607773303986,
      "epoch": 0.00421274354923644,
      "grad_norm": 115.99723815917969,
      "learning_rate": 3.5087719298245615e-06,
      "loss": 7.358461856842041,
      "mean_token_accuracy": 0.16679511591792107,
      "num_tokens": 98688.0,
      "step": 8
    },
    {
      "entropy": 1.5183644592761993,
      "epoch": 0.004739336492890996,
      "grad_norm": 106.1409912109375,
      "learning_rate": 4.010025062656641e-06,
      "loss": 6.668917655944824,
      "mean_token_accuracy": 0.18950679898262024,
      "num_tokens": 111024.0,
      "step": 9
    },
    {
      "entropy": 1.5116282999515533,
      "epoch": 0.0052659294365455505,
      "grad_norm": 110.57721710205078,
      "learning_rate": 4.511278195488722e-06,
      "loss": 6.99241304397583,
      "mean_token_accuracy": 0.18210845813155174,
      "num_tokens": 123360.0,
      "step": 10
    },
    {
      "entropy": 1.5269342362880707,
      "epoch": 0.005792522380200105,
      "grad_norm": 120.43307495117188,
      "learning_rate": 5.012531328320802e-06,
      "loss": 7.586757659912109,
      "mean_token_accuracy": 0.1462114453315735,
      "num_tokens": 135696.0,
      "step": 11
    },
    {
      "entropy": 1.5252384543418884,
      "epoch": 0.00631911532385466,
      "grad_norm": 111.8587875366211,
      "learning_rate": 5.5137844611528826e-06,
      "loss": 7.363627910614014,
      "mean_token_accuracy": 0.17107644863426685,
      "num_tokens": 148032.0,
      "step": 12
    },
    {
      "entropy": 1.5224651992321014,
      "epoch": 0.006845708267509215,
      "grad_norm": 107.8873291015625,
      "learning_rate": 6.015037593984962e-06,
      "loss": 6.356827735900879,
      "mean_token_accuracy": 0.19671257957816124,
      "num_tokens": 160368.0,
      "step": 13
    },
    {
      "entropy": 1.5416274964809418,
      "epoch": 0.00737230121116377,
      "grad_norm": 115.19085693359375,
      "learning_rate": 6.516290726817042e-06,
      "loss": 6.221941947937012,
      "mean_token_accuracy": 0.2142542451620102,
      "num_tokens": 172704.0,
      "step": 14
    },
    {
      "entropy": 1.5107212960720062,
      "epoch": 0.007898894154818325,
      "grad_norm": 110.73648834228516,
      "learning_rate": 7.017543859649123e-06,
      "loss": 6.397078514099121,
      "mean_token_accuracy": 0.20885847136378288,
      "num_tokens": 185040.0,
      "step": 15
    },
    {
      "entropy": 1.5597798526287079,
      "epoch": 0.00842548709847288,
      "grad_norm": 99.72173309326172,
      "learning_rate": 7.518796992481203e-06,
      "loss": 5.646764278411865,
      "mean_token_accuracy": 0.24349310994148254,
      "num_tokens": 197376.0,
      "step": 16
    },
    {
      "entropy": 1.5534575581550598,
      "epoch": 0.008952080042127435,
      "grad_norm": 104.43334197998047,
      "learning_rate": 8.020050125313283e-06,
      "loss": 5.892775535583496,
      "mean_token_accuracy": 0.22073492035269737,
      "num_tokens": 209712.0,
      "step": 17
    },
    {
      "entropy": 1.5676631033420563,
      "epoch": 0.009478672985781991,
      "grad_norm": 99.27777099609375,
      "learning_rate": 8.521303258145363e-06,
      "loss": 5.239416599273682,
      "mean_token_accuracy": 0.26043422892689705,
      "num_tokens": 222048.0,
      "step": 18
    },
    {
      "entropy": 1.5712612569332123,
      "epoch": 0.010005265929436546,
      "grad_norm": 94.56501770019531,
      "learning_rate": 9.022556390977444e-06,
      "loss": 5.018033981323242,
      "mean_token_accuracy": 0.2874941825866699,
      "num_tokens": 234384.0,
      "step": 19
    },
    {
      "entropy": 1.5856993794441223,
      "epoch": 0.010531858873091101,
      "grad_norm": 103.30012512207031,
      "learning_rate": 9.523809523809523e-06,
      "loss": 5.112401485443115,
      "mean_token_accuracy": 0.26686083525419235,
      "num_tokens": 246720.0,
      "step": 20
    },
    {
      "entropy": 1.542406052350998,
      "epoch": 0.011058451816745656,
      "grad_norm": 73.81751251220703,
      "learning_rate": 1.0025062656641604e-05,
      "loss": 3.298715353012085,
      "mean_token_accuracy": 0.42641731351614,
      "num_tokens": 259056.0,
      "step": 21
    },
    {
      "entropy": 1.5376673340797424,
      "epoch": 0.01158504476040021,
      "grad_norm": 74.847900390625,
      "learning_rate": 1.0526315789473684e-05,
      "loss": 3.5809030532836914,
      "mean_token_accuracy": 0.40752222388982773,
      "num_tokens": 271392.0,
      "step": 22
    },
    {
      "entropy": 1.548084020614624,
      "epoch": 0.012111637704054766,
      "grad_norm": 63.552825927734375,
      "learning_rate": 1.1027568922305765e-05,
      "loss": 3.100147247314453,
      "mean_token_accuracy": 0.45311493426561356,
      "num_tokens": 283728.0,
      "step": 23
    },
    {
      "entropy": 1.5953525304794312,
      "epoch": 0.01263823064770932,
      "grad_norm": 60.015655517578125,
      "learning_rate": 1.1528822055137844e-05,
      "loss": 2.783557176589966,
      "mean_token_accuracy": 0.47465962916612625,
      "num_tokens": 296064.0,
      "step": 24
    },
    {
      "entropy": 1.5308564603328705,
      "epoch": 0.013164823591363875,
      "grad_norm": 40.24797821044922,
      "learning_rate": 1.2030075187969925e-05,
      "loss": 2.171975612640381,
      "mean_token_accuracy": 0.5734862834215164,
      "num_tokens": 308400.0,
      "step": 25
    },
    {
      "entropy": 1.5149377882480621,
      "epoch": 0.01369141653501843,
      "grad_norm": 31.702476501464844,
      "learning_rate": 1.2531328320802006e-05,
      "loss": 1.907788634300232,
      "mean_token_accuracy": 0.6134172528982162,
      "num_tokens": 320736.0,
      "step": 26
    },
    {
      "entropy": 1.4917882978916168,
      "epoch": 0.014218009478672985,
      "grad_norm": 20.41485023498535,
      "learning_rate": 1.3032581453634085e-05,
      "loss": 1.6917057037353516,
      "mean_token_accuracy": 0.6387842893600464,
      "num_tokens": 333072.0,
      "step": 27
    },
    {
      "entropy": 1.4882407486438751,
      "epoch": 0.01474460242232754,
      "grad_norm": 15.065500259399414,
      "learning_rate": 1.3533834586466165e-05,
      "loss": 1.4968407154083252,
      "mean_token_accuracy": 0.6677049100399017,
      "num_tokens": 345408.0,
      "step": 28
    },
    {
      "entropy": 1.4370096921920776,
      "epoch": 0.015271195365982097,
      "grad_norm": 10.740882873535156,
      "learning_rate": 1.4035087719298246e-05,
      "loss": 1.3545355796813965,
      "mean_token_accuracy": 0.698340117931366,
      "num_tokens": 357744.0,
      "step": 29
    },
    {
      "entropy": 1.341864824295044,
      "epoch": 0.01579778830963665,
      "grad_norm": 3.3069381713867188,
      "learning_rate": 1.4536340852130325e-05,
      "loss": 1.0850293636322021,
      "mean_token_accuracy": 0.7386660128831863,
      "num_tokens": 370080.0,
      "step": 30
    },
    {
      "entropy": 1.4095737040042877,
      "epoch": 0.016324381253291206,
      "grad_norm": 2.9421474933624268,
      "learning_rate": 1.5037593984962406e-05,
      "loss": 1.1328743696212769,
      "mean_token_accuracy": 0.7404871881008148,
      "num_tokens": 382416.0,
      "step": 31
    },
    {
      "entropy": 1.4024023413658142,
      "epoch": 0.01685097419694576,
      "grad_norm": 2.5040762424468994,
      "learning_rate": 1.5538847117794486e-05,
      "loss": 1.1228951215744019,
      "mean_token_accuracy": 0.7428316622972488,
      "num_tokens": 394752.0,
      "step": 32
    },
    {
      "entropy": 1.3549567759037018,
      "epoch": 0.017377567140600316,
      "grad_norm": 1.7973337173461914,
      "learning_rate": 1.6040100250626565e-05,
      "loss": 1.0077903270721436,
      "mean_token_accuracy": 0.766022652387619,
      "num_tokens": 407088.0,
      "step": 33
    },
    {
      "entropy": 1.3836270570755005,
      "epoch": 0.01790416008425487,
      "grad_norm": 1.7406286001205444,
      "learning_rate": 1.6541353383458648e-05,
      "loss": 1.0222362279891968,
      "mean_token_accuracy": 0.7626231610774994,
      "num_tokens": 419424.0,
      "step": 34
    },
    {
      "entropy": 1.3947510421276093,
      "epoch": 0.018430753027909426,
      "grad_norm": 1.8865187168121338,
      "learning_rate": 1.7042606516290727e-05,
      "loss": 1.0371894836425781,
      "mean_token_accuracy": 0.7562028616666794,
      "num_tokens": 431760.0,
      "step": 35
    },
    {
      "entropy": 1.3465319573879242,
      "epoch": 0.018957345971563982,
      "grad_norm": 1.7011761665344238,
      "learning_rate": 1.7543859649122806e-05,
      "loss": 0.9494946599006653,
      "mean_token_accuracy": 0.7783531844615936,
      "num_tokens": 444096.0,
      "step": 36
    },
    {
      "entropy": 1.4580606818199158,
      "epoch": 0.019483938915218536,
      "grad_norm": 1.9021601676940918,
      "learning_rate": 1.8045112781954888e-05,
      "loss": 1.0398352146148682,
      "mean_token_accuracy": 0.7636931091547012,
      "num_tokens": 456432.0,
      "step": 37
    },
    {
      "entropy": 1.4095810949802399,
      "epoch": 0.020010531858873092,
      "grad_norm": 1.6176142692565918,
      "learning_rate": 1.8546365914786967e-05,
      "loss": 0.9648343324661255,
      "mean_token_accuracy": 0.763045608997345,
      "num_tokens": 468768.0,
      "step": 38
    },
    {
      "entropy": 1.4290408492088318,
      "epoch": 0.020537124802527645,
      "grad_norm": 1.802105188369751,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 0.9583951234817505,
      "mean_token_accuracy": 0.7706119865179062,
      "num_tokens": 481104.0,
      "step": 39
    },
    {
      "entropy": 1.4769366383552551,
      "epoch": 0.021063717746182202,
      "grad_norm": 1.8512225151062012,
      "learning_rate": 1.954887218045113e-05,
      "loss": 1.0114065408706665,
      "mean_token_accuracy": 0.7521056234836578,
      "num_tokens": 493440.0,
      "step": 40
    },
    {
      "entropy": 1.508185237646103,
      "epoch": 0.021590310689836755,
      "grad_norm": 2.076702356338501,
      "learning_rate": 2.0050125313283208e-05,
      "loss": 0.9636508226394653,
      "mean_token_accuracy": 0.7661653459072113,
      "num_tokens": 505776.0,
      "step": 41
    },
    {
      "entropy": 1.459254652261734,
      "epoch": 0.022116903633491312,
      "grad_norm": 2.349229335784912,
      "learning_rate": 2.0551378446115287e-05,
      "loss": 0.9206173419952393,
      "mean_token_accuracy": 0.7662864923477173,
      "num_tokens": 518112.0,
      "step": 42
    },
    {
      "entropy": 1.5238048434257507,
      "epoch": 0.022643496577145865,
      "grad_norm": 2.185512065887451,
      "learning_rate": 2.105263157894737e-05,
      "loss": 0.9314085245132446,
      "mean_token_accuracy": 0.7649829983711243,
      "num_tokens": 530448.0,
      "step": 43
    },
    {
      "entropy": 1.5003608763217926,
      "epoch": 0.02317008952080042,
      "grad_norm": 2.0811667442321777,
      "learning_rate": 2.1553884711779448e-05,
      "loss": 0.8671697974205017,
      "mean_token_accuracy": 0.7754391133785248,
      "num_tokens": 542784.0,
      "step": 44
    },
    {
      "entropy": 1.591546893119812,
      "epoch": 0.023696682464454975,
      "grad_norm": 2.0840373039245605,
      "learning_rate": 2.205513784461153e-05,
      "loss": 0.872833251953125,
      "mean_token_accuracy": 0.7740741670131683,
      "num_tokens": 555120.0,
      "step": 45
    },
    {
      "entropy": 1.6416102647781372,
      "epoch": 0.02422327540810953,
      "grad_norm": 2.033761501312256,
      "learning_rate": 2.255639097744361e-05,
      "loss": 0.8919705152511597,
      "mean_token_accuracy": 0.7629313617944717,
      "num_tokens": 567456.0,
      "step": 46
    },
    {
      "entropy": 1.6393465399742126,
      "epoch": 0.024749868351764088,
      "grad_norm": 1.7740504741668701,
      "learning_rate": 2.3057644110275688e-05,
      "loss": 0.8117689490318298,
      "mean_token_accuracy": 0.7816901206970215,
      "num_tokens": 579792.0,
      "step": 47
    },
    {
      "entropy": 1.6808819770812988,
      "epoch": 0.02527646129541864,
      "grad_norm": 1.4741381406784058,
      "learning_rate": 2.355889724310777e-05,
      "loss": 0.8110437393188477,
      "mean_token_accuracy": 0.7762878388166428,
      "num_tokens": 592128.0,
      "step": 48
    },
    {
      "entropy": 1.7813679873943329,
      "epoch": 0.025803054239073198,
      "grad_norm": 1.135923147201538,
      "learning_rate": 2.406015037593985e-05,
      "loss": 0.8597409129142761,
      "mean_token_accuracy": 0.7689527124166489,
      "num_tokens": 604464.0,
      "step": 49
    },
    {
      "entropy": 1.7125210165977478,
      "epoch": 0.02632964718272775,
      "grad_norm": 1.0037126541137695,
      "learning_rate": 2.456140350877193e-05,
      "loss": 0.7560303807258606,
      "mean_token_accuracy": 0.7888963371515274,
      "num_tokens": 616800.0,
      "step": 50
    },
    {
      "entropy": 1.7505145072937012,
      "epoch": 0.026856240126382307,
      "grad_norm": 0.9906253218650818,
      "learning_rate": 2.506265664160401e-05,
      "loss": 0.7666164636611938,
      "mean_token_accuracy": 0.7822892367839813,
      "num_tokens": 629136.0,
      "step": 51
    },
    {
      "entropy": 1.8206515610218048,
      "epoch": 0.02738283307003686,
      "grad_norm": 1.0637407302856445,
      "learning_rate": 2.556390977443609e-05,
      "loss": 0.7570075988769531,
      "mean_token_accuracy": 0.7857469767332077,
      "num_tokens": 641472.0,
      "step": 52
    },
    {
      "entropy": 1.9179592430591583,
      "epoch": 0.027909426013691417,
      "grad_norm": 0.994309663772583,
      "learning_rate": 2.606516290726817e-05,
      "loss": 0.8264732956886292,
      "mean_token_accuracy": 0.7659910470247269,
      "num_tokens": 653808.0,
      "step": 53
    },
    {
      "entropy": 1.8192859292030334,
      "epoch": 0.02843601895734597,
      "grad_norm": 1.283566951751709,
      "learning_rate": 2.656641604010025e-05,
      "loss": 0.7167642712593079,
      "mean_token_accuracy": 0.7935251444578171,
      "num_tokens": 666144.0,
      "step": 54
    },
    {
      "entropy": 1.8640032410621643,
      "epoch": 0.028962611901000527,
      "grad_norm": 0.993376612663269,
      "learning_rate": 2.706766917293233e-05,
      "loss": 0.732735276222229,
      "mean_token_accuracy": 0.7889553904533386,
      "num_tokens": 678480.0,
      "step": 55
    },
    {
      "entropy": 1.9309577941894531,
      "epoch": 0.02948920484465508,
      "grad_norm": 0.952996551990509,
      "learning_rate": 2.756892230576441e-05,
      "loss": 0.7183045148849487,
      "mean_token_accuracy": 0.7865996956825256,
      "num_tokens": 690816.0,
      "step": 56
    },
    {
      "entropy": 1.9969252347946167,
      "epoch": 0.030015797788309637,
      "grad_norm": 0.9141530990600586,
      "learning_rate": 2.8070175438596492e-05,
      "loss": 0.7127647399902344,
      "mean_token_accuracy": 0.7965414077043533,
      "num_tokens": 703152.0,
      "step": 57
    },
    {
      "entropy": 1.9162320494651794,
      "epoch": 0.030542390731964193,
      "grad_norm": 1.0311601161956787,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.6988133788108826,
      "mean_token_accuracy": 0.7964022308588028,
      "num_tokens": 715488.0,
      "step": 58
    },
    {
      "entropy": 1.9365825355052948,
      "epoch": 0.031068983675618746,
      "grad_norm": 0.8333176374435425,
      "learning_rate": 2.907268170426065e-05,
      "loss": 0.6844547986984253,
      "mean_token_accuracy": 0.8021096438169479,
      "num_tokens": 727824.0,
      "step": 59
    },
    {
      "entropy": 1.9246950447559357,
      "epoch": 0.0315955766192733,
      "grad_norm": 0.8664101362228394,
      "learning_rate": 2.9573934837092732e-05,
      "loss": 0.6528723239898682,
      "mean_token_accuracy": 0.8053467720746994,
      "num_tokens": 740160.0,
      "step": 60
    },
    {
      "entropy": 1.929794728755951,
      "epoch": 0.032122169562927856,
      "grad_norm": 0.9577063918113708,
      "learning_rate": 3.007518796992481e-05,
      "loss": 0.678918719291687,
      "mean_token_accuracy": 0.7989283800125122,
      "num_tokens": 752496.0,
      "step": 61
    },
    {
      "entropy": 1.826391577720642,
      "epoch": 0.03264876250658241,
      "grad_norm": 0.8492108583450317,
      "learning_rate": 3.0576441102756894e-05,
      "loss": 0.6723799109458923,
      "mean_token_accuracy": 0.7983379364013672,
      "num_tokens": 764832.0,
      "step": 62
    },
    {
      "entropy": 1.9241991639137268,
      "epoch": 0.03317535545023697,
      "grad_norm": 0.928464412689209,
      "learning_rate": 3.107769423558897e-05,
      "loss": 0.6530962586402893,
      "mean_token_accuracy": 0.8008216470479965,
      "num_tokens": 777168.0,
      "step": 63
    },
    {
      "entropy": 1.8967448770999908,
      "epoch": 0.03370194839389152,
      "grad_norm": 1.0670896768569946,
      "learning_rate": 3.157894736842105e-05,
      "loss": 0.6292267441749573,
      "mean_token_accuracy": 0.8003135621547699,
      "num_tokens": 789504.0,
      "step": 64
    },
    {
      "entropy": 1.8847980499267578,
      "epoch": 0.034228541337546076,
      "grad_norm": 1.0110414028167725,
      "learning_rate": 3.208020050125313e-05,
      "loss": 0.6133747100830078,
      "mean_token_accuracy": 0.8148374855518341,
      "num_tokens": 801840.0,
      "step": 65
    },
    {
      "entropy": 1.930319905281067,
      "epoch": 0.03475513428120063,
      "grad_norm": 1.023224949836731,
      "learning_rate": 3.258145363408521e-05,
      "loss": 0.6123053431510925,
      "mean_token_accuracy": 0.8167780637741089,
      "num_tokens": 814176.0,
      "step": 66
    },
    {
      "entropy": 1.9471323192119598,
      "epoch": 0.03528172722485519,
      "grad_norm": 1.1810884475708008,
      "learning_rate": 3.3082706766917295e-05,
      "loss": 0.6039740443229675,
      "mean_token_accuracy": 0.8179137706756592,
      "num_tokens": 826512.0,
      "step": 67
    },
    {
      "entropy": 1.8890258967876434,
      "epoch": 0.03580832016850974,
      "grad_norm": 0.9907887578010559,
      "learning_rate": 3.3583959899749374e-05,
      "loss": 0.599247932434082,
      "mean_token_accuracy": 0.8207112550735474,
      "num_tokens": 838848.0,
      "step": 68
    },
    {
      "entropy": 1.9326116144657135,
      "epoch": 0.036334913112164295,
      "grad_norm": 1.0037294626235962,
      "learning_rate": 3.4085213032581453e-05,
      "loss": 0.5630755424499512,
      "mean_token_accuracy": 0.8216072916984558,
      "num_tokens": 851184.0,
      "step": 69
    },
    {
      "entropy": 1.925311028957367,
      "epoch": 0.03686150605581885,
      "grad_norm": 1.1875734329223633,
      "learning_rate": 3.458646616541353e-05,
      "loss": 0.5733011364936829,
      "mean_token_accuracy": 0.8164849728345871,
      "num_tokens": 863520.0,
      "step": 70
    },
    {
      "entropy": 2.051402121782303,
      "epoch": 0.03738809899947341,
      "grad_norm": 1.162881851196289,
      "learning_rate": 3.508771929824561e-05,
      "loss": 0.550703763961792,
      "mean_token_accuracy": 0.8178133964538574,
      "num_tokens": 875856.0,
      "step": 71
    },
    {
      "entropy": 1.9634678363800049,
      "epoch": 0.037914691943127965,
      "grad_norm": 0.9865695834159851,
      "learning_rate": 3.55889724310777e-05,
      "loss": 0.5769374966621399,
      "mean_token_accuracy": 0.8114093840122223,
      "num_tokens": 888192.0,
      "step": 72
    },
    {
      "entropy": 1.9920983016490936,
      "epoch": 0.038441284886782515,
      "grad_norm": 1.0162739753723145,
      "learning_rate": 3.6090225563909776e-05,
      "loss": 0.5818509459495544,
      "mean_token_accuracy": 0.8134536445140839,
      "num_tokens": 900528.0,
      "step": 73
    },
    {
      "entropy": 1.9029254913330078,
      "epoch": 0.03896787783043707,
      "grad_norm": 0.9538004398345947,
      "learning_rate": 3.6591478696741855e-05,
      "loss": 0.5178959369659424,
      "mean_token_accuracy": 0.8257192969322205,
      "num_tokens": 912864.0,
      "step": 74
    },
    {
      "entropy": 1.967795580625534,
      "epoch": 0.03949447077409163,
      "grad_norm": 0.9578227996826172,
      "learning_rate": 3.7092731829573934e-05,
      "loss": 0.5342862010002136,
      "mean_token_accuracy": 0.8278465121984482,
      "num_tokens": 925200.0,
      "step": 75
    },
    {
      "entropy": 1.9465965926647186,
      "epoch": 0.040021063717746184,
      "grad_norm": 1.0253666639328003,
      "learning_rate": 3.759398496240601e-05,
      "loss": 0.5030800104141235,
      "mean_token_accuracy": 0.83804552257061,
      "num_tokens": 937536.0,
      "step": 76
    },
    {
      "entropy": 1.8636894226074219,
      "epoch": 0.040547656661400734,
      "grad_norm": 0.8184434771537781,
      "learning_rate": 3.809523809523809e-05,
      "loss": 0.45680901408195496,
      "mean_token_accuracy": 0.8485544174909592,
      "num_tokens": 949872.0,
      "step": 77
    },
    {
      "entropy": 1.928781270980835,
      "epoch": 0.04107424960505529,
      "grad_norm": 0.8536883592605591,
      "learning_rate": 3.859649122807018e-05,
      "loss": 0.4830615818500519,
      "mean_token_accuracy": 0.8397646248340607,
      "num_tokens": 962208.0,
      "step": 78
    },
    {
      "entropy": 1.8684503734111786,
      "epoch": 0.04160084254870985,
      "grad_norm": 0.868567168712616,
      "learning_rate": 3.909774436090226e-05,
      "loss": 0.49555566906929016,
      "mean_token_accuracy": 0.8393678069114685,
      "num_tokens": 974544.0,
      "step": 79
    },
    {
      "entropy": 1.832389086484909,
      "epoch": 0.042127435492364404,
      "grad_norm": 0.7834892868995667,
      "learning_rate": 3.9598997493734336e-05,
      "loss": 0.4530882239341736,
      "mean_token_accuracy": 0.8548341691493988,
      "num_tokens": 986880.0,
      "step": 80
    },
    {
      "entropy": 1.8249096870422363,
      "epoch": 0.04265402843601896,
      "grad_norm": 0.7752470374107361,
      "learning_rate": 4.0100250626566415e-05,
      "loss": 0.4578617215156555,
      "mean_token_accuracy": 0.8508521616458893,
      "num_tokens": 999216.0,
      "step": 81
    },
    {
      "entropy": 1.9078024625778198,
      "epoch": 0.04318062137967351,
      "grad_norm": 0.7781356573104858,
      "learning_rate": 4.0601503759398494e-05,
      "loss": 0.43578821420669556,
      "mean_token_accuracy": 0.8600046187639236,
      "num_tokens": 1011552.0,
      "step": 82
    },
    {
      "entropy": 1.8074560463428497,
      "epoch": 0.04370721432332807,
      "grad_norm": 0.7703754305839539,
      "learning_rate": 4.110275689223057e-05,
      "loss": 0.4543125629425049,
      "mean_token_accuracy": 0.8561971783638,
      "num_tokens": 1023888.0,
      "step": 83
    },
    {
      "entropy": 1.7470267415046692,
      "epoch": 0.044233807266982623,
      "grad_norm": 0.6839037537574768,
      "learning_rate": 4.160401002506266e-05,
      "loss": 0.4043883681297302,
      "mean_token_accuracy": 0.8632047474384308,
      "num_tokens": 1036224.0,
      "step": 84
    },
    {
      "entropy": 1.8056954443454742,
      "epoch": 0.04476040021063718,
      "grad_norm": 0.8745210766792297,
      "learning_rate": 4.210526315789474e-05,
      "loss": 0.4576418101787567,
      "mean_token_accuracy": 0.8515569567680359,
      "num_tokens": 1048560.0,
      "step": 85
    },
    {
      "entropy": 1.7271961271762848,
      "epoch": 0.04528699315429173,
      "grad_norm": 0.7987569570541382,
      "learning_rate": 4.260651629072682e-05,
      "loss": 0.4062059819698334,
      "mean_token_accuracy": 0.8644601851701736,
      "num_tokens": 1060896.0,
      "step": 86
    },
    {
      "entropy": 1.7829216122627258,
      "epoch": 0.045813586097946286,
      "grad_norm": 0.8260216116905212,
      "learning_rate": 4.3107769423558896e-05,
      "loss": 0.4285086989402771,
      "mean_token_accuracy": 0.8619498312473297,
      "num_tokens": 1073232.0,
      "step": 87
    },
    {
      "entropy": 1.720045566558838,
      "epoch": 0.04634017904160084,
      "grad_norm": 0.8638037443161011,
      "learning_rate": 4.3609022556390975e-05,
      "loss": 0.42514118552207947,
      "mean_token_accuracy": 0.8556364923715591,
      "num_tokens": 1085568.0,
      "step": 88
    },
    {
      "entropy": 1.749743402004242,
      "epoch": 0.0468667719852554,
      "grad_norm": 0.6750407218933105,
      "learning_rate": 4.411027568922306e-05,
      "loss": 0.3981553912162781,
      "mean_token_accuracy": 0.8708555996417999,
      "num_tokens": 1097904.0,
      "step": 89
    },
    {
      "entropy": 1.785173624753952,
      "epoch": 0.04739336492890995,
      "grad_norm": 0.7802124619483948,
      "learning_rate": 4.461152882205514e-05,
      "loss": 0.3845086097717285,
      "mean_token_accuracy": 0.8747637271881104,
      "num_tokens": 1110240.0,
      "step": 90
    },
    {
      "entropy": 1.7087633907794952,
      "epoch": 0.047919957872564506,
      "grad_norm": 0.8488918542861938,
      "learning_rate": 4.511278195488722e-05,
      "loss": 0.3644571006298065,
      "mean_token_accuracy": 0.8757548332214355,
      "num_tokens": 1122576.0,
      "step": 91
    },
    {
      "entropy": 1.6821248829364777,
      "epoch": 0.04844655081621906,
      "grad_norm": 0.7744936943054199,
      "learning_rate": 4.56140350877193e-05,
      "loss": 0.40767285227775574,
      "mean_token_accuracy": 0.8596755862236023,
      "num_tokens": 1134912.0,
      "step": 92
    },
    {
      "entropy": 1.7099383771419525,
      "epoch": 0.04897314375987362,
      "grad_norm": 0.9390520453453064,
      "learning_rate": 4.6115288220551377e-05,
      "loss": 0.4450380802154541,
      "mean_token_accuracy": 0.8517685830593109,
      "num_tokens": 1147248.0,
      "step": 93
    },
    {
      "entropy": 1.6644540429115295,
      "epoch": 0.049499736703528176,
      "grad_norm": 0.9382246136665344,
      "learning_rate": 4.6616541353383456e-05,
      "loss": 0.37574899196624756,
      "mean_token_accuracy": 0.8736520707607269,
      "num_tokens": 1159584.0,
      "step": 94
    },
    {
      "entropy": 1.7045432031154633,
      "epoch": 0.050026329647182725,
      "grad_norm": 0.784248948097229,
      "learning_rate": 4.711779448621554e-05,
      "loss": 0.39656203985214233,
      "mean_token_accuracy": 0.8673653900623322,
      "num_tokens": 1171920.0,
      "step": 95
    },
    {
      "entropy": 1.6858630180358887,
      "epoch": 0.05055292259083728,
      "grad_norm": 0.7153828144073486,
      "learning_rate": 4.761904761904762e-05,
      "loss": 0.3821033835411072,
      "mean_token_accuracy": 0.870804488658905,
      "num_tokens": 1184256.0,
      "step": 96
    },
    {
      "entropy": 1.5798078775405884,
      "epoch": 0.05107951553449184,
      "grad_norm": 0.6857785582542419,
      "learning_rate": 4.81203007518797e-05,
      "loss": 0.3593214452266693,
      "mean_token_accuracy": 0.8746844828128815,
      "num_tokens": 1196592.0,
      "step": 97
    },
    {
      "entropy": 1.670957088470459,
      "epoch": 0.051606108478146395,
      "grad_norm": 0.7506043910980225,
      "learning_rate": 4.862155388471178e-05,
      "loss": 0.3727911412715912,
      "mean_token_accuracy": 0.8756437748670578,
      "num_tokens": 1208928.0,
      "step": 98
    },
    {
      "entropy": 1.5233455896377563,
      "epoch": 0.052132701421800945,
      "grad_norm": 0.7735759019851685,
      "learning_rate": 4.912280701754386e-05,
      "loss": 0.37093132734298706,
      "mean_token_accuracy": 0.8704032748937607,
      "num_tokens": 1221264.0,
      "step": 99
    },
    {
      "entropy": 1.6192973852157593,
      "epoch": 0.0526592943654555,
      "grad_norm": 0.7478036880493164,
      "learning_rate": 4.9624060150375936e-05,
      "loss": 0.3426092863082886,
      "mean_token_accuracy": 0.8861435055732727,
      "num_tokens": 1233600.0,
      "step": 100
    },
    {
      "entropy": 1.493833750486374,
      "epoch": 0.05318588730911006,
      "grad_norm": 0.7044211030006409,
      "learning_rate": 5.012531328320802e-05,
      "loss": 0.35205507278442383,
      "mean_token_accuracy": 0.8778510093688965,
      "num_tokens": 1245936.0,
      "step": 101
    },
    {
      "entropy": 1.4939774870872498,
      "epoch": 0.053712480252764615,
      "grad_norm": 0.7720161080360413,
      "learning_rate": 5.06265664160401e-05,
      "loss": 0.3975054621696472,
      "mean_token_accuracy": 0.8641460090875626,
      "num_tokens": 1258272.0,
      "step": 102
    },
    {
      "entropy": 1.6101951897144318,
      "epoch": 0.05423907319641917,
      "grad_norm": 0.669315755367279,
      "learning_rate": 5.112781954887218e-05,
      "loss": 0.34713947772979736,
      "mean_token_accuracy": 0.8790434300899506,
      "num_tokens": 1270608.0,
      "step": 103
    },
    {
      "entropy": 1.5196483135223389,
      "epoch": 0.05476566614007372,
      "grad_norm": 0.8516085743904114,
      "learning_rate": 5.162907268170426e-05,
      "loss": 0.3601211607456207,
      "mean_token_accuracy": 0.8756401091814041,
      "num_tokens": 1282944.0,
      "step": 104
    },
    {
      "entropy": 1.4721749424934387,
      "epoch": 0.05529225908372828,
      "grad_norm": 0.8156314492225647,
      "learning_rate": 5.213032581453634e-05,
      "loss": 0.3697987198829651,
      "mean_token_accuracy": 0.8728637546300888,
      "num_tokens": 1295280.0,
      "step": 105
    },
    {
      "entropy": 1.4663786590099335,
      "epoch": 0.055818852027382834,
      "grad_norm": 0.6453195810317993,
      "learning_rate": 5.2631578947368424e-05,
      "loss": 0.3338157534599304,
      "mean_token_accuracy": 0.8864734172821045,
      "num_tokens": 1307616.0,
      "step": 106
    },
    {
      "entropy": 1.497151494026184,
      "epoch": 0.05634544497103739,
      "grad_norm": 0.7146667242050171,
      "learning_rate": 5.31328320802005e-05,
      "loss": 0.34923166036605835,
      "mean_token_accuracy": 0.8819292187690735,
      "num_tokens": 1319952.0,
      "step": 107
    },
    {
      "entropy": 1.5161625444889069,
      "epoch": 0.05687203791469194,
      "grad_norm": 0.8290141820907593,
      "learning_rate": 5.363408521303258e-05,
      "loss": 0.3371197581291199,
      "mean_token_accuracy": 0.8877773582935333,
      "num_tokens": 1332288.0,
      "step": 108
    },
    {
      "entropy": 1.4333459436893463,
      "epoch": 0.0573986308583465,
      "grad_norm": 0.8022440671920776,
      "learning_rate": 5.413533834586466e-05,
      "loss": 0.31702953577041626,
      "mean_token_accuracy": 0.8909593671560287,
      "num_tokens": 1344624.0,
      "step": 109
    },
    {
      "entropy": 1.4444147646427155,
      "epoch": 0.057925223802001054,
      "grad_norm": 0.7960955500602722,
      "learning_rate": 5.463659147869674e-05,
      "loss": 0.3525749742984772,
      "mean_token_accuracy": 0.8755318969488144,
      "num_tokens": 1356960.0,
      "step": 110
    },
    {
      "entropy": 1.469768613576889,
      "epoch": 0.05845181674565561,
      "grad_norm": 0.8183742165565491,
      "learning_rate": 5.513784461152882e-05,
      "loss": 0.3398601710796356,
      "mean_token_accuracy": 0.8840423971414566,
      "num_tokens": 1369296.0,
      "step": 111
    },
    {
      "entropy": 1.4481310844421387,
      "epoch": 0.05897840968931016,
      "grad_norm": 0.6801962852478027,
      "learning_rate": 5.5639097744360905e-05,
      "loss": 0.34869900345802307,
      "mean_token_accuracy": 0.877396434545517,
      "num_tokens": 1381632.0,
      "step": 112
    },
    {
      "entropy": 1.4586880207061768,
      "epoch": 0.05950500263296472,
      "grad_norm": 0.7105273008346558,
      "learning_rate": 5.6140350877192984e-05,
      "loss": 0.32431477308273315,
      "mean_token_accuracy": 0.8829907327890396,
      "num_tokens": 1393968.0,
      "step": 113
    },
    {
      "entropy": 1.4191592037677765,
      "epoch": 0.06003159557661927,
      "grad_norm": 0.8621596097946167,
      "learning_rate": 5.664160401002506e-05,
      "loss": 0.29889625310897827,
      "mean_token_accuracy": 0.8882533758878708,
      "num_tokens": 1406304.0,
      "step": 114
    },
    {
      "entropy": 1.431228756904602,
      "epoch": 0.06055818852027383,
      "grad_norm": 0.662341833114624,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.34662240743637085,
      "mean_token_accuracy": 0.8789393156766891,
      "num_tokens": 1418640.0,
      "step": 115
    },
    {
      "entropy": 1.317075937986374,
      "epoch": 0.061084781463928386,
      "grad_norm": 0.6940162777900696,
      "learning_rate": 5.764411027568922e-05,
      "loss": 0.3154875636100769,
      "mean_token_accuracy": 0.886825367808342,
      "num_tokens": 1430976.0,
      "step": 116
    },
    {
      "entropy": 1.3747568726539612,
      "epoch": 0.061611374407582936,
      "grad_norm": 0.6974264979362488,
      "learning_rate": 5.81453634085213e-05,
      "loss": 0.36218538880348206,
      "mean_token_accuracy": 0.8753509372472763,
      "num_tokens": 1443312.0,
      "step": 117
    },
    {
      "entropy": 1.3549829423427582,
      "epoch": 0.06213796735123749,
      "grad_norm": 0.7157356142997742,
      "learning_rate": 5.8646616541353386e-05,
      "loss": 0.32229024171829224,
      "mean_token_accuracy": 0.8855572938919067,
      "num_tokens": 1455648.0,
      "step": 118
    },
    {
      "entropy": 1.4570170640945435,
      "epoch": 0.06266456029489205,
      "grad_norm": 0.7710041999816895,
      "learning_rate": 5.9147869674185465e-05,
      "loss": 0.3285849094390869,
      "mean_token_accuracy": 0.8840549886226654,
      "num_tokens": 1467984.0,
      "step": 119
    },
    {
      "entropy": 1.4027460813522339,
      "epoch": 0.0631911532385466,
      "grad_norm": 0.6999377012252808,
      "learning_rate": 5.9649122807017544e-05,
      "loss": 0.3464536666870117,
      "mean_token_accuracy": 0.8799307644367218,
      "num_tokens": 1480320.0,
      "step": 120
    },
    {
      "entropy": 1.3301794826984406,
      "epoch": 0.06371774618220116,
      "grad_norm": 0.7836945056915283,
      "learning_rate": 6.015037593984962e-05,
      "loss": 0.3436964750289917,
      "mean_token_accuracy": 0.8764877319335938,
      "num_tokens": 1492656.0,
      "step": 121
    },
    {
      "entropy": 1.3585115671157837,
      "epoch": 0.06424433912585571,
      "grad_norm": 0.7850233316421509,
      "learning_rate": 6.06516290726817e-05,
      "loss": 0.35563230514526367,
      "mean_token_accuracy": 0.8757744282484055,
      "num_tokens": 1504992.0,
      "step": 122
    },
    {
      "entropy": 1.2973756194114685,
      "epoch": 0.06477093206951026,
      "grad_norm": 0.7351557612419128,
      "learning_rate": 6.115288220551379e-05,
      "loss": 0.3156384527683258,
      "mean_token_accuracy": 0.886470839381218,
      "num_tokens": 1517328.0,
      "step": 123
    },
    {
      "entropy": 1.3591448068618774,
      "epoch": 0.06529752501316483,
      "grad_norm": 0.7262402176856995,
      "learning_rate": 6.165413533834587e-05,
      "loss": 0.27420130372047424,
      "mean_token_accuracy": 0.904089480638504,
      "num_tokens": 1529664.0,
      "step": 124
    },
    {
      "entropy": 1.253787875175476,
      "epoch": 0.06582411795681938,
      "grad_norm": 0.6739073991775513,
      "learning_rate": 6.215538847117795e-05,
      "loss": 0.34957295656204224,
      "mean_token_accuracy": 0.879968523979187,
      "num_tokens": 1542000.0,
      "step": 125
    },
    {
      "entropy": 1.261105090379715,
      "epoch": 0.06635071090047394,
      "grad_norm": 0.7828674912452698,
      "learning_rate": 6.265664160401002e-05,
      "loss": 0.3064384460449219,
      "mean_token_accuracy": 0.8853085935115814,
      "num_tokens": 1554336.0,
      "step": 126
    },
    {
      "entropy": 1.2267013192176819,
      "epoch": 0.06687730384412849,
      "grad_norm": 0.9648973345756531,
      "learning_rate": 6.31578947368421e-05,
      "loss": 0.3406292498111725,
      "mean_token_accuracy": 0.8785828351974487,
      "num_tokens": 1566672.0,
      "step": 127
    },
    {
      "entropy": 1.2064116597175598,
      "epoch": 0.06740389678778304,
      "grad_norm": 0.8842466473579407,
      "learning_rate": 6.365914786967418e-05,
      "loss": 0.3578197658061981,
      "mean_token_accuracy": 0.8797773271799088,
      "num_tokens": 1579008.0,
      "step": 128
    },
    {
      "entropy": 1.1993357837200165,
      "epoch": 0.0679304897314376,
      "grad_norm": 0.8396037817001343,
      "learning_rate": 6.416040100250626e-05,
      "loss": 0.32377177476882935,
      "mean_token_accuracy": 0.8842920362949371,
      "num_tokens": 1591344.0,
      "step": 129
    },
    {
      "entropy": 1.2047476172447205,
      "epoch": 0.06845708267509215,
      "grad_norm": 0.6860671043395996,
      "learning_rate": 6.466165413533834e-05,
      "loss": 0.29577356576919556,
      "mean_token_accuracy": 0.896445706486702,
      "num_tokens": 1603680.0,
      "step": 130
    },
    {
      "entropy": 1.2046065032482147,
      "epoch": 0.06898367561874671,
      "grad_norm": 0.6543269157409668,
      "learning_rate": 6.516290726817042e-05,
      "loss": 0.2779792845249176,
      "mean_token_accuracy": 0.9030982404947281,
      "num_tokens": 1616016.0,
      "step": 131
    },
    {
      "entropy": 1.1669755578041077,
      "epoch": 0.06951026856240126,
      "grad_norm": 0.7457153797149658,
      "learning_rate": 6.566416040100251e-05,
      "loss": 0.3149452209472656,
      "mean_token_accuracy": 0.8876487761735916,
      "num_tokens": 1628352.0,
      "step": 132
    },
    {
      "entropy": 1.2609427571296692,
      "epoch": 0.07003686150605581,
      "grad_norm": 0.7485870122909546,
      "learning_rate": 6.616541353383459e-05,
      "loss": 0.31902027130126953,
      "mean_token_accuracy": 0.8813460171222687,
      "num_tokens": 1640688.0,
      "step": 133
    },
    {
      "entropy": 1.2575346231460571,
      "epoch": 0.07056345444971038,
      "grad_norm": 0.7213554382324219,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.2797701358795166,
      "mean_token_accuracy": 0.8987373113632202,
      "num_tokens": 1653024.0,
      "step": 134
    },
    {
      "entropy": 1.186894565820694,
      "epoch": 0.07109004739336493,
      "grad_norm": 0.7020873427391052,
      "learning_rate": 6.716791979949875e-05,
      "loss": 0.2995363771915436,
      "mean_token_accuracy": 0.8897402882575989,
      "num_tokens": 1665360.0,
      "step": 135
    },
    {
      "entropy": 1.161018431186676,
      "epoch": 0.07161664033701948,
      "grad_norm": 0.7155889868736267,
      "learning_rate": 6.766917293233083e-05,
      "loss": 0.2925366759300232,
      "mean_token_accuracy": 0.8951611965894699,
      "num_tokens": 1677696.0,
      "step": 136
    },
    {
      "entropy": 1.2650261223316193,
      "epoch": 0.07214323328067404,
      "grad_norm": 0.7642542123794556,
      "learning_rate": 6.817042606516291e-05,
      "loss": 0.2801384925842285,
      "mean_token_accuracy": 0.8974052667617798,
      "num_tokens": 1690032.0,
      "step": 137
    },
    {
      "entropy": 1.2530198991298676,
      "epoch": 0.07266982622432859,
      "grad_norm": 0.8588266372680664,
      "learning_rate": 6.867167919799499e-05,
      "loss": 0.3360725939273834,
      "mean_token_accuracy": 0.8794126808643341,
      "num_tokens": 1702368.0,
      "step": 138
    },
    {
      "entropy": 1.2518652081489563,
      "epoch": 0.07319641916798315,
      "grad_norm": 0.6655072569847107,
      "learning_rate": 6.917293233082706e-05,
      "loss": 0.2788892388343811,
      "mean_token_accuracy": 0.8994075357913971,
      "num_tokens": 1714704.0,
      "step": 139
    },
    {
      "entropy": 1.1829611957073212,
      "epoch": 0.0737230121116377,
      "grad_norm": 0.7172735333442688,
      "learning_rate": 6.967418546365914e-05,
      "loss": 0.2918061316013336,
      "mean_token_accuracy": 0.8959014713764191,
      "num_tokens": 1727040.0,
      "step": 140
    },
    {
      "entropy": 1.2773233950138092,
      "epoch": 0.07424960505529225,
      "grad_norm": 0.697880208492279,
      "learning_rate": 7.017543859649122e-05,
      "loss": 0.3217264711856842,
      "mean_token_accuracy": 0.8819008469581604,
      "num_tokens": 1739376.0,
      "step": 141
    },
    {
      "entropy": 1.2314070761203766,
      "epoch": 0.07477619799894682,
      "grad_norm": 0.701502799987793,
      "learning_rate": 7.06766917293233e-05,
      "loss": 0.29351311922073364,
      "mean_token_accuracy": 0.8908087462186813,
      "num_tokens": 1751712.0,
      "step": 142
    },
    {
      "entropy": 1.2544872462749481,
      "epoch": 0.07530279094260137,
      "grad_norm": 0.7028551697731018,
      "learning_rate": 7.11779448621554e-05,
      "loss": 0.3528595566749573,
      "mean_token_accuracy": 0.8716742098331451,
      "num_tokens": 1764048.0,
      "step": 143
    },
    {
      "entropy": 1.1984558403491974,
      "epoch": 0.07582938388625593,
      "grad_norm": 0.6910477876663208,
      "learning_rate": 7.167919799498747e-05,
      "loss": 0.3136734366416931,
      "mean_token_accuracy": 0.8870812356472015,
      "num_tokens": 1776384.0,
      "step": 144
    },
    {
      "entropy": 1.2033996284008026,
      "epoch": 0.07635597682991048,
      "grad_norm": 0.7239225506782532,
      "learning_rate": 7.218045112781955e-05,
      "loss": 0.2979337275028229,
      "mean_token_accuracy": 0.8915670961141586,
      "num_tokens": 1788720.0,
      "step": 145
    },
    {
      "entropy": 1.1562331318855286,
      "epoch": 0.07688256977356503,
      "grad_norm": 0.7481216192245483,
      "learning_rate": 7.268170426065163e-05,
      "loss": 0.2653358280658722,
      "mean_token_accuracy": 0.8980085849761963,
      "num_tokens": 1801056.0,
      "step": 146
    },
    {
      "entropy": 1.2064136862754822,
      "epoch": 0.07740916271721959,
      "grad_norm": 0.7776615023612976,
      "learning_rate": 7.318295739348371e-05,
      "loss": 0.2848292589187622,
      "mean_token_accuracy": 0.8949488550424576,
      "num_tokens": 1813392.0,
      "step": 147
    },
    {
      "entropy": 1.1387107968330383,
      "epoch": 0.07793575566087414,
      "grad_norm": 0.8243488669395447,
      "learning_rate": 7.368421052631579e-05,
      "loss": 0.31445616483688354,
      "mean_token_accuracy": 0.887312263250351,
      "num_tokens": 1825728.0,
      "step": 148
    },
    {
      "entropy": 1.0903229415416718,
      "epoch": 0.0784623486045287,
      "grad_norm": 0.775776207447052,
      "learning_rate": 7.418546365914787e-05,
      "loss": 0.2984585762023926,
      "mean_token_accuracy": 0.8886806517839432,
      "num_tokens": 1838064.0,
      "step": 149
    },
    {
      "entropy": 1.1340011656284332,
      "epoch": 0.07898894154818326,
      "grad_norm": 0.7941513657569885,
      "learning_rate": 7.468671679197995e-05,
      "loss": 0.27699777483940125,
      "mean_token_accuracy": 0.9033427834510803,
      "num_tokens": 1850400.0,
      "step": 150
    },
    {
      "entropy": 1.084587723016739,
      "epoch": 0.0795155344918378,
      "grad_norm": 0.7044249773025513,
      "learning_rate": 7.518796992481203e-05,
      "loss": 0.2815166711807251,
      "mean_token_accuracy": 0.8928717374801636,
      "num_tokens": 1862736.0,
      "step": 151
    },
    {
      "entropy": 1.0816469937562943,
      "epoch": 0.08004212743549237,
      "grad_norm": 0.8506439328193665,
      "learning_rate": 7.56892230576441e-05,
      "loss": 0.3135736584663391,
      "mean_token_accuracy": 0.8835454434156418,
      "num_tokens": 1875072.0,
      "step": 152
    },
    {
      "entropy": 1.037431538105011,
      "epoch": 0.08056872037914692,
      "grad_norm": 0.6357259750366211,
      "learning_rate": 7.619047619047618e-05,
      "loss": 0.2715918719768524,
      "mean_token_accuracy": 0.9003684967756271,
      "num_tokens": 1887408.0,
      "step": 153
    },
    {
      "entropy": 1.101006418466568,
      "epoch": 0.08109531332280147,
      "grad_norm": 0.7176199555397034,
      "learning_rate": 7.669172932330826e-05,
      "loss": 0.3140455186367035,
      "mean_token_accuracy": 0.8868316859006882,
      "num_tokens": 1899744.0,
      "step": 154
    },
    {
      "entropy": 1.048825979232788,
      "epoch": 0.08162190626645603,
      "grad_norm": 0.8264813423156738,
      "learning_rate": 7.719298245614036e-05,
      "loss": 0.27318042516708374,
      "mean_token_accuracy": 0.8995131105184555,
      "num_tokens": 1912080.0,
      "step": 155
    },
    {
      "entropy": 1.0369156897068024,
      "epoch": 0.08214849921011058,
      "grad_norm": 0.7105727791786194,
      "learning_rate": 7.769423558897244e-05,
      "loss": 0.287497341632843,
      "mean_token_accuracy": 0.8940722495317459,
      "num_tokens": 1924416.0,
      "step": 156
    },
    {
      "entropy": 1.0527066588401794,
      "epoch": 0.08267509215376515,
      "grad_norm": 0.7094339728355408,
      "learning_rate": 7.819548872180451e-05,
      "loss": 0.31165987253189087,
      "mean_token_accuracy": 0.8851522654294968,
      "num_tokens": 1936752.0,
      "step": 157
    },
    {
      "entropy": 1.018677532672882,
      "epoch": 0.0832016850974197,
      "grad_norm": 0.6412932276725769,
      "learning_rate": 7.869674185463659e-05,
      "loss": 0.28470396995544434,
      "mean_token_accuracy": 0.8961537182331085,
      "num_tokens": 1949088.0,
      "step": 158
    },
    {
      "entropy": 1.011895701289177,
      "epoch": 0.08372827804107424,
      "grad_norm": 0.6732510924339294,
      "learning_rate": 7.919799498746867e-05,
      "loss": 0.28573673963546753,
      "mean_token_accuracy": 0.8944952934980392,
      "num_tokens": 1961424.0,
      "step": 159
    },
    {
      "entropy": 1.011293813586235,
      "epoch": 0.08425487098472881,
      "grad_norm": 0.6785159707069397,
      "learning_rate": 7.969924812030075e-05,
      "loss": 0.30068039894104004,
      "mean_token_accuracy": 0.8911136984825134,
      "num_tokens": 1973760.0,
      "step": 160
    },
    {
      "entropy": 1.020560160279274,
      "epoch": 0.08478146392838336,
      "grad_norm": 0.7218092679977417,
      "learning_rate": 8.020050125313283e-05,
      "loss": 0.30739086866378784,
      "mean_token_accuracy": 0.885330468416214,
      "num_tokens": 1986096.0,
      "step": 161
    },
    {
      "entropy": 1.0597094595432281,
      "epoch": 0.08530805687203792,
      "grad_norm": 0.7374167442321777,
      "learning_rate": 8.070175438596491e-05,
      "loss": 0.31700241565704346,
      "mean_token_accuracy": 0.8844025731086731,
      "num_tokens": 1998432.0,
      "step": 162
    },
    {
      "entropy": 1.0140591859817505,
      "epoch": 0.08583464981569247,
      "grad_norm": 0.6328212022781372,
      "learning_rate": 8.120300751879699e-05,
      "loss": 0.282073050737381,
      "mean_token_accuracy": 0.9006371796131134,
      "num_tokens": 2010768.0,
      "step": 163
    },
    {
      "entropy": 1.0362461507320404,
      "epoch": 0.08636124275934702,
      "grad_norm": 0.6613935828208923,
      "learning_rate": 8.170426065162907e-05,
      "loss": 0.2826480567455292,
      "mean_token_accuracy": 0.8936625272035599,
      "num_tokens": 2023104.0,
      "step": 164
    },
    {
      "entropy": 1.0381528735160828,
      "epoch": 0.08688783570300158,
      "grad_norm": 0.6834626197814941,
      "learning_rate": 8.220551378446115e-05,
      "loss": 0.2852419912815094,
      "mean_token_accuracy": 0.8987163454294205,
      "num_tokens": 2035440.0,
      "step": 165
    },
    {
      "entropy": 1.0698944479227066,
      "epoch": 0.08741442864665613,
      "grad_norm": 0.7904614806175232,
      "learning_rate": 8.270676691729324e-05,
      "loss": 0.2937854826450348,
      "mean_token_accuracy": 0.8909425586462021,
      "num_tokens": 2047776.0,
      "step": 166
    },
    {
      "entropy": 1.040902853012085,
      "epoch": 0.08794102159031068,
      "grad_norm": 0.7362328767776489,
      "learning_rate": 8.320802005012532e-05,
      "loss": 0.2914920449256897,
      "mean_token_accuracy": 0.898249089717865,
      "num_tokens": 2060112.0,
      "step": 167
    },
    {
      "entropy": 1.028788536787033,
      "epoch": 0.08846761453396525,
      "grad_norm": 0.8054221868515015,
      "learning_rate": 8.37092731829574e-05,
      "loss": 0.3018620014190674,
      "mean_token_accuracy": 0.8883562982082367,
      "num_tokens": 2072448.0,
      "step": 168
    },
    {
      "entropy": 1.0401608049869537,
      "epoch": 0.0889942074776198,
      "grad_norm": 0.6363168954849243,
      "learning_rate": 8.421052631578948e-05,
      "loss": 0.2924948036670685,
      "mean_token_accuracy": 0.8931814581155777,
      "num_tokens": 2084784.0,
      "step": 169
    },
    {
      "entropy": 1.0388599783182144,
      "epoch": 0.08952080042127436,
      "grad_norm": 1.0536071062088013,
      "learning_rate": 8.471177944862155e-05,
      "loss": 0.3022274971008301,
      "mean_token_accuracy": 0.8892710357904434,
      "num_tokens": 2097120.0,
      "step": 170
    },
    {
      "entropy": 1.0794771611690521,
      "epoch": 0.09004739336492891,
      "grad_norm": 0.7485163807868958,
      "learning_rate": 8.521303258145363e-05,
      "loss": 0.28198155760765076,
      "mean_token_accuracy": 0.8907860517501831,
      "num_tokens": 2109456.0,
      "step": 171
    },
    {
      "entropy": 1.0737251937389374,
      "epoch": 0.09057398630858346,
      "grad_norm": 0.7381237745285034,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.2523871660232544,
      "mean_token_accuracy": 0.9074936509132385,
      "num_tokens": 2121792.0,
      "step": 172
    },
    {
      "entropy": 1.003856584429741,
      "epoch": 0.09110057925223802,
      "grad_norm": 0.7226340770721436,
      "learning_rate": 8.621553884711779e-05,
      "loss": 0.27109643816947937,
      "mean_token_accuracy": 0.9041073471307755,
      "num_tokens": 2134128.0,
      "step": 173
    },
    {
      "entropy": 1.1186200976371765,
      "epoch": 0.09162717219589257,
      "grad_norm": 0.7579658627510071,
      "learning_rate": 8.671679197994987e-05,
      "loss": 0.3053061366081238,
      "mean_token_accuracy": 0.8858212381601334,
      "num_tokens": 2146464.0,
      "step": 174
    },
    {
      "entropy": 1.1462504863739014,
      "epoch": 0.09215376513954714,
      "grad_norm": 0.6947624087333679,
      "learning_rate": 8.721804511278195e-05,
      "loss": 0.3097134232521057,
      "mean_token_accuracy": 0.8846383690834045,
      "num_tokens": 2158800.0,
      "step": 175
    },
    {
      "entropy": 1.1312821209430695,
      "epoch": 0.09268035808320169,
      "grad_norm": 0.6850878596305847,
      "learning_rate": 8.771929824561403e-05,
      "loss": 0.26429426670074463,
      "mean_token_accuracy": 0.8986905962228775,
      "num_tokens": 2171136.0,
      "step": 176
    },
    {
      "entropy": 1.146384358406067,
      "epoch": 0.09320695102685624,
      "grad_norm": 0.7358205914497375,
      "learning_rate": 8.822055137844612e-05,
      "loss": 0.2753527760505676,
      "mean_token_accuracy": 0.8983759731054306,
      "num_tokens": 2183472.0,
      "step": 177
    },
    {
      "entropy": 1.2575857639312744,
      "epoch": 0.0937335439705108,
      "grad_norm": 0.6449548006057739,
      "learning_rate": 8.87218045112782e-05,
      "loss": 0.2840855121612549,
      "mean_token_accuracy": 0.8939386010169983,
      "num_tokens": 2195808.0,
      "step": 178
    },
    {
      "entropy": 1.190169095993042,
      "epoch": 0.09426013691416535,
      "grad_norm": 0.7093657851219177,
      "learning_rate": 8.922305764411028e-05,
      "loss": 0.2701612114906311,
      "mean_token_accuracy": 0.9016707092523575,
      "num_tokens": 2208144.0,
      "step": 179
    },
    {
      "entropy": 1.229299396276474,
      "epoch": 0.0947867298578199,
      "grad_norm": 0.7190297245979309,
      "learning_rate": 8.972431077694236e-05,
      "loss": 0.25798100233078003,
      "mean_token_accuracy": 0.9037416577339172,
      "num_tokens": 2220480.0,
      "step": 180
    },
    {
      "entropy": 1.2668792605400085,
      "epoch": 0.09531332280147446,
      "grad_norm": 0.7405722737312317,
      "learning_rate": 9.022556390977444e-05,
      "loss": 0.2883971929550171,
      "mean_token_accuracy": 0.8925597667694092,
      "num_tokens": 2232816.0,
      "step": 181
    },
    {
      "entropy": 1.1805595457553864,
      "epoch": 0.09583991574512901,
      "grad_norm": 0.6785654425621033,
      "learning_rate": 9.072681704260652e-05,
      "loss": 0.2622534930706024,
      "mean_token_accuracy": 0.9043945819139481,
      "num_tokens": 2245152.0,
      "step": 182
    },
    {
      "entropy": 1.24412140250206,
      "epoch": 0.09636650868878358,
      "grad_norm": 0.6113075017929077,
      "learning_rate": 9.12280701754386e-05,
      "loss": 0.30303800106048584,
      "mean_token_accuracy": 0.8807209581136703,
      "num_tokens": 2257488.0,
      "step": 183
    },
    {
      "entropy": 1.1531569063663483,
      "epoch": 0.09689310163243813,
      "grad_norm": 0.6058000922203064,
      "learning_rate": 9.172932330827067e-05,
      "loss": 0.2888239622116089,
      "mean_token_accuracy": 0.8989417850971222,
      "num_tokens": 2269824.0,
      "step": 184
    },
    {
      "entropy": 1.1511635780334473,
      "epoch": 0.09741969457609267,
      "grad_norm": 0.6023234128952026,
      "learning_rate": 9.223057644110275e-05,
      "loss": 0.2515050172805786,
      "mean_token_accuracy": 0.9097182005643845,
      "num_tokens": 2282160.0,
      "step": 185
    },
    {
      "entropy": 1.1716180443763733,
      "epoch": 0.09794628751974724,
      "grad_norm": 0.6153926253318787,
      "learning_rate": 9.273182957393483e-05,
      "loss": 0.2679345905780792,
      "mean_token_accuracy": 0.8987378627061844,
      "num_tokens": 2294496.0,
      "step": 186
    },
    {
      "entropy": 1.1668370068073273,
      "epoch": 0.09847288046340179,
      "grad_norm": 0.6524267196655273,
      "learning_rate": 9.323308270676691e-05,
      "loss": 0.28671011328697205,
      "mean_token_accuracy": 0.8961494415998459,
      "num_tokens": 2306832.0,
      "step": 187
    },
    {
      "entropy": 1.1817428171634674,
      "epoch": 0.09899947340705635,
      "grad_norm": 0.6342366337776184,
      "learning_rate": 9.373433583959899e-05,
      "loss": 0.2956732511520386,
      "mean_token_accuracy": 0.8947313278913498,
      "num_tokens": 2319168.0,
      "step": 188
    },
    {
      "entropy": 1.118630200624466,
      "epoch": 0.0995260663507109,
      "grad_norm": 0.6613883972167969,
      "learning_rate": 9.423558897243108e-05,
      "loss": 0.24410809576511383,
      "mean_token_accuracy": 0.9081003367900848,
      "num_tokens": 2331504.0,
      "step": 189
    },
    {
      "entropy": 1.1118974089622498,
      "epoch": 0.10005265929436545,
      "grad_norm": 0.6532029509544373,
      "learning_rate": 9.473684210526316e-05,
      "loss": 0.2677920162677765,
      "mean_token_accuracy": 0.8997665196657181,
      "num_tokens": 2343840.0,
      "step": 190
    },
    {
      "entropy": 1.1150999963283539,
      "epoch": 0.10057925223802001,
      "grad_norm": 0.6919822692871094,
      "learning_rate": 9.523809523809524e-05,
      "loss": 0.2759622633457184,
      "mean_token_accuracy": 0.896903395652771,
      "num_tokens": 2356176.0,
      "step": 191
    },
    {
      "entropy": 1.0659538507461548,
      "epoch": 0.10110584518167456,
      "grad_norm": 0.6363855600357056,
      "learning_rate": 9.573934837092732e-05,
      "loss": 0.2608574628829956,
      "mean_token_accuracy": 0.8995112329721451,
      "num_tokens": 2368512.0,
      "step": 192
    },
    {
      "entropy": 1.0789634585380554,
      "epoch": 0.10163243812532911,
      "grad_norm": 0.7043361067771912,
      "learning_rate": 9.62406015037594e-05,
      "loss": 0.27143171429634094,
      "mean_token_accuracy": 0.897380143404007,
      "num_tokens": 2380848.0,
      "step": 193
    },
    {
      "entropy": 1.1381100416183472,
      "epoch": 0.10215903106898368,
      "grad_norm": 0.6552066206932068,
      "learning_rate": 9.674185463659148e-05,
      "loss": 0.26525720953941345,
      "mean_token_accuracy": 0.9070263504981995,
      "num_tokens": 2393184.0,
      "step": 194
    },
    {
      "entropy": 1.1865114271640778,
      "epoch": 0.10268562401263823,
      "grad_norm": 0.634785532951355,
      "learning_rate": 9.724310776942356e-05,
      "loss": 0.26738592982292175,
      "mean_token_accuracy": 0.9013326019048691,
      "num_tokens": 2405520.0,
      "step": 195
    },
    {
      "entropy": 1.1835360527038574,
      "epoch": 0.10321221695629279,
      "grad_norm": 0.7315617799758911,
      "learning_rate": 9.774436090225564e-05,
      "loss": 0.29927945137023926,
      "mean_token_accuracy": 0.8894186317920685,
      "num_tokens": 2417856.0,
      "step": 196
    },
    {
      "entropy": 1.1675716638565063,
      "epoch": 0.10373880989994734,
      "grad_norm": 0.5887401700019836,
      "learning_rate": 9.824561403508771e-05,
      "loss": 0.2657887935638428,
      "mean_token_accuracy": 0.9044163674116135,
      "num_tokens": 2430192.0,
      "step": 197
    },
    {
      "entropy": 1.2119200527668,
      "epoch": 0.10426540284360189,
      "grad_norm": 0.7186099290847778,
      "learning_rate": 9.87468671679198e-05,
      "loss": 0.24999147653579712,
      "mean_token_accuracy": 0.9089571237564087,
      "num_tokens": 2442528.0,
      "step": 198
    },
    {
      "entropy": 1.1587353944778442,
      "epoch": 0.10479199578725645,
      "grad_norm": 0.6930050253868103,
      "learning_rate": 9.924812030075187e-05,
      "loss": 0.28423306345939636,
      "mean_token_accuracy": 0.8938881307840347,
      "num_tokens": 2454864.0,
      "step": 199
    },
    {
      "entropy": 1.1999735534191132,
      "epoch": 0.105318588730911,
      "grad_norm": 0.6575484871864319,
      "learning_rate": 9.974937343358397e-05,
      "loss": 0.25892868638038635,
      "mean_token_accuracy": 0.9065304547548294,
      "num_tokens": 2467200.0,
      "step": 200
    },
    {
      "entropy": 1.1664292514324188,
      "epoch": 0.10584518167456557,
      "grad_norm": 0.6501355767250061,
      "learning_rate": 0.00010025062656641604,
      "loss": 0.2730686068534851,
      "mean_token_accuracy": 0.8991207629442215,
      "num_tokens": 2479536.0,
      "step": 201
    },
    {
      "entropy": 1.2156886756420135,
      "epoch": 0.10637177461822012,
      "grad_norm": 1.1883974075317383,
      "learning_rate": 0.00010075187969924814,
      "loss": 0.2882535755634308,
      "mean_token_accuracy": 0.8935256004333496,
      "num_tokens": 2491872.0,
      "step": 202
    },
    {
      "entropy": 1.1884984374046326,
      "epoch": 0.10689836756187467,
      "grad_norm": 0.6555978655815125,
      "learning_rate": 0.0001012531328320802,
      "loss": 0.2524234354496002,
      "mean_token_accuracy": 0.9072889387607574,
      "num_tokens": 2504208.0,
      "step": 203
    },
    {
      "entropy": 1.204279750585556,
      "epoch": 0.10742496050552923,
      "grad_norm": 0.6422716379165649,
      "learning_rate": 0.0001017543859649123,
      "loss": 0.2662729322910309,
      "mean_token_accuracy": 0.9040449559688568,
      "num_tokens": 2516544.0,
      "step": 204
    },
    {
      "entropy": 1.2619583904743195,
      "epoch": 0.10795155344918378,
      "grad_norm": 0.7150806188583374,
      "learning_rate": 0.00010225563909774436,
      "loss": 0.2756538689136505,
      "mean_token_accuracy": 0.8975429385900497,
      "num_tokens": 2528880.0,
      "step": 205
    },
    {
      "entropy": 1.265720009803772,
      "epoch": 0.10847814639283834,
      "grad_norm": 0.6873537302017212,
      "learning_rate": 0.00010275689223057645,
      "loss": 0.25472116470336914,
      "mean_token_accuracy": 0.9096198529005051,
      "num_tokens": 2541216.0,
      "step": 206
    },
    {
      "entropy": 1.269091546535492,
      "epoch": 0.10900473933649289,
      "grad_norm": 0.6951980590820312,
      "learning_rate": 0.00010325814536340852,
      "loss": 0.26120638847351074,
      "mean_token_accuracy": 0.9050850421190262,
      "num_tokens": 2553552.0,
      "step": 207
    },
    {
      "entropy": 1.269478291273117,
      "epoch": 0.10953133228014744,
      "grad_norm": 0.556747555732727,
      "learning_rate": 0.00010375939849624061,
      "loss": 0.25462740659713745,
      "mean_token_accuracy": 0.9035714715719223,
      "num_tokens": 2565888.0,
      "step": 208
    },
    {
      "entropy": 1.261308342218399,
      "epoch": 0.110057925223802,
      "grad_norm": 0.6084737777709961,
      "learning_rate": 0.00010426065162907268,
      "loss": 0.28605493903160095,
      "mean_token_accuracy": 0.8935003876686096,
      "num_tokens": 2578224.0,
      "step": 209
    },
    {
      "entropy": 1.3189480602741241,
      "epoch": 0.11058451816745656,
      "grad_norm": 0.5886352062225342,
      "learning_rate": 0.00010476190476190477,
      "loss": 0.24048885703086853,
      "mean_token_accuracy": 0.9100839495658875,
      "num_tokens": 2590560.0,
      "step": 210
    },
    {
      "entropy": 1.3389763832092285,
      "epoch": 0.1111111111111111,
      "grad_norm": 0.6794092059135437,
      "learning_rate": 0.00010526315789473685,
      "loss": 0.2771369516849518,
      "mean_token_accuracy": 0.890377551317215,
      "num_tokens": 2602896.0,
      "step": 211
    },
    {
      "entropy": 1.3012954592704773,
      "epoch": 0.11163770405476567,
      "grad_norm": 0.6500777006149292,
      "learning_rate": 0.00010576441102756893,
      "loss": 0.26194366812705994,
      "mean_token_accuracy": 0.9035488367080688,
      "num_tokens": 2615232.0,
      "step": 212
    },
    {
      "entropy": 1.3287411630153656,
      "epoch": 0.11216429699842022,
      "grad_norm": 0.7444258332252502,
      "learning_rate": 0.000106265664160401,
      "loss": 0.25705063343048096,
      "mean_token_accuracy": 0.8985376209020615,
      "num_tokens": 2627568.0,
      "step": 213
    },
    {
      "entropy": 1.3762761354446411,
      "epoch": 0.11269088994207478,
      "grad_norm": 0.6778383255004883,
      "learning_rate": 0.0001067669172932331,
      "loss": 0.26562660932540894,
      "mean_token_accuracy": 0.8980493247509003,
      "num_tokens": 2639904.0,
      "step": 214
    },
    {
      "entropy": 1.3341316282749176,
      "epoch": 0.11321748288572933,
      "grad_norm": 0.5709004998207092,
      "learning_rate": 0.00010726817042606516,
      "loss": 0.22839783132076263,
      "mean_token_accuracy": 0.9137850105762482,
      "num_tokens": 2652240.0,
      "step": 215
    },
    {
      "entropy": 1.397248774766922,
      "epoch": 0.11374407582938388,
      "grad_norm": 0.6649454236030579,
      "learning_rate": 0.00010776942355889726,
      "loss": 0.2758176624774933,
      "mean_token_accuracy": 0.8951268941164017,
      "num_tokens": 2664576.0,
      "step": 216
    },
    {
      "entropy": 1.3985490798950195,
      "epoch": 0.11427066877303844,
      "grad_norm": 0.6822545528411865,
      "learning_rate": 0.00010827067669172932,
      "loss": 0.2714245021343231,
      "mean_token_accuracy": 0.9002304524183273,
      "num_tokens": 2676912.0,
      "step": 217
    },
    {
      "entropy": 1.4322108924388885,
      "epoch": 0.114797261716693,
      "grad_norm": 0.6045425534248352,
      "learning_rate": 0.00010877192982456141,
      "loss": 0.27720269560813904,
      "mean_token_accuracy": 0.8973477184772491,
      "num_tokens": 2689248.0,
      "step": 218
    },
    {
      "entropy": 1.412218064069748,
      "epoch": 0.11532385466034756,
      "grad_norm": 0.5996334552764893,
      "learning_rate": 0.00010927318295739348,
      "loss": 0.2552666962146759,
      "mean_token_accuracy": 0.9052979648113251,
      "num_tokens": 2701584.0,
      "step": 219
    },
    {
      "entropy": 1.4080830514431,
      "epoch": 0.11585044760400211,
      "grad_norm": 0.5775550007820129,
      "learning_rate": 0.00010977443609022557,
      "loss": 0.26107677817344666,
      "mean_token_accuracy": 0.902474120259285,
      "num_tokens": 2713920.0,
      "step": 220
    },
    {
      "entropy": 1.4139901399612427,
      "epoch": 0.11637704054765666,
      "grad_norm": 0.5853016972541809,
      "learning_rate": 0.00011027568922305764,
      "loss": 0.2802305519580841,
      "mean_token_accuracy": 0.8949225097894669,
      "num_tokens": 2726256.0,
      "step": 221
    },
    {
      "entropy": 1.354727566242218,
      "epoch": 0.11690363349131122,
      "grad_norm": 0.6593271493911743,
      "learning_rate": 0.00011077694235588973,
      "loss": 0.25806909799575806,
      "mean_token_accuracy": 0.9009847342967987,
      "num_tokens": 2738592.0,
      "step": 222
    },
    {
      "entropy": 1.363867849111557,
      "epoch": 0.11743022643496577,
      "grad_norm": 0.5569132566452026,
      "learning_rate": 0.00011127819548872181,
      "loss": 0.2619836926460266,
      "mean_token_accuracy": 0.9014725536108017,
      "num_tokens": 2750928.0,
      "step": 223
    },
    {
      "entropy": 1.4332265555858612,
      "epoch": 0.11795681937862032,
      "grad_norm": 0.6189314723014832,
      "learning_rate": 0.00011177944862155389,
      "loss": 0.26161426305770874,
      "mean_token_accuracy": 0.8975115269422531,
      "num_tokens": 2763264.0,
      "step": 224
    },
    {
      "entropy": 1.3454554677009583,
      "epoch": 0.11848341232227488,
      "grad_norm": 0.6003779768943787,
      "learning_rate": 0.00011228070175438597,
      "loss": 0.2651001214981079,
      "mean_token_accuracy": 0.9023340046405792,
      "num_tokens": 2775600.0,
      "step": 225
    },
    {
      "entropy": 1.3587769269943237,
      "epoch": 0.11901000526592943,
      "grad_norm": 0.7411830425262451,
      "learning_rate": 0.00011278195488721806,
      "loss": 0.2888566851615906,
      "mean_token_accuracy": 0.895685225725174,
      "num_tokens": 2787936.0,
      "step": 226
    },
    {
      "entropy": 1.4554640650749207,
      "epoch": 0.119536598209584,
      "grad_norm": 0.6714136004447937,
      "learning_rate": 0.00011328320802005013,
      "loss": 0.2898600101470947,
      "mean_token_accuracy": 0.8905848562717438,
      "num_tokens": 2800272.0,
      "step": 227
    },
    {
      "entropy": 1.4274349808692932,
      "epoch": 0.12006319115323855,
      "grad_norm": 0.5972358584403992,
      "learning_rate": 0.00011378446115288222,
      "loss": 0.2719231843948364,
      "mean_token_accuracy": 0.8995204567909241,
      "num_tokens": 2812608.0,
      "step": 228
    },
    {
      "entropy": 1.3975107669830322,
      "epoch": 0.1205897840968931,
      "grad_norm": 0.7388063073158264,
      "learning_rate": 0.00011428571428571428,
      "loss": 0.2811048924922943,
      "mean_token_accuracy": 0.8961653858423233,
      "num_tokens": 2824944.0,
      "step": 229
    },
    {
      "entropy": 1.41668900847435,
      "epoch": 0.12111637704054766,
      "grad_norm": 0.6408892273902893,
      "learning_rate": 0.00011478696741854638,
      "loss": 0.23900066316127777,
      "mean_token_accuracy": 0.9131550490856171,
      "num_tokens": 2837280.0,
      "step": 230
    },
    {
      "entropy": 1.4389840960502625,
      "epoch": 0.12164296998420221,
      "grad_norm": 0.6279134154319763,
      "learning_rate": 0.00011528822055137844,
      "loss": 0.27471503615379333,
      "mean_token_accuracy": 0.8994272202253342,
      "num_tokens": 2849616.0,
      "step": 231
    },
    {
      "entropy": 1.3705064952373505,
      "epoch": 0.12216956292785677,
      "grad_norm": 0.7351282835006714,
      "learning_rate": 0.00011578947368421053,
      "loss": 0.2829359471797943,
      "mean_token_accuracy": 0.8966951966285706,
      "num_tokens": 2861952.0,
      "step": 232
    },
    {
      "entropy": 1.4880229830741882,
      "epoch": 0.12269615587151132,
      "grad_norm": 0.6098307371139526,
      "learning_rate": 0.0001162907268170426,
      "loss": 0.25372314453125,
      "mean_token_accuracy": 0.901044175028801,
      "num_tokens": 2874288.0,
      "step": 233
    },
    {
      "entropy": 1.4908549189567566,
      "epoch": 0.12322274881516587,
      "grad_norm": 0.6075619459152222,
      "learning_rate": 0.00011679197994987469,
      "loss": 0.2654881477355957,
      "mean_token_accuracy": 0.8972955793142319,
      "num_tokens": 2886624.0,
      "step": 234
    },
    {
      "entropy": 1.3783467411994934,
      "epoch": 0.12374934175882044,
      "grad_norm": 0.6030115485191345,
      "learning_rate": 0.00011729323308270677,
      "loss": 0.25111234188079834,
      "mean_token_accuracy": 0.902543842792511,
      "num_tokens": 2898960.0,
      "step": 235
    },
    {
      "entropy": 1.3683518171310425,
      "epoch": 0.12427593470247499,
      "grad_norm": 0.6671847105026245,
      "learning_rate": 0.00011779448621553886,
      "loss": 0.25656643509864807,
      "mean_token_accuracy": 0.9067094326019287,
      "num_tokens": 2911296.0,
      "step": 236
    },
    {
      "entropy": 1.4006175696849823,
      "epoch": 0.12480252764612954,
      "grad_norm": 0.5684525370597839,
      "learning_rate": 0.00011829573934837093,
      "loss": 0.2626839876174927,
      "mean_token_accuracy": 0.9033264219760895,
      "num_tokens": 2923632.0,
      "step": 237
    },
    {
      "entropy": 1.335418850183487,
      "epoch": 0.1253291205897841,
      "grad_norm": 0.5542851090431213,
      "learning_rate": 0.00011879699248120302,
      "loss": 0.24251079559326172,
      "mean_token_accuracy": 0.9084768146276474,
      "num_tokens": 2935968.0,
      "step": 238
    },
    {
      "entropy": 1.3918206691741943,
      "epoch": 0.12585571353343866,
      "grad_norm": 0.7893558740615845,
      "learning_rate": 0.00011929824561403509,
      "loss": 0.2818213999271393,
      "mean_token_accuracy": 0.8929243236780167,
      "num_tokens": 2948304.0,
      "step": 239
    },
    {
      "entropy": 1.3344398736953735,
      "epoch": 0.1263823064770932,
      "grad_norm": 0.62665194272995,
      "learning_rate": 0.00011979949874686718,
      "loss": 0.24543407559394836,
      "mean_token_accuracy": 0.9088149815797806,
      "num_tokens": 2960640.0,
      "step": 240
    },
    {
      "entropy": 1.362634390592575,
      "epoch": 0.12690889942074776,
      "grad_norm": 0.6178820133209229,
      "learning_rate": 0.00012030075187969925,
      "loss": 0.24784636497497559,
      "mean_token_accuracy": 0.9079759418964386,
      "num_tokens": 2972976.0,
      "step": 241
    },
    {
      "entropy": 1.3410207033157349,
      "epoch": 0.12743549236440233,
      "grad_norm": 0.6127734780311584,
      "learning_rate": 0.00012080200501253134,
      "loss": 0.276973694562912,
      "mean_token_accuracy": 0.900666818022728,
      "num_tokens": 2985312.0,
      "step": 242
    },
    {
      "entropy": 1.3882217109203339,
      "epoch": 0.12796208530805686,
      "grad_norm": 0.6211519837379456,
      "learning_rate": 0.0001213032581453634,
      "loss": 0.29095229506492615,
      "mean_token_accuracy": 0.8922038376331329,
      "num_tokens": 2997648.0,
      "step": 243
    },
    {
      "entropy": 1.3064784407615662,
      "epoch": 0.12848867825171142,
      "grad_norm": 0.6020708084106445,
      "learning_rate": 0.0001218045112781955,
      "loss": 0.25750190019607544,
      "mean_token_accuracy": 0.8979595750570297,
      "num_tokens": 3009984.0,
      "step": 244
    },
    {
      "entropy": 1.3404588103294373,
      "epoch": 0.129015271195366,
      "grad_norm": 0.5311986804008484,
      "learning_rate": 0.00012230576441102757,
      "loss": 0.2417442798614502,
      "mean_token_accuracy": 0.9036334156990051,
      "num_tokens": 3022320.0,
      "step": 245
    },
    {
      "entropy": 1.3250301778316498,
      "epoch": 0.12954186413902052,
      "grad_norm": 0.6158598065376282,
      "learning_rate": 0.00012280701754385965,
      "loss": 0.28592783212661743,
      "mean_token_accuracy": 0.8968044072389603,
      "num_tokens": 3034656.0,
      "step": 246
    },
    {
      "entropy": 1.2895326018333435,
      "epoch": 0.1300684570826751,
      "grad_norm": 0.5133572816848755,
      "learning_rate": 0.00012330827067669173,
      "loss": 0.26722657680511475,
      "mean_token_accuracy": 0.899601012468338,
      "num_tokens": 3046992.0,
      "step": 247
    },
    {
      "entropy": 1.2834957540035248,
      "epoch": 0.13059505002632965,
      "grad_norm": 0.5774098038673401,
      "learning_rate": 0.0001238095238095238,
      "loss": 0.2552984654903412,
      "mean_token_accuracy": 0.9021749496459961,
      "num_tokens": 3059328.0,
      "step": 248
    },
    {
      "entropy": 1.28067284822464,
      "epoch": 0.13112164296998421,
      "grad_norm": 0.5813867449760437,
      "learning_rate": 0.0001243107769423559,
      "loss": 0.26484355330467224,
      "mean_token_accuracy": 0.8958484381437302,
      "num_tokens": 3071664.0,
      "step": 249
    },
    {
      "entropy": 1.3292741775512695,
      "epoch": 0.13164823591363875,
      "grad_norm": 0.5437096357345581,
      "learning_rate": 0.00012481203007518797,
      "loss": 0.25312983989715576,
      "mean_token_accuracy": 0.9083140790462494,
      "num_tokens": 3084000.0,
      "step": 250
    },
    {
      "entropy": 1.217045933008194,
      "epoch": 0.1321748288572933,
      "grad_norm": 0.5642932057380676,
      "learning_rate": 0.00012531328320802005,
      "loss": 0.24098949134349823,
      "mean_token_accuracy": 0.9082589000463486,
      "num_tokens": 3096336.0,
      "step": 251
    },
    {
      "entropy": 1.3332138061523438,
      "epoch": 0.13270142180094788,
      "grad_norm": 0.5955415368080139,
      "learning_rate": 0.00012581453634085213,
      "loss": 0.2849401831626892,
      "mean_token_accuracy": 0.8916681408882141,
      "num_tokens": 3108672.0,
      "step": 252
    },
    {
      "entropy": 1.295847773551941,
      "epoch": 0.1332280147446024,
      "grad_norm": 0.6146616339683533,
      "learning_rate": 0.0001263157894736842,
      "loss": 0.2735283076763153,
      "mean_token_accuracy": 0.8989967852830887,
      "num_tokens": 3121008.0,
      "step": 253
    },
    {
      "entropy": 1.2074873447418213,
      "epoch": 0.13375460768825698,
      "grad_norm": 0.5243288278579712,
      "learning_rate": 0.0001268170426065163,
      "loss": 0.24956879019737244,
      "mean_token_accuracy": 0.9056222587823868,
      "num_tokens": 3133344.0,
      "step": 254
    },
    {
      "entropy": 1.2322421371936798,
      "epoch": 0.13428120063191154,
      "grad_norm": 0.5780612230300903,
      "learning_rate": 0.00012731829573934836,
      "loss": 0.22142812609672546,
      "mean_token_accuracy": 0.9111984372138977,
      "num_tokens": 3145680.0,
      "step": 255
    },
    {
      "entropy": 1.2550228536128998,
      "epoch": 0.13480779357556608,
      "grad_norm": 0.5654560923576355,
      "learning_rate": 0.00012781954887218047,
      "loss": 0.2577667236328125,
      "mean_token_accuracy": 0.9028536081314087,
      "num_tokens": 3158016.0,
      "step": 256
    },
    {
      "entropy": 1.2815439105033875,
      "epoch": 0.13533438651922064,
      "grad_norm": 0.5860424041748047,
      "learning_rate": 0.00012832080200501252,
      "loss": 0.24961405992507935,
      "mean_token_accuracy": 0.9073447585105896,
      "num_tokens": 3170352.0,
      "step": 257
    },
    {
      "entropy": 1.3496994376182556,
      "epoch": 0.1358609794628752,
      "grad_norm": 0.5496608018875122,
      "learning_rate": 0.00012882205513784463,
      "loss": 0.23788216710090637,
      "mean_token_accuracy": 0.913466677069664,
      "num_tokens": 3182688.0,
      "step": 258
    },
    {
      "entropy": 1.2379789352416992,
      "epoch": 0.13638757240652974,
      "grad_norm": 0.5407847166061401,
      "learning_rate": 0.00012932330827067668,
      "loss": 0.242479607462883,
      "mean_token_accuracy": 0.9110987186431885,
      "num_tokens": 3195024.0,
      "step": 259
    },
    {
      "entropy": 1.2678188979625702,
      "epoch": 0.1369141653501843,
      "grad_norm": 0.6954227089881897,
      "learning_rate": 0.0001298245614035088,
      "loss": 0.3035907447338104,
      "mean_token_accuracy": 0.8935103565454483,
      "num_tokens": 3207360.0,
      "step": 260
    },
    {
      "entropy": 1.2501021325588226,
      "epoch": 0.13744075829383887,
      "grad_norm": 0.6763582825660706,
      "learning_rate": 0.00013032581453634084,
      "loss": 0.2833873927593231,
      "mean_token_accuracy": 0.8930742889642715,
      "num_tokens": 3219696.0,
      "step": 261
    },
    {
      "entropy": 1.2093367278575897,
      "epoch": 0.13796735123749343,
      "grad_norm": 0.5583021640777588,
      "learning_rate": 0.00013082706766917294,
      "loss": 0.23031063377857208,
      "mean_token_accuracy": 0.9110294431447983,
      "num_tokens": 3232032.0,
      "step": 262
    },
    {
      "entropy": 1.2286772429943085,
      "epoch": 0.13849394418114797,
      "grad_norm": 0.5849270224571228,
      "learning_rate": 0.00013132832080200502,
      "loss": 0.27444857358932495,
      "mean_token_accuracy": 0.8965310454368591,
      "num_tokens": 3244368.0,
      "step": 263
    },
    {
      "entropy": 1.2541997134685516,
      "epoch": 0.13902053712480253,
      "grad_norm": 0.5210989713668823,
      "learning_rate": 0.0001318295739348371,
      "loss": 0.24464797973632812,
      "mean_token_accuracy": 0.9072559326887131,
      "num_tokens": 3256704.0,
      "step": 264
    },
    {
      "entropy": 1.2401838898658752,
      "epoch": 0.1395471300684571,
      "grad_norm": 0.5357401371002197,
      "learning_rate": 0.00013233082706766918,
      "loss": 0.24787980318069458,
      "mean_token_accuracy": 0.9098693877458572,
      "num_tokens": 3269040.0,
      "step": 265
    },
    {
      "entropy": 1.2660530507564545,
      "epoch": 0.14007372301211163,
      "grad_norm": 0.5940161943435669,
      "learning_rate": 0.00013283208020050126,
      "loss": 0.27411437034606934,
      "mean_token_accuracy": 0.8980700373649597,
      "num_tokens": 3281376.0,
      "step": 266
    },
    {
      "entropy": 1.241824746131897,
      "epoch": 0.1406003159557662,
      "grad_norm": 0.604097306728363,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.2866377830505371,
      "mean_token_accuracy": 0.8919961899518967,
      "num_tokens": 3293712.0,
      "step": 267
    },
    {
      "entropy": 1.1769486963748932,
      "epoch": 0.14112690889942076,
      "grad_norm": 0.7264558672904968,
      "learning_rate": 0.00013383458646616542,
      "loss": 0.26414263248443604,
      "mean_token_accuracy": 0.8967715352773666,
      "num_tokens": 3306048.0,
      "step": 268
    },
    {
      "entropy": 1.2283678948879242,
      "epoch": 0.1416535018430753,
      "grad_norm": 0.5465900301933289,
      "learning_rate": 0.0001343358395989975,
      "loss": 0.2366294264793396,
      "mean_token_accuracy": 0.9092733561992645,
      "num_tokens": 3318384.0,
      "step": 269
    },
    {
      "entropy": 1.262256383895874,
      "epoch": 0.14218009478672985,
      "grad_norm": 0.5315939784049988,
      "learning_rate": 0.00013483709273182958,
      "loss": 0.22708219289779663,
      "mean_token_accuracy": 0.914520725607872,
      "num_tokens": 3330720.0,
      "step": 270
    },
    {
      "entropy": 1.175988882780075,
      "epoch": 0.14270668773038442,
      "grad_norm": 0.5613946318626404,
      "learning_rate": 0.00013533834586466166,
      "loss": 0.22835803031921387,
      "mean_token_accuracy": 0.9081931263208389,
      "num_tokens": 3343056.0,
      "step": 271
    },
    {
      "entropy": 1.2391490042209625,
      "epoch": 0.14323328067403895,
      "grad_norm": 0.5381202101707458,
      "learning_rate": 0.00013583959899749373,
      "loss": 0.25535333156585693,
      "mean_token_accuracy": 0.9033481776714325,
      "num_tokens": 3355392.0,
      "step": 272
    },
    {
      "entropy": 1.2381133437156677,
      "epoch": 0.14375987361769352,
      "grad_norm": 0.6545175909996033,
      "learning_rate": 0.00013634085213032581,
      "loss": 0.268245667219162,
      "mean_token_accuracy": 0.9008642733097076,
      "num_tokens": 3367728.0,
      "step": 273
    },
    {
      "entropy": 1.2534950375556946,
      "epoch": 0.14428646656134808,
      "grad_norm": 0.6084462404251099,
      "learning_rate": 0.0001368421052631579,
      "loss": 0.241957426071167,
      "mean_token_accuracy": 0.9131230562925339,
      "num_tokens": 3380064.0,
      "step": 274
    },
    {
      "entropy": 1.2566357254981995,
      "epoch": 0.14481305950500264,
      "grad_norm": 0.5803132057189941,
      "learning_rate": 0.00013734335839598997,
      "loss": 0.24542784690856934,
      "mean_token_accuracy": 0.9083303511142731,
      "num_tokens": 3392400.0,
      "step": 275
    },
    {
      "entropy": 1.3298097252845764,
      "epoch": 0.14533965244865718,
      "grad_norm": 0.5581626296043396,
      "learning_rate": 0.00013784461152882208,
      "loss": 0.2682327330112457,
      "mean_token_accuracy": 0.8970459997653961,
      "num_tokens": 3404736.0,
      "step": 276
    },
    {
      "entropy": 1.2984472513198853,
      "epoch": 0.14586624539231174,
      "grad_norm": 0.5797234177589417,
      "learning_rate": 0.00013834586466165413,
      "loss": 0.2551107406616211,
      "mean_token_accuracy": 0.9025808274745941,
      "num_tokens": 3417072.0,
      "step": 277
    },
    {
      "entropy": 1.232091337442398,
      "epoch": 0.1463928383359663,
      "grad_norm": 0.56398606300354,
      "learning_rate": 0.00013884711779448624,
      "loss": 0.2523190379142761,
      "mean_token_accuracy": 0.9042512625455856,
      "num_tokens": 3429408.0,
      "step": 278
    },
    {
      "entropy": 1.364453375339508,
      "epoch": 0.14691943127962084,
      "grad_norm": 0.5884234309196472,
      "learning_rate": 0.0001393483709273183,
      "loss": 0.2788117229938507,
      "mean_token_accuracy": 0.90156190097332,
      "num_tokens": 3441744.0,
      "step": 279
    },
    {
      "entropy": 1.2391767501831055,
      "epoch": 0.1474460242232754,
      "grad_norm": 0.5528289079666138,
      "learning_rate": 0.0001398496240601504,
      "loss": 0.2562858760356903,
      "mean_token_accuracy": 0.905647337436676,
      "num_tokens": 3454080.0,
      "step": 280
    },
    {
      "entropy": 1.3013744056224823,
      "epoch": 0.14797261716692997,
      "grad_norm": 0.5353385806083679,
      "learning_rate": 0.00014035087719298245,
      "loss": 0.24675647914409637,
      "mean_token_accuracy": 0.9100503474473953,
      "num_tokens": 3466416.0,
      "step": 281
    },
    {
      "entropy": 1.2755855023860931,
      "epoch": 0.1484992101105845,
      "grad_norm": 0.5495297908782959,
      "learning_rate": 0.00014085213032581455,
      "loss": 0.2519223690032959,
      "mean_token_accuracy": 0.8966221362352371,
      "num_tokens": 3478752.0,
      "step": 282
    },
    {
      "entropy": 1.303219497203827,
      "epoch": 0.14902580305423907,
      "grad_norm": 0.5309876203536987,
      "learning_rate": 0.0001413533834586466,
      "loss": 0.24427109956741333,
      "mean_token_accuracy": 0.9075532704591751,
      "num_tokens": 3491088.0,
      "step": 283
    },
    {
      "entropy": 1.2614858448505402,
      "epoch": 0.14955239599789363,
      "grad_norm": 0.5807812809944153,
      "learning_rate": 0.0001418546365914787,
      "loss": 0.26809874176979065,
      "mean_token_accuracy": 0.899653896689415,
      "num_tokens": 3503424.0,
      "step": 284
    },
    {
      "entropy": 1.2611050605773926,
      "epoch": 0.1500789889415482,
      "grad_norm": 0.5592539310455322,
      "learning_rate": 0.0001423558897243108,
      "loss": 0.26588040590286255,
      "mean_token_accuracy": 0.8970725983381271,
      "num_tokens": 3515760.0,
      "step": 285
    },
    {
      "entropy": 1.2830645143985748,
      "epoch": 0.15060558188520273,
      "grad_norm": 0.5146520733833313,
      "learning_rate": 0.00014285714285714287,
      "loss": 0.24322782456874847,
      "mean_token_accuracy": 0.9055588096380234,
      "num_tokens": 3528096.0,
      "step": 286
    },
    {
      "entropy": 1.2292780876159668,
      "epoch": 0.1511321748288573,
      "grad_norm": 0.588441789150238,
      "learning_rate": 0.00014335839598997495,
      "loss": 0.2558269202709198,
      "mean_token_accuracy": 0.8964600563049316,
      "num_tokens": 3540432.0,
      "step": 287
    },
    {
      "entropy": 1.2410956919193268,
      "epoch": 0.15165876777251186,
      "grad_norm": 0.5761001706123352,
      "learning_rate": 0.00014385964912280703,
      "loss": 0.2511615455150604,
      "mean_token_accuracy": 0.9040048718452454,
      "num_tokens": 3552768.0,
      "step": 288
    },
    {
      "entropy": 1.3154540359973907,
      "epoch": 0.1521853607161664,
      "grad_norm": 0.6528503894805908,
      "learning_rate": 0.0001443609022556391,
      "loss": 0.2512635290622711,
      "mean_token_accuracy": 0.9024206399917603,
      "num_tokens": 3565104.0,
      "step": 289
    },
    {
      "entropy": 1.2947738468647003,
      "epoch": 0.15271195365982096,
      "grad_norm": 0.6061785221099854,
      "learning_rate": 0.00014486215538847118,
      "loss": 0.25342220067977905,
      "mean_token_accuracy": 0.9054669588804245,
      "num_tokens": 3577440.0,
      "step": 290
    },
    {
      "entropy": 1.2879831790924072,
      "epoch": 0.15323854660347552,
      "grad_norm": 0.5570182800292969,
      "learning_rate": 0.00014536340852130326,
      "loss": 0.27056393027305603,
      "mean_token_accuracy": 0.8992442786693573,
      "num_tokens": 3589776.0,
      "step": 291
    },
    {
      "entropy": 1.3449256420135498,
      "epoch": 0.15376513954713006,
      "grad_norm": 0.4950920045375824,
      "learning_rate": 0.00014586466165413534,
      "loss": 0.24097028374671936,
      "mean_token_accuracy": 0.9014376401901245,
      "num_tokens": 3602112.0,
      "step": 292
    },
    {
      "entropy": 1.4016221165657043,
      "epoch": 0.15429173249078462,
      "grad_norm": 0.554709792137146,
      "learning_rate": 0.00014636591478696742,
      "loss": 0.2270800769329071,
      "mean_token_accuracy": 0.9162011295557022,
      "num_tokens": 3614448.0,
      "step": 293
    },
    {
      "entropy": 1.514324426651001,
      "epoch": 0.15481832543443919,
      "grad_norm": 0.5531271696090698,
      "learning_rate": 0.0001468671679197995,
      "loss": 0.24683888256549835,
      "mean_token_accuracy": 0.9074130207300186,
      "num_tokens": 3626784.0,
      "step": 294
    },
    {
      "entropy": 1.471198707818985,
      "epoch": 0.15534491837809372,
      "grad_norm": 0.6712215542793274,
      "learning_rate": 0.00014736842105263158,
      "loss": 0.281759113073349,
      "mean_token_accuracy": 0.8955343812704086,
      "num_tokens": 3639120.0,
      "step": 295
    },
    {
      "entropy": 1.4630400836467743,
      "epoch": 0.15587151132174829,
      "grad_norm": 0.561416506767273,
      "learning_rate": 0.00014786967418546366,
      "loss": 0.275588721036911,
      "mean_token_accuracy": 0.8987540155649185,
      "num_tokens": 3651456.0,
      "step": 296
    },
    {
      "entropy": 1.491294264793396,
      "epoch": 0.15639810426540285,
      "grad_norm": 0.5332501530647278,
      "learning_rate": 0.00014837092731829574,
      "loss": 0.2502642273902893,
      "mean_token_accuracy": 0.9047496467828751,
      "num_tokens": 3663792.0,
      "step": 297
    },
    {
      "entropy": 1.3809879124164581,
      "epoch": 0.1569246972090574,
      "grad_norm": 0.499104768037796,
      "learning_rate": 0.00014887218045112784,
      "loss": 0.23079532384872437,
      "mean_token_accuracy": 0.909051313996315,
      "num_tokens": 3676128.0,
      "step": 298
    },
    {
      "entropy": 1.3202951848506927,
      "epoch": 0.15745129015271195,
      "grad_norm": 0.5114240646362305,
      "learning_rate": 0.0001493734335839599,
      "loss": 0.26147356629371643,
      "mean_token_accuracy": 0.9033501297235489,
      "num_tokens": 3688464.0,
      "step": 299
    },
    {
      "entropy": 1.3542028069496155,
      "epoch": 0.1579778830963665,
      "grad_norm": 0.492830753326416,
      "learning_rate": 0.000149874686716792,
      "loss": 0.2421473264694214,
      "mean_token_accuracy": 0.9081345945596695,
      "num_tokens": 3700800.0,
      "step": 300
    },
    {
      "entropy": 1.2910203635692596,
      "epoch": 0.15850447604002108,
      "grad_norm": 0.4862864911556244,
      "learning_rate": 0.00015037593984962405,
      "loss": 0.24877643585205078,
      "mean_token_accuracy": 0.9057342112064362,
      "num_tokens": 3713136.0,
      "step": 301
    },
    {
      "entropy": 1.3562625646591187,
      "epoch": 0.1590310689836756,
      "grad_norm": 0.6064998507499695,
      "learning_rate": 0.00015087719298245616,
      "loss": 0.2678510546684265,
      "mean_token_accuracy": 0.895992174744606,
      "num_tokens": 3725472.0,
      "step": 302
    },
    {
      "entropy": 1.3357418477535248,
      "epoch": 0.15955766192733017,
      "grad_norm": 0.5504618883132935,
      "learning_rate": 0.0001513784461152882,
      "loss": 0.24523413181304932,
      "mean_token_accuracy": 0.9103363454341888,
      "num_tokens": 3737808.0,
      "step": 303
    },
    {
      "entropy": 1.3335883915424347,
      "epoch": 0.16008425487098474,
      "grad_norm": 0.73444664478302,
      "learning_rate": 0.00015187969924812032,
      "loss": 0.2613740563392639,
      "mean_token_accuracy": 0.8996832072734833,
      "num_tokens": 3750144.0,
      "step": 304
    },
    {
      "entropy": 1.3423579335212708,
      "epoch": 0.16061084781463927,
      "grad_norm": 0.5342377424240112,
      "learning_rate": 0.00015238095238095237,
      "loss": 0.27072861790657043,
      "mean_token_accuracy": 0.8979348689317703,
      "num_tokens": 3762480.0,
      "step": 305
    },
    {
      "entropy": 1.2893436253070831,
      "epoch": 0.16113744075829384,
      "grad_norm": 0.49320632219314575,
      "learning_rate": 0.00015288220551378448,
      "loss": 0.24137352406978607,
      "mean_token_accuracy": 0.9123422801494598,
      "num_tokens": 3774816.0,
      "step": 306
    },
    {
      "entropy": 1.3107283115386963,
      "epoch": 0.1616640337019484,
      "grad_norm": 0.5037325620651245,
      "learning_rate": 0.00015338345864661653,
      "loss": 0.28284943103790283,
      "mean_token_accuracy": 0.8933527320623398,
      "num_tokens": 3787152.0,
      "step": 307
    },
    {
      "entropy": 1.2858748137950897,
      "epoch": 0.16219062664560294,
      "grad_norm": 0.7596031427383423,
      "learning_rate": 0.00015388471177944863,
      "loss": 0.2364031970500946,
      "mean_token_accuracy": 0.9070021212100983,
      "num_tokens": 3799488.0,
      "step": 308
    },
    {
      "entropy": 1.2649638950824738,
      "epoch": 0.1627172195892575,
      "grad_norm": 0.4712359607219696,
      "learning_rate": 0.0001543859649122807,
      "loss": 0.25866541266441345,
      "mean_token_accuracy": 0.8989972174167633,
      "num_tokens": 3811824.0,
      "step": 309
    },
    {
      "entropy": 1.1775002479553223,
      "epoch": 0.16324381253291206,
      "grad_norm": 0.47527384757995605,
      "learning_rate": 0.0001548872180451128,
      "loss": 0.22686654329299927,
      "mean_token_accuracy": 0.9112849235534668,
      "num_tokens": 3824160.0,
      "step": 310
    },
    {
      "entropy": 1.1848091185092926,
      "epoch": 0.16377040547656663,
      "grad_norm": 0.5009036064147949,
      "learning_rate": 0.00015538847117794487,
      "loss": 0.23199157416820526,
      "mean_token_accuracy": 0.9103835225105286,
      "num_tokens": 3836496.0,
      "step": 311
    },
    {
      "entropy": 1.1879796385765076,
      "epoch": 0.16429699842022116,
      "grad_norm": 0.530559778213501,
      "learning_rate": 0.00015588972431077695,
      "loss": 0.2687099277973175,
      "mean_token_accuracy": 0.9024157524108887,
      "num_tokens": 3848832.0,
      "step": 312
    },
    {
      "entropy": 1.1956203877925873,
      "epoch": 0.16482359136387573,
      "grad_norm": 0.5160781145095825,
      "learning_rate": 0.00015639097744360903,
      "loss": 0.24907854199409485,
      "mean_token_accuracy": 0.9046753495931625,
      "num_tokens": 3861168.0,
      "step": 313
    },
    {
      "entropy": 1.1685654819011688,
      "epoch": 0.1653501843075303,
      "grad_norm": 0.5312115550041199,
      "learning_rate": 0.0001568922305764411,
      "loss": 0.2576726973056793,
      "mean_token_accuracy": 0.9074427336454391,
      "num_tokens": 3873504.0,
      "step": 314
    },
    {
      "entropy": 1.1273577511310577,
      "epoch": 0.16587677725118483,
      "grad_norm": 0.6181286573410034,
      "learning_rate": 0.00015739348370927319,
      "loss": 0.23891520500183105,
      "mean_token_accuracy": 0.9119571298360825,
      "num_tokens": 3885840.0,
      "step": 315
    },
    {
      "entropy": 1.2541460990905762,
      "epoch": 0.1664033701948394,
      "grad_norm": 0.4766935408115387,
      "learning_rate": 0.00015789473684210527,
      "loss": 0.25483545660972595,
      "mean_token_accuracy": 0.8961910307407379,
      "num_tokens": 3898176.0,
      "step": 316
    },
    {
      "entropy": 1.1125807464122772,
      "epoch": 0.16692996313849395,
      "grad_norm": 0.5105007886886597,
      "learning_rate": 0.00015839598997493734,
      "loss": 0.2685958445072174,
      "mean_token_accuracy": 0.8989451825618744,
      "num_tokens": 3910512.0,
      "step": 317
    },
    {
      "entropy": 1.2121298015117645,
      "epoch": 0.1674565560821485,
      "grad_norm": 0.44646549224853516,
      "learning_rate": 0.00015889724310776942,
      "loss": 0.253686785697937,
      "mean_token_accuracy": 0.9083440899848938,
      "num_tokens": 3922848.0,
      "step": 318
    },
    {
      "entropy": 1.1011106967926025,
      "epoch": 0.16798314902580305,
      "grad_norm": 0.46656280755996704,
      "learning_rate": 0.0001593984962406015,
      "loss": 0.2927643060684204,
      "mean_token_accuracy": 0.8931492418050766,
      "num_tokens": 3935184.0,
      "step": 319
    },
    {
      "entropy": 1.1864506900310516,
      "epoch": 0.16850974196945762,
      "grad_norm": 0.4583514332771301,
      "learning_rate": 0.00015989974937343358,
      "loss": 0.22579318284988403,
      "mean_token_accuracy": 0.9128856211900711,
      "num_tokens": 3947520.0,
      "step": 320
    },
    {
      "entropy": 1.113357663154602,
      "epoch": 0.16903633491311215,
      "grad_norm": 0.5008531212806702,
      "learning_rate": 0.00016040100250626566,
      "loss": 0.2566073536872864,
      "mean_token_accuracy": 0.9067006707191467,
      "num_tokens": 3959856.0,
      "step": 321
    },
    {
      "entropy": 1.1666322350502014,
      "epoch": 0.16956292785676672,
      "grad_norm": 0.5016852021217346,
      "learning_rate": 0.00016090225563909777,
      "loss": 0.24807997047901154,
      "mean_token_accuracy": 0.9074774533510208,
      "num_tokens": 3972192.0,
      "step": 322
    },
    {
      "entropy": 1.0860579013824463,
      "epoch": 0.17008952080042128,
      "grad_norm": 0.5153301358222961,
      "learning_rate": 0.00016140350877192982,
      "loss": 0.2500810921192169,
      "mean_token_accuracy": 0.9049581736326218,
      "num_tokens": 3984528.0,
      "step": 323
    },
    {
      "entropy": 1.0754130184650421,
      "epoch": 0.17061611374407584,
      "grad_norm": 0.5485661625862122,
      "learning_rate": 0.00016190476190476192,
      "loss": 0.3087775707244873,
      "mean_token_accuracy": 0.8840746134519577,
      "num_tokens": 3996864.0,
      "step": 324
    },
    {
      "entropy": 1.097093939781189,
      "epoch": 0.17114270668773038,
      "grad_norm": 0.4524535834789276,
      "learning_rate": 0.00016240601503759398,
      "loss": 0.2556874454021454,
      "mean_token_accuracy": 0.9040686786174774,
      "num_tokens": 4009200.0,
      "step": 325
    },
    {
      "entropy": 1.0161993354558945,
      "epoch": 0.17166929963138494,
      "grad_norm": 0.5920532941818237,
      "learning_rate": 0.00016290726817042608,
      "loss": 0.2545970678329468,
      "mean_token_accuracy": 0.9038624912500381,
      "num_tokens": 4021536.0,
      "step": 326
    },
    {
      "entropy": 1.045365571975708,
      "epoch": 0.1721958925750395,
      "grad_norm": 0.5128290057182312,
      "learning_rate": 0.00016340852130325813,
      "loss": 0.23849724233150482,
      "mean_token_accuracy": 0.9063487350940704,
      "num_tokens": 4033872.0,
      "step": 327
    },
    {
      "entropy": 1.0524345636367798,
      "epoch": 0.17272248551869404,
      "grad_norm": 0.5058865547180176,
      "learning_rate": 0.00016390977443609024,
      "loss": 0.25231170654296875,
      "mean_token_accuracy": 0.9029942154884338,
      "num_tokens": 4046208.0,
      "step": 328
    },
    {
      "entropy": 1.1274868845939636,
      "epoch": 0.1732490784623486,
      "grad_norm": 0.5103746652603149,
      "learning_rate": 0.0001644110275689223,
      "loss": 0.26119422912597656,
      "mean_token_accuracy": 0.8970126211643219,
      "num_tokens": 4058544.0,
      "step": 329
    },
    {
      "entropy": 1.13772451877594,
      "epoch": 0.17377567140600317,
      "grad_norm": 0.463789701461792,
      "learning_rate": 0.0001649122807017544,
      "loss": 0.23172324895858765,
      "mean_token_accuracy": 0.9090371131896973,
      "num_tokens": 4070880.0,
      "step": 330
    },
    {
      "entropy": 1.1093823909759521,
      "epoch": 0.1743022643496577,
      "grad_norm": 0.4940638542175293,
      "learning_rate": 0.00016541353383458648,
      "loss": 0.25032839179039,
      "mean_token_accuracy": 0.9061925411224365,
      "num_tokens": 4083216.0,
      "step": 331
    },
    {
      "entropy": 1.0906598567962646,
      "epoch": 0.17482885729331227,
      "grad_norm": 0.5244042873382568,
      "learning_rate": 0.00016591478696741856,
      "loss": 0.2559994161128998,
      "mean_token_accuracy": 0.9056536555290222,
      "num_tokens": 4095552.0,
      "step": 332
    },
    {
      "entropy": 1.0064781606197357,
      "epoch": 0.17535545023696683,
      "grad_norm": 0.5328090786933899,
      "learning_rate": 0.00016641604010025064,
      "loss": 0.2423868030309677,
      "mean_token_accuracy": 0.9023227244615555,
      "num_tokens": 4107888.0,
      "step": 333
    },
    {
      "entropy": 0.9421130269765854,
      "epoch": 0.17588204318062137,
      "grad_norm": 0.4938746392726898,
      "learning_rate": 0.00016691729323308271,
      "loss": 0.2487686276435852,
      "mean_token_accuracy": 0.9072201699018478,
      "num_tokens": 4120224.0,
      "step": 334
    },
    {
      "entropy": 0.8471245914697647,
      "epoch": 0.17640863612427593,
      "grad_norm": 0.4719039499759674,
      "learning_rate": 0.0001674185463659148,
      "loss": 0.2677307426929474,
      "mean_token_accuracy": 0.89954973757267,
      "num_tokens": 4132560.0,
      "step": 335
    },
    {
      "entropy": 0.7532882988452911,
      "epoch": 0.1769352290679305,
      "grad_norm": 0.4836346209049225,
      "learning_rate": 0.00016791979949874687,
      "loss": 0.2508479356765747,
      "mean_token_accuracy": 0.9021095633506775,
      "num_tokens": 4144896.0,
      "step": 336
    },
    {
      "entropy": 0.6489329487085342,
      "epoch": 0.17746182201158506,
      "grad_norm": 0.5742446184158325,
      "learning_rate": 0.00016842105263157895,
      "loss": 0.2582406997680664,
      "mean_token_accuracy": 0.9013715237379074,
      "num_tokens": 4157232.0,
      "step": 337
    },
    {
      "entropy": 0.6537502855062485,
      "epoch": 0.1779884149552396,
      "grad_norm": 0.5390785336494446,
      "learning_rate": 0.00016892230576441103,
      "loss": 0.2649543881416321,
      "mean_token_accuracy": 0.8998610377311707,
      "num_tokens": 4169568.0,
      "step": 338
    },
    {
      "entropy": 0.6827399134635925,
      "epoch": 0.17851500789889416,
      "grad_norm": 0.507554829120636,
      "learning_rate": 0.0001694235588972431,
      "loss": 0.27018067240715027,
      "mean_token_accuracy": 0.8985839486122131,
      "num_tokens": 4181904.0,
      "step": 339
    },
    {
      "entropy": 0.7333462089300156,
      "epoch": 0.17904160084254872,
      "grad_norm": 0.4551506042480469,
      "learning_rate": 0.0001699248120300752,
      "loss": 0.25013625621795654,
      "mean_token_accuracy": 0.9111319333314896,
      "num_tokens": 4194240.0,
      "step": 340
    },
    {
      "entropy": 0.7495439946651459,
      "epoch": 0.17956819378620326,
      "grad_norm": 0.48456624150276184,
      "learning_rate": 0.00017042606516290727,
      "loss": 0.2707773745059967,
      "mean_token_accuracy": 0.8947316855192184,
      "num_tokens": 4206576.0,
      "step": 341
    },
    {
      "entropy": 0.7812775671482086,
      "epoch": 0.18009478672985782,
      "grad_norm": 0.4550190269947052,
      "learning_rate": 0.00017092731829573935,
      "loss": 0.2396947741508484,
      "mean_token_accuracy": 0.910267636179924,
      "num_tokens": 4218912.0,
      "step": 342
    },
    {
      "entropy": 0.9400374740362167,
      "epoch": 0.18062137967351238,
      "grad_norm": 0.5784738659858704,
      "learning_rate": 0.00017142857142857143,
      "loss": 0.26176217198371887,
      "mean_token_accuracy": 0.9009139388799667,
      "num_tokens": 4231248.0,
      "step": 343
    },
    {
      "entropy": 0.937469407916069,
      "epoch": 0.18114797261716692,
      "grad_norm": 0.5101773738861084,
      "learning_rate": 0.00017192982456140353,
      "loss": 0.222784623503685,
      "mean_token_accuracy": 0.9079811573028564,
      "num_tokens": 4243584.0,
      "step": 344
    },
    {
      "entropy": 0.9701919555664062,
      "epoch": 0.18167456556082148,
      "grad_norm": 0.4810556471347809,
      "learning_rate": 0.00017243107769423558,
      "loss": 0.25527724623680115,
      "mean_token_accuracy": 0.8977989107370377,
      "num_tokens": 4255920.0,
      "step": 345
    },
    {
      "entropy": 1.0421793162822723,
      "epoch": 0.18220115850447605,
      "grad_norm": 0.49446040391921997,
      "learning_rate": 0.0001729323308270677,
      "loss": 0.24790042638778687,
      "mean_token_accuracy": 0.9076082557439804,
      "num_tokens": 4268256.0,
      "step": 346
    },
    {
      "entropy": 1.0526159405708313,
      "epoch": 0.18272775144813058,
      "grad_norm": 0.5106586217880249,
      "learning_rate": 0.00017343358395989974,
      "loss": 0.258137047290802,
      "mean_token_accuracy": 0.9013570845127106,
      "num_tokens": 4280592.0,
      "step": 347
    },
    {
      "entropy": 1.0377429127693176,
      "epoch": 0.18325434439178515,
      "grad_norm": 0.5372883677482605,
      "learning_rate": 0.00017393483709273185,
      "loss": 0.2536155879497528,
      "mean_token_accuracy": 0.9079491943120956,
      "num_tokens": 4292928.0,
      "step": 348
    },
    {
      "entropy": 1.1226612329483032,
      "epoch": 0.1837809373354397,
      "grad_norm": 0.506536066532135,
      "learning_rate": 0.0001744360902255639,
      "loss": 0.2503521740436554,
      "mean_token_accuracy": 0.9043506681919098,
      "num_tokens": 4305264.0,
      "step": 349
    },
    {
      "entropy": 1.0955768823623657,
      "epoch": 0.18430753027909427,
      "grad_norm": 0.450717031955719,
      "learning_rate": 0.000174937343358396,
      "loss": 0.2611716687679291,
      "mean_token_accuracy": 0.9029702991247177,
      "num_tokens": 4317600.0,
      "step": 350
    },
    {
      "entropy": 1.1000354140996933,
      "epoch": 0.1848341232227488,
      "grad_norm": 0.5249627828598022,
      "learning_rate": 0.00017543859649122806,
      "loss": 0.23257189989089966,
      "mean_token_accuracy": 0.9120175540447235,
      "num_tokens": 4329936.0,
      "step": 351
    },
    {
      "entropy": 1.086414873600006,
      "epoch": 0.18536071616640337,
      "grad_norm": 0.488702654838562,
      "learning_rate": 0.00017593984962406016,
      "loss": 0.246709406375885,
      "mean_token_accuracy": 0.9029638320207596,
      "num_tokens": 4342272.0,
      "step": 352
    },
    {
      "entropy": 1.0042328536510468,
      "epoch": 0.18588730911005794,
      "grad_norm": 0.41849616169929504,
      "learning_rate": 0.00017644110275689224,
      "loss": 0.21574650704860687,
      "mean_token_accuracy": 0.9175253808498383,
      "num_tokens": 4354608.0,
      "step": 353
    },
    {
      "entropy": 1.0020455569028854,
      "epoch": 0.18641390205371247,
      "grad_norm": 0.46401116251945496,
      "learning_rate": 0.00017694235588972432,
      "loss": 0.2567673325538635,
      "mean_token_accuracy": 0.9054214805364609,
      "num_tokens": 4366944.0,
      "step": 354
    },
    {
      "entropy": 1.0092186033725739,
      "epoch": 0.18694049499736703,
      "grad_norm": 0.47048723697662354,
      "learning_rate": 0.0001774436090225564,
      "loss": 0.24711501598358154,
      "mean_token_accuracy": 0.9056050777435303,
      "num_tokens": 4379280.0,
      "step": 355
    },
    {
      "entropy": 0.9878545552492142,
      "epoch": 0.1874670879410216,
      "grad_norm": 0.47208353877067566,
      "learning_rate": 0.00017794486215538848,
      "loss": 0.2748403549194336,
      "mean_token_accuracy": 0.8964878469705582,
      "num_tokens": 4391616.0,
      "step": 356
    },
    {
      "entropy": 0.9895420670509338,
      "epoch": 0.18799368088467613,
      "grad_norm": 0.503665566444397,
      "learning_rate": 0.00017844611528822056,
      "loss": 0.24393996596336365,
      "mean_token_accuracy": 0.9081792533397675,
      "num_tokens": 4403952.0,
      "step": 357
    },
    {
      "entropy": 0.983001708984375,
      "epoch": 0.1885202738283307,
      "grad_norm": 0.5069933533668518,
      "learning_rate": 0.00017894736842105264,
      "loss": 0.2519129514694214,
      "mean_token_accuracy": 0.9055497497320175,
      "num_tokens": 4416288.0,
      "step": 358
    },
    {
      "entropy": 0.9950737804174423,
      "epoch": 0.18904686677198526,
      "grad_norm": 0.4974028766155243,
      "learning_rate": 0.00017944862155388472,
      "loss": 0.26243290305137634,
      "mean_token_accuracy": 0.8986903131008148,
      "num_tokens": 4428624.0,
      "step": 359
    },
    {
      "entropy": 1.0079210847616196,
      "epoch": 0.1895734597156398,
      "grad_norm": 0.5110583901405334,
      "learning_rate": 0.0001799498746867168,
      "loss": 0.2294146716594696,
      "mean_token_accuracy": 0.9131651520729065,
      "num_tokens": 4440960.0,
      "step": 360
    },
    {
      "entropy": 1.0026690363883972,
      "epoch": 0.19010005265929436,
      "grad_norm": 0.4598653316497803,
      "learning_rate": 0.00018045112781954887,
      "loss": 0.2722235321998596,
      "mean_token_accuracy": 0.8976626545190811,
      "num_tokens": 4453296.0,
      "step": 361
    },
    {
      "entropy": 1.0199629068374634,
      "epoch": 0.19062664560294892,
      "grad_norm": 0.4637596905231476,
      "learning_rate": 0.00018095238095238095,
      "loss": 0.2665867507457733,
      "mean_token_accuracy": 0.9000826925039291,
      "num_tokens": 4465632.0,
      "step": 362
    },
    {
      "entropy": 1.0425358265638351,
      "epoch": 0.1911532385466035,
      "grad_norm": 0.47371116280555725,
      "learning_rate": 0.00018145363408521303,
      "loss": 0.261046439409256,
      "mean_token_accuracy": 0.8978727906942368,
      "num_tokens": 4477968.0,
      "step": 363
    },
    {
      "entropy": 1.0688822865486145,
      "epoch": 0.19167983149025802,
      "grad_norm": 0.588598370552063,
      "learning_rate": 0.0001819548872180451,
      "loss": 0.2722393274307251,
      "mean_token_accuracy": 0.8954607844352722,
      "num_tokens": 4490304.0,
      "step": 364
    },
    {
      "entropy": 1.0311770141124725,
      "epoch": 0.1922064244339126,
      "grad_norm": 0.6127772331237793,
      "learning_rate": 0.0001824561403508772,
      "loss": 0.26557645201683044,
      "mean_token_accuracy": 0.898269534111023,
      "num_tokens": 4502640.0,
      "step": 365
    },
    {
      "entropy": 1.023191675543785,
      "epoch": 0.19273301737756715,
      "grad_norm": 0.4277988076210022,
      "learning_rate": 0.0001829573934837093,
      "loss": 0.25198060274124146,
      "mean_token_accuracy": 0.9041500091552734,
      "num_tokens": 4514976.0,
      "step": 366
    },
    {
      "entropy": 0.992491751909256,
      "epoch": 0.1932596103212217,
      "grad_norm": 0.4713631272315979,
      "learning_rate": 0.00018345864661654135,
      "loss": 0.25175848603248596,
      "mean_token_accuracy": 0.9065804481506348,
      "num_tokens": 4527312.0,
      "step": 367
    },
    {
      "entropy": 0.9969351142644882,
      "epoch": 0.19378620326487625,
      "grad_norm": 0.4729400873184204,
      "learning_rate": 0.00018395989974937345,
      "loss": 0.2526015341281891,
      "mean_token_accuracy": 0.9010139107704163,
      "num_tokens": 4539648.0,
      "step": 368
    },
    {
      "entropy": 1.0536664575338364,
      "epoch": 0.1943127962085308,
      "grad_norm": 0.5235168933868408,
      "learning_rate": 0.0001844611528822055,
      "loss": 0.292032927274704,
      "mean_token_accuracy": 0.8931012600660324,
      "num_tokens": 4551984.0,
      "step": 369
    },
    {
      "entropy": 1.042996197938919,
      "epoch": 0.19483938915218535,
      "grad_norm": 0.48083820939064026,
      "learning_rate": 0.0001849624060150376,
      "loss": 0.26578348875045776,
      "mean_token_accuracy": 0.9036569446325302,
      "num_tokens": 4564320.0,
      "step": 370
    },
    {
      "entropy": 1.0278095752000809,
      "epoch": 0.1953659820958399,
      "grad_norm": 0.39684435725212097,
      "learning_rate": 0.00018546365914786966,
      "loss": 0.23184379935264587,
      "mean_token_accuracy": 0.9090972989797592,
      "num_tokens": 4576656.0,
      "step": 371
    },
    {
      "entropy": 1.0532077848911285,
      "epoch": 0.19589257503949448,
      "grad_norm": 0.40189671516418457,
      "learning_rate": 0.00018596491228070177,
      "loss": 0.22099974751472473,
      "mean_token_accuracy": 0.9153455793857574,
      "num_tokens": 4588992.0,
      "step": 372
    },
    {
      "entropy": 1.090534657239914,
      "epoch": 0.196419167983149,
      "grad_norm": 0.4149918258190155,
      "learning_rate": 0.00018646616541353382,
      "loss": 0.26646703481674194,
      "mean_token_accuracy": 0.8960744440555573,
      "num_tokens": 4601328.0,
      "step": 373
    },
    {
      "entropy": 1.1020061373710632,
      "epoch": 0.19694576092680358,
      "grad_norm": 0.4702380299568176,
      "learning_rate": 0.00018696741854636593,
      "loss": 0.221252903342247,
      "mean_token_accuracy": 0.9114280194044113,
      "num_tokens": 4613664.0,
      "step": 374
    },
    {
      "entropy": 1.0703049004077911,
      "epoch": 0.19747235387045814,
      "grad_norm": 0.3877512216567993,
      "learning_rate": 0.00018746867167919798,
      "loss": 0.2393355667591095,
      "mean_token_accuracy": 0.9077306538820267,
      "num_tokens": 4626000.0,
      "step": 375
    },
    {
      "entropy": 1.1242592930793762,
      "epoch": 0.1979989468141127,
      "grad_norm": 0.4995574355125427,
      "learning_rate": 0.00018796992481203009,
      "loss": 0.26316261291503906,
      "mean_token_accuracy": 0.9027625322341919,
      "num_tokens": 4638336.0,
      "step": 376
    },
    {
      "entropy": 1.1577873229980469,
      "epoch": 0.19852553975776724,
      "grad_norm": 0.5234032273292542,
      "learning_rate": 0.00018847117794486217,
      "loss": 0.27375704050064087,
      "mean_token_accuracy": 0.8979371786117554,
      "num_tokens": 4650672.0,
      "step": 377
    },
    {
      "entropy": 1.1344494223594666,
      "epoch": 0.1990521327014218,
      "grad_norm": 0.5000519752502441,
      "learning_rate": 0.00018897243107769424,
      "loss": 0.26018521189689636,
      "mean_token_accuracy": 0.9034091979265213,
      "num_tokens": 4663008.0,
      "step": 378
    },
    {
      "entropy": 1.2858175039291382,
      "epoch": 0.19957872564507637,
      "grad_norm": 0.4786364436149597,
      "learning_rate": 0.00018947368421052632,
      "loss": 0.2609444260597229,
      "mean_token_accuracy": 0.8963579833507538,
      "num_tokens": 4675344.0,
      "step": 379
    },
    {
      "entropy": 1.205563873052597,
      "epoch": 0.2001053185887309,
      "grad_norm": 0.49225375056266785,
      "learning_rate": 0.0001899749373433584,
      "loss": 0.2702328562736511,
      "mean_token_accuracy": 0.89518603682518,
      "num_tokens": 4687680.0,
      "step": 380
    },
    {
      "entropy": 1.2623814940452576,
      "epoch": 0.20063191153238547,
      "grad_norm": 0.4847429692745209,
      "learning_rate": 0.00019047619047619048,
      "loss": 0.23678088188171387,
      "mean_token_accuracy": 0.9141177982091904,
      "num_tokens": 4700016.0,
      "step": 381
    },
    {
      "entropy": 1.2621397376060486,
      "epoch": 0.20115850447604003,
      "grad_norm": 0.42388200759887695,
      "learning_rate": 0.00019097744360902256,
      "loss": 0.22780729830265045,
      "mean_token_accuracy": 0.9121877998113632,
      "num_tokens": 4712352.0,
      "step": 382
    },
    {
      "entropy": 1.292501300573349,
      "epoch": 0.20168509741969456,
      "grad_norm": 0.4779922366142273,
      "learning_rate": 0.00019147869674185464,
      "loss": 0.2594977915287018,
      "mean_token_accuracy": 0.9015758037567139,
      "num_tokens": 4724688.0,
      "step": 383
    },
    {
      "entropy": 1.2592476904392242,
      "epoch": 0.20221169036334913,
      "grad_norm": 0.5031237006187439,
      "learning_rate": 0.00019197994987468672,
      "loss": 0.2655055820941925,
      "mean_token_accuracy": 0.8991453051567078,
      "num_tokens": 4737024.0,
      "step": 384
    },
    {
      "entropy": 1.2670154571533203,
      "epoch": 0.2027382833070037,
      "grad_norm": 0.43786391615867615,
      "learning_rate": 0.0001924812030075188,
      "loss": 0.2638092637062073,
      "mean_token_accuracy": 0.9021878391504288,
      "num_tokens": 4749360.0,
      "step": 385
    },
    {
      "entropy": 1.2625199258327484,
      "epoch": 0.20326487625065823,
      "grad_norm": 0.6012351512908936,
      "learning_rate": 0.00019298245614035088,
      "loss": 0.25398504734039307,
      "mean_token_accuracy": 0.9037431925535202,
      "num_tokens": 4761696.0,
      "step": 386
    },
    {
      "entropy": 1.2534445226192474,
      "epoch": 0.2037914691943128,
      "grad_norm": 0.48004376888275146,
      "learning_rate": 0.00019348370927318296,
      "loss": 0.24518710374832153,
      "mean_token_accuracy": 0.9088338762521744,
      "num_tokens": 4774032.0,
      "step": 387
    },
    {
      "entropy": 1.2239902317523956,
      "epoch": 0.20431806213796735,
      "grad_norm": 0.40316712856292725,
      "learning_rate": 0.00019398496240601503,
      "loss": 0.22658605873584747,
      "mean_token_accuracy": 0.9146187752485275,
      "num_tokens": 4786368.0,
      "step": 388
    },
    {
      "entropy": 1.236211746931076,
      "epoch": 0.20484465508162192,
      "grad_norm": 0.4805733859539032,
      "learning_rate": 0.0001944862155388471,
      "loss": 0.2645362615585327,
      "mean_token_accuracy": 0.899237260222435,
      "num_tokens": 4798704.0,
      "step": 389
    },
    {
      "entropy": 1.2168347239494324,
      "epoch": 0.20537124802527645,
      "grad_norm": 0.4017006754875183,
      "learning_rate": 0.00019498746867167922,
      "loss": 0.23945499956607819,
      "mean_token_accuracy": 0.9106440991163254,
      "num_tokens": 4811040.0,
      "step": 390
    },
    {
      "entropy": 1.2192648351192474,
      "epoch": 0.20589784096893102,
      "grad_norm": 0.4144308269023895,
      "learning_rate": 0.00019548872180451127,
      "loss": 0.24552568793296814,
      "mean_token_accuracy": 0.9093390554189682,
      "num_tokens": 4823376.0,
      "step": 391
    },
    {
      "entropy": 1.2436101734638214,
      "epoch": 0.20642443391258558,
      "grad_norm": 0.4212309420108795,
      "learning_rate": 0.00019598997493734338,
      "loss": 0.23153892159461975,
      "mean_token_accuracy": 0.9111403673887253,
      "num_tokens": 4835712.0,
      "step": 392
    },
    {
      "entropy": 1.255335807800293,
      "epoch": 0.20695102685624012,
      "grad_norm": 0.40789833664894104,
      "learning_rate": 0.00019649122807017543,
      "loss": 0.22183379530906677,
      "mean_token_accuracy": 0.9136195629835129,
      "num_tokens": 4848048.0,
      "step": 393
    },
    {
      "entropy": 1.2751134932041168,
      "epoch": 0.20747761979989468,
      "grad_norm": 0.4400269091129303,
      "learning_rate": 0.00019699248120300754,
      "loss": 0.24014121294021606,
      "mean_token_accuracy": 0.9085260182619095,
      "num_tokens": 4860384.0,
      "step": 394
    },
    {
      "entropy": 1.360796719789505,
      "epoch": 0.20800421274354924,
      "grad_norm": 0.4804284870624542,
      "learning_rate": 0.0001974937343358396,
      "loss": 0.25582095980644226,
      "mean_token_accuracy": 0.9006912261247635,
      "num_tokens": 4872720.0,
      "step": 395
    },
    {
      "entropy": 1.3161850571632385,
      "epoch": 0.20853080568720378,
      "grad_norm": 0.43244144320487976,
      "learning_rate": 0.0001979949874686717,
      "loss": 0.2349988967180252,
      "mean_token_accuracy": 0.9116800725460052,
      "num_tokens": 4885056.0,
      "step": 396
    },
    {
      "entropy": 1.3374731242656708,
      "epoch": 0.20905739863085834,
      "grad_norm": 0.4748205840587616,
      "learning_rate": 0.00019849624060150375,
      "loss": 0.22185081243515015,
      "mean_token_accuracy": 0.9125295132398605,
      "num_tokens": 4897392.0,
      "step": 397
    },
    {
      "entropy": 1.3775228261947632,
      "epoch": 0.2095839915745129,
      "grad_norm": 0.4581362307071686,
      "learning_rate": 0.00019899749373433585,
      "loss": 0.2567383646965027,
      "mean_token_accuracy": 0.9047663658857346,
      "num_tokens": 4909728.0,
      "step": 398
    },
    {
      "entropy": 1.4304566085338593,
      "epoch": 0.21011058451816747,
      "grad_norm": 0.5330026149749756,
      "learning_rate": 0.00019949874686716793,
      "loss": 0.24410131573677063,
      "mean_token_accuracy": 0.9004727154970169,
      "num_tokens": 4922064.0,
      "step": 399
    },
    {
      "entropy": 1.4176018834114075,
      "epoch": 0.210637177461822,
      "grad_norm": 0.5187347531318665,
      "learning_rate": 0.0002,
      "loss": 0.2622545063495636,
      "mean_token_accuracy": 0.8981441259384155,
      "num_tokens": 4934400.0,
      "step": 400
    },
    {
      "entropy": 1.3980787992477417,
      "epoch": 0.21116377040547657,
      "grad_norm": 0.4053288400173187,
      "learning_rate": 0.0001999999973286123,
      "loss": 0.23980936408042908,
      "mean_token_accuracy": 0.912782147526741,
      "num_tokens": 4946736.0,
      "step": 401
    },
    {
      "entropy": 1.4483689367771149,
      "epoch": 0.21169036334913113,
      "grad_norm": 0.45620113611221313,
      "learning_rate": 0.00019999998931444929,
      "loss": 0.25808754563331604,
      "mean_token_accuracy": 0.9030732810497284,
      "num_tokens": 4959072.0,
      "step": 402
    },
    {
      "entropy": 1.4470272064208984,
      "epoch": 0.21221695629278567,
      "grad_norm": 0.45807719230651855,
      "learning_rate": 0.0001999999759575115,
      "loss": 0.258093923330307,
      "mean_token_accuracy": 0.9026659429073334,
      "num_tokens": 4971408.0,
      "step": 403
    },
    {
      "entropy": 1.425654649734497,
      "epoch": 0.21274354923644023,
      "grad_norm": 0.4397832751274109,
      "learning_rate": 0.0001999999572577997,
      "loss": 0.24665014445781708,
      "mean_token_accuracy": 0.9062764793634415,
      "num_tokens": 4983744.0,
      "step": 404
    },
    {
      "entropy": 1.5687190890312195,
      "epoch": 0.2132701421800948,
      "grad_norm": 0.5367022156715393,
      "learning_rate": 0.00019999993321531494,
      "loss": 0.24174389243125916,
      "mean_token_accuracy": 0.9081769734621048,
      "num_tokens": 4996080.0,
      "step": 405
    },
    {
      "entropy": 1.405859649181366,
      "epoch": 0.21379673512374933,
      "grad_norm": 0.4007917046546936,
      "learning_rate": 0.00019999990383005872,
      "loss": 0.25402745604515076,
      "mean_token_accuracy": 0.9025277644395828,
      "num_tokens": 5008416.0,
      "step": 406
    },
    {
      "entropy": 1.4313717782497406,
      "epoch": 0.2143233280674039,
      "grad_norm": 0.4160802662372589,
      "learning_rate": 0.00019999986910203282,
      "loss": 0.24924606084823608,
      "mean_token_accuracy": 0.9043161869049072,
      "num_tokens": 5020752.0,
      "step": 407
    },
    {
      "entropy": 1.4701331555843353,
      "epoch": 0.21484992101105846,
      "grad_norm": 0.4316672086715698,
      "learning_rate": 0.00019999982903123921,
      "loss": 0.2638508379459381,
      "mean_token_accuracy": 0.8981801271438599,
      "num_tokens": 5033088.0,
      "step": 408
    },
    {
      "entropy": 1.4392390549182892,
      "epoch": 0.215376513954713,
      "grad_norm": 0.4544552266597748,
      "learning_rate": 0.00019999978361768031,
      "loss": 0.23990359902381897,
      "mean_token_accuracy": 0.9076667875051498,
      "num_tokens": 5045424.0,
      "step": 409
    },
    {
      "entropy": 1.4876410961151123,
      "epoch": 0.21590310689836756,
      "grad_norm": 0.4295142889022827,
      "learning_rate": 0.00019999973286135886,
      "loss": 0.21840089559555054,
      "mean_token_accuracy": 0.9214875847101212,
      "num_tokens": 5057760.0,
      "step": 410
    },
    {
      "entropy": 1.50434610247612,
      "epoch": 0.21642969984202212,
      "grad_norm": 0.4359295070171356,
      "learning_rate": 0.00019999967676227775,
      "loss": 0.2510741353034973,
      "mean_token_accuracy": 0.9052848815917969,
      "num_tokens": 5070096.0,
      "step": 411
    },
    {
      "entropy": 1.4829690754413605,
      "epoch": 0.21695629278567669,
      "grad_norm": 0.42792809009552,
      "learning_rate": 0.00019999961532044045,
      "loss": 0.2551060616970062,
      "mean_token_accuracy": 0.9011600762605667,
      "num_tokens": 5082432.0,
      "step": 412
    },
    {
      "entropy": 1.4852145910263062,
      "epoch": 0.21748288572933122,
      "grad_norm": 0.4182438850402832,
      "learning_rate": 0.00019999954853585052,
      "loss": 0.24789488315582275,
      "mean_token_accuracy": 0.9069723039865494,
      "num_tokens": 5094768.0,
      "step": 413
    },
    {
      "entropy": 1.4210355281829834,
      "epoch": 0.21800947867298578,
      "grad_norm": 0.42366713285446167,
      "learning_rate": 0.00019999947640851195,
      "loss": 0.24643118679523468,
      "mean_token_accuracy": 0.9047895669937134,
      "num_tokens": 5107104.0,
      "step": 414
    },
    {
      "entropy": 1.525164395570755,
      "epoch": 0.21853607161664035,
      "grad_norm": 0.4372459352016449,
      "learning_rate": 0.00019999939893842905,
      "loss": 0.2424328625202179,
      "mean_token_accuracy": 0.9087924510240555,
      "num_tokens": 5119440.0,
      "step": 415
    },
    {
      "entropy": 1.4822115898132324,
      "epoch": 0.21906266456029488,
      "grad_norm": 0.4659183919429779,
      "learning_rate": 0.00019999931612560637,
      "loss": 0.22431665658950806,
      "mean_token_accuracy": 0.9165952354669571,
      "num_tokens": 5131776.0,
      "step": 416
    },
    {
      "entropy": 1.5467821657657623,
      "epoch": 0.21958925750394945,
      "grad_norm": 0.44125568866729736,
      "learning_rate": 0.00019999922797004884,
      "loss": 0.23688746988773346,
      "mean_token_accuracy": 0.9075507670640945,
      "num_tokens": 5144112.0,
      "step": 417
    },
    {
      "entropy": 1.4780353605747223,
      "epoch": 0.220115850447604,
      "grad_norm": 0.43714895844459534,
      "learning_rate": 0.00019999913447176174,
      "loss": 0.24803104996681213,
      "mean_token_accuracy": 0.9086460769176483,
      "num_tokens": 5156448.0,
      "step": 418
    },
    {
      "entropy": 1.442844033241272,
      "epoch": 0.22064244339125855,
      "grad_norm": 0.5062248110771179,
      "learning_rate": 0.00019999903563075051,
      "loss": 0.2509956657886505,
      "mean_token_accuracy": 0.9084739983081818,
      "num_tokens": 5168784.0,
      "step": 419
    },
    {
      "entropy": 1.472544252872467,
      "epoch": 0.2211690363349131,
      "grad_norm": 0.44560346007347107,
      "learning_rate": 0.00019999893144702116,
      "loss": 0.2571001648902893,
      "mean_token_accuracy": 0.9050741344690323,
      "num_tokens": 5181120.0,
      "step": 420
    },
    {
      "entropy": 1.4741449654102325,
      "epoch": 0.22169562927856767,
      "grad_norm": 0.468660444021225,
      "learning_rate": 0.00019999882192057975,
      "loss": 0.2654499113559723,
      "mean_token_accuracy": 0.9030983746051788,
      "num_tokens": 5193456.0,
      "step": 421
    },
    {
      "entropy": 1.4696997106075287,
      "epoch": 0.2222222222222222,
      "grad_norm": 0.42363062500953674,
      "learning_rate": 0.0001999987070514329,
      "loss": 0.2351139783859253,
      "mean_token_accuracy": 0.9069055020809174,
      "num_tokens": 5205792.0,
      "step": 422
    },
    {
      "entropy": 1.452900618314743,
      "epoch": 0.22274881516587677,
      "grad_norm": 0.45147648453712463,
      "learning_rate": 0.00019999858683958726,
      "loss": 0.236272394657135,
      "mean_token_accuracy": 0.9084236770868301,
      "num_tokens": 5218128.0,
      "step": 423
    },
    {
      "entropy": 1.5096050798892975,
      "epoch": 0.22327540810953134,
      "grad_norm": 0.4349573850631714,
      "learning_rate": 0.00019999846128505015,
      "loss": 0.21908816695213318,
      "mean_token_accuracy": 0.9153110533952713,
      "num_tokens": 5230464.0,
      "step": 424
    },
    {
      "entropy": 1.5504009425640106,
      "epoch": 0.2238020010531859,
      "grad_norm": 0.40080899000167847,
      "learning_rate": 0.00019999833038782897,
      "loss": 0.21589110791683197,
      "mean_token_accuracy": 0.9171598702669144,
      "num_tokens": 5242800.0,
      "step": 425
    },
    {
      "entropy": 1.4936735033988953,
      "epoch": 0.22432859399684044,
      "grad_norm": 0.4071311056613922,
      "learning_rate": 0.0001999981941479314,
      "loss": 0.2125779092311859,
      "mean_token_accuracy": 0.9237974286079407,
      "num_tokens": 5255136.0,
      "step": 426
    },
    {
      "entropy": 1.5081810355186462,
      "epoch": 0.224855186940495,
      "grad_norm": 0.4242808222770691,
      "learning_rate": 0.0001999980525653656,
      "loss": 0.2447715699672699,
      "mean_token_accuracy": 0.9042525738477707,
      "num_tokens": 5267472.0,
      "step": 427
    },
    {
      "entropy": 1.5077386796474457,
      "epoch": 0.22538177988414956,
      "grad_norm": 0.5070521235466003,
      "learning_rate": 0.00019999790564014,
      "loss": 0.26209956407546997,
      "mean_token_accuracy": 0.9026446640491486,
      "num_tokens": 5279808.0,
      "step": 428
    },
    {
      "entropy": 1.4962070286273956,
      "epoch": 0.2259083728278041,
      "grad_norm": 0.6017259955406189,
      "learning_rate": 0.00019999775337226327,
      "loss": 0.3165000081062317,
      "mean_token_accuracy": 0.8868077397346497,
      "num_tokens": 5292144.0,
      "step": 429
    },
    {
      "entropy": 1.4911207258701324,
      "epoch": 0.22643496577145866,
      "grad_norm": 0.47006961703300476,
      "learning_rate": 0.00019999759576174448,
      "loss": 0.22799812257289886,
      "mean_token_accuracy": 0.9069435149431229,
      "num_tokens": 5304480.0,
      "step": 430
    },
    {
      "entropy": 1.5028018951416016,
      "epoch": 0.22696155871511323,
      "grad_norm": 0.4170990288257599,
      "learning_rate": 0.00019999743280859296,
      "loss": 0.21681472659111023,
      "mean_token_accuracy": 0.9184076339006424,
      "num_tokens": 5316816.0,
      "step": 431
    },
    {
      "entropy": 1.6698609590530396,
      "epoch": 0.22748815165876776,
      "grad_norm": 0.6547060012817383,
      "learning_rate": 0.0001999972645128184,
      "loss": 0.24768398702144623,
      "mean_token_accuracy": 0.9032211750745773,
      "num_tokens": 5329152.0,
      "step": 432
    },
    {
      "entropy": 1.549822211265564,
      "epoch": 0.22801474460242233,
      "grad_norm": 0.433370441198349,
      "learning_rate": 0.00019999709087443083,
      "loss": 0.24651046097278595,
      "mean_token_accuracy": 0.9072823375463486,
      "num_tokens": 5341488.0,
      "step": 433
    },
    {
      "entropy": 1.5724502205848694,
      "epoch": 0.2285413375460769,
      "grad_norm": 0.5081071853637695,
      "learning_rate": 0.00019999691189344045,
      "loss": 0.2539141774177551,
      "mean_token_accuracy": 0.901251494884491,
      "num_tokens": 5353824.0,
      "step": 434
    },
    {
      "entropy": 1.580730140209198,
      "epoch": 0.22906793048973143,
      "grad_norm": 0.4615112841129303,
      "learning_rate": 0.000199996727569858,
      "loss": 0.23179525136947632,
      "mean_token_accuracy": 0.9096536040306091,
      "num_tokens": 5366160.0,
      "step": 435
    },
    {
      "entropy": 1.6549272239208221,
      "epoch": 0.229594523433386,
      "grad_norm": 0.6388656497001648,
      "learning_rate": 0.00019999653790369438,
      "loss": 0.29721513390541077,
      "mean_token_accuracy": 0.8857980072498322,
      "num_tokens": 5378496.0,
      "step": 436
    },
    {
      "entropy": 1.6025878190994263,
      "epoch": 0.23012111637704055,
      "grad_norm": 0.49696317315101624,
      "learning_rate": 0.0001999963428949608,
      "loss": 0.2271474003791809,
      "mean_token_accuracy": 0.9138942360877991,
      "num_tokens": 5390832.0,
      "step": 437
    },
    {
      "entropy": 1.5836864113807678,
      "epoch": 0.23064770932069512,
      "grad_norm": 0.4876512289047241,
      "learning_rate": 0.00019999614254366895,
      "loss": 0.24330684542655945,
      "mean_token_accuracy": 0.9093976020812988,
      "num_tokens": 5403168.0,
      "step": 438
    },
    {
      "entropy": 1.5200332999229431,
      "epoch": 0.23117430226434965,
      "grad_norm": 0.6539078950881958,
      "learning_rate": 0.00019999593684983058,
      "loss": 0.2600172460079193,
      "mean_token_accuracy": 0.8999185413122177,
      "num_tokens": 5415504.0,
      "step": 439
    },
    {
      "entropy": 1.4733262360095978,
      "epoch": 0.23170089520800422,
      "grad_norm": 0.42358288168907166,
      "learning_rate": 0.000199995725813458,
      "loss": 0.22544077038764954,
      "mean_token_accuracy": 0.9120991379022598,
      "num_tokens": 5427840.0,
      "step": 440
    },
    {
      "entropy": 1.414179265499115,
      "epoch": 0.23222748815165878,
      "grad_norm": 0.49307864904403687,
      "learning_rate": 0.00019999550943456375,
      "loss": 0.24960950016975403,
      "mean_token_accuracy": 0.9004078656435013,
      "num_tokens": 5440176.0,
      "step": 441
    },
    {
      "entropy": 1.429827481508255,
      "epoch": 0.23275408109531331,
      "grad_norm": 0.505660891532898,
      "learning_rate": 0.00019999528771316057,
      "loss": 0.25559940934181213,
      "mean_token_accuracy": 0.9029601365327835,
      "num_tokens": 5452512.0,
      "step": 442
    },
    {
      "entropy": 1.4452406466007233,
      "epoch": 0.23328067403896788,
      "grad_norm": 0.5408805012702942,
      "learning_rate": 0.00019999506064926175,
      "loss": 0.2864495515823364,
      "mean_token_accuracy": 0.8919995576143265,
      "num_tokens": 5464848.0,
      "step": 443
    },
    {
      "entropy": 1.441390186548233,
      "epoch": 0.23380726698262244,
      "grad_norm": 0.49373185634613037,
      "learning_rate": 0.0001999948282428807,
      "loss": 0.2585368752479553,
      "mean_token_accuracy": 0.8959326446056366,
      "num_tokens": 5477184.0,
      "step": 444
    },
    {
      "entropy": 1.386150062084198,
      "epoch": 0.23433385992627698,
      "grad_norm": 0.4148918390274048,
      "learning_rate": 0.0001999945904940312,
      "loss": 0.2401098757982254,
      "mean_token_accuracy": 0.9103811830282211,
      "num_tokens": 5489520.0,
      "step": 445
    },
    {
      "entropy": 1.4309166073799133,
      "epoch": 0.23486045286993154,
      "grad_norm": 0.46061956882476807,
      "learning_rate": 0.00019999434740272743,
      "loss": 0.2515195310115814,
      "mean_token_accuracy": 0.9047931730747223,
      "num_tokens": 5501856.0,
      "step": 446
    },
    {
      "entropy": 1.4778549373149872,
      "epoch": 0.2353870458135861,
      "grad_norm": 0.543989896774292,
      "learning_rate": 0.00019999409896898372,
      "loss": 0.23118984699249268,
      "mean_token_accuracy": 0.9122479408979416,
      "num_tokens": 5514192.0,
      "step": 447
    },
    {
      "entropy": 1.3631019592285156,
      "epoch": 0.23591363875724064,
      "grad_norm": 0.39872297644615173,
      "learning_rate": 0.00019999384519281494,
      "loss": 0.23500953614711761,
      "mean_token_accuracy": 0.9080264270305634,
      "num_tokens": 5526528.0,
      "step": 448
    },
    {
      "entropy": 1.4472846686840057,
      "epoch": 0.2364402317008952,
      "grad_norm": 0.4705425500869751,
      "learning_rate": 0.00019999358607423608,
      "loss": 0.2556760609149933,
      "mean_token_accuracy": 0.903324156999588,
      "num_tokens": 5538864.0,
      "step": 449
    },
    {
      "entropy": 1.4216105043888092,
      "epoch": 0.23696682464454977,
      "grad_norm": 0.42123982310295105,
      "learning_rate": 0.00019999332161326253,
      "loss": 0.24162599444389343,
      "mean_token_accuracy": 0.906711995601654,
      "num_tokens": 5551200.0,
      "step": 450
    },
    {
      "entropy": 1.3214130699634552,
      "epoch": 0.23749341758820433,
      "grad_norm": 0.4052497148513794,
      "learning_rate": 0.00019999305180991002,
      "loss": 0.23021447658538818,
      "mean_token_accuracy": 0.9166329354047775,
      "num_tokens": 5563536.0,
      "step": 451
    },
    {
      "entropy": 1.2956161499023438,
      "epoch": 0.23802001053185887,
      "grad_norm": 0.4208316206932068,
      "learning_rate": 0.0001999927766641945,
      "loss": 0.25906720757484436,
      "mean_token_accuracy": 0.8939131498336792,
      "num_tokens": 5575872.0,
      "step": 452
    },
    {
      "entropy": 1.2186987400054932,
      "epoch": 0.23854660347551343,
      "grad_norm": 0.423001766204834,
      "learning_rate": 0.00019999249617613237,
      "loss": 0.2634647488594055,
      "mean_token_accuracy": 0.8950765579938889,
      "num_tokens": 5588208.0,
      "step": 453
    },
    {
      "entropy": 1.1696477234363556,
      "epoch": 0.239073196419168,
      "grad_norm": 0.7100384831428528,
      "learning_rate": 0.00019999221034574028,
      "loss": 0.2620895504951477,
      "mean_token_accuracy": 0.901919424533844,
      "num_tokens": 5600544.0,
      "step": 454
    },
    {
      "entropy": 1.1881844103336334,
      "epoch": 0.23959978936282253,
      "grad_norm": 0.40986204147338867,
      "learning_rate": 0.00019999191917303513,
      "loss": 0.23100420832633972,
      "mean_token_accuracy": 0.9076221436262131,
      "num_tokens": 5612880.0,
      "step": 455
    },
    {
      "entropy": 1.0672442317008972,
      "epoch": 0.2401263823064771,
      "grad_norm": 0.9633269309997559,
      "learning_rate": 0.0001999916226580343,
      "loss": 0.2286304533481598,
      "mean_token_accuracy": 0.9094352871179581,
      "num_tokens": 5625216.0,
      "step": 456
    },
    {
      "entropy": 1.0725638568401337,
      "epoch": 0.24065297525013166,
      "grad_norm": 0.5792625546455383,
      "learning_rate": 0.0001999913208007553,
      "loss": 0.21992720663547516,
      "mean_token_accuracy": 0.9132383912801743,
      "num_tokens": 5637552.0,
      "step": 457
    },
    {
      "entropy": 1.0756218880414963,
      "epoch": 0.2411795681937862,
      "grad_norm": 0.4889819622039795,
      "learning_rate": 0.0001999910136012161,
      "loss": 0.22047317028045654,
      "mean_token_accuracy": 0.9096969217061996,
      "num_tokens": 5649888.0,
      "step": 458
    },
    {
      "entropy": 1.025196298956871,
      "epoch": 0.24170616113744076,
      "grad_norm": 0.43335074186325073,
      "learning_rate": 0.00019999070105943494,
      "loss": 0.23299942910671234,
      "mean_token_accuracy": 0.9110541045665741,
      "num_tokens": 5662224.0,
      "step": 459
    },
    {
      "entropy": 1.0740464180707932,
      "epoch": 0.24223275408109532,
      "grad_norm": 0.4134856164455414,
      "learning_rate": 0.00019999038317543036,
      "loss": 0.2113601118326187,
      "mean_token_accuracy": 0.9125392287969589,
      "num_tokens": 5674560.0,
      "step": 460
    },
    {
      "entropy": 1.0400484502315521,
      "epoch": 0.24275934702474986,
      "grad_norm": 0.3933151364326477,
      "learning_rate": 0.00019999005994922125,
      "loss": 0.23023052513599396,
      "mean_token_accuracy": 0.9128931611776352,
      "num_tokens": 5686896.0,
      "step": 461
    },
    {
      "entropy": 1.033817708492279,
      "epoch": 0.24328593996840442,
      "grad_norm": 0.4203343093395233,
      "learning_rate": 0.00019998973138082675,
      "loss": 0.23599188029766083,
      "mean_token_accuracy": 0.9099869877099991,
      "num_tokens": 5699232.0,
      "step": 462
    },
    {
      "entropy": 1.0623035430908203,
      "epoch": 0.24381253291205898,
      "grad_norm": 0.41818881034851074,
      "learning_rate": 0.00019998939747026644,
      "loss": 0.23520024120807648,
      "mean_token_accuracy": 0.9101691842079163,
      "num_tokens": 5711568.0,
      "step": 463
    },
    {
      "entropy": 1.0741385221481323,
      "epoch": 0.24433912585571355,
      "grad_norm": 0.4350592792034149,
      "learning_rate": 0.00019998905821756006,
      "loss": 0.26694831252098083,
      "mean_token_accuracy": 0.8980891853570938,
      "num_tokens": 5723904.0,
      "step": 464
    },
    {
      "entropy": 1.1385951340198517,
      "epoch": 0.24486571879936808,
      "grad_norm": 0.4376484155654907,
      "learning_rate": 0.00019998871362272784,
      "loss": 0.25515180826187134,
      "mean_token_accuracy": 0.9086296558380127,
      "num_tokens": 5736240.0,
      "step": 465
    },
    {
      "entropy": 1.0884679853916168,
      "epoch": 0.24539231174302265,
      "grad_norm": 0.42011401057243347,
      "learning_rate": 0.00019998836368579013,
      "loss": 0.2314184159040451,
      "mean_token_accuracy": 0.9099607318639755,
      "num_tokens": 5748576.0,
      "step": 466
    },
    {
      "entropy": 1.09638212621212,
      "epoch": 0.2459189046866772,
      "grad_norm": 0.39627403020858765,
      "learning_rate": 0.0001999880084067678,
      "loss": 0.21417196094989777,
      "mean_token_accuracy": 0.9153980165719986,
      "num_tokens": 5760912.0,
      "step": 467
    },
    {
      "entropy": 1.091229110956192,
      "epoch": 0.24644549763033174,
      "grad_norm": 0.4235895872116089,
      "learning_rate": 0.00019998764778568192,
      "loss": 0.2682695686817169,
      "mean_token_accuracy": 0.9011466205120087,
      "num_tokens": 5773248.0,
      "step": 468
    },
    {
      "entropy": 1.1204252541065216,
      "epoch": 0.2469720905739863,
      "grad_norm": 0.43689507246017456,
      "learning_rate": 0.00019998728182255381,
      "loss": 0.24578194320201874,
      "mean_token_accuracy": 0.9062277525663376,
      "num_tokens": 5785584.0,
      "step": 469
    },
    {
      "entropy": 1.1320042610168457,
      "epoch": 0.24749868351764087,
      "grad_norm": 0.4219723641872406,
      "learning_rate": 0.0001999869105174053,
      "loss": 0.22462952136993408,
      "mean_token_accuracy": 0.9171530902385712,
      "num_tokens": 5797920.0,
      "step": 470
    },
    {
      "entropy": 1.1340222358703613,
      "epoch": 0.2480252764612954,
      "grad_norm": 0.3906649351119995,
      "learning_rate": 0.00019998653387025842,
      "loss": 0.2314276546239853,
      "mean_token_accuracy": 0.9133926779031754,
      "num_tokens": 5810256.0,
      "step": 471
    },
    {
      "entropy": 1.148337960243225,
      "epoch": 0.24855186940494997,
      "grad_norm": 0.435470312833786,
      "learning_rate": 0.00019998615188113547,
      "loss": 0.27237382531166077,
      "mean_token_accuracy": 0.8951744884252548,
      "num_tokens": 5822592.0,
      "step": 472
    },
    {
      "entropy": 1.084697663784027,
      "epoch": 0.24907846234860453,
      "grad_norm": 0.3843922019004822,
      "learning_rate": 0.00019998576455005918,
      "loss": 0.21877625584602356,
      "mean_token_accuracy": 0.9148665964603424,
      "num_tokens": 5834928.0,
      "step": 473
    },
    {
      "entropy": 1.1236861050128937,
      "epoch": 0.24960505529225907,
      "grad_norm": 0.3693842887878418,
      "learning_rate": 0.0001999853718770525,
      "loss": 0.24262937903404236,
      "mean_token_accuracy": 0.9089461863040924,
      "num_tokens": 5847264.0,
      "step": 474
    },
    {
      "entropy": 1.0875613689422607,
      "epoch": 0.25013164823591366,
      "grad_norm": 0.36348292231559753,
      "learning_rate": 0.0001999849738621388,
      "loss": 0.22801700234413147,
      "mean_token_accuracy": 0.9149875193834305,
      "num_tokens": 5859600.0,
      "step": 475
    },
    {
      "entropy": 1.1611447632312775,
      "epoch": 0.2506582411795682,
      "grad_norm": 0.3937731683254242,
      "learning_rate": 0.00019998457050534165,
      "loss": 0.2227628082036972,
      "mean_token_accuracy": 0.9130124449729919,
      "num_tokens": 5871936.0,
      "step": 476
    },
    {
      "entropy": 1.18001389503479,
      "epoch": 0.25118483412322273,
      "grad_norm": 0.38412168622016907,
      "learning_rate": 0.00019998416180668505,
      "loss": 0.25140437483787537,
      "mean_token_accuracy": 0.9021612256765366,
      "num_tokens": 5884272.0,
      "step": 477
    },
    {
      "entropy": 1.1783344447612762,
      "epoch": 0.2517114270668773,
      "grad_norm": 0.4068759083747864,
      "learning_rate": 0.00019998374776619316,
      "loss": 0.2507692575454712,
      "mean_token_accuracy": 0.9036175757646561,
      "num_tokens": 5896608.0,
      "step": 478
    },
    {
      "entropy": 1.2171072959899902,
      "epoch": 0.25223802001053186,
      "grad_norm": 0.38274940848350525,
      "learning_rate": 0.00019998332838389068,
      "loss": 0.23562754690647125,
      "mean_token_accuracy": 0.9149611443281174,
      "num_tokens": 5908944.0,
      "step": 479
    },
    {
      "entropy": 1.2112184464931488,
      "epoch": 0.2527646129541864,
      "grad_norm": 0.4171972870826721,
      "learning_rate": 0.00019998290365980247,
      "loss": 0.26383018493652344,
      "mean_token_accuracy": 0.8946961760520935,
      "num_tokens": 5921280.0,
      "step": 480
    },
    {
      "entropy": 1.1482765674591064,
      "epoch": 0.253291205897841,
      "grad_norm": 0.3705795407295227,
      "learning_rate": 0.0001999824735939537,
      "loss": 0.23539453744888306,
      "mean_token_accuracy": 0.9124044477939606,
      "num_tokens": 5933616.0,
      "step": 481
    },
    {
      "entropy": 1.1225126087665558,
      "epoch": 0.2538177988414955,
      "grad_norm": 0.38258257508277893,
      "learning_rate": 0.00019998203818636997,
      "loss": 0.24310466647148132,
      "mean_token_accuracy": 0.9069766700267792,
      "num_tokens": 5945952.0,
      "step": 482
    },
    {
      "entropy": 1.1831683218479156,
      "epoch": 0.25434439178515006,
      "grad_norm": 0.4023701548576355,
      "learning_rate": 0.00019998159743707706,
      "loss": 0.22924776375293732,
      "mean_token_accuracy": 0.910895898938179,
      "num_tokens": 5958288.0,
      "step": 483
    },
    {
      "entropy": 1.2125089168548584,
      "epoch": 0.25487098472880465,
      "grad_norm": 0.391679972410202,
      "learning_rate": 0.0001999811513461012,
      "loss": 0.23159608244895935,
      "mean_token_accuracy": 0.91130131483078,
      "num_tokens": 5970624.0,
      "step": 484
    },
    {
      "entropy": 1.1657118201255798,
      "epoch": 0.2553975776724592,
      "grad_norm": 0.36009886860847473,
      "learning_rate": 0.00019998069991346878,
      "loss": 0.22824552655220032,
      "mean_token_accuracy": 0.9051208347082138,
      "num_tokens": 5982960.0,
      "step": 485
    },
    {
      "entropy": 1.1666974127292633,
      "epoch": 0.2559241706161137,
      "grad_norm": 0.4200109839439392,
      "learning_rate": 0.00019998024313920668,
      "loss": 0.22612211108207703,
      "mean_token_accuracy": 0.9098036587238312,
      "num_tokens": 5995296.0,
      "step": 486
    },
    {
      "entropy": 1.1931614875793457,
      "epoch": 0.2564507635597683,
      "grad_norm": 0.3996848165988922,
      "learning_rate": 0.00019997978102334195,
      "loss": 0.23149043321609497,
      "mean_token_accuracy": 0.915941908955574,
      "num_tokens": 6007632.0,
      "step": 487
    },
    {
      "entropy": 1.2029556334018707,
      "epoch": 0.25697735650342285,
      "grad_norm": 0.381720632314682,
      "learning_rate": 0.00019997931356590212,
      "loss": 0.20601031184196472,
      "mean_token_accuracy": 0.9207601994276047,
      "num_tokens": 6019968.0,
      "step": 488
    },
    {
      "entropy": 1.286322921514511,
      "epoch": 0.2575039494470774,
      "grad_norm": 0.3947739601135254,
      "learning_rate": 0.00019997884076691484,
      "loss": 0.23165327310562134,
      "mean_token_accuracy": 0.9127722233533859,
      "num_tokens": 6032304.0,
      "step": 489
    },
    {
      "entropy": 1.1862512528896332,
      "epoch": 0.258030542390732,
      "grad_norm": 0.4733010232448578,
      "learning_rate": 0.00019997836262640825,
      "loss": 0.23455524444580078,
      "mean_token_accuracy": 0.9086931645870209,
      "num_tokens": 6044640.0,
      "step": 490
    },
    {
      "entropy": 1.2432229220867157,
      "epoch": 0.2585571353343865,
      "grad_norm": 0.5269920229911804,
      "learning_rate": 0.00019997787914441067,
      "loss": 0.25063106417655945,
      "mean_token_accuracy": 0.903441533446312,
      "num_tokens": 6056976.0,
      "step": 491
    },
    {
      "entropy": 1.1587136387825012,
      "epoch": 0.25908372827804105,
      "grad_norm": 0.3660036325454712,
      "learning_rate": 0.0001999773903209509,
      "loss": 0.23804998397827148,
      "mean_token_accuracy": 0.9118834733963013,
      "num_tokens": 6069312.0,
      "step": 492
    },
    {
      "entropy": 1.2006062865257263,
      "epoch": 0.25961032122169564,
      "grad_norm": 0.4517950117588043,
      "learning_rate": 0.00019997689615605785,
      "loss": 0.24688124656677246,
      "mean_token_accuracy": 0.9041164964437485,
      "num_tokens": 6081648.0,
      "step": 493
    },
    {
      "entropy": 1.2588125467300415,
      "epoch": 0.2601369141653502,
      "grad_norm": 0.3569098114967346,
      "learning_rate": 0.0001999763966497609,
      "loss": 0.22678923606872559,
      "mean_token_accuracy": 0.9117091447114944,
      "num_tokens": 6093984.0,
      "step": 494
    },
    {
      "entropy": 1.3271512985229492,
      "epoch": 0.26066350710900477,
      "grad_norm": 0.3427906036376953,
      "learning_rate": 0.00019997589180208972,
      "loss": 0.2427246868610382,
      "mean_token_accuracy": 0.91063092648983,
      "num_tokens": 6106320.0,
      "step": 495
    },
    {
      "entropy": 1.320880800485611,
      "epoch": 0.2611901000526593,
      "grad_norm": 0.41021794080734253,
      "learning_rate": 0.00019997538161307425,
      "loss": 0.24322479963302612,
      "mean_token_accuracy": 0.9047645032405853,
      "num_tokens": 6118656.0,
      "step": 496
    },
    {
      "entropy": 1.2419660687446594,
      "epoch": 0.26171669299631384,
      "grad_norm": 0.4706597328186035,
      "learning_rate": 0.00019997486608274478,
      "loss": 0.24686744809150696,
      "mean_token_accuracy": 0.9022967517375946,
      "num_tokens": 6130992.0,
      "step": 497
    },
    {
      "entropy": 1.2451711893081665,
      "epoch": 0.26224328593996843,
      "grad_norm": 0.37942013144493103,
      "learning_rate": 0.00019997434521113193,
      "loss": 0.2456897497177124,
      "mean_token_accuracy": 0.9084974527359009,
      "num_tokens": 6143328.0,
      "step": 498
    },
    {
      "entropy": 1.363191157579422,
      "epoch": 0.26276987888362296,
      "grad_norm": 0.4037798047065735,
      "learning_rate": 0.00019997381899826663,
      "loss": 0.24723505973815918,
      "mean_token_accuracy": 0.9025511145591736,
      "num_tokens": 6155664.0,
      "step": 499
    },
    {
      "entropy": 1.3585219979286194,
      "epoch": 0.2632964718272775,
      "grad_norm": 0.4076964855194092,
      "learning_rate": 0.0001999732874441801,
      "loss": 0.231999009847641,
      "mean_token_accuracy": 0.9121885448694229,
      "num_tokens": 6168000.0,
      "step": 500
    },
    {
      "entropy": 1.3322370648384094,
      "epoch": 0.2638230647709321,
      "grad_norm": 0.38084790110588074,
      "learning_rate": 0.0001999727505489039,
      "loss": 0.26456475257873535,
      "mean_token_accuracy": 0.9003604352474213,
      "num_tokens": 6180336.0,
      "step": 501
    },
    {
      "entropy": 1.3225562274456024,
      "epoch": 0.2643496577145866,
      "grad_norm": 0.3669094145298004,
      "learning_rate": 0.00019997220831246987,
      "loss": 0.22636903822422028,
      "mean_token_accuracy": 0.9135940670967102,
      "num_tokens": 6192672.0,
      "step": 502
    },
    {
      "entropy": 1.24677973985672,
      "epoch": 0.26487625065824116,
      "grad_norm": 0.36622145771980286,
      "learning_rate": 0.00019997166073491027,
      "loss": 0.2442116141319275,
      "mean_token_accuracy": 0.903778612613678,
      "num_tokens": 6205008.0,
      "step": 503
    },
    {
      "entropy": 1.2564740180969238,
      "epoch": 0.26540284360189575,
      "grad_norm": 0.37914055585861206,
      "learning_rate": 0.00019997110781625754,
      "loss": 0.250071257352829,
      "mean_token_accuracy": 0.905722975730896,
      "num_tokens": 6217344.0,
      "step": 504
    },
    {
      "entropy": 1.266231119632721,
      "epoch": 0.2659294365455503,
      "grad_norm": 0.4232172966003418,
      "learning_rate": 0.00019997054955654456,
      "loss": 0.22360444068908691,
      "mean_token_accuracy": 0.9164858758449554,
      "num_tokens": 6229680.0,
      "step": 505
    },
    {
      "entropy": 1.2842857539653778,
      "epoch": 0.2664560294892048,
      "grad_norm": 0.40517833828926086,
      "learning_rate": 0.0001999699859558044,
      "loss": 0.24367284774780273,
      "mean_token_accuracy": 0.906918466091156,
      "num_tokens": 6242016.0,
      "step": 506
    },
    {
      "entropy": 1.3325933814048767,
      "epoch": 0.2669826224328594,
      "grad_norm": 0.3894166648387909,
      "learning_rate": 0.00019996941701407058,
      "loss": 0.21899431943893433,
      "mean_token_accuracy": 0.9159570783376694,
      "num_tokens": 6254352.0,
      "step": 507
    },
    {
      "entropy": 1.2919807434082031,
      "epoch": 0.26750921537651395,
      "grad_norm": 0.3554909825325012,
      "learning_rate": 0.00019996884273137686,
      "loss": 0.22241441905498505,
      "mean_token_accuracy": 0.9181459695100784,
      "num_tokens": 6266688.0,
      "step": 508
    },
    {
      "entropy": 1.311550348997116,
      "epoch": 0.2680358083201685,
      "grad_norm": 0.36536744236946106,
      "learning_rate": 0.0001999682631077573,
      "loss": 0.2456498146057129,
      "mean_token_accuracy": 0.9058470577001572,
      "num_tokens": 6279024.0,
      "step": 509
    },
    {
      "entropy": 1.3708328008651733,
      "epoch": 0.2685624012638231,
      "grad_norm": 0.4042948782444,
      "learning_rate": 0.00019996767814324637,
      "loss": 0.266501247882843,
      "mean_token_accuracy": 0.900126188993454,
      "num_tokens": 6291360.0,
      "step": 510
    },
    {
      "entropy": 1.3850271105766296,
      "epoch": 0.2690889942074776,
      "grad_norm": 0.39048540592193604,
      "learning_rate": 0.00019996708783787874,
      "loss": 0.24780599772930145,
      "mean_token_accuracy": 0.9016763418912888,
      "num_tokens": 6303696.0,
      "step": 511
    },
    {
      "entropy": 1.3151749074459076,
      "epoch": 0.26961558715113215,
      "grad_norm": 0.37497779726982117,
      "learning_rate": 0.0001999664921916895,
      "loss": 0.23127570748329163,
      "mean_token_accuracy": 0.9075058400630951,
      "num_tokens": 6316032.0,
      "step": 512
    },
    {
      "entropy": 1.3835597038269043,
      "epoch": 0.27014218009478674,
      "grad_norm": 0.3880425691604614,
      "learning_rate": 0.00019996589120471392,
      "loss": 0.23898069560527802,
      "mean_token_accuracy": 0.9100230634212494,
      "num_tokens": 6328368.0,
      "step": 513
    },
    {
      "entropy": 1.4238722920417786,
      "epoch": 0.2706687730384413,
      "grad_norm": 0.36198124289512634,
      "learning_rate": 0.0001999652848769878,
      "loss": 0.22968879342079163,
      "mean_token_accuracy": 0.906474158167839,
      "num_tokens": 6340704.0,
      "step": 514
    },
    {
      "entropy": 1.517054557800293,
      "epoch": 0.2711953659820958,
      "grad_norm": 0.3657878637313843,
      "learning_rate": 0.00019996467320854703,
      "loss": 0.22068339586257935,
      "mean_token_accuracy": 0.9112901985645294,
      "num_tokens": 6353040.0,
      "step": 515
    },
    {
      "entropy": 1.5027157366275787,
      "epoch": 0.2717219589257504,
      "grad_norm": 0.34945791959762573,
      "learning_rate": 0.000199964056199428,
      "loss": 0.25596681237220764,
      "mean_token_accuracy": 0.9006428420543671,
      "num_tokens": 6365376.0,
      "step": 516
    },
    {
      "entropy": 1.5874098539352417,
      "epoch": 0.27224855186940494,
      "grad_norm": 0.3800085783004761,
      "learning_rate": 0.00019996343384966732,
      "loss": 0.25539156794548035,
      "mean_token_accuracy": 0.901904970407486,
      "num_tokens": 6377712.0,
      "step": 517
    },
    {
      "entropy": 1.5508817732334137,
      "epoch": 0.2727751448130595,
      "grad_norm": 0.3190958797931671,
      "learning_rate": 0.00019996280615930186,
      "loss": 0.235137477517128,
      "mean_token_accuracy": 0.9069496989250183,
      "num_tokens": 6390048.0,
      "step": 518
    },
    {
      "entropy": 1.5596497654914856,
      "epoch": 0.27330173775671407,
      "grad_norm": 0.35837146639823914,
      "learning_rate": 0.000199962173128369,
      "loss": 0.24638012051582336,
      "mean_token_accuracy": 0.9043673276901245,
      "num_tokens": 6402384.0,
      "step": 519
    },
    {
      "entropy": 1.544301688671112,
      "epoch": 0.2738283307003686,
      "grad_norm": 0.3943609595298767,
      "learning_rate": 0.00019996153475690623,
      "loss": 0.24823732674121857,
      "mean_token_accuracy": 0.9039914757013321,
      "num_tokens": 6414720.0,
      "step": 520
    },
    {
      "entropy": 1.4979600608348846,
      "epoch": 0.2743549236440232,
      "grad_norm": 0.3533165156841278,
      "learning_rate": 0.00019996089104495148,
      "loss": 0.21794329583644867,
      "mean_token_accuracy": 0.9170633852481842,
      "num_tokens": 6427056.0,
      "step": 521
    },
    {
      "entropy": 1.5348587334156036,
      "epoch": 0.27488151658767773,
      "grad_norm": 0.3516184091567993,
      "learning_rate": 0.00019996024199254295,
      "loss": 0.21856778860092163,
      "mean_token_accuracy": 0.9159553945064545,
      "num_tokens": 6439392.0,
      "step": 522
    },
    {
      "entropy": 1.5593447387218475,
      "epoch": 0.27540810953133227,
      "grad_norm": 0.3706720173358917,
      "learning_rate": 0.0001999595875997192,
      "loss": 0.24685126543045044,
      "mean_token_accuracy": 0.9023942798376083,
      "num_tokens": 6451728.0,
      "step": 523
    },
    {
      "entropy": 1.5938279330730438,
      "epoch": 0.27593470247498686,
      "grad_norm": 0.4374881088733673,
      "learning_rate": 0.00019995892786651905,
      "loss": 0.24188384413719177,
      "mean_token_accuracy": 0.904677078127861,
      "num_tokens": 6464064.0,
      "step": 524
    },
    {
      "entropy": 1.469886839389801,
      "epoch": 0.2764612954186414,
      "grad_norm": 0.46137046813964844,
      "learning_rate": 0.00019995826279298166,
      "loss": 0.23798640072345734,
      "mean_token_accuracy": 0.9109133780002594,
      "num_tokens": 6476400.0,
      "step": 525
    },
    {
      "entropy": 1.5873245596885681,
      "epoch": 0.27698788836229593,
      "grad_norm": 0.416920006275177,
      "learning_rate": 0.00019995759237914656,
      "loss": 0.25300133228302,
      "mean_token_accuracy": 0.9008941799402237,
      "num_tokens": 6488736.0,
      "step": 526
    },
    {
      "entropy": 1.5558778047561646,
      "epoch": 0.2775144813059505,
      "grad_norm": 0.36363670229911804,
      "learning_rate": 0.0001999569166250535,
      "loss": 0.22929418087005615,
      "mean_token_accuracy": 0.9127305895090103,
      "num_tokens": 6501072.0,
      "step": 527
    },
    {
      "entropy": 1.5871147811412811,
      "epoch": 0.27804107424960506,
      "grad_norm": 0.3830122947692871,
      "learning_rate": 0.00019995623553074258,
      "loss": 0.2472982108592987,
      "mean_token_accuracy": 0.9040207862854004,
      "num_tokens": 6513408.0,
      "step": 528
    },
    {
      "entropy": 1.6365769505500793,
      "epoch": 0.2785676671932596,
      "grad_norm": 0.3964959383010864,
      "learning_rate": 0.00019995554909625428,
      "loss": 0.25368693470954895,
      "mean_token_accuracy": 0.9007250666618347,
      "num_tokens": 6525744.0,
      "step": 529
    },
    {
      "entropy": 1.7017327845096588,
      "epoch": 0.2790942601369142,
      "grad_norm": 0.391290545463562,
      "learning_rate": 0.00019995485732162934,
      "loss": 0.2360502928495407,
      "mean_token_accuracy": 0.9084432125091553,
      "num_tokens": 6538080.0,
      "step": 530
    },
    {
      "entropy": 1.6393355131149292,
      "epoch": 0.2796208530805687,
      "grad_norm": 0.38507914543151855,
      "learning_rate": 0.0001999541602069088,
      "loss": 0.23786881566047668,
      "mean_token_accuracy": 0.9061485677957535,
      "num_tokens": 6550416.0,
      "step": 531
    },
    {
      "entropy": 1.6275135576725006,
      "epoch": 0.28014744602422326,
      "grad_norm": 0.37895116209983826,
      "learning_rate": 0.0001999534577521341,
      "loss": 0.23145025968551636,
      "mean_token_accuracy": 0.9090943783521652,
      "num_tokens": 6562752.0,
      "step": 532
    },
    {
      "entropy": 1.598475605249405,
      "epoch": 0.28067403896787785,
      "grad_norm": 0.4161452054977417,
      "learning_rate": 0.00019995274995734686,
      "loss": 0.2727176547050476,
      "mean_token_accuracy": 0.8956159949302673,
      "num_tokens": 6575088.0,
      "step": 533
    },
    {
      "entropy": 1.5213307440280914,
      "epoch": 0.2812006319115324,
      "grad_norm": 0.43801331520080566,
      "learning_rate": 0.00019995203682258915,
      "loss": 0.22940769791603088,
      "mean_token_accuracy": 0.9128211736679077,
      "num_tokens": 6587424.0,
      "step": 534
    },
    {
      "entropy": 1.634105533361435,
      "epoch": 0.2817272248551869,
      "grad_norm": 0.4054886996746063,
      "learning_rate": 0.0001999513183479033,
      "loss": 0.23942533135414124,
      "mean_token_accuracy": 0.9082587212324142,
      "num_tokens": 6599760.0,
      "step": 535
    },
    {
      "entropy": 1.4907733798027039,
      "epoch": 0.2822538177988415,
      "grad_norm": 0.40736711025238037,
      "learning_rate": 0.00019995059453333194,
      "loss": 0.22832639515399933,
      "mean_token_accuracy": 0.9135842621326447,
      "num_tokens": 6612096.0,
      "step": 536
    },
    {
      "entropy": 1.5832161009311676,
      "epoch": 0.28278041074249605,
      "grad_norm": 0.3980075418949127,
      "learning_rate": 0.0001999498653789181,
      "loss": 0.24887777864933014,
      "mean_token_accuracy": 0.9017034024000168,
      "num_tokens": 6624432.0,
      "step": 537
    },
    {
      "entropy": 1.4983101785182953,
      "epoch": 0.2833070036861506,
      "grad_norm": 0.3781083822250366,
      "learning_rate": 0.00019994913088470498,
      "loss": 0.23180519044399261,
      "mean_token_accuracy": 0.9052098244428635,
      "num_tokens": 6636768.0,
      "step": 538
    },
    {
      "entropy": 1.5062950551509857,
      "epoch": 0.2838335966298052,
      "grad_norm": 0.4003005623817444,
      "learning_rate": 0.00019994839105073623,
      "loss": 0.22549505531787872,
      "mean_token_accuracy": 0.9140617251396179,
      "num_tokens": 6649104.0,
      "step": 539
    },
    {
      "entropy": 1.5416783392429352,
      "epoch": 0.2843601895734597,
      "grad_norm": 0.39962485432624817,
      "learning_rate": 0.00019994764587705574,
      "loss": 0.2179470807313919,
      "mean_token_accuracy": 0.9161404520273209,
      "num_tokens": 6661440.0,
      "step": 540
    },
    {
      "entropy": 1.660192757844925,
      "epoch": 0.28488678251711425,
      "grad_norm": 0.42626917362213135,
      "learning_rate": 0.0001999468953637078,
      "loss": 0.26295173168182373,
      "mean_token_accuracy": 0.8983173221349716,
      "num_tokens": 6673776.0,
      "step": 541
    },
    {
      "entropy": 1.5149348974227905,
      "epoch": 0.28541337546076884,
      "grad_norm": 0.36926260590553284,
      "learning_rate": 0.00019994613951073692,
      "loss": 0.2277842015028,
      "mean_token_accuracy": 0.9144963473081589,
      "num_tokens": 6686112.0,
      "step": 542
    },
    {
      "entropy": 1.5601712763309479,
      "epoch": 0.2859399684044234,
      "grad_norm": 0.35751771926879883,
      "learning_rate": 0.00019994537831818799,
      "loss": 0.2134617567062378,
      "mean_token_accuracy": 0.9172465950250626,
      "num_tokens": 6698448.0,
      "step": 543
    },
    {
      "entropy": 1.629770964384079,
      "epoch": 0.2864665613480779,
      "grad_norm": 0.3708803951740265,
      "learning_rate": 0.00019994461178610617,
      "loss": 0.22911599278450012,
      "mean_token_accuracy": 0.9093553721904755,
      "num_tokens": 6710784.0,
      "step": 544
    },
    {
      "entropy": 1.5735207796096802,
      "epoch": 0.2869931542917325,
      "grad_norm": 0.3979873061180115,
      "learning_rate": 0.00019994383991453702,
      "loss": 0.2454529106616974,
      "mean_token_accuracy": 0.9057244658470154,
      "num_tokens": 6723120.0,
      "step": 545
    },
    {
      "entropy": 1.556952327489853,
      "epoch": 0.28751974723538704,
      "grad_norm": 0.36042284965515137,
      "learning_rate": 0.0001999430627035263,
      "loss": 0.22943763434886932,
      "mean_token_accuracy": 0.9095467925071716,
      "num_tokens": 6735456.0,
      "step": 546
    },
    {
      "entropy": 1.5784813165664673,
      "epoch": 0.2880463401790416,
      "grad_norm": 0.4354793131351471,
      "learning_rate": 0.0001999422801531202,
      "loss": 0.23101268708705902,
      "mean_token_accuracy": 0.9098020642995834,
      "num_tokens": 6747792.0,
      "step": 547
    },
    {
      "entropy": 1.6356432437896729,
      "epoch": 0.28857293312269616,
      "grad_norm": 0.38589903712272644,
      "learning_rate": 0.00019994149226336514,
      "loss": 0.2551157772541046,
      "mean_token_accuracy": 0.9008017480373383,
      "num_tokens": 6760128.0,
      "step": 548
    },
    {
      "entropy": 1.484334021806717,
      "epoch": 0.2890995260663507,
      "grad_norm": 0.3685605823993683,
      "learning_rate": 0.0001999406990343079,
      "loss": 0.21177145838737488,
      "mean_token_accuracy": 0.9134591519832611,
      "num_tokens": 6772464.0,
      "step": 549
    },
    {
      "entropy": 1.5352662205696106,
      "epoch": 0.2896261190100053,
      "grad_norm": 0.43439221382141113,
      "learning_rate": 0.00019993990046599555,
      "loss": 0.21738113462924957,
      "mean_token_accuracy": 0.9168129861354828,
      "num_tokens": 6784800.0,
      "step": 550
    },
    {
      "entropy": 1.4901257157325745,
      "epoch": 0.2901527119536598,
      "grad_norm": 0.40765848755836487,
      "learning_rate": 0.00019993909655847554,
      "loss": 0.2299802005290985,
      "mean_token_accuracy": 0.9086113125085831,
      "num_tokens": 6797136.0,
      "step": 551
    },
    {
      "entropy": 1.5647250413894653,
      "epoch": 0.29067930489731436,
      "grad_norm": 0.3714927136898041,
      "learning_rate": 0.0001999382873117956,
      "loss": 0.23209913074970245,
      "mean_token_accuracy": 0.912913978099823,
      "num_tokens": 6809472.0,
      "step": 552
    },
    {
      "entropy": 1.5570427775382996,
      "epoch": 0.29120589784096895,
      "grad_norm": 0.3235602080821991,
      "learning_rate": 0.0001999374727260037,
      "loss": 0.20470383763313293,
      "mean_token_accuracy": 0.9178998619318008,
      "num_tokens": 6821808.0,
      "step": 553
    },
    {
      "entropy": 1.6096443831920624,
      "epoch": 0.2917324907846235,
      "grad_norm": 0.3791026473045349,
      "learning_rate": 0.00019993665280114824,
      "loss": 0.23357072472572327,
      "mean_token_accuracy": 0.9100991189479828,
      "num_tokens": 6834144.0,
      "step": 554
    },
    {
      "entropy": 1.7047458291053772,
      "epoch": 0.292259083728278,
      "grad_norm": 0.4588649272918701,
      "learning_rate": 0.00019993582753727792,
      "loss": 0.23484769463539124,
      "mean_token_accuracy": 0.9089436531066895,
      "num_tokens": 6846480.0,
      "step": 555
    },
    {
      "entropy": 1.6045478284358978,
      "epoch": 0.2927856766719326,
      "grad_norm": 0.38022348284721375,
      "learning_rate": 0.00019993499693444168,
      "loss": 0.24132560193538666,
      "mean_token_accuracy": 0.9133987128734589,
      "num_tokens": 6858816.0,
      "step": 556
    },
    {
      "entropy": 1.5578656792640686,
      "epoch": 0.29331226961558715,
      "grad_norm": 0.3959854543209076,
      "learning_rate": 0.00019993416099268888,
      "loss": 0.2288515269756317,
      "mean_token_accuracy": 0.9094084948301315,
      "num_tokens": 6871152.0,
      "step": 557
    },
    {
      "entropy": 1.5771779716014862,
      "epoch": 0.2938388625592417,
      "grad_norm": 0.41283491253852844,
      "learning_rate": 0.00019993331971206911,
      "loss": 0.25233519077301025,
      "mean_token_accuracy": 0.8999712020158768,
      "num_tokens": 6883488.0,
      "step": 558
    },
    {
      "entropy": 1.5294803977012634,
      "epoch": 0.2943654555028963,
      "grad_norm": 0.41696879267692566,
      "learning_rate": 0.0001999324730926323,
      "loss": 0.2409447431564331,
      "mean_token_accuracy": 0.9075838178396225,
      "num_tokens": 6895824.0,
      "step": 559
    },
    {
      "entropy": 1.557918757200241,
      "epoch": 0.2948920484465508,
      "grad_norm": 0.8365952372550964,
      "learning_rate": 0.00019993162113442873,
      "loss": 0.26256802678108215,
      "mean_token_accuracy": 0.8970607817173004,
      "num_tokens": 6908160.0,
      "step": 560
    },
    {
      "entropy": 1.5486240088939667,
      "epoch": 0.29541864139020535,
      "grad_norm": 0.36394545435905457,
      "learning_rate": 0.000199930763837509,
      "loss": 0.23241116106510162,
      "mean_token_accuracy": 0.9164627194404602,
      "num_tokens": 6920496.0,
      "step": 561
    },
    {
      "entropy": 1.5366742312908173,
      "epoch": 0.29594523433385994,
      "grad_norm": 0.3717273771762848,
      "learning_rate": 0.00019992990120192393,
      "loss": 0.2316376268863678,
      "mean_token_accuracy": 0.9116377085447311,
      "num_tokens": 6932832.0,
      "step": 562
    },
    {
      "entropy": 1.6174998581409454,
      "epoch": 0.2964718272775145,
      "grad_norm": 0.3976577818393707,
      "learning_rate": 0.00019992903322772483,
      "loss": 0.2560444474220276,
      "mean_token_accuracy": 0.8992763608694077,
      "num_tokens": 6945168.0,
      "step": 563
    },
    {
      "entropy": 1.5432467758655548,
      "epoch": 0.296998420221169,
      "grad_norm": 0.32837730646133423,
      "learning_rate": 0.00019992815991496318,
      "loss": 0.2185848504304886,
      "mean_token_accuracy": 0.9149428755044937,
      "num_tokens": 6957504.0,
      "step": 564
    },
    {
      "entropy": 1.5321990847587585,
      "epoch": 0.2975250131648236,
      "grad_norm": 0.36741021275520325,
      "learning_rate": 0.00019992728126369078,
      "loss": 0.24631758034229279,
      "mean_token_accuracy": 0.9008835554122925,
      "num_tokens": 6969840.0,
      "step": 565
    },
    {
      "entropy": 1.6573955416679382,
      "epoch": 0.29805160610847814,
      "grad_norm": 0.40060925483703613,
      "learning_rate": 0.00019992639727395983,
      "loss": 0.24705615639686584,
      "mean_token_accuracy": 0.9055600762367249,
      "num_tokens": 6982176.0,
      "step": 566
    },
    {
      "entropy": 1.6146283447742462,
      "epoch": 0.2985781990521327,
      "grad_norm": 0.36694321036338806,
      "learning_rate": 0.00019992550794582282,
      "loss": 0.2382228970527649,
      "mean_token_accuracy": 0.9077549129724503,
      "num_tokens": 6994512.0,
      "step": 567
    },
    {
      "entropy": 1.608702838420868,
      "epoch": 0.29910479199578727,
      "grad_norm": 0.36138370633125305,
      "learning_rate": 0.00019992461327933252,
      "loss": 0.2497398406267166,
      "mean_token_accuracy": 0.9065199494361877,
      "num_tokens": 7006848.0,
      "step": 568
    },
    {
      "entropy": 1.5681868195533752,
      "epoch": 0.2996313849394418,
      "grad_norm": 0.3718649744987488,
      "learning_rate": 0.00019992371327454204,
      "loss": 0.22592690587043762,
      "mean_token_accuracy": 0.9139988124370575,
      "num_tokens": 7019184.0,
      "step": 569
    },
    {
      "entropy": 1.5541615784168243,
      "epoch": 0.3001579778830964,
      "grad_norm": 0.3475581407546997,
      "learning_rate": 0.00019992280793150485,
      "loss": 0.218974769115448,
      "mean_token_accuracy": 0.9153032898902893,
      "num_tokens": 7031520.0,
      "step": 570
    },
    {
      "entropy": 1.5086607038974762,
      "epoch": 0.30068457082675093,
      "grad_norm": 0.336376428604126,
      "learning_rate": 0.00019992189725027466,
      "loss": 0.2033504694700241,
      "mean_token_accuracy": 0.9149883091449738,
      "num_tokens": 7043856.0,
      "step": 571
    },
    {
      "entropy": 1.5713791847229004,
      "epoch": 0.30121116377040547,
      "grad_norm": 0.4335557520389557,
      "learning_rate": 0.0001999209812309055,
      "loss": 0.2449672371149063,
      "mean_token_accuracy": 0.9020196795463562,
      "num_tokens": 7056192.0,
      "step": 572
    },
    {
      "entropy": 1.5503774285316467,
      "epoch": 0.30173775671406006,
      "grad_norm": 0.3847261071205139,
      "learning_rate": 0.0001999200598734518,
      "loss": 0.23619771003723145,
      "mean_token_accuracy": 0.9090457707643509,
      "num_tokens": 7068528.0,
      "step": 573
    },
    {
      "entropy": 1.5812987089157104,
      "epoch": 0.3022643496577146,
      "grad_norm": 0.4005545675754547,
      "learning_rate": 0.00019991913317796825,
      "loss": 0.22766467928886414,
      "mean_token_accuracy": 0.9142022579908371,
      "num_tokens": 7080864.0,
      "step": 574
    },
    {
      "entropy": 1.5593630969524384,
      "epoch": 0.30279094260136913,
      "grad_norm": 0.3692858815193176,
      "learning_rate": 0.00019991820114450984,
      "loss": 0.24361425638198853,
      "mean_token_accuracy": 0.909025177359581,
      "num_tokens": 7093200.0,
      "step": 575
    },
    {
      "entropy": 1.639871895313263,
      "epoch": 0.3033175355450237,
      "grad_norm": 0.3423769176006317,
      "learning_rate": 0.0001999172637731319,
      "loss": 0.1989578753709793,
      "mean_token_accuracy": 0.9223005324602127,
      "num_tokens": 7105536.0,
      "step": 576
    },
    {
      "entropy": 1.5683670938014984,
      "epoch": 0.30384412848867826,
      "grad_norm": 0.3434232771396637,
      "learning_rate": 0.0001999163210638901,
      "loss": 0.23291000723838806,
      "mean_token_accuracy": 0.9105765074491501,
      "num_tokens": 7117872.0,
      "step": 577
    },
    {
      "entropy": 1.6403064131736755,
      "epoch": 0.3043707214323328,
      "grad_norm": 0.3877623975276947,
      "learning_rate": 0.00019991537301684037,
      "loss": 0.2589385509490967,
      "mean_token_accuracy": 0.9013812988996506,
      "num_tokens": 7130208.0,
      "step": 578
    },
    {
      "entropy": 1.577894926071167,
      "epoch": 0.3048973143759874,
      "grad_norm": 0.3848702013492584,
      "learning_rate": 0.00019991441963203906,
      "loss": 0.2235189825296402,
      "mean_token_accuracy": 0.9132665693759918,
      "num_tokens": 7142544.0,
      "step": 579
    },
    {
      "entropy": 1.594813734292984,
      "epoch": 0.3054239073196419,
      "grad_norm": 0.37452247738838196,
      "learning_rate": 0.00019991346090954268,
      "loss": 0.2309045046567917,
      "mean_token_accuracy": 0.9061081409454346,
      "num_tokens": 7154880.0,
      "step": 580
    },
    {
      "entropy": 1.6004383862018585,
      "epoch": 0.30595050026329645,
      "grad_norm": 0.36386990547180176,
      "learning_rate": 0.0001999124968494082,
      "loss": 0.22697995603084564,
      "mean_token_accuracy": 0.9153416603803635,
      "num_tokens": 7167216.0,
      "step": 581
    },
    {
      "entropy": 1.5263923108577728,
      "epoch": 0.30647709320695105,
      "grad_norm": 0.3861697316169739,
      "learning_rate": 0.00019991152745169283,
      "loss": 0.23612037301063538,
      "mean_token_accuracy": 0.9100482612848282,
      "num_tokens": 7179552.0,
      "step": 582
    },
    {
      "entropy": 1.4913224875926971,
      "epoch": 0.3070036861506056,
      "grad_norm": 0.37040355801582336,
      "learning_rate": 0.00019991055271645412,
      "loss": 0.2453731894493103,
      "mean_token_accuracy": 0.9096489399671555,
      "num_tokens": 7191888.0,
      "step": 583
    },
    {
      "entropy": 1.5870050489902496,
      "epoch": 0.3075302790942601,
      "grad_norm": 0.3563641309738159,
      "learning_rate": 0.00019990957264374992,
      "loss": 0.2088014781475067,
      "mean_token_accuracy": 0.9178295880556107,
      "num_tokens": 7204224.0,
      "step": 584
    },
    {
      "entropy": 1.5673131942749023,
      "epoch": 0.3080568720379147,
      "grad_norm": 0.42992061376571655,
      "learning_rate": 0.00019990858723363842,
      "loss": 0.22837956249713898,
      "mean_token_accuracy": 0.9150240123271942,
      "num_tokens": 7216560.0,
      "step": 585
    },
    {
      "entropy": 1.5697237253189087,
      "epoch": 0.30858346498156924,
      "grad_norm": 0.34907442331314087,
      "learning_rate": 0.00019990759648617814,
      "loss": 0.2061898112297058,
      "mean_token_accuracy": 0.9186743050813675,
      "num_tokens": 7228896.0,
      "step": 586
    },
    {
      "entropy": 1.5203896760940552,
      "epoch": 0.3091100579252238,
      "grad_norm": 0.3814384639263153,
      "learning_rate": 0.00019990660040142787,
      "loss": 0.23518475890159607,
      "mean_token_accuracy": 0.9073454886674881,
      "num_tokens": 7241232.0,
      "step": 587
    },
    {
      "entropy": 1.5959535539150238,
      "epoch": 0.30963665086887837,
      "grad_norm": 0.410187304019928,
      "learning_rate": 0.00019990559897944675,
      "loss": 0.2526550590991974,
      "mean_token_accuracy": 0.9053095877170563,
      "num_tokens": 7253568.0,
      "step": 588
    },
    {
      "entropy": 1.4929281175136566,
      "epoch": 0.3101632438125329,
      "grad_norm": 0.3563387989997864,
      "learning_rate": 0.00019990459222029422,
      "loss": 0.2284306138753891,
      "mean_token_accuracy": 0.9153739362955093,
      "num_tokens": 7265904.0,
      "step": 589
    },
    {
      "entropy": 1.5385482907295227,
      "epoch": 0.31068983675618744,
      "grad_norm": 0.3958924412727356,
      "learning_rate": 0.00019990358012403009,
      "loss": 0.25064951181411743,
      "mean_token_accuracy": 0.9019092172384262,
      "num_tokens": 7278240.0,
      "step": 590
    },
    {
      "entropy": 1.5795462429523468,
      "epoch": 0.31121642969984203,
      "grad_norm": 0.4616306722164154,
      "learning_rate": 0.00019990256269071435,
      "loss": 0.25259649753570557,
      "mean_token_accuracy": 0.9031090140342712,
      "num_tokens": 7290576.0,
      "step": 591
    },
    {
      "entropy": 1.5900325179100037,
      "epoch": 0.31174302264349657,
      "grad_norm": 0.36789095401763916,
      "learning_rate": 0.0001999015399204075,
      "loss": 0.239149808883667,
      "mean_token_accuracy": 0.9080104380846024,
      "num_tokens": 7302912.0,
      "step": 592
    },
    {
      "entropy": 1.595876008272171,
      "epoch": 0.3122696155871511,
      "grad_norm": 0.3929153084754944,
      "learning_rate": 0.0001999005118131702,
      "loss": 0.2852938175201416,
      "mean_token_accuracy": 0.8941920101642609,
      "num_tokens": 7315248.0,
      "step": 593
    },
    {
      "entropy": 1.616676688194275,
      "epoch": 0.3127962085308057,
      "grad_norm": 0.3501856327056885,
      "learning_rate": 0.00019989947836906346,
      "loss": 0.24000589549541473,
      "mean_token_accuracy": 0.9087683409452438,
      "num_tokens": 7327584.0,
      "step": 594
    },
    {
      "entropy": 1.6581095457077026,
      "epoch": 0.31332280147446023,
      "grad_norm": 0.42188331484794617,
      "learning_rate": 0.0001998984395881487,
      "loss": 0.24285097420215607,
      "mean_token_accuracy": 0.9076326787471771,
      "num_tokens": 7339920.0,
      "step": 595
    },
    {
      "entropy": 1.6323899328708649,
      "epoch": 0.3138493944181148,
      "grad_norm": 0.38462045788764954,
      "learning_rate": 0.00019989739547048753,
      "loss": 0.25823602080345154,
      "mean_token_accuracy": 0.9016465842723846,
      "num_tokens": 7352256.0,
      "step": 596
    },
    {
      "entropy": 1.648387372493744,
      "epoch": 0.31437598736176936,
      "grad_norm": 0.3550112545490265,
      "learning_rate": 0.00019989634601614198,
      "loss": 0.23165376484394073,
      "mean_token_accuracy": 0.9123150110244751,
      "num_tokens": 7364592.0,
      "step": 597
    },
    {
      "entropy": 1.541102021932602,
      "epoch": 0.3149025803054239,
      "grad_norm": 0.32691195607185364,
      "learning_rate": 0.0001998952912251743,
      "loss": 0.2173168659210205,
      "mean_token_accuracy": 0.9170237630605698,
      "num_tokens": 7376928.0,
      "step": 598
    },
    {
      "entropy": 1.5581879019737244,
      "epoch": 0.3154291732490785,
      "grad_norm": 0.37725362181663513,
      "learning_rate": 0.00019989423109764714,
      "loss": 0.22197510302066803,
      "mean_token_accuracy": 0.9138648211956024,
      "num_tokens": 7389264.0,
      "step": 599
    },
    {
      "entropy": 1.5590221285820007,
      "epoch": 0.315955766192733,
      "grad_norm": 0.36627396941185,
      "learning_rate": 0.0001998931656336234,
      "loss": 0.23069526255130768,
      "mean_token_accuracy": 0.9046057909727097,
      "num_tokens": 7401600.0,
      "step": 600
    },
    {
      "entropy": 1.617866188287735,
      "epoch": 0.31648235913638756,
      "grad_norm": 0.40031948685646057,
      "learning_rate": 0.00019989209483316637,
      "loss": 0.2336081862449646,
      "mean_token_accuracy": 0.9123480021953583,
      "num_tokens": 7413936.0,
      "step": 601
    },
    {
      "entropy": 1.53180992603302,
      "epoch": 0.31700895208004215,
      "grad_norm": 0.49535101652145386,
      "learning_rate": 0.00019989101869633962,
      "loss": 0.22568359971046448,
      "mean_token_accuracy": 0.9089675545692444,
      "num_tokens": 7426272.0,
      "step": 602
    },
    {
      "entropy": 1.5532157719135284,
      "epoch": 0.3175355450236967,
      "grad_norm": 0.4378284811973572,
      "learning_rate": 0.00019988993722320703,
      "loss": 0.2375514954328537,
      "mean_token_accuracy": 0.9061621576547623,
      "num_tokens": 7438608.0,
      "step": 603
    },
    {
      "entropy": 1.5265016555786133,
      "epoch": 0.3180621379673512,
      "grad_norm": 0.3522114157676697,
      "learning_rate": 0.0001998888504138327,
      "loss": 0.23949339985847473,
      "mean_token_accuracy": 0.9073717594146729,
      "num_tokens": 7450944.0,
      "step": 604
    },
    {
      "entropy": 1.61136794090271,
      "epoch": 0.3185887309110058,
      "grad_norm": 0.37977924942970276,
      "learning_rate": 0.00019988775826828133,
      "loss": 0.2230483740568161,
      "mean_token_accuracy": 0.9096570760011673,
      "num_tokens": 7463280.0,
      "step": 605
    },
    {
      "entropy": 1.593753159046173,
      "epoch": 0.31911532385466035,
      "grad_norm": 0.6494004130363464,
      "learning_rate": 0.0001998866607866176,
      "loss": 0.22851543128490448,
      "mean_token_accuracy": 0.9144780784845352,
      "num_tokens": 7475616.0,
      "step": 606
    },
    {
      "entropy": 1.5087667405605316,
      "epoch": 0.3196419167983149,
      "grad_norm": 0.3527071177959442,
      "learning_rate": 0.00019988555796890674,
      "loss": 0.20573341846466064,
      "mean_token_accuracy": 0.9186414480209351,
      "num_tokens": 7487952.0,
      "step": 607
    },
    {
      "entropy": 1.5721814334392548,
      "epoch": 0.3201685097419695,
      "grad_norm": 0.3536371886730194,
      "learning_rate": 0.00019988444981521418,
      "loss": 0.20877021551132202,
      "mean_token_accuracy": 0.9166864901781082,
      "num_tokens": 7500288.0,
      "step": 608
    },
    {
      "entropy": 1.5660102665424347,
      "epoch": 0.320695102685624,
      "grad_norm": 0.4063898026943207,
      "learning_rate": 0.00019988333632560572,
      "loss": 0.25388604402542114,
      "mean_token_accuracy": 0.9036387950181961,
      "num_tokens": 7512624.0,
      "step": 609
    },
    {
      "entropy": 1.5472122132778168,
      "epoch": 0.32122169562927855,
      "grad_norm": 0.3346962630748749,
      "learning_rate": 0.00019988221750014747,
      "loss": 0.2217264026403427,
      "mean_token_accuracy": 0.9168782234191895,
      "num_tokens": 7524960.0,
      "step": 610
    },
    {
      "entropy": 1.557081937789917,
      "epoch": 0.32174828857293314,
      "grad_norm": 0.37367674708366394,
      "learning_rate": 0.00019988109333890584,
      "loss": 0.21202939748764038,
      "mean_token_accuracy": 0.9218468219041824,
      "num_tokens": 7537296.0,
      "step": 611
    },
    {
      "entropy": 1.4869562089443207,
      "epoch": 0.3222748815165877,
      "grad_norm": 0.4604549705982208,
      "learning_rate": 0.00019987996384194755,
      "loss": 0.2245841920375824,
      "mean_token_accuracy": 0.9145530760288239,
      "num_tokens": 7549632.0,
      "step": 612
    },
    {
      "entropy": 1.515323966741562,
      "epoch": 0.3228014744602422,
      "grad_norm": 0.39376160502433777,
      "learning_rate": 0.0001998788290093397,
      "loss": 0.26140767335891724,
      "mean_token_accuracy": 0.8977334201335907,
      "num_tokens": 7561968.0,
      "step": 613
    },
    {
      "entropy": 1.487097293138504,
      "epoch": 0.3233280674038968,
      "grad_norm": 0.40344923734664917,
      "learning_rate": 0.00019987768884114962,
      "loss": 0.21170933544635773,
      "mean_token_accuracy": 0.9142201244831085,
      "num_tokens": 7574304.0,
      "step": 614
    },
    {
      "entropy": 1.4696341156959534,
      "epoch": 0.32385466034755134,
      "grad_norm": 0.3404286801815033,
      "learning_rate": 0.00019987654333744498,
      "loss": 0.2107430100440979,
      "mean_token_accuracy": 0.9169973582029343,
      "num_tokens": 7586640.0,
      "step": 615
    },
    {
      "entropy": 1.4763599038124084,
      "epoch": 0.3243812532912059,
      "grad_norm": 0.4449232518672943,
      "learning_rate": 0.00019987539249829381,
      "loss": 0.22219976782798767,
      "mean_token_accuracy": 0.9147885292768478,
      "num_tokens": 7598976.0,
      "step": 616
    },
    {
      "entropy": 1.4189346730709076,
      "epoch": 0.32490784623486046,
      "grad_norm": 0.33330264687538147,
      "learning_rate": 0.00019987423632376444,
      "loss": 0.20641736686229706,
      "mean_token_accuracy": 0.9154144078493118,
      "num_tokens": 7611312.0,
      "step": 617
    },
    {
      "entropy": 1.5400767922401428,
      "epoch": 0.325434439178515,
      "grad_norm": 0.3576372563838959,
      "learning_rate": 0.00019987307481392544,
      "loss": 0.21838723123073578,
      "mean_token_accuracy": 0.9165307581424713,
      "num_tokens": 7623648.0,
      "step": 618
    },
    {
      "entropy": 1.4934936463832855,
      "epoch": 0.32596103212216954,
      "grad_norm": 0.3858436942100525,
      "learning_rate": 0.00019987190796884582,
      "loss": 0.2230948656797409,
      "mean_token_accuracy": 0.9111486077308655,
      "num_tokens": 7635984.0,
      "step": 619
    },
    {
      "entropy": 1.5231453776359558,
      "epoch": 0.3264876250658241,
      "grad_norm": 0.3916998505592346,
      "learning_rate": 0.0001998707357885949,
      "loss": 0.2513001561164856,
      "mean_token_accuracy": 0.9021778851747513,
      "num_tokens": 7648320.0,
      "step": 620
    },
    {
      "entropy": 1.529598742723465,
      "epoch": 0.32701421800947866,
      "grad_norm": 0.768818736076355,
      "learning_rate": 0.00019986955827324215,
      "loss": 0.23995141685009003,
      "mean_token_accuracy": 0.9082267880439758,
      "num_tokens": 7660656.0,
      "step": 621
    },
    {
      "entropy": 1.56441992521286,
      "epoch": 0.32754081095313325,
      "grad_norm": 0.3836771249771118,
      "learning_rate": 0.0001998683754228575,
      "loss": 0.2571730613708496,
      "mean_token_accuracy": 0.905523344874382,
      "num_tokens": 7672992.0,
      "step": 622
    },
    {
      "entropy": 1.5032837986946106,
      "epoch": 0.3280674038967878,
      "grad_norm": 0.3536396920681,
      "learning_rate": 0.00019986718723751122,
      "loss": 0.2316400706768036,
      "mean_token_accuracy": 0.9121131896972656,
      "num_tokens": 7685328.0,
      "step": 623
    },
    {
      "entropy": 1.5374043881893158,
      "epoch": 0.3285939968404423,
      "grad_norm": 0.43927958607673645,
      "learning_rate": 0.00019986599371727383,
      "loss": 0.2222973108291626,
      "mean_token_accuracy": 0.9122229516506195,
      "num_tokens": 7697664.0,
      "step": 624
    },
    {
      "entropy": 1.5948379039764404,
      "epoch": 0.3291205897840969,
      "grad_norm": 0.3739503026008606,
      "learning_rate": 0.00019986479486221618,
      "loss": 0.23531857132911682,
      "mean_token_accuracy": 0.9108096957206726,
      "num_tokens": 7710000.0,
      "step": 625
    },
    {
      "entropy": 1.5133698880672455,
      "epoch": 0.32964718272775145,
      "grad_norm": 0.37042292952537537,
      "learning_rate": 0.00019986359067240942,
      "loss": 0.2119014859199524,
      "mean_token_accuracy": 0.9210023730993271,
      "num_tokens": 7722336.0,
      "step": 626
    },
    {
      "entropy": 1.565012902021408,
      "epoch": 0.330173775671406,
      "grad_norm": 0.36645272374153137,
      "learning_rate": 0.00019986238114792505,
      "loss": 0.22745972871780396,
      "mean_token_accuracy": 0.9089792966842651,
      "num_tokens": 7734672.0,
      "step": 627
    },
    {
      "entropy": 1.5374813377857208,
      "epoch": 0.3307003686150606,
      "grad_norm": 0.3918452858924866,
      "learning_rate": 0.00019986116628883485,
      "loss": 0.22233808040618896,
      "mean_token_accuracy": 0.91575026512146,
      "num_tokens": 7747008.0,
      "step": 628
    },
    {
      "entropy": 1.5814277529716492,
      "epoch": 0.3312269615587151,
      "grad_norm": 0.3956620395183563,
      "learning_rate": 0.00019985994609521098,
      "loss": 0.23237353563308716,
      "mean_token_accuracy": 0.9120265543460846,
      "num_tokens": 7759344.0,
      "step": 629
    },
    {
      "entropy": 1.607151210308075,
      "epoch": 0.33175355450236965,
      "grad_norm": 0.4154573380947113,
      "learning_rate": 0.00019985872056712585,
      "loss": 0.23763325810432434,
      "mean_token_accuracy": 0.9039438813924789,
      "num_tokens": 7771680.0,
      "step": 630
    },
    {
      "entropy": 1.5939337015151978,
      "epoch": 0.33228014744602424,
      "grad_norm": 0.5367919206619263,
      "learning_rate": 0.0001998574897046522,
      "loss": 0.22427549958229065,
      "mean_token_accuracy": 0.9124085754156113,
      "num_tokens": 7784016.0,
      "step": 631
    },
    {
      "entropy": 1.6544562876224518,
      "epoch": 0.3328067403896788,
      "grad_norm": 0.49991947412490845,
      "learning_rate": 0.00019985625350786313,
      "loss": 0.26143935322761536,
      "mean_token_accuracy": 0.8970800191164017,
      "num_tokens": 7796352.0,
      "step": 632
    },
    {
      "entropy": 1.567628562450409,
      "epoch": 0.3333333333333333,
      "grad_norm": 0.43659552931785583,
      "learning_rate": 0.00019985501197683202,
      "loss": 0.22153861820697784,
      "mean_token_accuracy": 0.9091014564037323,
      "num_tokens": 7808688.0,
      "step": 633
    },
    {
      "entropy": 1.6360341310501099,
      "epoch": 0.3338599262769879,
      "grad_norm": 0.406440794467926,
      "learning_rate": 0.00019985376511163255,
      "loss": 0.23250769078731537,
      "mean_token_accuracy": 0.9082053154706955,
      "num_tokens": 7821024.0,
      "step": 634
    },
    {
      "entropy": 1.617436408996582,
      "epoch": 0.33438651922064244,
      "grad_norm": 0.4746139347553253,
      "learning_rate": 0.00019985251291233872,
      "loss": 0.2448023557662964,
      "mean_token_accuracy": 0.9019272774457932,
      "num_tokens": 7833360.0,
      "step": 635
    },
    {
      "entropy": 1.5576602518558502,
      "epoch": 0.334913112164297,
      "grad_norm": 0.37426477670669556,
      "learning_rate": 0.00019985125537902497,
      "loss": 0.24818755686283112,
      "mean_token_accuracy": 0.9056387394666672,
      "num_tokens": 7845696.0,
      "step": 636
    },
    {
      "entropy": 1.5108153223991394,
      "epoch": 0.33543970510795157,
      "grad_norm": 0.36447635293006897,
      "learning_rate": 0.00019984999251176582,
      "loss": 0.23007383942604065,
      "mean_token_accuracy": 0.9102648496627808,
      "num_tokens": 7858032.0,
      "step": 637
    },
    {
      "entropy": 1.6435940563678741,
      "epoch": 0.3359662980516061,
      "grad_norm": 0.4201553165912628,
      "learning_rate": 0.00019984872431063632,
      "loss": 0.2524194121360779,
      "mean_token_accuracy": 0.9026057571172714,
      "num_tokens": 7870368.0,
      "step": 638
    },
    {
      "entropy": 1.5785358846187592,
      "epoch": 0.33649289099526064,
      "grad_norm": 0.34197092056274414,
      "learning_rate": 0.00019984745077571175,
      "loss": 0.22064879536628723,
      "mean_token_accuracy": 0.9168655872344971,
      "num_tokens": 7882704.0,
      "step": 639
    },
    {
      "entropy": 1.5749069154262543,
      "epoch": 0.33701948393891523,
      "grad_norm": 0.35670214891433716,
      "learning_rate": 0.00019984617190706768,
      "loss": 0.210345059633255,
      "mean_token_accuracy": 0.9159483909606934,
      "num_tokens": 7895040.0,
      "step": 640
    },
    {
      "entropy": 1.648300975561142,
      "epoch": 0.33754607688256977,
      "grad_norm": 0.5585328936576843,
      "learning_rate": 0.00019984488770478004,
      "loss": 0.26494815945625305,
      "mean_token_accuracy": 0.9005215167999268,
      "num_tokens": 7907376.0,
      "step": 641
    },
    {
      "entropy": 1.5516612827777863,
      "epoch": 0.3380726698262243,
      "grad_norm": 0.3812706172466278,
      "learning_rate": 0.0001998435981689251,
      "loss": 0.23725152015686035,
      "mean_token_accuracy": 0.9066549986600876,
      "num_tokens": 7919712.0,
      "step": 642
    },
    {
      "entropy": 1.538433849811554,
      "epoch": 0.3385992627698789,
      "grad_norm": 0.345907598733902,
      "learning_rate": 0.00019984230329957937,
      "loss": 0.21600767970085144,
      "mean_token_accuracy": 0.913017600774765,
      "num_tokens": 7932048.0,
      "step": 643
    },
    {
      "entropy": 1.5793551802635193,
      "epoch": 0.33912585571353343,
      "grad_norm": 0.375407874584198,
      "learning_rate": 0.00019984100309681973,
      "loss": 0.2297327220439911,
      "mean_token_accuracy": 0.9100670218467712,
      "num_tokens": 7944384.0,
      "step": 644
    },
    {
      "entropy": 1.6768275797367096,
      "epoch": 0.33965244865718797,
      "grad_norm": 0.3743014931678772,
      "learning_rate": 0.00019983969756072337,
      "loss": 0.23133614659309387,
      "mean_token_accuracy": 0.9111513495445251,
      "num_tokens": 7956720.0,
      "step": 645
    },
    {
      "entropy": 1.5642622113227844,
      "epoch": 0.34017904160084256,
      "grad_norm": 0.3964219093322754,
      "learning_rate": 0.00019983838669136782,
      "loss": 0.25138217210769653,
      "mean_token_accuracy": 0.9010957926511765,
      "num_tokens": 7969056.0,
      "step": 646
    },
    {
      "entropy": 1.562336266040802,
      "epoch": 0.3407056345444971,
      "grad_norm": 0.33525609970092773,
      "learning_rate": 0.00019983707048883083,
      "loss": 0.21662499010562897,
      "mean_token_accuracy": 0.9179873615503311,
      "num_tokens": 7981392.0,
      "step": 647
    },
    {
      "entropy": 1.5184325873851776,
      "epoch": 0.3412322274881517,
      "grad_norm": 0.36040404438972473,
      "learning_rate": 0.0001998357489531906,
      "loss": 0.21142053604125977,
      "mean_token_accuracy": 0.9129342883825302,
      "num_tokens": 7993728.0,
      "step": 648
    },
    {
      "entropy": 1.525155395269394,
      "epoch": 0.3417588204318062,
      "grad_norm": 0.37644830346107483,
      "learning_rate": 0.00019983442208452553,
      "loss": 0.24553003907203674,
      "mean_token_accuracy": 0.9037658125162125,
      "num_tokens": 8006064.0,
      "step": 649
    },
    {
      "entropy": 1.5811420381069183,
      "epoch": 0.34228541337546076,
      "grad_norm": 0.4055664837360382,
      "learning_rate": 0.00019983308988291446,
      "loss": 0.24890564382076263,
      "mean_token_accuracy": 0.9027184396982193,
      "num_tokens": 8018400.0,
      "step": 650
    },
    {
      "entropy": 1.5781408250331879,
      "epoch": 0.34281200631911535,
      "grad_norm": 0.37236547470092773,
      "learning_rate": 0.0001998317523484364,
      "loss": 0.2316565066576004,
      "mean_token_accuracy": 0.9119908362627029,
      "num_tokens": 8030736.0,
      "step": 651
    },
    {
      "entropy": 1.470947027206421,
      "epoch": 0.3433385992627699,
      "grad_norm": 0.3868812918663025,
      "learning_rate": 0.00019983040948117078,
      "loss": 0.2534410059452057,
      "mean_token_accuracy": 0.9021570980548859,
      "num_tokens": 8043072.0,
      "step": 652
    },
    {
      "entropy": 1.5348678827285767,
      "epoch": 0.3438651922064244,
      "grad_norm": 0.4387357234954834,
      "learning_rate": 0.00019982906128119732,
      "loss": 0.22310803830623627,
      "mean_token_accuracy": 0.9145376235246658,
      "num_tokens": 8055408.0,
      "step": 653
    },
    {
      "entropy": 1.6187977194786072,
      "epoch": 0.344391785150079,
      "grad_norm": 0.3973976671695709,
      "learning_rate": 0.00019982770774859608,
      "loss": 0.24628473818302155,
      "mean_token_accuracy": 0.9026829153299332,
      "num_tokens": 8067744.0,
      "step": 654
    },
    {
      "entropy": 1.552239865064621,
      "epoch": 0.34491837809373355,
      "grad_norm": 0.35203489661216736,
      "learning_rate": 0.00019982634888344737,
      "loss": 0.2246260643005371,
      "mean_token_accuracy": 0.9107431769371033,
      "num_tokens": 8080080.0,
      "step": 655
    },
    {
      "entropy": 1.5582615733146667,
      "epoch": 0.3454449710373881,
      "grad_norm": 0.35882458090782166,
      "learning_rate": 0.00019982498468583187,
      "loss": 0.24929207563400269,
      "mean_token_accuracy": 0.9079926460981369,
      "num_tokens": 8092416.0,
      "step": 656
    },
    {
      "entropy": 1.5417989492416382,
      "epoch": 0.3459715639810427,
      "grad_norm": 0.3812078833580017,
      "learning_rate": 0.00019982361515583056,
      "loss": 0.23627105355262756,
      "mean_token_accuracy": 0.9105822741985321,
      "num_tokens": 8104752.0,
      "step": 657
    },
    {
      "entropy": 1.5216515362262726,
      "epoch": 0.3464981569246972,
      "grad_norm": 0.425227552652359,
      "learning_rate": 0.00019982224029352477,
      "loss": 0.24528373777866364,
      "mean_token_accuracy": 0.9078731536865234,
      "num_tokens": 8117088.0,
      "step": 658
    },
    {
      "entropy": 1.5206952691078186,
      "epoch": 0.34702474986835175,
      "grad_norm": 0.33637750148773193,
      "learning_rate": 0.0001998208600989961,
      "loss": 0.2107471525669098,
      "mean_token_accuracy": 0.9135650843381882,
      "num_tokens": 8129424.0,
      "step": 659
    },
    {
      "entropy": 1.4702115952968597,
      "epoch": 0.34755134281200634,
      "grad_norm": 0.39711427688598633,
      "learning_rate": 0.00019981947457232646,
      "loss": 0.2600536048412323,
      "mean_token_accuracy": 0.900258481502533,
      "num_tokens": 8141760.0,
      "step": 660
    },
    {
      "entropy": 1.4251591563224792,
      "epoch": 0.34807793575566087,
      "grad_norm": 0.3307092487812042,
      "learning_rate": 0.00019981808371359816,
      "loss": 0.20669502019882202,
      "mean_token_accuracy": 0.9150898307561874,
      "num_tokens": 8154096.0,
      "step": 661
    },
    {
      "entropy": 1.4800917506217957,
      "epoch": 0.3486045286993154,
      "grad_norm": 0.3357905447483063,
      "learning_rate": 0.00019981668752289372,
      "loss": 0.21784240007400513,
      "mean_token_accuracy": 0.9145694226026535,
      "num_tokens": 8166432.0,
      "step": 662
    },
    {
      "entropy": 1.4939378798007965,
      "epoch": 0.34913112164297,
      "grad_norm": 0.3817315101623535,
      "learning_rate": 0.00019981528600029598,
      "loss": 0.23391678929328918,
      "mean_token_accuracy": 0.9120414704084396,
      "num_tokens": 8178768.0,
      "step": 663
    },
    {
      "entropy": 1.46476212143898,
      "epoch": 0.34965771458662454,
      "grad_norm": 0.38629287481307983,
      "learning_rate": 0.00019981387914588822,
      "loss": 0.2348746955394745,
      "mean_token_accuracy": 0.9141973257064819,
      "num_tokens": 8191104.0,
      "step": 664
    },
    {
      "entropy": 1.488828182220459,
      "epoch": 0.35018430753027907,
      "grad_norm": 0.43263551592826843,
      "learning_rate": 0.00019981246695975396,
      "loss": 0.24639199674129486,
      "mean_token_accuracy": 0.9060526490211487,
      "num_tokens": 8203440.0,
      "step": 665
    },
    {
      "entropy": 1.4127525091171265,
      "epoch": 0.35071090047393366,
      "grad_norm": 0.3648802936077118,
      "learning_rate": 0.00019981104944197698,
      "loss": 0.21946357190608978,
      "mean_token_accuracy": 0.9159405082464218,
      "num_tokens": 8215776.0,
      "step": 666
    },
    {
      "entropy": 1.4903503954410553,
      "epoch": 0.3512374934175882,
      "grad_norm": 0.34730279445648193,
      "learning_rate": 0.00019980962659264144,
      "loss": 0.2281692773103714,
      "mean_token_accuracy": 0.9098831713199615,
      "num_tokens": 8228112.0,
      "step": 667
    },
    {
      "entropy": 1.5247552990913391,
      "epoch": 0.35176408636124273,
      "grad_norm": 0.6796702742576599,
      "learning_rate": 0.00019980819841183185,
      "loss": 0.2541208267211914,
      "mean_token_accuracy": 0.899626687169075,
      "num_tokens": 8240448.0,
      "step": 668
    },
    {
      "entropy": 1.489711880683899,
      "epoch": 0.3522906793048973,
      "grad_norm": 0.33805036544799805,
      "learning_rate": 0.00019980676489963294,
      "loss": 0.2260848581790924,
      "mean_token_accuracy": 0.9087798148393631,
      "num_tokens": 8252784.0,
      "step": 669
    },
    {
      "entropy": 1.5027905106544495,
      "epoch": 0.35281727224855186,
      "grad_norm": 0.350353479385376,
      "learning_rate": 0.00019980532605612985,
      "loss": 0.21086406707763672,
      "mean_token_accuracy": 0.9210447818040848,
      "num_tokens": 8265120.0,
      "step": 670
    },
    {
      "entropy": 1.5674294233322144,
      "epoch": 0.35334386519220645,
      "grad_norm": 0.3340439200401306,
      "learning_rate": 0.00019980388188140798,
      "loss": 0.20525240898132324,
      "mean_token_accuracy": 0.922586515545845,
      "num_tokens": 8277456.0,
      "step": 671
    },
    {
      "entropy": 1.5054379403591156,
      "epoch": 0.353870458135861,
      "grad_norm": 0.3567858934402466,
      "learning_rate": 0.00019980243237555304,
      "loss": 0.2413492202758789,
      "mean_token_accuracy": 0.9065290242433548,
      "num_tokens": 8289792.0,
      "step": 672
    },
    {
      "entropy": 1.458409994840622,
      "epoch": 0.3543970510795155,
      "grad_norm": 0.3392413854598999,
      "learning_rate": 0.00019980097753865108,
      "loss": 0.21458403766155243,
      "mean_token_accuracy": 0.915942057967186,
      "num_tokens": 8302128.0,
      "step": 673
    },
    {
      "entropy": 1.5474618673324585,
      "epoch": 0.3549236440231701,
      "grad_norm": 0.3782627582550049,
      "learning_rate": 0.00019979951737078853,
      "loss": 0.24051563441753387,
      "mean_token_accuracy": 0.905216172337532,
      "num_tokens": 8314464.0,
      "step": 674
    },
    {
      "entropy": 1.453640639781952,
      "epoch": 0.35545023696682465,
      "grad_norm": 0.3473234474658966,
      "learning_rate": 0.000199798051872052,
      "loss": 0.22346456348896027,
      "mean_token_accuracy": 0.9117260426282883,
      "num_tokens": 8326800.0,
      "step": 675
    },
    {
      "entropy": 1.4625252783298492,
      "epoch": 0.3559768299104792,
      "grad_norm": 0.3387875258922577,
      "learning_rate": 0.0001997965810425285,
      "loss": 0.23459070920944214,
      "mean_token_accuracy": 0.9097730964422226,
      "num_tokens": 8339136.0,
      "step": 676
    },
    {
      "entropy": 1.4122467339038849,
      "epoch": 0.3565034228541338,
      "grad_norm": 0.3719231188297272,
      "learning_rate": 0.0001997951048823054,
      "loss": 0.21170935034751892,
      "mean_token_accuracy": 0.9177156686782837,
      "num_tokens": 8351472.0,
      "step": 677
    },
    {
      "entropy": 1.3950088024139404,
      "epoch": 0.3570300157977883,
      "grad_norm": 0.346206933259964,
      "learning_rate": 0.00019979362339147026,
      "loss": 0.24381937086582184,
      "mean_token_accuracy": 0.9096541851758957,
      "num_tokens": 8363808.0,
      "step": 678
    },
    {
      "entropy": 1.392151266336441,
      "epoch": 0.35755660874144285,
      "grad_norm": 0.4163087010383606,
      "learning_rate": 0.00019979213657011106,
      "loss": 0.23680301010608673,
      "mean_token_accuracy": 0.9078295677900314,
      "num_tokens": 8376144.0,
      "step": 679
    },
    {
      "entropy": 1.272907316684723,
      "epoch": 0.35808320168509744,
      "grad_norm": 0.4227026104927063,
      "learning_rate": 0.00019979064441831606,
      "loss": 0.2319648414850235,
      "mean_token_accuracy": 0.9085201323032379,
      "num_tokens": 8388480.0,
      "step": 680
    },
    {
      "entropy": 1.3540238738059998,
      "epoch": 0.358609794628752,
      "grad_norm": 0.41346514225006104,
      "learning_rate": 0.00019978914693617383,
      "loss": 0.228549063205719,
      "mean_token_accuracy": 0.9105959385633469,
      "num_tokens": 8400816.0,
      "step": 681
    },
    {
      "entropy": 1.360448271036148,
      "epoch": 0.3591363875724065,
      "grad_norm": 0.3442930281162262,
      "learning_rate": 0.0001997876441237733,
      "loss": 0.23922371864318848,
      "mean_token_accuracy": 0.9045034646987915,
      "num_tokens": 8413152.0,
      "step": 682
    },
    {
      "entropy": 1.2645371556282043,
      "epoch": 0.3596629805160611,
      "grad_norm": 0.3163891136646271,
      "learning_rate": 0.00019978613598120362,
      "loss": 0.23533323407173157,
      "mean_token_accuracy": 0.9094915986061096,
      "num_tokens": 8425488.0,
      "step": 683
    },
    {
      "entropy": 1.23646479845047,
      "epoch": 0.36018957345971564,
      "grad_norm": 0.3051406741142273,
      "learning_rate": 0.00019978462250855439,
      "loss": 0.23106315732002258,
      "mean_token_accuracy": 0.908999964594841,
      "num_tokens": 8437824.0,
      "step": 684
    },
    {
      "entropy": 1.2108670771121979,
      "epoch": 0.3607161664033702,
      "grad_norm": 0.3296745717525482,
      "learning_rate": 0.00019978310370591545,
      "loss": 0.21918885409832,
      "mean_token_accuracy": 0.9144444465637207,
      "num_tokens": 8450160.0,
      "step": 685
    },
    {
      "entropy": 1.2031162977218628,
      "epoch": 0.36124275934702477,
      "grad_norm": 0.34492868185043335,
      "learning_rate": 0.0001997815795733769,
      "loss": 0.23954619467258453,
      "mean_token_accuracy": 0.9077229052782059,
      "num_tokens": 8462496.0,
      "step": 686
    },
    {
      "entropy": 1.186975747346878,
      "epoch": 0.3617693522906793,
      "grad_norm": 0.3709777593612671,
      "learning_rate": 0.00019978005011102925,
      "loss": 0.2437351793050766,
      "mean_token_accuracy": 0.908468559384346,
      "num_tokens": 8474832.0,
      "step": 687
    },
    {
      "entropy": 1.2189928591251373,
      "epoch": 0.36229594523433384,
      "grad_norm": 0.39078283309936523,
      "learning_rate": 0.00019977851531896335,
      "loss": 0.23512142896652222,
      "mean_token_accuracy": 0.9079535007476807,
      "num_tokens": 8487168.0,
      "step": 688
    },
    {
      "entropy": 1.1402603685855865,
      "epoch": 0.36282253817798843,
      "grad_norm": 0.44031041860580444,
      "learning_rate": 0.00019977697519727025,
      "loss": 0.22278329730033875,
      "mean_token_accuracy": 0.9095389991998672,
      "num_tokens": 8499504.0,
      "step": 689
    },
    {
      "entropy": 1.2126906514167786,
      "epoch": 0.36334913112164297,
      "grad_norm": 0.4217199385166168,
      "learning_rate": 0.00019977542974604137,
      "loss": 0.2309587299823761,
      "mean_token_accuracy": 0.9086143374443054,
      "num_tokens": 8511840.0,
      "step": 690
    },
    {
      "entropy": 1.2279771864414215,
      "epoch": 0.3638757240652975,
      "grad_norm": 0.38159534335136414,
      "learning_rate": 0.00019977387896536847,
      "loss": 0.22851325571537018,
      "mean_token_accuracy": 0.9110317975282669,
      "num_tokens": 8524176.0,
      "step": 691
    },
    {
      "entropy": 1.1451427340507507,
      "epoch": 0.3644023170089521,
      "grad_norm": 0.38951390981674194,
      "learning_rate": 0.00019977232285534364,
      "loss": 0.22088123857975006,
      "mean_token_accuracy": 0.9204992055892944,
      "num_tokens": 8536512.0,
      "step": 692
    },
    {
      "entropy": 1.2615329325199127,
      "epoch": 0.36492890995260663,
      "grad_norm": 0.4139882028102875,
      "learning_rate": 0.0001997707614160592,
      "loss": 0.24705028533935547,
      "mean_token_accuracy": 0.9034076929092407,
      "num_tokens": 8548848.0,
      "step": 693
    },
    {
      "entropy": 1.283850759267807,
      "epoch": 0.36545550289626116,
      "grad_norm": 0.369973361492157,
      "learning_rate": 0.00019976919464760793,
      "loss": 0.22415512800216675,
      "mean_token_accuracy": 0.9107316434383392,
      "num_tokens": 8561184.0,
      "step": 694
    },
    {
      "entropy": 1.2292048037052155,
      "epoch": 0.36598209583991576,
      "grad_norm": 0.38250577449798584,
      "learning_rate": 0.00019976762255008277,
      "loss": 0.23568174242973328,
      "mean_token_accuracy": 0.9063956588506699,
      "num_tokens": 8573520.0,
      "step": 695
    },
    {
      "entropy": 1.2365350425243378,
      "epoch": 0.3665086887835703,
      "grad_norm": 0.3148071765899658,
      "learning_rate": 0.00019976604512357706,
      "loss": 0.1988360434770584,
      "mean_token_accuracy": 0.9224959909915924,
      "num_tokens": 8585856.0,
      "step": 696
    },
    {
      "entropy": 1.3029758036136627,
      "epoch": 0.3670352817272249,
      "grad_norm": 0.3744061589241028,
      "learning_rate": 0.00019976446236818444,
      "loss": 0.20495107769966125,
      "mean_token_accuracy": 0.9228791147470474,
      "num_tokens": 8598192.0,
      "step": 697
    },
    {
      "entropy": 1.2882287502288818,
      "epoch": 0.3675618746708794,
      "grad_norm": 0.36938560009002686,
      "learning_rate": 0.00019976287428399888,
      "loss": 0.21769095957279205,
      "mean_token_accuracy": 0.9142359048128128,
      "num_tokens": 8610528.0,
      "step": 698
    },
    {
      "entropy": 1.3306386470794678,
      "epoch": 0.36808846761453395,
      "grad_norm": 0.34892189502716064,
      "learning_rate": 0.00019976128087111468,
      "loss": 0.22379228472709656,
      "mean_token_accuracy": 0.910999983549118,
      "num_tokens": 8622864.0,
      "step": 699
    },
    {
      "entropy": 1.422240823507309,
      "epoch": 0.36861506055818855,
      "grad_norm": 0.4013206660747528,
      "learning_rate": 0.00019975968212962637,
      "loss": 0.24196377396583557,
      "mean_token_accuracy": 0.9052925109863281,
      "num_tokens": 8635200.0,
      "step": 700
    },
    {
      "entropy": 1.4155401587486267,
      "epoch": 0.3691416535018431,
      "grad_norm": 0.353211373090744,
      "learning_rate": 0.0001997580780596289,
      "loss": 0.23057395219802856,
      "mean_token_accuracy": 0.9112053066492081,
      "num_tokens": 8647536.0,
      "step": 701
    },
    {
      "entropy": 1.3654289841651917,
      "epoch": 0.3696682464454976,
      "grad_norm": 0.3298124372959137,
      "learning_rate": 0.0001997564686612175,
      "loss": 0.23379994928836823,
      "mean_token_accuracy": 0.9082982391119003,
      "num_tokens": 8659872.0,
      "step": 702
    },
    {
      "entropy": 1.4584643244743347,
      "epoch": 0.3701948393891522,
      "grad_norm": 0.3759617805480957,
      "learning_rate": 0.00019975485393448764,
      "loss": 0.23104792833328247,
      "mean_token_accuracy": 0.9065845310688019,
      "num_tokens": 8672208.0,
      "step": 703
    },
    {
      "entropy": 1.444229632616043,
      "epoch": 0.37072143233280674,
      "grad_norm": 0.3501303791999817,
      "learning_rate": 0.00019975323387953528,
      "loss": 0.2252783626317978,
      "mean_token_accuracy": 0.9078841656446457,
      "num_tokens": 8684544.0,
      "step": 704
    },
    {
      "entropy": 1.4961676001548767,
      "epoch": 0.3712480252764613,
      "grad_norm": 0.37864333391189575,
      "learning_rate": 0.00019975160849645656,
      "loss": 0.248760387301445,
      "mean_token_accuracy": 0.9001511335372925,
      "num_tokens": 8696880.0,
      "step": 705
    },
    {
      "entropy": 1.42547208070755,
      "epoch": 0.37177461822011587,
      "grad_norm": 0.34873253107070923,
      "learning_rate": 0.00019974997778534793,
      "loss": 0.21211083233356476,
      "mean_token_accuracy": 0.9166723936796188,
      "num_tokens": 8709216.0,
      "step": 706
    },
    {
      "entropy": 1.5042538046836853,
      "epoch": 0.3723012111637704,
      "grad_norm": 0.3550424873828888,
      "learning_rate": 0.00019974834174630622,
      "loss": 0.23886847496032715,
      "mean_token_accuracy": 0.907848909497261,
      "num_tokens": 8721552.0,
      "step": 707
    },
    {
      "entropy": 1.5081537961959839,
      "epoch": 0.37282780410742494,
      "grad_norm": 0.41367873549461365,
      "learning_rate": 0.00019974670037942855,
      "loss": 0.22889745235443115,
      "mean_token_accuracy": 0.9153233766555786,
      "num_tokens": 8733888.0,
      "step": 708
    },
    {
      "entropy": 1.3966407477855682,
      "epoch": 0.37335439705107953,
      "grad_norm": 0.3816620707511902,
      "learning_rate": 0.00019974505368481238,
      "loss": 0.22411897778511047,
      "mean_token_accuracy": 0.9170472025871277,
      "num_tokens": 8746224.0,
      "step": 709
    },
    {
      "entropy": 1.3528437912464142,
      "epoch": 0.37388098999473407,
      "grad_norm": 0.3909305930137634,
      "learning_rate": 0.00019974340166255543,
      "loss": 0.24883610010147095,
      "mean_token_accuracy": 0.9069965332746506,
      "num_tokens": 8758560.0,
      "step": 710
    },
    {
      "entropy": 1.3574591279029846,
      "epoch": 0.3744075829383886,
      "grad_norm": 0.3753831088542938,
      "learning_rate": 0.00019974174431275582,
      "loss": 0.21758557856082916,
      "mean_token_accuracy": 0.9138903617858887,
      "num_tokens": 8770896.0,
      "step": 711
    },
    {
      "entropy": 1.3783180713653564,
      "epoch": 0.3749341758820432,
      "grad_norm": 0.3737315833568573,
      "learning_rate": 0.0001997400816355119,
      "loss": 0.21359983086585999,
      "mean_token_accuracy": 0.9189057648181915,
      "num_tokens": 8783232.0,
      "step": 712
    },
    {
      "entropy": 1.3974202573299408,
      "epoch": 0.37546076882569773,
      "grad_norm": 0.349575012922287,
      "learning_rate": 0.00019973841363092232,
      "loss": 0.2197628617286682,
      "mean_token_accuracy": 0.910584107041359,
      "num_tokens": 8795568.0,
      "step": 713
    },
    {
      "entropy": 1.4738838374614716,
      "epoch": 0.37598736176935227,
      "grad_norm": 0.39151832461357117,
      "learning_rate": 0.0001997367402990862,
      "loss": 0.22688592970371246,
      "mean_token_accuracy": 0.9129604697227478,
      "num_tokens": 8807904.0,
      "step": 714
    },
    {
      "entropy": 1.4208093285560608,
      "epoch": 0.37651395471300686,
      "grad_norm": 0.41211023926734924,
      "learning_rate": 0.00019973506164010284,
      "loss": 0.20385678112506866,
      "mean_token_accuracy": 0.9227319210767746,
      "num_tokens": 8820240.0,
      "step": 715
    },
    {
      "entropy": 1.4047819375991821,
      "epoch": 0.3770405476566614,
      "grad_norm": 0.4299006164073944,
      "learning_rate": 0.00019973337765407187,
      "loss": 0.23750200867652893,
      "mean_token_accuracy": 0.9077501893043518,
      "num_tokens": 8832576.0,
      "step": 716
    },
    {
      "entropy": 1.374906837940216,
      "epoch": 0.37756714060031593,
      "grad_norm": 0.36846232414245605,
      "learning_rate": 0.00019973168834109327,
      "loss": 0.23285596072673798,
      "mean_token_accuracy": 0.9094402939081192,
      "num_tokens": 8844912.0,
      "step": 717
    },
    {
      "entropy": 1.4372902512550354,
      "epoch": 0.3780937335439705,
      "grad_norm": 0.3780461549758911,
      "learning_rate": 0.00019972999370126737,
      "loss": 0.23158109188079834,
      "mean_token_accuracy": 0.9099433869123459,
      "num_tokens": 8857248.0,
      "step": 718
    },
    {
      "entropy": 1.3874932825565338,
      "epoch": 0.37862032648762506,
      "grad_norm": 0.34537607431411743,
      "learning_rate": 0.00019972829373469467,
      "loss": 0.21864044666290283,
      "mean_token_accuracy": 0.9146191030740738,
      "num_tokens": 8869584.0,
      "step": 719
    },
    {
      "entropy": 1.374584138393402,
      "epoch": 0.3791469194312796,
      "grad_norm": 0.33857598900794983,
      "learning_rate": 0.00019972658844147621,
      "loss": 0.20933406054973602,
      "mean_token_accuracy": 0.917707160115242,
      "num_tokens": 8881920.0,
      "step": 720
    },
    {
      "entropy": 1.3695291578769684,
      "epoch": 0.3796735123749342,
      "grad_norm": 0.5717244148254395,
      "learning_rate": 0.0001997248778217131,
      "loss": 0.22078999876976013,
      "mean_token_accuracy": 0.9167369604110718,
      "num_tokens": 8894256.0,
      "step": 721
    },
    {
      "entropy": 1.26867213845253,
      "epoch": 0.3802001053185887,
      "grad_norm": 0.36367470026016235,
      "learning_rate": 0.000199723161875507,
      "loss": 0.23508650064468384,
      "mean_token_accuracy": 0.90509033203125,
      "num_tokens": 8906592.0,
      "step": 722
    },
    {
      "entropy": 1.3050034940242767,
      "epoch": 0.3807266982622433,
      "grad_norm": 0.3631988763809204,
      "learning_rate": 0.0001997214406029597,
      "loss": 0.22164851427078247,
      "mean_token_accuracy": 0.9132580459117889,
      "num_tokens": 8918928.0,
      "step": 723
    },
    {
      "entropy": 1.3181822001934052,
      "epoch": 0.38125329120589785,
      "grad_norm": 0.4014921486377716,
      "learning_rate": 0.00019971971400417342,
      "loss": 0.220194011926651,
      "mean_token_accuracy": 0.9143774807453156,
      "num_tokens": 8931264.0,
      "step": 724
    },
    {
      "entropy": 1.2655711770057678,
      "epoch": 0.3817798841495524,
      "grad_norm": 0.34297892451286316,
      "learning_rate": 0.00019971798207925063,
      "loss": 0.21592603623867035,
      "mean_token_accuracy": 0.9114037603139877,
      "num_tokens": 8943600.0,
      "step": 725
    },
    {
      "entropy": 1.3559162318706512,
      "epoch": 0.382306477093207,
      "grad_norm": 0.39595136046409607,
      "learning_rate": 0.00019971624482829415,
      "loss": 0.2135222852230072,
      "mean_token_accuracy": 0.9156046807765961,
      "num_tokens": 8955936.0,
      "step": 726
    },
    {
      "entropy": 1.4213075637817383,
      "epoch": 0.3828330700368615,
      "grad_norm": 0.35875996947288513,
      "learning_rate": 0.00019971450225140716,
      "loss": 0.22025166451931,
      "mean_token_accuracy": 0.9119539260864258,
      "num_tokens": 8968272.0,
      "step": 727
    },
    {
      "entropy": 1.2975660860538483,
      "epoch": 0.38335966298051605,
      "grad_norm": 0.33709245920181274,
      "learning_rate": 0.00019971275434869302,
      "loss": 0.2429068237543106,
      "mean_token_accuracy": 0.9056646823883057,
      "num_tokens": 8980608.0,
      "step": 728
    },
    {
      "entropy": 1.3944336771965027,
      "epoch": 0.38388625592417064,
      "grad_norm": 0.3580012321472168,
      "learning_rate": 0.0001997110011202556,
      "loss": 0.228425070643425,
      "mean_token_accuracy": 0.9088682681322098,
      "num_tokens": 8992944.0,
      "step": 729
    },
    {
      "entropy": 1.386294186115265,
      "epoch": 0.3844128488678252,
      "grad_norm": 0.5559288859367371,
      "learning_rate": 0.00019970924256619888,
      "loss": 0.22206491231918335,
      "mean_token_accuracy": 0.9095857888460159,
      "num_tokens": 9005280.0,
      "step": 730
    },
    {
      "entropy": 1.3968450129032135,
      "epoch": 0.3849394418114797,
      "grad_norm": 0.38346561789512634,
      "learning_rate": 0.0001997074786866273,
      "loss": 0.2516261637210846,
      "mean_token_accuracy": 0.9072631448507309,
      "num_tokens": 9017616.0,
      "step": 731
    },
    {
      "entropy": 1.3758391439914703,
      "epoch": 0.3854660347551343,
      "grad_norm": 0.4184466302394867,
      "learning_rate": 0.00019970570948164555,
      "loss": 0.2083946019411087,
      "mean_token_accuracy": 0.915515810251236,
      "num_tokens": 9029952.0,
      "step": 732
    },
    {
      "entropy": 1.3424640595912933,
      "epoch": 0.38599262769878884,
      "grad_norm": 0.3482263684272766,
      "learning_rate": 0.00019970393495135868,
      "loss": 0.23488564789295197,
      "mean_token_accuracy": 0.9109048992395401,
      "num_tokens": 9042288.0,
      "step": 733
    },
    {
      "entropy": 1.3400058150291443,
      "epoch": 0.3865192206424434,
      "grad_norm": 0.4521651268005371,
      "learning_rate": 0.00019970215509587205,
      "loss": 0.24672454595565796,
      "mean_token_accuracy": 0.9039757996797562,
      "num_tokens": 9054624.0,
      "step": 734
    },
    {
      "entropy": 1.3971166908740997,
      "epoch": 0.38704581358609796,
      "grad_norm": 0.3671362102031708,
      "learning_rate": 0.0001997003699152913,
      "loss": 0.22332459688186646,
      "mean_token_accuracy": 0.9150716066360474,
      "num_tokens": 9066960.0,
      "step": 735
    },
    {
      "entropy": 1.3570407629013062,
      "epoch": 0.3875724065297525,
      "grad_norm": 0.4007357358932495,
      "learning_rate": 0.00019969857940972235,
      "loss": 0.22975003719329834,
      "mean_token_accuracy": 0.9176786690950394,
      "num_tokens": 9079296.0,
      "step": 736
    },
    {
      "entropy": 1.3730257153511047,
      "epoch": 0.38809899947340704,
      "grad_norm": 0.3905254602432251,
      "learning_rate": 0.00019969678357927158,
      "loss": 0.22929853200912476,
      "mean_token_accuracy": 0.909871906042099,
      "num_tokens": 9091632.0,
      "step": 737
    },
    {
      "entropy": 1.3547758758068085,
      "epoch": 0.3886255924170616,
      "grad_norm": 0.36366337537765503,
      "learning_rate": 0.00019969498242404556,
      "loss": 0.22672298550605774,
      "mean_token_accuracy": 0.9105400443077087,
      "num_tokens": 9103968.0,
      "step": 738
    },
    {
      "entropy": 1.322388231754303,
      "epoch": 0.38915218536071616,
      "grad_norm": 0.3948267698287964,
      "learning_rate": 0.0001996931759441512,
      "loss": 0.21981076896190643,
      "mean_token_accuracy": 0.9127620607614517,
      "num_tokens": 9116304.0,
      "step": 739
    },
    {
      "entropy": 1.3535179495811462,
      "epoch": 0.3896787783043707,
      "grad_norm": 0.39007002115249634,
      "learning_rate": 0.00019969136413969577,
      "loss": 0.23180383443832397,
      "mean_token_accuracy": 0.9031165540218353,
      "num_tokens": 9128640.0,
      "step": 740
    },
    {
      "entropy": 1.395143300294876,
      "epoch": 0.3902053712480253,
      "grad_norm": 0.39624056220054626,
      "learning_rate": 0.00019968954701078678,
      "loss": 0.21380290389060974,
      "mean_token_accuracy": 0.9165746420621872,
      "num_tokens": 9140976.0,
      "step": 741
    },
    {
      "entropy": 1.3894523680210114,
      "epoch": 0.3907319641916798,
      "grad_norm": 0.37505805492401123,
      "learning_rate": 0.00019968772455753218,
      "loss": 0.21238526701927185,
      "mean_token_accuracy": 0.9176094681024551,
      "num_tokens": 9153312.0,
      "step": 742
    },
    {
      "entropy": 1.3369868099689484,
      "epoch": 0.39125855713533436,
      "grad_norm": 0.364109605550766,
      "learning_rate": 0.00019968589678004008,
      "loss": 0.2217993289232254,
      "mean_token_accuracy": 0.9145223498344421,
      "num_tokens": 9165648.0,
      "step": 743
    },
    {
      "entropy": 1.4128968715667725,
      "epoch": 0.39178515007898895,
      "grad_norm": 0.37613528966903687,
      "learning_rate": 0.00019968406367841903,
      "loss": 0.2087160050868988,
      "mean_token_accuracy": 0.9174602627754211,
      "num_tokens": 9177984.0,
      "step": 744
    },
    {
      "entropy": 1.3632179498672485,
      "epoch": 0.3923117430226435,
      "grad_norm": 0.3908347189426422,
      "learning_rate": 0.00019968222525277785,
      "loss": 0.24387134611606598,
      "mean_token_accuracy": 0.9096127450466156,
      "num_tokens": 9190320.0,
      "step": 745
    },
    {
      "entropy": 1.3598384857177734,
      "epoch": 0.392838335966298,
      "grad_norm": 0.4070599377155304,
      "learning_rate": 0.00019968038150322564,
      "loss": 0.22589103877544403,
      "mean_token_accuracy": 0.9176203459501266,
      "num_tokens": 9202656.0,
      "step": 746
    },
    {
      "entropy": 1.2700338661670685,
      "epoch": 0.3933649289099526,
      "grad_norm": 0.3903157711029053,
      "learning_rate": 0.0001996785324298719,
      "loss": 0.27620363235473633,
      "mean_token_accuracy": 0.89641934633255,
      "num_tokens": 9214992.0,
      "step": 747
    },
    {
      "entropy": 1.3561755120754242,
      "epoch": 0.39389152185360715,
      "grad_norm": 0.3912152647972107,
      "learning_rate": 0.00019967667803282637,
      "loss": 0.20860803127288818,
      "mean_token_accuracy": 0.918526291847229,
      "num_tokens": 9227328.0,
      "step": 748
    },
    {
      "entropy": 1.2694224119186401,
      "epoch": 0.39441811479726174,
      "grad_norm": 0.36671656370162964,
      "learning_rate": 0.00019967481831219917,
      "loss": 0.23243974149227142,
      "mean_token_accuracy": 0.9083072245121002,
      "num_tokens": 9239664.0,
      "step": 749
    },
    {
      "entropy": 1.2892687618732452,
      "epoch": 0.3949447077409163,
      "grad_norm": 0.33213678002357483,
      "learning_rate": 0.0001996729532681006,
      "loss": 0.22482578456401825,
      "mean_token_accuracy": 0.9108980298042297,
      "num_tokens": 9252000.0,
      "step": 750
    },
    {
      "entropy": 1.2784943878650665,
      "epoch": 0.3954713006845708,
      "grad_norm": 0.311139851808548,
      "learning_rate": 0.0001996710829006415,
      "loss": 0.22491638362407684,
      "mean_token_accuracy": 0.9146853238344193,
      "num_tokens": 9264336.0,
      "step": 751
    },
    {
      "entropy": 1.3392630815505981,
      "epoch": 0.3959978936282254,
      "grad_norm": 0.31522372364997864,
      "learning_rate": 0.00019966920720993286,
      "loss": 0.19653388857841492,
      "mean_token_accuracy": 0.9204227477312088,
      "num_tokens": 9276672.0,
      "step": 752
    },
    {
      "entropy": 1.3383756875991821,
      "epoch": 0.39652448657187994,
      "grad_norm": 0.353919118642807,
      "learning_rate": 0.00019966732619608598,
      "loss": 0.23490512371063232,
      "mean_token_accuracy": 0.9106432646512985,
      "num_tokens": 9289008.0,
      "step": 753
    },
    {
      "entropy": 1.3619754612445831,
      "epoch": 0.3970510795155345,
      "grad_norm": 0.40393969416618347,
      "learning_rate": 0.00019966543985921258,
      "loss": 0.25288036465644836,
      "mean_token_accuracy": 0.8984902203083038,
      "num_tokens": 9301344.0,
      "step": 754
    },
    {
      "entropy": 1.2913717925548553,
      "epoch": 0.39757767245918907,
      "grad_norm": 0.3196253478527069,
      "learning_rate": 0.00019966354819942462,
      "loss": 0.20618736743927002,
      "mean_token_accuracy": 0.9167482256889343,
      "num_tokens": 9313680.0,
      "step": 755
    },
    {
      "entropy": 1.3224665522575378,
      "epoch": 0.3981042654028436,
      "grad_norm": 0.32193470001220703,
      "learning_rate": 0.0001996616512168344,
      "loss": 0.21416081488132477,
      "mean_token_accuracy": 0.9158345758914948,
      "num_tokens": 9326016.0,
      "step": 756
    },
    {
      "entropy": 1.3358683288097382,
      "epoch": 0.39863085834649814,
      "grad_norm": 0.3324636220932007,
      "learning_rate": 0.00019965974891155458,
      "loss": 0.23043948411941528,
      "mean_token_accuracy": 0.9085040837526321,
      "num_tokens": 9338352.0,
      "step": 757
    },
    {
      "entropy": 1.3305124342441559,
      "epoch": 0.39915745129015273,
      "grad_norm": 0.34178489446640015,
      "learning_rate": 0.00019965784128369798,
      "loss": 0.22156858444213867,
      "mean_token_accuracy": 0.9111177027225494,
      "num_tokens": 9350688.0,
      "step": 758
    },
    {
      "entropy": 1.3993423879146576,
      "epoch": 0.39968404423380727,
      "grad_norm": 0.35117653012275696,
      "learning_rate": 0.00019965592833337797,
      "loss": 0.24540282785892487,
      "mean_token_accuracy": 0.9031884074211121,
      "num_tokens": 9363024.0,
      "step": 759
    },
    {
      "entropy": 1.363684356212616,
      "epoch": 0.4002106371774618,
      "grad_norm": 0.3588322699069977,
      "learning_rate": 0.000199654010060708,
      "loss": 0.23427176475524902,
      "mean_token_accuracy": 0.9089349061250687,
      "num_tokens": 9375360.0,
      "step": 760
    },
    {
      "entropy": 1.3599910140037537,
      "epoch": 0.4007372301211164,
      "grad_norm": 0.3444473147392273,
      "learning_rate": 0.000199652086465802,
      "loss": 0.2288491129875183,
      "mean_token_accuracy": 0.909742534160614,
      "num_tokens": 9387696.0,
      "step": 761
    },
    {
      "entropy": 1.3465708196163177,
      "epoch": 0.40126382306477093,
      "grad_norm": 0.3881551921367645,
      "learning_rate": 0.00019965015754877419,
      "loss": 0.21851959824562073,
      "mean_token_accuracy": 0.911541685461998,
      "num_tokens": 9400032.0,
      "step": 762
    },
    {
      "entropy": 1.3606258928775787,
      "epoch": 0.40179041600842547,
      "grad_norm": 0.3825187683105469,
      "learning_rate": 0.00019964822330973902,
      "loss": 0.24844150245189667,
      "mean_token_accuracy": 0.905233234167099,
      "num_tokens": 9412368.0,
      "step": 763
    },
    {
      "entropy": 1.4477742910385132,
      "epoch": 0.40231700895208006,
      "grad_norm": 0.4269307553768158,
      "learning_rate": 0.00019964628374881133,
      "loss": 0.21143238246440887,
      "mean_token_accuracy": 0.914307177066803,
      "num_tokens": 9424704.0,
      "step": 764
    },
    {
      "entropy": 1.4273262321949005,
      "epoch": 0.4028436018957346,
      "grad_norm": 0.477578341960907,
      "learning_rate": 0.0001996443388661063,
      "loss": 0.21041172742843628,
      "mean_token_accuracy": 0.9194318652153015,
      "num_tokens": 9437040.0,
      "step": 765
    },
    {
      "entropy": 1.42348974943161,
      "epoch": 0.40337019483938913,
      "grad_norm": 0.42482423782348633,
      "learning_rate": 0.00019964238866173933,
      "loss": 0.22475981712341309,
      "mean_token_accuracy": 0.9095583707094193,
      "num_tokens": 9449376.0,
      "step": 766
    },
    {
      "entropy": 1.3782070577144623,
      "epoch": 0.4038967877830437,
      "grad_norm": 0.3717379570007324,
      "learning_rate": 0.00019964043313582626,
      "loss": 0.2339317947626114,
      "mean_token_accuracy": 0.9064943641424179,
      "num_tokens": 9461712.0,
      "step": 767
    },
    {
      "entropy": 1.3753153681755066,
      "epoch": 0.40442338072669826,
      "grad_norm": 0.32700079679489136,
      "learning_rate": 0.00019963847228848312,
      "loss": 0.2289605587720871,
      "mean_token_accuracy": 0.9136441200971603,
      "num_tokens": 9474048.0,
      "step": 768
    },
    {
      "entropy": 1.3423367142677307,
      "epoch": 0.4049499736703528,
      "grad_norm": 0.4840228855609894,
      "learning_rate": 0.00019963650611982636,
      "loss": 0.24808013439178467,
      "mean_token_accuracy": 0.9013350456953049,
      "num_tokens": 9486384.0,
      "step": 769
    },
    {
      "entropy": 1.388725221157074,
      "epoch": 0.4054765666140074,
      "grad_norm": 0.3298131823539734,
      "learning_rate": 0.0001996345346299726,
      "loss": 0.21849092841148376,
      "mean_token_accuracy": 0.9153172671794891,
      "num_tokens": 9498720.0,
      "step": 770
    },
    {
      "entropy": 1.362036794424057,
      "epoch": 0.4060031595576619,
      "grad_norm": 0.423976331949234,
      "learning_rate": 0.00019963255781903902,
      "loss": 0.2337607443332672,
      "mean_token_accuracy": 0.9124665558338165,
      "num_tokens": 9511056.0,
      "step": 771
    },
    {
      "entropy": 1.321341723203659,
      "epoch": 0.40652975250131645,
      "grad_norm": 0.39548537135124207,
      "learning_rate": 0.00019963057568714288,
      "loss": 0.2241300493478775,
      "mean_token_accuracy": 0.9098886847496033,
      "num_tokens": 9523392.0,
      "step": 772
    },
    {
      "entropy": 1.2956216037273407,
      "epoch": 0.40705634544497105,
      "grad_norm": 0.36130490899086,
      "learning_rate": 0.00019962858823440184,
      "loss": 0.22631655633449554,
      "mean_token_accuracy": 0.9111677557229996,
      "num_tokens": 9535728.0,
      "step": 773
    },
    {
      "entropy": 1.3697873950004578,
      "epoch": 0.4075829383886256,
      "grad_norm": 0.3786010444164276,
      "learning_rate": 0.00019962659546093395,
      "loss": 0.22168497741222382,
      "mean_token_accuracy": 0.9110663533210754,
      "num_tokens": 9548064.0,
      "step": 774
    },
    {
      "entropy": 1.3025212287902832,
      "epoch": 0.4081095313322802,
      "grad_norm": 0.3891861140727997,
      "learning_rate": 0.00019962459736685745,
      "loss": 0.2645792067050934,
      "mean_token_accuracy": 0.8977373838424683,
      "num_tokens": 9560400.0,
      "step": 775
    },
    {
      "entropy": 1.3207987844944,
      "epoch": 0.4086361242759347,
      "grad_norm": 0.40660449862480164,
      "learning_rate": 0.000199622593952291,
      "loss": 0.23059073090553284,
      "mean_token_accuracy": 0.9096915870904922,
      "num_tokens": 9572736.0,
      "step": 776
    },
    {
      "entropy": 1.2878498435020447,
      "epoch": 0.40916271721958924,
      "grad_norm": 0.35745811462402344,
      "learning_rate": 0.00019962058521735346,
      "loss": 0.21263521909713745,
      "mean_token_accuracy": 0.9142482280731201,
      "num_tokens": 9585072.0,
      "step": 777
    },
    {
      "entropy": 1.2785497307777405,
      "epoch": 0.40968931016324384,
      "grad_norm": 0.3724184036254883,
      "learning_rate": 0.00019961857116216415,
      "loss": 0.2524154484272003,
      "mean_token_accuracy": 0.8991471230983734,
      "num_tokens": 9597408.0,
      "step": 778
    },
    {
      "entropy": 1.2408149540424347,
      "epoch": 0.41021590310689837,
      "grad_norm": 0.3448520600795746,
      "learning_rate": 0.00019961655178684263,
      "loss": 0.20971742272377014,
      "mean_token_accuracy": 0.9130227416753769,
      "num_tokens": 9609744.0,
      "step": 779
    },
    {
      "entropy": 1.2202830016613007,
      "epoch": 0.4107424960505529,
      "grad_norm": 0.37864333391189575,
      "learning_rate": 0.0001996145270915087,
      "loss": 0.23903977870941162,
      "mean_token_accuracy": 0.9029916822910309,
      "num_tokens": 9622080.0,
      "step": 780
    },
    {
      "entropy": 1.1748075485229492,
      "epoch": 0.4112690889942075,
      "grad_norm": 0.36790093779563904,
      "learning_rate": 0.0001996124970762827,
      "loss": 0.24326302111148834,
      "mean_token_accuracy": 0.9089493155479431,
      "num_tokens": 9634416.0,
      "step": 781
    },
    {
      "entropy": 1.1609198153018951,
      "epoch": 0.41179568193786203,
      "grad_norm": 0.3445703685283661,
      "learning_rate": 0.00019961046174128496,
      "loss": 0.22028180956840515,
      "mean_token_accuracy": 0.9107068032026291,
      "num_tokens": 9646752.0,
      "step": 782
    },
    {
      "entropy": 1.1026230156421661,
      "epoch": 0.41232227488151657,
      "grad_norm": 0.3440577983856201,
      "learning_rate": 0.00019960842108663644,
      "loss": 0.19753555953502655,
      "mean_token_accuracy": 0.9187215715646744,
      "num_tokens": 9659088.0,
      "step": 783
    },
    {
      "entropy": 1.1320343911647797,
      "epoch": 0.41284886782517116,
      "grad_norm": 0.38988834619522095,
      "learning_rate": 0.00019960637511245823,
      "loss": 0.24359619617462158,
      "mean_token_accuracy": 0.9055552929639816,
      "num_tokens": 9671424.0,
      "step": 784
    },
    {
      "entropy": 1.1358483731746674,
      "epoch": 0.4133754607688257,
      "grad_norm": 0.34981632232666016,
      "learning_rate": 0.0001996043238188718,
      "loss": 0.24487480521202087,
      "mean_token_accuracy": 0.9070784449577332,
      "num_tokens": 9683760.0,
      "step": 785
    },
    {
      "entropy": 1.1212266087532043,
      "epoch": 0.41390205371248023,
      "grad_norm": 0.3241615891456604,
      "learning_rate": 0.0001996022672059989,
      "loss": 0.2086082100868225,
      "mean_token_accuracy": 0.9190200418233871,
      "num_tokens": 9696096.0,
      "step": 786
    },
    {
      "entropy": 1.1631585955619812,
      "epoch": 0.4144286466561348,
      "grad_norm": 0.36856940388679504,
      "learning_rate": 0.00019960020527396165,
      "loss": 0.20457178354263306,
      "mean_token_accuracy": 0.9173056781291962,
      "num_tokens": 9708432.0,
      "step": 787
    },
    {
      "entropy": 1.1695051193237305,
      "epoch": 0.41495523959978936,
      "grad_norm": 0.3899996876716614,
      "learning_rate": 0.00019959813802288245,
      "loss": 0.24216774106025696,
      "mean_token_accuracy": 0.9103662818670273,
      "num_tokens": 9720768.0,
      "step": 788
    },
    {
      "entropy": 1.1804447770118713,
      "epoch": 0.4154818325434439,
      "grad_norm": 0.3940189778804779,
      "learning_rate": 0.000199596065452884,
      "loss": 0.2604479491710663,
      "mean_token_accuracy": 0.8987935930490494,
      "num_tokens": 9733104.0,
      "step": 789
    },
    {
      "entropy": 1.2298611104488373,
      "epoch": 0.4160084254870985,
      "grad_norm": 0.3992246687412262,
      "learning_rate": 0.00019959398756408937,
      "loss": 0.24308446049690247,
      "mean_token_accuracy": 0.9046371877193451,
      "num_tokens": 9745440.0,
      "step": 790
    },
    {
      "entropy": 1.130595326423645,
      "epoch": 0.416535018430753,
      "grad_norm": 0.32028424739837646,
      "learning_rate": 0.00019959190435662186,
      "loss": 0.21361656486988068,
      "mean_token_accuracy": 0.9175485223531723,
      "num_tokens": 9757776.0,
      "step": 791
    },
    {
      "entropy": 1.2031766474246979,
      "epoch": 0.41706161137440756,
      "grad_norm": 0.3323441445827484,
      "learning_rate": 0.00019958981583060518,
      "loss": 0.22529007494449615,
      "mean_token_accuracy": 0.9116253554821014,
      "num_tokens": 9770112.0,
      "step": 792
    },
    {
      "entropy": 1.1709198653697968,
      "epoch": 0.41758820431806215,
      "grad_norm": 0.3305797874927521,
      "learning_rate": 0.0001995877219861633,
      "loss": 0.21258684992790222,
      "mean_token_accuracy": 0.9117163866758347,
      "num_tokens": 9782448.0,
      "step": 793
    },
    {
      "entropy": 1.2462036609649658,
      "epoch": 0.4181147972617167,
      "grad_norm": 0.32563477754592896,
      "learning_rate": 0.00019958562282342056,
      "loss": 0.22679030895233154,
      "mean_token_accuracy": 0.9114914387464523,
      "num_tokens": 9794784.0,
      "step": 794
    },
    {
      "entropy": 1.1364922225475311,
      "epoch": 0.4186413902053712,
      "grad_norm": 0.3199447989463806,
      "learning_rate": 0.0001995835183425015,
      "loss": 0.22316469252109528,
      "mean_token_accuracy": 0.9091953933238983,
      "num_tokens": 9807120.0,
      "step": 795
    },
    {
      "entropy": 1.2300072014331818,
      "epoch": 0.4191679831490258,
      "grad_norm": 0.3554495871067047,
      "learning_rate": 0.0001995814085435311,
      "loss": 0.2148018628358841,
      "mean_token_accuracy": 0.9172432273626328,
      "num_tokens": 9819456.0,
      "step": 796
    },
    {
      "entropy": 1.2751358449459076,
      "epoch": 0.41969457609268035,
      "grad_norm": 0.3625534772872925,
      "learning_rate": 0.00019957929342663457,
      "loss": 0.217167928814888,
      "mean_token_accuracy": 0.9148837774991989,
      "num_tokens": 9831792.0,
      "step": 797
    },
    {
      "entropy": 1.2123389542102814,
      "epoch": 0.42022116903633494,
      "grad_norm": 0.3511140048503876,
      "learning_rate": 0.00019957717299193752,
      "loss": 0.21881210803985596,
      "mean_token_accuracy": 0.9147087782621384,
      "num_tokens": 9844128.0,
      "step": 798
    },
    {
      "entropy": 1.2617943584918976,
      "epoch": 0.4207477619799895,
      "grad_norm": 0.3321590721607208,
      "learning_rate": 0.0001995750472395658,
      "loss": 0.22508296370506287,
      "mean_token_accuracy": 0.9119736105203629,
      "num_tokens": 9856464.0,
      "step": 799
    },
    {
      "entropy": 1.2528738677501678,
      "epoch": 0.421274354923644,
      "grad_norm": 0.3493482768535614,
      "learning_rate": 0.00019957291616964565,
      "loss": 0.234869584441185,
      "mean_token_accuracy": 0.9126180708408356,
      "num_tokens": 9868800.0,
      "step": 800
    },
    {
      "entropy": 1.263050615787506,
      "epoch": 0.4218009478672986,
      "grad_norm": 0.33217278122901917,
      "learning_rate": 0.00019957077978230347,
      "loss": 0.22601178288459778,
      "mean_token_accuracy": 0.9101445376873016,
      "num_tokens": 9881136.0,
      "step": 801
    },
    {
      "entropy": 1.279297262430191,
      "epoch": 0.42232754081095314,
      "grad_norm": 0.3346133232116699,
      "learning_rate": 0.00019956863807766618,
      "loss": 0.22423239052295685,
      "mean_token_accuracy": 0.9134156703948975,
      "num_tokens": 9893472.0,
      "step": 802
    },
    {
      "entropy": 1.342239260673523,
      "epoch": 0.4228541337546077,
      "grad_norm": 0.32531458139419556,
      "learning_rate": 0.00019956649105586088,
      "loss": 0.23583942651748657,
      "mean_token_accuracy": 0.9080929756164551,
      "num_tokens": 9905808.0,
      "step": 803
    },
    {
      "entropy": 1.301451414823532,
      "epoch": 0.42338072669826227,
      "grad_norm": 0.3336106240749359,
      "learning_rate": 0.00019956433871701508,
      "loss": 0.23021520674228668,
      "mean_token_accuracy": 0.9132036864757538,
      "num_tokens": 9918144.0,
      "step": 804
    },
    {
      "entropy": 1.294978141784668,
      "epoch": 0.4239073196419168,
      "grad_norm": 0.36648303270339966,
      "learning_rate": 0.00019956218106125648,
      "loss": 0.216038778424263,
      "mean_token_accuracy": 0.9193701446056366,
      "num_tokens": 9930480.0,
      "step": 805
    },
    {
      "entropy": 1.3323890268802643,
      "epoch": 0.42443391258557134,
      "grad_norm": 0.5710987448692322,
      "learning_rate": 0.00019956001808871322,
      "loss": 0.21820521354675293,
      "mean_token_accuracy": 0.9192979484796524,
      "num_tokens": 9942816.0,
      "step": 806
    },
    {
      "entropy": 1.3704762756824493,
      "epoch": 0.42496050552922593,
      "grad_norm": 0.3389364778995514,
      "learning_rate": 0.00019955784979951366,
      "loss": 0.20125524699687958,
      "mean_token_accuracy": 0.9225403815507889,
      "num_tokens": 9955152.0,
      "step": 807
    },
    {
      "entropy": 1.361533671617508,
      "epoch": 0.42548709847288047,
      "grad_norm": 0.36687371134757996,
      "learning_rate": 0.00019955567619378653,
      "loss": 0.22097808122634888,
      "mean_token_accuracy": 0.9130203276872635,
      "num_tokens": 9967488.0,
      "step": 808
    },
    {
      "entropy": 1.3404709100723267,
      "epoch": 0.426013691416535,
      "grad_norm": 0.3457012474536896,
      "learning_rate": 0.00019955349727166088,
      "loss": 0.21268591284751892,
      "mean_token_accuracy": 0.9194072186946869,
      "num_tokens": 9979824.0,
      "step": 809
    },
    {
      "entropy": 1.3375678360462189,
      "epoch": 0.4265402843601896,
      "grad_norm": 0.3357281982898712,
      "learning_rate": 0.0001995513130332661,
      "loss": 0.21844780445098877,
      "mean_token_accuracy": 0.9134986102581024,
      "num_tokens": 9992160.0,
      "step": 810
    },
    {
      "entropy": 1.373364269733429,
      "epoch": 0.42706687730384413,
      "grad_norm": 0.3841971457004547,
      "learning_rate": 0.00019954912347873175,
      "loss": 0.2078622281551361,
      "mean_token_accuracy": 0.9202176630496979,
      "num_tokens": 10004496.0,
      "step": 811
    },
    {
      "entropy": 1.3305988311767578,
      "epoch": 0.42759347024749866,
      "grad_norm": 0.3852640986442566,
      "learning_rate": 0.0001995469286081879,
      "loss": 0.2230588048696518,
      "mean_token_accuracy": 0.9157514870166779,
      "num_tokens": 10016832.0,
      "step": 812
    },
    {
      "entropy": 1.3437947630882263,
      "epoch": 0.42812006319115326,
      "grad_norm": 0.390927791595459,
      "learning_rate": 0.0001995447284217648,
      "loss": 0.22504432499408722,
      "mean_token_accuracy": 0.911404550075531,
      "num_tokens": 10029168.0,
      "step": 813
    },
    {
      "entropy": 1.4292974174022675,
      "epoch": 0.4286466561348078,
      "grad_norm": 0.3649219274520874,
      "learning_rate": 0.00019954252291959313,
      "loss": 0.22576412558555603,
      "mean_token_accuracy": 0.9123170077800751,
      "num_tokens": 10041504.0,
      "step": 814
    },
    {
      "entropy": 1.3408824503421783,
      "epoch": 0.4291732490784623,
      "grad_norm": 0.4012393355369568,
      "learning_rate": 0.00019954031210180371,
      "loss": 0.2239188551902771,
      "mean_token_accuracy": 0.9098076522350311,
      "num_tokens": 10053840.0,
      "step": 815
    },
    {
      "entropy": 1.3955635130405426,
      "epoch": 0.4296998420221169,
      "grad_norm": 0.31125685572624207,
      "learning_rate": 0.00019953809596852786,
      "loss": 0.21482022106647491,
      "mean_token_accuracy": 0.9152773320674896,
      "num_tokens": 10066176.0,
      "step": 816
    },
    {
      "entropy": 1.372936338186264,
      "epoch": 0.43022643496577145,
      "grad_norm": 0.3509393632411957,
      "learning_rate": 0.00019953587451989712,
      "loss": 0.2275184839963913,
      "mean_token_accuracy": 0.9097452610731125,
      "num_tokens": 10078512.0,
      "step": 817
    },
    {
      "entropy": 1.2875484228134155,
      "epoch": 0.430753027909426,
      "grad_norm": 0.32160601019859314,
      "learning_rate": 0.00019953364775604336,
      "loss": 0.22542208433151245,
      "mean_token_accuracy": 0.9094801992177963,
      "num_tokens": 10090848.0,
      "step": 818
    },
    {
      "entropy": 1.387698233127594,
      "epoch": 0.4312796208530806,
      "grad_norm": 0.3713604807853699,
      "learning_rate": 0.00019953141567709878,
      "loss": 0.2560008466243744,
      "mean_token_accuracy": 0.9061192125082016,
      "num_tokens": 10103184.0,
      "step": 819
    },
    {
      "entropy": 1.3468024134635925,
      "epoch": 0.4318062137967351,
      "grad_norm": 0.30053895711898804,
      "learning_rate": 0.00019952917828319587,
      "loss": 0.21902483701705933,
      "mean_token_accuracy": 0.9183452129364014,
      "num_tokens": 10115520.0,
      "step": 820
    },
    {
      "entropy": 1.2196392714977264,
      "epoch": 0.43233280674038965,
      "grad_norm": 0.3135424256324768,
      "learning_rate": 0.00019952693557446748,
      "loss": 0.2191144973039627,
      "mean_token_accuracy": 0.9174018204212189,
      "num_tokens": 10127856.0,
      "step": 821
    },
    {
      "entropy": 1.341955155134201,
      "epoch": 0.43285939968404424,
      "grad_norm": 0.31046387553215027,
      "learning_rate": 0.00019952468755104672,
      "loss": 0.19744142889976501,
      "mean_token_accuracy": 0.9225213825702667,
      "num_tokens": 10140192.0,
      "step": 822
    },
    {
      "entropy": 1.2434613406658173,
      "epoch": 0.4333859926276988,
      "grad_norm": 0.3251499831676483,
      "learning_rate": 0.00019952243421306708,
      "loss": 0.22867801785469055,
      "mean_token_accuracy": 0.9065830409526825,
      "num_tokens": 10152528.0,
      "step": 823
    },
    {
      "entropy": 1.2479578852653503,
      "epoch": 0.43391258557135337,
      "grad_norm": 0.33372238278388977,
      "learning_rate": 0.0001995201755606622,
      "loss": 0.24028503894805908,
      "mean_token_accuracy": 0.9034052640199661,
      "num_tokens": 10164864.0,
      "step": 824
    },
    {
      "entropy": 1.2929136455059052,
      "epoch": 0.4344391785150079,
      "grad_norm": 0.3748015761375427,
      "learning_rate": 0.00019951791159396635,
      "loss": 0.22330275177955627,
      "mean_token_accuracy": 0.9107150882482529,
      "num_tokens": 10177200.0,
      "step": 825
    },
    {
      "entropy": 1.1717173159122467,
      "epoch": 0.43496577145866244,
      "grad_norm": 0.32337212562561035,
      "learning_rate": 0.00019951564231311382,
      "loss": 0.20805582404136658,
      "mean_token_accuracy": 0.9200474917888641,
      "num_tokens": 10189536.0,
      "step": 826
    },
    {
      "entropy": 1.2089639604091644,
      "epoch": 0.43549236440231703,
      "grad_norm": 0.3620331883430481,
      "learning_rate": 0.00019951336771823936,
      "loss": 0.24102292954921722,
      "mean_token_accuracy": 0.9084760844707489,
      "num_tokens": 10201872.0,
      "step": 827
    },
    {
      "entropy": 1.2690958082675934,
      "epoch": 0.43601895734597157,
      "grad_norm": 0.3436319828033447,
      "learning_rate": 0.00019951108780947793,
      "loss": 0.2396576702594757,
      "mean_token_accuracy": 0.9085531234741211,
      "num_tokens": 10214208.0,
      "step": 828
    },
    {
      "entropy": 1.1895755529403687,
      "epoch": 0.4365455502896261,
      "grad_norm": 0.3198072612285614,
      "learning_rate": 0.00019950880258696495,
      "loss": 0.21371476352214813,
      "mean_token_accuracy": 0.9214041531085968,
      "num_tokens": 10226544.0,
      "step": 829
    },
    {
      "entropy": 1.2306865453720093,
      "epoch": 0.4370721432332807,
      "grad_norm": 0.33646130561828613,
      "learning_rate": 0.00019950651205083607,
      "loss": 0.22665852308273315,
      "mean_token_accuracy": 0.9108365774154663,
      "num_tokens": 10238880.0,
      "step": 830
    },
    {
      "entropy": 1.2199313044548035,
      "epoch": 0.43759873617693523,
      "grad_norm": 0.2950493395328522,
      "learning_rate": 0.00019950421620122726,
      "loss": 0.19776280224323273,
      "mean_token_accuracy": 0.922358974814415,
      "num_tokens": 10251216.0,
      "step": 831
    },
    {
      "entropy": 1.2803155779838562,
      "epoch": 0.43812532912058977,
      "grad_norm": 0.3180263638496399,
      "learning_rate": 0.00019950191503827477,
      "loss": 0.2287098914384842,
      "mean_token_accuracy": 0.9091289192438126,
      "num_tokens": 10263552.0,
      "step": 832
    },
    {
      "entropy": 1.278805285692215,
      "epoch": 0.43865192206424436,
      "grad_norm": 0.30920109152793884,
      "learning_rate": 0.00019949960856211526,
      "loss": 0.22083017230033875,
      "mean_token_accuracy": 0.915079653263092,
      "num_tokens": 10275888.0,
      "step": 833
    },
    {
      "entropy": 1.255694329738617,
      "epoch": 0.4391785150078989,
      "grad_norm": 0.3112378716468811,
      "learning_rate": 0.00019949729677288568,
      "loss": 0.22883734107017517,
      "mean_token_accuracy": 0.9063798487186432,
      "num_tokens": 10288224.0,
      "step": 834
    },
    {
      "entropy": 1.2695356905460358,
      "epoch": 0.43970510795155343,
      "grad_norm": 0.36405643820762634,
      "learning_rate": 0.00019949497967072314,
      "loss": 0.25639811158180237,
      "mean_token_accuracy": 0.8983906358480453,
      "num_tokens": 10300560.0,
      "step": 835
    },
    {
      "entropy": 1.1852473616600037,
      "epoch": 0.440231700895208,
      "grad_norm": 0.3179355561733246,
      "learning_rate": 0.00019949265725576534,
      "loss": 0.2258102297782898,
      "mean_token_accuracy": 0.911256417632103,
      "num_tokens": 10312896.0,
      "step": 836
    },
    {
      "entropy": 1.2247830629348755,
      "epoch": 0.44075829383886256,
      "grad_norm": 0.30152472853660583,
      "learning_rate": 0.00019949032952815002,
      "loss": 0.22182312607765198,
      "mean_token_accuracy": 0.9133886694908142,
      "num_tokens": 10325232.0,
      "step": 837
    },
    {
      "entropy": 1.2358119487762451,
      "epoch": 0.4412848867825171,
      "grad_norm": 0.3537450134754181,
      "learning_rate": 0.00019948799648801546,
      "loss": 0.23890867829322815,
      "mean_token_accuracy": 0.9061432033777237,
      "num_tokens": 10337568.0,
      "step": 838
    },
    {
      "entropy": 1.255542904138565,
      "epoch": 0.4418114797261717,
      "grad_norm": 0.3185861110687256,
      "learning_rate": 0.00019948565813550012,
      "loss": 0.2323358654975891,
      "mean_token_accuracy": 0.910455733537674,
      "num_tokens": 10349904.0,
      "step": 839
    },
    {
      "entropy": 1.2060716450214386,
      "epoch": 0.4423380726698262,
      "grad_norm": 0.2943599820137024,
      "learning_rate": 0.00019948331447074282,
      "loss": 0.1973709762096405,
      "mean_token_accuracy": 0.9198663532733917,
      "num_tokens": 10362240.0,
      "step": 840
    },
    {
      "entropy": 1.23261758685112,
      "epoch": 0.44286466561348076,
      "grad_norm": 0.3526243269443512,
      "learning_rate": 0.00019948096549388269,
      "loss": 0.2500985562801361,
      "mean_token_accuracy": 0.9066997915506363,
      "num_tokens": 10374576.0,
      "step": 841
    },
    {
      "entropy": 1.237447440624237,
      "epoch": 0.44339125855713535,
      "grad_norm": 0.3191603422164917,
      "learning_rate": 0.00019947861120505914,
      "loss": 0.21682989597320557,
      "mean_token_accuracy": 0.914885938167572,
      "num_tokens": 10386912.0,
      "step": 842
    },
    {
      "entropy": 1.2385708093643188,
      "epoch": 0.4439178515007899,
      "grad_norm": 0.31065261363983154,
      "learning_rate": 0.000199476251604412,
      "loss": 0.22177095711231232,
      "mean_token_accuracy": 0.9108449369668961,
      "num_tokens": 10399248.0,
      "step": 843
    },
    {
      "entropy": 1.209405928850174,
      "epoch": 0.4444444444444444,
      "grad_norm": 0.3202918767929077,
      "learning_rate": 0.0001994738866920813,
      "loss": 0.23262163996696472,
      "mean_token_accuracy": 0.906676635146141,
      "num_tokens": 10411584.0,
      "step": 844
    },
    {
      "entropy": 1.2493412792682648,
      "epoch": 0.444971037388099,
      "grad_norm": 0.3377321660518646,
      "learning_rate": 0.00019947151646820745,
      "loss": 0.25681400299072266,
      "mean_token_accuracy": 0.9044294059276581,
      "num_tokens": 10423920.0,
      "step": 845
    },
    {
      "entropy": 1.2441616654396057,
      "epoch": 0.44549763033175355,
      "grad_norm": 0.3140577971935272,
      "learning_rate": 0.00019946914093293115,
      "loss": 0.23581147193908691,
      "mean_token_accuracy": 0.9107285887002945,
      "num_tokens": 10436256.0,
      "step": 846
    },
    {
      "entropy": 1.2595540583133698,
      "epoch": 0.4460242232754081,
      "grad_norm": 0.32932135462760925,
      "learning_rate": 0.00019946676008639343,
      "loss": 0.2249414175748825,
      "mean_token_accuracy": 0.9159797579050064,
      "num_tokens": 10448592.0,
      "step": 847
    },
    {
      "entropy": 1.2226338982582092,
      "epoch": 0.4465508162190627,
      "grad_norm": 0.3085028827190399,
      "learning_rate": 0.00019946437392873555,
      "loss": 0.22729279100894928,
      "mean_token_accuracy": 0.914769247174263,
      "num_tokens": 10460928.0,
      "step": 848
    },
    {
      "entropy": 1.2365905344486237,
      "epoch": 0.4470774091627172,
      "grad_norm": 0.33299946784973145,
      "learning_rate": 0.00019946198246009928,
      "loss": 0.23490425944328308,
      "mean_token_accuracy": 0.9082431495189667,
      "num_tokens": 10473264.0,
      "step": 849
    },
    {
      "entropy": 1.4061111807823181,
      "epoch": 0.4476040021063718,
      "grad_norm": 0.3621409833431244,
      "learning_rate": 0.00019945958568062656,
      "loss": 0.22818249464035034,
      "mean_token_accuracy": 0.9102335572242737,
      "num_tokens": 10485600.0,
      "step": 850
    },
    {
      "entropy": 1.2550583481788635,
      "epoch": 0.44813059505002634,
      "grad_norm": 0.33583348989486694,
      "learning_rate": 0.0001994571835904596,
      "loss": 0.2311660647392273,
      "mean_token_accuracy": 0.9055789560079575,
      "num_tokens": 10497936.0,
      "step": 851
    },
    {
      "entropy": 1.306765854358673,
      "epoch": 0.4486571879936809,
      "grad_norm": 0.34734323620796204,
      "learning_rate": 0.00019945477618974104,
      "loss": 0.24055415391921997,
      "mean_token_accuracy": 0.9080438315868378,
      "num_tokens": 10510272.0,
      "step": 852
    },
    {
      "entropy": 1.3021958768367767,
      "epoch": 0.44918378093733546,
      "grad_norm": 0.35933780670166016,
      "learning_rate": 0.00019945236347861383,
      "loss": 0.22972102463245392,
      "mean_token_accuracy": 0.9099899232387543,
      "num_tokens": 10522608.0,
      "step": 853
    },
    {
      "entropy": 1.2393949925899506,
      "epoch": 0.44971037388099,
      "grad_norm": 0.33583781123161316,
      "learning_rate": 0.00019944994545722115,
      "loss": 0.19537097215652466,
      "mean_token_accuracy": 0.9247924536466599,
      "num_tokens": 10534944.0,
      "step": 854
    },
    {
      "entropy": 1.2775700390338898,
      "epoch": 0.45023696682464454,
      "grad_norm": 0.33388838171958923,
      "learning_rate": 0.00019944752212570658,
      "loss": 0.20106476545333862,
      "mean_token_accuracy": 0.9226871877908707,
      "num_tokens": 10547280.0,
      "step": 855
    },
    {
      "entropy": 1.2714512348175049,
      "epoch": 0.4507635597682991,
      "grad_norm": 0.3518841862678528,
      "learning_rate": 0.00019944509348421394,
      "loss": 0.24091394245624542,
      "mean_token_accuracy": 0.9107283055782318,
      "num_tokens": 10559616.0,
      "step": 856
    },
    {
      "entropy": 1.3004609048366547,
      "epoch": 0.45129015271195366,
      "grad_norm": 0.31591442227363586,
      "learning_rate": 0.00019944265953288746,
      "loss": 0.22405283153057098,
      "mean_token_accuracy": 0.9140247255563736,
      "num_tokens": 10571952.0,
      "step": 857
    },
    {
      "entropy": 1.2649115324020386,
      "epoch": 0.4518167456556082,
      "grad_norm": 0.34270599484443665,
      "learning_rate": 0.00019944022027187157,
      "loss": 0.2242221236228943,
      "mean_token_accuracy": 0.9118231236934662,
      "num_tokens": 10584288.0,
      "step": 858
    },
    {
      "entropy": 1.20332869887352,
      "epoch": 0.4523433385992628,
      "grad_norm": 0.33847615122795105,
      "learning_rate": 0.0001994377757013111,
      "loss": 0.22582295536994934,
      "mean_token_accuracy": 0.9115125685930252,
      "num_tokens": 10596624.0,
      "step": 859
    },
    {
      "entropy": 1.2661672830581665,
      "epoch": 0.4528699315429173,
      "grad_norm": 0.34885501861572266,
      "learning_rate": 0.00019943532582135116,
      "loss": 0.2379041314125061,
      "mean_token_accuracy": 0.9074203372001648,
      "num_tokens": 10608960.0,
      "step": 860
    },
    {
      "entropy": 1.2014594674110413,
      "epoch": 0.45339652448657186,
      "grad_norm": 0.29767417907714844,
      "learning_rate": 0.0001994328706321372,
      "loss": 0.20825329422950745,
      "mean_token_accuracy": 0.9220836907625198,
      "num_tokens": 10621296.0,
      "step": 861
    },
    {
      "entropy": 1.1979358792304993,
      "epoch": 0.45392311743022645,
      "grad_norm": 0.33858057856559753,
      "learning_rate": 0.000199430410133815,
      "loss": 0.21688413619995117,
      "mean_token_accuracy": 0.9154618680477142,
      "num_tokens": 10633632.0,
      "step": 862
    },
    {
      "entropy": 1.1113716661930084,
      "epoch": 0.454449710373881,
      "grad_norm": 0.32101956009864807,
      "learning_rate": 0.00019942794432653053,
      "loss": 0.20722433924674988,
      "mean_token_accuracy": 0.9159047454595566,
      "num_tokens": 10645968.0,
      "step": 863
    },
    {
      "entropy": 1.1962774693965912,
      "epoch": 0.4549763033175355,
      "grad_norm": 0.34402385354042053,
      "learning_rate": 0.00019942547321043027,
      "loss": 0.2361735999584198,
      "mean_token_accuracy": 0.9091803878545761,
      "num_tokens": 10658304.0,
      "step": 864
    },
    {
      "entropy": 1.2290304005146027,
      "epoch": 0.4555028962611901,
      "grad_norm": 0.35236313939094543,
      "learning_rate": 0.00019942299678566086,
      "loss": 0.21090444922447205,
      "mean_token_accuracy": 0.9149055033922195,
      "num_tokens": 10670640.0,
      "step": 865
    },
    {
      "entropy": 1.1575745642185211,
      "epoch": 0.45602948920484465,
      "grad_norm": 0.34065473079681396,
      "learning_rate": 0.00019942051505236936,
      "loss": 0.2252214103937149,
      "mean_token_accuracy": 0.9156748950481415,
      "num_tokens": 10682976.0,
      "step": 866
    },
    {
      "entropy": 1.216041624546051,
      "epoch": 0.4565560821484992,
      "grad_norm": 0.34139883518218994,
      "learning_rate": 0.00019941802801070303,
      "loss": 0.2223672717809677,
      "mean_token_accuracy": 0.9103764742612839,
      "num_tokens": 10695312.0,
      "step": 867
    },
    {
      "entropy": 1.210105150938034,
      "epoch": 0.4570826750921538,
      "grad_norm": 0.33792680501937866,
      "learning_rate": 0.00019941553566080956,
      "loss": 0.21836510300636292,
      "mean_token_accuracy": 0.9169131815433502,
      "num_tokens": 10707648.0,
      "step": 868
    },
    {
      "entropy": 1.2123220264911652,
      "epoch": 0.4576092680358083,
      "grad_norm": 0.33836740255355835,
      "learning_rate": 0.0001994130380028369,
      "loss": 0.22877393662929535,
      "mean_token_accuracy": 0.9094105064868927,
      "num_tokens": 10719984.0,
      "step": 869
    },
    {
      "entropy": 1.2113963067531586,
      "epoch": 0.45813586097946285,
      "grad_norm": 0.3214447796344757,
      "learning_rate": 0.00019941053503693329,
      "loss": 0.22401154041290283,
      "mean_token_accuracy": 0.9101571440696716,
      "num_tokens": 10732320.0,
      "step": 870
    },
    {
      "entropy": 1.2003886699676514,
      "epoch": 0.45866245392311744,
      "grad_norm": 0.32083162665367126,
      "learning_rate": 0.00019940802676324734,
      "loss": 0.22345268726348877,
      "mean_token_accuracy": 0.909741148352623,
      "num_tokens": 10744656.0,
      "step": 871
    },
    {
      "entropy": 1.2443215548992157,
      "epoch": 0.459189046866772,
      "grad_norm": 0.32810378074645996,
      "learning_rate": 0.00019940551318192796,
      "loss": 0.22558192908763885,
      "mean_token_accuracy": 0.911412313580513,
      "num_tokens": 10756992.0,
      "step": 872
    },
    {
      "entropy": 1.2517528235912323,
      "epoch": 0.4597156398104265,
      "grad_norm": 0.2983804941177368,
      "learning_rate": 0.00019940299429312437,
      "loss": 0.22902357578277588,
      "mean_token_accuracy": 0.9084698557853699,
      "num_tokens": 10769328.0,
      "step": 873
    },
    {
      "entropy": 1.331667423248291,
      "epoch": 0.4602422327540811,
      "grad_norm": 0.33570387959480286,
      "learning_rate": 0.00019940047009698605,
      "loss": 0.21841317415237427,
      "mean_token_accuracy": 0.9151860028505325,
      "num_tokens": 10781664.0,
      "step": 874
    },
    {
      "entropy": 1.2408846318721771,
      "epoch": 0.46076882569773564,
      "grad_norm": 0.3269551992416382,
      "learning_rate": 0.00019939794059366294,
      "loss": 0.21816705167293549,
      "mean_token_accuracy": 0.9179367125034332,
      "num_tokens": 10794000.0,
      "step": 875
    },
    {
      "entropy": 1.2641226053237915,
      "epoch": 0.46129541864139023,
      "grad_norm": 0.31645506620407104,
      "learning_rate": 0.0001993954057833051,
      "loss": 0.22259601950645447,
      "mean_token_accuracy": 0.9116221219301224,
      "num_tokens": 10806336.0,
      "step": 876
    },
    {
      "entropy": 1.2864770293235779,
      "epoch": 0.46182201158504477,
      "grad_norm": 0.36219751834869385,
      "learning_rate": 0.0001993928656660631,
      "loss": 0.23597069084644318,
      "mean_token_accuracy": 0.9075523316860199,
      "num_tokens": 10818672.0,
      "step": 877
    },
    {
      "entropy": 1.284227877855301,
      "epoch": 0.4623486045286993,
      "grad_norm": 0.33684641122817993,
      "learning_rate": 0.00019939032024208762,
      "loss": 0.23688027262687683,
      "mean_token_accuracy": 0.90504090487957,
      "num_tokens": 10831008.0,
      "step": 878
    },
    {
      "entropy": 1.2863526940345764,
      "epoch": 0.4628751974723539,
      "grad_norm": 0.3741551637649536,
      "learning_rate": 0.0001993877695115299,
      "loss": 0.26197415590286255,
      "mean_token_accuracy": 0.8967071026563644,
      "num_tokens": 10843344.0,
      "step": 879
    },
    {
      "entropy": 1.3029287159442902,
      "epoch": 0.46340179041600843,
      "grad_norm": 0.33679425716400146,
      "learning_rate": 0.00019938521347454127,
      "loss": 0.2334819734096527,
      "mean_token_accuracy": 0.9123462438583374,
      "num_tokens": 10855680.0,
      "step": 880
    },
    {
      "entropy": 1.2353507280349731,
      "epoch": 0.46392838335966297,
      "grad_norm": 0.3251015543937683,
      "learning_rate": 0.00019938265213127344,
      "loss": 0.24127048254013062,
      "mean_token_accuracy": 0.9032625257968903,
      "num_tokens": 10868016.0,
      "step": 881
    },
    {
      "entropy": 1.2318426966667175,
      "epoch": 0.46445497630331756,
      "grad_norm": 0.3128981292247772,
      "learning_rate": 0.0001993800854818786,
      "loss": 0.23091557621955872,
      "mean_token_accuracy": 0.910790741443634,
      "num_tokens": 10880352.0,
      "step": 882
    },
    {
      "entropy": 1.1969189643859863,
      "epoch": 0.4649815692469721,
      "grad_norm": 0.2763598561286926,
      "learning_rate": 0.000199377513526509,
      "loss": 0.19723600149154663,
      "mean_token_accuracy": 0.9164901375770569,
      "num_tokens": 10892688.0,
      "step": 883
    },
    {
      "entropy": 1.3742352426052094,
      "epoch": 0.46550816219062663,
      "grad_norm": 0.36182069778442383,
      "learning_rate": 0.00019937493626531733,
      "loss": 0.23250778019428253,
      "mean_token_accuracy": 0.910800889134407,
      "num_tokens": 10905024.0,
      "step": 884
    },
    {
      "entropy": 1.287173181772232,
      "epoch": 0.4660347551342812,
      "grad_norm": 0.37185028195381165,
      "learning_rate": 0.00019937235369845666,
      "loss": 0.2612641453742981,
      "mean_token_accuracy": 0.8980638682842255,
      "num_tokens": 10917360.0,
      "step": 885
    },
    {
      "entropy": 1.256176471710205,
      "epoch": 0.46656134807793576,
      "grad_norm": 0.3193638026714325,
      "learning_rate": 0.00019936976582608023,
      "loss": 0.2168237864971161,
      "mean_token_accuracy": 0.9117200374603271,
      "num_tokens": 10929696.0,
      "step": 886
    },
    {
      "entropy": 1.3229191601276398,
      "epoch": 0.4670879410215903,
      "grad_norm": 0.3389144241809845,
      "learning_rate": 0.00019936717264834166,
      "loss": 0.22236081957817078,
      "mean_token_accuracy": 0.9158815294504166,
      "num_tokens": 10942032.0,
      "step": 887
    },
    {
      "entropy": 1.3444121778011322,
      "epoch": 0.4676145339652449,
      "grad_norm": 0.30865800380706787,
      "learning_rate": 0.00019936457416539497,
      "loss": 0.22498130798339844,
      "mean_token_accuracy": 0.9109358042478561,
      "num_tokens": 10954368.0,
      "step": 888
    },
    {
      "entropy": 1.3391337096691132,
      "epoch": 0.4681411269088994,
      "grad_norm": 0.34072619676589966,
      "learning_rate": 0.00019936197037739433,
      "loss": 0.25583821535110474,
      "mean_token_accuracy": 0.9030615091323853,
      "num_tokens": 10966704.0,
      "step": 889
    },
    {
      "entropy": 1.3236074447631836,
      "epoch": 0.46866771985255395,
      "grad_norm": 0.30708205699920654,
      "learning_rate": 0.00019935936128449437,
      "loss": 0.2282828986644745,
      "mean_token_accuracy": 0.9107722789049149,
      "num_tokens": 10979040.0,
      "step": 890
    },
    {
      "entropy": 1.3548866510391235,
      "epoch": 0.46919431279620855,
      "grad_norm": 0.3132648766040802,
      "learning_rate": 0.00019935674688684993,
      "loss": 0.21266606450080872,
      "mean_token_accuracy": 0.9191526621580124,
      "num_tokens": 10991376.0,
      "step": 891
    },
    {
      "entropy": 1.3176174461841583,
      "epoch": 0.4697209057398631,
      "grad_norm": 0.2955908477306366,
      "learning_rate": 0.00019935412718461625,
      "loss": 0.21996596455574036,
      "mean_token_accuracy": 0.9133886098861694,
      "num_tokens": 11003712.0,
      "step": 892
    },
    {
      "entropy": 1.3978124856948853,
      "epoch": 0.4702474986835176,
      "grad_norm": 0.29386526346206665,
      "learning_rate": 0.00019935150217794887,
      "loss": 0.21586540341377258,
      "mean_token_accuracy": 0.913035660982132,
      "num_tokens": 11016048.0,
      "step": 893
    },
    {
      "entropy": 1.3344171047210693,
      "epoch": 0.4707740916271722,
      "grad_norm": 0.307434618473053,
      "learning_rate": 0.00019934887186700352,
      "loss": 0.19684919714927673,
      "mean_token_accuracy": 0.9249100238084793,
      "num_tokens": 11028384.0,
      "step": 894
    },
    {
      "entropy": 1.2738686203956604,
      "epoch": 0.47130068457082674,
      "grad_norm": 0.30784371495246887,
      "learning_rate": 0.00019934623625193646,
      "loss": 0.23357078433036804,
      "mean_token_accuracy": 0.906443789601326,
      "num_tokens": 11040720.0,
      "step": 895
    },
    {
      "entropy": 1.3004032373428345,
      "epoch": 0.4718272775144813,
      "grad_norm": 0.3604377806186676,
      "learning_rate": 0.0001993435953329041,
      "loss": 0.23538318276405334,
      "mean_token_accuracy": 0.910987451672554,
      "num_tokens": 11053056.0,
      "step": 896
    },
    {
      "entropy": 1.349770426750183,
      "epoch": 0.47235387045813587,
      "grad_norm": 0.371415913105011,
      "learning_rate": 0.00019934094911006322,
      "loss": 0.23594705760478973,
      "mean_token_accuracy": 0.9082933962345123,
      "num_tokens": 11065392.0,
      "step": 897
    },
    {
      "entropy": 1.3280587792396545,
      "epoch": 0.4728804634017904,
      "grad_norm": 0.3518844544887543,
      "learning_rate": 0.0001993382975835709,
      "loss": 0.22077634930610657,
      "mean_token_accuracy": 0.9151550680398941,
      "num_tokens": 11077728.0,
      "step": 898
    },
    {
      "entropy": 1.2467398941516876,
      "epoch": 0.473407056345445,
      "grad_norm": 0.3485780358314514,
      "learning_rate": 0.00019933564075358455,
      "loss": 0.21681727468967438,
      "mean_token_accuracy": 0.917348250746727,
      "num_tokens": 11090064.0,
      "step": 899
    },
    {
      "entropy": 1.2497498095035553,
      "epoch": 0.47393364928909953,
      "grad_norm": 0.2795222997665405,
      "learning_rate": 0.0001993329786202619,
      "loss": 0.19592691957950592,
      "mean_token_accuracy": 0.92547307908535,
      "num_tokens": 11102400.0,
      "step": 900
    },
    {
      "entropy": 1.3175421953201294,
      "epoch": 0.47446024223275407,
      "grad_norm": 0.34323444962501526,
      "learning_rate": 0.00019933031118376097,
      "loss": 0.23341751098632812,
      "mean_token_accuracy": 0.9069218635559082,
      "num_tokens": 11114736.0,
      "step": 901
    },
    {
      "entropy": 1.2457618713378906,
      "epoch": 0.47498683517640866,
      "grad_norm": 0.3164939284324646,
      "learning_rate": 0.00019932763844424017,
      "loss": 0.2065175175666809,
      "mean_token_accuracy": 0.9177182614803314,
      "num_tokens": 11127072.0,
      "step": 902
    },
    {
      "entropy": 1.2637942731380463,
      "epoch": 0.4755134281200632,
      "grad_norm": 0.3504999279975891,
      "learning_rate": 0.00019932496040185808,
      "loss": 0.21458998322486877,
      "mean_token_accuracy": 0.9221720099449158,
      "num_tokens": 11139408.0,
      "step": 903
    },
    {
      "entropy": 1.2935825884342194,
      "epoch": 0.47604002106371773,
      "grad_norm": 0.37840691208839417,
      "learning_rate": 0.00019932227705677372,
      "loss": 0.227766752243042,
      "mean_token_accuracy": 0.9137209206819534,
      "num_tokens": 11151744.0,
      "step": 904
    },
    {
      "entropy": 1.3187288045883179,
      "epoch": 0.4765666140073723,
      "grad_norm": 0.35070836544036865,
      "learning_rate": 0.00019931958840914635,
      "loss": 0.23313553631305695,
      "mean_token_accuracy": 0.9072595983743668,
      "num_tokens": 11164080.0,
      "step": 905
    },
    {
      "entropy": 1.2109759449958801,
      "epoch": 0.47709320695102686,
      "grad_norm": 0.3268450200557709,
      "learning_rate": 0.00019931689445913564,
      "loss": 0.21801263093948364,
      "mean_token_accuracy": 0.9163070619106293,
      "num_tokens": 11176416.0,
      "step": 906
    },
    {
      "entropy": 1.2274765968322754,
      "epoch": 0.4776197998946814,
      "grad_norm": 0.3167557120323181,
      "learning_rate": 0.00019931419520690147,
      "loss": 0.21107861399650574,
      "mean_token_accuracy": 0.9160606116056442,
      "num_tokens": 11188752.0,
      "step": 907
    },
    {
      "entropy": 1.1903901398181915,
      "epoch": 0.478146392838336,
      "grad_norm": 0.3267144560813904,
      "learning_rate": 0.00019931149065260413,
      "loss": 0.20752331614494324,
      "mean_token_accuracy": 0.9146323353052139,
      "num_tokens": 11201088.0,
      "step": 908
    },
    {
      "entropy": 1.2301645576953888,
      "epoch": 0.4786729857819905,
      "grad_norm": 0.3323666751384735,
      "learning_rate": 0.0001993087807964041,
      "loss": 0.20132845640182495,
      "mean_token_accuracy": 0.9229601621627808,
      "num_tokens": 11213424.0,
      "step": 909
    },
    {
      "entropy": 1.2413881123065948,
      "epoch": 0.47919957872564506,
      "grad_norm": 0.3245013356208801,
      "learning_rate": 0.00019930606563846234,
      "loss": 0.19781675934791565,
      "mean_token_accuracy": 0.9200431853532791,
      "num_tokens": 11225760.0,
      "step": 910
    },
    {
      "entropy": 1.180581659078598,
      "epoch": 0.47972617166929965,
      "grad_norm": 0.31535807251930237,
      "learning_rate": 0.00019930334517893991,
      "loss": 0.2060268521308899,
      "mean_token_accuracy": 0.9165100157260895,
      "num_tokens": 11238096.0,
      "step": 911
    },
    {
      "entropy": 1.233038753271103,
      "epoch": 0.4802527646129542,
      "grad_norm": 0.32888907194137573,
      "learning_rate": 0.0001993006194179984,
      "loss": 0.22726945579051971,
      "mean_token_accuracy": 0.9112953692674637,
      "num_tokens": 11250432.0,
      "step": 912
    },
    {
      "entropy": 1.262090653181076,
      "epoch": 0.4807793575566087,
      "grad_norm": 0.3833845555782318,
      "learning_rate": 0.00019929788835579962,
      "loss": 0.23385398089885712,
      "mean_token_accuracy": 0.9076380431652069,
      "num_tokens": 11262768.0,
      "step": 913
    },
    {
      "entropy": 1.2252440452575684,
      "epoch": 0.4813059505002633,
      "grad_norm": 0.3054254949092865,
      "learning_rate": 0.00019929515199250565,
      "loss": 0.21175538003444672,
      "mean_token_accuracy": 0.91593237221241,
      "num_tokens": 11275104.0,
      "step": 914
    },
    {
      "entropy": 1.2347098588943481,
      "epoch": 0.48183254344391785,
      "grad_norm": 0.32731181383132935,
      "learning_rate": 0.00019929241032827898,
      "loss": 0.21469083428382874,
      "mean_token_accuracy": 0.9144894182682037,
      "num_tokens": 11287440.0,
      "step": 915
    },
    {
      "entropy": 1.2054485082626343,
      "epoch": 0.4823591363875724,
      "grad_norm": 0.3010466694831848,
      "learning_rate": 0.0001992896633632823,
      "loss": 0.2065753936767578,
      "mean_token_accuracy": 0.9242510497570038,
      "num_tokens": 11299776.0,
      "step": 916
    },
    {
      "entropy": 1.2271296083927155,
      "epoch": 0.482885729331227,
      "grad_norm": 0.42679160833358765,
      "learning_rate": 0.00019928691109767876,
      "loss": 0.23648393154144287,
      "mean_token_accuracy": 0.9071210771799088,
      "num_tokens": 11312112.0,
      "step": 917
    },
    {
      "entropy": 1.233128011226654,
      "epoch": 0.4834123222748815,
      "grad_norm": 0.3018331229686737,
      "learning_rate": 0.00019928415353163173,
      "loss": 0.20864740014076233,
      "mean_token_accuracy": 0.9222661405801773,
      "num_tokens": 11324448.0,
      "step": 918
    },
    {
      "entropy": 1.2823042571544647,
      "epoch": 0.48393891521853605,
      "grad_norm": 0.44017431139945984,
      "learning_rate": 0.00019928139066530487,
      "loss": 0.23257723450660706,
      "mean_token_accuracy": 0.9054065048694611,
      "num_tokens": 11336784.0,
      "step": 919
    },
    {
      "entropy": 1.2760326564311981,
      "epoch": 0.48446550816219064,
      "grad_norm": 0.3077127933502197,
      "learning_rate": 0.00019927862249886222,
      "loss": 0.22436192631721497,
      "mean_token_accuracy": 0.9136722385883331,
      "num_tokens": 11349120.0,
      "step": 920
    },
    {
      "entropy": 1.3157328069210052,
      "epoch": 0.4849921011058452,
      "grad_norm": 0.29968273639678955,
      "learning_rate": 0.0001992758490324681,
      "loss": 0.20738902688026428,
      "mean_token_accuracy": 0.9168381989002228,
      "num_tokens": 11361456.0,
      "step": 921
    },
    {
      "entropy": 1.3066685795783997,
      "epoch": 0.4855186940494997,
      "grad_norm": 0.31207534670829773,
      "learning_rate": 0.00019927307026628715,
      "loss": 0.2138848602771759,
      "mean_token_accuracy": 0.9126945734024048,
      "num_tokens": 11373792.0,
      "step": 922
    },
    {
      "entropy": 1.3030335009098053,
      "epoch": 0.4860452869931543,
      "grad_norm": 0.3280918598175049,
      "learning_rate": 0.00019927028620048438,
      "loss": 0.23367798328399658,
      "mean_token_accuracy": 0.9077742546796799,
      "num_tokens": 11386128.0,
      "step": 923
    },
    {
      "entropy": 1.2811558842658997,
      "epoch": 0.48657187993680884,
      "grad_norm": 0.376886248588562,
      "learning_rate": 0.00019926749683522502,
      "loss": 0.20860403776168823,
      "mean_token_accuracy": 0.9099331349134445,
      "num_tokens": 11398464.0,
      "step": 924
    },
    {
      "entropy": 1.2956900000572205,
      "epoch": 0.48709847288046343,
      "grad_norm": 0.2849150598049164,
      "learning_rate": 0.00019926470217067463,
      "loss": 0.20377135276794434,
      "mean_token_accuracy": 0.9249068945646286,
      "num_tokens": 11410800.0,
      "step": 925
    },
    {
      "entropy": 1.3335639834403992,
      "epoch": 0.48762506582411796,
      "grad_norm": 0.30451253056526184,
      "learning_rate": 0.00019926190220699916,
      "loss": 0.2137700468301773,
      "mean_token_accuracy": 0.9165246337652206,
      "num_tokens": 11423136.0,
      "step": 926
    },
    {
      "entropy": 1.3315201103687286,
      "epoch": 0.4881516587677725,
      "grad_norm": 0.31507113575935364,
      "learning_rate": 0.0001992590969443648,
      "loss": 0.22226491570472717,
      "mean_token_accuracy": 0.9110458046197891,
      "num_tokens": 11435472.0,
      "step": 927
    },
    {
      "entropy": 1.3439573049545288,
      "epoch": 0.4886782517114271,
      "grad_norm": 0.3123377561569214,
      "learning_rate": 0.00019925628638293815,
      "loss": 0.2124030739068985,
      "mean_token_accuracy": 0.9143670797348022,
      "num_tokens": 11447808.0,
      "step": 928
    },
    {
      "entropy": 1.373394101858139,
      "epoch": 0.4892048446550816,
      "grad_norm": 0.38396134972572327,
      "learning_rate": 0.00019925347052288594,
      "loss": 0.21086710691452026,
      "mean_token_accuracy": 0.9174065738916397,
      "num_tokens": 11460144.0,
      "step": 929
    },
    {
      "entropy": 1.3038064241409302,
      "epoch": 0.48973143759873616,
      "grad_norm": 0.3282279074192047,
      "learning_rate": 0.00019925064936437544,
      "loss": 0.19842374324798584,
      "mean_token_accuracy": 0.9203700721263885,
      "num_tokens": 11472480.0,
      "step": 930
    },
    {
      "entropy": 1.39743971824646,
      "epoch": 0.49025803054239075,
      "grad_norm": 0.3458907902240753,
      "learning_rate": 0.00019924782290757403,
      "loss": 0.21870574355125427,
      "mean_token_accuracy": 0.9133100211620331,
      "num_tokens": 11484816.0,
      "step": 931
    },
    {
      "entropy": 1.3737529814243317,
      "epoch": 0.4907846234860453,
      "grad_norm": 0.35361558198928833,
      "learning_rate": 0.0001992449911526496,
      "loss": 0.2228592187166214,
      "mean_token_accuracy": 0.908146470785141,
      "num_tokens": 11497152.0,
      "step": 932
    },
    {
      "entropy": 1.373944640159607,
      "epoch": 0.4913112164296998,
      "grad_norm": 0.3263792097568512,
      "learning_rate": 0.0001992421540997702,
      "loss": 0.19889512658119202,
      "mean_token_accuracy": 0.9241671562194824,
      "num_tokens": 11509488.0,
      "step": 933
    },
    {
      "entropy": 1.3812199532985687,
      "epoch": 0.4918378093733544,
      "grad_norm": 0.30469757318496704,
      "learning_rate": 0.00019923931174910421,
      "loss": 0.20764592289924622,
      "mean_token_accuracy": 0.9166074395179749,
      "num_tokens": 11521824.0,
      "step": 934
    },
    {
      "entropy": 1.3702083230018616,
      "epoch": 0.49236440231700895,
      "grad_norm": 0.3094841539859772,
      "learning_rate": 0.00019923646410082045,
      "loss": 0.22583043575286865,
      "mean_token_accuracy": 0.9111629873514175,
      "num_tokens": 11534160.0,
      "step": 935
    },
    {
      "entropy": 1.4550175070762634,
      "epoch": 0.4928909952606635,
      "grad_norm": 0.3479791581630707,
      "learning_rate": 0.0001992336111550879,
      "loss": 0.22404024004936218,
      "mean_token_accuracy": 0.9125335812568665,
      "num_tokens": 11546496.0,
      "step": 936
    },
    {
      "entropy": 1.369890809059143,
      "epoch": 0.4934175882043181,
      "grad_norm": 0.33549702167510986,
      "learning_rate": 0.00019923075291207595,
      "loss": 0.22361691296100616,
      "mean_token_accuracy": 0.9108781665563583,
      "num_tokens": 11558832.0,
      "step": 937
    },
    {
      "entropy": 1.4506123065948486,
      "epoch": 0.4939441811479726,
      "grad_norm": 0.31795117259025574,
      "learning_rate": 0.00019922788937195432,
      "loss": 0.20908354222774506,
      "mean_token_accuracy": 0.916689082980156,
      "num_tokens": 11571168.0,
      "step": 938
    },
    {
      "entropy": 1.4594894349575043,
      "epoch": 0.49447077409162715,
      "grad_norm": 0.3958785831928253,
      "learning_rate": 0.0001992250205348929,
      "loss": 0.22291752696037292,
      "mean_token_accuracy": 0.9057322442531586,
      "num_tokens": 11583504.0,
      "step": 939
    },
    {
      "entropy": 1.4367193281650543,
      "epoch": 0.49499736703528174,
      "grad_norm": 0.38530445098876953,
      "learning_rate": 0.00019922214640106207,
      "loss": 0.21757692098617554,
      "mean_token_accuracy": 0.9140817075967789,
      "num_tokens": 11595840.0,
      "step": 940
    },
    {
      "entropy": 1.4827219545841217,
      "epoch": 0.4955239599789363,
      "grad_norm": 0.36136746406555176,
      "learning_rate": 0.00019921926697063244,
      "loss": 0.245650514960289,
      "mean_token_accuracy": 0.9077811390161514,
      "num_tokens": 11608176.0,
      "step": 941
    },
    {
      "entropy": 1.4725195467472076,
      "epoch": 0.4960505529225908,
      "grad_norm": 0.3393801748752594,
      "learning_rate": 0.00019921638224377493,
      "loss": 0.22296787798404694,
      "mean_token_accuracy": 0.912161722779274,
      "num_tokens": 11620512.0,
      "step": 942
    },
    {
      "entropy": 1.5998758971691132,
      "epoch": 0.4965771458662454,
      "grad_norm": 0.31578320264816284,
      "learning_rate": 0.00019921349222066083,
      "loss": 0.21946173906326294,
      "mean_token_accuracy": 0.9090907126665115,
      "num_tokens": 11632848.0,
      "step": 943
    },
    {
      "entropy": 1.5493124723434448,
      "epoch": 0.49710373880989994,
      "grad_norm": 0.3594050705432892,
      "learning_rate": 0.00019921059690146165,
      "loss": 0.25090494751930237,
      "mean_token_accuracy": 0.899459958076477,
      "num_tokens": 11645184.0,
      "step": 944
    },
    {
      "entropy": 1.4591304063796997,
      "epoch": 0.4976303317535545,
      "grad_norm": 0.3398474454879761,
      "learning_rate": 0.0001992076962863493,
      "loss": 0.22868260741233826,
      "mean_token_accuracy": 0.9120910316705704,
      "num_tokens": 11657520.0,
      "step": 945
    },
    {
      "entropy": 1.5574268996715546,
      "epoch": 0.49815692469720907,
      "grad_norm": 0.37178367376327515,
      "learning_rate": 0.00019920479037549595,
      "loss": 0.24399921298027039,
      "mean_token_accuracy": 0.9063628762960434,
      "num_tokens": 11669856.0,
      "step": 946
    },
    {
      "entropy": 1.4919456243515015,
      "epoch": 0.4986835176408636,
      "grad_norm": 0.344330757856369,
      "learning_rate": 0.0001992018791690741,
      "loss": 0.2234281301498413,
      "mean_token_accuracy": 0.9139427691698074,
      "num_tokens": 11682192.0,
      "step": 947
    },
    {
      "entropy": 1.486481934785843,
      "epoch": 0.49921011058451814,
      "grad_norm": 0.3283190429210663,
      "learning_rate": 0.00019919896266725663,
      "loss": 0.21686026453971863,
      "mean_token_accuracy": 0.9156758487224579,
      "num_tokens": 11694528.0,
      "step": 948
    },
    {
      "entropy": 1.4495719075202942,
      "epoch": 0.49973670352817273,
      "grad_norm": 0.3260519802570343,
      "learning_rate": 0.00019919604087021664,
      "loss": 0.21722926199436188,
      "mean_token_accuracy": 0.9189446866512299,
      "num_tokens": 11706864.0,
      "step": 949
    },
    {
      "entropy": 1.5058691203594208,
      "epoch": 0.5002632964718273,
      "grad_norm": 0.34286144375801086,
      "learning_rate": 0.00019919311377812756,
      "loss": 0.2165110558271408,
      "mean_token_accuracy": 0.9119018763303757,
      "num_tokens": 11719200.0,
      "step": 950
    },
    {
      "entropy": 1.4864198863506317,
      "epoch": 0.5007898894154819,
      "grad_norm": 0.3339075446128845,
      "learning_rate": 0.00019919018139116318,
      "loss": 0.22437581419944763,
      "mean_token_accuracy": 0.9123266935348511,
      "num_tokens": 11731536.0,
      "step": 951
    },
    {
      "entropy": 1.5294865667819977,
      "epoch": 0.5013164823591364,
      "grad_norm": 0.3746489882469177,
      "learning_rate": 0.00019918724370949754,
      "loss": 0.2426229566335678,
      "mean_token_accuracy": 0.9112533032894135,
      "num_tokens": 11743872.0,
      "step": 952
    },
    {
      "entropy": 1.4867128133773804,
      "epoch": 0.5018430753027909,
      "grad_norm": 0.3308849036693573,
      "learning_rate": 0.00019918430073330513,
      "loss": 0.2087199091911316,
      "mean_token_accuracy": 0.919797733426094,
      "num_tokens": 11756208.0,
      "step": 953
    },
    {
      "entropy": 1.5136871337890625,
      "epoch": 0.5023696682464455,
      "grad_norm": 0.34608909487724304,
      "learning_rate": 0.00019918135246276052,
      "loss": 0.22718098759651184,
      "mean_token_accuracy": 0.9114474952220917,
      "num_tokens": 11768544.0,
      "step": 954
    },
    {
      "entropy": 1.5160588324069977,
      "epoch": 0.5028962611901,
      "grad_norm": 0.3665941655635834,
      "learning_rate": 0.00019917839889803884,
      "loss": 0.21780604124069214,
      "mean_token_accuracy": 0.9177397042512894,
      "num_tokens": 11780880.0,
      "step": 955
    },
    {
      "entropy": 1.4711733162403107,
      "epoch": 0.5034228541337546,
      "grad_norm": 0.3232526183128357,
      "learning_rate": 0.00019917544003931537,
      "loss": 0.20783714950084686,
      "mean_token_accuracy": 0.9170488715171814,
      "num_tokens": 11793216.0,
      "step": 956
    },
    {
      "entropy": 1.5061227977275848,
      "epoch": 0.5039494470774092,
      "grad_norm": 0.3460804522037506,
      "learning_rate": 0.00019917247588676582,
      "loss": 0.22913217544555664,
      "mean_token_accuracy": 0.9122664779424667,
      "num_tokens": 11805552.0,
      "step": 957
    },
    {
      "entropy": 1.4881491363048553,
      "epoch": 0.5044760400210637,
      "grad_norm": 0.38591158390045166,
      "learning_rate": 0.00019916950644056607,
      "loss": 0.2197885662317276,
      "mean_token_accuracy": 0.9077084809541702,
      "num_tokens": 11817888.0,
      "step": 958
    },
    {
      "entropy": 1.4673264026641846,
      "epoch": 0.5050026329647183,
      "grad_norm": 0.33065465092658997,
      "learning_rate": 0.00019916653170089246,
      "loss": 0.2319866120815277,
      "mean_token_accuracy": 0.9126220494508743,
      "num_tokens": 11830224.0,
      "step": 959
    },
    {
      "entropy": 1.4641340970993042,
      "epoch": 0.5055292259083728,
      "grad_norm": 0.36219480633735657,
      "learning_rate": 0.00019916355166792155,
      "loss": 0.23807652294635773,
      "mean_token_accuracy": 0.9052844792604446,
      "num_tokens": 11842560.0,
      "step": 960
    },
    {
      "entropy": 1.4770829379558563,
      "epoch": 0.5060558188520273,
      "grad_norm": 0.3402878940105438,
      "learning_rate": 0.00019916056634183027,
      "loss": 0.23622629046440125,
      "mean_token_accuracy": 0.9105163961648941,
      "num_tokens": 11854896.0,
      "step": 961
    },
    {
      "entropy": 1.510101705789566,
      "epoch": 0.506582411795682,
      "grad_norm": 0.38479122519493103,
      "learning_rate": 0.00019915757572279584,
      "loss": 0.26114726066589355,
      "mean_token_accuracy": 0.901023805141449,
      "num_tokens": 11867232.0,
      "step": 962
    },
    {
      "entropy": 1.462478369474411,
      "epoch": 0.5071090047393365,
      "grad_norm": 0.30232900381088257,
      "learning_rate": 0.0001991545798109958,
      "loss": 0.21362073719501495,
      "mean_token_accuracy": 0.9150588363409042,
      "num_tokens": 11879568.0,
      "step": 963
    },
    {
      "entropy": 1.4567637145519257,
      "epoch": 0.507635597682991,
      "grad_norm": 0.3392339050769806,
      "learning_rate": 0.00019915157860660797,
      "loss": 0.229543536901474,
      "mean_token_accuracy": 0.9123556017875671,
      "num_tokens": 11891904.0,
      "step": 964
    },
    {
      "entropy": 1.4373153448104858,
      "epoch": 0.5081621906266456,
      "grad_norm": 0.3237869441509247,
      "learning_rate": 0.00019914857210981052,
      "loss": 0.222116619348526,
      "mean_token_accuracy": 0.9098070710897446,
      "num_tokens": 11904240.0,
      "step": 965
    },
    {
      "entropy": 1.4922161400318146,
      "epoch": 0.5086887835703001,
      "grad_norm": 0.3090606927871704,
      "learning_rate": 0.00019914556032078198,
      "loss": 0.22452594339847565,
      "mean_token_accuracy": 0.9133640974760056,
      "num_tokens": 11916576.0,
      "step": 966
    },
    {
      "entropy": 1.4847496449947357,
      "epoch": 0.5092153765139548,
      "grad_norm": 0.31530138850212097,
      "learning_rate": 0.00019914254323970108,
      "loss": 0.21454353630542755,
      "mean_token_accuracy": 0.9175532162189484,
      "num_tokens": 11928912.0,
      "step": 967
    },
    {
      "entropy": 1.481870412826538,
      "epoch": 0.5097419694576093,
      "grad_norm": 0.3098122775554657,
      "learning_rate": 0.00019913952086674696,
      "loss": 0.21168866753578186,
      "mean_token_accuracy": 0.9122080057859421,
      "num_tokens": 11941248.0,
      "step": 968
    },
    {
      "entropy": 1.5705283880233765,
      "epoch": 0.5102685624012638,
      "grad_norm": 0.32009753584861755,
      "learning_rate": 0.000199136493202099,
      "loss": 0.20256565511226654,
      "mean_token_accuracy": 0.9237568825483322,
      "num_tokens": 11953584.0,
      "step": 969
    },
    {
      "entropy": 1.5321417450904846,
      "epoch": 0.5107951553449184,
      "grad_norm": 0.33614546060562134,
      "learning_rate": 0.000199133460245937,
      "loss": 0.2249448001384735,
      "mean_token_accuracy": 0.9105825424194336,
      "num_tokens": 11965920.0,
      "step": 970
    },
    {
      "entropy": 1.5060743689537048,
      "epoch": 0.5113217482885729,
      "grad_norm": 0.3201083540916443,
      "learning_rate": 0.00019913042199844097,
      "loss": 0.22582277655601501,
      "mean_token_accuracy": 0.9086000770330429,
      "num_tokens": 11978256.0,
      "step": 971
    },
    {
      "entropy": 1.5098620057106018,
      "epoch": 0.5118483412322274,
      "grad_norm": 0.739696204662323,
      "learning_rate": 0.00019912737845979126,
      "loss": 0.20600268244743347,
      "mean_token_accuracy": 0.9227887839078903,
      "num_tokens": 11990592.0,
      "step": 972
    },
    {
      "entropy": 1.5500882267951965,
      "epoch": 0.5123749341758821,
      "grad_norm": 0.3458639681339264,
      "learning_rate": 0.00019912432963016855,
      "loss": 0.23240959644317627,
      "mean_token_accuracy": 0.9070769846439362,
      "num_tokens": 12002928.0,
      "step": 973
    },
    {
      "entropy": 1.5514563918113708,
      "epoch": 0.5129015271195366,
      "grad_norm": 0.33823850750923157,
      "learning_rate": 0.0001991212755097539,
      "loss": 0.21927812695503235,
      "mean_token_accuracy": 0.9085335284471512,
      "num_tokens": 12015264.0,
      "step": 974
    },
    {
      "entropy": 1.4905177056789398,
      "epoch": 0.5134281200631912,
      "grad_norm": 0.3308093547821045,
      "learning_rate": 0.0001991182160987285,
      "loss": 0.22249096632003784,
      "mean_token_accuracy": 0.9114507138729095,
      "num_tokens": 12027600.0,
      "step": 975
    },
    {
      "entropy": 1.5092821419239044,
      "epoch": 0.5139547130068457,
      "grad_norm": 0.35571059584617615,
      "learning_rate": 0.0001991151513972741,
      "loss": 0.23242917656898499,
      "mean_token_accuracy": 0.9118809998035431,
      "num_tokens": 12039936.0,
      "step": 976
    },
    {
      "entropy": 1.426015704870224,
      "epoch": 0.5144813059505002,
      "grad_norm": 0.2942444086074829,
      "learning_rate": 0.0001991120814055725,
      "loss": 0.2060171216726303,
      "mean_token_accuracy": 0.9149444848299026,
      "num_tokens": 12052272.0,
      "step": 977
    },
    {
      "entropy": 1.4646040201187134,
      "epoch": 0.5150078988941548,
      "grad_norm": 0.36998897790908813,
      "learning_rate": 0.00019910900612380605,
      "loss": 0.23340070247650146,
      "mean_token_accuracy": 0.9063846617937088,
      "num_tokens": 12064608.0,
      "step": 978
    },
    {
      "entropy": 1.4659819900989532,
      "epoch": 0.5155344918378094,
      "grad_norm": 0.3255058526992798,
      "learning_rate": 0.00019910592555215725,
      "loss": 0.20911140739917755,
      "mean_token_accuracy": 0.9192948192358017,
      "num_tokens": 12076944.0,
      "step": 979
    },
    {
      "entropy": 1.5005815625190735,
      "epoch": 0.516061084781464,
      "grad_norm": 0.35735374689102173,
      "learning_rate": 0.000199102839690809,
      "loss": 0.2287176251411438,
      "mean_token_accuracy": 0.9081791937351227,
      "num_tokens": 12089280.0,
      "step": 980
    },
    {
      "entropy": 1.4197439849376678,
      "epoch": 0.5165876777251185,
      "grad_norm": 0.30643951892852783,
      "learning_rate": 0.00019909974853994452,
      "loss": 0.19968025386333466,
      "mean_token_accuracy": 0.9234030842781067,
      "num_tokens": 12101616.0,
      "step": 981
    },
    {
      "entropy": 1.4192424714565277,
      "epoch": 0.517114270668773,
      "grad_norm": 0.337433785200119,
      "learning_rate": 0.00019909665209974723,
      "loss": 0.23955072462558746,
      "mean_token_accuracy": 0.9109057635068893,
      "num_tokens": 12113952.0,
      "step": 982
    },
    {
      "entropy": 1.4805451035499573,
      "epoch": 0.5176408636124276,
      "grad_norm": 0.34595102071762085,
      "learning_rate": 0.00019909355037040104,
      "loss": 0.23975268006324768,
      "mean_token_accuracy": 0.9099203199148178,
      "num_tokens": 12126288.0,
      "step": 983
    },
    {
      "entropy": 1.4859175980091095,
      "epoch": 0.5181674565560821,
      "grad_norm": 0.33112627267837524,
      "learning_rate": 0.00019909044335209003,
      "loss": 0.23115341365337372,
      "mean_token_accuracy": 0.9144714772701263,
      "num_tokens": 12138624.0,
      "step": 984
    },
    {
      "entropy": 1.4870480000972748,
      "epoch": 0.5186940494997367,
      "grad_norm": 0.2882140576839447,
      "learning_rate": 0.00019908733104499868,
      "loss": 0.2003549188375473,
      "mean_token_accuracy": 0.9193922579288483,
      "num_tokens": 12150960.0,
      "step": 985
    },
    {
      "entropy": 1.5195731818675995,
      "epoch": 0.5192206424433913,
      "grad_norm": 0.33627068996429443,
      "learning_rate": 0.00019908421344931173,
      "loss": 0.2529386878013611,
      "mean_token_accuracy": 0.8942808210849762,
      "num_tokens": 12163296.0,
      "step": 986
    },
    {
      "entropy": 1.480097621679306,
      "epoch": 0.5197472353870458,
      "grad_norm": 1.469023585319519,
      "learning_rate": 0.0001990810905652142,
      "loss": 0.24130460619926453,
      "mean_token_accuracy": 0.9022433310747147,
      "num_tokens": 12175632.0,
      "step": 987
    },
    {
      "entropy": 1.498535007238388,
      "epoch": 0.5202738283307003,
      "grad_norm": 0.29503491520881653,
      "learning_rate": 0.00019907796239289154,
      "loss": 0.21136939525604248,
      "mean_token_accuracy": 0.9171615391969681,
      "num_tokens": 12187968.0,
      "step": 988
    },
    {
      "entropy": 1.5154505968093872,
      "epoch": 0.5208004212743549,
      "grad_norm": 0.29515334963798523,
      "learning_rate": 0.00019907482893252945,
      "loss": 0.22373729944229126,
      "mean_token_accuracy": 0.9104124307632446,
      "num_tokens": 12200304.0,
      "step": 989
    },
    {
      "entropy": 1.5455197095870972,
      "epoch": 0.5213270142180095,
      "grad_norm": 0.36879390478134155,
      "learning_rate": 0.00019907169018431394,
      "loss": 0.23778440058231354,
      "mean_token_accuracy": 0.9108938276767731,
      "num_tokens": 12212640.0,
      "step": 990
    },
    {
      "entropy": 1.5229522287845612,
      "epoch": 0.5218536071616641,
      "grad_norm": 0.33947938680648804,
      "learning_rate": 0.0001990685461484313,
      "loss": 0.22162161767482758,
      "mean_token_accuracy": 0.9117240607738495,
      "num_tokens": 12224976.0,
      "step": 991
    },
    {
      "entropy": 1.5108350217342377,
      "epoch": 0.5223802001053186,
      "grad_norm": 0.3344981074333191,
      "learning_rate": 0.00019906539682506823,
      "loss": 0.20931261777877808,
      "mean_token_accuracy": 0.9169653207063675,
      "num_tokens": 12237312.0,
      "step": 992
    },
    {
      "entropy": 1.508901059627533,
      "epoch": 0.5229067930489731,
      "grad_norm": 0.3136521279811859,
      "learning_rate": 0.00019906224221441168,
      "loss": 0.22339576482772827,
      "mean_token_accuracy": 0.9082980453968048,
      "num_tokens": 12249648.0,
      "step": 993
    },
    {
      "entropy": 1.5110598504543304,
      "epoch": 0.5234333859926277,
      "grad_norm": 0.3550313413143158,
      "learning_rate": 0.0001990590823166489,
      "loss": 0.2181922197341919,
      "mean_token_accuracy": 0.913140058517456,
      "num_tokens": 12261984.0,
      "step": 994
    },
    {
      "entropy": 1.4990539252758026,
      "epoch": 0.5239599789362822,
      "grad_norm": 0.3551686108112335,
      "learning_rate": 0.00019905591713196747,
      "loss": 0.21492283046245575,
      "mean_token_accuracy": 0.917353168129921,
      "num_tokens": 12274320.0,
      "step": 995
    },
    {
      "entropy": 1.6208187341690063,
      "epoch": 0.5244865718799369,
      "grad_norm": 0.3526499271392822,
      "learning_rate": 0.0001990527466605553,
      "loss": 0.2129640281200409,
      "mean_token_accuracy": 0.913350835442543,
      "num_tokens": 12286656.0,
      "step": 996
    },
    {
      "entropy": 1.5281141996383667,
      "epoch": 0.5250131648235914,
      "grad_norm": 0.33227595686912537,
      "learning_rate": 0.00019904957090260056,
      "loss": 0.21114130318164825,
      "mean_token_accuracy": 0.9172855019569397,
      "num_tokens": 12298992.0,
      "step": 997
    },
    {
      "entropy": 1.4690798819065094,
      "epoch": 0.5255397577672459,
      "grad_norm": 0.30048811435699463,
      "learning_rate": 0.00019904638985829187,
      "loss": 0.20892660319805145,
      "mean_token_accuracy": 0.9182561188936234,
      "num_tokens": 12311328.0,
      "step": 998
    },
    {
      "entropy": 1.6034342348575592,
      "epoch": 0.5260663507109005,
      "grad_norm": 0.3628803491592407,
      "learning_rate": 0.000199043203527818,
      "loss": 0.21608318388462067,
      "mean_token_accuracy": 0.913134828209877,
      "num_tokens": 12323664.0,
      "step": 999
    },
    {
      "entropy": 1.4999005496501923,
      "epoch": 0.526592943654555,
      "grad_norm": 0.3294629454612732,
      "learning_rate": 0.0001990400119113681,
      "loss": 0.21117916703224182,
      "mean_token_accuracy": 0.917900949716568,
      "num_tokens": 12336000.0,
      "step": 1000
    },
    {
      "entropy": 1.596118301153183,
      "epoch": 0.5271195365982095,
      "grad_norm": 0.3512589931488037,
      "learning_rate": 0.00019903681500913167,
      "loss": 0.20125192403793335,
      "mean_token_accuracy": 0.921380877494812,
      "num_tokens": 12348336.0,
      "step": 1001
    },
    {
      "entropy": 1.4794040620326996,
      "epoch": 0.5276461295418642,
      "grad_norm": 0.3046097457408905,
      "learning_rate": 0.0001990336128212985,
      "loss": 0.2075495421886444,
      "mean_token_accuracy": 0.9140082001686096,
      "num_tokens": 12360672.0,
      "step": 1002
    },
    {
      "entropy": 1.5259474217891693,
      "epoch": 0.5281727224855187,
      "grad_norm": 0.3376280963420868,
      "learning_rate": 0.00019903040534805866,
      "loss": 0.22057998180389404,
      "mean_token_accuracy": 0.9114288538694382,
      "num_tokens": 12373008.0,
      "step": 1003
    },
    {
      "entropy": 1.5479835271835327,
      "epoch": 0.5286993154291733,
      "grad_norm": 0.3668357729911804,
      "learning_rate": 0.00019902719258960253,
      "loss": 0.24602656066417694,
      "mean_token_accuracy": 0.9003023505210876,
      "num_tokens": 12385344.0,
      "step": 1004
    },
    {
      "entropy": 1.5563693940639496,
      "epoch": 0.5292259083728278,
      "grad_norm": 0.33542895317077637,
      "learning_rate": 0.0001990239745461209,
      "loss": 0.22432827949523926,
      "mean_token_accuracy": 0.9152787178754807,
      "num_tokens": 12397680.0,
      "step": 1005
    },
    {
      "entropy": 1.5360921025276184,
      "epoch": 0.5297525013164823,
      "grad_norm": 0.34227633476257324,
      "learning_rate": 0.00019902075121780473,
      "loss": 0.24711929261684418,
      "mean_token_accuracy": 0.9043864160776138,
      "num_tokens": 12410016.0,
      "step": 1006
    },
    {
      "entropy": 1.5392629504203796,
      "epoch": 0.5302790942601369,
      "grad_norm": 0.31668519973754883,
      "learning_rate": 0.00019901752260484545,
      "loss": 0.24319474399089813,
      "mean_token_accuracy": 0.9046953171491623,
      "num_tokens": 12422352.0,
      "step": 1007
    },
    {
      "entropy": 1.608776479959488,
      "epoch": 0.5308056872037915,
      "grad_norm": 0.3281152844429016,
      "learning_rate": 0.00019901428870743466,
      "loss": 0.2047930210828781,
      "mean_token_accuracy": 0.9202082604169846,
      "num_tokens": 12434688.0,
      "step": 1008
    },
    {
      "entropy": 1.6915300488471985,
      "epoch": 0.531332280147446,
      "grad_norm": 0.38227638602256775,
      "learning_rate": 0.0001990110495257644,
      "loss": 0.23451298475265503,
      "mean_token_accuracy": 0.9067147076129913,
      "num_tokens": 12447024.0,
      "step": 1009
    },
    {
      "entropy": 1.5676094889640808,
      "epoch": 0.5318588730911006,
      "grad_norm": 0.3091145157814026,
      "learning_rate": 0.0001990078050600269,
      "loss": 0.21618413925170898,
      "mean_token_accuracy": 0.9121405184268951,
      "num_tokens": 12459360.0,
      "step": 1010
    },
    {
      "entropy": 1.6247548460960388,
      "epoch": 0.5323854660347551,
      "grad_norm": 0.3261072337627411,
      "learning_rate": 0.0001990045553104148,
      "loss": 0.2148750126361847,
      "mean_token_accuracy": 0.9210677593946457,
      "num_tokens": 12471696.0,
      "step": 1011
    },
    {
      "entropy": 1.5738478302955627,
      "epoch": 0.5329120589784097,
      "grad_norm": 0.32214832305908203,
      "learning_rate": 0.00019900130027712099,
      "loss": 0.23918862640857697,
      "mean_token_accuracy": 0.9062842279672623,
      "num_tokens": 12484032.0,
      "step": 1012
    },
    {
      "entropy": 1.474539339542389,
      "epoch": 0.5334386519220643,
      "grad_norm": 0.346766859292984,
      "learning_rate": 0.00019899803996033876,
      "loss": 0.21119186282157898,
      "mean_token_accuracy": 0.918285146355629,
      "num_tokens": 12496368.0,
      "step": 1013
    },
    {
      "entropy": 1.5115889310836792,
      "epoch": 0.5339652448657188,
      "grad_norm": 0.3021303117275238,
      "learning_rate": 0.00019899477436026157,
      "loss": 0.19445011019706726,
      "mean_token_accuracy": 0.9237413257360458,
      "num_tokens": 12508704.0,
      "step": 1014
    },
    {
      "entropy": 1.492755115032196,
      "epoch": 0.5344918378093734,
      "grad_norm": 0.3711804449558258,
      "learning_rate": 0.00019899150347708335,
      "loss": 0.21977633237838745,
      "mean_token_accuracy": 0.9105686545372009,
      "num_tokens": 12521040.0,
      "step": 1015
    },
    {
      "entropy": 1.4154399931430817,
      "epoch": 0.5350184307530279,
      "grad_norm": 0.34001612663269043,
      "learning_rate": 0.00019898822731099827,
      "loss": 0.1881294697523117,
      "mean_token_accuracy": 0.924776166677475,
      "num_tokens": 12533376.0,
      "step": 1016
    },
    {
      "entropy": 1.41116601228714,
      "epoch": 0.5355450236966824,
      "grad_norm": 0.33592158555984497,
      "learning_rate": 0.00019898494586220077,
      "loss": 0.22032563388347626,
      "mean_token_accuracy": 0.9117150008678436,
      "num_tokens": 12545712.0,
      "step": 1017
    },
    {
      "entropy": 1.416001707315445,
      "epoch": 0.536071616640337,
      "grad_norm": 0.3635656535625458,
      "learning_rate": 0.00019898165913088568,
      "loss": 0.2379489541053772,
      "mean_token_accuracy": 0.9045498371124268,
      "num_tokens": 12558048.0,
      "step": 1018
    },
    {
      "entropy": 1.346794456243515,
      "epoch": 0.5365982095839916,
      "grad_norm": 0.41112419962882996,
      "learning_rate": 0.0001989783671172481,
      "loss": 0.24667346477508545,
      "mean_token_accuracy": 0.9044399708509445,
      "num_tokens": 12570384.0,
      "step": 1019
    },
    {
      "entropy": 1.417405605316162,
      "epoch": 0.5371248025276462,
      "grad_norm": 0.3971119821071625,
      "learning_rate": 0.00019897506982148353,
      "loss": 0.19713318347930908,
      "mean_token_accuracy": 0.9267271012067795,
      "num_tokens": 12582720.0,
      "step": 1020
    },
    {
      "entropy": 1.4534864723682404,
      "epoch": 0.5376513954713007,
      "grad_norm": 0.3351125717163086,
      "learning_rate": 0.00019897176724378762,
      "loss": 0.22281871736049652,
      "mean_token_accuracy": 0.9098447114229202,
      "num_tokens": 12595056.0,
      "step": 1021
    },
    {
      "entropy": 1.3609624803066254,
      "epoch": 0.5381779884149552,
      "grad_norm": 0.3119547665119171,
      "learning_rate": 0.00019896845938435642,
      "loss": 0.2145942896604538,
      "mean_token_accuracy": 0.9176490157842636,
      "num_tokens": 12607392.0,
      "step": 1022
    },
    {
      "entropy": 1.373078614473343,
      "epoch": 0.5387045813586098,
      "grad_norm": 0.3335905075073242,
      "learning_rate": 0.00019896514624338633,
      "loss": 0.20780502259731293,
      "mean_token_accuracy": 0.9145664125680923,
      "num_tokens": 12619728.0,
      "step": 1023
    },
    {
      "entropy": 1.45356023311615,
      "epoch": 0.5392311743022643,
      "grad_norm": 0.321328729391098,
      "learning_rate": 0.00019896182782107408,
      "loss": 0.21474064886569977,
      "mean_token_accuracy": 0.914635494351387,
      "num_tokens": 12632064.0,
      "step": 1024
    },
    {
      "entropy": 1.3756296038627625,
      "epoch": 0.539757767245919,
      "grad_norm": 0.3236009180545807,
      "learning_rate": 0.0001989585041176166,
      "loss": 0.19574275612831116,
      "mean_token_accuracy": 0.9214889109134674,
      "num_tokens": 12644400.0,
      "step": 1025
    },
    {
      "entropy": 1.374951332807541,
      "epoch": 0.5402843601895735,
      "grad_norm": 0.33436882495880127,
      "learning_rate": 0.00019895517513321125,
      "loss": 0.2343987077474594,
      "mean_token_accuracy": 0.9015287011861801,
      "num_tokens": 12656736.0,
      "step": 1026
    },
    {
      "entropy": 1.3473423719406128,
      "epoch": 0.540810953133228,
      "grad_norm": 0.30872446298599243,
      "learning_rate": 0.0001989518408680556,
      "loss": 0.23760074377059937,
      "mean_token_accuracy": 0.9077365547418594,
      "num_tokens": 12669072.0,
      "step": 1027
    },
    {
      "entropy": 1.4245812594890594,
      "epoch": 0.5413375460768826,
      "grad_norm": 0.32098084688186646,
      "learning_rate": 0.0001989485013223476,
      "loss": 0.22173723578453064,
      "mean_token_accuracy": 0.9116310328245163,
      "num_tokens": 12681408.0,
      "step": 1028
    },
    {
      "entropy": 1.357390284538269,
      "epoch": 0.5418641390205371,
      "grad_norm": 0.3540545701980591,
      "learning_rate": 0.0001989451564962855,
      "loss": 0.22398629784584045,
      "mean_token_accuracy": 0.914269283413887,
      "num_tokens": 12693744.0,
      "step": 1029
    },
    {
      "entropy": 1.4113155901432037,
      "epoch": 0.5423907319641916,
      "grad_norm": 0.3349270224571228,
      "learning_rate": 0.00019894180639006787,
      "loss": 0.23276042938232422,
      "mean_token_accuracy": 0.905420333147049,
      "num_tokens": 12706080.0,
      "step": 1030
    },
    {
      "entropy": 1.3748076260089874,
      "epoch": 0.5429173249078463,
      "grad_norm": 0.31247857213020325,
      "learning_rate": 0.0001989384510038936,
      "loss": 0.2479267716407776,
      "mean_token_accuracy": 0.8999170660972595,
      "num_tokens": 12718416.0,
      "step": 1031
    },
    {
      "entropy": 1.4230112135410309,
      "epoch": 0.5434439178515008,
      "grad_norm": 0.31677430868148804,
      "learning_rate": 0.0001989350903379619,
      "loss": 0.2388826310634613,
      "mean_token_accuracy": 0.9079335629940033,
      "num_tokens": 12730752.0,
      "step": 1032
    },
    {
      "entropy": 1.341791033744812,
      "epoch": 0.5439705107951553,
      "grad_norm": 0.30199915170669556,
      "learning_rate": 0.0001989317243924722,
      "loss": 0.21240849792957306,
      "mean_token_accuracy": 0.9131926447153091,
      "num_tokens": 12743088.0,
      "step": 1033
    },
    {
      "entropy": 1.3770472407341003,
      "epoch": 0.5444971037388099,
      "grad_norm": 0.28729885816574097,
      "learning_rate": 0.00019892835316762437,
      "loss": 0.20151323080062866,
      "mean_token_accuracy": 0.9193004071712494,
      "num_tokens": 12755424.0,
      "step": 1034
    },
    {
      "entropy": 1.3323063850402832,
      "epoch": 0.5450236966824644,
      "grad_norm": 0.3478204607963562,
      "learning_rate": 0.00019892497666361848,
      "loss": 0.22956590354442596,
      "mean_token_accuracy": 0.910034716129303,
      "num_tokens": 12767760.0,
      "step": 1035
    },
    {
      "entropy": 1.2777645885944366,
      "epoch": 0.545550289626119,
      "grad_norm": 0.32218146324157715,
      "learning_rate": 0.00019892159488065506,
      "loss": 0.2523505985736847,
      "mean_token_accuracy": 0.8999349474906921,
      "num_tokens": 12780096.0,
      "step": 1036
    },
    {
      "entropy": 1.3476167619228363,
      "epoch": 0.5460768825697736,
      "grad_norm": 0.3297399580478668,
      "learning_rate": 0.00019891820781893485,
      "loss": 0.24944739043712616,
      "mean_token_accuracy": 0.8990609049797058,
      "num_tokens": 12792432.0,
      "step": 1037
    },
    {
      "entropy": 1.3680584728717804,
      "epoch": 0.5466034755134281,
      "grad_norm": 0.31413063406944275,
      "learning_rate": 0.00019891481547865887,
      "loss": 0.23303711414337158,
      "mean_token_accuracy": 0.9072806537151337,
      "num_tokens": 12804768.0,
      "step": 1038
    },
    {
      "entropy": 1.2887639105319977,
      "epoch": 0.5471300684570827,
      "grad_norm": 0.30452004075050354,
      "learning_rate": 0.00019891141786002853,
      "loss": 0.20885013043880463,
      "mean_token_accuracy": 0.91634900867939,
      "num_tokens": 12817104.0,
      "step": 1039
    },
    {
      "entropy": 1.427742451429367,
      "epoch": 0.5476566614007372,
      "grad_norm": 0.3391040563583374,
      "learning_rate": 0.00019890801496324554,
      "loss": 0.21363744139671326,
      "mean_token_accuracy": 0.9168877154588699,
      "num_tokens": 12829440.0,
      "step": 1040
    },
    {
      "entropy": 1.4079940021038055,
      "epoch": 0.5481832543443917,
      "grad_norm": 0.34792712330818176,
      "learning_rate": 0.00019890460678851187,
      "loss": 0.23039647936820984,
      "mean_token_accuracy": 0.9067210853099823,
      "num_tokens": 12841776.0,
      "step": 1041
    },
    {
      "entropy": 1.362496793270111,
      "epoch": 0.5487098472880464,
      "grad_norm": 0.3340260684490204,
      "learning_rate": 0.00019890119333602988,
      "loss": 0.21935522556304932,
      "mean_token_accuracy": 0.9152266681194305,
      "num_tokens": 12854112.0,
      "step": 1042
    },
    {
      "entropy": 1.3945021033287048,
      "epoch": 0.5492364402317009,
      "grad_norm": 0.31123343110084534,
      "learning_rate": 0.00019889777460600221,
      "loss": 0.2165759801864624,
      "mean_token_accuracy": 0.9149927645921707,
      "num_tokens": 12866448.0,
      "step": 1043
    },
    {
      "entropy": 1.3361455798149109,
      "epoch": 0.5497630331753555,
      "grad_norm": 0.30849552154541016,
      "learning_rate": 0.0001988943505986318,
      "loss": 0.23334844410419464,
      "mean_token_accuracy": 0.9073198139667511,
      "num_tokens": 12878784.0,
      "step": 1044
    },
    {
      "entropy": 1.353363960981369,
      "epoch": 0.55028962611901,
      "grad_norm": 0.30394431948661804,
      "learning_rate": 0.00019889092131412188,
      "loss": 0.21873678267002106,
      "mean_token_accuracy": 0.9092230349779129,
      "num_tokens": 12891120.0,
      "step": 1045
    },
    {
      "entropy": 1.3730317652225494,
      "epoch": 0.5508162190626645,
      "grad_norm": 0.3051231801509857,
      "learning_rate": 0.0001988874867526761,
      "loss": 0.2074829638004303,
      "mean_token_accuracy": 0.9159383773803711,
      "num_tokens": 12903456.0,
      "step": 1046
    },
    {
      "entropy": 1.348529189825058,
      "epoch": 0.5513428120063191,
      "grad_norm": 0.34162041544914246,
      "learning_rate": 0.00019888404691449828,
      "loss": 0.20376741886138916,
      "mean_token_accuracy": 0.9213349372148514,
      "num_tokens": 12915792.0,
      "step": 1047
    },
    {
      "entropy": 1.3659497201442719,
      "epoch": 0.5518694049499737,
      "grad_norm": 0.3287021219730377,
      "learning_rate": 0.00019888060179979266,
      "loss": 0.22284625470638275,
      "mean_token_accuracy": 0.9160720854997635,
      "num_tokens": 12928128.0,
      "step": 1048
    },
    {
      "entropy": 1.39521923661232,
      "epoch": 0.5523959978936283,
      "grad_norm": 0.33031636476516724,
      "learning_rate": 0.00019887715140876373,
      "loss": 0.2400185465812683,
      "mean_token_accuracy": 0.9030187875032425,
      "num_tokens": 12940464.0,
      "step": 1049
    },
    {
      "entropy": 1.4470847249031067,
      "epoch": 0.5529225908372828,
      "grad_norm": 0.37168604135513306,
      "learning_rate": 0.00019887369574161633,
      "loss": 0.22982707619667053,
      "mean_token_accuracy": 0.9084116071462631,
      "num_tokens": 12952800.0,
      "step": 1050
    },
    {
      "entropy": 1.3384085595607758,
      "epoch": 0.5534491837809373,
      "grad_norm": 0.32396450638771057,
      "learning_rate": 0.00019887023479855565,
      "loss": 0.24000367522239685,
      "mean_token_accuracy": 0.9041632562875748,
      "num_tokens": 12965136.0,
      "step": 1051
    },
    {
      "entropy": 1.3531986474990845,
      "epoch": 0.5539757767245919,
      "grad_norm": 0.319603830575943,
      "learning_rate": 0.00019886676857978707,
      "loss": 0.23166413605213165,
      "mean_token_accuracy": 0.9047738164663315,
      "num_tokens": 12977472.0,
      "step": 1052
    },
    {
      "entropy": 1.3648565709590912,
      "epoch": 0.5545023696682464,
      "grad_norm": 0.3087165057659149,
      "learning_rate": 0.00019886329708551642,
      "loss": 0.21422043442726135,
      "mean_token_accuracy": 0.9170399606227875,
      "num_tokens": 12989808.0,
      "step": 1053
    },
    {
      "entropy": 1.4327314496040344,
      "epoch": 0.555028962611901,
      "grad_norm": 0.31154927611351013,
      "learning_rate": 0.00019885982031594973,
      "loss": 0.23706527054309845,
      "mean_token_accuracy": 0.9066223204135895,
      "num_tokens": 13002144.0,
      "step": 1054
    },
    {
      "entropy": 1.401264727115631,
      "epoch": 0.5555555555555556,
      "grad_norm": 0.30120137333869934,
      "learning_rate": 0.00019885633827129343,
      "loss": 0.21149972081184387,
      "mean_token_accuracy": 0.9169722944498062,
      "num_tokens": 13014480.0,
      "step": 1055
    },
    {
      "entropy": 1.4713716506958008,
      "epoch": 0.5560821484992101,
      "grad_norm": 0.39476147294044495,
      "learning_rate": 0.00019885285095175422,
      "loss": 0.26564526557922363,
      "mean_token_accuracy": 0.8999374806880951,
      "num_tokens": 13026816.0,
      "step": 1056
    },
    {
      "entropy": 1.4001291990280151,
      "epoch": 0.5566087414428647,
      "grad_norm": 0.3243412971496582,
      "learning_rate": 0.00019884935835753914,
      "loss": 0.2307104468345642,
      "mean_token_accuracy": 0.9069186598062515,
      "num_tokens": 13039152.0,
      "step": 1057
    },
    {
      "entropy": 1.4063750207424164,
      "epoch": 0.5571353343865192,
      "grad_norm": 0.3205982446670532,
      "learning_rate": 0.0001988458604888555,
      "loss": 0.2200152724981308,
      "mean_token_accuracy": 0.9117906540632248,
      "num_tokens": 13051488.0,
      "step": 1058
    },
    {
      "entropy": 1.4639129042625427,
      "epoch": 0.5576619273301737,
      "grad_norm": 0.35051432251930237,
      "learning_rate": 0.00019884235734591096,
      "loss": 0.21262915432453156,
      "mean_token_accuracy": 0.9155555963516235,
      "num_tokens": 13063824.0,
      "step": 1059
    },
    {
      "entropy": 1.3807141780853271,
      "epoch": 0.5581885202738284,
      "grad_norm": 0.31436532735824585,
      "learning_rate": 0.00019883884892891348,
      "loss": 0.23192916810512543,
      "mean_token_accuracy": 0.9132644683122635,
      "num_tokens": 13076160.0,
      "step": 1060
    },
    {
      "entropy": 1.4159002602100372,
      "epoch": 0.5587151132174829,
      "grad_norm": 0.35557860136032104,
      "learning_rate": 0.00019883533523807131,
      "loss": 0.2557951807975769,
      "mean_token_accuracy": 0.9019583463668823,
      "num_tokens": 13088496.0,
      "step": 1061
    },
    {
      "entropy": 1.4077826738357544,
      "epoch": 0.5592417061611374,
      "grad_norm": 0.35081014037132263,
      "learning_rate": 0.00019883181627359305,
      "loss": 0.21151600778102875,
      "mean_token_accuracy": 0.914755642414093,
      "num_tokens": 13100832.0,
      "step": 1062
    },
    {
      "entropy": 1.3742648661136627,
      "epoch": 0.559768299104792,
      "grad_norm": 0.28598442673683167,
      "learning_rate": 0.00019882829203568764,
      "loss": 0.21076026558876038,
      "mean_token_accuracy": 0.9148049652576447,
      "num_tokens": 13113168.0,
      "step": 1063
    },
    {
      "entropy": 1.4464651346206665,
      "epoch": 0.5602948920484465,
      "grad_norm": 0.29035648703575134,
      "learning_rate": 0.00019882476252456428,
      "loss": 0.20042768120765686,
      "mean_token_accuracy": 0.9182227253913879,
      "num_tokens": 13125504.0,
      "step": 1064
    },
    {
      "entropy": 1.3639024496078491,
      "epoch": 0.5608214849921012,
      "grad_norm": 0.3004502058029175,
      "learning_rate": 0.00019882122774043243,
      "loss": 0.2036202847957611,
      "mean_token_accuracy": 0.9217555373907089,
      "num_tokens": 13137840.0,
      "step": 1065
    },
    {
      "entropy": 1.3562072217464447,
      "epoch": 0.5613480779357557,
      "grad_norm": 0.29948773980140686,
      "learning_rate": 0.000198817687683502,
      "loss": 0.20943708717823029,
      "mean_token_accuracy": 0.9172091037034988,
      "num_tokens": 13150176.0,
      "step": 1066
    },
    {
      "entropy": 1.4171293675899506,
      "epoch": 0.5618746708794102,
      "grad_norm": 0.32766589522361755,
      "learning_rate": 0.00019881414235398313,
      "loss": 0.22548557817935944,
      "mean_token_accuracy": 0.9138119965791702,
      "num_tokens": 13162512.0,
      "step": 1067
    },
    {
      "entropy": 1.494034618139267,
      "epoch": 0.5624012638230648,
      "grad_norm": 0.3495429754257202,
      "learning_rate": 0.00019881059175208624,
      "loss": 0.24128882586956024,
      "mean_token_accuracy": 0.9064372777938843,
      "num_tokens": 13174848.0,
      "step": 1068
    },
    {
      "entropy": 1.4479554295539856,
      "epoch": 0.5629278567667193,
      "grad_norm": 0.3285626769065857,
      "learning_rate": 0.0001988070358780222,
      "loss": 0.1992790549993515,
      "mean_token_accuracy": 0.922189861536026,
      "num_tokens": 13187184.0,
      "step": 1069
    },
    {
      "entropy": 1.36091148853302,
      "epoch": 0.5634544497103738,
      "grad_norm": 0.29997125267982483,
      "learning_rate": 0.00019880347473200197,
      "loss": 0.19809836149215698,
      "mean_token_accuracy": 0.9212192296981812,
      "num_tokens": 13199520.0,
      "step": 1070
    },
    {
      "entropy": 1.4168705940246582,
      "epoch": 0.5639810426540285,
      "grad_norm": 0.320639044046402,
      "learning_rate": 0.0001987999083142371,
      "loss": 0.21112670004367828,
      "mean_token_accuracy": 0.9181769639253616,
      "num_tokens": 13211856.0,
      "step": 1071
    },
    {
      "entropy": 1.3809196650981903,
      "epoch": 0.564507635597683,
      "grad_norm": 0.36394959688186646,
      "learning_rate": 0.0001987963366249392,
      "loss": 0.23442226648330688,
      "mean_token_accuracy": 0.9071120172739029,
      "num_tokens": 13224192.0,
      "step": 1072
    },
    {
      "entropy": 1.3890978693962097,
      "epoch": 0.5650342285413376,
      "grad_norm": 0.34244871139526367,
      "learning_rate": 0.00019879275966432037,
      "loss": 0.22117844223976135,
      "mean_token_accuracy": 0.9115606546401978,
      "num_tokens": 13236528.0,
      "step": 1073
    },
    {
      "entropy": 1.3886411488056183,
      "epoch": 0.5655608214849921,
      "grad_norm": 0.35859042406082153,
      "learning_rate": 0.00019878917743259292,
      "loss": 0.21675634384155273,
      "mean_token_accuracy": 0.9194170236587524,
      "num_tokens": 13248864.0,
      "step": 1074
    },
    {
      "entropy": 1.4300493896007538,
      "epoch": 0.5660874144286466,
      "grad_norm": 0.34859365224838257,
      "learning_rate": 0.00019878558992996947,
      "loss": 0.24318671226501465,
      "mean_token_accuracy": 0.9037031382322311,
      "num_tokens": 13261200.0,
      "step": 1075
    },
    {
      "entropy": 1.3681704699993134,
      "epoch": 0.5666140073723012,
      "grad_norm": 0.3271603286266327,
      "learning_rate": 0.00019878199715666305,
      "loss": 0.2319377213716507,
      "mean_token_accuracy": 0.9129595458507538,
      "num_tokens": 13273536.0,
      "step": 1076
    },
    {
      "entropy": 1.395903766155243,
      "epoch": 0.5671406003159558,
      "grad_norm": 0.29173579812049866,
      "learning_rate": 0.00019877839911288693,
      "loss": 0.20177659392356873,
      "mean_token_accuracy": 0.9242661744356155,
      "num_tokens": 13285872.0,
      "step": 1077
    },
    {
      "entropy": 1.4401375949382782,
      "epoch": 0.5676671932596103,
      "grad_norm": 0.3181249499320984,
      "learning_rate": 0.0001987747957988547,
      "loss": 0.22763216495513916,
      "mean_token_accuracy": 0.907039076089859,
      "num_tokens": 13298208.0,
      "step": 1078
    },
    {
      "entropy": 1.3585200309753418,
      "epoch": 0.5681937862032649,
      "grad_norm": 0.2902856171131134,
      "learning_rate": 0.00019877118721478026,
      "loss": 0.19510802626609802,
      "mean_token_accuracy": 0.9210563600063324,
      "num_tokens": 13310544.0,
      "step": 1079
    },
    {
      "entropy": 1.3690218329429626,
      "epoch": 0.5687203791469194,
      "grad_norm": 0.31650200486183167,
      "learning_rate": 0.00019876757336087782,
      "loss": 0.23386812210083008,
      "mean_token_accuracy": 0.9065526723861694,
      "num_tokens": 13322880.0,
      "step": 1080
    },
    {
      "entropy": 1.4170444905757904,
      "epoch": 0.569246972090574,
      "grad_norm": 0.39458808302879333,
      "learning_rate": 0.00019876395423736192,
      "loss": 0.21256215870380402,
      "mean_token_accuracy": 0.9154652059078217,
      "num_tokens": 13335216.0,
      "step": 1081
    },
    {
      "entropy": 1.3711142838001251,
      "epoch": 0.5697735650342285,
      "grad_norm": 0.3102315366268158,
      "learning_rate": 0.00019876032984444744,
      "loss": 0.2107260674238205,
      "mean_token_accuracy": 0.9209572970867157,
      "num_tokens": 13347552.0,
      "step": 1082
    },
    {
      "entropy": 1.3468361794948578,
      "epoch": 0.5703001579778831,
      "grad_norm": 0.3204945921897888,
      "learning_rate": 0.00019875670018234946,
      "loss": 0.21677452325820923,
      "mean_token_accuracy": 0.9137840569019318,
      "num_tokens": 13359888.0,
      "step": 1083
    },
    {
      "entropy": 1.3397773206233978,
      "epoch": 0.5708267509215377,
      "grad_norm": 0.30542871356010437,
      "learning_rate": 0.00019875306525128354,
      "loss": 0.20801898837089539,
      "mean_token_accuracy": 0.9187730103731155,
      "num_tokens": 13372224.0,
      "step": 1084
    },
    {
      "entropy": 1.3542706668376923,
      "epoch": 0.5713533438651922,
      "grad_norm": 0.39070987701416016,
      "learning_rate": 0.00019874942505146542,
      "loss": 0.23829345405101776,
      "mean_token_accuracy": 0.9077901542186737,
      "num_tokens": 13384560.0,
      "step": 1085
    },
    {
      "entropy": 1.3789608180522919,
      "epoch": 0.5718799368088467,
      "grad_norm": 0.31589949131011963,
      "learning_rate": 0.00019874577958311124,
      "loss": 0.23594669997692108,
      "mean_token_accuracy": 0.9101516306400299,
      "num_tokens": 13396896.0,
      "step": 1086
    },
    {
      "entropy": 1.2831343114376068,
      "epoch": 0.5724065297525013,
      "grad_norm": 0.31056761741638184,
      "learning_rate": 0.00019874212884643733,
      "loss": 0.22809851169586182,
      "mean_token_accuracy": 0.9048083126544952,
      "num_tokens": 13409232.0,
      "step": 1087
    },
    {
      "entropy": 1.343775451183319,
      "epoch": 0.5729331226961558,
      "grad_norm": 0.35355332493782043,
      "learning_rate": 0.00019873847284166047,
      "loss": 0.24591763317584991,
      "mean_token_accuracy": 0.9045094102621078,
      "num_tokens": 13421568.0,
      "step": 1088
    },
    {
      "entropy": 1.3326263129711151,
      "epoch": 0.5734597156398105,
      "grad_norm": 0.30252182483673096,
      "learning_rate": 0.00019873481156899769,
      "loss": 0.2264258861541748,
      "mean_token_accuracy": 0.9123017936944962,
      "num_tokens": 13433904.0,
      "step": 1089
    },
    {
      "entropy": 1.4111762046813965,
      "epoch": 0.573986308583465,
      "grad_norm": 0.30441656708717346,
      "learning_rate": 0.00019873114502866633,
      "loss": 0.20945101976394653,
      "mean_token_accuracy": 0.9133919775485992,
      "num_tokens": 13446240.0,
      "step": 1090
    },
    {
      "entropy": 1.3507026731967926,
      "epoch": 0.5745129015271195,
      "grad_norm": 0.29725486040115356,
      "learning_rate": 0.00019872747322088405,
      "loss": 0.22033721208572388,
      "mean_token_accuracy": 0.9093347936868668,
      "num_tokens": 13458576.0,
      "step": 1091
    },
    {
      "entropy": 1.3604736626148224,
      "epoch": 0.5750394944707741,
      "grad_norm": 0.2988698482513428,
      "learning_rate": 0.00019872379614586884,
      "loss": 0.20934537053108215,
      "mean_token_accuracy": 0.919558510184288,
      "num_tokens": 13470912.0,
      "step": 1092
    },
    {
      "entropy": 1.2718960344791412,
      "epoch": 0.5755660874144286,
      "grad_norm": 0.296278178691864,
      "learning_rate": 0.00019872011380383896,
      "loss": 0.2263902723789215,
      "mean_token_accuracy": 0.9120451807975769,
      "num_tokens": 13483248.0,
      "step": 1093
    },
    {
      "entropy": 1.3046118319034576,
      "epoch": 0.5760926803580833,
      "grad_norm": 0.2931486666202545,
      "learning_rate": 0.00019871642619501302,
      "loss": 0.22945842146873474,
      "mean_token_accuracy": 0.9100485295057297,
      "num_tokens": 13495584.0,
      "step": 1094
    },
    {
      "entropy": 1.3241084516048431,
      "epoch": 0.5766192733017378,
      "grad_norm": 0.3063081204891205,
      "learning_rate": 0.00019871273331960996,
      "loss": 0.2523494362831116,
      "mean_token_accuracy": 0.9037816673517227,
      "num_tokens": 13507920.0,
      "step": 1095
    },
    {
      "entropy": 1.3798290491104126,
      "epoch": 0.5771458662453923,
      "grad_norm": 0.3451617956161499,
      "learning_rate": 0.00019870903517784898,
      "loss": 0.2181454300880432,
      "mean_token_accuracy": 0.9167459905147552,
      "num_tokens": 13520256.0,
      "step": 1096
    },
    {
      "entropy": 1.292254477739334,
      "epoch": 0.5776724591890469,
      "grad_norm": 0.29267409443855286,
      "learning_rate": 0.0001987053317699496,
      "loss": 0.2220621407032013,
      "mean_token_accuracy": 0.9128179252147675,
      "num_tokens": 13532592.0,
      "step": 1097
    },
    {
      "entropy": 1.2694779932498932,
      "epoch": 0.5781990521327014,
      "grad_norm": 0.2939811646938324,
      "learning_rate": 0.0001987016230961317,
      "loss": 0.21204796433448792,
      "mean_token_accuracy": 0.9153830707073212,
      "num_tokens": 13544928.0,
      "step": 1098
    },
    {
      "entropy": 1.301151692867279,
      "epoch": 0.5787256450763559,
      "grad_norm": 0.3276446461677551,
      "learning_rate": 0.0001986979091566154,
      "loss": 0.20669840276241302,
      "mean_token_accuracy": 0.9193602800369263,
      "num_tokens": 13557264.0,
      "step": 1099
    },
    {
      "entropy": 1.2780292332172394,
      "epoch": 0.5792522380200106,
      "grad_norm": 0.31635478138923645,
      "learning_rate": 0.00019869418995162124,
      "loss": 0.22475430369377136,
      "mean_token_accuracy": 0.9149479269981384,
      "num_tokens": 13569600.0,
      "step": 1100
    },
    {
      "entropy": 1.3376207053661346,
      "epoch": 0.5797788309636651,
      "grad_norm": 0.3281361162662506,
      "learning_rate": 0.00019869046548136996,
      "loss": 0.2165728509426117,
      "mean_token_accuracy": 0.9121999144554138,
      "num_tokens": 13581936.0,
      "step": 1101
    },
    {
      "entropy": 1.2792271375656128,
      "epoch": 0.5803054239073197,
      "grad_norm": 0.34059932827949524,
      "learning_rate": 0.00019868673574608266,
      "loss": 0.2282489687204361,
      "mean_token_accuracy": 0.9086829572916031,
      "num_tokens": 13594272.0,
      "step": 1102
    },
    {
      "entropy": 1.312345266342163,
      "epoch": 0.5808320168509742,
      "grad_norm": 0.3228605389595032,
      "learning_rate": 0.00019868300074598076,
      "loss": 0.22286418080329895,
      "mean_token_accuracy": 0.912465900182724,
      "num_tokens": 13606608.0,
      "step": 1103
    },
    {
      "entropy": 1.3179580569267273,
      "epoch": 0.5813586097946287,
      "grad_norm": 0.3180788457393646,
      "learning_rate": 0.000198679260481286,
      "loss": 0.2119407206773758,
      "mean_token_accuracy": 0.9113909304141998,
      "num_tokens": 13618944.0,
      "step": 1104
    },
    {
      "entropy": 1.3403850197792053,
      "epoch": 0.5818852027382833,
      "grad_norm": 0.3467418849468231,
      "learning_rate": 0.00019867551495222042,
      "loss": 0.23494362831115723,
      "mean_token_accuracy": 0.9045893847942352,
      "num_tokens": 13631280.0,
      "step": 1105
    },
    {
      "entropy": 1.3972817361354828,
      "epoch": 0.5824117956819379,
      "grad_norm": 0.289737343788147,
      "learning_rate": 0.00019867176415900635,
      "loss": 0.22767332196235657,
      "mean_token_accuracy": 0.9065064936876297,
      "num_tokens": 13643616.0,
      "step": 1106
    },
    {
      "entropy": 1.3646803498268127,
      "epoch": 0.5829383886255924,
      "grad_norm": 0.30822867155075073,
      "learning_rate": 0.00019866800810186644,
      "loss": 0.22304248809814453,
      "mean_token_accuracy": 0.9127557724714279,
      "num_tokens": 13655952.0,
      "step": 1107
    },
    {
      "entropy": 1.3917404413223267,
      "epoch": 0.583464981569247,
      "grad_norm": 0.3238813579082489,
      "learning_rate": 0.0001986642467810237,
      "loss": 0.23602177202701569,
      "mean_token_accuracy": 0.9076020121574402,
      "num_tokens": 13668288.0,
      "step": 1108
    },
    {
      "entropy": 1.4159584939479828,
      "epoch": 0.5839915745129015,
      "grad_norm": 0.3269672989845276,
      "learning_rate": 0.00019866048019670144,
      "loss": 0.22922644019126892,
      "mean_token_accuracy": 0.9118731915950775,
      "num_tokens": 13680624.0,
      "step": 1109
    },
    {
      "entropy": 1.426487773656845,
      "epoch": 0.584518167456556,
      "grad_norm": 0.33397728204727173,
      "learning_rate": 0.00019865670834912316,
      "loss": 0.21286353468894958,
      "mean_token_accuracy": 0.921152800321579,
      "num_tokens": 13692960.0,
      "step": 1110
    },
    {
      "entropy": 1.4119893908500671,
      "epoch": 0.5850447604002106,
      "grad_norm": 0.29753443598747253,
      "learning_rate": 0.00019865293123851288,
      "loss": 0.2373121976852417,
      "mean_token_accuracy": 0.9035847336053848,
      "num_tokens": 13705296.0,
      "step": 1111
    },
    {
      "entropy": 1.380676507949829,
      "epoch": 0.5855713533438652,
      "grad_norm": 0.2954482436180115,
      "learning_rate": 0.00019864914886509475,
      "loss": 0.21925240755081177,
      "mean_token_accuracy": 0.9149321764707565,
      "num_tokens": 13717632.0,
      "step": 1112
    },
    {
      "entropy": 1.382420837879181,
      "epoch": 0.5860979462875198,
      "grad_norm": 0.3165317177772522,
      "learning_rate": 0.00019864536122909336,
      "loss": 0.20962990820407867,
      "mean_token_accuracy": 0.9176686257123947,
      "num_tokens": 13729968.0,
      "step": 1113
    },
    {
      "entropy": 1.4016212224960327,
      "epoch": 0.5866245392311743,
      "grad_norm": 0.3741922974586487,
      "learning_rate": 0.00019864156833073352,
      "loss": 0.21471966803073883,
      "mean_token_accuracy": 0.9138137102127075,
      "num_tokens": 13742304.0,
      "step": 1114
    },
    {
      "entropy": 1.330253154039383,
      "epoch": 0.5871511321748288,
      "grad_norm": 0.30213749408721924,
      "learning_rate": 0.00019863777017024042,
      "loss": 0.21769596636295319,
      "mean_token_accuracy": 0.9100621193647385,
      "num_tokens": 13754640.0,
      "step": 1115
    },
    {
      "entropy": 1.353569656610489,
      "epoch": 0.5876777251184834,
      "grad_norm": 0.37241899967193604,
      "learning_rate": 0.00019863396674783953,
      "loss": 0.2180408537387848,
      "mean_token_accuracy": 0.9123483598232269,
      "num_tokens": 13766976.0,
      "step": 1116
    },
    {
      "entropy": 1.3220522999763489,
      "epoch": 0.588204318062138,
      "grad_norm": 0.33360689878463745,
      "learning_rate": 0.00019863015806375662,
      "loss": 0.21166455745697021,
      "mean_token_accuracy": 0.9151635468006134,
      "num_tokens": 13779312.0,
      "step": 1117
    },
    {
      "entropy": 1.4124476611614227,
      "epoch": 0.5887309110057926,
      "grad_norm": 0.35278213024139404,
      "learning_rate": 0.0001986263441182178,
      "loss": 0.24385729432106018,
      "mean_token_accuracy": 0.9091434925794601,
      "num_tokens": 13791648.0,
      "step": 1118
    },
    {
      "entropy": 1.3624746203422546,
      "epoch": 0.5892575039494471,
      "grad_norm": 0.303867906332016,
      "learning_rate": 0.00019862252491144954,
      "loss": 0.20682811737060547,
      "mean_token_accuracy": 0.9137058258056641,
      "num_tokens": 13803984.0,
      "step": 1119
    },
    {
      "entropy": 1.4214110970497131,
      "epoch": 0.5897840968931016,
      "grad_norm": 0.29519587755203247,
      "learning_rate": 0.00019861870044367844,
      "loss": 0.1955740600824356,
      "mean_token_accuracy": 0.9222785979509354,
      "num_tokens": 13816320.0,
      "step": 1120
    },
    {
      "entropy": 1.332068532705307,
      "epoch": 0.5903106898367562,
      "grad_norm": 0.31455355882644653,
      "learning_rate": 0.0001986148707151316,
      "loss": 0.22069820761680603,
      "mean_token_accuracy": 0.9075255244970322,
      "num_tokens": 13828656.0,
      "step": 1121
    },
    {
      "entropy": 1.288136750459671,
      "epoch": 0.5908372827804107,
      "grad_norm": 0.3047003448009491,
      "learning_rate": 0.0001986110357260364,
      "loss": 0.23481610417366028,
      "mean_token_accuracy": 0.9039241820573807,
      "num_tokens": 13840992.0,
      "step": 1122
    },
    {
      "entropy": 1.3215467631816864,
      "epoch": 0.5913638757240653,
      "grad_norm": 0.2980950176715851,
      "learning_rate": 0.0001986071954766205,
      "loss": 0.20938631892204285,
      "mean_token_accuracy": 0.9155066460371017,
      "num_tokens": 13853328.0,
      "step": 1123
    },
    {
      "entropy": 1.3901076912879944,
      "epoch": 0.5918904686677199,
      "grad_norm": 0.3429953455924988,
      "learning_rate": 0.0001986033499671118,
      "loss": 0.24023953080177307,
      "mean_token_accuracy": 0.9067379236221313,
      "num_tokens": 13865664.0,
      "step": 1124
    },
    {
      "entropy": 1.3358671963214874,
      "epoch": 0.5924170616113744,
      "grad_norm": 0.2808769643306732,
      "learning_rate": 0.00019859949919773866,
      "loss": 0.20975762605667114,
      "mean_token_accuracy": 0.9166984409093857,
      "num_tokens": 13878000.0,
      "step": 1125
    },
    {
      "entropy": 1.3196144104003906,
      "epoch": 0.592943654555029,
      "grad_norm": 0.3139532506465912,
      "learning_rate": 0.0001985956431687296,
      "loss": 0.23033276200294495,
      "mean_token_accuracy": 0.9116202145814896,
      "num_tokens": 13890336.0,
      "step": 1126
    },
    {
      "entropy": 1.3117577731609344,
      "epoch": 0.5934702474986835,
      "grad_norm": 0.27732840180397034,
      "learning_rate": 0.0001985917818803136,
      "loss": 0.20888479053974152,
      "mean_token_accuracy": 0.9136437624692917,
      "num_tokens": 13902672.0,
      "step": 1127
    },
    {
      "entropy": 1.3565605282783508,
      "epoch": 0.593996840442338,
      "grad_norm": 0.3072391152381897,
      "learning_rate": 0.00019858791533271993,
      "loss": 0.2347339540719986,
      "mean_token_accuracy": 0.9031410962343216,
      "num_tokens": 13915008.0,
      "step": 1128
    },
    {
      "entropy": 1.3418332636356354,
      "epoch": 0.5945234333859927,
      "grad_norm": 0.306152880191803,
      "learning_rate": 0.000198584043526178,
      "loss": 0.21512483060359955,
      "mean_token_accuracy": 0.913850799202919,
      "num_tokens": 13927344.0,
      "step": 1129
    },
    {
      "entropy": 1.2864646911621094,
      "epoch": 0.5950500263296472,
      "grad_norm": 0.272443026304245,
      "learning_rate": 0.0001985801664609177,
      "loss": 0.19553211331367493,
      "mean_token_accuracy": 0.9215691983699799,
      "num_tokens": 13939680.0,
      "step": 1130
    },
    {
      "entropy": 1.3621837496757507,
      "epoch": 0.5955766192733017,
      "grad_norm": 0.30602437257766724,
      "learning_rate": 0.00019857628413716923,
      "loss": 0.2275530993938446,
      "mean_token_accuracy": 0.9101060777902603,
      "num_tokens": 13952016.0,
      "step": 1131
    },
    {
      "entropy": 1.3566639721393585,
      "epoch": 0.5961032122169563,
      "grad_norm": 0.3048968017101288,
      "learning_rate": 0.00019857239655516302,
      "loss": 0.20348000526428223,
      "mean_token_accuracy": 0.9176989495754242,
      "num_tokens": 13964352.0,
      "step": 1132
    },
    {
      "entropy": 1.3097732663154602,
      "epoch": 0.5966298051606108,
      "grad_norm": 0.2780429720878601,
      "learning_rate": 0.00019856850371512987,
      "loss": 0.21352221071720123,
      "mean_token_accuracy": 0.9180653989315033,
      "num_tokens": 13976688.0,
      "step": 1133
    },
    {
      "entropy": 1.3601540923118591,
      "epoch": 0.5971563981042654,
      "grad_norm": 0.28900468349456787,
      "learning_rate": 0.00019856460561730086,
      "loss": 0.21711350977420807,
      "mean_token_accuracy": 0.9169211685657501,
      "num_tokens": 13989024.0,
      "step": 1134
    },
    {
      "entropy": 1.3538919687271118,
      "epoch": 0.59768299104792,
      "grad_norm": 0.34999901056289673,
      "learning_rate": 0.00019856070226190745,
      "loss": 0.24132667481899261,
      "mean_token_accuracy": 0.9076129645109177,
      "num_tokens": 14001360.0,
      "step": 1135
    },
    {
      "entropy": 1.3285188376903534,
      "epoch": 0.5982095839915745,
      "grad_norm": 0.3087072968482971,
      "learning_rate": 0.00019855679364918125,
      "loss": 0.21430730819702148,
      "mean_token_accuracy": 0.911715641617775,
      "num_tokens": 14013696.0,
      "step": 1136
    },
    {
      "entropy": 1.3584640622138977,
      "epoch": 0.5987361769352291,
      "grad_norm": 0.34012794494628906,
      "learning_rate": 0.00019855287977935438,
      "loss": 0.22795352339744568,
      "mean_token_accuracy": 0.9065137207508087,
      "num_tokens": 14026032.0,
      "step": 1137
    },
    {
      "entropy": 1.337429255247116,
      "epoch": 0.5992627698788836,
      "grad_norm": 0.30005964636802673,
      "learning_rate": 0.0001985489606526592,
      "loss": 0.22258292138576508,
      "mean_token_accuracy": 0.9104499816894531,
      "num_tokens": 14038368.0,
      "step": 1138
    },
    {
      "entropy": 1.3470830917358398,
      "epoch": 0.5997893628225381,
      "grad_norm": 0.33669254183769226,
      "learning_rate": 0.0001985450362693283,
      "loss": 0.24716174602508545,
      "mean_token_accuracy": 0.9071019142866135,
      "num_tokens": 14050704.0,
      "step": 1139
    },
    {
      "entropy": 1.3219377994537354,
      "epoch": 0.6003159557661928,
      "grad_norm": 0.28407081961631775,
      "learning_rate": 0.0001985411066295947,
      "loss": 0.2014906108379364,
      "mean_token_accuracy": 0.9218366146087646,
      "num_tokens": 14063040.0,
      "step": 1140
    },
    {
      "entropy": 1.3444631695747375,
      "epoch": 0.6008425487098473,
      "grad_norm": 0.3870190978050232,
      "learning_rate": 0.00019853717173369163,
      "loss": 0.231156125664711,
      "mean_token_accuracy": 0.9088256508111954,
      "num_tokens": 14075376.0,
      "step": 1141
    },
    {
      "entropy": 1.308059424161911,
      "epoch": 0.6013691416535019,
      "grad_norm": 0.2819710373878479,
      "learning_rate": 0.0001985332315818527,
      "loss": 0.20217713713645935,
      "mean_token_accuracy": 0.9211951047182083,
      "num_tokens": 14087712.0,
      "step": 1142
    },
    {
      "entropy": 1.3603175282478333,
      "epoch": 0.6018957345971564,
      "grad_norm": 0.2916105389595032,
      "learning_rate": 0.00019852928617431186,
      "loss": 0.21151602268218994,
      "mean_token_accuracy": 0.9143334627151489,
      "num_tokens": 14100048.0,
      "step": 1143
    },
    {
      "entropy": 1.4045506119728088,
      "epoch": 0.6024223275408109,
      "grad_norm": 0.33014437556266785,
      "learning_rate": 0.00019852533551130324,
      "loss": 0.2289315164089203,
      "mean_token_accuracy": 0.9064499735832214,
      "num_tokens": 14112384.0,
      "step": 1144
    },
    {
      "entropy": 1.3394192457199097,
      "epoch": 0.6029489204844655,
      "grad_norm": 0.28447818756103516,
      "learning_rate": 0.00019852137959306144,
      "loss": 0.21860811114311218,
      "mean_token_accuracy": 0.9082520753145218,
      "num_tokens": 14124720.0,
      "step": 1145
    },
    {
      "entropy": 1.391217827796936,
      "epoch": 0.6034755134281201,
      "grad_norm": 0.28422555327415466,
      "learning_rate": 0.00019851741841982128,
      "loss": 0.19729138910770416,
      "mean_token_accuracy": 0.9245949536561966,
      "num_tokens": 14137056.0,
      "step": 1146
    },
    {
      "entropy": 1.3858129382133484,
      "epoch": 0.6040021063717746,
      "grad_norm": 0.29647180438041687,
      "learning_rate": 0.00019851345199181793,
      "loss": 0.22420856356620789,
      "mean_token_accuracy": 0.9095530211925507,
      "num_tokens": 14149392.0,
      "step": 1147
    },
    {
      "entropy": 1.3382080793380737,
      "epoch": 0.6045286993154292,
      "grad_norm": 0.3809289336204529,
      "learning_rate": 0.0001985094803092868,
      "loss": 0.23012962937355042,
      "mean_token_accuracy": 0.9103907644748688,
      "num_tokens": 14161728.0,
      "step": 1148
    },
    {
      "entropy": 1.252241313457489,
      "epoch": 0.6050552922590837,
      "grad_norm": 0.28908881545066833,
      "learning_rate": 0.00019850550337246366,
      "loss": 0.21694348752498627,
      "mean_token_accuracy": 0.9180784374475479,
      "num_tokens": 14174064.0,
      "step": 1149
    },
    {
      "entropy": 1.3115107417106628,
      "epoch": 0.6055818852027383,
      "grad_norm": 0.30919113755226135,
      "learning_rate": 0.00019850152118158472,
      "loss": 0.22491689026355743,
      "mean_token_accuracy": 0.9089210480451584,
      "num_tokens": 14186400.0,
      "step": 1150
    },
    {
      "entropy": 1.2686559855937958,
      "epoch": 0.6061084781463928,
      "grad_norm": 0.27780571579933167,
      "learning_rate": 0.0001984975337368862,
      "loss": 0.18563629686832428,
      "mean_token_accuracy": 0.9245605766773224,
      "num_tokens": 14198736.0,
      "step": 1151
    },
    {
      "entropy": 1.272880494594574,
      "epoch": 0.6066350710900474,
      "grad_norm": 0.31458860635757446,
      "learning_rate": 0.00019849354103860495,
      "loss": 0.21442924439907074,
      "mean_token_accuracy": 0.9128790646791458,
      "num_tokens": 14211072.0,
      "step": 1152
    },
    {
      "entropy": 1.2540790140628815,
      "epoch": 0.607161664033702,
      "grad_norm": 0.32044321298599243,
      "learning_rate": 0.000198489543086978,
      "loss": 0.20413605868816376,
      "mean_token_accuracy": 0.9179537445306778,
      "num_tokens": 14223408.0,
      "step": 1153
    },
    {
      "entropy": 1.3018059134483337,
      "epoch": 0.6076882569773565,
      "grad_norm": 0.3418496251106262,
      "learning_rate": 0.00019848553988224254,
      "loss": 0.2165253758430481,
      "mean_token_accuracy": 0.9125253260135651,
      "num_tokens": 14235744.0,
      "step": 1154
    },
    {
      "entropy": 1.2327319979667664,
      "epoch": 0.608214849921011,
      "grad_norm": 0.3597237467765808,
      "learning_rate": 0.0001984815314246364,
      "loss": 0.235613614320755,
      "mean_token_accuracy": 0.9087760597467422,
      "num_tokens": 14248080.0,
      "step": 1155
    },
    {
      "entropy": 1.3383413255214691,
      "epoch": 0.6087414428646656,
      "grad_norm": 0.36206018924713135,
      "learning_rate": 0.00019847751771439738,
      "loss": 0.2203393429517746,
      "mean_token_accuracy": 0.9124666899442673,
      "num_tokens": 14260416.0,
      "step": 1156
    },
    {
      "entropy": 1.2537713646888733,
      "epoch": 0.6092680358083201,
      "grad_norm": 0.314043253660202,
      "learning_rate": 0.00019847349875176383,
      "loss": 0.18635031580924988,
      "mean_token_accuracy": 0.924136608839035,
      "num_tokens": 14272752.0,
      "step": 1157
    },
    {
      "entropy": 1.2804813086986542,
      "epoch": 0.6097946287519748,
      "grad_norm": 0.3333822786808014,
      "learning_rate": 0.00019846947453697436,
      "loss": 0.24032224714756012,
      "mean_token_accuracy": 0.9084908664226532,
      "num_tokens": 14285088.0,
      "step": 1158
    },
    {
      "entropy": 1.2433983385562897,
      "epoch": 0.6103212216956293,
      "grad_norm": 0.3367704451084137,
      "learning_rate": 0.0001984654450702678,
      "loss": 0.21996420621871948,
      "mean_token_accuracy": 0.9129030555486679,
      "num_tokens": 14297424.0,
      "step": 1159
    },
    {
      "entropy": 1.2872591018676758,
      "epoch": 0.6108478146392838,
      "grad_norm": 0.3126697838306427,
      "learning_rate": 0.00019846141035188334,
      "loss": 0.22938281297683716,
      "mean_token_accuracy": 0.9084669649600983,
      "num_tokens": 14309760.0,
      "step": 1160
    },
    {
      "entropy": 1.2282376289367676,
      "epoch": 0.6113744075829384,
      "grad_norm": 0.3045027554035187,
      "learning_rate": 0.0001984573703820606,
      "loss": 0.21839705109596252,
      "mean_token_accuracy": 0.9129580110311508,
      "num_tokens": 14322096.0,
      "step": 1161
    },
    {
      "entropy": 1.296276867389679,
      "epoch": 0.6119010005265929,
      "grad_norm": 0.3050879240036011,
      "learning_rate": 0.00019845332516103933,
      "loss": 0.21910154819488525,
      "mean_token_accuracy": 0.9160716086626053,
      "num_tokens": 14334432.0,
      "step": 1162
    },
    {
      "entropy": 1.2552114129066467,
      "epoch": 0.6124275934702474,
      "grad_norm": 0.3606356978416443,
      "learning_rate": 0.0001984492746890597,
      "loss": 0.23291565477848053,
      "mean_token_accuracy": 0.9024920612573624,
      "num_tokens": 14346768.0,
      "step": 1163
    },
    {
      "entropy": 1.2350418269634247,
      "epoch": 0.6129541864139021,
      "grad_norm": 0.30588603019714355,
      "learning_rate": 0.00019844521896636213,
      "loss": 0.22299546003341675,
      "mean_token_accuracy": 0.9098654091358185,
      "num_tokens": 14359104.0,
      "step": 1164
    },
    {
      "entropy": 1.2168787717819214,
      "epoch": 0.6134807793575566,
      "grad_norm": 0.30036792159080505,
      "learning_rate": 0.00019844115799318743,
      "loss": 0.2024330198764801,
      "mean_token_accuracy": 0.9177137911319733,
      "num_tokens": 14371440.0,
      "step": 1165
    },
    {
      "entropy": 1.2549520134925842,
      "epoch": 0.6140073723012112,
      "grad_norm": 0.38305899500846863,
      "learning_rate": 0.00019843709176977666,
      "loss": 0.2291533648967743,
      "mean_token_accuracy": 0.912639394402504,
      "num_tokens": 14383776.0,
      "step": 1166
    },
    {
      "entropy": 1.235708236694336,
      "epoch": 0.6145339652448657,
      "grad_norm": 0.34467607736587524,
      "learning_rate": 0.0001984330202963712,
      "loss": 0.22410570085048676,
      "mean_token_accuracy": 0.9121714979410172,
      "num_tokens": 14396112.0,
      "step": 1167
    },
    {
      "entropy": 1.1345283687114716,
      "epoch": 0.6150605581885202,
      "grad_norm": 0.3024910092353821,
      "learning_rate": 0.0001984289435732127,
      "loss": 0.22202812135219574,
      "mean_token_accuracy": 0.913068488240242,
      "num_tokens": 14408448.0,
      "step": 1168
    },
    {
      "entropy": 1.154352843761444,
      "epoch": 0.6155871511321749,
      "grad_norm": 0.3543850779533386,
      "learning_rate": 0.00019842486160054327,
      "loss": 0.24754247069358826,
      "mean_token_accuracy": 0.9041948169469833,
      "num_tokens": 14420784.0,
      "step": 1169
    },
    {
      "entropy": 1.1256769597530365,
      "epoch": 0.6161137440758294,
      "grad_norm": 0.31246209144592285,
      "learning_rate": 0.00019842077437860517,
      "loss": 0.23024949431419373,
      "mean_token_accuracy": 0.9097237884998322,
      "num_tokens": 14433120.0,
      "step": 1170
    },
    {
      "entropy": 1.2304538488388062,
      "epoch": 0.616640337019484,
      "grad_norm": 0.35418546199798584,
      "learning_rate": 0.00019841668190764106,
      "loss": 0.23257943987846375,
      "mean_token_accuracy": 0.9058277457952499,
      "num_tokens": 14445456.0,
      "step": 1171
    },
    {
      "entropy": 1.1736994087696075,
      "epoch": 0.6171669299631385,
      "grad_norm": 0.29421883821487427,
      "learning_rate": 0.0001984125841878939,
      "loss": 0.2265184223651886,
      "mean_token_accuracy": 0.9127582311630249,
      "num_tokens": 14457792.0,
      "step": 1172
    },
    {
      "entropy": 1.210179328918457,
      "epoch": 0.617693522906793,
      "grad_norm": 0.3278239667415619,
      "learning_rate": 0.0001984084812196069,
      "loss": 0.2220735400915146,
      "mean_token_accuracy": 0.9109156280755997,
      "num_tokens": 14470128.0,
      "step": 1173
    },
    {
      "entropy": 1.1588793694972992,
      "epoch": 0.6182201158504476,
      "grad_norm": 0.27815452218055725,
      "learning_rate": 0.00019840437300302366,
      "loss": 0.22251342236995697,
      "mean_token_accuracy": 0.9095259606838226,
      "num_tokens": 14482464.0,
      "step": 1174
    },
    {
      "entropy": 1.1791687309741974,
      "epoch": 0.6187467087941022,
      "grad_norm": 0.2694774270057678,
      "learning_rate": 0.00019840025953838804,
      "loss": 0.2061459869146347,
      "mean_token_accuracy": 0.9203435033559799,
      "num_tokens": 14494800.0,
      "step": 1175
    },
    {
      "entropy": 1.1984158754348755,
      "epoch": 0.6192733017377567,
      "grad_norm": 0.29139313101768494,
      "learning_rate": 0.00019839614082594424,
      "loss": 0.22982501983642578,
      "mean_token_accuracy": 0.911991760134697,
      "num_tokens": 14507136.0,
      "step": 1176
    },
    {
      "entropy": 1.1957634687423706,
      "epoch": 0.6197998946814113,
      "grad_norm": 0.2952525019645691,
      "learning_rate": 0.0001983920168659368,
      "loss": 0.20150825381278992,
      "mean_token_accuracy": 0.9190460294485092,
      "num_tokens": 14519472.0,
      "step": 1177
    },
    {
      "entropy": 1.2356001436710358,
      "epoch": 0.6203264876250658,
      "grad_norm": 0.3485545516014099,
      "learning_rate": 0.0001983878876586105,
      "loss": 0.22730165719985962,
      "mean_token_accuracy": 0.9079895615577698,
      "num_tokens": 14531808.0,
      "step": 1178
    },
    {
      "entropy": 1.2191669344902039,
      "epoch": 0.6208530805687204,
      "grad_norm": 0.31337442994117737,
      "learning_rate": 0.00019838375320421046,
      "loss": 0.2090841382741928,
      "mean_token_accuracy": 0.9179579317569733,
      "num_tokens": 14544144.0,
      "step": 1179
    },
    {
      "entropy": 1.1627593338489532,
      "epoch": 0.6213796735123749,
      "grad_norm": 0.3373863995075226,
      "learning_rate": 0.00019837961350298213,
      "loss": 0.22159543633460999,
      "mean_token_accuracy": 0.9035765677690506,
      "num_tokens": 14556480.0,
      "step": 1180
    },
    {
      "entropy": 1.244479089975357,
      "epoch": 0.6219062664560295,
      "grad_norm": 0.2939426898956299,
      "learning_rate": 0.0001983754685551713,
      "loss": 0.20940116047859192,
      "mean_token_accuracy": 0.9177102446556091,
      "num_tokens": 14568816.0,
      "step": 1181
    },
    {
      "entropy": 1.2316089570522308,
      "epoch": 0.6224328593996841,
      "grad_norm": 0.3053775727748871,
      "learning_rate": 0.00019837131836102396,
      "loss": 0.21764273941516876,
      "mean_token_accuracy": 0.9150272458791733,
      "num_tokens": 14581152.0,
      "step": 1182
    },
    {
      "entropy": 1.2752362787723541,
      "epoch": 0.6229594523433386,
      "grad_norm": 0.3304329812526703,
      "learning_rate": 0.00019836716292078647,
      "loss": 0.23420026898384094,
      "mean_token_accuracy": 0.9085260033607483,
      "num_tokens": 14593488.0,
      "step": 1183
    },
    {
      "entropy": 1.2659085094928741,
      "epoch": 0.6234860452869931,
      "grad_norm": 0.3893950879573822,
      "learning_rate": 0.00019836300223470562,
      "loss": 0.2608282268047333,
      "mean_token_accuracy": 0.9021032303571701,
      "num_tokens": 14605824.0,
      "step": 1184
    },
    {
      "entropy": 1.3032302856445312,
      "epoch": 0.6240126382306477,
      "grad_norm": 0.3007699251174927,
      "learning_rate": 0.00019835883630302836,
      "loss": 0.20315539836883545,
      "mean_token_accuracy": 0.9167050868272781,
      "num_tokens": 14618160.0,
      "step": 1185
    },
    {
      "entropy": 1.2084311544895172,
      "epoch": 0.6245392311743022,
      "grad_norm": 0.32948681712150574,
      "learning_rate": 0.00019835466512600197,
      "loss": 0.23484407365322113,
      "mean_token_accuracy": 0.9104520976543427,
      "num_tokens": 14630496.0,
      "step": 1186
    },
    {
      "entropy": 1.2006751894950867,
      "epoch": 0.6250658241179569,
      "grad_norm": 0.2981775104999542,
      "learning_rate": 0.00019835048870387405,
      "loss": 0.22286085784435272,
      "mean_token_accuracy": 0.9095976203680038,
      "num_tokens": 14642832.0,
      "step": 1187
    },
    {
      "entropy": 1.3096867799758911,
      "epoch": 0.6255924170616114,
      "grad_norm": 0.3300210237503052,
      "learning_rate": 0.0001983463070368926,
      "loss": 0.22102200984954834,
      "mean_token_accuracy": 0.9119229018688202,
      "num_tokens": 14655168.0,
      "step": 1188
    },
    {
      "entropy": 1.2840961813926697,
      "epoch": 0.6261190100052659,
      "grad_norm": 0.31997406482696533,
      "learning_rate": 0.00019834212012530579,
      "loss": 0.23797491192817688,
      "mean_token_accuracy": 0.9043692797422409,
      "num_tokens": 14667504.0,
      "step": 1189
    },
    {
      "entropy": 1.2393836975097656,
      "epoch": 0.6266456029489205,
      "grad_norm": 0.30698251724243164,
      "learning_rate": 0.00019833792796936222,
      "loss": 0.20287004113197327,
      "mean_token_accuracy": 0.9166252613067627,
      "num_tokens": 14679840.0,
      "step": 1190
    },
    {
      "entropy": 1.3439719676971436,
      "epoch": 0.627172195892575,
      "grad_norm": 0.32229262590408325,
      "learning_rate": 0.00019833373056931072,
      "loss": 0.23332646489143372,
      "mean_token_accuracy": 0.9090562462806702,
      "num_tokens": 14692176.0,
      "step": 1191
    },
    {
      "entropy": 1.2390953600406647,
      "epoch": 0.6276987888362296,
      "grad_norm": 0.3261661231517792,
      "learning_rate": 0.00019832952792540054,
      "loss": 0.22384727001190186,
      "mean_token_accuracy": 0.9121635258197784,
      "num_tokens": 14704512.0,
      "step": 1192
    },
    {
      "entropy": 1.3028371036052704,
      "epoch": 0.6282253817798842,
      "grad_norm": 0.31324324011802673,
      "learning_rate": 0.00019832532003788108,
      "loss": 0.2185819447040558,
      "mean_token_accuracy": 0.9167534857988358,
      "num_tokens": 14716848.0,
      "step": 1193
    },
    {
      "entropy": 1.1885470747947693,
      "epoch": 0.6287519747235387,
      "grad_norm": 0.46663784980773926,
      "learning_rate": 0.00019832110690700213,
      "loss": 0.21501871943473816,
      "mean_token_accuracy": 0.9109776020050049,
      "num_tokens": 14729184.0,
      "step": 1194
    },
    {
      "entropy": 1.2453204691410065,
      "epoch": 0.6292785676671933,
      "grad_norm": 0.3183780014514923,
      "learning_rate": 0.00019831688853301388,
      "loss": 0.22543299198150635,
      "mean_token_accuracy": 0.910496860742569,
      "num_tokens": 14741520.0,
      "step": 1195
    },
    {
      "entropy": 1.2583057284355164,
      "epoch": 0.6298051606108478,
      "grad_norm": 0.3381344676017761,
      "learning_rate": 0.0001983126649161667,
      "loss": 0.214624285697937,
      "mean_token_accuracy": 0.9137195497751236,
      "num_tokens": 14753856.0,
      "step": 1196
    },
    {
      "entropy": 1.2429248094558716,
      "epoch": 0.6303317535545023,
      "grad_norm": 0.30227339267730713,
      "learning_rate": 0.00019830843605671131,
      "loss": 0.2062695324420929,
      "mean_token_accuracy": 0.9234135448932648,
      "num_tokens": 14766192.0,
      "step": 1197
    },
    {
      "entropy": 1.220680981874466,
      "epoch": 0.630858346498157,
      "grad_norm": 0.3083505630493164,
      "learning_rate": 0.00019830420195489877,
      "loss": 0.2090727537870407,
      "mean_token_accuracy": 0.9170975387096405,
      "num_tokens": 14778528.0,
      "step": 1198
    },
    {
      "entropy": 1.248794138431549,
      "epoch": 0.6313849394418115,
      "grad_norm": 0.3802277743816376,
      "learning_rate": 0.00019829996261098046,
      "loss": 0.2367159128189087,
      "mean_token_accuracy": 0.9018937945365906,
      "num_tokens": 14790864.0,
      "step": 1199
    },
    {
      "entropy": 1.3259137272834778,
      "epoch": 0.631911532385466,
      "grad_norm": 0.414235919713974,
      "learning_rate": 0.00019829571802520803,
      "loss": 0.2376767247915268,
      "mean_token_accuracy": 0.9016633927822113,
      "num_tokens": 14803200.0,
      "step": 1200
    },
    {
      "entropy": 1.3644030392169952,
      "epoch": 0.6324381253291206,
      "grad_norm": 0.3521367013454437,
      "learning_rate": 0.0001982914681978334,
      "loss": 0.23597122728824615,
      "mean_token_accuracy": 0.9043200761079788,
      "num_tokens": 14815536.0,
      "step": 1201
    },
    {
      "entropy": 1.3481078147888184,
      "epoch": 0.6329647182727751,
      "grad_norm": 0.2994113266468048,
      "learning_rate": 0.000198287213129109,
      "loss": 0.21383647620677948,
      "mean_token_accuracy": 0.9209251701831818,
      "num_tokens": 14827872.0,
      "step": 1202
    },
    {
      "entropy": 1.3808801472187042,
      "epoch": 0.6334913112164297,
      "grad_norm": 0.2982345223426819,
      "learning_rate": 0.00019828295281928727,
      "loss": 0.2155206948518753,
      "mean_token_accuracy": 0.9116237312555313,
      "num_tokens": 14840208.0,
      "step": 1203
    },
    {
      "entropy": 1.418733924627304,
      "epoch": 0.6340179041600843,
      "grad_norm": 0.3023967742919922,
      "learning_rate": 0.00019827868726862117,
      "loss": 0.23393575847148895,
      "mean_token_accuracy": 0.9095605909824371,
      "num_tokens": 14852544.0,
      "step": 1204
    },
    {
      "entropy": 1.5514957010746002,
      "epoch": 0.6345444971037388,
      "grad_norm": 0.40304750204086304,
      "learning_rate": 0.00019827441647736393,
      "loss": 0.26255497336387634,
      "mean_token_accuracy": 0.9011336117982864,
      "num_tokens": 14864880.0,
      "step": 1205
    },
    {
      "entropy": 1.4219785034656525,
      "epoch": 0.6350710900473934,
      "grad_norm": 0.291746586561203,
      "learning_rate": 0.00019827014044576912,
      "loss": 0.21620029211044312,
      "mean_token_accuracy": 0.9115499705076218,
      "num_tokens": 14877216.0,
      "step": 1206
    },
    {
      "entropy": 1.3420428931713104,
      "epoch": 0.6355976829910479,
      "grad_norm": 0.3108786940574646,
      "learning_rate": 0.0001982658591740905,
      "loss": 0.23703286051750183,
      "mean_token_accuracy": 0.9016107618808746,
      "num_tokens": 14889552.0,
      "step": 1207
    },
    {
      "entropy": 1.4392370581626892,
      "epoch": 0.6361242759347024,
      "grad_norm": 0.27034372091293335,
      "learning_rate": 0.00019826157266258232,
      "loss": 0.21097946166992188,
      "mean_token_accuracy": 0.9190766364336014,
      "num_tokens": 14901888.0,
      "step": 1208
    },
    {
      "entropy": 1.4509558379650116,
      "epoch": 0.636650868878357,
      "grad_norm": 0.29659944772720337,
      "learning_rate": 0.000198257280911499,
      "loss": 0.22469229996204376,
      "mean_token_accuracy": 0.913391038775444,
      "num_tokens": 14914224.0,
      "step": 1209
    },
    {
      "entropy": 1.3904311954975128,
      "epoch": 0.6371774618220116,
      "grad_norm": 0.2719598412513733,
      "learning_rate": 0.00019825298392109529,
      "loss": 0.18848726153373718,
      "mean_token_accuracy": 0.9262688905000687,
      "num_tokens": 14926560.0,
      "step": 1210
    },
    {
      "entropy": 1.4359396994113922,
      "epoch": 0.6377040547656662,
      "grad_norm": 0.3138201832771301,
      "learning_rate": 0.00019824868169162631,
      "loss": 0.24162685871124268,
      "mean_token_accuracy": 0.9020742177963257,
      "num_tokens": 14938896.0,
      "step": 1211
    },
    {
      "entropy": 1.3890084326267242,
      "epoch": 0.6382306477093207,
      "grad_norm": 0.2986134886741638,
      "learning_rate": 0.00019824437422334744,
      "loss": 0.22484566271305084,
      "mean_token_accuracy": 0.9077206701040268,
      "num_tokens": 14951232.0,
      "step": 1212
    },
    {
      "entropy": 1.378806233406067,
      "epoch": 0.6387572406529752,
      "grad_norm": 0.29134130477905273,
      "learning_rate": 0.00019824006151651443,
      "loss": 0.2171209454536438,
      "mean_token_accuracy": 0.9133529961109161,
      "num_tokens": 14963568.0,
      "step": 1213
    },
    {
      "entropy": 1.363432914018631,
      "epoch": 0.6392838335966298,
      "grad_norm": 0.3041214346885681,
      "learning_rate": 0.00019823574357138323,
      "loss": 0.24093735218048096,
      "mean_token_accuracy": 0.903191551566124,
      "num_tokens": 14975904.0,
      "step": 1214
    },
    {
      "entropy": 1.432945430278778,
      "epoch": 0.6398104265402843,
      "grad_norm": 0.30195045471191406,
      "learning_rate": 0.00019823142038821027,
      "loss": 0.20907755196094513,
      "mean_token_accuracy": 0.9161103218793869,
      "num_tokens": 14988240.0,
      "step": 1215
    },
    {
      "entropy": 1.3210029900074005,
      "epoch": 0.640337019483939,
      "grad_norm": 0.29575979709625244,
      "learning_rate": 0.00019822709196725208,
      "loss": 0.21090644598007202,
      "mean_token_accuracy": 0.9149498790502548,
      "num_tokens": 15000576.0,
      "step": 1216
    },
    {
      "entropy": 1.3699910044670105,
      "epoch": 0.6408636124275935,
      "grad_norm": 0.2982396185398102,
      "learning_rate": 0.00019822275830876568,
      "loss": 0.2054290473461151,
      "mean_token_accuracy": 0.9217698276042938,
      "num_tokens": 15012912.0,
      "step": 1217
    },
    {
      "entropy": 1.41866934299469,
      "epoch": 0.641390205371248,
      "grad_norm": 0.3208104968070984,
      "learning_rate": 0.00019821841941300834,
      "loss": 0.24424780905246735,
      "mean_token_accuracy": 0.9118457287549973,
      "num_tokens": 15025248.0,
      "step": 1218
    },
    {
      "entropy": 1.399036020040512,
      "epoch": 0.6419167983149026,
      "grad_norm": 0.3094969689846039,
      "learning_rate": 0.00019821407528023758,
      "loss": 0.23681235313415527,
      "mean_token_accuracy": 0.9031997323036194,
      "num_tokens": 15037584.0,
      "step": 1219
    },
    {
      "entropy": 1.3915925920009613,
      "epoch": 0.6424433912585571,
      "grad_norm": 0.2973760962486267,
      "learning_rate": 0.00019820972591071131,
      "loss": 0.19640693068504333,
      "mean_token_accuracy": 0.9181762486696243,
      "num_tokens": 15049920.0,
      "step": 1220
    },
    {
      "entropy": 1.4664425551891327,
      "epoch": 0.6429699842022117,
      "grad_norm": 0.7199912667274475,
      "learning_rate": 0.0001982053713046878,
      "loss": 0.2480940818786621,
      "mean_token_accuracy": 0.9039055407047272,
      "num_tokens": 15062256.0,
      "step": 1221
    },
    {
      "entropy": 1.4178178012371063,
      "epoch": 0.6434965771458663,
      "grad_norm": 0.3677431643009186,
      "learning_rate": 0.00019820101146242547,
      "loss": 0.22347916662693024,
      "mean_token_accuracy": 0.9104437232017517,
      "num_tokens": 15074592.0,
      "step": 1222
    },
    {
      "entropy": 1.3942118883132935,
      "epoch": 0.6440231700895208,
      "grad_norm": 0.3363230228424072,
      "learning_rate": 0.00019819664638418314,
      "loss": 0.23758456110954285,
      "mean_token_accuracy": 0.9113318026065826,
      "num_tokens": 15086928.0,
      "step": 1223
    },
    {
      "entropy": 1.3070717453956604,
      "epoch": 0.6445497630331753,
      "grad_norm": 0.3119392693042755,
      "learning_rate": 0.00019819227607021996,
      "loss": 0.1808784306049347,
      "mean_token_accuracy": 0.9277928322553635,
      "num_tokens": 15099264.0,
      "step": 1224
    },
    {
      "entropy": 1.3485881388187408,
      "epoch": 0.6450763559768299,
      "grad_norm": 0.2953029274940491,
      "learning_rate": 0.00019818790052079542,
      "loss": 0.21690060198307037,
      "mean_token_accuracy": 0.9107033312320709,
      "num_tokens": 15111600.0,
      "step": 1225
    },
    {
      "entropy": 1.3310377597808838,
      "epoch": 0.6456029489204844,
      "grad_norm": 0.30317792296409607,
      "learning_rate": 0.0001981835197361692,
      "loss": 0.21382620930671692,
      "mean_token_accuracy": 0.9134241938591003,
      "num_tokens": 15123936.0,
      "step": 1226
    },
    {
      "entropy": 1.3547466397285461,
      "epoch": 0.6461295418641391,
      "grad_norm": 0.31380537152290344,
      "learning_rate": 0.00019817913371660136,
      "loss": 0.23309272527694702,
      "mean_token_accuracy": 0.9060655981302261,
      "num_tokens": 15136272.0,
      "step": 1227
    },
    {
      "entropy": 1.359354555606842,
      "epoch": 0.6466561348077936,
      "grad_norm": 0.3240332007408142,
      "learning_rate": 0.00019817474246235233,
      "loss": 0.2206256240606308,
      "mean_token_accuracy": 0.9170451760292053,
      "num_tokens": 15148608.0,
      "step": 1228
    },
    {
      "entropy": 1.3810060322284698,
      "epoch": 0.6471827277514481,
      "grad_norm": 0.28998929262161255,
      "learning_rate": 0.00019817034597368276,
      "loss": 0.20593759417533875,
      "mean_token_accuracy": 0.9143964052200317,
      "num_tokens": 15160944.0,
      "step": 1229
    },
    {
      "entropy": 1.3853793740272522,
      "epoch": 0.6477093206951027,
      "grad_norm": 0.3167085349559784,
      "learning_rate": 0.00019816594425085365,
      "loss": 0.214578315615654,
      "mean_token_accuracy": 0.9154220819473267,
      "num_tokens": 15173280.0,
      "step": 1230
    },
    {
      "entropy": 1.3485651314258575,
      "epoch": 0.6482359136387572,
      "grad_norm": 0.32996588945388794,
      "learning_rate": 0.00019816153729412627,
      "loss": 0.227751225233078,
      "mean_token_accuracy": 0.9076825082302094,
      "num_tokens": 15185616.0,
      "step": 1231
    },
    {
      "entropy": 1.2957139611244202,
      "epoch": 0.6487625065824117,
      "grad_norm": 0.2826898396015167,
      "learning_rate": 0.0001981571251037623,
      "loss": 0.2006268948316574,
      "mean_token_accuracy": 0.9202533811330795,
      "num_tokens": 15197952.0,
      "step": 1232
    },
    {
      "entropy": 1.42320716381073,
      "epoch": 0.6492890995260664,
      "grad_norm": 0.3413523733615875,
      "learning_rate": 0.00019815270768002357,
      "loss": 0.2385195791721344,
      "mean_token_accuracy": 0.9027537554502487,
      "num_tokens": 15210288.0,
      "step": 1233
    },
    {
      "entropy": 1.3331094086170197,
      "epoch": 0.6498156924697209,
      "grad_norm": 0.3069290220737457,
      "learning_rate": 0.00019814828502317245,
      "loss": 0.2170781046152115,
      "mean_token_accuracy": 0.9153832644224167,
      "num_tokens": 15222624.0,
      "step": 1234
    },
    {
      "entropy": 1.365331083536148,
      "epoch": 0.6503422854133755,
      "grad_norm": 0.2763271629810333,
      "learning_rate": 0.0001981438571334714,
      "loss": 0.21132177114486694,
      "mean_token_accuracy": 0.9158383458852768,
      "num_tokens": 15234960.0,
      "step": 1235
    },
    {
      "entropy": 1.3776979446411133,
      "epoch": 0.65086887835703,
      "grad_norm": 0.3232009708881378,
      "learning_rate": 0.0001981394240111833,
      "loss": 0.23016487061977386,
      "mean_token_accuracy": 0.9127575606107712,
      "num_tokens": 15247296.0,
      "step": 1236
    },
    {
      "entropy": 1.42231023311615,
      "epoch": 0.6513954713006845,
      "grad_norm": 0.2981938421726227,
      "learning_rate": 0.00019813498565657127,
      "loss": 0.22353443503379822,
      "mean_token_accuracy": 0.9140705019235611,
      "num_tokens": 15259632.0,
      "step": 1237
    },
    {
      "entropy": 1.4192762076854706,
      "epoch": 0.6519220642443391,
      "grad_norm": 0.30204325914382935,
      "learning_rate": 0.00019813054206989884,
      "loss": 0.22629618644714355,
      "mean_token_accuracy": 0.9069414585828781,
      "num_tokens": 15271968.0,
      "step": 1238
    },
    {
      "entropy": 1.4479241073131561,
      "epoch": 0.6524486571879937,
      "grad_norm": 0.2993687689304352,
      "learning_rate": 0.00019812609325142982,
      "loss": 0.22675678133964539,
      "mean_token_accuracy": 0.9094991534948349,
      "num_tokens": 15284304.0,
      "step": 1239
    },
    {
      "entropy": 1.4620944261550903,
      "epoch": 0.6529752501316483,
      "grad_norm": 0.3134925961494446,
      "learning_rate": 0.00019812163920142827,
      "loss": 0.22805798053741455,
      "mean_token_accuracy": 0.9107698351144791,
      "num_tokens": 15296640.0,
      "step": 1240
    },
    {
      "entropy": 1.495926171541214,
      "epoch": 0.6535018430753028,
      "grad_norm": 0.3140438199043274,
      "learning_rate": 0.00019811717992015862,
      "loss": 0.2086748480796814,
      "mean_token_accuracy": 0.9102585017681122,
      "num_tokens": 15308976.0,
      "step": 1241
    },
    {
      "entropy": 1.4446808993816376,
      "epoch": 0.6540284360189573,
      "grad_norm": 0.30744868516921997,
      "learning_rate": 0.00019811271540788556,
      "loss": 0.22943398356437683,
      "mean_token_accuracy": 0.9074350297451019,
      "num_tokens": 15321312.0,
      "step": 1242
    },
    {
      "entropy": 1.5100261270999908,
      "epoch": 0.6545550289626119,
      "grad_norm": 0.3204497694969177,
      "learning_rate": 0.00019810824566487417,
      "loss": 0.22844550013542175,
      "mean_token_accuracy": 0.9073818922042847,
      "num_tokens": 15333648.0,
      "step": 1243
    },
    {
      "entropy": 1.4075362086296082,
      "epoch": 0.6550816219062665,
      "grad_norm": 0.32697567343711853,
      "learning_rate": 0.00019810377069138976,
      "loss": 0.23277214169502258,
      "mean_token_accuracy": 0.9124523550271988,
      "num_tokens": 15345984.0,
      "step": 1244
    },
    {
      "entropy": 1.474784404039383,
      "epoch": 0.655608214849921,
      "grad_norm": 0.31703367829322815,
      "learning_rate": 0.000198099290487698,
      "loss": 0.2128317952156067,
      "mean_token_accuracy": 0.9103437811136246,
      "num_tokens": 15358320.0,
      "step": 1245
    },
    {
      "entropy": 1.3723844587802887,
      "epoch": 0.6561348077935756,
      "grad_norm": 0.3136475384235382,
      "learning_rate": 0.0001980948050540648,
      "loss": 0.2106158286333084,
      "mean_token_accuracy": 0.9139569699764252,
      "num_tokens": 15370656.0,
      "step": 1246
    },
    {
      "entropy": 1.4297202229499817,
      "epoch": 0.6566614007372301,
      "grad_norm": 0.31566086411476135,
      "learning_rate": 0.00019809031439075653,
      "loss": 0.22323954105377197,
      "mean_token_accuracy": 0.912299782037735,
      "num_tokens": 15382992.0,
      "step": 1247
    },
    {
      "entropy": 1.404792070388794,
      "epoch": 0.6571879936808847,
      "grad_norm": 0.30582812428474426,
      "learning_rate": 0.00019808581849803972,
      "loss": 0.19535206258296967,
      "mean_token_accuracy": 0.926054835319519,
      "num_tokens": 15395328.0,
      "step": 1248
    },
    {
      "entropy": 1.365869253873825,
      "epoch": 0.6577145866245392,
      "grad_norm": 0.32905054092407227,
      "learning_rate": 0.00019808131737618126,
      "loss": 0.206635981798172,
      "mean_token_accuracy": 0.9179205745458603,
      "num_tokens": 15407664.0,
      "step": 1249
    },
    {
      "entropy": 1.4380927085876465,
      "epoch": 0.6582411795681938,
      "grad_norm": 0.3363628685474396,
      "learning_rate": 0.00019807681102544834,
      "loss": 0.23559972643852234,
      "mean_token_accuracy": 0.9073036760091782,
      "num_tokens": 15420000.0,
      "step": 1250
    },
    {
      "entropy": 1.4528064727783203,
      "epoch": 0.6587677725118484,
      "grad_norm": 0.3616321384906769,
      "learning_rate": 0.00019807229944610854,
      "loss": 0.23041245341300964,
      "mean_token_accuracy": 0.9012827575206757,
      "num_tokens": 15432336.0,
      "step": 1251
    },
    {
      "entropy": 1.4460172951221466,
      "epoch": 0.6592943654555029,
      "grad_norm": 0.305955171585083,
      "learning_rate": 0.00019806778263842964,
      "loss": 0.21666109561920166,
      "mean_token_accuracy": 0.9198573529720306,
      "num_tokens": 15444672.0,
      "step": 1252
    },
    {
      "entropy": 1.500756323337555,
      "epoch": 0.6598209583991574,
      "grad_norm": 0.29015445709228516,
      "learning_rate": 0.00019806326060267977,
      "loss": 0.21642570197582245,
      "mean_token_accuracy": 0.9154843240976334,
      "num_tokens": 15457008.0,
      "step": 1253
    },
    {
      "entropy": 1.4731870293617249,
      "epoch": 0.660347551342812,
      "grad_norm": 0.3006648123264313,
      "learning_rate": 0.00019805873333912739,
      "loss": 0.20880235731601715,
      "mean_token_accuracy": 0.924048438668251,
      "num_tokens": 15469344.0,
      "step": 1254
    },
    {
      "entropy": 1.5191810727119446,
      "epoch": 0.6608741442864665,
      "grad_norm": 0.290721595287323,
      "learning_rate": 0.00019805420084804124,
      "loss": 0.21848249435424805,
      "mean_token_accuracy": 0.9141736477613449,
      "num_tokens": 15481680.0,
      "step": 1255
    },
    {
      "entropy": 1.521433562040329,
      "epoch": 0.6614007372301212,
      "grad_norm": 0.3009195327758789,
      "learning_rate": 0.00019804966312969046,
      "loss": 0.20721524953842163,
      "mean_token_accuracy": 0.9165780544281006,
      "num_tokens": 15494016.0,
      "step": 1256
    },
    {
      "entropy": 1.5230248272418976,
      "epoch": 0.6619273301737757,
      "grad_norm": 0.2907300293445587,
      "learning_rate": 0.00019804512018434432,
      "loss": 0.2220079004764557,
      "mean_token_accuracy": 0.9127085506916046,
      "num_tokens": 15506352.0,
      "step": 1257
    },
    {
      "entropy": 1.4415357112884521,
      "epoch": 0.6624539231174302,
      "grad_norm": 0.27609995007514954,
      "learning_rate": 0.00019804057201227259,
      "loss": 0.19916732609272003,
      "mean_token_accuracy": 0.9211323857307434,
      "num_tokens": 15518688.0,
      "step": 1258
    },
    {
      "entropy": 1.5071959793567657,
      "epoch": 0.6629805160610848,
      "grad_norm": 0.3255380690097809,
      "learning_rate": 0.0001980360186137452,
      "loss": 0.23832152783870697,
      "mean_token_accuracy": 0.9030174165964127,
      "num_tokens": 15531024.0,
      "step": 1259
    },
    {
      "entropy": 1.544082134962082,
      "epoch": 0.6635071090047393,
      "grad_norm": 0.31669536232948303,
      "learning_rate": 0.00019803145998903253,
      "loss": 0.2193489968776703,
      "mean_token_accuracy": 0.9140938073396683,
      "num_tokens": 15543360.0,
      "step": 1260
    },
    {
      "entropy": 1.5186286568641663,
      "epoch": 0.6640337019483938,
      "grad_norm": 0.3087945282459259,
      "learning_rate": 0.00019802689613840514,
      "loss": 0.2145516723394394,
      "mean_token_accuracy": 0.9134328663349152,
      "num_tokens": 15555696.0,
      "step": 1261
    },
    {
      "entropy": 1.4552082121372223,
      "epoch": 0.6645602948920485,
      "grad_norm": 0.3144809901714325,
      "learning_rate": 0.000198022327062134,
      "loss": 0.22542014718055725,
      "mean_token_accuracy": 0.9080185741186142,
      "num_tokens": 15568032.0,
      "step": 1262
    },
    {
      "entropy": 1.4759101569652557,
      "epoch": 0.665086887835703,
      "grad_norm": 0.28055956959724426,
      "learning_rate": 0.00019801775276049035,
      "loss": 0.19411805272102356,
      "mean_token_accuracy": 0.9219205677509308,
      "num_tokens": 15580368.0,
      "step": 1263
    },
    {
      "entropy": 1.5172317326068878,
      "epoch": 0.6656134807793576,
      "grad_norm": 0.28847187757492065,
      "learning_rate": 0.00019801317323374574,
      "loss": 0.1965804398059845,
      "mean_token_accuracy": 0.9217233210802078,
      "num_tokens": 15592704.0,
      "step": 1264
    },
    {
      "entropy": 1.4631735682487488,
      "epoch": 0.6661400737230121,
      "grad_norm": 0.3277401924133301,
      "learning_rate": 0.000198008588482172,
      "loss": 0.23770709335803986,
      "mean_token_accuracy": 0.9039840698242188,
      "num_tokens": 15605040.0,
      "step": 1265
    },
    {
      "entropy": 1.5213152170181274,
      "epoch": 0.6666666666666666,
      "grad_norm": 0.31430521607398987,
      "learning_rate": 0.00019800399850604129,
      "loss": 0.23058706521987915,
      "mean_token_accuracy": 0.9074577987194061,
      "num_tokens": 15617376.0,
      "step": 1266
    },
    {
      "entropy": 1.4705300331115723,
      "epoch": 0.6671932596103213,
      "grad_norm": 0.3256817162036896,
      "learning_rate": 0.0001979994033056261,
      "loss": 0.25320321321487427,
      "mean_token_accuracy": 0.9007576107978821,
      "num_tokens": 15629712.0,
      "step": 1267
    },
    {
      "entropy": 1.530697613954544,
      "epoch": 0.6677198525539758,
      "grad_norm": 0.30975228548049927,
      "learning_rate": 0.00019799480288119927,
      "loss": 0.23081722855567932,
      "mean_token_accuracy": 0.9080140292644501,
      "num_tokens": 15642048.0,
      "step": 1268
    },
    {
      "entropy": 1.495768964290619,
      "epoch": 0.6682464454976303,
      "grad_norm": 0.29344475269317627,
      "learning_rate": 0.00019799019723303388,
      "loss": 0.2015610784292221,
      "mean_token_accuracy": 0.9178611040115356,
      "num_tokens": 15654384.0,
      "step": 1269
    },
    {
      "entropy": 1.5803856551647186,
      "epoch": 0.6687730384412849,
      "grad_norm": 0.3023964464664459,
      "learning_rate": 0.00019798558636140333,
      "loss": 0.2174011915922165,
      "mean_token_accuracy": 0.9140703827142715,
      "num_tokens": 15666720.0,
      "step": 1270
    },
    {
      "entropy": 1.4897505939006805,
      "epoch": 0.6692996313849394,
      "grad_norm": 0.2858726978302002,
      "learning_rate": 0.0001979809702665813,
      "loss": 0.20024171471595764,
      "mean_token_accuracy": 0.9191962629556656,
      "num_tokens": 15679056.0,
      "step": 1271
    },
    {
      "entropy": 1.5215353667736053,
      "epoch": 0.669826224328594,
      "grad_norm": 0.27068790793418884,
      "learning_rate": 0.00019797634894884184,
      "loss": 0.21988946199417114,
      "mean_token_accuracy": 0.9098561853170395,
      "num_tokens": 15691392.0,
      "step": 1272
    },
    {
      "entropy": 1.5361709892749786,
      "epoch": 0.6703528172722486,
      "grad_norm": 0.3324735164642334,
      "learning_rate": 0.00019797172240845935,
      "loss": 0.2674825191497803,
      "mean_token_accuracy": 0.8955648243427277,
      "num_tokens": 15703728.0,
      "step": 1273
    },
    {
      "entropy": 1.5240019261837006,
      "epoch": 0.6708794102159031,
      "grad_norm": 0.3050387501716614,
      "learning_rate": 0.00019796709064570842,
      "loss": 0.22774918377399445,
      "mean_token_accuracy": 0.9072832018136978,
      "num_tokens": 15716064.0,
      "step": 1274
    },
    {
      "entropy": 1.4996045529842377,
      "epoch": 0.6714060031595577,
      "grad_norm": 0.2832599878311157,
      "learning_rate": 0.00019796245366086404,
      "loss": 0.21256539225578308,
      "mean_token_accuracy": 0.9113875329494476,
      "num_tokens": 15728400.0,
      "step": 1275
    },
    {
      "entropy": 1.537344366312027,
      "epoch": 0.6719325961032122,
      "grad_norm": 0.28976351022720337,
      "learning_rate": 0.00019795781145420148,
      "loss": 0.2061801552772522,
      "mean_token_accuracy": 0.9169371575117111,
      "num_tokens": 15740736.0,
      "step": 1276
    },
    {
      "entropy": 1.5299006402492523,
      "epoch": 0.6724591890468667,
      "grad_norm": 0.2772320806980133,
      "learning_rate": 0.0001979531640259963,
      "loss": 0.20862804353237152,
      "mean_token_accuracy": 0.9104678928852081,
      "num_tokens": 15753072.0,
      "step": 1277
    },
    {
      "entropy": 1.5044286847114563,
      "epoch": 0.6729857819905213,
      "grad_norm": 0.3096766769886017,
      "learning_rate": 0.0001979485113765244,
      "loss": 0.23148494958877563,
      "mean_token_accuracy": 0.9055089354515076,
      "num_tokens": 15765408.0,
      "step": 1278
    },
    {
      "entropy": 1.492220789194107,
      "epoch": 0.6735123749341759,
      "grad_norm": 0.2892164885997772,
      "learning_rate": 0.00019794385350606198,
      "loss": 0.2099100798368454,
      "mean_token_accuracy": 0.9140574485063553,
      "num_tokens": 15777744.0,
      "step": 1279
    },
    {
      "entropy": 1.5607218742370605,
      "epoch": 0.6740389678778305,
      "grad_norm": 0.3277181088924408,
      "learning_rate": 0.00019793919041488555,
      "loss": 0.22673308849334717,
      "mean_token_accuracy": 0.9092708230018616,
      "num_tokens": 15790080.0,
      "step": 1280
    },
    {
      "entropy": 1.5162492990493774,
      "epoch": 0.674565560821485,
      "grad_norm": 0.3172418475151062,
      "learning_rate": 0.00019793452210327194,
      "loss": 0.23533746600151062,
      "mean_token_accuracy": 0.9022314250469208,
      "num_tokens": 15802416.0,
      "step": 1281
    },
    {
      "entropy": 1.4783676862716675,
      "epoch": 0.6750921537651395,
      "grad_norm": 0.2982933521270752,
      "learning_rate": 0.00019792984857149826,
      "loss": 0.21463748812675476,
      "mean_token_accuracy": 0.9117985665798187,
      "num_tokens": 15814752.0,
      "step": 1282
    },
    {
      "entropy": 1.5162720084190369,
      "epoch": 0.6756187467087941,
      "grad_norm": 0.2976648211479187,
      "learning_rate": 0.000197925169819842,
      "loss": 0.21266265213489532,
      "mean_token_accuracy": 0.9171696603298187,
      "num_tokens": 15827088.0,
      "step": 1283
    },
    {
      "entropy": 1.5761696994304657,
      "epoch": 0.6761453396524486,
      "grad_norm": 0.32240986824035645,
      "learning_rate": 0.00019792048584858082,
      "loss": 0.22523874044418335,
      "mean_token_accuracy": 0.9125792235136032,
      "num_tokens": 15839424.0,
      "step": 1284
    },
    {
      "entropy": 1.4458520412445068,
      "epoch": 0.6766719325961033,
      "grad_norm": 0.2936963140964508,
      "learning_rate": 0.00019791579665799286,
      "loss": 0.20535065233707428,
      "mean_token_accuracy": 0.9165598750114441,
      "num_tokens": 15851760.0,
      "step": 1285
    },
    {
      "entropy": 1.4983311593532562,
      "epoch": 0.6771985255397578,
      "grad_norm": 0.34449514746665955,
      "learning_rate": 0.0001979111022483565,
      "loss": 0.21712952852249146,
      "mean_token_accuracy": 0.9153823703527451,
      "num_tokens": 15864096.0,
      "step": 1286
    },
    {
      "entropy": 1.562139093875885,
      "epoch": 0.6777251184834123,
      "grad_norm": 0.31743091344833374,
      "learning_rate": 0.00019790640261995034,
      "loss": 0.219345360994339,
      "mean_token_accuracy": 0.9084994047880173,
      "num_tokens": 15876432.0,
      "step": 1287
    },
    {
      "entropy": 1.5375787019729614,
      "epoch": 0.6782517114270669,
      "grad_norm": 0.27115342020988464,
      "learning_rate": 0.00019790169777305345,
      "loss": 0.19260238111019135,
      "mean_token_accuracy": 0.9227229058742523,
      "num_tokens": 15888768.0,
      "step": 1288
    },
    {
      "entropy": 1.4687424898147583,
      "epoch": 0.6787783043707214,
      "grad_norm": 0.302287220954895,
      "learning_rate": 0.00019789698770794506,
      "loss": 0.24450846016407013,
      "mean_token_accuracy": 0.9040905237197876,
      "num_tokens": 15901104.0,
      "step": 1289
    },
    {
      "entropy": 1.5125727355480194,
      "epoch": 0.6793048973143759,
      "grad_norm": 0.3316330313682556,
      "learning_rate": 0.00019789227242490481,
      "loss": 0.23443403840065002,
      "mean_token_accuracy": 0.9077270179986954,
      "num_tokens": 15913440.0,
      "step": 1290
    },
    {
      "entropy": 1.457822471857071,
      "epoch": 0.6798314902580306,
      "grad_norm": 0.3082807958126068,
      "learning_rate": 0.00019788755192421266,
      "loss": 0.2276977300643921,
      "mean_token_accuracy": 0.907697319984436,
      "num_tokens": 15925776.0,
      "step": 1291
    },
    {
      "entropy": 1.5687705874443054,
      "epoch": 0.6803580832016851,
      "grad_norm": 0.3347495198249817,
      "learning_rate": 0.00019788282620614877,
      "loss": 0.23658336699008942,
      "mean_token_accuracy": 0.9071597903966904,
      "num_tokens": 15938112.0,
      "step": 1292
    },
    {
      "entropy": 1.4747570753097534,
      "epoch": 0.6808846761453397,
      "grad_norm": 0.31832975149154663,
      "learning_rate": 0.0001978780952709937,
      "loss": 0.19530466198921204,
      "mean_token_accuracy": 0.9262362122535706,
      "num_tokens": 15950448.0,
      "step": 1293
    },
    {
      "entropy": 1.5592786967754364,
      "epoch": 0.6814112690889942,
      "grad_norm": 0.29860836267471313,
      "learning_rate": 0.00019787335911902835,
      "loss": 0.21259617805480957,
      "mean_token_accuracy": 0.9156482219696045,
      "num_tokens": 15962784.0,
      "step": 1294
    },
    {
      "entropy": 1.590408980846405,
      "epoch": 0.6819378620326487,
      "grad_norm": 0.33426377177238464,
      "learning_rate": 0.0001978686177505338,
      "loss": 0.250851035118103,
      "mean_token_accuracy": 0.9034130126237869,
      "num_tokens": 15975120.0,
      "step": 1295
    },
    {
      "entropy": 1.520596295595169,
      "epoch": 0.6824644549763034,
      "grad_norm": 0.31754234433174133,
      "learning_rate": 0.00019786387116579155,
      "loss": 0.23741202056407928,
      "mean_token_accuracy": 0.9074528068304062,
      "num_tokens": 15987456.0,
      "step": 1296
    },
    {
      "entropy": 1.589228630065918,
      "epoch": 0.6829910479199579,
      "grad_norm": 0.334957480430603,
      "learning_rate": 0.0001978591193650834,
      "loss": 0.2431160807609558,
      "mean_token_accuracy": 0.9039677083492279,
      "num_tokens": 15999792.0,
      "step": 1297
    },
    {
      "entropy": 1.5251929759979248,
      "epoch": 0.6835176408636124,
      "grad_norm": 0.26925790309906006,
      "learning_rate": 0.00019785436234869138,
      "loss": 0.18518362939357758,
      "mean_token_accuracy": 0.9277700483798981,
      "num_tokens": 16012128.0,
      "step": 1298
    },
    {
      "entropy": 1.49588942527771,
      "epoch": 0.684044233807267,
      "grad_norm": 0.2995278835296631,
      "learning_rate": 0.00019784960011689793,
      "loss": 0.21347451210021973,
      "mean_token_accuracy": 0.9186176806688309,
      "num_tokens": 16024464.0,
      "step": 1299
    },
    {
      "entropy": 1.5090012848377228,
      "epoch": 0.6845708267509215,
      "grad_norm": 0.2900152802467346,
      "learning_rate": 0.00019784483266998575,
      "loss": 0.20770123600959778,
      "mean_token_accuracy": 0.9164092987775803,
      "num_tokens": 16036800.0,
      "step": 1300
    },
    {
      "entropy": 1.524385005235672,
      "epoch": 0.685097419694576,
      "grad_norm": 0.2801123857498169,
      "learning_rate": 0.00019784006000823787,
      "loss": 0.1997857242822647,
      "mean_token_accuracy": 0.9173883497714996,
      "num_tokens": 16049136.0,
      "step": 1301
    },
    {
      "entropy": 1.5324156284332275,
      "epoch": 0.6856240126382307,
      "grad_norm": 0.30303382873535156,
      "learning_rate": 0.0001978352821319376,
      "loss": 0.21537630259990692,
      "mean_token_accuracy": 0.9123150855302811,
      "num_tokens": 16061472.0,
      "step": 1302
    },
    {
      "entropy": 1.538516879081726,
      "epoch": 0.6861506055818852,
      "grad_norm": 0.2965938448905945,
      "learning_rate": 0.00019783049904136855,
      "loss": 0.20010440051555634,
      "mean_token_accuracy": 0.9172840118408203,
      "num_tokens": 16073808.0,
      "step": 1303
    },
    {
      "entropy": 1.4958137571811676,
      "epoch": 0.6866771985255398,
      "grad_norm": 0.29906049370765686,
      "learning_rate": 0.0001978257107368147,
      "loss": 0.20351183414459229,
      "mean_token_accuracy": 0.921799048781395,
      "num_tokens": 16086144.0,
      "step": 1304
    },
    {
      "entropy": 1.4531968832015991,
      "epoch": 0.6872037914691943,
      "grad_norm": 0.31886446475982666,
      "learning_rate": 0.0001978209172185603,
      "loss": 0.21320727467536926,
      "mean_token_accuracy": 0.9105631560087204,
      "num_tokens": 16098480.0,
      "step": 1305
    },
    {
      "entropy": 1.3787581324577332,
      "epoch": 0.6877303844128488,
      "grad_norm": 0.2898115813732147,
      "learning_rate": 0.0001978161184868899,
      "loss": 0.2124798595905304,
      "mean_token_accuracy": 0.9114013463258743,
      "num_tokens": 16110816.0,
      "step": 1306
    },
    {
      "entropy": 1.4378518164157867,
      "epoch": 0.6882569773565034,
      "grad_norm": 0.31470921635627747,
      "learning_rate": 0.00019781131454208833,
      "loss": 0.19701991975307465,
      "mean_token_accuracy": 0.9186731576919556,
      "num_tokens": 16123152.0,
      "step": 1307
    },
    {
      "entropy": 1.4289041459560394,
      "epoch": 0.688783570300158,
      "grad_norm": 0.3303619921207428,
      "learning_rate": 0.0001978065053844409,
      "loss": 0.23253998160362244,
      "mean_token_accuracy": 0.9104226678609848,
      "num_tokens": 16135488.0,
      "step": 1308
    },
    {
      "entropy": 1.458989679813385,
      "epoch": 0.6893101632438126,
      "grad_norm": 0.30710119009017944,
      "learning_rate": 0.00019780169101423295,
      "loss": 0.2202616184949875,
      "mean_token_accuracy": 0.9115623980760574,
      "num_tokens": 16147824.0,
      "step": 1309
    },
    {
      "entropy": 1.5141521096229553,
      "epoch": 0.6898367561874671,
      "grad_norm": 0.3152952194213867,
      "learning_rate": 0.0001977968714317504,
      "loss": 0.19381463527679443,
      "mean_token_accuracy": 0.9204765856266022,
      "num_tokens": 16160160.0,
      "step": 1310
    },
    {
      "entropy": 1.4804642498493195,
      "epoch": 0.6903633491311216,
      "grad_norm": 0.3526984751224518,
      "learning_rate": 0.0001977920466372793,
      "loss": 0.22401507198810577,
      "mean_token_accuracy": 0.9142041653394699,
      "num_tokens": 16172496.0,
      "step": 1311
    },
    {
      "entropy": 1.4126333594322205,
      "epoch": 0.6908899420747762,
      "grad_norm": 0.2999390661716461,
      "learning_rate": 0.00019778721663110603,
      "loss": 0.212639719247818,
      "mean_token_accuracy": 0.9140689373016357,
      "num_tokens": 16184832.0,
      "step": 1312
    },
    {
      "entropy": 1.4407996237277985,
      "epoch": 0.6914165350184307,
      "grad_norm": 0.3175979554653168,
      "learning_rate": 0.00019778238141351744,
      "loss": 0.23567378520965576,
      "mean_token_accuracy": 0.9058562517166138,
      "num_tokens": 16197168.0,
      "step": 1313
    },
    {
      "entropy": 1.4304053783416748,
      "epoch": 0.6919431279620853,
      "grad_norm": 0.32357415556907654,
      "learning_rate": 0.00019777754098480047,
      "loss": 0.23323741555213928,
      "mean_token_accuracy": 0.9068403393030167,
      "num_tokens": 16209504.0,
      "step": 1314
    },
    {
      "entropy": 1.4463219940662384,
      "epoch": 0.6924697209057399,
      "grad_norm": 0.32085898518562317,
      "learning_rate": 0.0001977726953452425,
      "loss": 0.25303539633750916,
      "mean_token_accuracy": 0.9010388404130936,
      "num_tokens": 16221840.0,
      "step": 1315
    },
    {
      "entropy": 1.5001619458198547,
      "epoch": 0.6929963138493944,
      "grad_norm": 0.30218154191970825,
      "learning_rate": 0.0001977678444951312,
      "loss": 0.20458348095417023,
      "mean_token_accuracy": 0.9198221117258072,
      "num_tokens": 16234176.0,
      "step": 1316
    },
    {
      "entropy": 1.4169287383556366,
      "epoch": 0.693522906793049,
      "grad_norm": 0.2987150549888611,
      "learning_rate": 0.0001977629884347545,
      "loss": 0.20915895700454712,
      "mean_token_accuracy": 0.914911612868309,
      "num_tokens": 16246512.0,
      "step": 1317
    },
    {
      "entropy": 1.4708910584449768,
      "epoch": 0.6940494997367035,
      "grad_norm": 0.2975226938724518,
      "learning_rate": 0.00019775812716440073,
      "loss": 0.21582108736038208,
      "mean_token_accuracy": 0.916588231921196,
      "num_tokens": 16258848.0,
      "step": 1318
    },
    {
      "entropy": 1.395564466714859,
      "epoch": 0.6945760926803581,
      "grad_norm": 0.26800790429115295,
      "learning_rate": 0.00019775326068435843,
      "loss": 0.19701655209064484,
      "mean_token_accuracy": 0.9204675257205963,
      "num_tokens": 16271184.0,
      "step": 1319
    },
    {
      "entropy": 1.4954661428928375,
      "epoch": 0.6951026856240127,
      "grad_norm": 0.3218332529067993,
      "learning_rate": 0.00019774838899491647,
      "loss": 0.20654654502868652,
      "mean_token_accuracy": 0.9205126464366913,
      "num_tokens": 16283520.0,
      "step": 1320
    },
    {
      "entropy": 1.475353330373764,
      "epoch": 0.6956292785676672,
      "grad_norm": 0.29199621081352234,
      "learning_rate": 0.00019774351209636413,
      "loss": 0.22376279532909393,
      "mean_token_accuracy": 0.9100230783224106,
      "num_tokens": 16295856.0,
      "step": 1321
    },
    {
      "entropy": 1.5148694515228271,
      "epoch": 0.6961558715113217,
      "grad_norm": 0.30520734190940857,
      "learning_rate": 0.00019773862998899086,
      "loss": 0.2176750898361206,
      "mean_token_accuracy": 0.9201380461454391,
      "num_tokens": 16308192.0,
      "step": 1322
    },
    {
      "entropy": 1.5539270341396332,
      "epoch": 0.6966824644549763,
      "grad_norm": 0.32190483808517456,
      "learning_rate": 0.00019773374267308653,
      "loss": 0.2056635320186615,
      "mean_token_accuracy": 0.914083868265152,
      "num_tokens": 16320528.0,
      "step": 1323
    },
    {
      "entropy": 1.4479952156543732,
      "epoch": 0.6972090573986308,
      "grad_norm": 0.3086356222629547,
      "learning_rate": 0.00019772885014894125,
      "loss": 0.19541175663471222,
      "mean_token_accuracy": 0.9208300113677979,
      "num_tokens": 16332864.0,
      "step": 1324
    },
    {
      "entropy": 1.4255986511707306,
      "epoch": 0.6977356503422855,
      "grad_norm": 0.29759716987609863,
      "learning_rate": 0.00019772395241684547,
      "loss": 0.22268646955490112,
      "mean_token_accuracy": 0.9130487442016602,
      "num_tokens": 16345200.0,
      "step": 1325
    },
    {
      "entropy": 1.417414665222168,
      "epoch": 0.69826224328594,
      "grad_norm": 0.30235084891319275,
      "learning_rate": 0.00019771904947708991,
      "loss": 0.19872839748859406,
      "mean_token_accuracy": 0.9202670305967331,
      "num_tokens": 16357536.0,
      "step": 1326
    },
    {
      "entropy": 1.462871015071869,
      "epoch": 0.6987888362295945,
      "grad_norm": 0.330930233001709,
      "learning_rate": 0.00019771414132996565,
      "loss": 0.21214503049850464,
      "mean_token_accuracy": 0.9157495498657227,
      "num_tokens": 16369872.0,
      "step": 1327
    },
    {
      "entropy": 1.4870843291282654,
      "epoch": 0.6993154291732491,
      "grad_norm": 0.33824852108955383,
      "learning_rate": 0.00019770922797576407,
      "loss": 0.23521599173545837,
      "mean_token_accuracy": 0.9104333966970444,
      "num_tokens": 16382208.0,
      "step": 1328
    },
    {
      "entropy": 1.4692321121692657,
      "epoch": 0.6998420221169036,
      "grad_norm": 0.2853015959262848,
      "learning_rate": 0.0001977043094147768,
      "loss": 0.21163907647132874,
      "mean_token_accuracy": 0.9183786511421204,
      "num_tokens": 16394544.0,
      "step": 1329
    },
    {
      "entropy": 1.4160318076610565,
      "epoch": 0.7003686150605581,
      "grad_norm": 0.3194182217121124,
      "learning_rate": 0.00019769938564729585,
      "loss": 0.22332850098609924,
      "mean_token_accuracy": 0.9086870551109314,
      "num_tokens": 16406880.0,
      "step": 1330
    },
    {
      "entropy": 1.5249937772750854,
      "epoch": 0.7008952080042128,
      "grad_norm": 0.2979271709918976,
      "learning_rate": 0.00019769445667361358,
      "loss": 0.20148411393165588,
      "mean_token_accuracy": 0.919349730014801,
      "num_tokens": 16419216.0,
      "step": 1331
    },
    {
      "entropy": 1.475309818983078,
      "epoch": 0.7014218009478673,
      "grad_norm": 0.31436359882354736,
      "learning_rate": 0.00019768952249402252,
      "loss": 0.2226470410823822,
      "mean_token_accuracy": 0.9143447577953339,
      "num_tokens": 16431552.0,
      "step": 1332
    },
    {
      "entropy": 1.45591801404953,
      "epoch": 0.7019483938915219,
      "grad_norm": 0.3134812116622925,
      "learning_rate": 0.00019768458310881557,
      "loss": 0.239861398935318,
      "mean_token_accuracy": 0.9030255079269409,
      "num_tokens": 16443888.0,
      "step": 1333
    },
    {
      "entropy": 1.4746840596199036,
      "epoch": 0.7024749868351764,
      "grad_norm": 0.2945224642753601,
      "learning_rate": 0.00019767963851828604,
      "loss": 0.2150000035762787,
      "mean_token_accuracy": 0.9179409146308899,
      "num_tokens": 16456224.0,
      "step": 1334
    },
    {
      "entropy": 1.4820148646831512,
      "epoch": 0.7030015797788309,
      "grad_norm": 0.30797162652015686,
      "learning_rate": 0.00019767468872272737,
      "loss": 0.22350315749645233,
      "mean_token_accuracy": 0.9135662913322449,
      "num_tokens": 16468560.0,
      "step": 1335
    },
    {
      "entropy": 1.5698617100715637,
      "epoch": 0.7035281727224855,
      "grad_norm": 0.3179170489311218,
      "learning_rate": 0.00019766973372243343,
      "loss": 0.20203739404678345,
      "mean_token_accuracy": 0.9202521592378616,
      "num_tokens": 16480896.0,
      "step": 1336
    },
    {
      "entropy": 1.5165068209171295,
      "epoch": 0.7040547656661401,
      "grad_norm": 0.282023549079895,
      "learning_rate": 0.0001976647735176984,
      "loss": 0.2072068601846695,
      "mean_token_accuracy": 0.9170822501182556,
      "num_tokens": 16493232.0,
      "step": 1337
    },
    {
      "entropy": 1.5599988996982574,
      "epoch": 0.7045813586097947,
      "grad_norm": 0.32171323895454407,
      "learning_rate": 0.0001976598081088167,
      "loss": 0.2177916020154953,
      "mean_token_accuracy": 0.9149684607982635,
      "num_tokens": 16505568.0,
      "step": 1338
    },
    {
      "entropy": 1.566039115190506,
      "epoch": 0.7051079515534492,
      "grad_norm": 0.2976767420768738,
      "learning_rate": 0.00019765483749608313,
      "loss": 0.20423297584056854,
      "mean_token_accuracy": 0.9218804687261581,
      "num_tokens": 16517904.0,
      "step": 1339
    },
    {
      "entropy": 1.5926516354084015,
      "epoch": 0.7056345444971037,
      "grad_norm": 0.3847099244594574,
      "learning_rate": 0.00019764986167979275,
      "loss": 0.22431136667728424,
      "mean_token_accuracy": 0.9091393053531647,
      "num_tokens": 16530240.0,
      "step": 1340
    },
    {
      "entropy": 1.5083985030651093,
      "epoch": 0.7061611374407583,
      "grad_norm": 0.3162590563297272,
      "learning_rate": 0.0001976448806602409,
      "loss": 0.2287883162498474,
      "mean_token_accuracy": 0.9064637273550034,
      "num_tokens": 16542576.0,
      "step": 1341
    },
    {
      "entropy": 1.4989595413208008,
      "epoch": 0.7066877303844129,
      "grad_norm": 0.3000263571739197,
      "learning_rate": 0.00019763989443772337,
      "loss": 0.21327601373195648,
      "mean_token_accuracy": 0.9155562520027161,
      "num_tokens": 16554912.0,
      "step": 1342
    },
    {
      "entropy": 1.5128290951251984,
      "epoch": 0.7072143233280674,
      "grad_norm": 0.2836988568305969,
      "learning_rate": 0.00019763490301253608,
      "loss": 0.19906045496463776,
      "mean_token_accuracy": 0.9238555133342743,
      "num_tokens": 16567248.0,
      "step": 1343
    },
    {
      "entropy": 1.5032199919223785,
      "epoch": 0.707740916271722,
      "grad_norm": 0.2904132306575775,
      "learning_rate": 0.0001976299063849754,
      "loss": 0.21531090140342712,
      "mean_token_accuracy": 0.9095672369003296,
      "num_tokens": 16579584.0,
      "step": 1344
    },
    {
      "entropy": 1.587888479232788,
      "epoch": 0.7082675092153765,
      "grad_norm": 0.3187963664531708,
      "learning_rate": 0.00019762490455533792,
      "loss": 0.21344661712646484,
      "mean_token_accuracy": 0.9137663394212723,
      "num_tokens": 16591920.0,
      "step": 1345
    },
    {
      "entropy": 1.5427050590515137,
      "epoch": 0.708794102159031,
      "grad_norm": 0.2940306067466736,
      "learning_rate": 0.00019761989752392053,
      "loss": 0.1942479908466339,
      "mean_token_accuracy": 0.9190978854894638,
      "num_tokens": 16604256.0,
      "step": 1346
    },
    {
      "entropy": 1.5427431166172028,
      "epoch": 0.7093206951026856,
      "grad_norm": 0.3436744809150696,
      "learning_rate": 0.00019761488529102057,
      "loss": 0.2245483100414276,
      "mean_token_accuracy": 0.909973606467247,
      "num_tokens": 16616592.0,
      "step": 1347
    },
    {
      "entropy": 1.4897418022155762,
      "epoch": 0.7098472880463402,
      "grad_norm": 0.34564313292503357,
      "learning_rate": 0.0001976098678569355,
      "loss": 0.23059901595115662,
      "mean_token_accuracy": 0.9069952964782715,
      "num_tokens": 16628928.0,
      "step": 1348
    },
    {
      "entropy": 1.5078385174274445,
      "epoch": 0.7103738809899948,
      "grad_norm": 0.32199203968048096,
      "learning_rate": 0.0001976048452219632,
      "loss": 0.22921977937221527,
      "mean_token_accuracy": 0.9082883149385452,
      "num_tokens": 16641264.0,
      "step": 1349
    },
    {
      "entropy": 1.5083224773406982,
      "epoch": 0.7109004739336493,
      "grad_norm": 0.3183170258998871,
      "learning_rate": 0.00019759981738640185,
      "loss": 0.22415608167648315,
      "mean_token_accuracy": 0.9100949168205261,
      "num_tokens": 16653600.0,
      "step": 1350
    },
    {
      "entropy": 1.5398178398609161,
      "epoch": 0.7114270668773038,
      "grad_norm": 0.30399900674819946,
      "learning_rate": 0.0001975947843505499,
      "loss": 0.22306974232196808,
      "mean_token_accuracy": 0.9138671904802322,
      "num_tokens": 16665936.0,
      "step": 1351
    },
    {
      "entropy": 1.535412222146988,
      "epoch": 0.7119536598209584,
      "grad_norm": 0.2907712757587433,
      "learning_rate": 0.00019758974611470615,
      "loss": 0.2197205275297165,
      "mean_token_accuracy": 0.9130839556455612,
      "num_tokens": 16678272.0,
      "step": 1352
    },
    {
      "entropy": 1.4855974614620209,
      "epoch": 0.7124802527646129,
      "grad_norm": 0.28864648938179016,
      "learning_rate": 0.0001975847026791697,
      "loss": 0.21561571955680847,
      "mean_token_accuracy": 0.9168848842382431,
      "num_tokens": 16690608.0,
      "step": 1353
    },
    {
      "entropy": 1.453218251466751,
      "epoch": 0.7130068457082676,
      "grad_norm": 0.3135017454624176,
      "learning_rate": 0.00019757965404423994,
      "loss": 0.2011297047138214,
      "mean_token_accuracy": 0.9190320074558258,
      "num_tokens": 16702944.0,
      "step": 1354
    },
    {
      "entropy": 1.4790661036968231,
      "epoch": 0.7135334386519221,
      "grad_norm": 0.3001917004585266,
      "learning_rate": 0.00019757460021021655,
      "loss": 0.23639211058616638,
      "mean_token_accuracy": 0.9105001240968704,
      "num_tokens": 16715280.0,
      "step": 1355
    },
    {
      "entropy": 1.496882289648056,
      "epoch": 0.7140600315955766,
      "grad_norm": 0.31785574555397034,
      "learning_rate": 0.00019756954117739956,
      "loss": 0.2306048721075058,
      "mean_token_accuracy": 0.9150114953517914,
      "num_tokens": 16727616.0,
      "step": 1356
    },
    {
      "entropy": 1.4668599963188171,
      "epoch": 0.7145866245392312,
      "grad_norm": 0.29470139741897583,
      "learning_rate": 0.00019756447694608932,
      "loss": 0.21168065071105957,
      "mean_token_accuracy": 0.9137952625751495,
      "num_tokens": 16739952.0,
      "step": 1357
    },
    {
      "entropy": 1.422877013683319,
      "epoch": 0.7151132174828857,
      "grad_norm": 0.27277153730392456,
      "learning_rate": 0.00019755940751658643,
      "loss": 0.20774061977863312,
      "mean_token_accuracy": 0.9166775196790695,
      "num_tokens": 16752288.0,
      "step": 1358
    },
    {
      "entropy": 1.4244271516799927,
      "epoch": 0.7156398104265402,
      "grad_norm": 0.31778132915496826,
      "learning_rate": 0.00019755433288919187,
      "loss": 0.231976717710495,
      "mean_token_accuracy": 0.9075390994548798,
      "num_tokens": 16764624.0,
      "step": 1359
    },
    {
      "entropy": 1.4513772428035736,
      "epoch": 0.7161664033701949,
      "grad_norm": 0.30516210198402405,
      "learning_rate": 0.0001975492530642069,
      "loss": 0.20754879713058472,
      "mean_token_accuracy": 0.9216670095920563,
      "num_tokens": 16776960.0,
      "step": 1360
    },
    {
      "entropy": 1.4767080545425415,
      "epoch": 0.7166929963138494,
      "grad_norm": 0.3156355917453766,
      "learning_rate": 0.00019754416804193298,
      "loss": 0.20898932218551636,
      "mean_token_accuracy": 0.9173450469970703,
      "num_tokens": 16789296.0,
      "step": 1361
    },
    {
      "entropy": 1.4257088005542755,
      "epoch": 0.717219589257504,
      "grad_norm": 0.33443206548690796,
      "learning_rate": 0.0001975390778226721,
      "loss": 0.237446129322052,
      "mean_token_accuracy": 0.9077172726392746,
      "num_tokens": 16801632.0,
      "step": 1362
    },
    {
      "entropy": 1.393272042274475,
      "epoch": 0.7177461822011585,
      "grad_norm": 0.30247074365615845,
      "learning_rate": 0.00019753398240672637,
      "loss": 0.21317076683044434,
      "mean_token_accuracy": 0.9131140112876892,
      "num_tokens": 16813968.0,
      "step": 1363
    },
    {
      "entropy": 1.431013286113739,
      "epoch": 0.718272775144813,
      "grad_norm": 0.29114142060279846,
      "learning_rate": 0.0001975288817943983,
      "loss": 0.19287830591201782,
      "mean_token_accuracy": 0.9247073382139206,
      "num_tokens": 16826304.0,
      "step": 1364
    },
    {
      "entropy": 1.3813497424125671,
      "epoch": 0.7187993680884676,
      "grad_norm": 0.2997731566429138,
      "learning_rate": 0.00019752377598599067,
      "loss": 0.2021138221025467,
      "mean_token_accuracy": 0.9203190356492996,
      "num_tokens": 16838640.0,
      "step": 1365
    },
    {
      "entropy": 1.4252038598060608,
      "epoch": 0.7193259610321222,
      "grad_norm": 0.3252887725830078,
      "learning_rate": 0.0001975186649818066,
      "loss": 0.20873992145061493,
      "mean_token_accuracy": 0.912327915430069,
      "num_tokens": 16850976.0,
      "step": 1366
    },
    {
      "entropy": 1.4511545896530151,
      "epoch": 0.7198525539757767,
      "grad_norm": 0.34324178099632263,
      "learning_rate": 0.00019751354878214946,
      "loss": 0.21222266554832458,
      "mean_token_accuracy": 0.9157299846410751,
      "num_tokens": 16863312.0,
      "step": 1367
    },
    {
      "entropy": 1.3630057871341705,
      "epoch": 0.7203791469194313,
      "grad_norm": 0.37623706459999084,
      "learning_rate": 0.00019750842738732304,
      "loss": 0.2201295793056488,
      "mean_token_accuracy": 0.9165831059217453,
      "num_tokens": 16875648.0,
      "step": 1368
    },
    {
      "entropy": 1.3505005836486816,
      "epoch": 0.7209057398630858,
      "grad_norm": 0.30142372846603394,
      "learning_rate": 0.0001975033007976313,
      "loss": 0.22212043404579163,
      "mean_token_accuracy": 0.9122855514287949,
      "num_tokens": 16887984.0,
      "step": 1369
    },
    {
      "entropy": 1.4169504642486572,
      "epoch": 0.7214323328067404,
      "grad_norm": 0.3326164782047272,
      "learning_rate": 0.0001974981690133786,
      "loss": 0.22192654013633728,
      "mean_token_accuracy": 0.9108520150184631,
      "num_tokens": 16900320.0,
      "step": 1370
    },
    {
      "entropy": 1.367185264825821,
      "epoch": 0.721958925750395,
      "grad_norm": 0.33256617188453674,
      "learning_rate": 0.00019749303203486958,
      "loss": 0.24168333411216736,
      "mean_token_accuracy": 0.9047919362783432,
      "num_tokens": 16912656.0,
      "step": 1371
    },
    {
      "entropy": 1.3614123463630676,
      "epoch": 0.7224855186940495,
      "grad_norm": 0.3197641670703888,
      "learning_rate": 0.00019748788986240917,
      "loss": 0.23665747046470642,
      "mean_token_accuracy": 0.9026706665754318,
      "num_tokens": 16924992.0,
      "step": 1372
    },
    {
      "entropy": 1.3906526267528534,
      "epoch": 0.7230121116377041,
      "grad_norm": 0.29753339290618896,
      "learning_rate": 0.00019748274249630269,
      "loss": 0.19765980541706085,
      "mean_token_accuracy": 0.9235783219337463,
      "num_tokens": 16937328.0,
      "step": 1373
    },
    {
      "entropy": 1.3462652564048767,
      "epoch": 0.7235387045813586,
      "grad_norm": 0.2804873585700989,
      "learning_rate": 0.00019747758993685568,
      "loss": 0.22064946591854095,
      "mean_token_accuracy": 0.9110289812088013,
      "num_tokens": 16949664.0,
      "step": 1374
    },
    {
      "entropy": 1.4082630276679993,
      "epoch": 0.7240652975250131,
      "grad_norm": 0.3095674216747284,
      "learning_rate": 0.000197472432184374,
      "loss": 0.20987483859062195,
      "mean_token_accuracy": 0.9188564568758011,
      "num_tokens": 16962000.0,
      "step": 1375
    },
    {
      "entropy": 1.304635614156723,
      "epoch": 0.7245918904686677,
      "grad_norm": 0.2778051197528839,
      "learning_rate": 0.00019746726923916387,
      "loss": 0.21651524305343628,
      "mean_token_accuracy": 0.9125857651233673,
      "num_tokens": 16974336.0,
      "step": 1376
    },
    {
      "entropy": 1.3015822768211365,
      "epoch": 0.7251184834123223,
      "grad_norm": 0.2806200683116913,
      "learning_rate": 0.00019746210110153172,
      "loss": 0.22142334282398224,
      "mean_token_accuracy": 0.911768451333046,
      "num_tokens": 16986672.0,
      "step": 1377
    },
    {
      "entropy": 1.3526243567466736,
      "epoch": 0.7256450763559769,
      "grad_norm": 0.3183799684047699,
      "learning_rate": 0.0001974569277717844,
      "loss": 0.2256852686405182,
      "mean_token_accuracy": 0.9073309451341629,
      "num_tokens": 16999008.0,
      "step": 1378
    },
    {
      "entropy": 1.3906278908252716,
      "epoch": 0.7261716692996314,
      "grad_norm": 0.2876151204109192,
      "learning_rate": 0.00019745174925022904,
      "loss": 0.2028118073940277,
      "mean_token_accuracy": 0.923457607626915,
      "num_tokens": 17011344.0,
      "step": 1379
    },
    {
      "entropy": 1.350665271282196,
      "epoch": 0.7266982622432859,
      "grad_norm": 0.33748704195022583,
      "learning_rate": 0.00019744656553717305,
      "loss": 0.2175031304359436,
      "mean_token_accuracy": 0.9140133857727051,
      "num_tokens": 17023680.0,
      "step": 1380
    },
    {
      "entropy": 1.3705949485301971,
      "epoch": 0.7272248551869405,
      "grad_norm": 0.3174329400062561,
      "learning_rate": 0.00019744137663292412,
      "loss": 0.20505183935165405,
      "mean_token_accuracy": 0.916661724448204,
      "num_tokens": 17036016.0,
      "step": 1381
    },
    {
      "entropy": 1.3908927738666534,
      "epoch": 0.727751448130595,
      "grad_norm": 0.31065794825553894,
      "learning_rate": 0.00019743618253779033,
      "loss": 0.23529362678527832,
      "mean_token_accuracy": 0.910541832447052,
      "num_tokens": 17048352.0,
      "step": 1382
    },
    {
      "entropy": 1.3258641958236694,
      "epoch": 0.7282780410742496,
      "grad_norm": 0.26997655630111694,
      "learning_rate": 0.00019743098325207995,
      "loss": 0.18447992205619812,
      "mean_token_accuracy": 0.9224408715963364,
      "num_tokens": 17060688.0,
      "step": 1383
    },
    {
      "entropy": 1.3685748875141144,
      "epoch": 0.7288046340179042,
      "grad_norm": 0.3111379146575928,
      "learning_rate": 0.00019742577877610173,
      "loss": 0.22453925013542175,
      "mean_token_accuracy": 0.9133676290512085,
      "num_tokens": 17073024.0,
      "step": 1384
    },
    {
      "entropy": 1.42083340883255,
      "epoch": 0.7293312269615587,
      "grad_norm": 0.331272691488266,
      "learning_rate": 0.00019742056911016456,
      "loss": 0.23528197407722473,
      "mean_token_accuracy": 0.9075358361005783,
      "num_tokens": 17085360.0,
      "step": 1385
    },
    {
      "entropy": 1.4835380911827087,
      "epoch": 0.7298578199052133,
      "grad_norm": 0.30057939887046814,
      "learning_rate": 0.0001974153542545777,
      "loss": 0.19702467322349548,
      "mean_token_accuracy": 0.9221197068691254,
      "num_tokens": 17097696.0,
      "step": 1386
    },
    {
      "entropy": 1.412318229675293,
      "epoch": 0.7303844128488678,
      "grad_norm": 0.314919650554657,
      "learning_rate": 0.00019741013420965077,
      "loss": 0.22496280074119568,
      "mean_token_accuracy": 0.9107683897018433,
      "num_tokens": 17110032.0,
      "step": 1387
    },
    {
      "entropy": 1.5203742980957031,
      "epoch": 0.7309110057925223,
      "grad_norm": 0.4578418433666229,
      "learning_rate": 0.00019740490897569366,
      "loss": 0.22902803122997284,
      "mean_token_accuracy": 0.9130762368440628,
      "num_tokens": 17122368.0,
      "step": 1388
    },
    {
      "entropy": 1.496030032634735,
      "epoch": 0.731437598736177,
      "grad_norm": 0.29055219888687134,
      "learning_rate": 0.00019739967855301652,
      "loss": 0.21085382997989655,
      "mean_token_accuracy": 0.9122315496206284,
      "num_tokens": 17134704.0,
      "step": 1389
    },
    {
      "entropy": 1.4665674567222595,
      "epoch": 0.7319641916798315,
      "grad_norm": 0.2788514494895935,
      "learning_rate": 0.0001973944429419299,
      "loss": 0.20301197469234467,
      "mean_token_accuracy": 0.9215195178985596,
      "num_tokens": 17147040.0,
      "step": 1390
    },
    {
      "entropy": 1.5158916115760803,
      "epoch": 0.732490784623486,
      "grad_norm": 0.3139953315258026,
      "learning_rate": 0.00019738920214274454,
      "loss": 0.20764051377773285,
      "mean_token_accuracy": 0.915527880191803,
      "num_tokens": 17159376.0,
      "step": 1391
    },
    {
      "entropy": 1.5249950289726257,
      "epoch": 0.7330173775671406,
      "grad_norm": 0.30590322613716125,
      "learning_rate": 0.00019738395615577161,
      "loss": 0.22601057589054108,
      "mean_token_accuracy": 0.9111659973859787,
      "num_tokens": 17171712.0,
      "step": 1392
    },
    {
      "entropy": 1.468850463628769,
      "epoch": 0.7335439705107951,
      "grad_norm": 0.311515212059021,
      "learning_rate": 0.00019737870498132248,
      "loss": 0.2210465669631958,
      "mean_token_accuracy": 0.9097635447978973,
      "num_tokens": 17184048.0,
      "step": 1393
    },
    {
      "entropy": 1.5013999938964844,
      "epoch": 0.7340705634544498,
      "grad_norm": 0.3101951479911804,
      "learning_rate": 0.00019737344861970896,
      "loss": 0.21400384604930878,
      "mean_token_accuracy": 0.9109934121370316,
      "num_tokens": 17196384.0,
      "step": 1394
    },
    {
      "entropy": 1.4603233933448792,
      "epoch": 0.7345971563981043,
      "grad_norm": 0.26839688420295715,
      "learning_rate": 0.00019736818707124303,
      "loss": 0.1959075927734375,
      "mean_token_accuracy": 0.9210042357444763,
      "num_tokens": 17208720.0,
      "step": 1395
    },
    {
      "entropy": 1.513748586177826,
      "epoch": 0.7351237493417588,
      "grad_norm": 0.3097459673881531,
      "learning_rate": 0.00019736292033623704,
      "loss": 0.2226242572069168,
      "mean_token_accuracy": 0.9143490940332413,
      "num_tokens": 17221056.0,
      "step": 1396
    },
    {
      "entropy": 1.4970153272151947,
      "epoch": 0.7356503422854134,
      "grad_norm": 0.3178339898586273,
      "learning_rate": 0.00019735764841500368,
      "loss": 0.22419294714927673,
      "mean_token_accuracy": 0.9115694165229797,
      "num_tokens": 17233392.0,
      "step": 1397
    },
    {
      "entropy": 1.4637455344200134,
      "epoch": 0.7361769352290679,
      "grad_norm": 0.3103068768978119,
      "learning_rate": 0.0001973523713078559,
      "loss": 0.2213975489139557,
      "mean_token_accuracy": 0.908700242638588,
      "num_tokens": 17245728.0,
      "step": 1398
    },
    {
      "entropy": 1.4048950970172882,
      "epoch": 0.7367035281727224,
      "grad_norm": 0.3316826820373535,
      "learning_rate": 0.00019734708901510695,
      "loss": 0.22181765735149384,
      "mean_token_accuracy": 0.9124202877283096,
      "num_tokens": 17258064.0,
      "step": 1399
    },
    {
      "entropy": 1.400584727525711,
      "epoch": 0.7372301211163771,
      "grad_norm": 0.30288171768188477,
      "learning_rate": 0.0001973418015370704,
      "loss": 0.23632709681987762,
      "mean_token_accuracy": 0.903136819601059,
      "num_tokens": 17270400.0,
      "step": 1400
    },
    {
      "entropy": 1.4175618290901184,
      "epoch": 0.7377567140600316,
      "grad_norm": 0.2811354696750641,
      "learning_rate": 0.00019733650887406018,
      "loss": 0.21318787336349487,
      "mean_token_accuracy": 0.9126076847314835,
      "num_tokens": 17282736.0,
      "step": 1401
    },
    {
      "entropy": 1.4151119589805603,
      "epoch": 0.7382833070036862,
      "grad_norm": 0.2701869010925293,
      "learning_rate": 0.00019733121102639048,
      "loss": 0.21690967679023743,
      "mean_token_accuracy": 0.9093466997146606,
      "num_tokens": 17295072.0,
      "step": 1402
    },
    {
      "entropy": 1.3918836414813995,
      "epoch": 0.7388098999473407,
      "grad_norm": 0.27023303508758545,
      "learning_rate": 0.00019732590799437577,
      "loss": 0.18338678777217865,
      "mean_token_accuracy": 0.9283169507980347,
      "num_tokens": 17307408.0,
      "step": 1403
    },
    {
      "entropy": 1.4320692718029022,
      "epoch": 0.7393364928909952,
      "grad_norm": 0.26482948660850525,
      "learning_rate": 0.00019732059977833088,
      "loss": 0.2143803983926773,
      "mean_token_accuracy": 0.9123003780841827,
      "num_tokens": 17319744.0,
      "step": 1404
    },
    {
      "entropy": 1.4552482962608337,
      "epoch": 0.7398630858346498,
      "grad_norm": 0.3084104359149933,
      "learning_rate": 0.00019731528637857094,
      "loss": 0.21455544233322144,
      "mean_token_accuracy": 0.9187665581703186,
      "num_tokens": 17332080.0,
      "step": 1405
    },
    {
      "entropy": 1.362955391407013,
      "epoch": 0.7403896787783044,
      "grad_norm": 0.2885385751724243,
      "learning_rate": 0.00019730996779541133,
      "loss": 0.2053847312927246,
      "mean_token_accuracy": 0.9195720255374908,
      "num_tokens": 17344416.0,
      "step": 1406
    },
    {
      "entropy": 1.3902620077133179,
      "epoch": 0.740916271721959,
      "grad_norm": 0.31122514605522156,
      "learning_rate": 0.0001973046440291678,
      "loss": 0.24145708978176117,
      "mean_token_accuracy": 0.9063299000263214,
      "num_tokens": 17356752.0,
      "step": 1407
    },
    {
      "entropy": 1.3320231139659882,
      "epoch": 0.7414428646656135,
      "grad_norm": 0.285426527261734,
      "learning_rate": 0.00019729931508015647,
      "loss": 0.202953040599823,
      "mean_token_accuracy": 0.9166394174098969,
      "num_tokens": 17369088.0,
      "step": 1408
    },
    {
      "entropy": 1.3167757391929626,
      "epoch": 0.741969457609268,
      "grad_norm": 0.3052743375301361,
      "learning_rate": 0.00019729398094869358,
      "loss": 0.23250755667686462,
      "mean_token_accuracy": 0.909424439072609,
      "num_tokens": 17381424.0,
      "step": 1409
    },
    {
      "entropy": 1.3000437319278717,
      "epoch": 0.7424960505529226,
      "grad_norm": 0.33441463112831116,
      "learning_rate": 0.00019728864163509583,
      "loss": 0.2160688042640686,
      "mean_token_accuracy": 0.9094254523515701,
      "num_tokens": 17393760.0,
      "step": 1410
    },
    {
      "entropy": 1.3271216750144958,
      "epoch": 0.7430226434965771,
      "grad_norm": 0.2913570702075958,
      "learning_rate": 0.00019728329713968016,
      "loss": 0.1972869336605072,
      "mean_token_accuracy": 0.9217556267976761,
      "num_tokens": 17406096.0,
      "step": 1411
    },
    {
      "entropy": 1.3546923398971558,
      "epoch": 0.7435492364402317,
      "grad_norm": 0.31364765763282776,
      "learning_rate": 0.00019727794746276391,
      "loss": 0.21945282816886902,
      "mean_token_accuracy": 0.9129824340343475,
      "num_tokens": 17418432.0,
      "step": 1412
    },
    {
      "entropy": 1.3115905821323395,
      "epoch": 0.7440758293838863,
      "grad_norm": 0.29262346029281616,
      "learning_rate": 0.00019727259260466458,
      "loss": 0.22911304235458374,
      "mean_token_accuracy": 0.9053215533494949,
      "num_tokens": 17430768.0,
      "step": 1413
    },
    {
      "entropy": 1.2379891574382782,
      "epoch": 0.7446024223275408,
      "grad_norm": 0.3259003162384033,
      "learning_rate": 0.0001972672325657001,
      "loss": 0.19887758791446686,
      "mean_token_accuracy": 0.9181624054908752,
      "num_tokens": 17443104.0,
      "step": 1414
    },
    {
      "entropy": 1.3647005259990692,
      "epoch": 0.7451290152711953,
      "grad_norm": 0.4454919695854187,
      "learning_rate": 0.00019726186734618867,
      "loss": 0.22842802107334137,
      "mean_token_accuracy": 0.9110278934240341,
      "num_tokens": 17455440.0,
      "step": 1415
    },
    {
      "entropy": 1.3383563160896301,
      "epoch": 0.7456556082148499,
      "grad_norm": 0.30020466446876526,
      "learning_rate": 0.00019725649694644876,
      "loss": 0.22231580317020416,
      "mean_token_accuracy": 0.9075456112623215,
      "num_tokens": 17467776.0,
      "step": 1416
    },
    {
      "entropy": 1.3714047372341156,
      "epoch": 0.7461822011585044,
      "grad_norm": 0.29304370284080505,
      "learning_rate": 0.0001972511213667992,
      "loss": 0.21095791459083557,
      "mean_token_accuracy": 0.9166263341903687,
      "num_tokens": 17480112.0,
      "step": 1417
    },
    {
      "entropy": 1.3512816727161407,
      "epoch": 0.7467087941021591,
      "grad_norm": 0.3256070613861084,
      "learning_rate": 0.0001972457406075591,
      "loss": 0.20153497159481049,
      "mean_token_accuracy": 0.9194765239953995,
      "num_tokens": 17492448.0,
      "step": 1418
    },
    {
      "entropy": 1.3885486125946045,
      "epoch": 0.7472353870458136,
      "grad_norm": 0.3231737017631531,
      "learning_rate": 0.00019724035466904788,
      "loss": 0.2227013111114502,
      "mean_token_accuracy": 0.9109441936016083,
      "num_tokens": 17504784.0,
      "step": 1419
    },
    {
      "entropy": 1.3755838871002197,
      "epoch": 0.7477619799894681,
      "grad_norm": 0.29787546396255493,
      "learning_rate": 0.0001972349635515853,
      "loss": 0.216201514005661,
      "mean_token_accuracy": 0.9155356884002686,
      "num_tokens": 17517120.0,
      "step": 1420
    },
    {
      "entropy": 1.437903732061386,
      "epoch": 0.7482885729331227,
      "grad_norm": 0.33993855118751526,
      "learning_rate": 0.00019722956725549138,
      "loss": 0.21523010730743408,
      "mean_token_accuracy": 0.9125061631202698,
      "num_tokens": 17529456.0,
      "step": 1421
    },
    {
      "entropy": 1.333466112613678,
      "epoch": 0.7488151658767772,
      "grad_norm": 0.33743515610694885,
      "learning_rate": 0.00019722416578108642,
      "loss": 0.22355730831623077,
      "mean_token_accuracy": 0.9136557281017303,
      "num_tokens": 17541792.0,
      "step": 1422
    },
    {
      "entropy": 1.360717535018921,
      "epoch": 0.7493417588204319,
      "grad_norm": 0.2994014024734497,
      "learning_rate": 0.00019721875912869118,
      "loss": 0.19619853794574738,
      "mean_token_accuracy": 0.9230263978242874,
      "num_tokens": 17554128.0,
      "step": 1423
    },
    {
      "entropy": 1.2954363822937012,
      "epoch": 0.7498683517640864,
      "grad_norm": 0.29989591240882874,
      "learning_rate": 0.0001972133472986265,
      "loss": 0.2181311547756195,
      "mean_token_accuracy": 0.9170015305280685,
      "num_tokens": 17566464.0,
      "step": 1424
    },
    {
      "entropy": 1.380856215953827,
      "epoch": 0.7503949447077409,
      "grad_norm": 0.28764021396636963,
      "learning_rate": 0.00019720793029121377,
      "loss": 0.21031011641025543,
      "mean_token_accuracy": 0.9177572429180145,
      "num_tokens": 17578800.0,
      "step": 1425
    },
    {
      "entropy": 1.3035545349121094,
      "epoch": 0.7509215376513955,
      "grad_norm": 0.2988784611225128,
      "learning_rate": 0.00019720250810677446,
      "loss": 0.20998632907867432,
      "mean_token_accuracy": 0.9130989611148834,
      "num_tokens": 17591136.0,
      "step": 1426
    },
    {
      "entropy": 1.289101094007492,
      "epoch": 0.75144813059505,
      "grad_norm": 0.49617183208465576,
      "learning_rate": 0.0001971970807456305,
      "loss": 0.21026289463043213,
      "mean_token_accuracy": 0.9138355553150177,
      "num_tokens": 17603472.0,
      "step": 1427
    },
    {
      "entropy": 1.3593553006649017,
      "epoch": 0.7519747235387045,
      "grad_norm": 0.3344785273075104,
      "learning_rate": 0.0001971916482081041,
      "loss": 0.20339743793010712,
      "mean_token_accuracy": 0.9177296459674835,
      "num_tokens": 17615808.0,
      "step": 1428
    },
    {
      "entropy": 1.3161483705043793,
      "epoch": 0.7525013164823592,
      "grad_norm": 0.3118573725223541,
      "learning_rate": 0.00019718621049451772,
      "loss": 0.2201179563999176,
      "mean_token_accuracy": 0.9167029410600662,
      "num_tokens": 17628144.0,
      "step": 1429
    },
    {
      "entropy": 1.3809407949447632,
      "epoch": 0.7530279094260137,
      "grad_norm": 0.3208526074886322,
      "learning_rate": 0.00019718076760519423,
      "loss": 0.23530805110931396,
      "mean_token_accuracy": 0.9093427956104279,
      "num_tokens": 17640480.0,
      "step": 1430
    },
    {
      "entropy": 1.3892112076282501,
      "epoch": 0.7535545023696683,
      "grad_norm": 0.2703491747379303,
      "learning_rate": 0.00019717531954045663,
      "loss": 0.2121361941099167,
      "mean_token_accuracy": 0.916591003537178,
      "num_tokens": 17652816.0,
      "step": 1431
    },
    {
      "entropy": 1.464809536933899,
      "epoch": 0.7540810953133228,
      "grad_norm": 0.2980828285217285,
      "learning_rate": 0.00019716986630062842,
      "loss": 0.21547271311283112,
      "mean_token_accuracy": 0.9138219356536865,
      "num_tokens": 17665152.0,
      "step": 1432
    },
    {
      "entropy": 1.4611221253871918,
      "epoch": 0.7546076882569773,
      "grad_norm": 0.3190062642097473,
      "learning_rate": 0.00019716440788603332,
      "loss": 0.22262822091579437,
      "mean_token_accuracy": 0.9171289652585983,
      "num_tokens": 17677488.0,
      "step": 1433
    },
    {
      "entropy": 1.4677778780460358,
      "epoch": 0.7551342812006319,
      "grad_norm": 0.2803551256656647,
      "learning_rate": 0.00019715894429699537,
      "loss": 0.2030690461397171,
      "mean_token_accuracy": 0.9211814850568771,
      "num_tokens": 17689824.0,
      "step": 1434
    },
    {
      "entropy": 1.546930730342865,
      "epoch": 0.7556608741442865,
      "grad_norm": 0.3004758954048157,
      "learning_rate": 0.00019715347553383888,
      "loss": 0.21929655969142914,
      "mean_token_accuracy": 0.9147268533706665,
      "num_tokens": 17702160.0,
      "step": 1435
    },
    {
      "entropy": 1.458657145500183,
      "epoch": 0.756187467087941,
      "grad_norm": 0.2918381094932556,
      "learning_rate": 0.00019714800159688852,
      "loss": 0.20412471890449524,
      "mean_token_accuracy": 0.9208769351243973,
      "num_tokens": 17714496.0,
      "step": 1436
    },
    {
      "entropy": 1.4661605954170227,
      "epoch": 0.7567140600315956,
      "grad_norm": 0.28824299573898315,
      "learning_rate": 0.00019714252248646924,
      "loss": 0.19737893342971802,
      "mean_token_accuracy": 0.9271862953901291,
      "num_tokens": 17726832.0,
      "step": 1437
    },
    {
      "entropy": 1.4540847837924957,
      "epoch": 0.7572406529752501,
      "grad_norm": 0.2962500751018524,
      "learning_rate": 0.00019713703820290634,
      "loss": 0.21638396382331848,
      "mean_token_accuracy": 0.9159590601921082,
      "num_tokens": 17739168.0,
      "step": 1438
    },
    {
      "entropy": 1.4231859147548676,
      "epoch": 0.7577672459189047,
      "grad_norm": 0.29495999217033386,
      "learning_rate": 0.00019713154874652532,
      "loss": 0.2159111201763153,
      "mean_token_accuracy": 0.9145191609859467,
      "num_tokens": 17751504.0,
      "step": 1439
    },
    {
      "entropy": 1.3984394371509552,
      "epoch": 0.7582938388625592,
      "grad_norm": 0.3085009753704071,
      "learning_rate": 0.0001971260541176521,
      "loss": 0.2231639325618744,
      "mean_token_accuracy": 0.9079516977071762,
      "num_tokens": 17763840.0,
      "step": 1440
    },
    {
      "entropy": 1.4092283248901367,
      "epoch": 0.7588204318062138,
      "grad_norm": 0.31046876311302185,
      "learning_rate": 0.00019712055431661286,
      "loss": 0.21453410387039185,
      "mean_token_accuracy": 0.916040763258934,
      "num_tokens": 17776176.0,
      "step": 1441
    },
    {
      "entropy": 1.3918389678001404,
      "epoch": 0.7593470247498684,
      "grad_norm": 0.3046187460422516,
      "learning_rate": 0.00019711504934373408,
      "loss": 0.20514801144599915,
      "mean_token_accuracy": 0.9181930124759674,
      "num_tokens": 17788512.0,
      "step": 1442
    },
    {
      "entropy": 1.3869245946407318,
      "epoch": 0.7598736176935229,
      "grad_norm": 0.31358590722084045,
      "learning_rate": 0.00019710953919934256,
      "loss": 0.22413912415504456,
      "mean_token_accuracy": 0.9102539569139481,
      "num_tokens": 17800848.0,
      "step": 1443
    },
    {
      "entropy": 1.3643006086349487,
      "epoch": 0.7604002106371774,
      "grad_norm": 0.27224990725517273,
      "learning_rate": 0.00019710402388376544,
      "loss": 0.20147213339805603,
      "mean_token_accuracy": 0.9196710139513016,
      "num_tokens": 17813184.0,
      "step": 1444
    },
    {
      "entropy": 1.39336559176445,
      "epoch": 0.760926803580832,
      "grad_norm": 0.3104258179664612,
      "learning_rate": 0.00019709850339733008,
      "loss": 0.21504773199558258,
      "mean_token_accuracy": 0.9135936945676804,
      "num_tokens": 17825520.0,
      "step": 1445
    },
    {
      "entropy": 1.3767728209495544,
      "epoch": 0.7614533965244866,
      "grad_norm": 0.3152048587799072,
      "learning_rate": 0.0001970929777403642,
      "loss": 0.221971333026886,
      "mean_token_accuracy": 0.911506250500679,
      "num_tokens": 17837856.0,
      "step": 1446
    },
    {
      "entropy": 1.4271937906742096,
      "epoch": 0.7619799894681412,
      "grad_norm": 0.3253970444202423,
      "learning_rate": 0.00019708744691319588,
      "loss": 0.2253606915473938,
      "mean_token_accuracy": 0.9094466418027878,
      "num_tokens": 17850192.0,
      "step": 1447
    },
    {
      "entropy": 1.3933376669883728,
      "epoch": 0.7625065824117957,
      "grad_norm": 0.3199850022792816,
      "learning_rate": 0.00019708191091615343,
      "loss": 0.21662366390228271,
      "mean_token_accuracy": 0.9175620973110199,
      "num_tokens": 17862528.0,
      "step": 1448
    },
    {
      "entropy": 1.433942288160324,
      "epoch": 0.7630331753554502,
      "grad_norm": 0.2839816212654114,
      "learning_rate": 0.00019707636974956546,
      "loss": 0.20464417338371277,
      "mean_token_accuracy": 0.9186283946037292,
      "num_tokens": 17874864.0,
      "step": 1449
    },
    {
      "entropy": 1.3883062601089478,
      "epoch": 0.7635597682991048,
      "grad_norm": 0.3059125542640686,
      "learning_rate": 0.00019707082341376093,
      "loss": 0.21790815889835358,
      "mean_token_accuracy": 0.9127152860164642,
      "num_tokens": 17887200.0,
      "step": 1450
    },
    {
      "entropy": 1.4840424954891205,
      "epoch": 0.7640863612427593,
      "grad_norm": 0.317634642124176,
      "learning_rate": 0.0001970652719090691,
      "loss": 0.22078271210193634,
      "mean_token_accuracy": 0.9099593609571457,
      "num_tokens": 17899536.0,
      "step": 1451
    },
    {
      "entropy": 1.373614490032196,
      "epoch": 0.764612954186414,
      "grad_norm": 0.2764304280281067,
      "learning_rate": 0.00019705971523581957,
      "loss": 0.20861467719078064,
      "mean_token_accuracy": 0.9210799187421799,
      "num_tokens": 17911872.0,
      "step": 1452
    },
    {
      "entropy": 1.4546839594841003,
      "epoch": 0.7651395471300685,
      "grad_norm": 0.30617302656173706,
      "learning_rate": 0.00019705415339434212,
      "loss": 0.21506822109222412,
      "mean_token_accuracy": 0.9104336351156235,
      "num_tokens": 17924208.0,
      "step": 1453
    },
    {
      "entropy": 1.46762016415596,
      "epoch": 0.765666140073723,
      "grad_norm": 0.33141979575157166,
      "learning_rate": 0.000197048586384967,
      "loss": 0.23449833691120148,
      "mean_token_accuracy": 0.909214973449707,
      "num_tokens": 17936544.0,
      "step": 1454
    },
    {
      "entropy": 1.3759736120700836,
      "epoch": 0.7661927330173776,
      "grad_norm": 0.28817251324653625,
      "learning_rate": 0.00019704301420802467,
      "loss": 0.2031758576631546,
      "mean_token_accuracy": 0.9166503101587296,
      "num_tokens": 17948880.0,
      "step": 1455
    },
    {
      "entropy": 1.4604673385620117,
      "epoch": 0.7667193259610321,
      "grad_norm": 0.28507548570632935,
      "learning_rate": 0.0001970374368638459,
      "loss": 0.21905720233917236,
      "mean_token_accuracy": 0.9121448248624802,
      "num_tokens": 17961216.0,
      "step": 1456
    },
    {
      "entropy": 1.4368546605110168,
      "epoch": 0.7672459189046866,
      "grad_norm": 0.3244202435016632,
      "learning_rate": 0.00019703185435276179,
      "loss": 0.21986661851406097,
      "mean_token_accuracy": 0.9103575944900513,
      "num_tokens": 17973552.0,
      "step": 1457
    },
    {
      "entropy": 1.4183049201965332,
      "epoch": 0.7677725118483413,
      "grad_norm": 0.3483145833015442,
      "learning_rate": 0.00019702626667510376,
      "loss": 0.2258511334657669,
      "mean_token_accuracy": 0.9121571332216263,
      "num_tokens": 17985888.0,
      "step": 1458
    },
    {
      "entropy": 1.4387763142585754,
      "epoch": 0.7682991047919958,
      "grad_norm": 0.30231183767318726,
      "learning_rate": 0.00019702067383120352,
      "loss": 0.20727157592773438,
      "mean_token_accuracy": 0.9170671254396439,
      "num_tokens": 17998224.0,
      "step": 1459
    },
    {
      "entropy": 1.3991906344890594,
      "epoch": 0.7688256977356503,
      "grad_norm": 0.3081129789352417,
      "learning_rate": 0.00019701507582139304,
      "loss": 0.2180902510881424,
      "mean_token_accuracy": 0.916274681687355,
      "num_tokens": 18010560.0,
      "step": 1460
    },
    {
      "entropy": 1.4099707305431366,
      "epoch": 0.7693522906793049,
      "grad_norm": 0.27477627992630005,
      "learning_rate": 0.00019700947264600466,
      "loss": 0.19322465360164642,
      "mean_token_accuracy": 0.9191512167453766,
      "num_tokens": 18022896.0,
      "step": 1461
    },
    {
      "entropy": 1.4558897018432617,
      "epoch": 0.7698788836229594,
      "grad_norm": 0.2677430212497711,
      "learning_rate": 0.00019700386430537105,
      "loss": 0.21823228895664215,
      "mean_token_accuracy": 0.9123710244894028,
      "num_tokens": 18035232.0,
      "step": 1462
    },
    {
      "entropy": 1.4174329936504364,
      "epoch": 0.770405476566614,
      "grad_norm": 0.2713851034641266,
      "learning_rate": 0.00019699825079982512,
      "loss": 0.19724853336811066,
      "mean_token_accuracy": 0.9200254529714584,
      "num_tokens": 18047568.0,
      "step": 1463
    },
    {
      "entropy": 1.4575957953929901,
      "epoch": 0.7709320695102686,
      "grad_norm": 0.30220046639442444,
      "learning_rate": 0.00019699263212970007,
      "loss": 0.20754371583461761,
      "mean_token_accuracy": 0.9198533892631531,
      "num_tokens": 18059904.0,
      "step": 1464
    },
    {
      "entropy": 1.450849562883377,
      "epoch": 0.7714586624539231,
      "grad_norm": 0.28346338868141174,
      "learning_rate": 0.0001969870082953295,
      "loss": 0.19828161597251892,
      "mean_token_accuracy": 0.9172644019126892,
      "num_tokens": 18072240.0,
      "step": 1465
    },
    {
      "entropy": 1.4658414423465729,
      "epoch": 0.7719852553975777,
      "grad_norm": 0.3004949986934662,
      "learning_rate": 0.00019698137929704724,
      "loss": 0.20767343044281006,
      "mean_token_accuracy": 0.9210254102945328,
      "num_tokens": 18084576.0,
      "step": 1466
    },
    {
      "entropy": 1.4050607681274414,
      "epoch": 0.7725118483412322,
      "grad_norm": 0.2896452248096466,
      "learning_rate": 0.00019697574513518747,
      "loss": 0.20795601606369019,
      "mean_token_accuracy": 0.914929062128067,
      "num_tokens": 18096912.0,
      "step": 1467
    },
    {
      "entropy": 1.4351738393306732,
      "epoch": 0.7730384412848867,
      "grad_norm": 0.34425756335258484,
      "learning_rate": 0.00019697010581008463,
      "loss": 0.21601206064224243,
      "mean_token_accuracy": 0.9149073958396912,
      "num_tokens": 18109248.0,
      "step": 1468
    },
    {
      "entropy": 1.3905524611473083,
      "epoch": 0.7735650342285414,
      "grad_norm": 0.3009505271911621,
      "learning_rate": 0.00019696446132207353,
      "loss": 0.19988638162612915,
      "mean_token_accuracy": 0.9213133901357651,
      "num_tokens": 18121584.0,
      "step": 1469
    },
    {
      "entropy": 1.4411927461624146,
      "epoch": 0.7740916271721959,
      "grad_norm": 0.3239229619503021,
      "learning_rate": 0.0001969588116714892,
      "loss": 0.2117571234703064,
      "mean_token_accuracy": 0.9106622785329819,
      "num_tokens": 18133920.0,
      "step": 1470
    },
    {
      "entropy": 1.4064708650112152,
      "epoch": 0.7746182201158505,
      "grad_norm": 0.3362599015235901,
      "learning_rate": 0.00019695315685866707,
      "loss": 0.23721855878829956,
      "mean_token_accuracy": 0.9070688337087631,
      "num_tokens": 18146256.0,
      "step": 1471
    },
    {
      "entropy": 1.4237781167030334,
      "epoch": 0.775144813059505,
      "grad_norm": 0.3095615804195404,
      "learning_rate": 0.00019694749688394281,
      "loss": 0.2411169707775116,
      "mean_token_accuracy": 0.9083186686038971,
      "num_tokens": 18158592.0,
      "step": 1472
    },
    {
      "entropy": 1.3879849016666412,
      "epoch": 0.7756714060031595,
      "grad_norm": 0.28509050607681274,
      "learning_rate": 0.00019694183174765245,
      "loss": 0.22019074857234955,
      "mean_token_accuracy": 0.911543145775795,
      "num_tokens": 18170928.0,
      "step": 1473
    },
    {
      "entropy": 1.3992516100406647,
      "epoch": 0.7761979989468141,
      "grad_norm": 0.3135905861854553,
      "learning_rate": 0.00019693616145013227,
      "loss": 0.2088148593902588,
      "mean_token_accuracy": 0.9130579680204391,
      "num_tokens": 18183264.0,
      "step": 1474
    },
    {
      "entropy": 1.3637616336345673,
      "epoch": 0.7767245918904687,
      "grad_norm": 0.26418524980545044,
      "learning_rate": 0.00019693048599171887,
      "loss": 0.20479831099510193,
      "mean_token_accuracy": 0.916004553437233,
      "num_tokens": 18195600.0,
      "step": 1475
    },
    {
      "entropy": 1.4390030205249786,
      "epoch": 0.7772511848341233,
      "grad_norm": 0.290937215089798,
      "learning_rate": 0.0001969248053727492,
      "loss": 0.21481530368328094,
      "mean_token_accuracy": 0.9162030220031738,
      "num_tokens": 18207936.0,
      "step": 1476
    },
    {
      "entropy": 1.4204131364822388,
      "epoch": 0.7777777777777778,
      "grad_norm": 0.3285602331161499,
      "learning_rate": 0.00019691911959356044,
      "loss": 0.19588018953800201,
      "mean_token_accuracy": 0.9232761710882187,
      "num_tokens": 18220272.0,
      "step": 1477
    },
    {
      "entropy": 1.4693542420864105,
      "epoch": 0.7783043707214323,
      "grad_norm": 0.3064950704574585,
      "learning_rate": 0.0001969134286544902,
      "loss": 0.20760740339756012,
      "mean_token_accuracy": 0.9132848978042603,
      "num_tokens": 18232608.0,
      "step": 1478
    },
    {
      "entropy": 1.3828904628753662,
      "epoch": 0.7788309636650869,
      "grad_norm": 0.3061995506286621,
      "learning_rate": 0.00019690773255587625,
      "loss": 0.2265746295452118,
      "mean_token_accuracy": 0.9108180850744247,
      "num_tokens": 18244944.0,
      "step": 1479
    },
    {
      "entropy": 1.3849403858184814,
      "epoch": 0.7793575566087414,
      "grad_norm": 0.3251616656780243,
      "learning_rate": 0.00019690203129805672,
      "loss": 0.2233261615037918,
      "mean_token_accuracy": 0.9125779420137405,
      "num_tokens": 18257280.0,
      "step": 1480
    },
    {
      "entropy": 1.3147023022174835,
      "epoch": 0.779884149552396,
      "grad_norm": 0.31116747856140137,
      "learning_rate": 0.00019689632488137016,
      "loss": 0.24036768078804016,
      "mean_token_accuracy": 0.906961664557457,
      "num_tokens": 18269616.0,
      "step": 1481
    },
    {
      "entropy": 1.3186087906360626,
      "epoch": 0.7804107424960506,
      "grad_norm": 0.30934205651283264,
      "learning_rate": 0.0001968906133061552,
      "loss": 0.21348203718662262,
      "mean_token_accuracy": 0.9163684844970703,
      "num_tokens": 18281952.0,
      "step": 1482
    },
    {
      "entropy": 1.3171938359737396,
      "epoch": 0.7809373354397051,
      "grad_norm": 0.3076079487800598,
      "learning_rate": 0.00019688489657275103,
      "loss": 0.22806760668754578,
      "mean_token_accuracy": 0.9050341248512268,
      "num_tokens": 18294288.0,
      "step": 1483
    },
    {
      "entropy": 1.331948697566986,
      "epoch": 0.7814639283833597,
      "grad_norm": 0.3263510465621948,
      "learning_rate": 0.0001968791746814969,
      "loss": 0.2278282344341278,
      "mean_token_accuracy": 0.9075683802366257,
      "num_tokens": 18306624.0,
      "step": 1484
    },
    {
      "entropy": 1.3975997269153595,
      "epoch": 0.7819905213270142,
      "grad_norm": 0.37345680594444275,
      "learning_rate": 0.00019687344763273255,
      "loss": 0.22596792876720428,
      "mean_token_accuracy": 0.913162499666214,
      "num_tokens": 18318960.0,
      "step": 1485
    },
    {
      "entropy": 1.3315426707267761,
      "epoch": 0.7825171142706687,
      "grad_norm": 0.28193777799606323,
      "learning_rate": 0.00019686771542679797,
      "loss": 0.1983056217432022,
      "mean_token_accuracy": 0.9222036451101303,
      "num_tokens": 18331296.0,
      "step": 1486
    },
    {
      "entropy": 1.4003748893737793,
      "epoch": 0.7830437072143234,
      "grad_norm": 0.30775102972984314,
      "learning_rate": 0.00019686197806403343,
      "loss": 0.22822785377502441,
      "mean_token_accuracy": 0.9109926074743271,
      "num_tokens": 18343632.0,
      "step": 1487
    },
    {
      "entropy": 1.3931090831756592,
      "epoch": 0.7835703001579779,
      "grad_norm": 0.2824244201183319,
      "learning_rate": 0.0001968562355447795,
      "loss": 0.20619836449623108,
      "mean_token_accuracy": 0.9165951460599899,
      "num_tokens": 18355968.0,
      "step": 1488
    },
    {
      "entropy": 1.3029468953609467,
      "epoch": 0.7840968931016324,
      "grad_norm": 0.26663464307785034,
      "learning_rate": 0.00019685048786937713,
      "loss": 0.2070493847131729,
      "mean_token_accuracy": 0.91603122651577,
      "num_tokens": 18368304.0,
      "step": 1489
    },
    {
      "entropy": 1.4118830561637878,
      "epoch": 0.784623486045287,
      "grad_norm": 0.2948478162288666,
      "learning_rate": 0.00019684473503816746,
      "loss": 0.2121935933828354,
      "mean_token_accuracy": 0.9169209152460098,
      "num_tokens": 18380640.0,
      "step": 1490
    },
    {
      "entropy": 1.5040743947029114,
      "epoch": 0.7851500789889415,
      "grad_norm": 0.33927392959594727,
      "learning_rate": 0.00019683897705149208,
      "loss": 0.23608259856700897,
      "mean_token_accuracy": 0.9112606197595596,
      "num_tokens": 18392976.0,
      "step": 1491
    },
    {
      "entropy": 1.388082355260849,
      "epoch": 0.785676671932596,
      "grad_norm": 0.2748982012271881,
      "learning_rate": 0.0001968332139096927,
      "loss": 0.21956929564476013,
      "mean_token_accuracy": 0.9111993908882141,
      "num_tokens": 18405312.0,
      "step": 1492
    },
    {
      "entropy": 1.419054538011551,
      "epoch": 0.7862032648762507,
      "grad_norm": 0.2975718379020691,
      "learning_rate": 0.00019682744561311163,
      "loss": 0.20622003078460693,
      "mean_token_accuracy": 0.91482774913311,
      "num_tokens": 18417648.0,
      "step": 1493
    },
    {
      "entropy": 1.3681580722332,
      "epoch": 0.7867298578199052,
      "grad_norm": 0.30167657136917114,
      "learning_rate": 0.00019682167216209106,
      "loss": 0.22527143359184265,
      "mean_token_accuracy": 0.9118600189685822,
      "num_tokens": 18429984.0,
      "step": 1494
    },
    {
      "entropy": 1.3992347717285156,
      "epoch": 0.7872564507635598,
      "grad_norm": 0.3023481070995331,
      "learning_rate": 0.00019681589355697394,
      "loss": 0.20631316304206848,
      "mean_token_accuracy": 0.9144344329833984,
      "num_tokens": 18442320.0,
      "step": 1495
    },
    {
      "entropy": 1.2978556156158447,
      "epoch": 0.7877830437072143,
      "grad_norm": 0.275492399930954,
      "learning_rate": 0.00019681010979810318,
      "loss": 0.2021515965461731,
      "mean_token_accuracy": 0.9172556847333908,
      "num_tokens": 18454656.0,
      "step": 1496
    },
    {
      "entropy": 1.2966987490653992,
      "epoch": 0.7883096366508688,
      "grad_norm": 0.3461536169052124,
      "learning_rate": 0.00019680432088582217,
      "loss": 0.22693704068660736,
      "mean_token_accuracy": 0.9091721177101135,
      "num_tokens": 18466992.0,
      "step": 1497
    },
    {
      "entropy": 1.304772973060608,
      "epoch": 0.7888362295945235,
      "grad_norm": 0.32779547572135925,
      "learning_rate": 0.00019679852682047457,
      "loss": 0.21427524089813232,
      "mean_token_accuracy": 0.9151846766471863,
      "num_tokens": 18479328.0,
      "step": 1498
    },
    {
      "entropy": 1.2882267534732819,
      "epoch": 0.789362822538178,
      "grad_norm": 0.3116947114467621,
      "learning_rate": 0.00019679272760240435,
      "loss": 0.20784853398799896,
      "mean_token_accuracy": 0.9149093627929688,
      "num_tokens": 18491664.0,
      "step": 1499
    },
    {
      "entropy": 1.263739138841629,
      "epoch": 0.7898894154818326,
      "grad_norm": 0.443194717168808,
      "learning_rate": 0.0001967869232319557,
      "loss": 0.23977030813694,
      "mean_token_accuracy": 0.9084160029888153,
      "num_tokens": 18504000.0,
      "step": 1500
    },
    {
      "entropy": 1.2093487679958344,
      "epoch": 0.7904160084254871,
      "grad_norm": 0.29420503973960876,
      "learning_rate": 0.0001967811137094733,
      "loss": 0.22335118055343628,
      "mean_token_accuracy": 0.909618929028511,
      "num_tokens": 18516336.0,
      "step": 1501
    },
    {
      "entropy": 1.271395206451416,
      "epoch": 0.7909426013691416,
      "grad_norm": 0.5013836026191711,
      "learning_rate": 0.00019677529903530197,
      "loss": 0.2110762596130371,
      "mean_token_accuracy": 0.9183898121118546,
      "num_tokens": 18528672.0,
      "step": 1502
    },
    {
      "entropy": 1.2578312754631042,
      "epoch": 0.7914691943127962,
      "grad_norm": 0.37272414565086365,
      "learning_rate": 0.0001967694792097869,
      "loss": 0.22164121270179749,
      "mean_token_accuracy": 0.9116562008857727,
      "num_tokens": 18541008.0,
      "step": 1503
    },
    {
      "entropy": 1.2144514918327332,
      "epoch": 0.7919957872564508,
      "grad_norm": 0.3145962059497833,
      "learning_rate": 0.0001967636542332736,
      "loss": 0.22305795550346375,
      "mean_token_accuracy": 0.9156023114919662,
      "num_tokens": 18553344.0,
      "step": 1504
    },
    {
      "entropy": 1.2401383221149445,
      "epoch": 0.7925223802001053,
      "grad_norm": 0.2835533022880554,
      "learning_rate": 0.00019675782410610778,
      "loss": 0.19279327988624573,
      "mean_token_accuracy": 0.926255002617836,
      "num_tokens": 18565680.0,
      "step": 1505
    },
    {
      "entropy": 1.2049389481544495,
      "epoch": 0.7930489731437599,
      "grad_norm": 0.2632310092449188,
      "learning_rate": 0.00019675198882863567,
      "loss": 0.2010316103696823,
      "mean_token_accuracy": 0.91452956199646,
      "num_tokens": 18578016.0,
      "step": 1506
    },
    {
      "entropy": 1.249726414680481,
      "epoch": 0.7935755660874144,
      "grad_norm": 0.30521342158317566,
      "learning_rate": 0.0001967461484012036,
      "loss": 0.23277787864208221,
      "mean_token_accuracy": 0.9076357632875443,
      "num_tokens": 18590352.0,
      "step": 1507
    },
    {
      "entropy": 1.1978009939193726,
      "epoch": 0.794102159031069,
      "grad_norm": 0.2853034734725952,
      "learning_rate": 0.00019674030282415826,
      "loss": 0.2127259522676468,
      "mean_token_accuracy": 0.9155232310295105,
      "num_tokens": 18602688.0,
      "step": 1508
    },
    {
      "entropy": 1.25815749168396,
      "epoch": 0.7946287519747235,
      "grad_norm": 0.28645238280296326,
      "learning_rate": 0.00019673445209784677,
      "loss": 0.22002597153186798,
      "mean_token_accuracy": 0.9130334705114365,
      "num_tokens": 18615024.0,
      "step": 1509
    },
    {
      "entropy": 1.3180667161941528,
      "epoch": 0.7951553449183781,
      "grad_norm": 0.2906380891799927,
      "learning_rate": 0.00019672859622261633,
      "loss": 0.22877947986125946,
      "mean_token_accuracy": 0.912013441324234,
      "num_tokens": 18627360.0,
      "step": 1510
    },
    {
      "entropy": 1.2853264808654785,
      "epoch": 0.7956819378620327,
      "grad_norm": 0.28856080770492554,
      "learning_rate": 0.00019672273519881465,
      "loss": 0.2222394347190857,
      "mean_token_accuracy": 0.9078328162431717,
      "num_tokens": 18639696.0,
      "step": 1511
    },
    {
      "entropy": 1.2696054577827454,
      "epoch": 0.7962085308056872,
      "grad_norm": 0.30683693289756775,
      "learning_rate": 0.0001967168690267896,
      "loss": 0.22363853454589844,
      "mean_token_accuracy": 0.9149999022483826,
      "num_tokens": 18652032.0,
      "step": 1512
    },
    {
      "entropy": 1.272040843963623,
      "epoch": 0.7967351237493417,
      "grad_norm": 0.3236723840236664,
      "learning_rate": 0.0001967109977068895,
      "loss": 0.21476420760154724,
      "mean_token_accuracy": 0.9179616123437881,
      "num_tokens": 18664368.0,
      "step": 1513
    },
    {
      "entropy": 1.260970801115036,
      "epoch": 0.7972617166929963,
      "grad_norm": 0.30413469672203064,
      "learning_rate": 0.00019670512123946284,
      "loss": 0.22492626309394836,
      "mean_token_accuracy": 0.9107203483581543,
      "num_tokens": 18676704.0,
      "step": 1514
    },
    {
      "entropy": 1.2300290167331696,
      "epoch": 0.7977883096366508,
      "grad_norm": 0.2895817458629608,
      "learning_rate": 0.00019669923962485847,
      "loss": 0.2138168066740036,
      "mean_token_accuracy": 0.9164409637451172,
      "num_tokens": 18689040.0,
      "step": 1515
    },
    {
      "entropy": 1.2246987521648407,
      "epoch": 0.7983149025803055,
      "grad_norm": 0.30868765711784363,
      "learning_rate": 0.0001966933528634256,
      "loss": 0.2062627673149109,
      "mean_token_accuracy": 0.9158836156129837,
      "num_tokens": 18701376.0,
      "step": 1516
    },
    {
      "entropy": 1.24353489279747,
      "epoch": 0.79884149552396,
      "grad_norm": 0.2896019518375397,
      "learning_rate": 0.00019668746095551364,
      "loss": 0.2127014845609665,
      "mean_token_accuracy": 0.9154282063245773,
      "num_tokens": 18713712.0,
      "step": 1517
    },
    {
      "entropy": 1.219551146030426,
      "epoch": 0.7993680884676145,
      "grad_norm": 0.5301105380058289,
      "learning_rate": 0.0001966815639014724,
      "loss": 0.20765289664268494,
      "mean_token_accuracy": 0.9193731993436813,
      "num_tokens": 18726048.0,
      "step": 1518
    },
    {
      "entropy": 1.2230626046657562,
      "epoch": 0.7998946814112691,
      "grad_norm": 0.2891480624675751,
      "learning_rate": 0.00019667566170165194,
      "loss": 0.20567816495895386,
      "mean_token_accuracy": 0.9149514883756638,
      "num_tokens": 18738384.0,
      "step": 1519
    },
    {
      "entropy": 1.1955215632915497,
      "epoch": 0.8004212743549236,
      "grad_norm": 0.3952074646949768,
      "learning_rate": 0.00019666975435640258,
      "loss": 0.23236501216888428,
      "mean_token_accuracy": 0.910706028342247,
      "num_tokens": 18750720.0,
      "step": 1520
    },
    {
      "entropy": 1.1208319664001465,
      "epoch": 0.8009478672985783,
      "grad_norm": 0.3569393455982208,
      "learning_rate": 0.00019666384186607506,
      "loss": 0.2112501859664917,
      "mean_token_accuracy": 0.9126569777727127,
      "num_tokens": 18763056.0,
      "step": 1521
    },
    {
      "entropy": 1.1377290189266205,
      "epoch": 0.8014744602422328,
      "grad_norm": 0.2929931879043579,
      "learning_rate": 0.00019665792423102037,
      "loss": 0.22051754593849182,
      "mean_token_accuracy": 0.9091715216636658,
      "num_tokens": 18775392.0,
      "step": 1522
    },
    {
      "entropy": 1.1271924078464508,
      "epoch": 0.8020010531858873,
      "grad_norm": 0.3465351164340973,
      "learning_rate": 0.0001966520014515898,
      "loss": 0.22917094826698303,
      "mean_token_accuracy": 0.9075794517993927,
      "num_tokens": 18787728.0,
      "step": 1523
    },
    {
      "entropy": 1.1986294984817505,
      "epoch": 0.8025276461295419,
      "grad_norm": 0.32447630167007446,
      "learning_rate": 0.00019664607352813494,
      "loss": 0.23350387811660767,
      "mean_token_accuracy": 0.9076796770095825,
      "num_tokens": 18800064.0,
      "step": 1524
    },
    {
      "entropy": 1.1514316499233246,
      "epoch": 0.8030542390731964,
      "grad_norm": 0.3304944932460785,
      "learning_rate": 0.00019664014046100775,
      "loss": 0.21476618945598602,
      "mean_token_accuracy": 0.913004919886589,
      "num_tokens": 18812400.0,
      "step": 1525
    },
    {
      "entropy": 1.1340425908565521,
      "epoch": 0.8035808320168509,
      "grad_norm": 0.5441511869430542,
      "learning_rate": 0.00019663420225056035,
      "loss": 0.22214046120643616,
      "mean_token_accuracy": 0.9116765111684799,
      "num_tokens": 18824736.0,
      "step": 1526
    },
    {
      "entropy": 1.0562350451946259,
      "epoch": 0.8041074249605056,
      "grad_norm": 0.3077342212200165,
      "learning_rate": 0.00019662825889714533,
      "loss": 0.21319684386253357,
      "mean_token_accuracy": 0.918745294213295,
      "num_tokens": 18837072.0,
      "step": 1527
    },
    {
      "entropy": 1.0274389684200287,
      "epoch": 0.8046340179041601,
      "grad_norm": 0.3354491889476776,
      "learning_rate": 0.0001966223104011155,
      "loss": 0.2296486645936966,
      "mean_token_accuracy": 0.9057023078203201,
      "num_tokens": 18849408.0,
      "step": 1528
    },
    {
      "entropy": 1.0815512537956238,
      "epoch": 0.8051606108478147,
      "grad_norm": 0.3144749402999878,
      "learning_rate": 0.00019661635676282394,
      "loss": 0.229911670088768,
      "mean_token_accuracy": 0.9100077450275421,
      "num_tokens": 18861744.0,
      "step": 1529
    },
    {
      "entropy": 1.0930194556713104,
      "epoch": 0.8056872037914692,
      "grad_norm": 0.300818532705307,
      "learning_rate": 0.00019661039798262415,
      "loss": 0.20994170010089874,
      "mean_token_accuracy": 0.9187563508749008,
      "num_tokens": 18874080.0,
      "step": 1530
    },
    {
      "entropy": 1.0869486331939697,
      "epoch": 0.8062137967351237,
      "grad_norm": 0.28593873977661133,
      "learning_rate": 0.00019660443406086984,
      "loss": 0.20231741666793823,
      "mean_token_accuracy": 0.918257012963295,
      "num_tokens": 18886416.0,
      "step": 1531
    },
    {
      "entropy": 1.0835871398448944,
      "epoch": 0.8067403896787783,
      "grad_norm": 0.7419033050537109,
      "learning_rate": 0.000196598464997915,
      "loss": 0.2157197892665863,
      "mean_token_accuracy": 0.9135694205760956,
      "num_tokens": 18898752.0,
      "step": 1532
    },
    {
      "entropy": 1.1274356842041016,
      "epoch": 0.8072669826224329,
      "grad_norm": 0.31080278754234314,
      "learning_rate": 0.00019659249079411405,
      "loss": 0.22466787695884705,
      "mean_token_accuracy": 0.9107064604759216,
      "num_tokens": 18911088.0,
      "step": 1533
    },
    {
      "entropy": 1.0774615406990051,
      "epoch": 0.8077935755660874,
      "grad_norm": 0.28806376457214355,
      "learning_rate": 0.00019658651144982163,
      "loss": 0.2336619794368744,
      "mean_token_accuracy": 0.9125310033559799,
      "num_tokens": 18923424.0,
      "step": 1534
    },
    {
      "entropy": 1.0493130534887314,
      "epoch": 0.808320168509742,
      "grad_norm": 0.2861652672290802,
      "learning_rate": 0.0001965805269653927,
      "loss": 0.20115122199058533,
      "mean_token_accuracy": 0.919718474149704,
      "num_tokens": 18935760.0,
      "step": 1535
    },
    {
      "entropy": 1.0706775784492493,
      "epoch": 0.8088467614533965,
      "grad_norm": 0.29678940773010254,
      "learning_rate": 0.0001965745373411825,
      "loss": 0.21643559634685516,
      "mean_token_accuracy": 0.9111261665821075,
      "num_tokens": 18948096.0,
      "step": 1536
    },
    {
      "entropy": 1.0705525279045105,
      "epoch": 0.809373354397051,
      "grad_norm": 0.3051356077194214,
      "learning_rate": 0.0001965685425775466,
      "loss": 0.2151341289281845,
      "mean_token_accuracy": 0.9130971729755402,
      "num_tokens": 18960432.0,
      "step": 1537
    },
    {
      "entropy": 1.0797742903232574,
      "epoch": 0.8098999473407056,
      "grad_norm": 0.29696813225746155,
      "learning_rate": 0.0001965625426748409,
      "loss": 0.22983068227767944,
      "mean_token_accuracy": 0.9089944660663605,
      "num_tokens": 18972768.0,
      "step": 1538
    },
    {
      "entropy": 1.0420788824558258,
      "epoch": 0.8104265402843602,
      "grad_norm": 0.29141098260879517,
      "learning_rate": 0.00019655653763342155,
      "loss": 0.20703990757465363,
      "mean_token_accuracy": 0.9171910136938095,
      "num_tokens": 18985104.0,
      "step": 1539
    },
    {
      "entropy": 1.0945038497447968,
      "epoch": 0.8109531332280148,
      "grad_norm": 0.30959945917129517,
      "learning_rate": 0.00019655052745364509,
      "loss": 0.24617335200309753,
      "mean_token_accuracy": 0.9053667187690735,
      "num_tokens": 18997440.0,
      "step": 1540
    },
    {
      "entropy": 1.0674882233142853,
      "epoch": 0.8114797261716693,
      "grad_norm": 0.29151442646980286,
      "learning_rate": 0.00019654451213586824,
      "loss": 0.202132448554039,
      "mean_token_accuracy": 0.9202746897935867,
      "num_tokens": 19009776.0,
      "step": 1541
    },
    {
      "entropy": 1.1482883095741272,
      "epoch": 0.8120063191153238,
      "grad_norm": 0.3008717894554138,
      "learning_rate": 0.00019653849168044815,
      "loss": 0.2195252776145935,
      "mean_token_accuracy": 0.9127016216516495,
      "num_tokens": 19022112.0,
      "step": 1542
    },
    {
      "entropy": 1.0573867559432983,
      "epoch": 0.8125329120589784,
      "grad_norm": 0.2779742479324341,
      "learning_rate": 0.00019653246608774215,
      "loss": 0.20425541698932648,
      "mean_token_accuracy": 0.9176970571279526,
      "num_tokens": 19034448.0,
      "step": 1543
    },
    {
      "entropy": 1.1284627318382263,
      "epoch": 0.8130595050026329,
      "grad_norm": 0.3272855281829834,
      "learning_rate": 0.00019652643535810803,
      "loss": 0.2418060004711151,
      "mean_token_accuracy": 0.9060062915086746,
      "num_tokens": 19046784.0,
      "step": 1544
    },
    {
      "entropy": 1.160320371389389,
      "epoch": 0.8135860979462876,
      "grad_norm": 0.28571900725364685,
      "learning_rate": 0.00019652039949190372,
      "loss": 0.19945061206817627,
      "mean_token_accuracy": 0.9213592410087585,
      "num_tokens": 19059120.0,
      "step": 1545
    },
    {
      "entropy": 1.0708126425743103,
      "epoch": 0.8141126908899421,
      "grad_norm": 0.28214171528816223,
      "learning_rate": 0.00019651435848948762,
      "loss": 0.21407712996006012,
      "mean_token_accuracy": 0.9169910848140717,
      "num_tokens": 19071456.0,
      "step": 1546
    },
    {
      "entropy": 1.071442037820816,
      "epoch": 0.8146392838335966,
      "grad_norm": 0.27851834893226624,
      "learning_rate": 0.00019650831235121824,
      "loss": 0.20925593376159668,
      "mean_token_accuracy": 0.9159252196550369,
      "num_tokens": 19083792.0,
      "step": 1547
    },
    {
      "entropy": 1.1155974566936493,
      "epoch": 0.8151658767772512,
      "grad_norm": 0.3253529369831085,
      "learning_rate": 0.00019650226107745461,
      "loss": 0.21849502623081207,
      "mean_token_accuracy": 0.9110303372144699,
      "num_tokens": 19096128.0,
      "step": 1548
    },
    {
      "entropy": 1.0896004736423492,
      "epoch": 0.8156924697209057,
      "grad_norm": 0.28845205903053284,
      "learning_rate": 0.0001964962046685559,
      "loss": 0.22848929464817047,
      "mean_token_accuracy": 0.9086662530899048,
      "num_tokens": 19108464.0,
      "step": 1549
    },
    {
      "entropy": 1.0635098963975906,
      "epoch": 0.8162190626645603,
      "grad_norm": 0.3507218658924103,
      "learning_rate": 0.00019649014312488164,
      "loss": 0.22886288166046143,
      "mean_token_accuracy": 0.9107942432165146,
      "num_tokens": 19120800.0,
      "step": 1550
    },
    {
      "entropy": 1.0771049559116364,
      "epoch": 0.8167456556082149,
      "grad_norm": 0.3061995506286621,
      "learning_rate": 0.0001964840764467917,
      "loss": 0.22587770223617554,
      "mean_token_accuracy": 0.907668799161911,
      "num_tokens": 19133136.0,
      "step": 1551
    },
    {
      "entropy": 1.0604461431503296,
      "epoch": 0.8172722485518694,
      "grad_norm": 0.29044315218925476,
      "learning_rate": 0.00019647800463464622,
      "loss": 0.2082800567150116,
      "mean_token_accuracy": 0.9161125719547272,
      "num_tokens": 19145472.0,
      "step": 1552
    },
    {
      "entropy": 1.0522086769342422,
      "epoch": 0.817798841495524,
      "grad_norm": 0.301567405462265,
      "learning_rate": 0.0001964719276888056,
      "loss": 0.20523683726787567,
      "mean_token_accuracy": 0.9130432307720184,
      "num_tokens": 19157808.0,
      "step": 1553
    },
    {
      "entropy": 1.02540722489357,
      "epoch": 0.8183254344391785,
      "grad_norm": 0.29618892073631287,
      "learning_rate": 0.00019646584560963065,
      "loss": 0.22347863018512726,
      "mean_token_accuracy": 0.9131524860858917,
      "num_tokens": 19170144.0,
      "step": 1554
    },
    {
      "entropy": 1.0311861038208008,
      "epoch": 0.818852027382833,
      "grad_norm": 0.27861911058425903,
      "learning_rate": 0.00019645975839748244,
      "loss": 0.182900071144104,
      "mean_token_accuracy": 0.9252242147922516,
      "num_tokens": 19182480.0,
      "step": 1555
    },
    {
      "entropy": 1.0627757012844086,
      "epoch": 0.8193786203264877,
      "grad_norm": 0.3110608458518982,
      "learning_rate": 0.00019645366605272228,
      "loss": 0.23385684192180634,
      "mean_token_accuracy": 0.9053423404693604,
      "num_tokens": 19194816.0,
      "step": 1556
    },
    {
      "entropy": 1.0807601511478424,
      "epoch": 0.8199052132701422,
      "grad_norm": 0.31477105617523193,
      "learning_rate": 0.00019644756857571186,
      "loss": 0.22126808762550354,
      "mean_token_accuracy": 0.9116723388433456,
      "num_tokens": 19207152.0,
      "step": 1557
    },
    {
      "entropy": 1.079503744840622,
      "epoch": 0.8204318062137967,
      "grad_norm": 0.314552366733551,
      "learning_rate": 0.00019644146596681312,
      "loss": 0.22594426572322845,
      "mean_token_accuracy": 0.9112756550312042,
      "num_tokens": 19219488.0,
      "step": 1558
    },
    {
      "entropy": 1.0542688369750977,
      "epoch": 0.8209583991574513,
      "grad_norm": 0.30733972787857056,
      "learning_rate": 0.0001964353582263884,
      "loss": 0.22227591276168823,
      "mean_token_accuracy": 0.9121880233287811,
      "num_tokens": 19231824.0,
      "step": 1559
    },
    {
      "entropy": 1.0992242991924286,
      "epoch": 0.8214849921011058,
      "grad_norm": 0.29605984687805176,
      "learning_rate": 0.0001964292453548002,
      "loss": 0.22202306985855103,
      "mean_token_accuracy": 0.9160446226596832,
      "num_tokens": 19244160.0,
      "step": 1560
    },
    {
      "entropy": 1.1573190689086914,
      "epoch": 0.8220115850447604,
      "grad_norm": 0.3093453347682953,
      "learning_rate": 0.00019642312735241148,
      "loss": 0.22188301384449005,
      "mean_token_accuracy": 0.9117791652679443,
      "num_tokens": 19256496.0,
      "step": 1561
    },
    {
      "entropy": 1.0756475329399109,
      "epoch": 0.822538177988415,
      "grad_norm": 0.29175224900245667,
      "learning_rate": 0.00019641700421958543,
      "loss": 0.20454362034797668,
      "mean_token_accuracy": 0.9208622127771378,
      "num_tokens": 19268832.0,
      "step": 1562
    },
    {
      "entropy": 1.1309452950954437,
      "epoch": 0.8230647709320695,
      "grad_norm": 0.3055620789527893,
      "learning_rate": 0.00019641087595668547,
      "loss": 0.24093538522720337,
      "mean_token_accuracy": 0.9080805480480194,
      "num_tokens": 19281168.0,
      "step": 1563
    },
    {
      "entropy": 1.0666328370571136,
      "epoch": 0.8235913638757241,
      "grad_norm": 0.26273226737976074,
      "learning_rate": 0.00019640474256407545,
      "loss": 0.18816737830638885,
      "mean_token_accuracy": 0.9207744300365448,
      "num_tokens": 19293504.0,
      "step": 1564
    },
    {
      "entropy": 1.1276142299175262,
      "epoch": 0.8241179568193786,
      "grad_norm": 0.3050276041030884,
      "learning_rate": 0.0001963986040421195,
      "loss": 0.21849283576011658,
      "mean_token_accuracy": 0.9142372459173203,
      "num_tokens": 19305840.0,
      "step": 1565
    },
    {
      "entropy": 1.0887978076934814,
      "epoch": 0.8246445497630331,
      "grad_norm": 0.2735901176929474,
      "learning_rate": 0.00019639246039118198,
      "loss": 0.20344410836696625,
      "mean_token_accuracy": 0.9216662347316742,
      "num_tokens": 19318176.0,
      "step": 1566
    },
    {
      "entropy": 1.0880299508571625,
      "epoch": 0.8251711427066877,
      "grad_norm": 0.2992630898952484,
      "learning_rate": 0.00019638631161162761,
      "loss": 0.20155635476112366,
      "mean_token_accuracy": 0.9177606403827667,
      "num_tokens": 19330512.0,
      "step": 1567
    },
    {
      "entropy": 1.0990647375583649,
      "epoch": 0.8256977356503423,
      "grad_norm": 0.29473257064819336,
      "learning_rate": 0.00019638015770382142,
      "loss": 0.20519070327281952,
      "mean_token_accuracy": 0.9206506013870239,
      "num_tokens": 19342848.0,
      "step": 1568
    },
    {
      "entropy": 1.0590737760066986,
      "epoch": 0.8262243285939969,
      "grad_norm": 0.28194659948349,
      "learning_rate": 0.00019637399866812873,
      "loss": 0.20845578610897064,
      "mean_token_accuracy": 0.9180682003498077,
      "num_tokens": 19355184.0,
      "step": 1569
    },
    {
      "entropy": 1.1114600002765656,
      "epoch": 0.8267509215376514,
      "grad_norm": 0.3110455870628357,
      "learning_rate": 0.00019636783450491517,
      "loss": 0.2090432196855545,
      "mean_token_accuracy": 0.9182581901550293,
      "num_tokens": 19367520.0,
      "step": 1570
    },
    {
      "entropy": 1.159915953874588,
      "epoch": 0.8272775144813059,
      "grad_norm": 0.3254052400588989,
      "learning_rate": 0.00019636166521454668,
      "loss": 0.21491852402687073,
      "mean_token_accuracy": 0.9105751514434814,
      "num_tokens": 19379856.0,
      "step": 1571
    },
    {
      "entropy": 1.1465847194194794,
      "epoch": 0.8278041074249605,
      "grad_norm": 0.28690120577812195,
      "learning_rate": 0.00019635549079738946,
      "loss": 0.21204861998558044,
      "mean_token_accuracy": 0.9141225218772888,
      "num_tokens": 19392192.0,
      "step": 1572
    },
    {
      "entropy": 1.1238433420658112,
      "epoch": 0.8283307003686151,
      "grad_norm": 0.3004522919654846,
      "learning_rate": 0.0001963493112538101,
      "loss": 0.22797521948814392,
      "mean_token_accuracy": 0.9102497845888138,
      "num_tokens": 19404528.0,
      "step": 1573
    },
    {
      "entropy": 1.1517022848129272,
      "epoch": 0.8288572933122696,
      "grad_norm": 0.3072613477706909,
      "learning_rate": 0.00019634312658417538,
      "loss": 0.2027738243341446,
      "mean_token_accuracy": 0.9213471561670303,
      "num_tokens": 19416864.0,
      "step": 1574
    },
    {
      "entropy": 1.1321575045585632,
      "epoch": 0.8293838862559242,
      "grad_norm": 0.33203771710395813,
      "learning_rate": 0.0001963369367888525,
      "loss": 0.23440833389759064,
      "mean_token_accuracy": 0.9052640497684479,
      "num_tokens": 19429200.0,
      "step": 1575
    },
    {
      "entropy": 1.1507805287837982,
      "epoch": 0.8299104791995787,
      "grad_norm": 0.28893306851387024,
      "learning_rate": 0.00019633074186820886,
      "loss": 0.21041327714920044,
      "mean_token_accuracy": 0.9186779856681824,
      "num_tokens": 19441536.0,
      "step": 1576
    },
    {
      "entropy": 1.1217812597751617,
      "epoch": 0.8304370721432333,
      "grad_norm": 0.286567747592926,
      "learning_rate": 0.00019632454182261228,
      "loss": 0.20225471258163452,
      "mean_token_accuracy": 0.9235644489526749,
      "num_tokens": 19453872.0,
      "step": 1577
    },
    {
      "entropy": 1.1518227756023407,
      "epoch": 0.8309636650868878,
      "grad_norm": 0.30317631363868713,
      "learning_rate": 0.00019631833665243075,
      "loss": 0.22191138565540314,
      "mean_token_accuracy": 0.9106118381023407,
      "num_tokens": 19466208.0,
      "step": 1578
    },
    {
      "entropy": 1.164818674325943,
      "epoch": 0.8314902580305424,
      "grad_norm": 0.3264894187450409,
      "learning_rate": 0.0001963121263580327,
      "loss": 0.22096458077430725,
      "mean_token_accuracy": 0.9108817130327225,
      "num_tokens": 19478544.0,
      "step": 1579
    },
    {
      "entropy": 1.0784786641597748,
      "epoch": 0.832016850974197,
      "grad_norm": 0.2952364683151245,
      "learning_rate": 0.0001963059109397868,
      "loss": 0.2409459352493286,
      "mean_token_accuracy": 0.9058417230844498,
      "num_tokens": 19490880.0,
      "step": 1580
    },
    {
      "entropy": 1.0973447263240814,
      "epoch": 0.8325434439178515,
      "grad_norm": 0.2781769931316376,
      "learning_rate": 0.00019629969039806195,
      "loss": 0.2172403782606125,
      "mean_token_accuracy": 0.9110274314880371,
      "num_tokens": 19503216.0,
      "step": 1581
    },
    {
      "entropy": 1.1422053277492523,
      "epoch": 0.833070036861506,
      "grad_norm": 0.2971686124801636,
      "learning_rate": 0.0001962934647332275,
      "loss": 0.21011781692504883,
      "mean_token_accuracy": 0.9153971076011658,
      "num_tokens": 19515552.0,
      "step": 1582
    },
    {
      "entropy": 1.1796616911888123,
      "epoch": 0.8335966298051606,
      "grad_norm": 0.29507434368133545,
      "learning_rate": 0.000196287233945653,
      "loss": 0.19983485341072083,
      "mean_token_accuracy": 0.9158152490854263,
      "num_tokens": 19527888.0,
      "step": 1583
    },
    {
      "entropy": 1.1146294176578522,
      "epoch": 0.8341232227488151,
      "grad_norm": 0.275584876537323,
      "learning_rate": 0.00019628099803570837,
      "loss": 0.2035273015499115,
      "mean_token_accuracy": 0.9163656979799271,
      "num_tokens": 19540224.0,
      "step": 1584
    },
    {
      "entropy": 1.1470642983913422,
      "epoch": 0.8346498156924698,
      "grad_norm": 0.2765231430530548,
      "learning_rate": 0.00019627475700376374,
      "loss": 0.22747623920440674,
      "mean_token_accuracy": 0.9137639552354813,
      "num_tokens": 19552560.0,
      "step": 1585
    },
    {
      "entropy": 1.1012212336063385,
      "epoch": 0.8351764086361243,
      "grad_norm": 0.28456148505210876,
      "learning_rate": 0.00019626851085018964,
      "loss": 0.2105616182088852,
      "mean_token_accuracy": 0.9195501506328583,
      "num_tokens": 19564896.0,
      "step": 1586
    },
    {
      "entropy": 1.106357455253601,
      "epoch": 0.8357030015797788,
      "grad_norm": 0.32695838809013367,
      "learning_rate": 0.00019626225957535684,
      "loss": 0.22337135672569275,
      "mean_token_accuracy": 0.9134061336517334,
      "num_tokens": 19577232.0,
      "step": 1587
    },
    {
      "entropy": 1.169274091720581,
      "epoch": 0.8362295945234334,
      "grad_norm": 0.336927205324173,
      "learning_rate": 0.0001962560031796365,
      "loss": 0.2262648195028305,
      "mean_token_accuracy": 0.9084394127130508,
      "num_tokens": 19589568.0,
      "step": 1588
    },
    {
      "entropy": 1.1865193247795105,
      "epoch": 0.8367561874670879,
      "grad_norm": 0.3452165722846985,
      "learning_rate": 0.00019624974166339998,
      "loss": 0.24824325740337372,
      "mean_token_accuracy": 0.9022375047206879,
      "num_tokens": 19601904.0,
      "step": 1589
    },
    {
      "entropy": 1.1062462031841278,
      "epoch": 0.8372827804107424,
      "grad_norm": 0.3023000955581665,
      "learning_rate": 0.000196243475027019,
      "loss": 0.23354491591453552,
      "mean_token_accuracy": 0.9097914397716522,
      "num_tokens": 19614240.0,
      "step": 1590
    },
    {
      "entropy": 1.1293524503707886,
      "epoch": 0.8378093733543971,
      "grad_norm": 0.2942156195640564,
      "learning_rate": 0.00019623720327086557,
      "loss": 0.22122958302497864,
      "mean_token_accuracy": 0.9109482020139694,
      "num_tokens": 19626576.0,
      "step": 1591
    },
    {
      "entropy": 1.124775916337967,
      "epoch": 0.8383359662980516,
      "grad_norm": 0.315075546503067,
      "learning_rate": 0.00019623092639531198,
      "loss": 0.22557595372200012,
      "mean_token_accuracy": 0.9101526886224747,
      "num_tokens": 19638912.0,
      "step": 1592
    },
    {
      "entropy": 1.148577243089676,
      "epoch": 0.8388625592417062,
      "grad_norm": 0.43819931149482727,
      "learning_rate": 0.0001962246444007309,
      "loss": 0.22852541506290436,
      "mean_token_accuracy": 0.9057890176773071,
      "num_tokens": 19651248.0,
      "step": 1593
    },
    {
      "entropy": 1.14993816614151,
      "epoch": 0.8393891521853607,
      "grad_norm": 0.31310200691223145,
      "learning_rate": 0.00019621835728749525,
      "loss": 0.2385086715221405,
      "mean_token_accuracy": 0.9071990847587585,
      "num_tokens": 19663584.0,
      "step": 1594
    },
    {
      "entropy": 1.1644853949546814,
      "epoch": 0.8399157451290152,
      "grad_norm": 0.2970283031463623,
      "learning_rate": 0.00019621206505597823,
      "loss": 0.2200639396905899,
      "mean_token_accuracy": 0.9170584827661514,
      "num_tokens": 19675920.0,
      "step": 1595
    },
    {
      "entropy": 1.1328399777412415,
      "epoch": 0.8404423380726699,
      "grad_norm": 0.3323354125022888,
      "learning_rate": 0.0001962057677065534,
      "loss": 0.22121144831180573,
      "mean_token_accuracy": 0.9114271998405457,
      "num_tokens": 19688256.0,
      "step": 1596
    },
    {
      "entropy": 1.0829459726810455,
      "epoch": 0.8409689310163244,
      "grad_norm": 0.28420135378837585,
      "learning_rate": 0.0001961994652395946,
      "loss": 0.2405356764793396,
      "mean_token_accuracy": 0.9100402593612671,
      "num_tokens": 19700592.0,
      "step": 1597
    },
    {
      "entropy": 1.0924118757247925,
      "epoch": 0.841495523959979,
      "grad_norm": 0.25400787591934204,
      "learning_rate": 0.00019619315765547594,
      "loss": 0.18612799048423767,
      "mean_token_accuracy": 0.9215864688158035,
      "num_tokens": 19712928.0,
      "step": 1598
    },
    {
      "entropy": 1.1357889473438263,
      "epoch": 0.8420221169036335,
      "grad_norm": 0.2727976143360138,
      "learning_rate": 0.00019618684495457188,
      "loss": 0.22136437892913818,
      "mean_token_accuracy": 0.9126606434583664,
      "num_tokens": 19725264.0,
      "step": 1599
    },
    {
      "entropy": 1.0817617177963257,
      "epoch": 0.842548709847288,
      "grad_norm": 0.2677607238292694,
      "learning_rate": 0.0001961805271372572,
      "loss": 0.20533126592636108,
      "mean_token_accuracy": 0.9148357510566711,
      "num_tokens": 19737600.0,
      "step": 1600
    },
    {
      "entropy": 1.1030884683132172,
      "epoch": 0.8430753027909426,
      "grad_norm": 0.2927640378475189,
      "learning_rate": 0.0001961742042039069,
      "loss": 0.22252412140369415,
      "mean_token_accuracy": 0.9111676663160324,
      "num_tokens": 19749936.0,
      "step": 1601
    },
    {
      "entropy": 1.1529828608036041,
      "epoch": 0.8436018957345972,
      "grad_norm": 0.30727458000183105,
      "learning_rate": 0.00019616787615489635,
      "loss": 0.2029101848602295,
      "mean_token_accuracy": 0.9227928221225739,
      "num_tokens": 19762272.0,
      "step": 1602
    },
    {
      "entropy": 1.0316384136676788,
      "epoch": 0.8441284886782517,
      "grad_norm": 0.27791184186935425,
      "learning_rate": 0.00019616154299060122,
      "loss": 0.21758908033370972,
      "mean_token_accuracy": 0.9157789200544357,
      "num_tokens": 19774608.0,
      "step": 1603
    },
    {
      "entropy": 1.0185693204402924,
      "epoch": 0.8446550816219063,
      "grad_norm": 0.2891453504562378,
      "learning_rate": 0.00019615520471139749,
      "loss": 0.19542157649993896,
      "mean_token_accuracy": 0.9184477776288986,
      "num_tokens": 19786944.0,
      "step": 1604
    },
    {
      "entropy": 1.010437250137329,
      "epoch": 0.8451816745655608,
      "grad_norm": 0.29489409923553467,
      "learning_rate": 0.00019614886131766137,
      "loss": 0.2418351173400879,
      "mean_token_accuracy": 0.9019056260585785,
      "num_tokens": 19799280.0,
      "step": 1605
    },
    {
      "entropy": 1.0607281923294067,
      "epoch": 0.8457082675092154,
      "grad_norm": 0.2774660587310791,
      "learning_rate": 0.00019614251280976948,
      "loss": 0.1955130398273468,
      "mean_token_accuracy": 0.9208610653877258,
      "num_tokens": 19811616.0,
      "step": 1606
    },
    {
      "entropy": 1.0406393706798553,
      "epoch": 0.8462348604528699,
      "grad_norm": 0.3048405647277832,
      "learning_rate": 0.00019613615918809868,
      "loss": 0.24297615885734558,
      "mean_token_accuracy": 0.9006519466638565,
      "num_tokens": 19823952.0,
      "step": 1607
    },
    {
      "entropy": 0.9977871924638748,
      "epoch": 0.8467614533965245,
      "grad_norm": 0.29823529720306396,
      "learning_rate": 0.00019612980045302617,
      "loss": 0.22520115971565247,
      "mean_token_accuracy": 0.9101865142583847,
      "num_tokens": 19836288.0,
      "step": 1608
    },
    {
      "entropy": 0.9967030137777328,
      "epoch": 0.8472880463401791,
      "grad_norm": 0.30385443568229675,
      "learning_rate": 0.0001961234366049294,
      "loss": 0.2113213837146759,
      "mean_token_accuracy": 0.9132890105247498,
      "num_tokens": 19848624.0,
      "step": 1609
    },
    {
      "entropy": 0.9831344783306122,
      "epoch": 0.8478146392838336,
      "grad_norm": 0.2813466787338257,
      "learning_rate": 0.00019611706764418617,
      "loss": 0.20107880234718323,
      "mean_token_accuracy": 0.9161145985126495,
      "num_tokens": 19860960.0,
      "step": 1610
    },
    {
      "entropy": 1.0240805000066757,
      "epoch": 0.8483412322274881,
      "grad_norm": 0.3142811954021454,
      "learning_rate": 0.00019611069357117455,
      "loss": 0.20909224450588226,
      "mean_token_accuracy": 0.9144964218139648,
      "num_tokens": 19873296.0,
      "step": 1611
    },
    {
      "entropy": 0.9786220341920853,
      "epoch": 0.8488678251711427,
      "grad_norm": 0.2860144376754761,
      "learning_rate": 0.00019610431438627296,
      "loss": 0.20540790259838104,
      "mean_token_accuracy": 0.9215816557407379,
      "num_tokens": 19885632.0,
      "step": 1612
    },
    {
      "entropy": 1.0295808017253876,
      "epoch": 0.8493944181147972,
      "grad_norm": 0.2819706201553345,
      "learning_rate": 0.00019609793008986003,
      "loss": 0.2058688849210739,
      "mean_token_accuracy": 0.9158887714147568,
      "num_tokens": 19897968.0,
      "step": 1613
    },
    {
      "entropy": 1.0835894644260406,
      "epoch": 0.8499210110584519,
      "grad_norm": 0.2932409346103668,
      "learning_rate": 0.00019609154068231486,
      "loss": 0.2193165421485901,
      "mean_token_accuracy": 0.9130483567714691,
      "num_tokens": 19910304.0,
      "step": 1614
    },
    {
      "entropy": 1.0591089725494385,
      "epoch": 0.8504476040021064,
      "grad_norm": 0.3336801528930664,
      "learning_rate": 0.00019608514616401668,
      "loss": 0.2534202039241791,
      "mean_token_accuracy": 0.9020056575536728,
      "num_tokens": 19922640.0,
      "step": 1615
    },
    {
      "entropy": 1.0837358832359314,
      "epoch": 0.8509741969457609,
      "grad_norm": 0.2906700372695923,
      "learning_rate": 0.00019607874653534513,
      "loss": 0.21131698787212372,
      "mean_token_accuracy": 0.9210411906242371,
      "num_tokens": 19934976.0,
      "step": 1616
    },
    {
      "entropy": 1.0396023094654083,
      "epoch": 0.8515007898894155,
      "grad_norm": 0.29053011536598206,
      "learning_rate": 0.0001960723417966801,
      "loss": 0.21710500121116638,
      "mean_token_accuracy": 0.914475828409195,
      "num_tokens": 19947312.0,
      "step": 1617
    },
    {
      "entropy": 1.0233092904090881,
      "epoch": 0.85202738283307,
      "grad_norm": 0.26935815811157227,
      "learning_rate": 0.00019606593194840177,
      "loss": 0.22018460929393768,
      "mean_token_accuracy": 0.9134867042303085,
      "num_tokens": 19959648.0,
      "step": 1618
    },
    {
      "entropy": 1.0701540112495422,
      "epoch": 0.8525539757767245,
      "grad_norm": 0.28057801723480225,
      "learning_rate": 0.00019605951699089075,
      "loss": 0.21612004935741425,
      "mean_token_accuracy": 0.9129879325628281,
      "num_tokens": 19971984.0,
      "step": 1619
    },
    {
      "entropy": 1.012107402086258,
      "epoch": 0.8530805687203792,
      "grad_norm": 0.2693917155265808,
      "learning_rate": 0.00019605309692452774,
      "loss": 0.2114534080028534,
      "mean_token_accuracy": 0.9149720519781113,
      "num_tokens": 19984320.0,
      "step": 1620
    },
    {
      "entropy": 1.0413274019956589,
      "epoch": 0.8536071616640337,
      "grad_norm": 0.27927207946777344,
      "learning_rate": 0.00019604667174969394,
      "loss": 0.21135683357715607,
      "mean_token_accuracy": 0.9143811762332916,
      "num_tokens": 19996656.0,
      "step": 1621
    },
    {
      "entropy": 0.9798619896173477,
      "epoch": 0.8541337546076883,
      "grad_norm": 0.2772925794124603,
      "learning_rate": 0.00019604024146677079,
      "loss": 0.19996821880340576,
      "mean_token_accuracy": 0.9175093472003937,
      "num_tokens": 20008992.0,
      "step": 1622
    },
    {
      "entropy": 0.9713831841945648,
      "epoch": 0.8546603475513428,
      "grad_norm": 0.2909282445907593,
      "learning_rate": 0.00019603380607613994,
      "loss": 0.21484866738319397,
      "mean_token_accuracy": 0.9112430512905121,
      "num_tokens": 20021328.0,
      "step": 1623
    },
    {
      "entropy": 0.9760904163122177,
      "epoch": 0.8551869404949973,
      "grad_norm": 0.2880821228027344,
      "learning_rate": 0.0001960273655781835,
      "loss": 0.21467523276805878,
      "mean_token_accuracy": 0.9168452471494675,
      "num_tokens": 20033664.0,
      "step": 1624
    },
    {
      "entropy": 1.0048719644546509,
      "epoch": 0.855713533438652,
      "grad_norm": 0.3039538860321045,
      "learning_rate": 0.00019602091997328376,
      "loss": 0.21055804193019867,
      "mean_token_accuracy": 0.9170923382043839,
      "num_tokens": 20046000.0,
      "step": 1625
    },
    {
      "entropy": 0.95078045129776,
      "epoch": 0.8562401263823065,
      "grad_norm": 0.28427428007125854,
      "learning_rate": 0.00019601446926182335,
      "loss": 0.2352895438671112,
      "mean_token_accuracy": 0.9064240008592606,
      "num_tokens": 20058336.0,
      "step": 1626
    },
    {
      "entropy": 0.9472018629312515,
      "epoch": 0.856766719325961,
      "grad_norm": 0.2894728183746338,
      "learning_rate": 0.00019600801344418526,
      "loss": 0.2225104421377182,
      "mean_token_accuracy": 0.913215383887291,
      "num_tokens": 20070672.0,
      "step": 1627
    },
    {
      "entropy": 1.0047412514686584,
      "epoch": 0.8572933122696156,
      "grad_norm": 0.30747485160827637,
      "learning_rate": 0.00019600155252075268,
      "loss": 0.2196165919303894,
      "mean_token_accuracy": 0.9105159193277359,
      "num_tokens": 20083008.0,
      "step": 1628
    },
    {
      "entropy": 1.0216589123010635,
      "epoch": 0.8578199052132701,
      "grad_norm": 0.2717617452144623,
      "learning_rate": 0.0001959950864919092,
      "loss": 0.20522895455360413,
      "mean_token_accuracy": 0.9180753976106644,
      "num_tokens": 20095344.0,
      "step": 1629
    },
    {
      "entropy": 1.0321899056434631,
      "epoch": 0.8583464981569247,
      "grad_norm": 0.30837494134902954,
      "learning_rate": 0.00019598861535803863,
      "loss": 0.21629376709461212,
      "mean_token_accuracy": 0.9151955991983414,
      "num_tokens": 20107680.0,
      "step": 1630
    },
    {
      "entropy": 1.0596579015254974,
      "epoch": 0.8588730911005793,
      "grad_norm": 0.3152981996536255,
      "learning_rate": 0.00019598213911952515,
      "loss": 0.24829894304275513,
      "mean_token_accuracy": 0.9071104824542999,
      "num_tokens": 20120016.0,
      "step": 1631
    },
    {
      "entropy": 1.027356281876564,
      "epoch": 0.8593996840442338,
      "grad_norm": 0.2758905589580536,
      "learning_rate": 0.00019597565777675324,
      "loss": 0.20423409342765808,
      "mean_token_accuracy": 0.919173076748848,
      "num_tokens": 20132352.0,
      "step": 1632
    },
    {
      "entropy": 1.0339451730251312,
      "epoch": 0.8599262769878884,
      "grad_norm": 0.2790839672088623,
      "learning_rate": 0.0001959691713301076,
      "loss": 0.20338216423988342,
      "mean_token_accuracy": 0.916516900062561,
      "num_tokens": 20144688.0,
      "step": 1633
    },
    {
      "entropy": 1.0149255990982056,
      "epoch": 0.8604528699315429,
      "grad_norm": 0.2958633005619049,
      "learning_rate": 0.00019596267977997332,
      "loss": 0.22986331582069397,
      "mean_token_accuracy": 0.9077225029468536,
      "num_tokens": 20157024.0,
      "step": 1634
    },
    {
      "entropy": 0.9824042469263077,
      "epoch": 0.8609794628751974,
      "grad_norm": 0.2640792429447174,
      "learning_rate": 0.0001959561831267358,
      "loss": 0.20994797348976135,
      "mean_token_accuracy": 0.9186595976352692,
      "num_tokens": 20169360.0,
      "step": 1635
    },
    {
      "entropy": 0.9684455096721649,
      "epoch": 0.861506055818852,
      "grad_norm": 0.2797081768512726,
      "learning_rate": 0.00019594968137078068,
      "loss": 0.21638056635856628,
      "mean_token_accuracy": 0.9145815968513489,
      "num_tokens": 20181696.0,
      "step": 1636
    },
    {
      "entropy": 0.9729262888431549,
      "epoch": 0.8620326487625066,
      "grad_norm": 0.2565400004386902,
      "learning_rate": 0.0001959431745124939,
      "loss": 0.20372052490711212,
      "mean_token_accuracy": 0.9224633723497391,
      "num_tokens": 20194032.0,
      "step": 1637
    },
    {
      "entropy": 0.9809018820524216,
      "epoch": 0.8625592417061612,
      "grad_norm": 0.3174954354763031,
      "learning_rate": 0.0001959366625522618,
      "loss": 0.2331572026014328,
      "mean_token_accuracy": 0.9117663353681564,
      "num_tokens": 20206368.0,
      "step": 1638
    },
    {
      "entropy": 0.9628456979990005,
      "epoch": 0.8630858346498157,
      "grad_norm": 0.2894734740257263,
      "learning_rate": 0.0001959301454904709,
      "loss": 0.2223999798297882,
      "mean_token_accuracy": 0.9094413071870804,
      "num_tokens": 20218704.0,
      "step": 1639
    },
    {
      "entropy": 1.0126699656248093,
      "epoch": 0.8636124275934702,
      "grad_norm": 0.2930248975753784,
      "learning_rate": 0.0001959236233275081,
      "loss": 0.21124093234539032,
      "mean_token_accuracy": 0.9162924438714981,
      "num_tokens": 20231040.0,
      "step": 1640
    },
    {
      "entropy": 0.9787211865186691,
      "epoch": 0.8641390205371248,
      "grad_norm": 0.30070269107818604,
      "learning_rate": 0.00019591709606376059,
      "loss": 0.21121284365653992,
      "mean_token_accuracy": 0.9159329831600189,
      "num_tokens": 20243376.0,
      "step": 1641
    },
    {
      "entropy": 1.010118991136551,
      "epoch": 0.8646656134807793,
      "grad_norm": 0.28400495648384094,
      "learning_rate": 0.00019591056369961586,
      "loss": 0.21099403500556946,
      "mean_token_accuracy": 0.9190017580986023,
      "num_tokens": 20255712.0,
      "step": 1642
    },
    {
      "entropy": 0.9496536999940872,
      "epoch": 0.865192206424434,
      "grad_norm": 0.2675454914569855,
      "learning_rate": 0.00019590402623546167,
      "loss": 0.2049626260995865,
      "mean_token_accuracy": 0.9156690835952759,
      "num_tokens": 20268048.0,
      "step": 1643
    },
    {
      "entropy": 0.9466981887817383,
      "epoch": 0.8657187993680885,
      "grad_norm": 0.30046844482421875,
      "learning_rate": 0.00019589748367168612,
      "loss": 0.21190345287322998,
      "mean_token_accuracy": 0.9161060154438019,
      "num_tokens": 20280384.0,
      "step": 1644
    },
    {
      "entropy": 1.0232085585594177,
      "epoch": 0.866245392311743,
      "grad_norm": 0.3621319532394409,
      "learning_rate": 0.00019589093600867763,
      "loss": 0.2193715125322342,
      "mean_token_accuracy": 0.9074915945529938,
      "num_tokens": 20292720.0,
      "step": 1645
    },
    {
      "entropy": 0.993518516421318,
      "epoch": 0.8667719852553976,
      "grad_norm": 0.28956446051597595,
      "learning_rate": 0.00019588438324682488,
      "loss": 0.21250395476818085,
      "mean_token_accuracy": 0.9168424904346466,
      "num_tokens": 20305056.0,
      "step": 1646
    },
    {
      "entropy": 0.9817381650209427,
      "epoch": 0.8672985781990521,
      "grad_norm": 0.34931275248527527,
      "learning_rate": 0.00019587782538651687,
      "loss": 0.2343052327632904,
      "mean_token_accuracy": 0.9059536755084991,
      "num_tokens": 20317392.0,
      "step": 1647
    },
    {
      "entropy": 0.9840113967657089,
      "epoch": 0.8678251711427067,
      "grad_norm": 0.304169237613678,
      "learning_rate": 0.00019587126242814288,
      "loss": 0.20760950446128845,
      "mean_token_accuracy": 0.9196920245885849,
      "num_tokens": 20329728.0,
      "step": 1648
    },
    {
      "entropy": 0.99486044049263,
      "epoch": 0.8683517640863613,
      "grad_norm": 0.3464246094226837,
      "learning_rate": 0.00019586469437209256,
      "loss": 0.25210103392601013,
      "mean_token_accuracy": 0.9036094397306442,
      "num_tokens": 20342064.0,
      "step": 1649
    },
    {
      "entropy": 0.9775729477405548,
      "epoch": 0.8688783570300158,
      "grad_norm": 0.27859246730804443,
      "learning_rate": 0.00019585812121875577,
      "loss": 0.21410948038101196,
      "mean_token_accuracy": 0.9136857688426971,
      "num_tokens": 20354400.0,
      "step": 1650
    },
    {
      "entropy": 1.0246230065822601,
      "epoch": 0.8694049499736703,
      "grad_norm": 0.28361114859580994,
      "learning_rate": 0.00019585154296852277,
      "loss": 0.19681544601917267,
      "mean_token_accuracy": 0.9178670197725296,
      "num_tokens": 20366736.0,
      "step": 1651
    },
    {
      "entropy": 0.9918327033519745,
      "epoch": 0.8699315429173249,
      "grad_norm": 0.2778252065181732,
      "learning_rate": 0.00019584495962178403,
      "loss": 0.20562246441841125,
      "mean_token_accuracy": 0.9203788787126541,
      "num_tokens": 20379072.0,
      "step": 1652
    },
    {
      "entropy": 1.0197946578264236,
      "epoch": 0.8704581358609794,
      "grad_norm": 0.2764199674129486,
      "learning_rate": 0.0001958383711789304,
      "loss": 0.2005048245191574,
      "mean_token_accuracy": 0.9199100285768509,
      "num_tokens": 20391408.0,
      "step": 1653
    },
    {
      "entropy": 0.946838304400444,
      "epoch": 0.8709847288046341,
      "grad_norm": 0.318099707365036,
      "learning_rate": 0.00019583177764035295,
      "loss": 0.18645739555358887,
      "mean_token_accuracy": 0.9281531125307083,
      "num_tokens": 20403744.0,
      "step": 1654
    },
    {
      "entropy": 0.9969734102487564,
      "epoch": 0.8715113217482886,
      "grad_norm": 0.330425888299942,
      "learning_rate": 0.00019582517900644313,
      "loss": 0.2349887192249298,
      "mean_token_accuracy": 0.9082667678594589,
      "num_tokens": 20416080.0,
      "step": 1655
    },
    {
      "entropy": 0.9850341528654099,
      "epoch": 0.8720379146919431,
      "grad_norm": 0.30549484491348267,
      "learning_rate": 0.00019581857527759265,
      "loss": 0.22454282641410828,
      "mean_token_accuracy": 0.9096111208200455,
      "num_tokens": 20428416.0,
      "step": 1656
    },
    {
      "entropy": 0.9726726412773132,
      "epoch": 0.8725645076355977,
      "grad_norm": 0.26101022958755493,
      "learning_rate": 0.00019581196645419358,
      "loss": 0.18027372658252716,
      "mean_token_accuracy": 0.9280071556568146,
      "num_tokens": 20440752.0,
      "step": 1657
    },
    {
      "entropy": 1.045421838760376,
      "epoch": 0.8730911005792522,
      "grad_norm": 0.31144458055496216,
      "learning_rate": 0.0001958053525366382,
      "loss": 0.22132211923599243,
      "mean_token_accuracy": 0.9151443690061569,
      "num_tokens": 20453088.0,
      "step": 1658
    },
    {
      "entropy": 1.0683381259441376,
      "epoch": 0.8736176935229067,
      "grad_norm": 0.34451693296432495,
      "learning_rate": 0.0001957987335253191,
      "loss": 0.2497258186340332,
      "mean_token_accuracy": 0.9017534852027893,
      "num_tokens": 20465424.0,
      "step": 1659
    },
    {
      "entropy": 1.0424925982952118,
      "epoch": 0.8741442864665614,
      "grad_norm": 0.305476576089859,
      "learning_rate": 0.00019579210942062932,
      "loss": 0.21568341553211212,
      "mean_token_accuracy": 0.913795217871666,
      "num_tokens": 20477760.0,
      "step": 1660
    },
    {
      "entropy": 1.077263206243515,
      "epoch": 0.8746708794102159,
      "grad_norm": 0.28648725152015686,
      "learning_rate": 0.00019578548022296203,
      "loss": 0.18887600302696228,
      "mean_token_accuracy": 0.9262412637472153,
      "num_tokens": 20490096.0,
      "step": 1661
    },
    {
      "entropy": 1.0266597121953964,
      "epoch": 0.8751974723538705,
      "grad_norm": 0.31522682309150696,
      "learning_rate": 0.00019577884593271076,
      "loss": 0.21652202308177948,
      "mean_token_accuracy": 0.907884418964386,
      "num_tokens": 20502432.0,
      "step": 1662
    },
    {
      "entropy": 1.0532887279987335,
      "epoch": 0.875724065297525,
      "grad_norm": 0.25786280632019043,
      "learning_rate": 0.00019577220655026938,
      "loss": 0.18014520406723022,
      "mean_token_accuracy": 0.9304206520318985,
      "num_tokens": 20514768.0,
      "step": 1663
    },
    {
      "entropy": 1.0302300453186035,
      "epoch": 0.8762506582411795,
      "grad_norm": 0.27384263277053833,
      "learning_rate": 0.000195765562076032,
      "loss": 0.19573301076889038,
      "mean_token_accuracy": 0.9229860901832581,
      "num_tokens": 20527104.0,
      "step": 1664
    },
    {
      "entropy": 1.0142173171043396,
      "epoch": 0.8767772511848341,
      "grad_norm": 0.2829652428627014,
      "learning_rate": 0.0001957589125103931,
      "loss": 0.19770190119743347,
      "mean_token_accuracy": 0.9212857335805893,
      "num_tokens": 20539440.0,
      "step": 1665
    },
    {
      "entropy": 1.030619502067566,
      "epoch": 0.8773038441284887,
      "grad_norm": 0.31889650225639343,
      "learning_rate": 0.0001957522578537474,
      "loss": 0.21395039558410645,
      "mean_token_accuracy": 0.9180440455675125,
      "num_tokens": 20551776.0,
      "step": 1666
    },
    {
      "entropy": 1.0783992409706116,
      "epoch": 0.8778304370721433,
      "grad_norm": 0.31347182393074036,
      "learning_rate": 0.00019574559810648994,
      "loss": 0.23077845573425293,
      "mean_token_accuracy": 0.9107988774776459,
      "num_tokens": 20564112.0,
      "step": 1667
    },
    {
      "entropy": 1.0769988298416138,
      "epoch": 0.8783570300157978,
      "grad_norm": 0.30257898569107056,
      "learning_rate": 0.00019573893326901607,
      "loss": 0.18793995678424835,
      "mean_token_accuracy": 0.9216104596853256,
      "num_tokens": 20576448.0,
      "step": 1668
    },
    {
      "entropy": 1.041060984134674,
      "epoch": 0.8788836229594523,
      "grad_norm": 0.29523128271102905,
      "learning_rate": 0.0001957322633417215,
      "loss": 0.20431648194789886,
      "mean_token_accuracy": 0.9192978888750076,
      "num_tokens": 20588784.0,
      "step": 1669
    },
    {
      "entropy": 1.0586915016174316,
      "epoch": 0.8794102159031069,
      "grad_norm": 0.29364022612571716,
      "learning_rate": 0.0001957255883250021,
      "loss": 0.22029076516628265,
      "mean_token_accuracy": 0.9125624299049377,
      "num_tokens": 20601120.0,
      "step": 1670
    },
    {
      "entropy": 1.0633411705493927,
      "epoch": 0.8799368088467614,
      "grad_norm": 0.2905363440513611,
      "learning_rate": 0.00019571890821925417,
      "loss": 0.19489197432994843,
      "mean_token_accuracy": 0.9242742508649826,
      "num_tokens": 20613456.0,
      "step": 1671
    },
    {
      "entropy": 1.135309487581253,
      "epoch": 0.880463401790416,
      "grad_norm": 0.3044449985027313,
      "learning_rate": 0.0001957122230248743,
      "loss": 0.19300708174705505,
      "mean_token_accuracy": 0.9245816022157669,
      "num_tokens": 20625792.0,
      "step": 1672
    },
    {
      "entropy": 1.0445529222488403,
      "epoch": 0.8809899947340706,
      "grad_norm": 0.28844696283340454,
      "learning_rate": 0.00019570553274225928,
      "loss": 0.20883029699325562,
      "mean_token_accuracy": 0.9140314012765884,
      "num_tokens": 20638128.0,
      "step": 1673
    },
    {
      "entropy": 1.1107747852802277,
      "epoch": 0.8815165876777251,
      "grad_norm": 0.32846325635910034,
      "learning_rate": 0.00019569883737180634,
      "loss": 0.24370679259300232,
      "mean_token_accuracy": 0.901743620634079,
      "num_tokens": 20650464.0,
      "step": 1674
    },
    {
      "entropy": 1.1209677755832672,
      "epoch": 0.8820431806213797,
      "grad_norm": 0.2889225482940674,
      "learning_rate": 0.0001956921369139129,
      "loss": 0.21621105074882507,
      "mean_token_accuracy": 0.9143183678388596,
      "num_tokens": 20662800.0,
      "step": 1675
    },
    {
      "entropy": 1.0856188535690308,
      "epoch": 0.8825697735650342,
      "grad_norm": 0.27882513403892517,
      "learning_rate": 0.00019568543136897675,
      "loss": 0.21793431043624878,
      "mean_token_accuracy": 0.9142801910638809,
      "num_tokens": 20675136.0,
      "step": 1676
    },
    {
      "entropy": 1.0983311533927917,
      "epoch": 0.8830963665086888,
      "grad_norm": 0.272402822971344,
      "learning_rate": 0.00019567872073739595,
      "loss": 0.20574277639389038,
      "mean_token_accuracy": 0.9181714653968811,
      "num_tokens": 20687472.0,
      "step": 1677
    },
    {
      "entropy": 1.0844581425189972,
      "epoch": 0.8836229594523434,
      "grad_norm": 0.27611178159713745,
      "learning_rate": 0.0001956720050195689,
      "loss": 0.2087741196155548,
      "mean_token_accuracy": 0.9139288067817688,
      "num_tokens": 20699808.0,
      "step": 1678
    },
    {
      "entropy": 1.12508624792099,
      "epoch": 0.8841495523959979,
      "grad_norm": 0.29753515124320984,
      "learning_rate": 0.0001956652842158942,
      "loss": 0.20983164012432098,
      "mean_token_accuracy": 0.9156063199043274,
      "num_tokens": 20712144.0,
      "step": 1679
    },
    {
      "entropy": 1.1237779259681702,
      "epoch": 0.8846761453396524,
      "grad_norm": 0.2853649854660034,
      "learning_rate": 0.00019565855832677086,
      "loss": 0.20436659455299377,
      "mean_token_accuracy": 0.9203456342220306,
      "num_tokens": 20724480.0,
      "step": 1680
    },
    {
      "entropy": 1.1044580936431885,
      "epoch": 0.885202738283307,
      "grad_norm": 0.2854003608226776,
      "learning_rate": 0.00019565182735259822,
      "loss": 0.2195129096508026,
      "mean_token_accuracy": 0.9125242233276367,
      "num_tokens": 20736816.0,
      "step": 1681
    },
    {
      "entropy": 1.084385484457016,
      "epoch": 0.8857293312269615,
      "grad_norm": 0.29433006048202515,
      "learning_rate": 0.0001956450912937758,
      "loss": 0.2089007943868637,
      "mean_token_accuracy": 0.9135125875473022,
      "num_tokens": 20749152.0,
      "step": 1682
    },
    {
      "entropy": 1.0872920453548431,
      "epoch": 0.8862559241706162,
      "grad_norm": 0.2716350555419922,
      "learning_rate": 0.00019563835015070348,
      "loss": 0.20034612715244293,
      "mean_token_accuracy": 0.9140070527791977,
      "num_tokens": 20761488.0,
      "step": 1683
    },
    {
      "entropy": 1.0698755085468292,
      "epoch": 0.8867825171142707,
      "grad_norm": 0.263379842042923,
      "learning_rate": 0.00019563160392378144,
      "loss": 0.2013968825340271,
      "mean_token_accuracy": 0.920297846198082,
      "num_tokens": 20773824.0,
      "step": 1684
    },
    {
      "entropy": 1.0688921809196472,
      "epoch": 0.8873091100579252,
      "grad_norm": 0.2941676676273346,
      "learning_rate": 0.00019562485261341017,
      "loss": 0.222881019115448,
      "mean_token_accuracy": 0.9110155403614044,
      "num_tokens": 20786160.0,
      "step": 1685
    },
    {
      "entropy": 1.0462908446788788,
      "epoch": 0.8878357030015798,
      "grad_norm": 0.2833232581615448,
      "learning_rate": 0.00019561809621999047,
      "loss": 0.2037080079317093,
      "mean_token_accuracy": 0.9212964028120041,
      "num_tokens": 20798496.0,
      "step": 1686
    },
    {
      "entropy": 1.0370090901851654,
      "epoch": 0.8883622959452343,
      "grad_norm": 0.299941748380661,
      "learning_rate": 0.0001956113347439234,
      "loss": 0.21904562413692474,
      "mean_token_accuracy": 0.9075719863176346,
      "num_tokens": 20810832.0,
      "step": 1687
    },
    {
      "entropy": 1.0448128283023834,
      "epoch": 0.8888888888888888,
      "grad_norm": 0.2878952920436859,
      "learning_rate": 0.0001956045681856104,
      "loss": 0.2038077563047409,
      "mean_token_accuracy": 0.9187713265419006,
      "num_tokens": 20823168.0,
      "step": 1688
    },
    {
      "entropy": 1.0190322399139404,
      "epoch": 0.8894154818325435,
      "grad_norm": 0.34769734740257263,
      "learning_rate": 0.0001955977965454531,
      "loss": 0.213621124625206,
      "mean_token_accuracy": 0.9181492775678635,
      "num_tokens": 20835504.0,
      "step": 1689
    },
    {
      "entropy": 1.0234404504299164,
      "epoch": 0.889942074776198,
      "grad_norm": 0.2829277813434601,
      "learning_rate": 0.00019559101982385356,
      "loss": 0.20716841518878937,
      "mean_token_accuracy": 0.9201241731643677,
      "num_tokens": 20847840.0,
      "step": 1690
    },
    {
      "entropy": 1.0015476644039154,
      "epoch": 0.8904686677198526,
      "grad_norm": 0.26036950945854187,
      "learning_rate": 0.000195584238021214,
      "loss": 0.18766967952251434,
      "mean_token_accuracy": 0.9250665605068207,
      "num_tokens": 20860176.0,
      "step": 1691
    },
    {
      "entropy": 1.032362923026085,
      "epoch": 0.8909952606635071,
      "grad_norm": 0.4818808138370514,
      "learning_rate": 0.00019557745113793704,
      "loss": 0.19411087036132812,
      "mean_token_accuracy": 0.921613797545433,
      "num_tokens": 20872512.0,
      "step": 1692
    },
    {
      "entropy": 0.9709571003913879,
      "epoch": 0.8915218536071616,
      "grad_norm": 0.2804003655910492,
      "learning_rate": 0.0001955706591744256,
      "loss": 0.21858839690685272,
      "mean_token_accuracy": 0.9120522737503052,
      "num_tokens": 20884848.0,
      "step": 1693
    },
    {
      "entropy": 1.0739005506038666,
      "epoch": 0.8920484465508162,
      "grad_norm": 0.2946876585483551,
      "learning_rate": 0.00019556386213108288,
      "loss": 0.19359217584133148,
      "mean_token_accuracy": 0.925596609711647,
      "num_tokens": 20897184.0,
      "step": 1694
    },
    {
      "entropy": 1.0181286334991455,
      "epoch": 0.8925750394944708,
      "grad_norm": 0.3197228014469147,
      "learning_rate": 0.00019555706000831235,
      "loss": 0.2175355702638626,
      "mean_token_accuracy": 0.9139997512102127,
      "num_tokens": 20909520.0,
      "step": 1695
    },
    {
      "entropy": 1.0345196276903152,
      "epoch": 0.8931016324381253,
      "grad_norm": 0.29833653569221497,
      "learning_rate": 0.00019555025280651786,
      "loss": 0.2255219966173172,
      "mean_token_accuracy": 0.9102495014667511,
      "num_tokens": 20921856.0,
      "step": 1696
    },
    {
      "entropy": 0.9587903320789337,
      "epoch": 0.8936282253817799,
      "grad_norm": 0.2722357511520386,
      "learning_rate": 0.00019554344052610348,
      "loss": 0.19479139149188995,
      "mean_token_accuracy": 0.9188268184661865,
      "num_tokens": 20934192.0,
      "step": 1697
    },
    {
      "entropy": 0.9713253229856491,
      "epoch": 0.8941548183254344,
      "grad_norm": 0.2900646924972534,
      "learning_rate": 0.0001955366231674736,
      "loss": 0.20535482466220856,
      "mean_token_accuracy": 0.9150903820991516,
      "num_tokens": 20946528.0,
      "step": 1698
    },
    {
      "entropy": 1.020512416958809,
      "epoch": 0.894681411269089,
      "grad_norm": 0.2782817780971527,
      "learning_rate": 0.00019552980073103297,
      "loss": 0.20126226544380188,
      "mean_token_accuracy": 0.9161560386419296,
      "num_tokens": 20958864.0,
      "step": 1699
    },
    {
      "entropy": 0.9827070981264114,
      "epoch": 0.8952080042127436,
      "grad_norm": 0.2738768458366394,
      "learning_rate": 0.00019552297321718658,
      "loss": 0.19939669966697693,
      "mean_token_accuracy": 0.9180780202150345,
      "num_tokens": 20971200.0,
      "step": 1700
    },
    {
      "entropy": 0.9992471039295197,
      "epoch": 0.8957345971563981,
      "grad_norm": 0.31194329261779785,
      "learning_rate": 0.0001955161406263397,
      "loss": 0.23202218115329742,
      "mean_token_accuracy": 0.9068114459514618,
      "num_tokens": 20983536.0,
      "step": 1701
    },
    {
      "entropy": 0.9976187497377396,
      "epoch": 0.8962611901000527,
      "grad_norm": 0.29019710421562195,
      "learning_rate": 0.00019550930295889803,
      "loss": 0.21963413059711456,
      "mean_token_accuracy": 0.9133173823356628,
      "num_tokens": 20995872.0,
      "step": 1702
    },
    {
      "entropy": 1.0274995416402817,
      "epoch": 0.8967877830437072,
      "grad_norm": 0.2910099923610687,
      "learning_rate": 0.00019550246021526736,
      "loss": 0.20295658707618713,
      "mean_token_accuracy": 0.9164696484804153,
      "num_tokens": 21008208.0,
      "step": 1703
    },
    {
      "entropy": 1.0305428206920624,
      "epoch": 0.8973143759873617,
      "grad_norm": 0.3525075912475586,
      "learning_rate": 0.000195495612395854,
      "loss": 0.20155680179595947,
      "mean_token_accuracy": 0.9174475073814392,
      "num_tokens": 21020544.0,
      "step": 1704
    },
    {
      "entropy": 1.0068664401769638,
      "epoch": 0.8978409689310163,
      "grad_norm": 0.33528581261634827,
      "learning_rate": 0.00019548875950106448,
      "loss": 0.22776372730731964,
      "mean_token_accuracy": 0.906059592962265,
      "num_tokens": 21032880.0,
      "step": 1705
    },
    {
      "entropy": 1.0565652251243591,
      "epoch": 0.8983675618746709,
      "grad_norm": 0.29654526710510254,
      "learning_rate": 0.00019548190153130553,
      "loss": 0.23316293954849243,
      "mean_token_accuracy": 0.9070921689271927,
      "num_tokens": 21045216.0,
      "step": 1706
    },
    {
      "entropy": 0.9845771491527557,
      "epoch": 0.8988941548183255,
      "grad_norm": 0.3172268271446228,
      "learning_rate": 0.00019547503848698435,
      "loss": 0.22646287083625793,
      "mean_token_accuracy": 0.9135229587554932,
      "num_tokens": 21057552.0,
      "step": 1707
    },
    {
      "entropy": 1.014034628868103,
      "epoch": 0.89942074776198,
      "grad_norm": 0.27241358160972595,
      "learning_rate": 0.00019546817036850827,
      "loss": 0.21421730518341064,
      "mean_token_accuracy": 0.9127627313137054,
      "num_tokens": 21069888.0,
      "step": 1708
    },
    {
      "entropy": 1.044843077659607,
      "epoch": 0.8999473407056345,
      "grad_norm": 0.27422401309013367,
      "learning_rate": 0.00019546129717628512,
      "loss": 0.22259175777435303,
      "mean_token_accuracy": 0.90959133207798,
      "num_tokens": 21082224.0,
      "step": 1709
    },
    {
      "entropy": 1.0896000117063522,
      "epoch": 0.9004739336492891,
      "grad_norm": 0.31072503328323364,
      "learning_rate": 0.00019545441891072283,
      "loss": 0.21754461526870728,
      "mean_token_accuracy": 0.9161411225795746,
      "num_tokens": 21094560.0,
      "step": 1710
    },
    {
      "entropy": 1.0415829718112946,
      "epoch": 0.9010005265929436,
      "grad_norm": 0.261515736579895,
      "learning_rate": 0.00019544753557222975,
      "loss": 0.1926898956298828,
      "mean_token_accuracy": 0.9213140159845352,
      "num_tokens": 21106896.0,
      "step": 1711
    },
    {
      "entropy": 1.0076445639133453,
      "epoch": 0.9015271195365983,
      "grad_norm": 0.26134979724884033,
      "learning_rate": 0.00019544064716121452,
      "loss": 0.20023104548454285,
      "mean_token_accuracy": 0.9194119721651077,
      "num_tokens": 21119232.0,
      "step": 1712
    },
    {
      "entropy": 1.007362276315689,
      "epoch": 0.9020537124802528,
      "grad_norm": 0.2691597044467926,
      "learning_rate": 0.00019543375367808604,
      "loss": 0.19764183461666107,
      "mean_token_accuracy": 0.9207281470298767,
      "num_tokens": 21131568.0,
      "step": 1713
    },
    {
      "entropy": 1.0264256298542023,
      "epoch": 0.9025803054239073,
      "grad_norm": 0.2841010093688965,
      "learning_rate": 0.00019542685512325357,
      "loss": 0.19019414484500885,
      "mean_token_accuracy": 0.9279275834560394,
      "num_tokens": 21143904.0,
      "step": 1714
    },
    {
      "entropy": 1.0494853258132935,
      "epoch": 0.9031068983675619,
      "grad_norm": 0.2881733477115631,
      "learning_rate": 0.00019541995149712662,
      "loss": 0.19694221019744873,
      "mean_token_accuracy": 0.9203800857067108,
      "num_tokens": 21156240.0,
      "step": 1715
    },
    {
      "entropy": 1.029950574040413,
      "epoch": 0.9036334913112164,
      "grad_norm": 0.2856560945510864,
      "learning_rate": 0.00019541304280011498,
      "loss": 0.208457350730896,
      "mean_token_accuracy": 0.9130570143461227,
      "num_tokens": 21168576.0,
      "step": 1716
    },
    {
      "entropy": 1.036653459072113,
      "epoch": 0.9041600842548709,
      "grad_norm": 0.35959112644195557,
      "learning_rate": 0.00019540612903262887,
      "loss": 0.2388800084590912,
      "mean_token_accuracy": 0.909120500087738,
      "num_tokens": 21180912.0,
      "step": 1717
    },
    {
      "entropy": 1.00953871011734,
      "epoch": 0.9046866771985256,
      "grad_norm": 0.3067711293697357,
      "learning_rate": 0.00019539921019507862,
      "loss": 0.2145732194185257,
      "mean_token_accuracy": 0.9115498811006546,
      "num_tokens": 21193248.0,
      "step": 1718
    },
    {
      "entropy": 0.9901564866304398,
      "epoch": 0.9052132701421801,
      "grad_norm": 0.3036840856075287,
      "learning_rate": 0.00019539228628787501,
      "loss": 0.2193184792995453,
      "mean_token_accuracy": 0.9104121774435043,
      "num_tokens": 21205584.0,
      "step": 1719
    },
    {
      "entropy": 0.9771833568811417,
      "epoch": 0.9057398630858347,
      "grad_norm": 0.32093900442123413,
      "learning_rate": 0.00019538535731142907,
      "loss": 0.23529593646526337,
      "mean_token_accuracy": 0.9093421250581741,
      "num_tokens": 21217920.0,
      "step": 1720
    },
    {
      "entropy": 1.011559009552002,
      "epoch": 0.9062664560294892,
      "grad_norm": 0.2692308723926544,
      "learning_rate": 0.00019537842326615215,
      "loss": 0.20593476295471191,
      "mean_token_accuracy": 0.9199935644865036,
      "num_tokens": 21230256.0,
      "step": 1721
    },
    {
      "entropy": 1.073755070567131,
      "epoch": 0.9067930489731437,
      "grad_norm": 0.3065417408943176,
      "learning_rate": 0.00019537148415245582,
      "loss": 0.23454323410987854,
      "mean_token_accuracy": 0.9069335609674454,
      "num_tokens": 21242592.0,
      "step": 1722
    },
    {
      "entropy": 1.0648082047700882,
      "epoch": 0.9073196419167984,
      "grad_norm": 0.32037290930747986,
      "learning_rate": 0.0001953645399707521,
      "loss": 0.2138725221157074,
      "mean_token_accuracy": 0.9131258726119995,
      "num_tokens": 21254928.0,
      "step": 1723
    },
    {
      "entropy": 1.0663966238498688,
      "epoch": 0.9078462348604529,
      "grad_norm": 0.2899954915046692,
      "learning_rate": 0.00019535759072145318,
      "loss": 0.204481303691864,
      "mean_token_accuracy": 0.9170624017715454,
      "num_tokens": 21267264.0,
      "step": 1724
    },
    {
      "entropy": 1.020171880722046,
      "epoch": 0.9083728278041074,
      "grad_norm": 0.29145675897598267,
      "learning_rate": 0.0001953506364049716,
      "loss": 0.21512804925441742,
      "mean_token_accuracy": 0.9173563122749329,
      "num_tokens": 21279600.0,
      "step": 1725
    },
    {
      "entropy": 1.0235641449689865,
      "epoch": 0.908899420747762,
      "grad_norm": 0.27262571454048157,
      "learning_rate": 0.00019534367702172016,
      "loss": 0.22479593753814697,
      "mean_token_accuracy": 0.9110193848609924,
      "num_tokens": 21291936.0,
      "step": 1726
    },
    {
      "entropy": 0.9845803081989288,
      "epoch": 0.9094260136914165,
      "grad_norm": 0.2786709666252136,
      "learning_rate": 0.00019533671257211205,
      "loss": 0.2018449455499649,
      "mean_token_accuracy": 0.9166059494018555,
      "num_tokens": 21304272.0,
      "step": 1727
    },
    {
      "entropy": 0.9907389134168625,
      "epoch": 0.909952606635071,
      "grad_norm": 0.2637747526168823,
      "learning_rate": 0.00019532974305656075,
      "loss": 0.211765855550766,
      "mean_token_accuracy": 0.9151495099067688,
      "num_tokens": 21316608.0,
      "step": 1728
    },
    {
      "entropy": 1.0142314583063126,
      "epoch": 0.9104791995787257,
      "grad_norm": 0.2817028760910034,
      "learning_rate": 0.0001953227684754799,
      "loss": 0.19784370064735413,
      "mean_token_accuracy": 0.9207914024591446,
      "num_tokens": 21328944.0,
      "step": 1729
    },
    {
      "entropy": 1.0088147670030594,
      "epoch": 0.9110057925223802,
      "grad_norm": 0.2688204050064087,
      "learning_rate": 0.00019531578882928357,
      "loss": 0.20449689030647278,
      "mean_token_accuracy": 0.9207023978233337,
      "num_tokens": 21341280.0,
      "step": 1730
    },
    {
      "entropy": 0.9528838694095612,
      "epoch": 0.9115323854660348,
      "grad_norm": 0.27442365884780884,
      "learning_rate": 0.00019530880411838616,
      "loss": 0.21661211550235748,
      "mean_token_accuracy": 0.9121255576610565,
      "num_tokens": 21353616.0,
      "step": 1731
    },
    {
      "entropy": 0.9663679003715515,
      "epoch": 0.9120589784096893,
      "grad_norm": 0.2818276584148407,
      "learning_rate": 0.00019530181434320224,
      "loss": 0.21009200811386108,
      "mean_token_accuracy": 0.9155483096837997,
      "num_tokens": 21365952.0,
      "step": 1732
    },
    {
      "entropy": 0.9394736588001251,
      "epoch": 0.9125855713533438,
      "grad_norm": 0.3315548598766327,
      "learning_rate": 0.0001952948195041468,
      "loss": 0.21075570583343506,
      "mean_token_accuracy": 0.9181785732507706,
      "num_tokens": 21378288.0,
      "step": 1733
    },
    {
      "entropy": 0.9442544430494308,
      "epoch": 0.9131121642969984,
      "grad_norm": 0.3284794092178345,
      "learning_rate": 0.000195287819601635,
      "loss": 0.23087412118911743,
      "mean_token_accuracy": 0.9092219769954681,
      "num_tokens": 21390624.0,
      "step": 1734
    },
    {
      "entropy": 0.9372933954000473,
      "epoch": 0.913638757240653,
      "grad_norm": 0.29973915219306946,
      "learning_rate": 0.0001952808146360825,
      "loss": 0.2058921754360199,
      "mean_token_accuracy": 0.9160513281822205,
      "num_tokens": 21402960.0,
      "step": 1735
    },
    {
      "entropy": 0.9260263890028,
      "epoch": 0.9141653501843076,
      "grad_norm": 0.2917967438697815,
      "learning_rate": 0.00019527380460790508,
      "loss": 0.2268896996974945,
      "mean_token_accuracy": 0.9107523709535599,
      "num_tokens": 21415296.0,
      "step": 1736
    },
    {
      "entropy": 0.9555350244045258,
      "epoch": 0.9146919431279621,
      "grad_norm": 0.2836836874485016,
      "learning_rate": 0.0001952667895175189,
      "loss": 0.21083997189998627,
      "mean_token_accuracy": 0.9108854085206985,
      "num_tokens": 21427632.0,
      "step": 1737
    },
    {
      "entropy": 0.9710684269666672,
      "epoch": 0.9152185360716166,
      "grad_norm": 0.3263842761516571,
      "learning_rate": 0.00019525976936534035,
      "loss": 0.19704914093017578,
      "mean_token_accuracy": 0.9213050603866577,
      "num_tokens": 21439968.0,
      "step": 1738
    },
    {
      "entropy": 0.920185998082161,
      "epoch": 0.9157451290152712,
      "grad_norm": 0.3038593828678131,
      "learning_rate": 0.00019525274415178626,
      "loss": 0.2302832156419754,
      "mean_token_accuracy": 0.9091469496488571,
      "num_tokens": 21452304.0,
      "step": 1739
    },
    {
      "entropy": 0.9555569142103195,
      "epoch": 0.9162717219589257,
      "grad_norm": 0.3209346532821655,
      "learning_rate": 0.00019524571387727365,
      "loss": 0.2268507480621338,
      "mean_token_accuracy": 0.9097347557544708,
      "num_tokens": 21464640.0,
      "step": 1740
    },
    {
      "entropy": 0.9613929092884064,
      "epoch": 0.9167983149025803,
      "grad_norm": 0.28639957308769226,
      "learning_rate": 0.0001952386785422199,
      "loss": 0.22456036508083344,
      "mean_token_accuracy": 0.9144959151744843,
      "num_tokens": 21476976.0,
      "step": 1741
    },
    {
      "entropy": 0.9676374346017838,
      "epoch": 0.9173249078462349,
      "grad_norm": 0.29118600487709045,
      "learning_rate": 0.00019523163814704253,
      "loss": 0.21779820322990417,
      "mean_token_accuracy": 0.9170442819595337,
      "num_tokens": 21489312.0,
      "step": 1742
    },
    {
      "entropy": 0.9580910950899124,
      "epoch": 0.9178515007898894,
      "grad_norm": 0.28892767429351807,
      "learning_rate": 0.0001952245926921596,
      "loss": 0.1965646892786026,
      "mean_token_accuracy": 0.9236303865909576,
      "num_tokens": 21501648.0,
      "step": 1743
    },
    {
      "entropy": 1.0211364477872849,
      "epoch": 0.918378093733544,
      "grad_norm": 0.2978745996952057,
      "learning_rate": 0.00019521754217798935,
      "loss": 0.2231602966785431,
      "mean_token_accuracy": 0.9130188077688217,
      "num_tokens": 21513984.0,
      "step": 1744
    },
    {
      "entropy": 0.9940465688705444,
      "epoch": 0.9189046866771985,
      "grad_norm": 0.2930096685886383,
      "learning_rate": 0.00019521048660495026,
      "loss": 0.19159580767154694,
      "mean_token_accuracy": 0.9195211082696915,
      "num_tokens": 21526320.0,
      "step": 1745
    },
    {
      "entropy": 0.9247337281703949,
      "epoch": 0.919431279620853,
      "grad_norm": 0.2832118570804596,
      "learning_rate": 0.00019520342597346125,
      "loss": 0.21378204226493835,
      "mean_token_accuracy": 0.9141468554735184,
      "num_tokens": 21538656.0,
      "step": 1746
    },
    {
      "entropy": 0.9416433870792389,
      "epoch": 0.9199578725645077,
      "grad_norm": 0.3025250732898712,
      "learning_rate": 0.00019519636028394148,
      "loss": 0.21570643782615662,
      "mean_token_accuracy": 0.9150727838277817,
      "num_tokens": 21550992.0,
      "step": 1747
    },
    {
      "entropy": 0.9537800550460815,
      "epoch": 0.9204844655081622,
      "grad_norm": 0.29714229702949524,
      "learning_rate": 0.00019518928953681032,
      "loss": 0.22764158248901367,
      "mean_token_accuracy": 0.9034122675657272,
      "num_tokens": 21563328.0,
      "step": 1748
    },
    {
      "entropy": 0.9719788879156113,
      "epoch": 0.9210110584518167,
      "grad_norm": 0.2960829734802246,
      "learning_rate": 0.00019518221373248756,
      "loss": 0.2084151953458786,
      "mean_token_accuracy": 0.9151151925325394,
      "num_tokens": 21575664.0,
      "step": 1749
    },
    {
      "entropy": 0.9682374745607376,
      "epoch": 0.9215376513954713,
      "grad_norm": 0.2902142405509949,
      "learning_rate": 0.00019517513287139326,
      "loss": 0.19477756321430206,
      "mean_token_accuracy": 0.9197756201028824,
      "num_tokens": 21588000.0,
      "step": 1750
    },
    {
      "entropy": 0.9519971162080765,
      "epoch": 0.9220642443391258,
      "grad_norm": 0.28657373785972595,
      "learning_rate": 0.00019516804695394777,
      "loss": 0.23340506851673126,
      "mean_token_accuracy": 0.9101651012897491,
      "num_tokens": 21600336.0,
      "step": 1751
    },
    {
      "entropy": 0.9670501947402954,
      "epoch": 0.9225908372827805,
      "grad_norm": 0.2798555791378021,
      "learning_rate": 0.00019516095598057174,
      "loss": 0.2065662145614624,
      "mean_token_accuracy": 0.9186369478702545,
      "num_tokens": 21612672.0,
      "step": 1752
    },
    {
      "entropy": 0.9457984417676926,
      "epoch": 0.923117430226435,
      "grad_norm": 0.2589358985424042,
      "learning_rate": 0.00019515385995168608,
      "loss": 0.1857452392578125,
      "mean_token_accuracy": 0.9278716593980789,
      "num_tokens": 21625008.0,
      "step": 1753
    },
    {
      "entropy": 0.923014685511589,
      "epoch": 0.9236440231700895,
      "grad_norm": 0.29148226976394653,
      "learning_rate": 0.00019514675886771207,
      "loss": 0.21651901304721832,
      "mean_token_accuracy": 0.9152826964855194,
      "num_tokens": 21637344.0,
      "step": 1754
    },
    {
      "entropy": 0.95480976998806,
      "epoch": 0.9241706161137441,
      "grad_norm": 0.28343141078948975,
      "learning_rate": 0.00019513965272907128,
      "loss": 0.2175363004207611,
      "mean_token_accuracy": 0.9101854413747787,
      "num_tokens": 21649680.0,
      "step": 1755
    },
    {
      "entropy": 0.9613447189331055,
      "epoch": 0.9246972090573986,
      "grad_norm": 0.3009392023086548,
      "learning_rate": 0.0001951325415361855,
      "loss": 0.22403854131698608,
      "mean_token_accuracy": 0.9168571084737778,
      "num_tokens": 21662016.0,
      "step": 1756
    },
    {
      "entropy": 0.9929066747426987,
      "epoch": 0.9252238020010531,
      "grad_norm": 0.3036598563194275,
      "learning_rate": 0.00019512542528947696,
      "loss": 0.22976119816303253,
      "mean_token_accuracy": 0.9101551920175552,
      "num_tokens": 21674352.0,
      "step": 1757
    },
    {
      "entropy": 0.968064159154892,
      "epoch": 0.9257503949447078,
      "grad_norm": 0.28782951831817627,
      "learning_rate": 0.00019511830398936809,
      "loss": 0.21798846125602722,
      "mean_token_accuracy": 0.9074330925941467,
      "num_tokens": 21686688.0,
      "step": 1758
    },
    {
      "entropy": 0.9647489339113235,
      "epoch": 0.9262769878883623,
      "grad_norm": 0.2734207212924957,
      "learning_rate": 0.00019511117763628156,
      "loss": 0.209343820810318,
      "mean_token_accuracy": 0.9190388172864914,
      "num_tokens": 21699024.0,
      "step": 1759
    },
    {
      "entropy": 0.9990984350442886,
      "epoch": 0.9268035808320169,
      "grad_norm": 0.26921236515045166,
      "learning_rate": 0.00019510404623064053,
      "loss": 0.21170470118522644,
      "mean_token_accuracy": 0.9162604361772537,
      "num_tokens": 21711360.0,
      "step": 1760
    },
    {
      "entropy": 0.9623449593782425,
      "epoch": 0.9273301737756714,
      "grad_norm": 0.2714099884033203,
      "learning_rate": 0.00019509690977286825,
      "loss": 0.21800348162651062,
      "mean_token_accuracy": 0.9147970378398895,
      "num_tokens": 21723696.0,
      "step": 1761
    },
    {
      "entropy": 1.0402387380599976,
      "epoch": 0.9278567667193259,
      "grad_norm": 0.2736240327358246,
      "learning_rate": 0.00019508976826338844,
      "loss": 0.22173887491226196,
      "mean_token_accuracy": 0.9075192958116531,
      "num_tokens": 21736032.0,
      "step": 1762
    },
    {
      "entropy": 1.0271132439374924,
      "epoch": 0.9283833596629805,
      "grad_norm": 0.2877287268638611,
      "learning_rate": 0.00019508262170262502,
      "loss": 0.21932274103164673,
      "mean_token_accuracy": 0.915712833404541,
      "num_tokens": 21748368.0,
      "step": 1763
    },
    {
      "entropy": 1.034477338194847,
      "epoch": 0.9289099526066351,
      "grad_norm": 0.2779938280582428,
      "learning_rate": 0.0001950754700910023,
      "loss": 0.207020103931427,
      "mean_token_accuracy": 0.9147273898124695,
      "num_tokens": 21760704.0,
      "step": 1764
    },
    {
      "entropy": 0.9738915860652924,
      "epoch": 0.9294365455502897,
      "grad_norm": 0.25146281719207764,
      "learning_rate": 0.00019506831342894473,
      "loss": 0.1815354973077774,
      "mean_token_accuracy": 0.9243048876523972,
      "num_tokens": 21773040.0,
      "step": 1765
    },
    {
      "entropy": 0.9640511274337769,
      "epoch": 0.9299631384939442,
      "grad_norm": 0.2794640362262726,
      "learning_rate": 0.0001950611517168772,
      "loss": 0.222852423787117,
      "mean_token_accuracy": 0.9066295772790909,
      "num_tokens": 21785376.0,
      "step": 1766
    },
    {
      "entropy": 0.9703959077596664,
      "epoch": 0.9304897314375987,
      "grad_norm": 0.2606351971626282,
      "learning_rate": 0.00019505398495522487,
      "loss": 0.1913529932498932,
      "mean_token_accuracy": 0.9248643219470978,
      "num_tokens": 21797712.0,
      "step": 1767
    },
    {
      "entropy": 0.9996339529752731,
      "epoch": 0.9310163243812533,
      "grad_norm": 0.310641884803772,
      "learning_rate": 0.00019504681314441323,
      "loss": 0.22644834220409393,
      "mean_token_accuracy": 0.9118537902832031,
      "num_tokens": 21810048.0,
      "step": 1768
    },
    {
      "entropy": 0.9964617937803268,
      "epoch": 0.9315429173249078,
      "grad_norm": 0.2875900864601135,
      "learning_rate": 0.00019503963628486793,
      "loss": 0.2038813978433609,
      "mean_token_accuracy": 0.9186841398477554,
      "num_tokens": 21822384.0,
      "step": 1769
    },
    {
      "entropy": 0.9495928287506104,
      "epoch": 0.9320695102685624,
      "grad_norm": 0.2897108495235443,
      "learning_rate": 0.00019503245437701508,
      "loss": 0.20874524116516113,
      "mean_token_accuracy": 0.9164410084486008,
      "num_tokens": 21834720.0,
      "step": 1770
    },
    {
      "entropy": 1.0277469605207443,
      "epoch": 0.932596103212217,
      "grad_norm": 0.31645241379737854,
      "learning_rate": 0.00019502526742128104,
      "loss": 0.2254372090101242,
      "mean_token_accuracy": 0.9089788496494293,
      "num_tokens": 21847056.0,
      "step": 1771
    },
    {
      "entropy": 1.041675180196762,
      "epoch": 0.9331226961558715,
      "grad_norm": 0.3042741119861603,
      "learning_rate": 0.00019501807541809243,
      "loss": 0.21151864528656006,
      "mean_token_accuracy": 0.9152505397796631,
      "num_tokens": 21859392.0,
      "step": 1772
    },
    {
      "entropy": 1.0295332968235016,
      "epoch": 0.933649289099526,
      "grad_norm": 0.32914313673973083,
      "learning_rate": 0.00019501087836787623,
      "loss": 0.2500017285346985,
      "mean_token_accuracy": 0.8985483199357986,
      "num_tokens": 21871728.0,
      "step": 1773
    },
    {
      "entropy": 1.0453828126192093,
      "epoch": 0.9341758820431806,
      "grad_norm": 0.2983553111553192,
      "learning_rate": 0.00019500367627105965,
      "loss": 0.21561893820762634,
      "mean_token_accuracy": 0.9130748361349106,
      "num_tokens": 21884064.0,
      "step": 1774
    },
    {
      "entropy": 1.002024069428444,
      "epoch": 0.9347024749868352,
      "grad_norm": 0.28427883982658386,
      "learning_rate": 0.00019499646912807023,
      "loss": 0.22381159663200378,
      "mean_token_accuracy": 0.9057275652885437,
      "num_tokens": 21896400.0,
      "step": 1775
    },
    {
      "entropy": 1.0608263909816742,
      "epoch": 0.9352290679304898,
      "grad_norm": 0.27167659997940063,
      "learning_rate": 0.00019498925693933584,
      "loss": 0.1875414401292801,
      "mean_token_accuracy": 0.9234858453273773,
      "num_tokens": 21908736.0,
      "step": 1776
    },
    {
      "entropy": 1.0062582045793533,
      "epoch": 0.9357556608741443,
      "grad_norm": 0.2636836767196655,
      "learning_rate": 0.00019498203970528462,
      "loss": 0.21502971649169922,
      "mean_token_accuracy": 0.919221356511116,
      "num_tokens": 21921072.0,
      "step": 1777
    },
    {
      "entropy": 1.006988525390625,
      "epoch": 0.9362822538177988,
      "grad_norm": 0.26056134700775146,
      "learning_rate": 0.00019497481742634503,
      "loss": 0.201970174908638,
      "mean_token_accuracy": 0.9196654409170151,
      "num_tokens": 21933408.0,
      "step": 1778
    },
    {
      "entropy": 1.0629092454910278,
      "epoch": 0.9368088467614534,
      "grad_norm": 0.29083991050720215,
      "learning_rate": 0.00019496759010294577,
      "loss": 0.22228607535362244,
      "mean_token_accuracy": 0.9100936055183411,
      "num_tokens": 21945744.0,
      "step": 1779
    },
    {
      "entropy": 1.0019027441740036,
      "epoch": 0.9373354397051079,
      "grad_norm": 0.2699200212955475,
      "learning_rate": 0.00019496035773551592,
      "loss": 0.22549045085906982,
      "mean_token_accuracy": 0.9094872325658798,
      "num_tokens": 21958080.0,
      "step": 1780
    },
    {
      "entropy": 1.0057822614908218,
      "epoch": 0.9378620326487626,
      "grad_norm": 0.28440120816230774,
      "learning_rate": 0.00019495312032448485,
      "loss": 0.22345775365829468,
      "mean_token_accuracy": 0.9118607193231583,
      "num_tokens": 21970416.0,
      "step": 1781
    },
    {
      "entropy": 1.0094308406114578,
      "epoch": 0.9383886255924171,
      "grad_norm": 0.2834523320198059,
      "learning_rate": 0.00019494587787028216,
      "loss": 0.22185727953910828,
      "mean_token_accuracy": 0.9113683104515076,
      "num_tokens": 21982752.0,
      "step": 1782
    },
    {
      "entropy": 0.9964211732149124,
      "epoch": 0.9389152185360716,
      "grad_norm": 0.2954987585544586,
      "learning_rate": 0.00019493863037333778,
      "loss": 0.20775803923606873,
      "mean_token_accuracy": 0.9154636561870575,
      "num_tokens": 21995088.0,
      "step": 1783
    },
    {
      "entropy": 1.0346784442663193,
      "epoch": 0.9394418114797262,
      "grad_norm": 0.2964785695075989,
      "learning_rate": 0.000194931377834082,
      "loss": 0.2272047996520996,
      "mean_token_accuracy": 0.9083570241928101,
      "num_tokens": 22007424.0,
      "step": 1784
    },
    {
      "entropy": 1.0281512439250946,
      "epoch": 0.9399684044233807,
      "grad_norm": 0.26846325397491455,
      "learning_rate": 0.00019492412025294534,
      "loss": 0.20067930221557617,
      "mean_token_accuracy": 0.9195989966392517,
      "num_tokens": 22019760.0,
      "step": 1785
    },
    {
      "entropy": 0.9764789193868637,
      "epoch": 0.9404949973670352,
      "grad_norm": 0.26383689045906067,
      "learning_rate": 0.0001949168576303586,
      "loss": 0.2051767259836197,
      "mean_token_accuracy": 0.92010198533535,
      "num_tokens": 22032096.0,
      "step": 1786
    },
    {
      "entropy": 1.078786700963974,
      "epoch": 0.9410215903106899,
      "grad_norm": 0.3011668920516968,
      "learning_rate": 0.000194909589966753,
      "loss": 0.221777081489563,
      "mean_token_accuracy": 0.9121461063623428,
      "num_tokens": 22044432.0,
      "step": 1787
    },
    {
      "entropy": 0.9719114750623703,
      "epoch": 0.9415481832543444,
      "grad_norm": 0.27153825759887695,
      "learning_rate": 0.0001949023172625599,
      "loss": 0.21883930265903473,
      "mean_token_accuracy": 0.913441613316536,
      "num_tokens": 22056768.0,
      "step": 1788
    },
    {
      "entropy": 0.9923989027738571,
      "epoch": 0.942074776197999,
      "grad_norm": 0.3212338984012604,
      "learning_rate": 0.00019489503951821107,
      "loss": 0.22175338864326477,
      "mean_token_accuracy": 0.9168316125869751,
      "num_tokens": 22069104.0,
      "step": 1789
    },
    {
      "entropy": 0.9783156663179398,
      "epoch": 0.9426013691416535,
      "grad_norm": 0.273110568523407,
      "learning_rate": 0.00019488775673413857,
      "loss": 0.21147552132606506,
      "mean_token_accuracy": 0.9130014330148697,
      "num_tokens": 22081440.0,
      "step": 1790
    },
    {
      "entropy": 0.9893038272857666,
      "epoch": 0.943127962085308,
      "grad_norm": 0.2790793478488922,
      "learning_rate": 0.00019488046891077474,
      "loss": 0.22615540027618408,
      "mean_token_accuracy": 0.9109261482954025,
      "num_tokens": 22093776.0,
      "step": 1791
    },
    {
      "entropy": 1.0380553901195526,
      "epoch": 0.9436545550289626,
      "grad_norm": 0.27198460698127747,
      "learning_rate": 0.00019487317604855212,
      "loss": 0.205999955534935,
      "mean_token_accuracy": 0.9175222963094711,
      "num_tokens": 22106112.0,
      "step": 1792
    },
    {
      "entropy": 1.0116278380155563,
      "epoch": 0.9441811479726172,
      "grad_norm": 0.2635519504547119,
      "learning_rate": 0.00019486587814790377,
      "loss": 0.21772751212120056,
      "mean_token_accuracy": 0.9136025160551071,
      "num_tokens": 22118448.0,
      "step": 1793
    },
    {
      "entropy": 1.0237055569887161,
      "epoch": 0.9447077409162717,
      "grad_norm": 0.2741040289402008,
      "learning_rate": 0.00019485857520926288,
      "loss": 0.20760180056095123,
      "mean_token_accuracy": 0.9132292419672012,
      "num_tokens": 22130784.0,
      "step": 1794
    },
    {
      "entropy": 1.0196455419063568,
      "epoch": 0.9452343338599263,
      "grad_norm": 0.27313318848609924,
      "learning_rate": 0.00019485126723306292,
      "loss": 0.18969658017158508,
      "mean_token_accuracy": 0.9258754402399063,
      "num_tokens": 22143120.0,
      "step": 1795
    },
    {
      "entropy": 1.0403377711772919,
      "epoch": 0.9457609268035808,
      "grad_norm": 0.2604946494102478,
      "learning_rate": 0.00019484395421973783,
      "loss": 0.2132149338722229,
      "mean_token_accuracy": 0.9197222590446472,
      "num_tokens": 22155456.0,
      "step": 1796
    },
    {
      "entropy": 0.9670400768518448,
      "epoch": 0.9462875197472354,
      "grad_norm": 0.24395477771759033,
      "learning_rate": 0.00019483663616972165,
      "loss": 0.19094271957874298,
      "mean_token_accuracy": 0.9200141131877899,
      "num_tokens": 22167792.0,
      "step": 1797
    },
    {
      "entropy": 1.0197023451328278,
      "epoch": 0.94681411269089,
      "grad_norm": 0.2558318078517914,
      "learning_rate": 0.00019482931308344888,
      "loss": 0.19818465411663055,
      "mean_token_accuracy": 0.9231192022562027,
      "num_tokens": 22180128.0,
      "step": 1798
    },
    {
      "entropy": 1.0280962586402893,
      "epoch": 0.9473407056345445,
      "grad_norm": 0.28250062465667725,
      "learning_rate": 0.00019482198496135417,
      "loss": 0.22149549424648285,
      "mean_token_accuracy": 0.9105774611234665,
      "num_tokens": 22192464.0,
      "step": 1799
    },
    {
      "entropy": 1.004937931895256,
      "epoch": 0.9478672985781991,
      "grad_norm": 0.2890641391277313,
      "learning_rate": 0.00019481465180387264,
      "loss": 0.2117823362350464,
      "mean_token_accuracy": 0.9159420728683472,
      "num_tokens": 22204800.0,
      "step": 1800
    },
    {
      "entropy": 0.9912999272346497,
      "epoch": 0.9483938915218536,
      "grad_norm": 0.2721498906612396,
      "learning_rate": 0.00019480731361143955,
      "loss": 0.19752754271030426,
      "mean_token_accuracy": 0.9188252538442612,
      "num_tokens": 22217136.0,
      "step": 1801
    },
    {
      "entropy": 0.9911512136459351,
      "epoch": 0.9489204844655081,
      "grad_norm": 0.2967390716075897,
      "learning_rate": 0.00019479997038449054,
      "loss": 0.2183246612548828,
      "mean_token_accuracy": 0.908792570233345,
      "num_tokens": 22229472.0,
      "step": 1802
    },
    {
      "entropy": 1.0003108382225037,
      "epoch": 0.9494470774091627,
      "grad_norm": 0.281955748796463,
      "learning_rate": 0.00019479262212346153,
      "loss": 0.20015354454517365,
      "mean_token_accuracy": 0.9192506521940231,
      "num_tokens": 22241808.0,
      "step": 1803
    },
    {
      "entropy": 1.005830705165863,
      "epoch": 0.9499736703528173,
      "grad_norm": 0.29906165599823,
      "learning_rate": 0.00019478526882878876,
      "loss": 0.2205047309398651,
      "mean_token_accuracy": 0.9138208776712418,
      "num_tokens": 22254144.0,
      "step": 1804
    },
    {
      "entropy": 0.9975222796201706,
      "epoch": 0.9505002632964719,
      "grad_norm": 0.297961950302124,
      "learning_rate": 0.00019477791050090875,
      "loss": 0.21634206175804138,
      "mean_token_accuracy": 0.9120276570320129,
      "num_tokens": 22266480.0,
      "step": 1805
    },
    {
      "entropy": 0.9912213832139969,
      "epoch": 0.9510268562401264,
      "grad_norm": 0.3019638955593109,
      "learning_rate": 0.00019477054714025832,
      "loss": 0.21199895441532135,
      "mean_token_accuracy": 0.9165389537811279,
      "num_tokens": 22278816.0,
      "step": 1806
    },
    {
      "entropy": 0.9327168315649033,
      "epoch": 0.9515534491837809,
      "grad_norm": 0.288906991481781,
      "learning_rate": 0.00019476317874727456,
      "loss": 0.23043318092823029,
      "mean_token_accuracy": 0.9077083468437195,
      "num_tokens": 22291152.0,
      "step": 1807
    },
    {
      "entropy": 0.9759026318788528,
      "epoch": 0.9520800421274355,
      "grad_norm": 0.2820807993412018,
      "learning_rate": 0.00019475580532239497,
      "loss": 0.21610046923160553,
      "mean_token_accuracy": 0.914526030421257,
      "num_tokens": 22303488.0,
      "step": 1808
    },
    {
      "entropy": 1.01944038271904,
      "epoch": 0.95260663507109,
      "grad_norm": 0.29718121886253357,
      "learning_rate": 0.00019474842686605716,
      "loss": 0.2213006466627121,
      "mean_token_accuracy": 0.9078305512666702,
      "num_tokens": 22315824.0,
      "step": 1809
    },
    {
      "entropy": 0.9788577258586884,
      "epoch": 0.9531332280147446,
      "grad_norm": 0.2710515558719635,
      "learning_rate": 0.00019474104337869924,
      "loss": 0.19855360686779022,
      "mean_token_accuracy": 0.9186171144247055,
      "num_tokens": 22328160.0,
      "step": 1810
    },
    {
      "entropy": 0.9749794155359268,
      "epoch": 0.9536598209583992,
      "grad_norm": 0.2699875235557556,
      "learning_rate": 0.00019473365486075944,
      "loss": 0.20002411305904388,
      "mean_token_accuracy": 0.917415052652359,
      "num_tokens": 22340496.0,
      "step": 1811
    },
    {
      "entropy": 1.0002324134111404,
      "epoch": 0.9541864139020537,
      "grad_norm": 0.2768368422985077,
      "learning_rate": 0.00019472626131267646,
      "loss": 0.2179412692785263,
      "mean_token_accuracy": 0.9109683334827423,
      "num_tokens": 22352832.0,
      "step": 1812
    },
    {
      "entropy": 1.0184978395700455,
      "epoch": 0.9547130068457083,
      "grad_norm": 0.2990346848964691,
      "learning_rate": 0.00019471886273488915,
      "loss": 0.22559918463230133,
      "mean_token_accuracy": 0.911268338561058,
      "num_tokens": 22365168.0,
      "step": 1813
    },
    {
      "entropy": 0.9672580063343048,
      "epoch": 0.9552395997893628,
      "grad_norm": 0.258120059967041,
      "learning_rate": 0.00019471145912783671,
      "loss": 0.19848020374774933,
      "mean_token_accuracy": 0.9207747876644135,
      "num_tokens": 22377504.0,
      "step": 1814
    },
    {
      "entropy": 1.0332013070583344,
      "epoch": 0.9557661927330173,
      "grad_norm": 0.27658164501190186,
      "learning_rate": 0.00019470405049195876,
      "loss": 0.22540630400180817,
      "mean_token_accuracy": 0.909706637263298,
      "num_tokens": 22389840.0,
      "step": 1815
    },
    {
      "entropy": 1.0286743193864822,
      "epoch": 0.956292785676672,
      "grad_norm": 0.30504515767097473,
      "learning_rate": 0.00019469663682769491,
      "loss": 0.22958287596702576,
      "mean_token_accuracy": 0.90970279276371,
      "num_tokens": 22402176.0,
      "step": 1816
    },
    {
      "entropy": 1.0396344661712646,
      "epoch": 0.9568193786203265,
      "grad_norm": 0.3122229278087616,
      "learning_rate": 0.00019468921813548546,
      "loss": 0.21201902627944946,
      "mean_token_accuracy": 0.912709191441536,
      "num_tokens": 22414512.0,
      "step": 1817
    },
    {
      "entropy": 0.9904589354991913,
      "epoch": 0.957345971563981,
      "grad_norm": 0.2821761667728424,
      "learning_rate": 0.00019468179441577073,
      "loss": 0.21789240837097168,
      "mean_token_accuracy": 0.9137752056121826,
      "num_tokens": 22426848.0,
      "step": 1818
    },
    {
      "entropy": 0.9968847781419754,
      "epoch": 0.9578725645076356,
      "grad_norm": 0.2896675765514374,
      "learning_rate": 0.0001946743656689914,
      "loss": 0.19366274774074554,
      "mean_token_accuracy": 0.9199696481227875,
      "num_tokens": 22439184.0,
      "step": 1819
    },
    {
      "entropy": 0.9972859472036362,
      "epoch": 0.9583991574512901,
      "grad_norm": 0.2833546996116638,
      "learning_rate": 0.00019466693189558852,
      "loss": 0.20901280641555786,
      "mean_token_accuracy": 0.9163765460252762,
      "num_tokens": 22451520.0,
      "step": 1820
    },
    {
      "entropy": 1.0221098363399506,
      "epoch": 0.9589257503949447,
      "grad_norm": 0.30042803287506104,
      "learning_rate": 0.00019465949309600334,
      "loss": 0.20722414553165436,
      "mean_token_accuracy": 0.9166164398193359,
      "num_tokens": 22463856.0,
      "step": 1821
    },
    {
      "entropy": 0.9847268313169479,
      "epoch": 0.9594523433385993,
      "grad_norm": 0.29877185821533203,
      "learning_rate": 0.00019465204927067754,
      "loss": 0.22702287137508392,
      "mean_token_accuracy": 0.910624086856842,
      "num_tokens": 22476192.0,
      "step": 1822
    },
    {
      "entropy": 1.0212393552064896,
      "epoch": 0.9599789362822538,
      "grad_norm": 0.3099231421947479,
      "learning_rate": 0.00019464460042005293,
      "loss": 0.23366589844226837,
      "mean_token_accuracy": 0.9104152172803879,
      "num_tokens": 22488528.0,
      "step": 1823
    },
    {
      "entropy": 1.0240531265735626,
      "epoch": 0.9605055292259084,
      "grad_norm": 0.2815970182418823,
      "learning_rate": 0.00019463714654457172,
      "loss": 0.19396552443504333,
      "mean_token_accuracy": 0.9191655963659286,
      "num_tokens": 22500864.0,
      "step": 1824
    },
    {
      "entropy": 0.9908776879310608,
      "epoch": 0.9610321221695629,
      "grad_norm": 0.25329092144966125,
      "learning_rate": 0.00019462968764467648,
      "loss": 0.19574449956417084,
      "mean_token_accuracy": 0.9221374094486237,
      "num_tokens": 22513200.0,
      "step": 1825
    },
    {
      "entropy": 1.0275763273239136,
      "epoch": 0.9615587151132174,
      "grad_norm": 0.29698899388313293,
      "learning_rate": 0.00019462222372080992,
      "loss": 0.23368221521377563,
      "mean_token_accuracy": 0.9036354571580887,
      "num_tokens": 22525536.0,
      "step": 1826
    },
    {
      "entropy": 1.0337569862604141,
      "epoch": 0.9620853080568721,
      "grad_norm": 0.3343221843242645,
      "learning_rate": 0.00019461475477341512,
      "loss": 0.21144059300422668,
      "mean_token_accuracy": 0.9137226939201355,
      "num_tokens": 22537872.0,
      "step": 1827
    },
    {
      "entropy": 1.0447833985090256,
      "epoch": 0.9626119010005266,
      "grad_norm": 0.3130665421485901,
      "learning_rate": 0.0001946072808029355,
      "loss": 0.2065916359424591,
      "mean_token_accuracy": 0.9174522012472153,
      "num_tokens": 22550208.0,
      "step": 1828
    },
    {
      "entropy": 1.0395597666502,
      "epoch": 0.9631384939441812,
      "grad_norm": 0.26665323972702026,
      "learning_rate": 0.00019459980180981478,
      "loss": 0.21482327580451965,
      "mean_token_accuracy": 0.9130977690219879,
      "num_tokens": 22562544.0,
      "step": 1829
    },
    {
      "entropy": 1.0779733955860138,
      "epoch": 0.9636650868878357,
      "grad_norm": 0.28825950622558594,
      "learning_rate": 0.0001945923177944969,
      "loss": 0.21779173612594604,
      "mean_token_accuracy": 0.9156653732061386,
      "num_tokens": 22574880.0,
      "step": 1830
    },
    {
      "entropy": 1.0036635249853134,
      "epoch": 0.9641916798314902,
      "grad_norm": 0.2941460907459259,
      "learning_rate": 0.0001945848287574262,
      "loss": 0.2330627739429474,
      "mean_token_accuracy": 0.9059428721666336,
      "num_tokens": 22587216.0,
      "step": 1831
    },
    {
      "entropy": 0.9797258973121643,
      "epoch": 0.9647182727751448,
      "grad_norm": 0.25932562351226807,
      "learning_rate": 0.00019457733469904715,
      "loss": 0.2051321119070053,
      "mean_token_accuracy": 0.9192904084920883,
      "num_tokens": 22599552.0,
      "step": 1832
    },
    {
      "entropy": 1.039179965853691,
      "epoch": 0.9652448657187994,
      "grad_norm": 0.26872527599334717,
      "learning_rate": 0.0001945698356198047,
      "loss": 0.21343429386615753,
      "mean_token_accuracy": 0.9149598628282547,
      "num_tokens": 22611888.0,
      "step": 1833
    },
    {
      "entropy": 1.0027782917022705,
      "epoch": 0.965771458662454,
      "grad_norm": 0.27663007378578186,
      "learning_rate": 0.00019456233152014406,
      "loss": 0.19677159190177917,
      "mean_token_accuracy": 0.9209509938955307,
      "num_tokens": 22624224.0,
      "step": 1834
    },
    {
      "entropy": 1.0312834680080414,
      "epoch": 0.9662980516061085,
      "grad_norm": 0.28490519523620605,
      "learning_rate": 0.00019455482240051064,
      "loss": 0.21659842133522034,
      "mean_token_accuracy": 0.9108777791261673,
      "num_tokens": 22636560.0,
      "step": 1835
    },
    {
      "entropy": 1.012413963675499,
      "epoch": 0.966824644549763,
      "grad_norm": 0.27517637610435486,
      "learning_rate": 0.00019454730826135022,
      "loss": 0.22000275552272797,
      "mean_token_accuracy": 0.9186523258686066,
      "num_tokens": 22648896.0,
      "step": 1836
    },
    {
      "entropy": 0.9954932332038879,
      "epoch": 0.9673512374934176,
      "grad_norm": 0.27069538831710815,
      "learning_rate": 0.0001945397891031089,
      "loss": 0.19973544776439667,
      "mean_token_accuracy": 0.9211115539073944,
      "num_tokens": 22661232.0,
      "step": 1837
    },
    {
      "entropy": 1.0200046300888062,
      "epoch": 0.9678778304370721,
      "grad_norm": 0.2899874150753021,
      "learning_rate": 0.00019453226492623305,
      "loss": 0.22812676429748535,
      "mean_token_accuracy": 0.9122350364923477,
      "num_tokens": 22673568.0,
      "step": 1838
    },
    {
      "entropy": 0.9972085207700729,
      "epoch": 0.9684044233807267,
      "grad_norm": 0.29899537563323975,
      "learning_rate": 0.0001945247357311693,
      "loss": 0.2259579598903656,
      "mean_token_accuracy": 0.9079723209142685,
      "num_tokens": 22685904.0,
      "step": 1839
    },
    {
      "entropy": 1.0059181898832321,
      "epoch": 0.9689310163243813,
      "grad_norm": 0.29635098576545715,
      "learning_rate": 0.00019451720151836467,
      "loss": 0.23771284520626068,
      "mean_token_accuracy": 0.9140820056200027,
      "num_tokens": 22698240.0,
      "step": 1840
    },
    {
      "entropy": 1.0503579378128052,
      "epoch": 0.9694576092680358,
      "grad_norm": 0.29462575912475586,
      "learning_rate": 0.00019450966228826635,
      "loss": 0.19687220454216003,
      "mean_token_accuracy": 0.9218025654554367,
      "num_tokens": 22710576.0,
      "step": 1841
    },
    {
      "entropy": 0.9612717479467392,
      "epoch": 0.9699842022116903,
      "grad_norm": 0.27696606516838074,
      "learning_rate": 0.00019450211804132196,
      "loss": 0.19762445986270905,
      "mean_token_accuracy": 0.9190530776977539,
      "num_tokens": 22722912.0,
      "step": 1842
    },
    {
      "entropy": 0.9952682256698608,
      "epoch": 0.9705107951553449,
      "grad_norm": 0.2652910053730011,
      "learning_rate": 0.00019449456877797933,
      "loss": 0.21299602091312408,
      "mean_token_accuracy": 0.9156957715749741,
      "num_tokens": 22735248.0,
      "step": 1843
    },
    {
      "entropy": 0.9986753612756729,
      "epoch": 0.9710373880989994,
      "grad_norm": 0.2620997726917267,
      "learning_rate": 0.00019448701449868662,
      "loss": 0.1993142068386078,
      "mean_token_accuracy": 0.9197477102279663,
      "num_tokens": 22747584.0,
      "step": 1844
    },
    {
      "entropy": 0.9899931252002716,
      "epoch": 0.9715639810426541,
      "grad_norm": 0.29060348868370056,
      "learning_rate": 0.0001944794552038923,
      "loss": 0.2212766706943512,
      "mean_token_accuracy": 0.9078968465328217,
      "num_tokens": 22759920.0,
      "step": 1845
    },
    {
      "entropy": 1.0022216588258743,
      "epoch": 0.9720905739863086,
      "grad_norm": 0.30724501609802246,
      "learning_rate": 0.00019447189089404513,
      "loss": 0.2398059368133545,
      "mean_token_accuracy": 0.9064589291810989,
      "num_tokens": 22772256.0,
      "step": 1846
    },
    {
      "entropy": 0.997131809592247,
      "epoch": 0.9726171669299631,
      "grad_norm": 0.286475270986557,
      "learning_rate": 0.0001944643215695941,
      "loss": 0.21739161014556885,
      "mean_token_accuracy": 0.9086156636476517,
      "num_tokens": 22784592.0,
      "step": 1847
    },
    {
      "entropy": 0.9620841443538666,
      "epoch": 0.9731437598736177,
      "grad_norm": 0.2802620828151703,
      "learning_rate": 0.00019445674723098864,
      "loss": 0.18609827756881714,
      "mean_token_accuracy": 0.9273043125867844,
      "num_tokens": 22796928.0,
      "step": 1848
    },
    {
      "entropy": 0.9614392817020416,
      "epoch": 0.9736703528172722,
      "grad_norm": 0.2547474205493927,
      "learning_rate": 0.0001944491678786783,
      "loss": 0.20111055672168732,
      "mean_token_accuracy": 0.9191482216119766,
      "num_tokens": 22809264.0,
      "step": 1849
    },
    {
      "entropy": 0.9734077900648117,
      "epoch": 0.9741969457609269,
      "grad_norm": 0.2975533902645111,
      "learning_rate": 0.00019444158351311308,
      "loss": 0.20715618133544922,
      "mean_token_accuracy": 0.9150333106517792,
      "num_tokens": 22821600.0,
      "step": 1850
    },
    {
      "entropy": 0.9761214107275009,
      "epoch": 0.9747235387045814,
      "grad_norm": 0.28768840432167053,
      "learning_rate": 0.0001944339941347432,
      "loss": 0.18795141577720642,
      "mean_token_accuracy": 0.9215870052576065,
      "num_tokens": 22833936.0,
      "step": 1851
    },
    {
      "entropy": 0.9937604367733002,
      "epoch": 0.9752501316482359,
      "grad_norm": 0.29926928877830505,
      "learning_rate": 0.00019442639974401923,
      "loss": 0.21389541029930115,
      "mean_token_accuracy": 0.917677566409111,
      "num_tokens": 22846272.0,
      "step": 1852
    },
    {
      "entropy": 0.9756381958723068,
      "epoch": 0.9757767245918905,
      "grad_norm": 0.298100084066391,
      "learning_rate": 0.00019441880034139198,
      "loss": 0.21645361185073853,
      "mean_token_accuracy": 0.91536945104599,
      "num_tokens": 22858608.0,
      "step": 1853
    },
    {
      "entropy": 0.9328114837408066,
      "epoch": 0.976303317535545,
      "grad_norm": 0.27972036600112915,
      "learning_rate": 0.0001944111959273126,
      "loss": 0.20870962738990784,
      "mean_token_accuracy": 0.9177060127258301,
      "num_tokens": 22870944.0,
      "step": 1854
    },
    {
      "entropy": 0.9690214693546295,
      "epoch": 0.9768299104791995,
      "grad_norm": 0.28668034076690674,
      "learning_rate": 0.00019440358650223244,
      "loss": 0.20322152972221375,
      "mean_token_accuracy": 0.9198893010616302,
      "num_tokens": 22883280.0,
      "step": 1855
    },
    {
      "entropy": 0.9374023228883743,
      "epoch": 0.9773565034228542,
      "grad_norm": 0.2798213064670563,
      "learning_rate": 0.00019439597206660336,
      "loss": 0.20974165201187134,
      "mean_token_accuracy": 0.9153610616922379,
      "num_tokens": 22895616.0,
      "step": 1856
    },
    {
      "entropy": 0.988965705037117,
      "epoch": 0.9778830963665087,
      "grad_norm": 0.2775600254535675,
      "learning_rate": 0.00019438835262087728,
      "loss": 0.19997943937778473,
      "mean_token_accuracy": 0.9216296821832657,
      "num_tokens": 22907952.0,
      "step": 1857
    },
    {
      "entropy": 1.0289342552423477,
      "epoch": 0.9784096893101633,
      "grad_norm": 0.3432585299015045,
      "learning_rate": 0.00019438072816550654,
      "loss": 0.22643734514713287,
      "mean_token_accuracy": 0.9054546356201172,
      "num_tokens": 22920288.0,
      "step": 1858
    },
    {
      "entropy": 0.9867872446775436,
      "epoch": 0.9789362822538178,
      "grad_norm": 0.29411035776138306,
      "learning_rate": 0.0001943730987009438,
      "loss": 0.22154730558395386,
      "mean_token_accuracy": 0.9093334376811981,
      "num_tokens": 22932624.0,
      "step": 1859
    },
    {
      "entropy": 1.0116321295499802,
      "epoch": 0.9794628751974723,
      "grad_norm": 0.25875324010849,
      "learning_rate": 0.00019436546422764198,
      "loss": 0.19839854538440704,
      "mean_token_accuracy": 0.9156231582164764,
      "num_tokens": 22944960.0,
      "step": 1860
    },
    {
      "entropy": 1.0039302557706833,
      "epoch": 0.9799894681411269,
      "grad_norm": 0.29284337162971497,
      "learning_rate": 0.00019435782474605423,
      "loss": 0.2374696135520935,
      "mean_token_accuracy": 0.905433177947998,
      "num_tokens": 22957296.0,
      "step": 1861
    },
    {
      "entropy": 1.0000396072864532,
      "epoch": 0.9805160610847815,
      "grad_norm": 0.24808640778064728,
      "learning_rate": 0.00019435018025663412,
      "loss": 0.18588939309120178,
      "mean_token_accuracy": 0.9234057366847992,
      "num_tokens": 22969632.0,
      "step": 1862
    },
    {
      "entropy": 1.0562945902347565,
      "epoch": 0.981042654028436,
      "grad_norm": 0.2809024453163147,
      "learning_rate": 0.00019434253075983543,
      "loss": 0.19566890597343445,
      "mean_token_accuracy": 0.9169919043779373,
      "num_tokens": 22981968.0,
      "step": 1863
    },
    {
      "entropy": 1.066685050725937,
      "epoch": 0.9815692469720906,
      "grad_norm": 0.2744297385215759,
      "learning_rate": 0.0001943348762561123,
      "loss": 0.21286125481128693,
      "mean_token_accuracy": 0.911152720451355,
      "num_tokens": 22994304.0,
      "step": 1864
    },
    {
      "entropy": 1.0672977268695831,
      "epoch": 0.9820958399157451,
      "grad_norm": 0.30773451924324036,
      "learning_rate": 0.00019432721674591906,
      "loss": 0.22466342151165009,
      "mean_token_accuracy": 0.9027170091867447,
      "num_tokens": 23006640.0,
      "step": 1865
    },
    {
      "entropy": 0.9817476272583008,
      "epoch": 0.9826224328593997,
      "grad_norm": 0.25929689407348633,
      "learning_rate": 0.00019431955222971048,
      "loss": 0.20428617298603058,
      "mean_token_accuracy": 0.9160020500421524,
      "num_tokens": 23018976.0,
      "step": 1866
    },
    {
      "entropy": 1.0060690939426422,
      "epoch": 0.9831490258030542,
      "grad_norm": 0.3303799629211426,
      "learning_rate": 0.00019431188270794153,
      "loss": 0.22543492913246155,
      "mean_token_accuracy": 0.9112392216920853,
      "num_tokens": 23031312.0,
      "step": 1867
    },
    {
      "entropy": 0.9948325008153915,
      "epoch": 0.9836756187467088,
      "grad_norm": 0.2678799331188202,
      "learning_rate": 0.00019430420818106753,
      "loss": 0.19793486595153809,
      "mean_token_accuracy": 0.9236938953399658,
      "num_tokens": 23043648.0,
      "step": 1868
    },
    {
      "entropy": 1.0155895799398422,
      "epoch": 0.9842022116903634,
      "grad_norm": 0.32058584690093994,
      "learning_rate": 0.00019429652864954403,
      "loss": 0.22886891663074493,
      "mean_token_accuracy": 0.9127451777458191,
      "num_tokens": 23055984.0,
      "step": 1869
    },
    {
      "entropy": 1.0170871764421463,
      "epoch": 0.9847288046340179,
      "grad_norm": 0.2957039177417755,
      "learning_rate": 0.00019428884411382694,
      "loss": 0.21584120392799377,
      "mean_token_accuracy": 0.9092356264591217,
      "num_tokens": 23068320.0,
      "step": 1870
    },
    {
      "entropy": 1.0231945216655731,
      "epoch": 0.9852553975776724,
      "grad_norm": 0.2651207149028778,
      "learning_rate": 0.00019428115457437249,
      "loss": 0.18167884647846222,
      "mean_token_accuracy": 0.9309497475624084,
      "num_tokens": 23080656.0,
      "step": 1871
    },
    {
      "entropy": 0.9718558192253113,
      "epoch": 0.985781990521327,
      "grad_norm": 0.29894301295280457,
      "learning_rate": 0.0001942734600316371,
      "loss": 0.20839130878448486,
      "mean_token_accuracy": 0.9109772592782974,
      "num_tokens": 23092992.0,
      "step": 1872
    },
    {
      "entropy": 1.0036133825778961,
      "epoch": 0.9863085834649815,
      "grad_norm": 0.28074634075164795,
      "learning_rate": 0.00019426576048607754,
      "loss": 0.21850639581680298,
      "mean_token_accuracy": 0.9151745438575745,
      "num_tokens": 23105328.0,
      "step": 1873
    },
    {
      "entropy": 1.0208822190761566,
      "epoch": 0.9868351764086362,
      "grad_norm": 0.27776047587394714,
      "learning_rate": 0.00019425805593815094,
      "loss": 0.1931566298007965,
      "mean_token_accuracy": 0.9207221418619156,
      "num_tokens": 23117664.0,
      "step": 1874
    },
    {
      "entropy": 0.9643638432025909,
      "epoch": 0.9873617693522907,
      "grad_norm": 0.28227487206459045,
      "learning_rate": 0.00019425034638831466,
      "loss": 0.2174389511346817,
      "mean_token_accuracy": 0.9123106747865677,
      "num_tokens": 23130000.0,
      "step": 1875
    },
    {
      "entropy": 1.0151402354240417,
      "epoch": 0.9878883622959452,
      "grad_norm": 0.3007744252681732,
      "learning_rate": 0.00019424263183702634,
      "loss": 0.23054783046245575,
      "mean_token_accuracy": 0.9074221700429916,
      "num_tokens": 23142336.0,
      "step": 1876
    },
    {
      "entropy": 1.023296281695366,
      "epoch": 0.9884149552395998,
      "grad_norm": 0.2875635623931885,
      "learning_rate": 0.000194234912284744,
      "loss": 0.2183448076248169,
      "mean_token_accuracy": 0.9146157503128052,
      "num_tokens": 23154672.0,
      "step": 1877
    },
    {
      "entropy": 1.025291085243225,
      "epoch": 0.9889415481832543,
      "grad_norm": 0.2833808660507202,
      "learning_rate": 0.00019422718773192584,
      "loss": 0.2020779550075531,
      "mean_token_accuracy": 0.9207105487585068,
      "num_tokens": 23167008.0,
      "step": 1878
    },
    {
      "entropy": 1.0463321805000305,
      "epoch": 0.989468141126909,
      "grad_norm": 0.27166518568992615,
      "learning_rate": 0.0001942194581790305,
      "loss": 0.21112844347953796,
      "mean_token_accuracy": 0.9151988327503204,
      "num_tokens": 23179344.0,
      "step": 1879
    },
    {
      "entropy": 1.0174481570720673,
      "epoch": 0.9899947340705635,
      "grad_norm": 0.28479474782943726,
      "learning_rate": 0.00019421172362651677,
      "loss": 0.2219792902469635,
      "mean_token_accuracy": 0.9093788266181946,
      "num_tokens": 23191680.0,
      "step": 1880
    },
    {
      "entropy": 1.068659245967865,
      "epoch": 0.990521327014218,
      "grad_norm": 0.27825888991355896,
      "learning_rate": 0.00019420398407484383,
      "loss": 0.2115180492401123,
      "mean_token_accuracy": 0.913215160369873,
      "num_tokens": 23204016.0,
      "step": 1881
    },
    {
      "entropy": 1.0102509111166,
      "epoch": 0.9910479199578726,
      "grad_norm": 0.2567676901817322,
      "learning_rate": 0.00019419623952447113,
      "loss": 0.1962418258190155,
      "mean_token_accuracy": 0.9181914627552032,
      "num_tokens": 23216352.0,
      "step": 1882
    },
    {
      "entropy": 1.0041260570287704,
      "epoch": 0.9915745129015271,
      "grad_norm": 0.2607569992542267,
      "learning_rate": 0.00019418848997585842,
      "loss": 0.18007442355155945,
      "mean_token_accuracy": 0.9283247590065002,
      "num_tokens": 23228688.0,
      "step": 1883
    },
    {
      "entropy": 1.0119262784719467,
      "epoch": 0.9921011058451816,
      "grad_norm": 0.2691285312175751,
      "learning_rate": 0.00019418073542946577,
      "loss": 0.2185424566268921,
      "mean_token_accuracy": 0.9108522534370422,
      "num_tokens": 23241024.0,
      "step": 1884
    },
    {
      "entropy": 1.0200906693935394,
      "epoch": 0.9926276987888363,
      "grad_norm": 0.2762972414493561,
      "learning_rate": 0.0001941729758857535,
      "loss": 0.21047905087471008,
      "mean_token_accuracy": 0.921312689781189,
      "num_tokens": 23253360.0,
      "step": 1885
    },
    {
      "entropy": 1.0015190541744232,
      "epoch": 0.9931542917324908,
      "grad_norm": 0.2789957523345947,
      "learning_rate": 0.0001941652113451822,
      "loss": 0.22564658522605896,
      "mean_token_accuracy": 0.9052274376153946,
      "num_tokens": 23265696.0,
      "step": 1886
    },
    {
      "entropy": 1.0212842226028442,
      "epoch": 0.9936808846761453,
      "grad_norm": 0.2787906527519226,
      "learning_rate": 0.00019415744180821293,
      "loss": 0.19954261183738708,
      "mean_token_accuracy": 0.9225759655237198,
      "num_tokens": 23278032.0,
      "step": 1887
    },
    {
      "entropy": 1.0256837606430054,
      "epoch": 0.9942074776197999,
      "grad_norm": 0.3047766089439392,
      "learning_rate": 0.0001941496672753068,
      "loss": 0.22101764380931854,
      "mean_token_accuracy": 0.9100571721792221,
      "num_tokens": 23290368.0,
      "step": 1888
    },
    {
      "entropy": 1.0017095357179642,
      "epoch": 0.9947340705634544,
      "grad_norm": 0.3079693913459778,
      "learning_rate": 0.00019414188774692542,
      "loss": 0.23877866566181183,
      "mean_token_accuracy": 0.9044640958309174,
      "num_tokens": 23302704.0,
      "step": 1889
    },
    {
      "entropy": 0.9979955106973648,
      "epoch": 0.995260663507109,
      "grad_norm": 0.28942957520484924,
      "learning_rate": 0.00019413410322353055,
      "loss": 0.21703916788101196,
      "mean_token_accuracy": 0.911283403635025,
      "num_tokens": 23315040.0,
      "step": 1890
    },
    {
      "entropy": 1.000881314277649,
      "epoch": 0.9957872564507636,
      "grad_norm": 0.2723085582256317,
      "learning_rate": 0.00019412631370558433,
      "loss": 0.2083275318145752,
      "mean_token_accuracy": 0.9137948602437973,
      "num_tokens": 23327376.0,
      "step": 1891
    },
    {
      "entropy": 0.9760307371616364,
      "epoch": 0.9963138493944181,
      "grad_norm": 0.28710952401161194,
      "learning_rate": 0.00019411851919354925,
      "loss": 0.2206597775220871,
      "mean_token_accuracy": 0.913771316409111,
      "num_tokens": 23339712.0,
      "step": 1892
    },
    {
      "entropy": 1.0313685238361359,
      "epoch": 0.9968404423380727,
      "grad_norm": 0.2784387767314911,
      "learning_rate": 0.00019411071968788792,
      "loss": 0.20836740732192993,
      "mean_token_accuracy": 0.9158921986818314,
      "num_tokens": 23352048.0,
      "step": 1893
    },
    {
      "entropy": 0.9972709268331528,
      "epoch": 0.9973670352817272,
      "grad_norm": 0.2760670781135559,
      "learning_rate": 0.00019410291518906337,
      "loss": 0.215082049369812,
      "mean_token_accuracy": 0.9125083088874817,
      "num_tokens": 23364384.0,
      "step": 1894
    },
    {
      "entropy": 0.9749780297279358,
      "epoch": 0.9978936282253817,
      "grad_norm": 0.2756492793560028,
      "learning_rate": 0.00019409510569753897,
      "loss": 0.19914628565311432,
      "mean_token_accuracy": 0.9158760160207748,
      "num_tokens": 23376720.0,
      "step": 1895
    },
    {
      "entropy": 0.9882865399122238,
      "epoch": 0.9984202211690363,
      "grad_norm": 0.3229084610939026,
      "learning_rate": 0.00019408729121377828,
      "loss": 0.21740438044071198,
      "mean_token_accuracy": 0.9149737656116486,
      "num_tokens": 23389056.0,
      "step": 1896
    },
    {
      "entropy": 0.979081466794014,
      "epoch": 0.9989468141126909,
      "grad_norm": 0.2839930057525635,
      "learning_rate": 0.00019407947173824522,
      "loss": 0.20839889347553253,
      "mean_token_accuracy": 0.9204780906438828,
      "num_tokens": 23401392.0,
      "step": 1897
    },
    {
      "entropy": 0.9407005459070206,
      "epoch": 0.9994734070563455,
      "grad_norm": 0.28035539388656616,
      "learning_rate": 0.00019407164727140396,
      "loss": 0.2062290906906128,
      "mean_token_accuracy": 0.9152461290359497,
      "num_tokens": 23413728.0,
      "step": 1898
    },
    {
      "entropy": 0.9624819904565811,
      "epoch": 1.0,
      "grad_norm": 0.29015156626701355,
      "learning_rate": 0.00019406381781371902,
      "loss": 0.21500550210475922,
      "mean_token_accuracy": 0.915779635310173,
      "num_tokens": 23425036.0,
      "step": 1899
    },
    {
      "epoch": 1.0,
      "eval_entropy": 0.925739836008565,
      "eval_loss": 0.21148517727851868,
      "eval_mean_token_accuracy": 0.9153817107703826,
      "eval_num_tokens": 23425036.0,
      "eval_runtime": 664.4914,
      "eval_samples_per_second": 8.573,
      "eval_steps_per_second": 2.858,
      "step": 1899
    },
    {
      "entropy": 0.9483114928007126,
      "epoch": 1.0005265929436546,
      "grad_norm": 0.28408128023147583,
      "learning_rate": 0.00019405598336565518,
      "loss": 0.2106991708278656,
      "mean_token_accuracy": 0.9160860329866409,
      "num_tokens": 23437372.0,
      "step": 1900
    },
    {
      "entropy": 0.898838147521019,
      "epoch": 1.001053185887309,
      "grad_norm": 0.27440792322158813,
      "learning_rate": 0.00019404814392767753,
      "loss": 0.21290592849254608,
      "mean_token_accuracy": 0.9125117361545563,
      "num_tokens": 23449708.0,
      "step": 1901
    },
    {
      "entropy": 0.9791853725910187,
      "epoch": 1.0015797788309637,
      "grad_norm": 0.28151366114616394,
      "learning_rate": 0.00019404029950025143,
      "loss": 0.19366838037967682,
      "mean_token_accuracy": 0.9241389185190201,
      "num_tokens": 23462044.0,
      "step": 1902
    },
    {
      "entropy": 0.9103992432355881,
      "epoch": 1.0021063717746181,
      "grad_norm": 0.29249274730682373,
      "learning_rate": 0.00019403245008384255,
      "loss": 0.21274180710315704,
      "mean_token_accuracy": 0.9132009893655777,
      "num_tokens": 23474380.0,
      "step": 1903
    },
    {
      "entropy": 0.9207734614610672,
      "epoch": 1.0026329647182728,
      "grad_norm": 0.29525914788246155,
      "learning_rate": 0.00019402459567891694,
      "loss": 0.20322710275650024,
      "mean_token_accuracy": 0.9193697869777679,
      "num_tokens": 23486716.0,
      "step": 1904
    },
    {
      "entropy": 0.9738444536924362,
      "epoch": 1.0031595576619274,
      "grad_norm": 0.28035998344421387,
      "learning_rate": 0.00019401673628594078,
      "loss": 0.1997321993112564,
      "mean_token_accuracy": 0.915515661239624,
      "num_tokens": 23499052.0,
      "step": 1905
    },
    {
      "entropy": 0.9290015399456024,
      "epoch": 1.0036861506055819,
      "grad_norm": 0.2815547287464142,
      "learning_rate": 0.00019400887190538068,
      "loss": 0.20885860919952393,
      "mean_token_accuracy": 0.9181715399026871,
      "num_tokens": 23511388.0,
      "step": 1906
    },
    {
      "entropy": 0.9481154382228851,
      "epoch": 1.0042127435492365,
      "grad_norm": 0.2717452645301819,
      "learning_rate": 0.0001940010025377035,
      "loss": 0.20289872586727142,
      "mean_token_accuracy": 0.9163729101419449,
      "num_tokens": 23523724.0,
      "step": 1907
    },
    {
      "entropy": 0.9266571402549744,
      "epoch": 1.004739336492891,
      "grad_norm": 0.3059215843677521,
      "learning_rate": 0.00019399312818337637,
      "loss": 0.20683327317237854,
      "mean_token_accuracy": 0.9123108834028244,
      "num_tokens": 23536060.0,
      "step": 1908
    },
    {
      "entropy": 0.9232073873281479,
      "epoch": 1.0052659294365456,
      "grad_norm": 0.2654300034046173,
      "learning_rate": 0.00019398524884286678,
      "loss": 0.18700408935546875,
      "mean_token_accuracy": 0.9244769513607025,
      "num_tokens": 23548396.0,
      "step": 1909
    },
    {
      "entropy": 0.9462733566761017,
      "epoch": 1.0057925223802,
      "grad_norm": 0.28558823466300964,
      "learning_rate": 0.0001939773645166425,
      "loss": 0.20201043784618378,
      "mean_token_accuracy": 0.9207958728075027,
      "num_tokens": 23560732.0,
      "step": 1910
    },
    {
      "entropy": 0.9496246725320816,
      "epoch": 1.0063191153238547,
      "grad_norm": 0.27963462471961975,
      "learning_rate": 0.00019396947520517154,
      "loss": 0.20976471900939941,
      "mean_token_accuracy": 0.9165468662977219,
      "num_tokens": 23573068.0,
      "step": 1911
    },
    {
      "entropy": 0.9894253462553024,
      "epoch": 1.0068457082675093,
      "grad_norm": 0.2891998291015625,
      "learning_rate": 0.00019396158090892224,
      "loss": 0.2190844714641571,
      "mean_token_accuracy": 0.9077310711145401,
      "num_tokens": 23585404.0,
      "step": 1912
    },
    {
      "entropy": 1.0001903474330902,
      "epoch": 1.0073723012111637,
      "grad_norm": 0.2947920560836792,
      "learning_rate": 0.00019395368162836323,
      "loss": 0.21794044971466064,
      "mean_token_accuracy": 0.9169338941574097,
      "num_tokens": 23597740.0,
      "step": 1913
    },
    {
      "entropy": 0.9664654582738876,
      "epoch": 1.0078988941548184,
      "grad_norm": 0.270565390586853,
      "learning_rate": 0.00019394577736396346,
      "loss": 0.19095301628112793,
      "mean_token_accuracy": 0.9217241108417511,
      "num_tokens": 23610076.0,
      "step": 1914
    },
    {
      "entropy": 0.9208107441663742,
      "epoch": 1.0084254870984728,
      "grad_norm": 0.27241528034210205,
      "learning_rate": 0.0001939378681161922,
      "loss": 0.18645267188549042,
      "mean_token_accuracy": 0.9272374659776688,
      "num_tokens": 23622412.0,
      "step": 1915
    },
    {
      "entropy": 1.0111596435308456,
      "epoch": 1.0089520800421274,
      "grad_norm": 0.3164159655570984,
      "learning_rate": 0.00019392995388551888,
      "loss": 0.22077365219593048,
      "mean_token_accuracy": 0.9118006080389023,
      "num_tokens": 23634748.0,
      "step": 1916
    },
    {
      "entropy": 0.944818839430809,
      "epoch": 1.009478672985782,
      "grad_norm": 0.28582748770713806,
      "learning_rate": 0.00019392203467241342,
      "loss": 0.2104344218969345,
      "mean_token_accuracy": 0.9131231158971786,
      "num_tokens": 23647084.0,
      "step": 1917
    },
    {
      "entropy": 1.0247685611248016,
      "epoch": 1.0100052659294365,
      "grad_norm": 0.2851823568344116,
      "learning_rate": 0.00019391411047734589,
      "loss": 0.21748843789100647,
      "mean_token_accuracy": 0.9151061922311783,
      "num_tokens": 23659420.0,
      "step": 1918
    },
    {
      "entropy": 0.9971954673528671,
      "epoch": 1.0105318588730912,
      "grad_norm": 0.2717190682888031,
      "learning_rate": 0.00019390618130078673,
      "loss": 0.2043081521987915,
      "mean_token_accuracy": 0.9173936247825623,
      "num_tokens": 23671756.0,
      "step": 1919
    },
    {
      "entropy": 0.9666483253240585,
      "epoch": 1.0110584518167456,
      "grad_norm": 0.2750032842159271,
      "learning_rate": 0.00019389824714320658,
      "loss": 0.19933579862117767,
      "mean_token_accuracy": 0.9193840771913528,
      "num_tokens": 23684092.0,
      "step": 1920
    },
    {
      "entropy": 0.997409462928772,
      "epoch": 1.0115850447604002,
      "grad_norm": 0.2951802611351013,
      "learning_rate": 0.0001938903080050765,
      "loss": 0.2241615653038025,
      "mean_token_accuracy": 0.9114827215671539,
      "num_tokens": 23696428.0,
      "step": 1921
    },
    {
      "entropy": 1.0097313821315765,
      "epoch": 1.0121116377040547,
      "grad_norm": 0.29612311720848083,
      "learning_rate": 0.00019388236388686783,
      "loss": 0.2130715399980545,
      "mean_token_accuracy": 0.9157524257898331,
      "num_tokens": 23708764.0,
      "step": 1922
    },
    {
      "entropy": 1.029169887304306,
      "epoch": 1.0126382306477093,
      "grad_norm": 0.28184011578559875,
      "learning_rate": 0.00019387441478905206,
      "loss": 0.20634552836418152,
      "mean_token_accuracy": 0.9152942299842834,
      "num_tokens": 23721100.0,
      "step": 1923
    },
    {
      "entropy": 1.0215208977460861,
      "epoch": 1.013164823591364,
      "grad_norm": 0.2626985013484955,
      "learning_rate": 0.00019386646071210118,
      "loss": 0.18794600665569305,
      "mean_token_accuracy": 0.9244592189788818,
      "num_tokens": 23733436.0,
      "step": 1924
    },
    {
      "entropy": 1.03949436545372,
      "epoch": 1.0136914165350184,
      "grad_norm": 0.2628861665725708,
      "learning_rate": 0.00019385850165648733,
      "loss": 0.2007792592048645,
      "mean_token_accuracy": 0.9187650829553604,
      "num_tokens": 23745772.0,
      "step": 1925
    },
    {
      "entropy": 0.9779924899339676,
      "epoch": 1.014218009478673,
      "grad_norm": 0.28041279315948486,
      "learning_rate": 0.000193850537622683,
      "loss": 0.1995813548564911,
      "mean_token_accuracy": 0.9189839065074921,
      "num_tokens": 23758108.0,
      "step": 1926
    },
    {
      "entropy": 1.0496584177017212,
      "epoch": 1.0147446024223274,
      "grad_norm": 0.30866992473602295,
      "learning_rate": 0.00019384256861116096,
      "loss": 0.2425222247838974,
      "mean_token_accuracy": 0.9028880447149277,
      "num_tokens": 23770444.0,
      "step": 1927
    },
    {
      "entropy": 1.0065599381923676,
      "epoch": 1.015271195365982,
      "grad_norm": 0.26788726449012756,
      "learning_rate": 0.0001938345946223943,
      "loss": 0.1882321834564209,
      "mean_token_accuracy": 0.9198320806026459,
      "num_tokens": 23782780.0,
      "step": 1928
    },
    {
      "entropy": 1.0039747208356857,
      "epoch": 1.0157977883096367,
      "grad_norm": 0.2758983373641968,
      "learning_rate": 0.00019382661565685638,
      "loss": 0.20896543562412262,
      "mean_token_accuracy": 0.9133927524089813,
      "num_tokens": 23795116.0,
      "step": 1929
    },
    {
      "entropy": 0.9877621978521347,
      "epoch": 1.0163243812532912,
      "grad_norm": 0.27864596247673035,
      "learning_rate": 0.00019381863171502088,
      "loss": 0.18458585441112518,
      "mean_token_accuracy": 0.9242655336856842,
      "num_tokens": 23807452.0,
      "step": 1930
    },
    {
      "entropy": 1.00676129758358,
      "epoch": 1.0168509741969458,
      "grad_norm": 0.27991485595703125,
      "learning_rate": 0.00019381064279736174,
      "loss": 0.21812650561332703,
      "mean_token_accuracy": 0.915228009223938,
      "num_tokens": 23819788.0,
      "step": 1931
    },
    {
      "entropy": 0.9534656554460526,
      "epoch": 1.0173775671406002,
      "grad_norm": 0.2649945914745331,
      "learning_rate": 0.00019380264890435324,
      "loss": 0.21227021515369415,
      "mean_token_accuracy": 0.9195111691951752,
      "num_tokens": 23832124.0,
      "step": 1932
    },
    {
      "entropy": 1.000964567065239,
      "epoch": 1.0179041600842549,
      "grad_norm": 0.268867164850235,
      "learning_rate": 0.0001937946500364699,
      "loss": 0.20481450855731964,
      "mean_token_accuracy": 0.9178451746702194,
      "num_tokens": 23844460.0,
      "step": 1933
    },
    {
      "entropy": 1.000000774860382,
      "epoch": 1.0184307530279095,
      "grad_norm": 0.27489760518074036,
      "learning_rate": 0.00019378664619418653,
      "loss": 0.20561310648918152,
      "mean_token_accuracy": 0.915080651640892,
      "num_tokens": 23856796.0,
      "step": 1934
    },
    {
      "entropy": 0.9716245830059052,
      "epoch": 1.018957345971564,
      "grad_norm": 0.306060791015625,
      "learning_rate": 0.00019377863737797839,
      "loss": 0.21995992958545685,
      "mean_token_accuracy": 0.910162016749382,
      "num_tokens": 23869132.0,
      "step": 1935
    },
    {
      "entropy": 0.9378920793533325,
      "epoch": 1.0194839389152186,
      "grad_norm": 0.26929163932800293,
      "learning_rate": 0.00019377062358832083,
      "loss": 0.18971963226795197,
      "mean_token_accuracy": 0.921707347035408,
      "num_tokens": 23881468.0,
      "step": 1936
    },
    {
      "entropy": 0.9584116488695145,
      "epoch": 1.020010531858873,
      "grad_norm": 0.27198344469070435,
      "learning_rate": 0.0001937626048256896,
      "loss": 0.19669067859649658,
      "mean_token_accuracy": 0.9219775944948196,
      "num_tokens": 23893804.0,
      "step": 1937
    },
    {
      "entropy": 0.9765763580799103,
      "epoch": 1.0205371248025277,
      "grad_norm": 0.2800286114215851,
      "learning_rate": 0.0001937545810905607,
      "loss": 0.20201562345027924,
      "mean_token_accuracy": 0.9161263704299927,
      "num_tokens": 23906140.0,
      "step": 1938
    },
    {
      "entropy": 0.9443774819374084,
      "epoch": 1.021063717746182,
      "grad_norm": 0.28303396701812744,
      "learning_rate": 0.0001937465523834105,
      "loss": 0.19613805413246155,
      "mean_token_accuracy": 0.9168961495161057,
      "num_tokens": 23918476.0,
      "step": 1939
    },
    {
      "entropy": 0.9929839223623276,
      "epoch": 1.0215903106898367,
      "grad_norm": 0.317105770111084,
      "learning_rate": 0.00019373851870471562,
      "loss": 0.20963701605796814,
      "mean_token_accuracy": 0.9184058755636215,
      "num_tokens": 23930812.0,
      "step": 1940
    },
    {
      "entropy": 0.9796905666589737,
      "epoch": 1.0221169036334914,
      "grad_norm": 0.29658782482147217,
      "learning_rate": 0.0001937304800549529,
      "loss": 0.2231239676475525,
      "mean_token_accuracy": 0.9108584076166153,
      "num_tokens": 23943148.0,
      "step": 1941
    },
    {
      "entropy": 0.9595655649900436,
      "epoch": 1.0226434965771458,
      "grad_norm": 0.25836655497550964,
      "learning_rate": 0.00019372243643459963,
      "loss": 0.18731528520584106,
      "mean_token_accuracy": 0.9224317222833633,
      "num_tokens": 23955484.0,
      "step": 1942
    },
    {
      "entropy": 0.9837910383939743,
      "epoch": 1.0231700895208005,
      "grad_norm": 0.2838844060897827,
      "learning_rate": 0.0001937143878441333,
      "loss": 0.21538090705871582,
      "mean_token_accuracy": 0.9147688299417496,
      "num_tokens": 23967820.0,
      "step": 1943
    },
    {
      "entropy": 1.002283364534378,
      "epoch": 1.0236966824644549,
      "grad_norm": 0.2761163115501404,
      "learning_rate": 0.00019370633428403164,
      "loss": 0.17965167760849,
      "mean_token_accuracy": 0.9276806563138962,
      "num_tokens": 23980156.0,
      "step": 1944
    },
    {
      "entropy": 0.9750987738370895,
      "epoch": 1.0242232754081095,
      "grad_norm": 0.2677707374095917,
      "learning_rate": 0.00019369827575477282,
      "loss": 0.1916501522064209,
      "mean_token_accuracy": 0.9206420630216599,
      "num_tokens": 23992492.0,
      "step": 1945
    },
    {
      "entropy": 1.0182987600564957,
      "epoch": 1.0247498683517642,
      "grad_norm": 0.2850121259689331,
      "learning_rate": 0.00019369021225683518,
      "loss": 0.2081344723701477,
      "mean_token_accuracy": 0.9149588048458099,
      "num_tokens": 24004828.0,
      "step": 1946
    },
    {
      "entropy": 0.9728025048971176,
      "epoch": 1.0252764612954186,
      "grad_norm": 0.2776131331920624,
      "learning_rate": 0.00019368214379069745,
      "loss": 0.20609930157661438,
      "mean_token_accuracy": 0.9178636223077774,
      "num_tokens": 24017164.0,
      "step": 1947
    },
    {
      "entropy": 0.9742907732725143,
      "epoch": 1.0258030542390733,
      "grad_norm": 0.29074424505233765,
      "learning_rate": 0.0001936740703568386,
      "loss": 0.21784421801567078,
      "mean_token_accuracy": 0.9106293469667435,
      "num_tokens": 24029500.0,
      "step": 1948
    },
    {
      "entropy": 0.9815188944339752,
      "epoch": 1.0263296471827277,
      "grad_norm": 0.2835097908973694,
      "learning_rate": 0.0001936659919557378,
      "loss": 0.20473387837409973,
      "mean_token_accuracy": 0.9137338846921921,
      "num_tokens": 24041836.0,
      "step": 1949
    },
    {
      "entropy": 0.9593153893947601,
      "epoch": 1.0268562401263823,
      "grad_norm": 0.27597561478614807,
      "learning_rate": 0.00019365790858787477,
      "loss": 0.21293160319328308,
      "mean_token_accuracy": 0.9137197583913803,
      "num_tokens": 24054172.0,
      "step": 1950
    },
    {
      "entropy": 1.006902426481247,
      "epoch": 1.0273828330700367,
      "grad_norm": 0.29685065150260925,
      "learning_rate": 0.00019364982025372925,
      "loss": 0.21124479174613953,
      "mean_token_accuracy": 0.9155112355947495,
      "num_tokens": 24066508.0,
      "step": 1951
    },
    {
      "entropy": 0.9826443940401077,
      "epoch": 1.0279094260136914,
      "grad_norm": 0.28184542059898376,
      "learning_rate": 0.0001936417269537815,
      "loss": 0.18985268473625183,
      "mean_token_accuracy": 0.9251315295696259,
      "num_tokens": 24078844.0,
      "step": 1952
    },
    {
      "entropy": 0.9592526108026505,
      "epoch": 1.028436018957346,
      "grad_norm": 0.2665354013442993,
      "learning_rate": 0.00019363362868851186,
      "loss": 0.17802020907402039,
      "mean_token_accuracy": 0.923170417547226,
      "num_tokens": 24091180.0,
      "step": 1953
    },
    {
      "entropy": 0.9736118167638779,
      "epoch": 1.0289626119010005,
      "grad_norm": 0.2990165650844574,
      "learning_rate": 0.00019362552545840121,
      "loss": 0.21102318167686462,
      "mean_token_accuracy": 0.918361246585846,
      "num_tokens": 24103516.0,
      "step": 1954
    },
    {
      "entropy": 0.9543041735887527,
      "epoch": 1.0294892048446551,
      "grad_norm": 0.2993450462818146,
      "learning_rate": 0.00019361741726393048,
      "loss": 0.20676571130752563,
      "mean_token_accuracy": 0.9157939106225967,
      "num_tokens": 24115852.0,
      "step": 1955
    },
    {
      "entropy": 0.9188299477100372,
      "epoch": 1.0300157977883095,
      "grad_norm": 0.28595927357673645,
      "learning_rate": 0.00019360930410558107,
      "loss": 0.20469802618026733,
      "mean_token_accuracy": 0.9192691892385483,
      "num_tokens": 24128188.0,
      "step": 1956
    },
    {
      "entropy": 0.9500316083431244,
      "epoch": 1.0305423907319642,
      "grad_norm": 0.3056456744670868,
      "learning_rate": 0.00019360118598383454,
      "loss": 0.21120858192443848,
      "mean_token_accuracy": 0.9139054864645004,
      "num_tokens": 24140524.0,
      "step": 1957
    },
    {
      "entropy": 0.9614865332841873,
      "epoch": 1.0310689836756188,
      "grad_norm": 0.3015027642250061,
      "learning_rate": 0.00019359306289917292,
      "loss": 0.19627857208251953,
      "mean_token_accuracy": 0.9196411669254303,
      "num_tokens": 24152860.0,
      "step": 1958
    },
    {
      "entropy": 0.9348547607660294,
      "epoch": 1.0315955766192733,
      "grad_norm": 0.2926042377948761,
      "learning_rate": 0.00019358493485207833,
      "loss": 0.1929434835910797,
      "mean_token_accuracy": 0.9180857390165329,
      "num_tokens": 24165196.0,
      "step": 1959
    },
    {
      "entropy": 0.8882603049278259,
      "epoch": 1.032122169562928,
      "grad_norm": 0.2788626551628113,
      "learning_rate": 0.00019357680184303334,
      "loss": 0.2019919753074646,
      "mean_token_accuracy": 0.9170414507389069,
      "num_tokens": 24177532.0,
      "step": 1960
    },
    {
      "entropy": 0.9481528550386429,
      "epoch": 1.0326487625065823,
      "grad_norm": 0.30911776423454285,
      "learning_rate": 0.00019356866387252076,
      "loss": 0.20932908356189728,
      "mean_token_accuracy": 0.9150265455245972,
      "num_tokens": 24189868.0,
      "step": 1961
    },
    {
      "entropy": 0.9632448703050613,
      "epoch": 1.033175355450237,
      "grad_norm": 0.29422178864479065,
      "learning_rate": 0.00019356052094102367,
      "loss": 0.1939503401517868,
      "mean_token_accuracy": 0.9227122515439987,
      "num_tokens": 24202204.0,
      "step": 1962
    },
    {
      "entropy": 0.9328284412622452,
      "epoch": 1.0337019483938916,
      "grad_norm": 0.3035720884799957,
      "learning_rate": 0.0001935523730490255,
      "loss": 0.21952201426029205,
      "mean_token_accuracy": 0.9107953608036041,
      "num_tokens": 24214540.0,
      "step": 1963
    },
    {
      "entropy": 0.9676711559295654,
      "epoch": 1.034228541337546,
      "grad_norm": 0.2915896773338318,
      "learning_rate": 0.0001935442201970099,
      "loss": 0.19596201181411743,
      "mean_token_accuracy": 0.9193816781044006,
      "num_tokens": 24226876.0,
      "step": 1964
    },
    {
      "entropy": 0.9730495363473892,
      "epoch": 1.0347551342812007,
      "grad_norm": 0.28182631731033325,
      "learning_rate": 0.0001935360623854609,
      "loss": 0.20961566269397736,
      "mean_token_accuracy": 0.9185821712017059,
      "num_tokens": 24239212.0,
      "step": 1965
    },
    {
      "entropy": 1.023067906498909,
      "epoch": 1.0352817272248551,
      "grad_norm": 0.30114588141441345,
      "learning_rate": 0.00019352789961486273,
      "loss": 0.23333251476287842,
      "mean_token_accuracy": 0.9085756540298462,
      "num_tokens": 24251548.0,
      "step": 1966
    },
    {
      "entropy": 0.9701975286006927,
      "epoch": 1.0358083201685098,
      "grad_norm": 0.2716124355792999,
      "learning_rate": 0.00019351973188570004,
      "loss": 0.20822694897651672,
      "mean_token_accuracy": 0.9222011566162109,
      "num_tokens": 24263884.0,
      "step": 1967
    },
    {
      "entropy": 0.9808885604143143,
      "epoch": 1.0363349131121642,
      "grad_norm": 0.26052355766296387,
      "learning_rate": 0.0001935115591984576,
      "loss": 0.1830577850341797,
      "mean_token_accuracy": 0.9277938008308411,
      "num_tokens": 24276220.0,
      "step": 1968
    },
    {
      "entropy": 1.0236380398273468,
      "epoch": 1.0368615060558188,
      "grad_norm": 0.2982364296913147,
      "learning_rate": 0.00019350338155362066,
      "loss": 0.19804757833480835,
      "mean_token_accuracy": 0.9196522831916809,
      "num_tokens": 24288556.0,
      "step": 1969
    },
    {
      "entropy": 1.0457386374473572,
      "epoch": 1.0373880989994735,
      "grad_norm": 0.30198392271995544,
      "learning_rate": 0.00019349519895167468,
      "loss": 0.20456969738006592,
      "mean_token_accuracy": 0.9179401248693466,
      "num_tokens": 24300892.0,
      "step": 1970
    },
    {
      "entropy": 1.0371754169464111,
      "epoch": 1.037914691943128,
      "grad_norm": 0.2852479815483093,
      "learning_rate": 0.00019348701139310537,
      "loss": 0.22744318842887878,
      "mean_token_accuracy": 0.9061823189258575,
      "num_tokens": 24313228.0,
      "step": 1971
    },
    {
      "entropy": 1.0361986309289932,
      "epoch": 1.0384412848867826,
      "grad_norm": 0.29646071791648865,
      "learning_rate": 0.00019347881887839878,
      "loss": 0.22257165610790253,
      "mean_token_accuracy": 0.9108520895242691,
      "num_tokens": 24325564.0,
      "step": 1972
    },
    {
      "entropy": 1.043959766626358,
      "epoch": 1.038967877830437,
      "grad_norm": 0.3205704092979431,
      "learning_rate": 0.00019347062140804126,
      "loss": 0.21183185279369354,
      "mean_token_accuracy": 0.9146294444799423,
      "num_tokens": 24337900.0,
      "step": 1973
    },
    {
      "entropy": 1.0562986433506012,
      "epoch": 1.0394944707740916,
      "grad_norm": 0.31516483426094055,
      "learning_rate": 0.00019346241898251945,
      "loss": 0.222466379404068,
      "mean_token_accuracy": 0.912394255399704,
      "num_tokens": 24350236.0,
      "step": 1974
    },
    {
      "entropy": 0.9892242848873138,
      "epoch": 1.0400210637177463,
      "grad_norm": 0.27365726232528687,
      "learning_rate": 0.0001934542116023203,
      "loss": 0.18739140033721924,
      "mean_token_accuracy": 0.924969956278801,
      "num_tokens": 24362572.0,
      "step": 1975
    },
    {
      "entropy": 1.006972685456276,
      "epoch": 1.0405476566614007,
      "grad_norm": 0.2618653178215027,
      "learning_rate": 0.00019344599926793098,
      "loss": 0.2140425145626068,
      "mean_token_accuracy": 0.9133986085653305,
      "num_tokens": 24374908.0,
      "step": 1976
    },
    {
      "entropy": 1.0172683149576187,
      "epoch": 1.0410742496050553,
      "grad_norm": 0.29427164793014526,
      "learning_rate": 0.00019343778197983912,
      "loss": 0.22519932687282562,
      "mean_token_accuracy": 0.9123007953166962,
      "num_tokens": 24387244.0,
      "step": 1977
    },
    {
      "entropy": 1.01324824988842,
      "epoch": 1.0416008425487098,
      "grad_norm": 0.2959752678871155,
      "learning_rate": 0.00019342955973853236,
      "loss": 0.2103637009859085,
      "mean_token_accuracy": 0.9110966324806213,
      "num_tokens": 24399580.0,
      "step": 1978
    },
    {
      "entropy": 0.9949903935194016,
      "epoch": 1.0421274354923644,
      "grad_norm": 0.26972147822380066,
      "learning_rate": 0.00019342133254449898,
      "loss": 0.19294336438179016,
      "mean_token_accuracy": 0.9218278974294662,
      "num_tokens": 24411916.0,
      "step": 1979
    },
    {
      "entropy": 1.003416359424591,
      "epoch": 1.042654028436019,
      "grad_norm": 0.29061558842658997,
      "learning_rate": 0.00019341310039822725,
      "loss": 0.1943337321281433,
      "mean_token_accuracy": 0.9220097661018372,
      "num_tokens": 24424252.0,
      "step": 1980
    },
    {
      "entropy": 1.0185414999723434,
      "epoch": 1.0431806213796735,
      "grad_norm": 0.2810025215148926,
      "learning_rate": 0.00019340486330020595,
      "loss": 0.1928320974111557,
      "mean_token_accuracy": 0.9200875461101532,
      "num_tokens": 24436588.0,
      "step": 1981
    },
    {
      "entropy": 1.0032877922058105,
      "epoch": 1.0437072143233281,
      "grad_norm": 0.27209043502807617,
      "learning_rate": 0.00019339662125092403,
      "loss": 0.19464010000228882,
      "mean_token_accuracy": 0.9196325391530991,
      "num_tokens": 24448924.0,
      "step": 1982
    },
    {
      "entropy": 1.0001914203166962,
      "epoch": 1.0442338072669826,
      "grad_norm": 0.3038516342639923,
      "learning_rate": 0.00019338837425087078,
      "loss": 0.21659797430038452,
      "mean_token_accuracy": 0.9114441424608231,
      "num_tokens": 24461260.0,
      "step": 1983
    },
    {
      "entropy": 0.9755618721246719,
      "epoch": 1.0447604002106372,
      "grad_norm": 0.27092164754867554,
      "learning_rate": 0.00019338012230053574,
      "loss": 0.19371184706687927,
      "mean_token_accuracy": 0.9238433092832565,
      "num_tokens": 24473596.0,
      "step": 1984
    },
    {
      "entropy": 0.9904940128326416,
      "epoch": 1.0452869931542916,
      "grad_norm": 0.29529231786727905,
      "learning_rate": 0.00019337186540040885,
      "loss": 0.20442527532577515,
      "mean_token_accuracy": 0.9141305983066559,
      "num_tokens": 24485932.0,
      "step": 1985
    },
    {
      "entropy": 0.9658221155405045,
      "epoch": 1.0458135860979463,
      "grad_norm": 0.27508610486984253,
      "learning_rate": 0.00019336360355098024,
      "loss": 0.21024343371391296,
      "mean_token_accuracy": 0.9153925180435181,
      "num_tokens": 24498268.0,
      "step": 1986
    },
    {
      "entropy": 0.9924428313970566,
      "epoch": 1.046340179041601,
      "grad_norm": 0.28747668862342834,
      "learning_rate": 0.00019335533675274037,
      "loss": 0.18813030421733856,
      "mean_token_accuracy": 0.9244283735752106,
      "num_tokens": 24510604.0,
      "step": 1987
    },
    {
      "entropy": 1.0350338965654373,
      "epoch": 1.0468667719852554,
      "grad_norm": 0.28181594610214233,
      "learning_rate": 0.00019334706500617998,
      "loss": 0.18969137966632843,
      "mean_token_accuracy": 0.9228011220693588,
      "num_tokens": 24522940.0,
      "step": 1988
    },
    {
      "entropy": 1.017800748348236,
      "epoch": 1.04739336492891,
      "grad_norm": 0.2933719754219055,
      "learning_rate": 0.00019333878831179012,
      "loss": 0.2083795666694641,
      "mean_token_accuracy": 0.9186853021383286,
      "num_tokens": 24535276.0,
      "step": 1989
    },
    {
      "entropy": 1.0671946704387665,
      "epoch": 1.0479199578725644,
      "grad_norm": 0.28361761569976807,
      "learning_rate": 0.00019333050667006213,
      "loss": 0.20133714377880096,
      "mean_token_accuracy": 0.9176107496023178,
      "num_tokens": 24547612.0,
      "step": 1990
    },
    {
      "entropy": 1.0133258700370789,
      "epoch": 1.048446550816219,
      "grad_norm": 0.2936132550239563,
      "learning_rate": 0.00019332222008148765,
      "loss": 0.23258453607559204,
      "mean_token_accuracy": 0.9048744589090347,
      "num_tokens": 24559948.0,
      "step": 1991
    },
    {
      "entropy": 1.1020651161670685,
      "epoch": 1.0489731437598737,
      "grad_norm": 0.29475727677345276,
      "learning_rate": 0.0001933139285465586,
      "loss": 0.23158571124076843,
      "mean_token_accuracy": 0.90408755838871,
      "num_tokens": 24572284.0,
      "step": 1992
    },
    {
      "entropy": 1.0560840368270874,
      "epoch": 1.0494997367035281,
      "grad_norm": 0.2781185805797577,
      "learning_rate": 0.00019330563206576725,
      "loss": 0.21855907142162323,
      "mean_token_accuracy": 0.9118623584508896,
      "num_tokens": 24584620.0,
      "step": 1993
    },
    {
      "entropy": 1.1237812638282776,
      "epoch": 1.0500263296471828,
      "grad_norm": 0.3516911268234253,
      "learning_rate": 0.000193297330639606,
      "loss": 0.21132902801036835,
      "mean_token_accuracy": 0.9161721616983414,
      "num_tokens": 24596956.0,
      "step": 1994
    },
    {
      "entropy": 1.0761407017707825,
      "epoch": 1.0505529225908372,
      "grad_norm": 0.2794252038002014,
      "learning_rate": 0.00019328902426856776,
      "loss": 0.20890802145004272,
      "mean_token_accuracy": 0.920161560177803,
      "num_tokens": 24609292.0,
      "step": 1995
    },
    {
      "entropy": 1.049127846956253,
      "epoch": 1.0510795155344919,
      "grad_norm": 0.2529737651348114,
      "learning_rate": 0.00019328071295314557,
      "loss": 0.1920773684978485,
      "mean_token_accuracy": 0.9240875691175461,
      "num_tokens": 24621628.0,
      "step": 1996
    },
    {
      "entropy": 1.1565322875976562,
      "epoch": 1.0516061084781465,
      "grad_norm": 0.28871044516563416,
      "learning_rate": 0.00019327239669383287,
      "loss": 0.21732774376869202,
      "mean_token_accuracy": 0.9144384115934372,
      "num_tokens": 24633964.0,
      "step": 1997
    },
    {
      "entropy": 1.0519833862781525,
      "epoch": 1.052132701421801,
      "grad_norm": 0.25382453203201294,
      "learning_rate": 0.00019326407549112332,
      "loss": 0.2088373750448227,
      "mean_token_accuracy": 0.9129265993833542,
      "num_tokens": 24646300.0,
      "step": 1998
    },
    {
      "entropy": 1.0678929388523102,
      "epoch": 1.0526592943654556,
      "grad_norm": 0.2656494081020355,
      "learning_rate": 0.0001932557493455109,
      "loss": 0.19079969823360443,
      "mean_token_accuracy": 0.9278986901044846,
      "num_tokens": 24658636.0,
      "step": 1999
    },
    {
      "entropy": 1.074799358844757,
      "epoch": 1.05318588730911,
      "grad_norm": 0.27760738134384155,
      "learning_rate": 0.00019324741825748988,
      "loss": 0.18772991001605988,
      "mean_token_accuracy": 0.9245434403419495,
      "num_tokens": 24670972.0,
      "step": 2000
    },
    {
      "entropy": 1.0773050487041473,
      "epoch": 1.0537124802527646,
      "grad_norm": 0.27600982785224915,
      "learning_rate": 0.0001932390822275549,
      "loss": 0.20754915475845337,
      "mean_token_accuracy": 0.9160582423210144,
      "num_tokens": 24683308.0,
      "step": 2001
    },
    {
      "entropy": 1.0425165295600891,
      "epoch": 1.054239073196419,
      "grad_norm": 0.2946891188621521,
      "learning_rate": 0.0001932307412562007,
      "loss": 0.1997680515050888,
      "mean_token_accuracy": 0.9208045303821564,
      "num_tokens": 24695644.0,
      "step": 2002
    },
    {
      "entropy": 1.0058126598596573,
      "epoch": 1.0547656661400737,
      "grad_norm": 0.29050102829933167,
      "learning_rate": 0.00019322239534392253,
      "loss": 0.20790837705135345,
      "mean_token_accuracy": 0.9103892594575882,
      "num_tokens": 24707980.0,
      "step": 2003
    },
    {
      "entropy": 0.9951896965503693,
      "epoch": 1.0552922590837284,
      "grad_norm": 0.3011538088321686,
      "learning_rate": 0.0001932140444912158,
      "loss": 0.20860427618026733,
      "mean_token_accuracy": 0.9159162491559982,
      "num_tokens": 24720316.0,
      "step": 2004
    },
    {
      "entropy": 1.0219779312610626,
      "epoch": 1.0558188520273828,
      "grad_norm": 0.2890591323375702,
      "learning_rate": 0.00019320568869857623,
      "loss": 0.22163572907447815,
      "mean_token_accuracy": 0.911671981215477,
      "num_tokens": 24732652.0,
      "step": 2005
    },
    {
      "entropy": 1.0054317712783813,
      "epoch": 1.0563454449710374,
      "grad_norm": 0.31731894612312317,
      "learning_rate": 0.00019319732796649992,
      "loss": 0.21079890429973602,
      "mean_token_accuracy": 0.9145800471305847,
      "num_tokens": 24744988.0,
      "step": 2006
    },
    {
      "entropy": 0.9920244216918945,
      "epoch": 1.0568720379146919,
      "grad_norm": 0.26154065132141113,
      "learning_rate": 0.00019318896229548313,
      "loss": 0.1758606880903244,
      "mean_token_accuracy": 0.9275805801153183,
      "num_tokens": 24757324.0,
      "step": 2007
    },
    {
      "entropy": 1.013138324022293,
      "epoch": 1.0573986308583465,
      "grad_norm": 0.2615264058113098,
      "learning_rate": 0.00019318059168602251,
      "loss": 0.18549133837223053,
      "mean_token_accuracy": 0.9265227019786835,
      "num_tokens": 24769660.0,
      "step": 2008
    },
    {
      "entropy": 1.0628323256969452,
      "epoch": 1.0579252238020012,
      "grad_norm": 0.3117140233516693,
      "learning_rate": 0.000193172216138615,
      "loss": 0.22904282808303833,
      "mean_token_accuracy": 0.9080474376678467,
      "num_tokens": 24781996.0,
      "step": 2009
    },
    {
      "entropy": 1.061298355460167,
      "epoch": 1.0584518167456556,
      "grad_norm": 0.2991147041320801,
      "learning_rate": 0.00019316383565375777,
      "loss": 0.19954919815063477,
      "mean_token_accuracy": 0.9186259061098099,
      "num_tokens": 24794332.0,
      "step": 2010
    },
    {
      "entropy": 1.0400668680667877,
      "epoch": 1.0589784096893102,
      "grad_norm": 0.29817700386047363,
      "learning_rate": 0.00019315545023194835,
      "loss": 0.21616995334625244,
      "mean_token_accuracy": 0.912378340959549,
      "num_tokens": 24806668.0,
      "step": 2011
    },
    {
      "entropy": 1.0405495166778564,
      "epoch": 1.0595050026329647,
      "grad_norm": 0.2854476571083069,
      "learning_rate": 0.0001931470598736845,
      "loss": 0.2017708569765091,
      "mean_token_accuracy": 0.91714146733284,
      "num_tokens": 24819004.0,
      "step": 2012
    },
    {
      "entropy": 1.0748549401760101,
      "epoch": 1.0600315955766193,
      "grad_norm": 0.30470097064971924,
      "learning_rate": 0.00019313866457946433,
      "loss": 0.20843440294265747,
      "mean_token_accuracy": 0.917302206158638,
      "num_tokens": 24831340.0,
      "step": 2013
    },
    {
      "entropy": 1.0435102880001068,
      "epoch": 1.0605581885202737,
      "grad_norm": 0.29722654819488525,
      "learning_rate": 0.0001931302643497862,
      "loss": 0.2125845104455948,
      "mean_token_accuracy": 0.9157662689685822,
      "num_tokens": 24843676.0,
      "step": 2014
    },
    {
      "entropy": 1.0407266318798065,
      "epoch": 1.0610847814639284,
      "grad_norm": 0.27866822481155396,
      "learning_rate": 0.0001931218591851488,
      "loss": 0.1866873800754547,
      "mean_token_accuracy": 0.9244244992733002,
      "num_tokens": 24856012.0,
      "step": 2015
    },
    {
      "entropy": 1.0343786031007767,
      "epoch": 1.061611374407583,
      "grad_norm": 0.29595327377319336,
      "learning_rate": 0.00019311344908605108,
      "loss": 0.21080759167671204,
      "mean_token_accuracy": 0.9200511127710342,
      "num_tokens": 24868348.0,
      "step": 2016
    },
    {
      "entropy": 1.0134167224168777,
      "epoch": 1.0621379673512374,
      "grad_norm": 0.2973070740699768,
      "learning_rate": 0.0001931050340529923,
      "loss": 0.21294423937797546,
      "mean_token_accuracy": 0.9145906716585159,
      "num_tokens": 24880684.0,
      "step": 2017
    },
    {
      "entropy": 1.129615068435669,
      "epoch": 1.062664560294892,
      "grad_norm": 0.29881370067596436,
      "learning_rate": 0.00019309661408647207,
      "loss": 0.20720480382442474,
      "mean_token_accuracy": 0.9196062386035919,
      "num_tokens": 24893020.0,
      "step": 2018
    },
    {
      "entropy": 1.0598470270633698,
      "epoch": 1.0631911532385465,
      "grad_norm": 0.26026830077171326,
      "learning_rate": 0.00019308818918699016,
      "loss": 0.17674966156482697,
      "mean_token_accuracy": 0.9275426715612411,
      "num_tokens": 24905356.0,
      "step": 2019
    },
    {
      "entropy": 1.045887529850006,
      "epoch": 1.0637177461822012,
      "grad_norm": 0.25472280383110046,
      "learning_rate": 0.00019307975935504672,
      "loss": 0.17633673548698425,
      "mean_token_accuracy": 0.9277019053697586,
      "num_tokens": 24917692.0,
      "step": 2020
    },
    {
      "entropy": 1.0263851881027222,
      "epoch": 1.0642443391258558,
      "grad_norm": 0.2858225107192993,
      "learning_rate": 0.00019307132459114222,
      "loss": 0.20337790250778198,
      "mean_token_accuracy": 0.9162404388189316,
      "num_tokens": 24930028.0,
      "step": 2021
    },
    {
      "entropy": 1.0703861713409424,
      "epoch": 1.0647709320695102,
      "grad_norm": 0.3153981566429138,
      "learning_rate": 0.00019306288489577734,
      "loss": 0.217787504196167,
      "mean_token_accuracy": 0.9114295095205307,
      "num_tokens": 24942364.0,
      "step": 2022
    },
    {
      "entropy": 1.0497725158929825,
      "epoch": 1.0652975250131649,
      "grad_norm": 0.29816338419914246,
      "learning_rate": 0.0001930544402694531,
      "loss": 0.21639864146709442,
      "mean_token_accuracy": 0.9137828648090363,
      "num_tokens": 24954700.0,
      "step": 2023
    },
    {
      "entropy": 1.0218316316604614,
      "epoch": 1.0658241179568193,
      "grad_norm": 0.32789114117622375,
      "learning_rate": 0.00019304599071267082,
      "loss": 0.22291415929794312,
      "mean_token_accuracy": 0.9094216227531433,
      "num_tokens": 24967036.0,
      "step": 2024
    },
    {
      "entropy": 1.0307366997003555,
      "epoch": 1.066350710900474,
      "grad_norm": 0.2897968888282776,
      "learning_rate": 0.0001930375362259321,
      "loss": 0.20538388192653656,
      "mean_token_accuracy": 0.9126376360654831,
      "num_tokens": 24979372.0,
      "step": 2025
    },
    {
      "entropy": 1.0069323182106018,
      "epoch": 1.0668773038441284,
      "grad_norm": 0.258139044046402,
      "learning_rate": 0.00019302907680973888,
      "loss": 0.19350063800811768,
      "mean_token_accuracy": 0.9235964566469193,
      "num_tokens": 24991708.0,
      "step": 2026
    },
    {
      "entropy": 0.9675715416669846,
      "epoch": 1.067403896787783,
      "grad_norm": 0.28487148880958557,
      "learning_rate": 0.00019302061246459325,
      "loss": 0.2119164764881134,
      "mean_token_accuracy": 0.9105935990810394,
      "num_tokens": 25004044.0,
      "step": 2027
    },
    {
      "entropy": 1.0363286584615707,
      "epoch": 1.0679304897314377,
      "grad_norm": 0.30349263548851013,
      "learning_rate": 0.00019301214319099774,
      "loss": 0.20691387355327606,
      "mean_token_accuracy": 0.9157197773456573,
      "num_tokens": 25016380.0,
      "step": 2028
    },
    {
      "entropy": 1.0630298554897308,
      "epoch": 1.068457082675092,
      "grad_norm": 0.30097535252571106,
      "learning_rate": 0.0001930036689894551,
      "loss": 0.20818623900413513,
      "mean_token_accuracy": 0.9164175242185593,
      "num_tokens": 25028716.0,
      "step": 2029
    },
    {
      "entropy": 0.996089369058609,
      "epoch": 1.0689836756187467,
      "grad_norm": 0.28663763403892517,
      "learning_rate": 0.0001929951898604685,
      "loss": 0.2216407209634781,
      "mean_token_accuracy": 0.9137296378612518,
      "num_tokens": 25041052.0,
      "step": 2030
    },
    {
      "entropy": 1.0333895981311798,
      "epoch": 1.0695102685624012,
      "grad_norm": 0.2872811555862427,
      "learning_rate": 0.00019298670580454114,
      "loss": 0.19668258726596832,
      "mean_token_accuracy": 0.9209001511335373,
      "num_tokens": 25053388.0,
      "step": 2031
    },
    {
      "entropy": 1.050906777381897,
      "epoch": 1.0700368615060558,
      "grad_norm": 0.29872143268585205,
      "learning_rate": 0.00019297821682217676,
      "loss": 0.21886757016181946,
      "mean_token_accuracy": 0.9144444167613983,
      "num_tokens": 25065724.0,
      "step": 2032
    },
    {
      "entropy": 0.9869663417339325,
      "epoch": 1.0705634544497105,
      "grad_norm": 0.26373037695884705,
      "learning_rate": 0.00019296972291387926,
      "loss": 0.21317996084690094,
      "mean_token_accuracy": 0.913250133395195,
      "num_tokens": 25078060.0,
      "step": 2033
    },
    {
      "entropy": 0.996940404176712,
      "epoch": 1.0710900473933649,
      "grad_norm": 0.27339690923690796,
      "learning_rate": 0.00019296122408015294,
      "loss": 0.21194380521774292,
      "mean_token_accuracy": 0.9130052775144577,
      "num_tokens": 25090396.0,
      "step": 2034
    },
    {
      "entropy": 1.0184991508722305,
      "epoch": 1.0716166403370195,
      "grad_norm": 0.2744937539100647,
      "learning_rate": 0.00019295272032150224,
      "loss": 0.19760127365589142,
      "mean_token_accuracy": 0.9164283275604248,
      "num_tokens": 25102732.0,
      "step": 2035
    },
    {
      "entropy": 1.0074135214090347,
      "epoch": 1.072143233280674,
      "grad_norm": 0.28469040989875793,
      "learning_rate": 0.00019294421163843203,
      "loss": 0.2201453596353531,
      "mean_token_accuracy": 0.9090453386306763,
      "num_tokens": 25115068.0,
      "step": 2036
    },
    {
      "entropy": 0.9988047778606415,
      "epoch": 1.0726698262243286,
      "grad_norm": 0.26611679792404175,
      "learning_rate": 0.00019293569803144743,
      "loss": 0.21755945682525635,
      "mean_token_accuracy": 0.9131471365690231,
      "num_tokens": 25127404.0,
      "step": 2037
    },
    {
      "entropy": 0.9802228659391403,
      "epoch": 1.0731964191679833,
      "grad_norm": 0.27171045541763306,
      "learning_rate": 0.00019292717950105382,
      "loss": 0.21177193522453308,
      "mean_token_accuracy": 0.9141269624233246,
      "num_tokens": 25139740.0,
      "step": 2038
    },
    {
      "entropy": 1.0224299430847168,
      "epoch": 1.0737230121116377,
      "grad_norm": 0.304115355014801,
      "learning_rate": 0.00019291865604775688,
      "loss": 0.21834969520568848,
      "mean_token_accuracy": 0.9116663038730621,
      "num_tokens": 25152076.0,
      "step": 2039
    },
    {
      "entropy": 1.0109304785728455,
      "epoch": 1.0742496050552923,
      "grad_norm": 0.2777882218360901,
      "learning_rate": 0.00019291012767206261,
      "loss": 0.19935154914855957,
      "mean_token_accuracy": 0.9226459413766861,
      "num_tokens": 25164412.0,
      "step": 2040
    },
    {
      "entropy": 1.0325984209775925,
      "epoch": 1.0747761979989467,
      "grad_norm": 0.3269108235836029,
      "learning_rate": 0.00019290159437447733,
      "loss": 0.23336344957351685,
      "mean_token_accuracy": 0.9101394861936569,
      "num_tokens": 25176748.0,
      "step": 2041
    },
    {
      "entropy": 1.0169707685709,
      "epoch": 1.0753027909426014,
      "grad_norm": 0.2711583077907562,
      "learning_rate": 0.00019289305615550754,
      "loss": 0.19230744242668152,
      "mean_token_accuracy": 0.9222860485315323,
      "num_tokens": 25189084.0,
      "step": 2042
    },
    {
      "entropy": 0.9706145972013474,
      "epoch": 1.0758293838862558,
      "grad_norm": 0.26547273993492126,
      "learning_rate": 0.0001928845130156602,
      "loss": 0.20648162066936493,
      "mean_token_accuracy": 0.9119333028793335,
      "num_tokens": 25201420.0,
      "step": 2043
    },
    {
      "entropy": 1.017429158091545,
      "epoch": 1.0763559768299105,
      "grad_norm": 0.2850116491317749,
      "learning_rate": 0.00019287596495544233,
      "loss": 0.2100791037082672,
      "mean_token_accuracy": 0.9145694822072983,
      "num_tokens": 25213756.0,
      "step": 2044
    },
    {
      "entropy": 0.9998975098133087,
      "epoch": 1.0768825697735651,
      "grad_norm": 0.2765794098377228,
      "learning_rate": 0.0001928674119753615,
      "loss": 0.2017804980278015,
      "mean_token_accuracy": 0.9189972132444382,
      "num_tokens": 25226092.0,
      "step": 2045
    },
    {
      "entropy": 1.041368529200554,
      "epoch": 1.0774091627172195,
      "grad_norm": 0.2918790280818939,
      "learning_rate": 0.00019285885407592542,
      "loss": 0.20694515109062195,
      "mean_token_accuracy": 0.9102893173694611,
      "num_tokens": 25238428.0,
      "step": 2046
    },
    {
      "entropy": 0.9665294885635376,
      "epoch": 1.0779357556608742,
      "grad_norm": 0.2802746891975403,
      "learning_rate": 0.0001928502912576421,
      "loss": 0.2127482146024704,
      "mean_token_accuracy": 0.9124037772417068,
      "num_tokens": 25250764.0,
      "step": 2047
    },
    {
      "entropy": 1.009907528758049,
      "epoch": 1.0784623486045286,
      "grad_norm": 0.3065343201160431,
      "learning_rate": 0.0001928417235210199,
      "loss": 0.23531584441661835,
      "mean_token_accuracy": 0.9090160578489304,
      "num_tokens": 25263100.0,
      "step": 2048
    },
    {
      "entropy": 1.0422001481056213,
      "epoch": 1.0789889415481833,
      "grad_norm": 0.295613169670105,
      "learning_rate": 0.00019283315086656737,
      "loss": 0.20345695316791534,
      "mean_token_accuracy": 0.9175281077623367,
      "num_tokens": 25275436.0,
      "step": 2049
    },
    {
      "entropy": 1.0016635805368423,
      "epoch": 1.079515534491838,
      "grad_norm": 0.27219337224960327,
      "learning_rate": 0.0001928245732947935,
      "loss": 0.19025292992591858,
      "mean_token_accuracy": 0.9241680800914764,
      "num_tokens": 25287772.0,
      "step": 2050
    },
    {
      "entropy": 0.9872867614030838,
      "epoch": 1.0800421274354923,
      "grad_norm": 0.2815146744251251,
      "learning_rate": 0.00019281599080620741,
      "loss": 0.2239912748336792,
      "mean_token_accuracy": 0.9086337685585022,
      "num_tokens": 25300108.0,
      "step": 2051
    },
    {
      "entropy": 1.0149886310100555,
      "epoch": 1.080568720379147,
      "grad_norm": 0.2981141209602356,
      "learning_rate": 0.0001928074034013187,
      "loss": 0.21916988492012024,
      "mean_token_accuracy": 0.9091440141201019,
      "num_tokens": 25312444.0,
      "step": 2052
    },
    {
      "entropy": 0.9956868588924408,
      "epoch": 1.0810953133228014,
      "grad_norm": 0.28481054306030273,
      "learning_rate": 0.00019279881108063705,
      "loss": 0.21595682203769684,
      "mean_token_accuracy": 0.912951722741127,
      "num_tokens": 25324780.0,
      "step": 2053
    },
    {
      "entropy": 1.0007793605327606,
      "epoch": 1.081621906266456,
      "grad_norm": 0.281215637922287,
      "learning_rate": 0.00019279021384467255,
      "loss": 0.21096494793891907,
      "mean_token_accuracy": 0.9124733507633209,
      "num_tokens": 25337116.0,
      "step": 2054
    },
    {
      "entropy": 1.0316517055034637,
      "epoch": 1.0821484992101107,
      "grad_norm": 0.4712248742580414,
      "learning_rate": 0.00019278161169393563,
      "loss": 0.21723365783691406,
      "mean_token_accuracy": 0.9101579338312149,
      "num_tokens": 25349452.0,
      "step": 2055
    },
    {
      "entropy": 1.036922350525856,
      "epoch": 1.0826750921537651,
      "grad_norm": 0.27001330256462097,
      "learning_rate": 0.0001927730046289369,
      "loss": 0.2058403491973877,
      "mean_token_accuracy": 0.9116992652416229,
      "num_tokens": 25361788.0,
      "step": 2056
    },
    {
      "entropy": 1.043273389339447,
      "epoch": 1.0832016850974198,
      "grad_norm": 0.29966092109680176,
      "learning_rate": 0.0001927643926501873,
      "loss": 0.2038857638835907,
      "mean_token_accuracy": 0.9206432849168777,
      "num_tokens": 25374124.0,
      "step": 2057
    },
    {
      "entropy": 1.1009390354156494,
      "epoch": 1.0837282780410742,
      "grad_norm": 0.295094758272171,
      "learning_rate": 0.0001927557757581981,
      "loss": 0.21461978554725647,
      "mean_token_accuracy": 0.9137238711118698,
      "num_tokens": 25386460.0,
      "step": 2058
    },
    {
      "entropy": 1.0440688282251358,
      "epoch": 1.0842548709847288,
      "grad_norm": 0.2719572186470032,
      "learning_rate": 0.00019274715395348085,
      "loss": 0.19468249380588531,
      "mean_token_accuracy": 0.9245257526636124,
      "num_tokens": 25398796.0,
      "step": 2059
    },
    {
      "entropy": 1.095595896244049,
      "epoch": 1.0847814639283833,
      "grad_norm": 0.30201783776283264,
      "learning_rate": 0.00019273852723654736,
      "loss": 0.2008979767560959,
      "mean_token_accuracy": 0.9223850518465042,
      "num_tokens": 25411132.0,
      "step": 2060
    },
    {
      "entropy": 1.072413593530655,
      "epoch": 1.085308056872038,
      "grad_norm": 0.3010813891887665,
      "learning_rate": 0.00019272989560790974,
      "loss": 0.1963309794664383,
      "mean_token_accuracy": 0.9216117411851883,
      "num_tokens": 25423468.0,
      "step": 2061
    },
    {
      "entropy": 1.0537616908550262,
      "epoch": 1.0858346498156926,
      "grad_norm": 0.4835893213748932,
      "learning_rate": 0.00019272125906808038,
      "loss": 0.2151128351688385,
      "mean_token_accuracy": 0.9163444191217422,
      "num_tokens": 25435804.0,
      "step": 2062
    },
    {
      "entropy": 1.0696274489164352,
      "epoch": 1.086361242759347,
      "grad_norm": 0.2978211045265198,
      "learning_rate": 0.00019271261761757202,
      "loss": 0.19762234389781952,
      "mean_token_accuracy": 0.9217347949743271,
      "num_tokens": 25448140.0,
      "step": 2063
    },
    {
      "entropy": 1.0669613778591156,
      "epoch": 1.0868878357030016,
      "grad_norm": 0.27142512798309326,
      "learning_rate": 0.00019270397125689765,
      "loss": 0.20439964532852173,
      "mean_token_accuracy": 0.9173642098903656,
      "num_tokens": 25460476.0,
      "step": 2064
    },
    {
      "entropy": 1.0918617248535156,
      "epoch": 1.087414428646656,
      "grad_norm": 0.2899603545665741,
      "learning_rate": 0.00019269531998657055,
      "loss": 0.2081996202468872,
      "mean_token_accuracy": 0.916422575712204,
      "num_tokens": 25472812.0,
      "step": 2065
    },
    {
      "entropy": 1.0441029965877533,
      "epoch": 1.0879410215903107,
      "grad_norm": 0.2680945098400116,
      "learning_rate": 0.00019268666380710425,
      "loss": 0.19378285109996796,
      "mean_token_accuracy": 0.9220094531774521,
      "num_tokens": 25485148.0,
      "step": 2066
    },
    {
      "entropy": 1.0455826669931412,
      "epoch": 1.0884676145339653,
      "grad_norm": 0.2643810212612152,
      "learning_rate": 0.0001926780027190127,
      "loss": 0.19173350930213928,
      "mean_token_accuracy": 0.9210256785154343,
      "num_tokens": 25497484.0,
      "step": 2067
    },
    {
      "entropy": 1.169319897890091,
      "epoch": 1.0889942074776198,
      "grad_norm": 0.3351131081581116,
      "learning_rate": 0.00019266933672280998,
      "loss": 0.20055952668190002,
      "mean_token_accuracy": 0.9239787012338638,
      "num_tokens": 25509820.0,
      "step": 2068
    },
    {
      "entropy": 0.9933416992425919,
      "epoch": 1.0895208004212744,
      "grad_norm": 0.2722816467285156,
      "learning_rate": 0.00019266066581901056,
      "loss": 0.18528851866722107,
      "mean_token_accuracy": 0.925559476017952,
      "num_tokens": 25522156.0,
      "step": 2069
    },
    {
      "entropy": 1.0143402963876724,
      "epoch": 1.0900473933649288,
      "grad_norm": 0.2853977680206299,
      "learning_rate": 0.00019265199000812922,
      "loss": 0.20564962923526764,
      "mean_token_accuracy": 0.918012797832489,
      "num_tokens": 25534492.0,
      "step": 2070
    },
    {
      "entropy": 1.0507995039224625,
      "epoch": 1.0905739863085835,
      "grad_norm": 0.2650372385978699,
      "learning_rate": 0.00019264330929068098,
      "loss": 0.19438353180885315,
      "mean_token_accuracy": 0.9210219085216522,
      "num_tokens": 25546828.0,
      "step": 2071
    },
    {
      "entropy": 1.0700190365314484,
      "epoch": 1.0911005792522381,
      "grad_norm": 0.2895369529724121,
      "learning_rate": 0.00019263462366718112,
      "loss": 0.20609940588474274,
      "mean_token_accuracy": 0.9171310365200043,
      "num_tokens": 25559164.0,
      "step": 2072
    },
    {
      "entropy": 1.0044712573289871,
      "epoch": 1.0916271721958926,
      "grad_norm": 0.2867662310600281,
      "learning_rate": 0.0001926259331381453,
      "loss": 0.1890471875667572,
      "mean_token_accuracy": 0.9236837476491928,
      "num_tokens": 25571500.0,
      "step": 2073
    },
    {
      "entropy": 1.0000011622905731,
      "epoch": 1.0921537651395472,
      "grad_norm": 0.2822493612766266,
      "learning_rate": 0.00019261723770408942,
      "loss": 0.20667140185832977,
      "mean_token_accuracy": 0.9199023395776749,
      "num_tokens": 25583836.0,
      "step": 2074
    },
    {
      "entropy": 0.966169461607933,
      "epoch": 1.0926803580832016,
      "grad_norm": 0.28859975934028625,
      "learning_rate": 0.00019260853736552963,
      "loss": 0.21461637318134308,
      "mean_token_accuracy": 0.909663125872612,
      "num_tokens": 25596172.0,
      "step": 2075
    },
    {
      "entropy": 1.0433668792247772,
      "epoch": 1.0932069510268563,
      "grad_norm": 0.2675197422504425,
      "learning_rate": 0.0001925998321229825,
      "loss": 0.20010057091712952,
      "mean_token_accuracy": 0.9186754375696182,
      "num_tokens": 25608508.0,
      "step": 2076
    },
    {
      "entropy": 1.08010932803154,
      "epoch": 1.0937335439705107,
      "grad_norm": 0.2648799419403076,
      "learning_rate": 0.00019259112197696473,
      "loss": 0.1912347376346588,
      "mean_token_accuracy": 0.923126220703125,
      "num_tokens": 25620844.0,
      "step": 2077
    },
    {
      "entropy": 1.0481482297182083,
      "epoch": 1.0942601369141653,
      "grad_norm": 0.26581716537475586,
      "learning_rate": 0.00019258240692799342,
      "loss": 0.2124476432800293,
      "mean_token_accuracy": 0.9156401306390762,
      "num_tokens": 25633180.0,
      "step": 2078
    },
    {
      "entropy": 0.991655707359314,
      "epoch": 1.09478672985782,
      "grad_norm": 0.24418914318084717,
      "learning_rate": 0.00019257368697658597,
      "loss": 0.1879653036594391,
      "mean_token_accuracy": 0.9245227575302124,
      "num_tokens": 25645516.0,
      "step": 2079
    },
    {
      "entropy": 1.029534325003624,
      "epoch": 1.0953133228014744,
      "grad_norm": 0.27712008357048035,
      "learning_rate": 0.00019256496212326,
      "loss": 0.19567881524562836,
      "mean_token_accuracy": 0.9220569729804993,
      "num_tokens": 25657852.0,
      "step": 2080
    },
    {
      "entropy": 1.0455122739076614,
      "epoch": 1.095839915745129,
      "grad_norm": 0.25748133659362793,
      "learning_rate": 0.0001925562323685334,
      "loss": 0.19087985157966614,
      "mean_token_accuracy": 0.9210804253816605,
      "num_tokens": 25670188.0,
      "step": 2081
    },
    {
      "entropy": 1.053975835442543,
      "epoch": 1.0963665086887835,
      "grad_norm": 0.2648088037967682,
      "learning_rate": 0.0001925474977129245,
      "loss": 0.20259462296962738,
      "mean_token_accuracy": 0.9170144945383072,
      "num_tokens": 25682524.0,
      "step": 2082
    },
    {
      "entropy": 1.1167932152748108,
      "epoch": 1.0968931016324381,
      "grad_norm": 0.3016040325164795,
      "learning_rate": 0.0001925387581569518,
      "loss": 0.2160032093524933,
      "mean_token_accuracy": 0.9073923379182816,
      "num_tokens": 25694860.0,
      "step": 2083
    },
    {
      "entropy": 1.0568408221006393,
      "epoch": 1.0974196945760926,
      "grad_norm": 0.29051339626312256,
      "learning_rate": 0.0001925300137011341,
      "loss": 0.1947801113128662,
      "mean_token_accuracy": 0.9265160709619522,
      "num_tokens": 25707196.0,
      "step": 2084
    },
    {
      "entropy": 1.0293535590171814,
      "epoch": 1.0979462875197472,
      "grad_norm": 0.28292426466941833,
      "learning_rate": 0.00019252126434599046,
      "loss": 0.20448973774909973,
      "mean_token_accuracy": 0.9163498878479004,
      "num_tokens": 25719532.0,
      "step": 2085
    },
    {
      "entropy": 0.9932720512151718,
      "epoch": 1.0984728804634019,
      "grad_norm": 0.268726110458374,
      "learning_rate": 0.00019251251009204037,
      "loss": 0.20217624306678772,
      "mean_token_accuracy": 0.9180169999599457,
      "num_tokens": 25731868.0,
      "step": 2086
    },
    {
      "entropy": 1.03742016851902,
      "epoch": 1.0989994734070563,
      "grad_norm": 0.2555520534515381,
      "learning_rate": 0.00019250375093980346,
      "loss": 0.19689497351646423,
      "mean_token_accuracy": 0.922407329082489,
      "num_tokens": 25744204.0,
      "step": 2087
    },
    {
      "entropy": 1.0640356242656708,
      "epoch": 1.099526066350711,
      "grad_norm": 0.28077611327171326,
      "learning_rate": 0.00019249498688979973,
      "loss": 0.2088557481765747,
      "mean_token_accuracy": 0.9098563194274902,
      "num_tokens": 25756540.0,
      "step": 2088
    },
    {
      "entropy": 1.059511423110962,
      "epoch": 1.1000526592943654,
      "grad_norm": 0.27445870637893677,
      "learning_rate": 0.00019248621794254944,
      "loss": 0.20993183553218842,
      "mean_token_accuracy": 0.916303813457489,
      "num_tokens": 25768876.0,
      "step": 2089
    },
    {
      "entropy": 1.0723062455654144,
      "epoch": 1.10057925223802,
      "grad_norm": 0.2826826274394989,
      "learning_rate": 0.00019247744409857316,
      "loss": 0.20269325375556946,
      "mean_token_accuracy": 0.9222795218229294,
      "num_tokens": 25781212.0,
      "step": 2090
    },
    {
      "entropy": 1.0870884507894516,
      "epoch": 1.1011058451816746,
      "grad_norm": 0.27127087116241455,
      "learning_rate": 0.00019246866535839173,
      "loss": 0.20817485451698303,
      "mean_token_accuracy": 0.916502445936203,
      "num_tokens": 25793548.0,
      "step": 2091
    },
    {
      "entropy": 1.037529081106186,
      "epoch": 1.101632438125329,
      "grad_norm": 0.28745195269584656,
      "learning_rate": 0.0001924598817225263,
      "loss": 0.22985900938510895,
      "mean_token_accuracy": 0.9097397327423096,
      "num_tokens": 25805884.0,
      "step": 2092
    },
    {
      "entropy": 1.0101779103279114,
      "epoch": 1.1021590310689837,
      "grad_norm": 0.26528680324554443,
      "learning_rate": 0.00019245109319149833,
      "loss": 0.19186118245124817,
      "mean_token_accuracy": 0.923490360379219,
      "num_tokens": 25818220.0,
      "step": 2093
    },
    {
      "entropy": 1.0416104793548584,
      "epoch": 1.1026856240126381,
      "grad_norm": 0.2599078118801117,
      "learning_rate": 0.0001924422997658295,
      "loss": 0.2087230086326599,
      "mean_token_accuracy": 0.9174604564905167,
      "num_tokens": 25830556.0,
      "step": 2094
    },
    {
      "entropy": 1.071406990289688,
      "epoch": 1.1032122169562928,
      "grad_norm": 0.27701428532600403,
      "learning_rate": 0.00019243350144604185,
      "loss": 0.21016359329223633,
      "mean_token_accuracy": 0.9129587262868881,
      "num_tokens": 25842892.0,
      "step": 2095
    },
    {
      "entropy": 1.0269631147384644,
      "epoch": 1.1037388098999474,
      "grad_norm": 0.28013738989830017,
      "learning_rate": 0.00019242469823265769,
      "loss": 0.20927533507347107,
      "mean_token_accuracy": 0.9170541018247604,
      "num_tokens": 25855228.0,
      "step": 2096
    },
    {
      "entropy": 1.076466590166092,
      "epoch": 1.1042654028436019,
      "grad_norm": 0.2795049548149109,
      "learning_rate": 0.00019241589012619953,
      "loss": 0.20179404318332672,
      "mean_token_accuracy": 0.9215785712003708,
      "num_tokens": 25867564.0,
      "step": 2097
    },
    {
      "entropy": 1.097602665424347,
      "epoch": 1.1047919957872565,
      "grad_norm": 0.2958826720714569,
      "learning_rate": 0.00019240707712719042,
      "loss": 0.22175639867782593,
      "mean_token_accuracy": 0.9060819000005722,
      "num_tokens": 25879900.0,
      "step": 2098
    },
    {
      "entropy": 1.074698805809021,
      "epoch": 1.105318588730911,
      "grad_norm": 0.26773980259895325,
      "learning_rate": 0.0001923982592361534,
      "loss": 0.18457193672657013,
      "mean_token_accuracy": 0.923177182674408,
      "num_tokens": 25892236.0,
      "step": 2099
    },
    {
      "entropy": 1.1188512742519379,
      "epoch": 1.1058451816745656,
      "grad_norm": 0.2522232234477997,
      "learning_rate": 0.000192389436453612,
      "loss": 0.19157402217388153,
      "mean_token_accuracy": 0.921467125415802,
      "num_tokens": 25904572.0,
      "step": 2100
    },
    {
      "entropy": 1.101550132036209,
      "epoch": 1.10637177461822,
      "grad_norm": 0.268435537815094,
      "learning_rate": 0.00019238060878008995,
      "loss": 0.1853233426809311,
      "mean_token_accuracy": 0.9239556342363358,
      "num_tokens": 25916908.0,
      "step": 2101
    },
    {
      "entropy": 1.1288717091083527,
      "epoch": 1.1068983675618747,
      "grad_norm": 0.30907148122787476,
      "learning_rate": 0.00019237177621611134,
      "loss": 0.21557700634002686,
      "mean_token_accuracy": 0.912951648235321,
      "num_tokens": 25929244.0,
      "step": 2102
    },
    {
      "entropy": 1.0122136920690536,
      "epoch": 1.1074249605055293,
      "grad_norm": 0.2754114866256714,
      "learning_rate": 0.00019236293876220042,
      "loss": 0.22462958097457886,
      "mean_token_accuracy": 0.9148844480514526,
      "num_tokens": 25941580.0,
      "step": 2103
    },
    {
      "entropy": 1.0024352222681046,
      "epoch": 1.1079515534491837,
      "grad_norm": 0.28880998492240906,
      "learning_rate": 0.0001923540964188819,
      "loss": 0.19037997722625732,
      "mean_token_accuracy": 0.9214117079973221,
      "num_tokens": 25953916.0,
      "step": 2104
    },
    {
      "entropy": 1.090283751487732,
      "epoch": 1.1084781463928384,
      "grad_norm": 0.3156434893608093,
      "learning_rate": 0.0001923452491866807,
      "loss": 0.22172510623931885,
      "mean_token_accuracy": 0.9104326665401459,
      "num_tokens": 25966252.0,
      "step": 2105
    },
    {
      "entropy": 1.0386274009943008,
      "epoch": 1.1090047393364928,
      "grad_norm": 0.28844019770622253,
      "learning_rate": 0.00019233639706612196,
      "loss": 0.19381242990493774,
      "mean_token_accuracy": 0.920799121260643,
      "num_tokens": 25978588.0,
      "step": 2106
    },
    {
      "entropy": 1.0614490807056427,
      "epoch": 1.1095313322801474,
      "grad_norm": 0.2973669767379761,
      "learning_rate": 0.00019232754005773123,
      "loss": 0.206210196018219,
      "mean_token_accuracy": 0.9167844653129578,
      "num_tokens": 25990924.0,
      "step": 2107
    },
    {
      "entropy": 1.0339891910552979,
      "epoch": 1.110057925223802,
      "grad_norm": 0.29104986786842346,
      "learning_rate": 0.00019231867816203428,
      "loss": 0.20842497050762177,
      "mean_token_accuracy": 0.9144662916660309,
      "num_tokens": 26003260.0,
      "step": 2108
    },
    {
      "entropy": 1.0249072164297104,
      "epoch": 1.1105845181674565,
      "grad_norm": 0.28756386041641235,
      "learning_rate": 0.0001923098113795572,
      "loss": 0.2018943727016449,
      "mean_token_accuracy": 0.9205996543169022,
      "num_tokens": 26015596.0,
      "step": 2109
    },
    {
      "entropy": 1.017577350139618,
      "epoch": 1.1111111111111112,
      "grad_norm": 0.30264243483543396,
      "learning_rate": 0.0001923009397108264,
      "loss": 0.20579789578914642,
      "mean_token_accuracy": 0.9167483001947403,
      "num_tokens": 26027932.0,
      "step": 2110
    },
    {
      "entropy": 1.0238764435052872,
      "epoch": 1.1116377040547656,
      "grad_norm": 0.27734920382499695,
      "learning_rate": 0.00019229206315636845,
      "loss": 0.19919277727603912,
      "mean_token_accuracy": 0.9176904708147049,
      "num_tokens": 26040268.0,
      "step": 2111
    },
    {
      "entropy": 1.002189740538597,
      "epoch": 1.1121642969984202,
      "grad_norm": 0.2930961847305298,
      "learning_rate": 0.00019228318171671035,
      "loss": 0.19597947597503662,
      "mean_token_accuracy": 0.9197946488857269,
      "num_tokens": 26052604.0,
      "step": 2112
    },
    {
      "entropy": 0.9857950955629349,
      "epoch": 1.1126908899420749,
      "grad_norm": 0.28313517570495605,
      "learning_rate": 0.00019227429539237935,
      "loss": 0.2069045752286911,
      "mean_token_accuracy": 0.9212967753410339,
      "num_tokens": 26064940.0,
      "step": 2113
    },
    {
      "entropy": 1.0099002569913864,
      "epoch": 1.1132174828857293,
      "grad_norm": 0.2959808111190796,
      "learning_rate": 0.00019226540418390293,
      "loss": 0.21741098165512085,
      "mean_token_accuracy": 0.9100223034620285,
      "num_tokens": 26077276.0,
      "step": 2114
    },
    {
      "entropy": 0.9958448708057404,
      "epoch": 1.113744075829384,
      "grad_norm": 0.2769598364830017,
      "learning_rate": 0.00019225650809180897,
      "loss": 0.20837773382663727,
      "mean_token_accuracy": 0.918271854519844,
      "num_tokens": 26089612.0,
      "step": 2115
    },
    {
      "entropy": 1.014888659119606,
      "epoch": 1.1142706687730384,
      "grad_norm": 0.29256078600883484,
      "learning_rate": 0.00019224760711662555,
      "loss": 0.20603659749031067,
      "mean_token_accuracy": 0.9164493978023529,
      "num_tokens": 26101948.0,
      "step": 2116
    },
    {
      "entropy": 1.005908653140068,
      "epoch": 1.114797261716693,
      "grad_norm": 0.2789181172847748,
      "learning_rate": 0.00019223870125888108,
      "loss": 0.19784985482692719,
      "mean_token_accuracy": 0.9218984693288803,
      "num_tokens": 26114284.0,
      "step": 2117
    },
    {
      "entropy": 1.0451974421739578,
      "epoch": 1.1153238546603474,
      "grad_norm": 0.27004557847976685,
      "learning_rate": 0.0001922297905191042,
      "loss": 0.19614283740520477,
      "mean_token_accuracy": 0.9270102232694626,
      "num_tokens": 26126620.0,
      "step": 2118
    },
    {
      "entropy": 1.091399073600769,
      "epoch": 1.115850447604002,
      "grad_norm": 0.28412166237831116,
      "learning_rate": 0.000192220874897824,
      "loss": 0.18808285892009735,
      "mean_token_accuracy": 0.9227005243301392,
      "num_tokens": 26138956.0,
      "step": 2119
    },
    {
      "entropy": 1.0639157593250275,
      "epoch": 1.1163770405476567,
      "grad_norm": 0.29297924041748047,
      "learning_rate": 0.0001922119543955696,
      "loss": 0.20654083788394928,
      "mean_token_accuracy": 0.9152026325464249,
      "num_tokens": 26151292.0,
      "step": 2120
    },
    {
      "entropy": 1.0102118253707886,
      "epoch": 1.1169036334913112,
      "grad_norm": 0.27283284068107605,
      "learning_rate": 0.0001922030290128707,
      "loss": 0.19149716198444366,
      "mean_token_accuracy": 0.9258815199136734,
      "num_tokens": 26163628.0,
      "step": 2121
    },
    {
      "entropy": 1.0362121015787125,
      "epoch": 1.1174302264349658,
      "grad_norm": 0.28490447998046875,
      "learning_rate": 0.00019219409875025705,
      "loss": 0.1954336166381836,
      "mean_token_accuracy": 0.9174920916557312,
      "num_tokens": 26175964.0,
      "step": 2122
    },
    {
      "entropy": 1.0734621286392212,
      "epoch": 1.1179568193786202,
      "grad_norm": 0.289915531873703,
      "learning_rate": 0.0001921851636082588,
      "loss": 0.20870134234428406,
      "mean_token_accuracy": 0.9131461679935455,
      "num_tokens": 26188300.0,
      "step": 2123
    },
    {
      "entropy": 1.1021492779254913,
      "epoch": 1.1184834123222749,
      "grad_norm": 0.30639079213142395,
      "learning_rate": 0.0001921762235874064,
      "loss": 0.21529828011989594,
      "mean_token_accuracy": 0.9132403880357742,
      "num_tokens": 26200636.0,
      "step": 2124
    },
    {
      "entropy": 1.041254460811615,
      "epoch": 1.1190100052659295,
      "grad_norm": 0.24776063859462738,
      "learning_rate": 0.00019216727868823062,
      "loss": 0.19100171327590942,
      "mean_token_accuracy": 0.9220355451107025,
      "num_tokens": 26212972.0,
      "step": 2125
    },
    {
      "entropy": 1.0897882282733917,
      "epoch": 1.119536598209584,
      "grad_norm": 0.28905171155929565,
      "learning_rate": 0.00019215832891126237,
      "loss": 0.19190385937690735,
      "mean_token_accuracy": 0.9242671728134155,
      "num_tokens": 26225308.0,
      "step": 2126
    },
    {
      "entropy": 1.1073324233293533,
      "epoch": 1.1200631911532386,
      "grad_norm": 0.2728095054626465,
      "learning_rate": 0.000192149374257033,
      "loss": 0.2034318745136261,
      "mean_token_accuracy": 0.9145397841930389,
      "num_tokens": 26237644.0,
      "step": 2127
    },
    {
      "entropy": 1.1713103353977203,
      "epoch": 1.120589784096893,
      "grad_norm": 0.31770387291908264,
      "learning_rate": 0.00019214041472607408,
      "loss": 0.22776833176612854,
      "mean_token_accuracy": 0.9096698760986328,
      "num_tokens": 26249980.0,
      "step": 2128
    },
    {
      "entropy": 1.169573962688446,
      "epoch": 1.1211163770405477,
      "grad_norm": 0.3100426495075226,
      "learning_rate": 0.00019213145031891747,
      "loss": 0.19859901070594788,
      "mean_token_accuracy": 0.9183387011289597,
      "num_tokens": 26262316.0,
      "step": 2129
    },
    {
      "entropy": 1.0875117480754852,
      "epoch": 1.1216429699842023,
      "grad_norm": 0.27162665128707886,
      "learning_rate": 0.00019212248103609537,
      "loss": 0.20907218754291534,
      "mean_token_accuracy": 0.9196607321500778,
      "num_tokens": 26274652.0,
      "step": 2130
    },
    {
      "entropy": 1.084071934223175,
      "epoch": 1.1221695629278567,
      "grad_norm": 0.2759605944156647,
      "learning_rate": 0.00019211350687814023,
      "loss": 0.18926078081130981,
      "mean_token_accuracy": 0.9211393296718597,
      "num_tokens": 26286988.0,
      "step": 2131
    },
    {
      "entropy": 1.1398064196109772,
      "epoch": 1.1226961558715114,
      "grad_norm": 0.28818970918655396,
      "learning_rate": 0.00019210452784558478,
      "loss": 0.20041367411613464,
      "mean_token_accuracy": 0.9223820567131042,
      "num_tokens": 26299324.0,
      "step": 2132
    },
    {
      "entropy": 1.0466588735580444,
      "epoch": 1.1232227488151658,
      "grad_norm": 0.29245075583457947,
      "learning_rate": 0.000192095543938962,
      "loss": 0.20363174378871918,
      "mean_token_accuracy": 0.9165193289518356,
      "num_tokens": 26311660.0,
      "step": 2133
    },
    {
      "entropy": 1.0405475795269012,
      "epoch": 1.1237493417588205,
      "grad_norm": 0.2925106883049011,
      "learning_rate": 0.00019208655515880532,
      "loss": 0.1999765932559967,
      "mean_token_accuracy": 0.9169381856918335,
      "num_tokens": 26323996.0,
      "step": 2134
    },
    {
      "entropy": 1.04118974506855,
      "epoch": 1.1242759347024749,
      "grad_norm": 0.26943859457969666,
      "learning_rate": 0.0001920775615056483,
      "loss": 0.19177857041358948,
      "mean_token_accuracy": 0.9220841228961945,
      "num_tokens": 26336332.0,
      "step": 2135
    },
    {
      "entropy": 1.0740750879049301,
      "epoch": 1.1248025276461295,
      "grad_norm": 0.2784332036972046,
      "learning_rate": 0.0001920685629800248,
      "loss": 0.18828029930591583,
      "mean_token_accuracy": 0.9239522963762283,
      "num_tokens": 26348668.0,
      "step": 2136
    },
    {
      "entropy": 1.032617449760437,
      "epoch": 1.1253291205897842,
      "grad_norm": 0.28828391432762146,
      "learning_rate": 0.00019205955958246905,
      "loss": 0.20189997553825378,
      "mean_token_accuracy": 0.9197202920913696,
      "num_tokens": 26361004.0,
      "step": 2137
    },
    {
      "entropy": 1.1041496396064758,
      "epoch": 1.1258557135334386,
      "grad_norm": 0.3010398745536804,
      "learning_rate": 0.00019205055131351557,
      "loss": 0.20388422906398773,
      "mean_token_accuracy": 0.9182727336883545,
      "num_tokens": 26373340.0,
      "step": 2138
    },
    {
      "entropy": 1.0780200958251953,
      "epoch": 1.1263823064770933,
      "grad_norm": 0.30870407819747925,
      "learning_rate": 0.00019204153817369905,
      "loss": 0.21839921176433563,
      "mean_token_accuracy": 0.9158252775669098,
      "num_tokens": 26385676.0,
      "step": 2139
    },
    {
      "entropy": 1.090522974729538,
      "epoch": 1.1269088994207477,
      "grad_norm": 0.3072364330291748,
      "learning_rate": 0.00019203252016355458,
      "loss": 0.1994020640850067,
      "mean_token_accuracy": 0.9187403917312622,
      "num_tokens": 26398012.0,
      "step": 2140
    },
    {
      "entropy": 1.1098020374774933,
      "epoch": 1.1274354923644023,
      "grad_norm": 0.2888351082801819,
      "learning_rate": 0.00019202349728361754,
      "loss": 0.2039174735546112,
      "mean_token_accuracy": 0.9148209691047668,
      "num_tokens": 26410348.0,
      "step": 2141
    },
    {
      "entropy": 1.1232051849365234,
      "epoch": 1.1279620853080567,
      "grad_norm": 0.2976013123989105,
      "learning_rate": 0.0001920144695344235,
      "loss": 0.2292768359184265,
      "mean_token_accuracy": 0.9084900766611099,
      "num_tokens": 26422684.0,
      "step": 2142
    },
    {
      "entropy": 1.1429337561130524,
      "epoch": 1.1284886782517114,
      "grad_norm": 0.2777968645095825,
      "learning_rate": 0.00019200543691650845,
      "loss": 0.19244031608104706,
      "mean_token_accuracy": 0.9237828403711319,
      "num_tokens": 26435020.0,
      "step": 2143
    },
    {
      "entropy": 1.096487671136856,
      "epoch": 1.129015271195366,
      "grad_norm": 0.25569429993629456,
      "learning_rate": 0.00019199639943040856,
      "loss": 0.1880742460489273,
      "mean_token_accuracy": 0.924060508608818,
      "num_tokens": 26447356.0,
      "step": 2144
    },
    {
      "entropy": 1.1004936397075653,
      "epoch": 1.1295418641390205,
      "grad_norm": 0.25819581747055054,
      "learning_rate": 0.00019198735707666036,
      "loss": 0.18885719776153564,
      "mean_token_accuracy": 0.9271882772445679,
      "num_tokens": 26459692.0,
      "step": 2145
    },
    {
      "entropy": 1.1160069704055786,
      "epoch": 1.1300684570826751,
      "grad_norm": 0.28844067454338074,
      "learning_rate": 0.00019197830985580064,
      "loss": 0.2207649052143097,
      "mean_token_accuracy": 0.9111542254686356,
      "num_tokens": 26472028.0,
      "step": 2146
    },
    {
      "entropy": 1.1166575253009796,
      "epoch": 1.1305950500263298,
      "grad_norm": 0.26548051834106445,
      "learning_rate": 0.00019196925776836644,
      "loss": 0.20018815994262695,
      "mean_token_accuracy": 0.9172813147306442,
      "num_tokens": 26484364.0,
      "step": 2147
    },
    {
      "entropy": 1.1224709153175354,
      "epoch": 1.1311216429699842,
      "grad_norm": 0.26083195209503174,
      "learning_rate": 0.00019196020081489517,
      "loss": 0.19701796770095825,
      "mean_token_accuracy": 0.9212660789489746,
      "num_tokens": 26496700.0,
      "step": 2148
    },
    {
      "entropy": 1.1387836039066315,
      "epoch": 1.1316482359136388,
      "grad_norm": 0.2727262079715729,
      "learning_rate": 0.00019195113899592446,
      "loss": 0.18492068350315094,
      "mean_token_accuracy": 0.9268793612718582,
      "num_tokens": 26509036.0,
      "step": 2149
    },
    {
      "entropy": 1.1890142858028412,
      "epoch": 1.1321748288572933,
      "grad_norm": 0.28917086124420166,
      "learning_rate": 0.0001919420723119923,
      "loss": 0.21750162541866302,
      "mean_token_accuracy": 0.9132700264453888,
      "num_tokens": 26521372.0,
      "step": 2150
    },
    {
      "entropy": 1.1991948783397675,
      "epoch": 1.132701421800948,
      "grad_norm": 0.3479141294956207,
      "learning_rate": 0.00019193300076363688,
      "loss": 0.21669164299964905,
      "mean_token_accuracy": 0.9105279594659805,
      "num_tokens": 26533708.0,
      "step": 2151
    },
    {
      "entropy": 1.166500449180603,
      "epoch": 1.1332280147446023,
      "grad_norm": 0.27059152722358704,
      "learning_rate": 0.00019192392435139676,
      "loss": 0.1777915358543396,
      "mean_token_accuracy": 0.9269713759422302,
      "num_tokens": 26546044.0,
      "step": 2152
    },
    {
      "entropy": 1.1897345781326294,
      "epoch": 1.133754607688257,
      "grad_norm": 0.27053096890449524,
      "learning_rate": 0.00019191484307581075,
      "loss": 0.1803973764181137,
      "mean_token_accuracy": 0.9253678023815155,
      "num_tokens": 26558380.0,
      "step": 2153
    },
    {
      "entropy": 1.1454325318336487,
      "epoch": 1.1342812006319116,
      "grad_norm": 0.3126206696033478,
      "learning_rate": 0.00019190575693741794,
      "loss": 0.20110175013542175,
      "mean_token_accuracy": 0.9191780984401703,
      "num_tokens": 26570716.0,
      "step": 2154
    },
    {
      "entropy": 1.1290227174758911,
      "epoch": 1.134807793575566,
      "grad_norm": 0.2912094295024872,
      "learning_rate": 0.0001918966659367577,
      "loss": 0.19985133409500122,
      "mean_token_accuracy": 0.9185344576835632,
      "num_tokens": 26583052.0,
      "step": 2155
    },
    {
      "entropy": 1.1607117354869843,
      "epoch": 1.1353343865192207,
      "grad_norm": 0.3160911202430725,
      "learning_rate": 0.00019188757007436978,
      "loss": 0.21099629998207092,
      "mean_token_accuracy": 0.914495974779129,
      "num_tokens": 26595388.0,
      "step": 2156
    },
    {
      "entropy": 1.1428423523902893,
      "epoch": 1.1358609794628751,
      "grad_norm": 0.30592912435531616,
      "learning_rate": 0.00019187846935079405,
      "loss": 0.20286978781223297,
      "mean_token_accuracy": 0.919879361987114,
      "num_tokens": 26607724.0,
      "step": 2157
    },
    {
      "entropy": 1.1966476142406464,
      "epoch": 1.1363875724065298,
      "grad_norm": 0.305442750453949,
      "learning_rate": 0.00019186936376657085,
      "loss": 0.19738367199897766,
      "mean_token_accuracy": 0.9184410721063614,
      "num_tokens": 26620060.0,
      "step": 2158
    },
    {
      "entropy": 1.1541268825531006,
      "epoch": 1.1369141653501842,
      "grad_norm": 0.32751592993736267,
      "learning_rate": 0.00019186025332224068,
      "loss": 0.22538770735263824,
      "mean_token_accuracy": 0.9071175456047058,
      "num_tokens": 26632396.0,
      "step": 2159
    },
    {
      "entropy": 1.1819856464862823,
      "epoch": 1.1374407582938388,
      "grad_norm": 0.3354586958885193,
      "learning_rate": 0.00019185113801834438,
      "loss": 0.2322450429201126,
      "mean_token_accuracy": 0.9149561524391174,
      "num_tokens": 26644732.0,
      "step": 2160
    },
    {
      "entropy": 1.2278803288936615,
      "epoch": 1.1379673512374935,
      "grad_norm": 0.39081871509552,
      "learning_rate": 0.0001918420178554231,
      "loss": 0.2215672731399536,
      "mean_token_accuracy": 0.9058908820152283,
      "num_tokens": 26657068.0,
      "step": 2161
    },
    {
      "entropy": 1.1993200778961182,
      "epoch": 1.138493944181148,
      "grad_norm": 0.30215588212013245,
      "learning_rate": 0.00019183289283401821,
      "loss": 0.21090206503868103,
      "mean_token_accuracy": 0.9153352379798889,
      "num_tokens": 26669404.0,
      "step": 2162
    },
    {
      "entropy": 1.2832166850566864,
      "epoch": 1.1390205371248026,
      "grad_norm": 0.28006455302238464,
      "learning_rate": 0.00019182376295467143,
      "loss": 0.19265320897102356,
      "mean_token_accuracy": 0.9220689386129379,
      "num_tokens": 26681740.0,
      "step": 2163
    },
    {
      "entropy": 1.2227767407894135,
      "epoch": 1.1395471300684572,
      "grad_norm": 0.2612974941730499,
      "learning_rate": 0.0001918146282179248,
      "loss": 0.19369927048683167,
      "mean_token_accuracy": 0.9246975183486938,
      "num_tokens": 26694076.0,
      "step": 2164
    },
    {
      "entropy": 1.2275767028331757,
      "epoch": 1.1400737230121116,
      "grad_norm": 0.2888461649417877,
      "learning_rate": 0.00019180548862432047,
      "loss": 0.22057251632213593,
      "mean_token_accuracy": 0.9090530127286911,
      "num_tokens": 26706412.0,
      "step": 2165
    },
    {
      "entropy": 1.3238163888454437,
      "epoch": 1.1406003159557663,
      "grad_norm": 0.2778269350528717,
      "learning_rate": 0.0001917963441744011,
      "loss": 0.21366479992866516,
      "mean_token_accuracy": 0.9129827171564102,
      "num_tokens": 26718748.0,
      "step": 2166
    },
    {
      "entropy": 1.2234577536582947,
      "epoch": 1.1411269088994207,
      "grad_norm": 0.2672150135040283,
      "learning_rate": 0.0001917871948687096,
      "loss": 0.20926086604595184,
      "mean_token_accuracy": 0.9146319776773453,
      "num_tokens": 26731084.0,
      "step": 2167
    },
    {
      "entropy": 1.285551905632019,
      "epoch": 1.1416535018430753,
      "grad_norm": 0.2621254026889801,
      "learning_rate": 0.00019177804070778893,
      "loss": 0.19555458426475525,
      "mean_token_accuracy": 0.9218839704990387,
      "num_tokens": 26743420.0,
      "step": 2168
    },
    {
      "entropy": 1.2906259596347809,
      "epoch": 1.1421800947867298,
      "grad_norm": 0.2878972291946411,
      "learning_rate": 0.00019176888169218265,
      "loss": 0.20401382446289062,
      "mean_token_accuracy": 0.9181995242834091,
      "num_tokens": 26755756.0,
      "step": 2169
    },
    {
      "entropy": 1.2635590136051178,
      "epoch": 1.1427066877303844,
      "grad_norm": 0.27613574266433716,
      "learning_rate": 0.0001917597178224345,
      "loss": 0.1996793895959854,
      "mean_token_accuracy": 0.9244147539138794,
      "num_tokens": 26768092.0,
      "step": 2170
    },
    {
      "entropy": 1.2653245031833649,
      "epoch": 1.143233280674039,
      "grad_norm": 0.26447412371635437,
      "learning_rate": 0.00019175054909908838,
      "loss": 0.19309335947036743,
      "mean_token_accuracy": 0.9219875186681747,
      "num_tokens": 26780428.0,
      "step": 2171
    },
    {
      "entropy": 1.2837787866592407,
      "epoch": 1.1437598736176935,
      "grad_norm": 0.27839452028274536,
      "learning_rate": 0.00019174137552268865,
      "loss": 0.19706347584724426,
      "mean_token_accuracy": 0.9179331064224243,
      "num_tokens": 26792764.0,
      "step": 2172
    },
    {
      "entropy": 1.284653127193451,
      "epoch": 1.1442864665613481,
      "grad_norm": 0.31177014112472534,
      "learning_rate": 0.0001917321970937799,
      "loss": 0.20646242797374725,
      "mean_token_accuracy": 0.9135401993989944,
      "num_tokens": 26805100.0,
      "step": 2173
    },
    {
      "entropy": 1.2473564743995667,
      "epoch": 1.1448130595050026,
      "grad_norm": 0.29228970408439636,
      "learning_rate": 0.00019172301381290696,
      "loss": 0.19468794763088226,
      "mean_token_accuracy": 0.9253511428833008,
      "num_tokens": 26817436.0,
      "step": 2174
    },
    {
      "entropy": 1.2133914828300476,
      "epoch": 1.1453396524486572,
      "grad_norm": 0.27723586559295654,
      "learning_rate": 0.000191713825680615,
      "loss": 0.19811224937438965,
      "mean_token_accuracy": 0.9239038378000259,
      "num_tokens": 26829772.0,
      "step": 2175
    },
    {
      "entropy": 1.2187740504741669,
      "epoch": 1.1458662453923116,
      "grad_norm": 0.32399874925613403,
      "learning_rate": 0.0001917046326974495,
      "loss": 0.21336714923381805,
      "mean_token_accuracy": 0.9163462072610855,
      "num_tokens": 26842108.0,
      "step": 2176
    },
    {
      "entropy": 1.1896620392799377,
      "epoch": 1.1463928383359663,
      "grad_norm": 0.2760239541530609,
      "learning_rate": 0.00019169543486395612,
      "loss": 0.19351127743721008,
      "mean_token_accuracy": 0.920502170920372,
      "num_tokens": 26854444.0,
      "step": 2177
    },
    {
      "entropy": 1.3020657300949097,
      "epoch": 1.146919431279621,
      "grad_norm": 0.33600714802742004,
      "learning_rate": 0.00019168623218068098,
      "loss": 0.19245027005672455,
      "mean_token_accuracy": 0.9246270656585693,
      "num_tokens": 26866780.0,
      "step": 2178
    },
    {
      "entropy": 1.2935131192207336,
      "epoch": 1.1474460242232754,
      "grad_norm": 0.28109049797058105,
      "learning_rate": 0.00019167702464817032,
      "loss": 0.18623016774654388,
      "mean_token_accuracy": 0.9251067191362381,
      "num_tokens": 26879116.0,
      "step": 2179
    },
    {
      "entropy": 1.2790460288524628,
      "epoch": 1.14797261716693,
      "grad_norm": 0.3010100722312927,
      "learning_rate": 0.00019166781226697077,
      "loss": 0.20548678934574127,
      "mean_token_accuracy": 0.9172774851322174,
      "num_tokens": 26891452.0,
      "step": 2180
    },
    {
      "entropy": 1.3008805811405182,
      "epoch": 1.1484992101105844,
      "grad_norm": 0.32880863547325134,
      "learning_rate": 0.00019165859503762918,
      "loss": 0.19846193492412567,
      "mean_token_accuracy": 0.9145328849554062,
      "num_tokens": 26903788.0,
      "step": 2181
    },
    {
      "entropy": 1.2794018983840942,
      "epoch": 1.149025803054239,
      "grad_norm": 0.3064977526664734,
      "learning_rate": 0.00019164937296069275,
      "loss": 0.20974700152873993,
      "mean_token_accuracy": 0.9109003394842148,
      "num_tokens": 26916124.0,
      "step": 2182
    },
    {
      "entropy": 1.2440880835056305,
      "epoch": 1.1495523959978937,
      "grad_norm": 0.27806052565574646,
      "learning_rate": 0.00019164014603670896,
      "loss": 0.19501592218875885,
      "mean_token_accuracy": 0.9204808473587036,
      "num_tokens": 26928460.0,
      "step": 2183
    },
    {
      "entropy": 1.3518921732902527,
      "epoch": 1.1500789889415481,
      "grad_norm": 0.321502149105072,
      "learning_rate": 0.0001916309142662255,
      "loss": 0.22053658962249756,
      "mean_token_accuracy": 0.9092803299427032,
      "num_tokens": 26940796.0,
      "step": 2184
    },
    {
      "entropy": 1.2679640352725983,
      "epoch": 1.1506055818852028,
      "grad_norm": 0.2918587625026703,
      "learning_rate": 0.00019162167764979048,
      "loss": 0.21362480521202087,
      "mean_token_accuracy": 0.9154315143823624,
      "num_tokens": 26953132.0,
      "step": 2185
    },
    {
      "entropy": 1.3025008738040924,
      "epoch": 1.1511321748288572,
      "grad_norm": 0.38866928219795227,
      "learning_rate": 0.00019161243618795213,
      "loss": 0.27079999446868896,
      "mean_token_accuracy": 0.896030992269516,
      "num_tokens": 26965468.0,
      "step": 2186
    },
    {
      "entropy": 1.2515160739421844,
      "epoch": 1.1516587677725119,
      "grad_norm": 0.2795846462249756,
      "learning_rate": 0.00019160318988125917,
      "loss": 0.20014791190624237,
      "mean_token_accuracy": 0.9226631373167038,
      "num_tokens": 26977804.0,
      "step": 2187
    },
    {
      "entropy": 1.2235470414161682,
      "epoch": 1.1521853607161665,
      "grad_norm": 0.27590784430503845,
      "learning_rate": 0.0001915939387302604,
      "loss": 0.1861528754234314,
      "mean_token_accuracy": 0.9201293885707855,
      "num_tokens": 26990140.0,
      "step": 2188
    },
    {
      "entropy": 1.222849279642105,
      "epoch": 1.152711953659821,
      "grad_norm": 0.2712383568286896,
      "learning_rate": 0.00019158468273550508,
      "loss": 0.19584113359451294,
      "mean_token_accuracy": 0.9204514920711517,
      "num_tokens": 27002476.0,
      "step": 2189
    },
    {
      "entropy": 1.2125034630298615,
      "epoch": 1.1532385466034756,
      "grad_norm": 0.2759631872177124,
      "learning_rate": 0.00019157542189754268,
      "loss": 0.19963940978050232,
      "mean_token_accuracy": 0.9220093786716461,
      "num_tokens": 27014812.0,
      "step": 2190
    },
    {
      "entropy": 1.1839116215705872,
      "epoch": 1.15376513954713,
      "grad_norm": 0.25730669498443604,
      "learning_rate": 0.00019156615621692292,
      "loss": 0.18340635299682617,
      "mean_token_accuracy": 0.9252117276191711,
      "num_tokens": 27027148.0,
      "step": 2191
    },
    {
      "entropy": 1.202635020017624,
      "epoch": 1.1542917324907846,
      "grad_norm": 0.2737926244735718,
      "learning_rate": 0.00019155688569419586,
      "loss": 0.206609845161438,
      "mean_token_accuracy": 0.9209761023521423,
      "num_tokens": 27039484.0,
      "step": 2192
    },
    {
      "entropy": 1.168206125497818,
      "epoch": 1.154818325434439,
      "grad_norm": 0.26999542117118835,
      "learning_rate": 0.00019154761032991185,
      "loss": 0.20011095702648163,
      "mean_token_accuracy": 0.9189373552799225,
      "num_tokens": 27051820.0,
      "step": 2193
    },
    {
      "entropy": 1.2162071764469147,
      "epoch": 1.1553449183780937,
      "grad_norm": 0.25214529037475586,
      "learning_rate": 0.00019153833012462148,
      "loss": 0.19554975628852844,
      "mean_token_accuracy": 0.9181893914937973,
      "num_tokens": 27064156.0,
      "step": 2194
    },
    {
      "entropy": 1.1848547458648682,
      "epoch": 1.1558715113217484,
      "grad_norm": 0.2536722719669342,
      "learning_rate": 0.00019152904507887573,
      "loss": 0.19084079563617706,
      "mean_token_accuracy": 0.9256320744752884,
      "num_tokens": 27076492.0,
      "step": 2195
    },
    {
      "entropy": 1.1943599879741669,
      "epoch": 1.1563981042654028,
      "grad_norm": 0.2896130084991455,
      "learning_rate": 0.00019151975519322576,
      "loss": 0.19871902465820312,
      "mean_token_accuracy": 0.9247763454914093,
      "num_tokens": 27088828.0,
      "step": 2196
    },
    {
      "entropy": 1.2219892144203186,
      "epoch": 1.1569246972090574,
      "grad_norm": 0.287672221660614,
      "learning_rate": 0.00019151046046822305,
      "loss": 0.20714130997657776,
      "mean_token_accuracy": 0.918072521686554,
      "num_tokens": 27101164.0,
      "step": 2197
    },
    {
      "entropy": 1.2384284734725952,
      "epoch": 1.1574512901527119,
      "grad_norm": 0.2936769425868988,
      "learning_rate": 0.00019150116090441938,
      "loss": 0.21063557267189026,
      "mean_token_accuracy": 0.9142559766769409,
      "num_tokens": 27113500.0,
      "step": 2198
    },
    {
      "entropy": 1.2583527266979218,
      "epoch": 1.1579778830963665,
      "grad_norm": 0.30406489968299866,
      "learning_rate": 0.00019149185650236682,
      "loss": 0.2245735228061676,
      "mean_token_accuracy": 0.9095106869935989,
      "num_tokens": 27125836.0,
      "step": 2199
    },
    {
      "entropy": 1.2499944269657135,
      "epoch": 1.1585044760400212,
      "grad_norm": 0.30430343747138977,
      "learning_rate": 0.0001914825472626177,
      "loss": 0.21744973957538605,
      "mean_token_accuracy": 0.9074316918849945,
      "num_tokens": 27138172.0,
      "step": 2200
    },
    {
      "entropy": 1.1596157252788544,
      "epoch": 1.1590310689836756,
      "grad_norm": 0.29872894287109375,
      "learning_rate": 0.00019147323318572466,
      "loss": 0.20363301038742065,
      "mean_token_accuracy": 0.9211505800485611,
      "num_tokens": 27150508.0,
      "step": 2201
    },
    {
      "entropy": 1.2048721611499786,
      "epoch": 1.1595576619273302,
      "grad_norm": 0.28121620416641235,
      "learning_rate": 0.00019146391427224063,
      "loss": 0.20417368412017822,
      "mean_token_accuracy": 0.9163524359464645,
      "num_tokens": 27162844.0,
      "step": 2202
    },
    {
      "entropy": 1.2005835473537445,
      "epoch": 1.1600842548709847,
      "grad_norm": 0.2875327169895172,
      "learning_rate": 0.00019145459052271883,
      "loss": 0.1831551492214203,
      "mean_token_accuracy": 0.9263263195753098,
      "num_tokens": 27175180.0,
      "step": 2203
    },
    {
      "entropy": 1.1740988194942474,
      "epoch": 1.1606108478146393,
      "grad_norm": 0.30328717827796936,
      "learning_rate": 0.00019144526193771275,
      "loss": 0.2115185409784317,
      "mean_token_accuracy": 0.9105354696512222,
      "num_tokens": 27187516.0,
      "step": 2204
    },
    {
      "entropy": 1.1883420646190643,
      "epoch": 1.161137440758294,
      "grad_norm": 0.3002709448337555,
      "learning_rate": 0.00019143592851777617,
      "loss": 0.1997799128293991,
      "mean_token_accuracy": 0.9204158633947372,
      "num_tokens": 27199852.0,
      "step": 2205
    },
    {
      "entropy": 1.186734914779663,
      "epoch": 1.1616640337019484,
      "grad_norm": 0.2615937888622284,
      "learning_rate": 0.00019142659026346315,
      "loss": 0.20187148451805115,
      "mean_token_accuracy": 0.9159397482872009,
      "num_tokens": 27212188.0,
      "step": 2206
    },
    {
      "entropy": 1.1872017085552216,
      "epoch": 1.162190626645603,
      "grad_norm": 0.26249435544013977,
      "learning_rate": 0.00019141724717532802,
      "loss": 0.18397313356399536,
      "mean_token_accuracy": 0.9237561523914337,
      "num_tokens": 27224524.0,
      "step": 2207
    },
    {
      "entropy": 1.1571813225746155,
      "epoch": 1.1627172195892574,
      "grad_norm": 0.27942758798599243,
      "learning_rate": 0.0001914078992539255,
      "loss": 0.18515753746032715,
      "mean_token_accuracy": 0.9259720593690872,
      "num_tokens": 27236860.0,
      "step": 2208
    },
    {
      "entropy": 1.1432051360607147,
      "epoch": 1.163243812532912,
      "grad_norm": 0.2963655889034271,
      "learning_rate": 0.00019139854649981048,
      "loss": 0.21573269367218018,
      "mean_token_accuracy": 0.9100171327590942,
      "num_tokens": 27249196.0,
      "step": 2209
    },
    {
      "entropy": 1.188007116317749,
      "epoch": 1.1637704054765665,
      "grad_norm": 0.31860122084617615,
      "learning_rate": 0.00019138918891353817,
      "loss": 0.2418171912431717,
      "mean_token_accuracy": 0.9069865345954895,
      "num_tokens": 27261532.0,
      "step": 2210
    },
    {
      "entropy": 1.1004420518875122,
      "epoch": 1.1642969984202212,
      "grad_norm": 0.2776052951812744,
      "learning_rate": 0.00019137982649566407,
      "loss": 0.2053711712360382,
      "mean_token_accuracy": 0.9153793305158615,
      "num_tokens": 27273868.0,
      "step": 2211
    },
    {
      "entropy": 1.1410734951496124,
      "epoch": 1.1648235913638758,
      "grad_norm": 0.28460225462913513,
      "learning_rate": 0.00019137045924674402,
      "loss": 0.21728798747062683,
      "mean_token_accuracy": 0.9123410433530807,
      "num_tokens": 27286204.0,
      "step": 2212
    },
    {
      "entropy": 1.1587063372135162,
      "epoch": 1.1653501843075302,
      "grad_norm": 0.2642859220504761,
      "learning_rate": 0.0001913610871673341,
      "loss": 0.18721605837345123,
      "mean_token_accuracy": 0.9224630892276764,
      "num_tokens": 27298540.0,
      "step": 2213
    },
    {
      "entropy": 1.1271196007728577,
      "epoch": 1.1658767772511849,
      "grad_norm": 0.28819388151168823,
      "learning_rate": 0.00019135171025799056,
      "loss": 0.18717646598815918,
      "mean_token_accuracy": 0.9211658984422684,
      "num_tokens": 27310876.0,
      "step": 2214
    },
    {
      "entropy": 1.1337847411632538,
      "epoch": 1.1664033701948393,
      "grad_norm": 0.2941300570964813,
      "learning_rate": 0.0001913423285192702,
      "loss": 0.21211886405944824,
      "mean_token_accuracy": 0.9153169095516205,
      "num_tokens": 27323212.0,
      "step": 2215
    },
    {
      "entropy": 1.1438775956630707,
      "epoch": 1.166929963138494,
      "grad_norm": 0.2863110303878784,
      "learning_rate": 0.0001913329419517299,
      "loss": 0.2077072560787201,
      "mean_token_accuracy": 0.9118431657552719,
      "num_tokens": 27335548.0,
      "step": 2216
    },
    {
      "entropy": 1.1338017284870148,
      "epoch": 1.1674565560821484,
      "grad_norm": 0.2920835316181183,
      "learning_rate": 0.0001913235505559268,
      "loss": 0.19436489045619965,
      "mean_token_accuracy": 0.9221871793270111,
      "num_tokens": 27347884.0,
      "step": 2217
    },
    {
      "entropy": 1.142452985048294,
      "epoch": 1.167983149025803,
      "grad_norm": 0.26582619547843933,
      "learning_rate": 0.00019131415433241855,
      "loss": 0.19146628677845,
      "mean_token_accuracy": 0.9218153953552246,
      "num_tokens": 27360220.0,
      "step": 2218
    },
    {
      "entropy": 1.1075855195522308,
      "epoch": 1.1685097419694577,
      "grad_norm": 0.2855731248855591,
      "learning_rate": 0.0001913047532817629,
      "loss": 0.20022396743297577,
      "mean_token_accuracy": 0.919172614812851,
      "num_tokens": 27372556.0,
      "step": 2219
    },
    {
      "entropy": 1.161133885383606,
      "epoch": 1.169036334913112,
      "grad_norm": 0.2820180356502533,
      "learning_rate": 0.0001912953474045179,
      "loss": 0.2025647759437561,
      "mean_token_accuracy": 0.9174589961767197,
      "num_tokens": 27384892.0,
      "step": 2220
    },
    {
      "entropy": 1.145233541727066,
      "epoch": 1.1695629278567667,
      "grad_norm": 0.2780614495277405,
      "learning_rate": 0.00019128593670124198,
      "loss": 0.20629477500915527,
      "mean_token_accuracy": 0.9178284406661987,
      "num_tokens": 27397228.0,
      "step": 2221
    },
    {
      "entropy": 1.1302664279937744,
      "epoch": 1.1700895208004214,
      "grad_norm": 0.28989875316619873,
      "learning_rate": 0.00019127652117249374,
      "loss": 0.20143242180347443,
      "mean_token_accuracy": 0.9146946966648102,
      "num_tokens": 27409564.0,
      "step": 2222
    },
    {
      "entropy": 1.145494520664215,
      "epoch": 1.1706161137440758,
      "grad_norm": 0.280318945646286,
      "learning_rate": 0.0001912671008188322,
      "loss": 0.18556565046310425,
      "mean_token_accuracy": 0.9243800044059753,
      "num_tokens": 27421900.0,
      "step": 2223
    },
    {
      "entropy": 1.122549831867218,
      "epoch": 1.1711427066877305,
      "grad_norm": 0.28058454394340515,
      "learning_rate": 0.0001912576756408165,
      "loss": 0.19919465482234955,
      "mean_token_accuracy": 0.9234228730201721,
      "num_tokens": 27434236.0,
      "step": 2224
    },
    {
      "entropy": 1.1942664980888367,
      "epoch": 1.1716692996313849,
      "grad_norm": 0.3011435866355896,
      "learning_rate": 0.00019124824563900622,
      "loss": 0.2100905328989029,
      "mean_token_accuracy": 0.9161610305309296,
      "num_tokens": 27446572.0,
      "step": 2225
    },
    {
      "entropy": 1.1235324144363403,
      "epoch": 1.1721958925750395,
      "grad_norm": 0.2678482234477997,
      "learning_rate": 0.00019123881081396113,
      "loss": 0.18935970962047577,
      "mean_token_accuracy": 0.9210453182458878,
      "num_tokens": 27458908.0,
      "step": 2226
    },
    {
      "entropy": 1.1233053505420685,
      "epoch": 1.172722485518694,
      "grad_norm": 0.3001168966293335,
      "learning_rate": 0.00019122937116624133,
      "loss": 0.1962110847234726,
      "mean_token_accuracy": 0.9216617047786713,
      "num_tokens": 27471244.0,
      "step": 2227
    },
    {
      "entropy": 1.1534164547920227,
      "epoch": 1.1732490784623486,
      "grad_norm": 0.3095707893371582,
      "learning_rate": 0.00019121992669640726,
      "loss": 0.2072276473045349,
      "mean_token_accuracy": 0.9137547612190247,
      "num_tokens": 27483580.0,
      "step": 2228
    },
    {
      "entropy": 1.1193904280662537,
      "epoch": 1.1737756714060033,
      "grad_norm": 0.26677873730659485,
      "learning_rate": 0.00019121047740501947,
      "loss": 0.20390596985816956,
      "mean_token_accuracy": 0.9185405522584915,
      "num_tokens": 27495916.0,
      "step": 2229
    },
    {
      "entropy": 1.1349280178546906,
      "epoch": 1.1743022643496577,
      "grad_norm": 0.2756739854812622,
      "learning_rate": 0.000191201023292639,
      "loss": 0.2036488950252533,
      "mean_token_accuracy": 0.9196960180997849,
      "num_tokens": 27508252.0,
      "step": 2230
    },
    {
      "entropy": 1.1713554561138153,
      "epoch": 1.1748288572933123,
      "grad_norm": 0.30103445053100586,
      "learning_rate": 0.00019119156435982703,
      "loss": 0.20072484016418457,
      "mean_token_accuracy": 0.925217941403389,
      "num_tokens": 27520588.0,
      "step": 2231
    },
    {
      "entropy": 1.1370952129364014,
      "epoch": 1.1753554502369667,
      "grad_norm": 0.32296401262283325,
      "learning_rate": 0.00019118210060714512,
      "loss": 0.23248368501663208,
      "mean_token_accuracy": 0.9066440016031265,
      "num_tokens": 27532924.0,
      "step": 2232
    },
    {
      "entropy": 1.152237057685852,
      "epoch": 1.1758820431806214,
      "grad_norm": 0.2729482650756836,
      "learning_rate": 0.00019117263203515504,
      "loss": 0.19771021604537964,
      "mean_token_accuracy": 0.9137565195560455,
      "num_tokens": 27545260.0,
      "step": 2233
    },
    {
      "entropy": 1.175236165523529,
      "epoch": 1.1764086361242758,
      "grad_norm": 0.27823638916015625,
      "learning_rate": 0.00019116315864441897,
      "loss": 0.20588554441928864,
      "mean_token_accuracy": 0.9182644933462143,
      "num_tokens": 27557596.0,
      "step": 2234
    },
    {
      "entropy": 1.19709312915802,
      "epoch": 1.1769352290679305,
      "grad_norm": 0.2955763041973114,
      "learning_rate": 0.00019115368043549916,
      "loss": 0.1897638589143753,
      "mean_token_accuracy": 0.9216989874839783,
      "num_tokens": 27569932.0,
      "step": 2235
    },
    {
      "entropy": 1.2332687377929688,
      "epoch": 1.1774618220115851,
      "grad_norm": 0.3177914023399353,
      "learning_rate": 0.00019114419740895837,
      "loss": 0.23029568791389465,
      "mean_token_accuracy": 0.906041145324707,
      "num_tokens": 27582268.0,
      "step": 2236
    },
    {
      "entropy": 1.1908305287361145,
      "epoch": 1.1779884149552395,
      "grad_norm": 0.2913646996021271,
      "learning_rate": 0.00019113470956535952,
      "loss": 0.20150071382522583,
      "mean_token_accuracy": 0.9204053580760956,
      "num_tokens": 27594604.0,
      "step": 2237
    },
    {
      "entropy": 1.2038147449493408,
      "epoch": 1.1785150078988942,
      "grad_norm": 0.28544673323631287,
      "learning_rate": 0.00019112521690526583,
      "loss": 0.18555422127246857,
      "mean_token_accuracy": 0.924328088760376,
      "num_tokens": 27606940.0,
      "step": 2238
    },
    {
      "entropy": 1.2000732123851776,
      "epoch": 1.1790416008425488,
      "grad_norm": 0.2675821781158447,
      "learning_rate": 0.00019111571942924085,
      "loss": 0.19991934299468994,
      "mean_token_accuracy": 0.9188268631696701,
      "num_tokens": 27619276.0,
      "step": 2239
    },
    {
      "entropy": 1.206215888261795,
      "epoch": 1.1795681937862033,
      "grad_norm": 0.2979394495487213,
      "learning_rate": 0.00019110621713784842,
      "loss": 0.19255416095256805,
      "mean_token_accuracy": 0.9227475672960281,
      "num_tokens": 27631612.0,
      "step": 2240
    },
    {
      "entropy": 1.1973096132278442,
      "epoch": 1.180094786729858,
      "grad_norm": 0.27704933285713196,
      "learning_rate": 0.0001910967100316526,
      "loss": 0.18796265125274658,
      "mean_token_accuracy": 0.9225670397281647,
      "num_tokens": 27643948.0,
      "step": 2241
    },
    {
      "entropy": 1.2093007564544678,
      "epoch": 1.1806213796735123,
      "grad_norm": 0.29337137937545776,
      "learning_rate": 0.00019108719811121772,
      "loss": 0.193947434425354,
      "mean_token_accuracy": 0.9231081902980804,
      "num_tokens": 27656284.0,
      "step": 2242
    },
    {
      "entropy": 1.2430245876312256,
      "epoch": 1.181147972617167,
      "grad_norm": 0.3377665579319,
      "learning_rate": 0.00019107768137710852,
      "loss": 0.21937604248523712,
      "mean_token_accuracy": 0.9147068560123444,
      "num_tokens": 27668620.0,
      "step": 2243
    },
    {
      "entropy": 1.247610479593277,
      "epoch": 1.1816745655608214,
      "grad_norm": 0.32605549693107605,
      "learning_rate": 0.00019106815982988996,
      "loss": 0.22798651456832886,
      "mean_token_accuracy": 0.9099006354808807,
      "num_tokens": 27680956.0,
      "step": 2244
    },
    {
      "entropy": 1.1888035833835602,
      "epoch": 1.182201158504476,
      "grad_norm": 0.30127960443496704,
      "learning_rate": 0.00019105863347012724,
      "loss": 0.21623851358890533,
      "mean_token_accuracy": 0.9116876125335693,
      "num_tokens": 27693292.0,
      "step": 2245
    },
    {
      "entropy": 1.1904528439044952,
      "epoch": 1.1827277514481307,
      "grad_norm": 0.26522204279899597,
      "learning_rate": 0.0001910491022983859,
      "loss": 0.19360768795013428,
      "mean_token_accuracy": 0.9212395548820496,
      "num_tokens": 27705628.0,
      "step": 2246
    },
    {
      "entropy": 1.190793126821518,
      "epoch": 1.1832543443917851,
      "grad_norm": 0.28089118003845215,
      "learning_rate": 0.00019103956631523177,
      "loss": 0.20212800800800323,
      "mean_token_accuracy": 0.918514221906662,
      "num_tokens": 27717964.0,
      "step": 2247
    },
    {
      "entropy": 1.1978855729103088,
      "epoch": 1.1837809373354398,
      "grad_norm": 0.29225775599479675,
      "learning_rate": 0.00019103002552123087,
      "loss": 0.21059700846672058,
      "mean_token_accuracy": 0.913848802447319,
      "num_tokens": 27730300.0,
      "step": 2248
    },
    {
      "entropy": 1.1586676836013794,
      "epoch": 1.1843075302790942,
      "grad_norm": 0.2872113287448883,
      "learning_rate": 0.00019102047991694966,
      "loss": 0.21480488777160645,
      "mean_token_accuracy": 0.9116514921188354,
      "num_tokens": 27742636.0,
      "step": 2249
    },
    {
      "entropy": 1.1448751091957092,
      "epoch": 1.1848341232227488,
      "grad_norm": 0.29005923867225647,
      "learning_rate": 0.00019101092950295478,
      "loss": 0.2099185287952423,
      "mean_token_accuracy": 0.9138695448637009,
      "num_tokens": 27754972.0,
      "step": 2250
    },
    {
      "entropy": 1.137921154499054,
      "epoch": 1.1853607161664033,
      "grad_norm": 0.27500849962234497,
      "learning_rate": 0.0001910013742798132,
      "loss": 0.1960453987121582,
      "mean_token_accuracy": 0.9165975153446198,
      "num_tokens": 27767308.0,
      "step": 2251
    },
    {
      "entropy": 1.0970045328140259,
      "epoch": 1.185887309110058,
      "grad_norm": 0.26834315061569214,
      "learning_rate": 0.0001909918142480921,
      "loss": 0.2065771222114563,
      "mean_token_accuracy": 0.9169502705335617,
      "num_tokens": 27779644.0,
      "step": 2252
    },
    {
      "entropy": 1.1106819808483124,
      "epoch": 1.1864139020537126,
      "grad_norm": 0.29660072922706604,
      "learning_rate": 0.0001909822494083591,
      "loss": 0.19244039058685303,
      "mean_token_accuracy": 0.9193838387727737,
      "num_tokens": 27791980.0,
      "step": 2253
    },
    {
      "entropy": 1.0947784334421158,
      "epoch": 1.186940494997367,
      "grad_norm": 0.3122023642063141,
      "learning_rate": 0.0001909726797611819,
      "loss": 0.2027546465396881,
      "mean_token_accuracy": 0.916048526763916,
      "num_tokens": 27804316.0,
      "step": 2254
    },
    {
      "entropy": 1.077293038368225,
      "epoch": 1.1874670879410216,
      "grad_norm": 0.29647305607795715,
      "learning_rate": 0.00019096310530712865,
      "loss": 0.19166618585586548,
      "mean_token_accuracy": 0.9190387576818466,
      "num_tokens": 27816652.0,
      "step": 2255
    },
    {
      "entropy": 1.0975450575351715,
      "epoch": 1.187993680884676,
      "grad_norm": 0.30608993768692017,
      "learning_rate": 0.00019095352604676775,
      "loss": 0.20418736338615417,
      "mean_token_accuracy": 0.9154408276081085,
      "num_tokens": 27828988.0,
      "step": 2256
    },
    {
      "entropy": 1.0858297049999237,
      "epoch": 1.1885202738283307,
      "grad_norm": 0.2871757447719574,
      "learning_rate": 0.00019094394198066784,
      "loss": 0.20662033557891846,
      "mean_token_accuracy": 0.9137272536754608,
      "num_tokens": 27841324.0,
      "step": 2257
    },
    {
      "entropy": 1.098552942276001,
      "epoch": 1.1890468667719853,
      "grad_norm": 0.31691616773605347,
      "learning_rate": 0.0001909343531093979,
      "loss": 0.19544214010238647,
      "mean_token_accuracy": 0.9170811921358109,
      "num_tokens": 27853660.0,
      "step": 2258
    },
    {
      "entropy": 1.0465008020401,
      "epoch": 1.1895734597156398,
      "grad_norm": 0.274053156375885,
      "learning_rate": 0.00019092475943352708,
      "loss": 0.18214793503284454,
      "mean_token_accuracy": 0.9284517616033554,
      "num_tokens": 27865996.0,
      "step": 2259
    },
    {
      "entropy": 1.0639517903327942,
      "epoch": 1.1901000526592944,
      "grad_norm": 0.3095509111881256,
      "learning_rate": 0.000190915160953625,
      "loss": 0.21801498532295227,
      "mean_token_accuracy": 0.9145664125680923,
      "num_tokens": 27878332.0,
      "step": 2260
    },
    {
      "entropy": 1.0542739927768707,
      "epoch": 1.1906266456029488,
      "grad_norm": 0.29980573058128357,
      "learning_rate": 0.0001909055576702614,
      "loss": 0.21734799444675446,
      "mean_token_accuracy": 0.9118731617927551,
      "num_tokens": 27890668.0,
      "step": 2261
    },
    {
      "entropy": 1.0719305872917175,
      "epoch": 1.1911532385466035,
      "grad_norm": 0.29808852076530457,
      "learning_rate": 0.00019089594958400638,
      "loss": 0.19167831540107727,
      "mean_token_accuracy": 0.9198271036148071,
      "num_tokens": 27903004.0,
      "step": 2262
    },
    {
      "entropy": 1.083087533712387,
      "epoch": 1.1916798314902581,
      "grad_norm": 0.2985793948173523,
      "learning_rate": 0.00019088633669543035,
      "loss": 0.20173481106758118,
      "mean_token_accuracy": 0.9191490411758423,
      "num_tokens": 27915340.0,
      "step": 2263
    },
    {
      "entropy": 1.0608322620391846,
      "epoch": 1.1922064244339126,
      "grad_norm": 0.3635084927082062,
      "learning_rate": 0.00019087671900510394,
      "loss": 0.19898994266986847,
      "mean_token_accuracy": 0.9207655042409897,
      "num_tokens": 27927676.0,
      "step": 2264
    },
    {
      "entropy": 1.0396868586540222,
      "epoch": 1.1927330173775672,
      "grad_norm": 0.2880086898803711,
      "learning_rate": 0.00019086709651359817,
      "loss": 0.2001173496246338,
      "mean_token_accuracy": 0.9208283871412277,
      "num_tokens": 27940012.0,
      "step": 2265
    },
    {
      "entropy": 1.0880854725837708,
      "epoch": 1.1932596103212216,
      "grad_norm": 0.30705323815345764,
      "learning_rate": 0.00019085746922148413,
      "loss": 0.22503894567489624,
      "mean_token_accuracy": 0.9097017347812653,
      "num_tokens": 27952348.0,
      "step": 2266
    },
    {
      "entropy": 1.0872746706008911,
      "epoch": 1.1937862032648763,
      "grad_norm": 0.27415189146995544,
      "learning_rate": 0.00019084783712933345,
      "loss": 0.19403532147407532,
      "mean_token_accuracy": 0.9209066331386566,
      "num_tokens": 27964684.0,
      "step": 2267
    },
    {
      "entropy": 1.0447344183921814,
      "epoch": 1.1943127962085307,
      "grad_norm": 0.2834930121898651,
      "learning_rate": 0.0001908382002377179,
      "loss": 0.2041846364736557,
      "mean_token_accuracy": 0.919536292552948,
      "num_tokens": 27977020.0,
      "step": 2268
    },
    {
      "entropy": 1.0546895861625671,
      "epoch": 1.1948393891521853,
      "grad_norm": 0.2582406997680664,
      "learning_rate": 0.00019082855854720955,
      "loss": 0.19313426315784454,
      "mean_token_accuracy": 0.920450747013092,
      "num_tokens": 27989356.0,
      "step": 2269
    },
    {
      "entropy": 1.1272867023944855,
      "epoch": 1.19536598209584,
      "grad_norm": 0.28488391637802124,
      "learning_rate": 0.0001908189120583808,
      "loss": 0.22456082701683044,
      "mean_token_accuracy": 0.9063926041126251,
      "num_tokens": 28001692.0,
      "step": 2270
    },
    {
      "entropy": 1.071615993976593,
      "epoch": 1.1958925750394944,
      "grad_norm": 0.262404203414917,
      "learning_rate": 0.00019080926077180425,
      "loss": 0.18905514478683472,
      "mean_token_accuracy": 0.9255506098270416,
      "num_tokens": 28014028.0,
      "step": 2271
    },
    {
      "entropy": 1.0979838967323303,
      "epoch": 1.196419167983149,
      "grad_norm": 0.2802709639072418,
      "learning_rate": 0.00019079960468805293,
      "loss": 0.1837414801120758,
      "mean_token_accuracy": 0.9260729998350143,
      "num_tokens": 28026364.0,
      "step": 2272
    },
    {
      "entropy": 1.0705920308828354,
      "epoch": 1.1969457609268035,
      "grad_norm": 0.2867828607559204,
      "learning_rate": 0.00019078994380769998,
      "loss": 0.2067106068134308,
      "mean_token_accuracy": 0.9144153445959091,
      "num_tokens": 28038700.0,
      "step": 2273
    },
    {
      "entropy": 1.0631752610206604,
      "epoch": 1.1974723538704581,
      "grad_norm": 0.2956964075565338,
      "learning_rate": 0.000190780278131319,
      "loss": 0.22864574193954468,
      "mean_token_accuracy": 0.9114911556243896,
      "num_tokens": 28051036.0,
      "step": 2274
    },
    {
      "entropy": 1.0306233167648315,
      "epoch": 1.1979989468141128,
      "grad_norm": 0.27961382269859314,
      "learning_rate": 0.00019077060765948368,
      "loss": 0.2214936763048172,
      "mean_token_accuracy": 0.910724937915802,
      "num_tokens": 28063372.0,
      "step": 2275
    },
    {
      "entropy": 1.0444636046886444,
      "epoch": 1.1985255397577672,
      "grad_norm": 0.2704748511314392,
      "learning_rate": 0.00019076093239276813,
      "loss": 0.1987978219985962,
      "mean_token_accuracy": 0.9210948199033737,
      "num_tokens": 28075708.0,
      "step": 2276
    },
    {
      "entropy": 1.0565994679927826,
      "epoch": 1.1990521327014219,
      "grad_norm": 0.2979368269443512,
      "learning_rate": 0.00019075125233174674,
      "loss": 0.20775671303272247,
      "mean_token_accuracy": 0.915656104683876,
      "num_tokens": 28088044.0,
      "step": 2277
    },
    {
      "entropy": 1.0443108528852463,
      "epoch": 1.1995787256450763,
      "grad_norm": 0.28842073678970337,
      "learning_rate": 0.0001907415674769942,
      "loss": 0.2142159640789032,
      "mean_token_accuracy": 0.9145304262638092,
      "num_tokens": 28100380.0,
      "step": 2278
    },
    {
      "entropy": 1.0332957953214645,
      "epoch": 1.200105318588731,
      "grad_norm": 0.2912449538707733,
      "learning_rate": 0.00019073187782908538,
      "loss": 0.20978567004203796,
      "mean_token_accuracy": 0.9145148694515228,
      "num_tokens": 28112716.0,
      "step": 2279
    },
    {
      "entropy": 1.0637232959270477,
      "epoch": 1.2006319115323856,
      "grad_norm": 0.27986404299736023,
      "learning_rate": 0.0001907221833885955,
      "loss": 0.20642036199569702,
      "mean_token_accuracy": 0.9175934493541718,
      "num_tokens": 28125052.0,
      "step": 2280
    },
    {
      "entropy": 1.0875510275363922,
      "epoch": 1.20115850447604,
      "grad_norm": 0.2697252333164215,
      "learning_rate": 0.00019071248415610008,
      "loss": 0.17736025154590607,
      "mean_token_accuracy": 0.9242502599954605,
      "num_tokens": 28137388.0,
      "step": 2281
    },
    {
      "entropy": 1.0523739904165268,
      "epoch": 1.2016850974196946,
      "grad_norm": 0.28700074553489685,
      "learning_rate": 0.0001907027801321749,
      "loss": 0.19527116417884827,
      "mean_token_accuracy": 0.9207979440689087,
      "num_tokens": 28149724.0,
      "step": 2282
    },
    {
      "entropy": 1.0521668642759323,
      "epoch": 1.202211690363349,
      "grad_norm": 0.33191800117492676,
      "learning_rate": 0.00019069307131739608,
      "loss": 0.20707619190216064,
      "mean_token_accuracy": 0.9119413197040558,
      "num_tokens": 28162060.0,
      "step": 2283
    },
    {
      "entropy": 1.0997390151023865,
      "epoch": 1.2027382833070037,
      "grad_norm": 0.2936375141143799,
      "learning_rate": 0.00019068335771233987,
      "loss": 0.1859363317489624,
      "mean_token_accuracy": 0.925885409116745,
      "num_tokens": 28174396.0,
      "step": 2284
    },
    {
      "entropy": 1.0267358273267746,
      "epoch": 1.2032648762506581,
      "grad_norm": 0.31446945667266846,
      "learning_rate": 0.000190673639317583,
      "loss": 0.20978572964668274,
      "mean_token_accuracy": 0.9154664278030396,
      "num_tokens": 28186732.0,
      "step": 2285
    },
    {
      "entropy": 1.0447548627853394,
      "epoch": 1.2037914691943128,
      "grad_norm": 0.26753443479537964,
      "learning_rate": 0.0001906639161337024,
      "loss": 0.18991108238697052,
      "mean_token_accuracy": 0.9253087639808655,
      "num_tokens": 28199068.0,
      "step": 2286
    },
    {
      "entropy": 1.019596055150032,
      "epoch": 1.2043180621379674,
      "grad_norm": 0.2849011719226837,
      "learning_rate": 0.00019065418816127517,
      "loss": 0.19186937808990479,
      "mean_token_accuracy": 0.9237010776996613,
      "num_tokens": 28211404.0,
      "step": 2287
    },
    {
      "entropy": 1.0712246000766754,
      "epoch": 1.2048446550816219,
      "grad_norm": 0.3231803774833679,
      "learning_rate": 0.0001906444554008789,
      "loss": 0.21262578666210175,
      "mean_token_accuracy": 0.9131361246109009,
      "num_tokens": 28223740.0,
      "step": 2288
    },
    {
      "entropy": 1.0217514485120773,
      "epoch": 1.2053712480252765,
      "grad_norm": 0.27426403760910034,
      "learning_rate": 0.00019063471785309136,
      "loss": 0.19608943164348602,
      "mean_token_accuracy": 0.9183037132024765,
      "num_tokens": 28236076.0,
      "step": 2289
    },
    {
      "entropy": 1.0533497035503387,
      "epoch": 1.205897840968931,
      "grad_norm": 0.31602412462234497,
      "learning_rate": 0.0001906249755184906,
      "loss": 0.20852188766002655,
      "mean_token_accuracy": 0.9144260287284851,
      "num_tokens": 28248412.0,
      "step": 2290
    },
    {
      "entropy": 1.0375554859638214,
      "epoch": 1.2064244339125856,
      "grad_norm": 0.2901093065738678,
      "learning_rate": 0.00019061522839765495,
      "loss": 0.20651787519454956,
      "mean_token_accuracy": 0.9151993989944458,
      "num_tokens": 28260748.0,
      "step": 2291
    },
    {
      "entropy": 1.047476589679718,
      "epoch": 1.20695102685624,
      "grad_norm": 0.28905728459358215,
      "learning_rate": 0.00019060547649116304,
      "loss": 0.18046444654464722,
      "mean_token_accuracy": 0.924500435590744,
      "num_tokens": 28273084.0,
      "step": 2292
    },
    {
      "entropy": 1.0409999787807465,
      "epoch": 1.2074776197998947,
      "grad_norm": 0.26790550351142883,
      "learning_rate": 0.00019059571979959382,
      "loss": 0.182978093624115,
      "mean_token_accuracy": 0.9302794337272644,
      "num_tokens": 28285420.0,
      "step": 2293
    },
    {
      "entropy": 0.9823900163173676,
      "epoch": 1.2080042127435493,
      "grad_norm": 0.29670295119285583,
      "learning_rate": 0.00019058595832352647,
      "loss": 0.19820117950439453,
      "mean_token_accuracy": 0.9207546710968018,
      "num_tokens": 28297756.0,
      "step": 2294
    },
    {
      "entropy": 1.0482214242219925,
      "epoch": 1.2085308056872037,
      "grad_norm": 0.2754944860935211,
      "learning_rate": 0.00019057619206354042,
      "loss": 0.1841028481721878,
      "mean_token_accuracy": 0.9217727184295654,
      "num_tokens": 28310092.0,
      "step": 2295
    },
    {
      "entropy": 0.979021742939949,
      "epoch": 1.2090573986308584,
      "grad_norm": 0.2850244641304016,
      "learning_rate": 0.00019056642102021555,
      "loss": 0.19173939526081085,
      "mean_token_accuracy": 0.920306608080864,
      "num_tokens": 28322428.0,
      "step": 2296
    },
    {
      "entropy": 1.0073926001787186,
      "epoch": 1.209583991574513,
      "grad_norm": 0.2759486734867096,
      "learning_rate": 0.00019055664519413177,
      "loss": 0.19916388392448425,
      "mean_token_accuracy": 0.9175459295511246,
      "num_tokens": 28334764.0,
      "step": 2297
    },
    {
      "entropy": 0.9477061033248901,
      "epoch": 1.2101105845181674,
      "grad_norm": 0.3503093719482422,
      "learning_rate": 0.00019054686458586952,
      "loss": 0.20742762088775635,
      "mean_token_accuracy": 0.9118971973657608,
      "num_tokens": 28347100.0,
      "step": 2298
    },
    {
      "entropy": 1.0388128459453583,
      "epoch": 1.210637177461822,
      "grad_norm": 0.3258967697620392,
      "learning_rate": 0.00019053707919600938,
      "loss": 0.22573794424533844,
      "mean_token_accuracy": 0.9091981053352356,
      "num_tokens": 28359436.0,
      "step": 2299
    },
    {
      "entropy": 0.9342583864927292,
      "epoch": 1.2111637704054765,
      "grad_norm": 0.2741185426712036,
      "learning_rate": 0.00019052728902513224,
      "loss": 0.1891242414712906,
      "mean_token_accuracy": 0.920941025018692,
      "num_tokens": 28371772.0,
      "step": 2300
    },
    {
      "entropy": 0.9953952133655548,
      "epoch": 1.2116903633491312,
      "grad_norm": 0.29077091813087463,
      "learning_rate": 0.00019051749407381934,
      "loss": 0.2002788484096527,
      "mean_token_accuracy": 0.9163627922534943,
      "num_tokens": 28384108.0,
      "step": 2301
    },
    {
      "entropy": 0.9766784757375717,
      "epoch": 1.2122169562927856,
      "grad_norm": 0.301318496465683,
      "learning_rate": 0.00019050769434265206,
      "loss": 0.20140445232391357,
      "mean_token_accuracy": 0.9168388992547989,
      "num_tokens": 28396444.0,
      "step": 2302
    },
    {
      "entropy": 0.9402145892381668,
      "epoch": 1.2127435492364402,
      "grad_norm": 0.2892124056816101,
      "learning_rate": 0.00019049788983221223,
      "loss": 0.19553864002227783,
      "mean_token_accuracy": 0.9176475703716278,
      "num_tokens": 28408780.0,
      "step": 2303
    },
    {
      "entropy": 0.9864234030246735,
      "epoch": 1.2132701421800949,
      "grad_norm": 0.2760343551635742,
      "learning_rate": 0.00019048808054308183,
      "loss": 0.18409714102745056,
      "mean_token_accuracy": 0.9212307929992676,
      "num_tokens": 28421116.0,
      "step": 2304
    },
    {
      "entropy": 1.0085313767194748,
      "epoch": 1.2137967351237493,
      "grad_norm": 0.31804919242858887,
      "learning_rate": 0.00019047826647584324,
      "loss": 0.21396562457084656,
      "mean_token_accuracy": 0.9174764901399612,
      "num_tokens": 28433452.0,
      "step": 2305
    },
    {
      "entropy": 0.9060227423906326,
      "epoch": 1.214323328067404,
      "grad_norm": 0.26906898617744446,
      "learning_rate": 0.000190468447631079,
      "loss": 0.1767490655183792,
      "mean_token_accuracy": 0.9256879538297653,
      "num_tokens": 28445788.0,
      "step": 2306
    },
    {
      "entropy": 0.9600940495729446,
      "epoch": 1.2148499210110584,
      "grad_norm": 0.31325221061706543,
      "learning_rate": 0.00019045862400937207,
      "loss": 0.2050899714231491,
      "mean_token_accuracy": 0.915095791220665,
      "num_tokens": 28458124.0,
      "step": 2307
    },
    {
      "entropy": 0.9618964791297913,
      "epoch": 1.215376513954713,
      "grad_norm": 0.28453946113586426,
      "learning_rate": 0.00019044879561130553,
      "loss": 0.19918674230575562,
      "mean_token_accuracy": 0.9192324429750443,
      "num_tokens": 28470460.0,
      "step": 2308
    },
    {
      "entropy": 0.9584704488515854,
      "epoch": 1.2159031068983674,
      "grad_norm": 0.27916356921195984,
      "learning_rate": 0.00019043896243746292,
      "loss": 0.20780225098133087,
      "mean_token_accuracy": 0.9173972904682159,
      "num_tokens": 28482796.0,
      "step": 2309
    },
    {
      "entropy": 0.9718508273363113,
      "epoch": 1.216429699842022,
      "grad_norm": 0.3944566249847412,
      "learning_rate": 0.00019042912448842794,
      "loss": 0.20700976252555847,
      "mean_token_accuracy": 0.9170730262994766,
      "num_tokens": 28495132.0,
      "step": 2310
    },
    {
      "entropy": 0.9472010433673859,
      "epoch": 1.2169562927856767,
      "grad_norm": 0.27791833877563477,
      "learning_rate": 0.0001904192817647846,
      "loss": 0.19973602890968323,
      "mean_token_accuracy": 0.9199665784835815,
      "num_tokens": 28507468.0,
      "step": 2311
    },
    {
      "entropy": 0.9158832877874374,
      "epoch": 1.2174828857293312,
      "grad_norm": 0.2705928385257721,
      "learning_rate": 0.00019040943426711725,
      "loss": 0.1925109326839447,
      "mean_token_accuracy": 0.922896683216095,
      "num_tokens": 28519804.0,
      "step": 2312
    },
    {
      "entropy": 0.9301202595233917,
      "epoch": 1.2180094786729858,
      "grad_norm": 0.29519709944725037,
      "learning_rate": 0.00019039958199601044,
      "loss": 0.2137046456336975,
      "mean_token_accuracy": 0.9163611382246017,
      "num_tokens": 28532140.0,
      "step": 2313
    },
    {
      "entropy": 0.9268923699855804,
      "epoch": 1.2185360716166405,
      "grad_norm": 0.29034554958343506,
      "learning_rate": 0.00019038972495204906,
      "loss": 0.20945435762405396,
      "mean_token_accuracy": 0.9179003089666367,
      "num_tokens": 28544476.0,
      "step": 2314
    },
    {
      "entropy": 0.9507464468479156,
      "epoch": 1.2190626645602949,
      "grad_norm": 0.3054998517036438,
      "learning_rate": 0.00019037986313581823,
      "loss": 0.21049703657627106,
      "mean_token_accuracy": 0.9139607399702072,
      "num_tokens": 28556812.0,
      "step": 2315
    },
    {
      "entropy": 0.9595097601413727,
      "epoch": 1.2195892575039495,
      "grad_norm": 0.28608593344688416,
      "learning_rate": 0.0001903699965479034,
      "loss": 0.22186635434627533,
      "mean_token_accuracy": 0.9084164053201675,
      "num_tokens": 28569148.0,
      "step": 2316
    },
    {
      "entropy": 0.9086242020130157,
      "epoch": 1.220115850447604,
      "grad_norm": 0.2904057502746582,
      "learning_rate": 0.00019036012518889036,
      "loss": 0.1948484480381012,
      "mean_token_accuracy": 0.9190005958080292,
      "num_tokens": 28581484.0,
      "step": 2317
    },
    {
      "entropy": 0.9179560393095016,
      "epoch": 1.2206424433912586,
      "grad_norm": 0.2642764151096344,
      "learning_rate": 0.00019035024905936503,
      "loss": 0.18806758522987366,
      "mean_token_accuracy": 0.9178005754947662,
      "num_tokens": 28593820.0,
      "step": 2318
    },
    {
      "entropy": 0.9741402119398117,
      "epoch": 1.221169036334913,
      "grad_norm": 0.28891104459762573,
      "learning_rate": 0.00019034036815991374,
      "loss": 0.21470314264297485,
      "mean_token_accuracy": 0.9158063977956772,
      "num_tokens": 28606156.0,
      "step": 2319
    },
    {
      "entropy": 0.9723116159439087,
      "epoch": 1.2216956292785677,
      "grad_norm": 0.2928186058998108,
      "learning_rate": 0.00019033048249112304,
      "loss": 0.21727263927459717,
      "mean_token_accuracy": 0.9125364273786545,
      "num_tokens": 28618492.0,
      "step": 2320
    },
    {
      "entropy": 1.0076178759336472,
      "epoch": 1.2222222222222223,
      "grad_norm": 0.2738136649131775,
      "learning_rate": 0.00019032059205357978,
      "loss": 0.20682492852210999,
      "mean_token_accuracy": 0.9168647974729538,
      "num_tokens": 28630828.0,
      "step": 2321
    },
    {
      "entropy": 0.9914740473031998,
      "epoch": 1.2227488151658767,
      "grad_norm": 0.32728976011276245,
      "learning_rate": 0.00019031069684787117,
      "loss": 0.23216472566127777,
      "mean_token_accuracy": 0.9085246026515961,
      "num_tokens": 28643164.0,
      "step": 2322
    },
    {
      "entropy": 1.019994467496872,
      "epoch": 1.2232754081095314,
      "grad_norm": 0.3143840432167053,
      "learning_rate": 0.00019030079687458454,
      "loss": 0.2100144624710083,
      "mean_token_accuracy": 0.9146460592746735,
      "num_tokens": 28655500.0,
      "step": 2323
    },
    {
      "entropy": 1.0059214681386948,
      "epoch": 1.2238020010531858,
      "grad_norm": 0.285477876663208,
      "learning_rate": 0.0001902908921343076,
      "loss": 0.19599957764148712,
      "mean_token_accuracy": 0.9203824996948242,
      "num_tokens": 28667836.0,
      "step": 2324
    },
    {
      "entropy": 1.0299699008464813,
      "epoch": 1.2243285939968405,
      "grad_norm": 0.30807819962501526,
      "learning_rate": 0.00019028098262762844,
      "loss": 0.2146599292755127,
      "mean_token_accuracy": 0.9128930121660233,
      "num_tokens": 28680172.0,
      "step": 2325
    },
    {
      "entropy": 0.9639817774295807,
      "epoch": 1.2248551869404949,
      "grad_norm": 0.2883492112159729,
      "learning_rate": 0.00019027106835513519,
      "loss": 0.2116781324148178,
      "mean_token_accuracy": 0.9104322791099548,
      "num_tokens": 28692508.0,
      "step": 2326
    },
    {
      "entropy": 0.9987800866365433,
      "epoch": 1.2253817798841495,
      "grad_norm": 0.2772632837295532,
      "learning_rate": 0.00019026114931741648,
      "loss": 0.18406018614768982,
      "mean_token_accuracy": 0.9260071665048599,
      "num_tokens": 28704844.0,
      "step": 2327
    },
    {
      "entropy": 0.9284749180078506,
      "epoch": 1.2259083728278042,
      "grad_norm": 0.26899397373199463,
      "learning_rate": 0.00019025122551506117,
      "loss": 0.19465957581996918,
      "mean_token_accuracy": 0.9211094230413437,
      "num_tokens": 28717180.0,
      "step": 2328
    },
    {
      "entropy": 0.9498527348041534,
      "epoch": 1.2264349657714586,
      "grad_norm": 0.2672523260116577,
      "learning_rate": 0.0001902412969486583,
      "loss": 0.1857932209968567,
      "mean_token_accuracy": 0.923532709479332,
      "num_tokens": 28729516.0,
      "step": 2329
    },
    {
      "entropy": 0.9485073536634445,
      "epoch": 1.2269615587151133,
      "grad_norm": 0.6426500678062439,
      "learning_rate": 0.00019023136361879733,
      "loss": 0.20064154267311096,
      "mean_token_accuracy": 0.9190724045038223,
      "num_tokens": 28741852.0,
      "step": 2330
    },
    {
      "entropy": 0.9678667783737183,
      "epoch": 1.2274881516587677,
      "grad_norm": 0.28017935156822205,
      "learning_rate": 0.00019022142552606788,
      "loss": 0.19379380345344543,
      "mean_token_accuracy": 0.9230899214744568,
      "num_tokens": 28754188.0,
      "step": 2331
    },
    {
      "entropy": 0.9232030957937241,
      "epoch": 1.2280147446024223,
      "grad_norm": 0.2922697365283966,
      "learning_rate": 0.00019021148267106002,
      "loss": 0.21339812874794006,
      "mean_token_accuracy": 0.9169651716947556,
      "num_tokens": 28766524.0,
      "step": 2332
    },
    {
      "entropy": 0.9682541638612747,
      "epoch": 1.228541337546077,
      "grad_norm": 0.2776217758655548,
      "learning_rate": 0.00019020153505436388,
      "loss": 0.1879439353942871,
      "mean_token_accuracy": 0.9254468381404877,
      "num_tokens": 28778860.0,
      "step": 2333
    },
    {
      "entropy": 0.9788610488176346,
      "epoch": 1.2290679304897314,
      "grad_norm": 0.2904343903064728,
      "learning_rate": 0.00019019158267657014,
      "loss": 0.20155103504657745,
      "mean_token_accuracy": 0.915040597319603,
      "num_tokens": 28791196.0,
      "step": 2334
    },
    {
      "entropy": 0.9573673754930496,
      "epoch": 1.229594523433386,
      "grad_norm": 0.2809135913848877,
      "learning_rate": 0.00019018162553826947,
      "loss": 0.2041783183813095,
      "mean_token_accuracy": 0.91671222448349,
      "num_tokens": 28803532.0,
      "step": 2335
    },
    {
      "entropy": 0.9262150526046753,
      "epoch": 1.2301211163770405,
      "grad_norm": 0.2890893220901489,
      "learning_rate": 0.00019017166364005303,
      "loss": 0.1984890252351761,
      "mean_token_accuracy": 0.9185997843742371,
      "num_tokens": 28815868.0,
      "step": 2336
    },
    {
      "entropy": 0.9690218567848206,
      "epoch": 1.2306477093206951,
      "grad_norm": 0.30195772647857666,
      "learning_rate": 0.0001901616969825122,
      "loss": 0.19511699676513672,
      "mean_token_accuracy": 0.9191469699144363,
      "num_tokens": 28828204.0,
      "step": 2337
    },
    {
      "entropy": 0.9442458599805832,
      "epoch": 1.2311743022643498,
      "grad_norm": 0.3288037180900574,
      "learning_rate": 0.00019015172556623863,
      "loss": 0.23278209567070007,
      "mean_token_accuracy": 0.9055567979812622,
      "num_tokens": 28840540.0,
      "step": 2338
    },
    {
      "entropy": 0.9516178071498871,
      "epoch": 1.2317008952080042,
      "grad_norm": 0.3043895363807678,
      "learning_rate": 0.00019014174939182432,
      "loss": 0.20317667722702026,
      "mean_token_accuracy": 0.9137967079877853,
      "num_tokens": 28852876.0,
      "step": 2339
    },
    {
      "entropy": 1.0066047459840775,
      "epoch": 1.2322274881516588,
      "grad_norm": 0.28470954298973083,
      "learning_rate": 0.0001901317684598614,
      "loss": 0.17943891882896423,
      "mean_token_accuracy": 0.9240919649600983,
      "num_tokens": 28865212.0,
      "step": 2340
    },
    {
      "entropy": 0.9875893741846085,
      "epoch": 1.2327540810953133,
      "grad_norm": 0.279293954372406,
      "learning_rate": 0.00019012178277094246,
      "loss": 0.18953418731689453,
      "mean_token_accuracy": 0.9218238890171051,
      "num_tokens": 28877548.0,
      "step": 2341
    },
    {
      "entropy": 1.0479088127613068,
      "epoch": 1.233280674038968,
      "grad_norm": 0.3232957720756531,
      "learning_rate": 0.00019011179232566026,
      "loss": 0.19601483643054962,
      "mean_token_accuracy": 0.9135144054889679,
      "num_tokens": 28889884.0,
      "step": 2342
    },
    {
      "entropy": 1.0485905408859253,
      "epoch": 1.2338072669826223,
      "grad_norm": 0.2966122031211853,
      "learning_rate": 0.00019010179712460786,
      "loss": 0.214534193277359,
      "mean_token_accuracy": 0.9167278558015823,
      "num_tokens": 28902220.0,
      "step": 2343
    },
    {
      "entropy": 1.0693544447422028,
      "epoch": 1.234333859926277,
      "grad_norm": 0.2795875668525696,
      "learning_rate": 0.00019009179716837865,
      "loss": 0.1979323923587799,
      "mean_token_accuracy": 0.9212673157453537,
      "num_tokens": 28914556.0,
      "step": 2344
    },
    {
      "entropy": 1.1432946920394897,
      "epoch": 1.2348604528699316,
      "grad_norm": 0.32812783122062683,
      "learning_rate": 0.00019008179245756624,
      "loss": 0.20545022189617157,
      "mean_token_accuracy": 0.9139831513166428,
      "num_tokens": 28926892.0,
      "step": 2345
    },
    {
      "entropy": 1.156277745962143,
      "epoch": 1.235387045813586,
      "grad_norm": 0.35038238763809204,
      "learning_rate": 0.00019007178299276453,
      "loss": 0.22025498747825623,
      "mean_token_accuracy": 0.9107816964387894,
      "num_tokens": 28939228.0,
      "step": 2346
    },
    {
      "entropy": 1.167900800704956,
      "epoch": 1.2359136387572407,
      "grad_norm": 0.2617517411708832,
      "learning_rate": 0.00019006176877456782,
      "loss": 0.1741318702697754,
      "mean_token_accuracy": 0.9294529408216476,
      "num_tokens": 28951564.0,
      "step": 2347
    },
    {
      "entropy": 1.2039166390895844,
      "epoch": 1.2364402317008951,
      "grad_norm": 0.3141527473926544,
      "learning_rate": 0.0001900517498035705,
      "loss": 0.21124760806560516,
      "mean_token_accuracy": 0.9103933870792389,
      "num_tokens": 28963900.0,
      "step": 2348
    },
    {
      "entropy": 1.1911627054214478,
      "epoch": 1.2369668246445498,
      "grad_norm": 0.3090423047542572,
      "learning_rate": 0.00019004172608036736,
      "loss": 0.20455247163772583,
      "mean_token_accuracy": 0.9178474098443985,
      "num_tokens": 28976236.0,
      "step": 2349
    },
    {
      "entropy": 1.2294414341449738,
      "epoch": 1.2374934175882044,
      "grad_norm": 0.42555737495422363,
      "learning_rate": 0.0001900316976055535,
      "loss": 0.1899854689836502,
      "mean_token_accuracy": 0.9238343685865402,
      "num_tokens": 28988572.0,
      "step": 2350
    },
    {
      "entropy": 1.2140373587608337,
      "epoch": 1.2380200105318588,
      "grad_norm": 0.3391142189502716,
      "learning_rate": 0.00019002166437972418,
      "loss": 0.212956041097641,
      "mean_token_accuracy": 0.9184004813432693,
      "num_tokens": 29000908.0,
      "step": 2351
    },
    {
      "entropy": 1.245664894580841,
      "epoch": 1.2385466034755135,
      "grad_norm": 0.2943424880504608,
      "learning_rate": 0.00019001162640347506,
      "loss": 0.19764168560504913,
      "mean_token_accuracy": 0.9214494824409485,
      "num_tokens": 29013244.0,
      "step": 2352
    },
    {
      "entropy": 1.2262179851531982,
      "epoch": 1.239073196419168,
      "grad_norm": 0.29305392503738403,
      "learning_rate": 0.000190001583677402,
      "loss": 0.21708795428276062,
      "mean_token_accuracy": 0.9119549542665482,
      "num_tokens": 29025580.0,
      "step": 2353
    },
    {
      "entropy": 1.2130177021026611,
      "epoch": 1.2395997893628226,
      "grad_norm": 0.2500525414943695,
      "learning_rate": 0.00018999153620210122,
      "loss": 0.1909836232662201,
      "mean_token_accuracy": 0.9247071444988251,
      "num_tokens": 29037916.0,
      "step": 2354
    },
    {
      "entropy": 1.2663672864437103,
      "epoch": 1.2401263823064772,
      "grad_norm": 0.2888049781322479,
      "learning_rate": 0.00018998148397816918,
      "loss": 0.19668535888195038,
      "mean_token_accuracy": 0.9236019253730774,
      "num_tokens": 29050252.0,
      "step": 2355
    },
    {
      "entropy": 1.2552369832992554,
      "epoch": 1.2406529752501316,
      "grad_norm": 0.30252373218536377,
      "learning_rate": 0.00018997142700620257,
      "loss": 0.21202778816223145,
      "mean_token_accuracy": 0.9104820787906647,
      "num_tokens": 29062588.0,
      "step": 2356
    },
    {
      "entropy": 1.2437193095684052,
      "epoch": 1.2411795681937863,
      "grad_norm": 0.2685464322566986,
      "learning_rate": 0.0001899613652867985,
      "loss": 0.19783562421798706,
      "mean_token_accuracy": 0.9259221404790878,
      "num_tokens": 29074924.0,
      "step": 2357
    },
    {
      "entropy": 1.2458664774894714,
      "epoch": 1.2417061611374407,
      "grad_norm": 0.27750658988952637,
      "learning_rate": 0.00018995129882055418,
      "loss": 0.19608265161514282,
      "mean_token_accuracy": 0.9223610162734985,
      "num_tokens": 29087260.0,
      "step": 2358
    },
    {
      "entropy": 1.28984996676445,
      "epoch": 1.2422327540810953,
      "grad_norm": 0.3127368688583374,
      "learning_rate": 0.00018994122760806724,
      "loss": 0.2381632924079895,
      "mean_token_accuracy": 0.906804770231247,
      "num_tokens": 29099596.0,
      "step": 2359
    },
    {
      "entropy": 1.2786964178085327,
      "epoch": 1.2427593470247498,
      "grad_norm": 0.29435664415359497,
      "learning_rate": 0.00018993115164993556,
      "loss": 0.2056436836719513,
      "mean_token_accuracy": 0.9188906252384186,
      "num_tokens": 29111932.0,
      "step": 2360
    },
    {
      "entropy": 1.296678066253662,
      "epoch": 1.2432859399684044,
      "grad_norm": 0.2973683476448059,
      "learning_rate": 0.00018992107094675727,
      "loss": 0.21866685152053833,
      "mean_token_accuracy": 0.9110830724239349,
      "num_tokens": 29124268.0,
      "step": 2361
    },
    {
      "entropy": 1.2541703283786774,
      "epoch": 1.243812532912059,
      "grad_norm": 0.2897864580154419,
      "learning_rate": 0.00018991098549913084,
      "loss": 0.18205136060714722,
      "mean_token_accuracy": 0.9259398877620697,
      "num_tokens": 29136604.0,
      "step": 2362
    },
    {
      "entropy": 1.2319467067718506,
      "epoch": 1.2443391258557135,
      "grad_norm": 0.2912856638431549,
      "learning_rate": 0.00018990089530765493,
      "loss": 0.21475206315517426,
      "mean_token_accuracy": 0.9117068946361542,
      "num_tokens": 29148940.0,
      "step": 2363
    },
    {
      "entropy": 1.1983669102191925,
      "epoch": 1.2448657187993681,
      "grad_norm": 0.30084285140037537,
      "learning_rate": 0.00018989080037292856,
      "loss": 0.20513667166233063,
      "mean_token_accuracy": 0.9175577163696289,
      "num_tokens": 29161276.0,
      "step": 2364
    },
    {
      "entropy": 1.208276093006134,
      "epoch": 1.2453923117430226,
      "grad_norm": 0.28459402918815613,
      "learning_rate": 0.00018988070069555105,
      "loss": 0.20270894467830658,
      "mean_token_accuracy": 0.9216506034135818,
      "num_tokens": 29173612.0,
      "step": 2365
    },
    {
      "entropy": 1.1837323307991028,
      "epoch": 1.2459189046866772,
      "grad_norm": 0.27861863374710083,
      "learning_rate": 0.0001898705962761219,
      "loss": 0.19915355741977692,
      "mean_token_accuracy": 0.9196107387542725,
      "num_tokens": 29185948.0,
      "step": 2366
    },
    {
      "entropy": 1.1845957934856415,
      "epoch": 1.2464454976303316,
      "grad_norm": 0.28316664695739746,
      "learning_rate": 0.00018986048711524097,
      "loss": 0.21355314552783966,
      "mean_token_accuracy": 0.9129188060760498,
      "num_tokens": 29198284.0,
      "step": 2367
    },
    {
      "entropy": 1.1677678227424622,
      "epoch": 1.2469720905739863,
      "grad_norm": 0.3210926651954651,
      "learning_rate": 0.00018985037321350836,
      "loss": 0.20634154975414276,
      "mean_token_accuracy": 0.9156036972999573,
      "num_tokens": 29210620.0,
      "step": 2368
    },
    {
      "entropy": 1.2081373929977417,
      "epoch": 1.247498683517641,
      "grad_norm": 0.32751724123954773,
      "learning_rate": 0.00018984025457152452,
      "loss": 0.22158020734786987,
      "mean_token_accuracy": 0.9142793416976929,
      "num_tokens": 29222956.0,
      "step": 2369
    },
    {
      "entropy": 1.198670506477356,
      "epoch": 1.2480252764612954,
      "grad_norm": 0.3396313488483429,
      "learning_rate": 0.0001898301311898901,
      "loss": 0.22338370978832245,
      "mean_token_accuracy": 0.9130024313926697,
      "num_tokens": 29235292.0,
      "step": 2370
    },
    {
      "entropy": 1.187466412782669,
      "epoch": 1.24855186940495,
      "grad_norm": 0.330683171749115,
      "learning_rate": 0.00018982000306920607,
      "loss": 0.2252286970615387,
      "mean_token_accuracy": 0.9099649637937546,
      "num_tokens": 29247628.0,
      "step": 2371
    },
    {
      "entropy": 1.1201028525829315,
      "epoch": 1.2490784623486046,
      "grad_norm": 0.2788279950618744,
      "learning_rate": 0.00018980987021007372,
      "loss": 0.197609081864357,
      "mean_token_accuracy": 0.919813871383667,
      "num_tokens": 29259964.0,
      "step": 2372
    },
    {
      "entropy": 1.1749643087387085,
      "epoch": 1.249605055292259,
      "grad_norm": 0.27696964144706726,
      "learning_rate": 0.00018979973261309453,
      "loss": 0.19340327382087708,
      "mean_token_accuracy": 0.9178819060325623,
      "num_tokens": 29272300.0,
      "step": 2373
    },
    {
      "entropy": 1.171377807855606,
      "epoch": 1.2501316482359137,
      "grad_norm": 0.26827797293663025,
      "learning_rate": 0.0001897895902788703,
      "loss": 0.20257636904716492,
      "mean_token_accuracy": 0.9170994907617569,
      "num_tokens": 29284636.0,
      "step": 2374
    },
    {
      "entropy": 1.1583305597305298,
      "epoch": 1.2506582411795681,
      "grad_norm": 0.26647472381591797,
      "learning_rate": 0.00018977944320800315,
      "loss": 0.19090743362903595,
      "mean_token_accuracy": 0.9245351999998093,
      "num_tokens": 29296972.0,
      "step": 2375
    },
    {
      "entropy": 1.2192396521568298,
      "epoch": 1.2511848341232228,
      "grad_norm": 0.29227685928344727,
      "learning_rate": 0.00018976929140109542,
      "loss": 0.21741797029972076,
      "mean_token_accuracy": 0.911442294716835,
      "num_tokens": 29309308.0,
      "step": 2376
    },
    {
      "entropy": 1.120053082704544,
      "epoch": 1.2517114270668772,
      "grad_norm": 0.27130043506622314,
      "learning_rate": 0.00018975913485874983,
      "loss": 0.2002987265586853,
      "mean_token_accuracy": 0.9195822030305862,
      "num_tokens": 29321644.0,
      "step": 2377
    },
    {
      "entropy": 1.1931759715080261,
      "epoch": 1.2522380200105319,
      "grad_norm": 0.28110983967781067,
      "learning_rate": 0.00018974897358156926,
      "loss": 0.18119770288467407,
      "mean_token_accuracy": 0.921199381351471,
      "num_tokens": 29333980.0,
      "step": 2378
    },
    {
      "entropy": 1.1507650911808014,
      "epoch": 1.2527646129541865,
      "grad_norm": 0.3061140775680542,
      "learning_rate": 0.00018973880757015696,
      "loss": 0.20350392162799835,
      "mean_token_accuracy": 0.918747752904892,
      "num_tokens": 29346316.0,
      "step": 2379
    },
    {
      "entropy": 1.1738670468330383,
      "epoch": 1.253291205897841,
      "grad_norm": 0.27520132064819336,
      "learning_rate": 0.00018972863682511639,
      "loss": 0.19713857769966125,
      "mean_token_accuracy": 0.9176778495311737,
      "num_tokens": 29358652.0,
      "step": 2380
    },
    {
      "entropy": 1.1170638799667358,
      "epoch": 1.2538177988414956,
      "grad_norm": 0.28535959124565125,
      "learning_rate": 0.00018971846134705137,
      "loss": 0.21452412009239197,
      "mean_token_accuracy": 0.9095005989074707,
      "num_tokens": 29370988.0,
      "step": 2381
    },
    {
      "entropy": 1.1119027435779572,
      "epoch": 1.25434439178515,
      "grad_norm": 0.2742728888988495,
      "learning_rate": 0.00018970828113656593,
      "loss": 0.1901988387107849,
      "mean_token_accuracy": 0.921889141201973,
      "num_tokens": 29383324.0,
      "step": 2382
    },
    {
      "entropy": 1.0820831060409546,
      "epoch": 1.2548709847288047,
      "grad_norm": 0.28090232610702515,
      "learning_rate": 0.00018969809619426439,
      "loss": 0.19445692002773285,
      "mean_token_accuracy": 0.9170260429382324,
      "num_tokens": 29395660.0,
      "step": 2383
    },
    {
      "entropy": 1.055662214756012,
      "epoch": 1.255397577672459,
      "grad_norm": 0.27719929814338684,
      "learning_rate": 0.00018968790652075136,
      "loss": 0.1977691501379013,
      "mean_token_accuracy": 0.919326663017273,
      "num_tokens": 29407996.0,
      "step": 2384
    },
    {
      "entropy": 1.1013965904712677,
      "epoch": 1.2559241706161137,
      "grad_norm": 0.31086885929107666,
      "learning_rate": 0.00018967771211663182,
      "loss": 0.2363457977771759,
      "mean_token_accuracy": 0.906139463186264,
      "num_tokens": 29420332.0,
      "step": 2385
    },
    {
      "entropy": 1.090101808309555,
      "epoch": 1.2564507635597684,
      "grad_norm": 0.2687731087207794,
      "learning_rate": 0.00018966751298251093,
      "loss": 0.18311643600463867,
      "mean_token_accuracy": 0.9264214485883713,
      "num_tokens": 29432668.0,
      "step": 2386
    },
    {
      "entropy": 1.1113143563270569,
      "epoch": 1.2569773565034228,
      "grad_norm": 0.3031775951385498,
      "learning_rate": 0.0001896573091189941,
      "loss": 0.20669670403003693,
      "mean_token_accuracy": 0.9185007810592651,
      "num_tokens": 29445004.0,
      "step": 2387
    },
    {
      "entropy": 1.0604256093502045,
      "epoch": 1.2575039494470774,
      "grad_norm": 0.28647610545158386,
      "learning_rate": 0.00018964710052668714,
      "loss": 0.20895907282829285,
      "mean_token_accuracy": 0.9176440387964249,
      "num_tokens": 29457340.0,
      "step": 2388
    },
    {
      "entropy": 1.094488948583603,
      "epoch": 1.258030542390732,
      "grad_norm": 0.28277674317359924,
      "learning_rate": 0.000189636887206196,
      "loss": 0.1958828866481781,
      "mean_token_accuracy": 0.9168824553489685,
      "num_tokens": 29469676.0,
      "step": 2389
    },
    {
      "entropy": 1.0930402874946594,
      "epoch": 1.2585571353343865,
      "grad_norm": 0.3038513660430908,
      "learning_rate": 0.00018962666915812702,
      "loss": 0.22608797252178192,
      "mean_token_accuracy": 0.9073146730661392,
      "num_tokens": 29482012.0,
      "step": 2390
    },
    {
      "entropy": 1.090045154094696,
      "epoch": 1.259083728278041,
      "grad_norm": 0.27166202664375305,
      "learning_rate": 0.00018961644638308683,
      "loss": 0.1975858360528946,
      "mean_token_accuracy": 0.9241401851177216,
      "num_tokens": 29494348.0,
      "step": 2391
    },
    {
      "entropy": 1.0796633809804916,
      "epoch": 1.2596103212216956,
      "grad_norm": 0.2842889130115509,
      "learning_rate": 0.00018960621888168224,
      "loss": 0.19585753977298737,
      "mean_token_accuracy": 0.9207384586334229,
      "num_tokens": 29506684.0,
      "step": 2392
    },
    {
      "entropy": 1.1113807559013367,
      "epoch": 1.2601369141653502,
      "grad_norm": 0.27799156308174133,
      "learning_rate": 0.00018959598665452043,
      "loss": 0.20170974731445312,
      "mean_token_accuracy": 0.9220674335956573,
      "num_tokens": 29519020.0,
      "step": 2393
    },
    {
      "entropy": 1.1154912114143372,
      "epoch": 1.2606635071090047,
      "grad_norm": 0.2918773889541626,
      "learning_rate": 0.0001895857497022088,
      "loss": 0.2008930742740631,
      "mean_token_accuracy": 0.9212677925825119,
      "num_tokens": 29531356.0,
      "step": 2394
    },
    {
      "entropy": 1.0571572482585907,
      "epoch": 1.2611901000526593,
      "grad_norm": 0.28213104605674744,
      "learning_rate": 0.00018957550802535503,
      "loss": 0.21011856198310852,
      "mean_token_accuracy": 0.9190024733543396,
      "num_tokens": 29543692.0,
      "step": 2395
    },
    {
      "entropy": 1.0876768827438354,
      "epoch": 1.261716692996314,
      "grad_norm": 0.3166556656360626,
      "learning_rate": 0.00018956526162456722,
      "loss": 0.21292051672935486,
      "mean_token_accuracy": 0.9117415696382523,
      "num_tokens": 29556028.0,
      "step": 2396
    },
    {
      "entropy": 1.1123850643634796,
      "epoch": 1.2622432859399684,
      "grad_norm": 0.3177889585494995,
      "learning_rate": 0.00018955501050045353,
      "loss": 0.21554914116859436,
      "mean_token_accuracy": 0.9117074012756348,
      "num_tokens": 29568364.0,
      "step": 2397
    },
    {
      "entropy": 1.121145874261856,
      "epoch": 1.262769878883623,
      "grad_norm": 0.28797921538352966,
      "learning_rate": 0.00018954475465362256,
      "loss": 0.2110595405101776,
      "mean_token_accuracy": 0.910643994808197,
      "num_tokens": 29580700.0,
      "step": 2398
    },
    {
      "entropy": 1.081002026796341,
      "epoch": 1.2632964718272774,
      "grad_norm": 0.2767108678817749,
      "learning_rate": 0.00018953449408468312,
      "loss": 0.20874468982219696,
      "mean_token_accuracy": 0.916566014289856,
      "num_tokens": 29593036.0,
      "step": 2399
    },
    {
      "entropy": 1.0670734643936157,
      "epoch": 1.263823064770932,
      "grad_norm": 0.2667304277420044,
      "learning_rate": 0.00018952422879424437,
      "loss": 0.21136702597141266,
      "mean_token_accuracy": 0.915779709815979,
      "num_tokens": 29605372.0,
      "step": 2400
    },
    {
      "entropy": 1.1176138818264008,
      "epoch": 1.2643496577145865,
      "grad_norm": 0.28897732496261597,
      "learning_rate": 0.0001895139587829156,
      "loss": 0.20740735530853271,
      "mean_token_accuracy": 0.9172630310058594,
      "num_tokens": 29617708.0,
      "step": 2401
    },
    {
      "entropy": 1.0523147583007812,
      "epoch": 1.2648762506582412,
      "grad_norm": 0.28087154030799866,
      "learning_rate": 0.00018950368405130653,
      "loss": 0.20881010591983795,
      "mean_token_accuracy": 0.9143330603837967,
      "num_tokens": 29630044.0,
      "step": 2402
    },
    {
      "entropy": 1.073082596063614,
      "epoch": 1.2654028436018958,
      "grad_norm": 0.2673327624797821,
      "learning_rate": 0.00018949340460002717,
      "loss": 0.20516178011894226,
      "mean_token_accuracy": 0.9163246154785156,
      "num_tokens": 29642380.0,
      "step": 2403
    },
    {
      "entropy": 0.9944493472576141,
      "epoch": 1.2659294365455502,
      "grad_norm": 0.23029980063438416,
      "learning_rate": 0.00018948312042968768,
      "loss": 0.18642765283584595,
      "mean_token_accuracy": 0.9236330091953278,
      "num_tokens": 29654716.0,
      "step": 2404
    },
    {
      "entropy": 1.0669757723808289,
      "epoch": 1.2664560294892049,
      "grad_norm": 0.26822593808174133,
      "learning_rate": 0.0001894728315408986,
      "loss": 0.20285025238990784,
      "mean_token_accuracy": 0.9172052592039108,
      "num_tokens": 29667052.0,
      "step": 2405
    },
    {
      "entropy": 1.0667217671871185,
      "epoch": 1.2669826224328595,
      "grad_norm": 0.28592145442962646,
      "learning_rate": 0.00018946253793427075,
      "loss": 0.21876372396945953,
      "mean_token_accuracy": 0.9099210351705551,
      "num_tokens": 29679388.0,
      "step": 2406
    },
    {
      "entropy": 1.0617501437664032,
      "epoch": 1.267509215376514,
      "grad_norm": 0.2814294099807739,
      "learning_rate": 0.0001894522396104151,
      "loss": 0.21466723084449768,
      "mean_token_accuracy": 0.9166467934846878,
      "num_tokens": 29691724.0,
      "step": 2407
    },
    {
      "entropy": 1.0295425057411194,
      "epoch": 1.2680358083201684,
      "grad_norm": 0.2818832993507385,
      "learning_rate": 0.0001894419365699431,
      "loss": 0.2018805742263794,
      "mean_token_accuracy": 0.9214745163917542,
      "num_tokens": 29704060.0,
      "step": 2408
    },
    {
      "entropy": 0.9962182641029358,
      "epoch": 1.268562401263823,
      "grad_norm": 0.2606944441795349,
      "learning_rate": 0.00018943162881346633,
      "loss": 0.20163223147392273,
      "mean_token_accuracy": 0.9152373224496841,
      "num_tokens": 29716396.0,
      "step": 2409
    },
    {
      "entropy": 1.0103175342082977,
      "epoch": 1.2690889942074777,
      "grad_norm": 0.2832614779472351,
      "learning_rate": 0.00018942131634159672,
      "loss": 0.20267333090305328,
      "mean_token_accuracy": 0.917528048157692,
      "num_tokens": 29728732.0,
      "step": 2410
    },
    {
      "entropy": 1.030224397778511,
      "epoch": 1.269615587151132,
      "grad_norm": 0.28863924741744995,
      "learning_rate": 0.00018941099915494646,
      "loss": 0.20709314942359924,
      "mean_token_accuracy": 0.9195363223552704,
      "num_tokens": 29741068.0,
      "step": 2411
    },
    {
      "entropy": 0.9850256592035294,
      "epoch": 1.2701421800947867,
      "grad_norm": 0.33532777428627014,
      "learning_rate": 0.00018940067725412807,
      "loss": 0.22378413379192352,
      "mean_token_accuracy": 0.9103842526674271,
      "num_tokens": 29753404.0,
      "step": 2412
    },
    {
      "entropy": 0.9951465576887131,
      "epoch": 1.2706687730384414,
      "grad_norm": 0.2885618209838867,
      "learning_rate": 0.0001893903506397542,
      "loss": 0.20284360647201538,
      "mean_token_accuracy": 0.9174323529005051,
      "num_tokens": 29765740.0,
      "step": 2413
    },
    {
      "entropy": 0.9759911596775055,
      "epoch": 1.2711953659820958,
      "grad_norm": 0.30822882056236267,
      "learning_rate": 0.00018938001931243797,
      "loss": 0.21964068710803986,
      "mean_token_accuracy": 0.9072186797857285,
      "num_tokens": 29778076.0,
      "step": 2414
    },
    {
      "entropy": 0.9762609899044037,
      "epoch": 1.2717219589257505,
      "grad_norm": 0.2687565088272095,
      "learning_rate": 0.00018936968327279264,
      "loss": 0.17709898948669434,
      "mean_token_accuracy": 0.9282322078943253,
      "num_tokens": 29790412.0,
      "step": 2415
    },
    {
      "entropy": 0.9670789986848831,
      "epoch": 1.2722485518694049,
      "grad_norm": 0.28542351722717285,
      "learning_rate": 0.00018935934252143182,
      "loss": 0.1963491588830948,
      "mean_token_accuracy": 0.9188527315855026,
      "num_tokens": 29802748.0,
      "step": 2416
    },
    {
      "entropy": 0.9343395233154297,
      "epoch": 1.2727751448130595,
      "grad_norm": 0.28767991065979004,
      "learning_rate": 0.00018934899705896937,
      "loss": 0.2048400342464447,
      "mean_token_accuracy": 0.9184893816709518,
      "num_tokens": 29815084.0,
      "step": 2417
    },
    {
      "entropy": 0.965802863240242,
      "epoch": 1.273301737756714,
      "grad_norm": 0.2734317481517792,
      "learning_rate": 0.00018933864688601946,
      "loss": 0.19449257850646973,
      "mean_token_accuracy": 0.9198505729436874,
      "num_tokens": 29827420.0,
      "step": 2418
    },
    {
      "entropy": 0.9207088947296143,
      "epoch": 1.2738283307003686,
      "grad_norm": 0.28382739424705505,
      "learning_rate": 0.00018932829200319647,
      "loss": 0.21684856712818146,
      "mean_token_accuracy": 0.9112024903297424,
      "num_tokens": 29839756.0,
      "step": 2419
    },
    {
      "entropy": 0.9559458643198013,
      "epoch": 1.2743549236440233,
      "grad_norm": 0.26874884963035583,
      "learning_rate": 0.00018931793241111519,
      "loss": 0.200919508934021,
      "mean_token_accuracy": 0.9189273715019226,
      "num_tokens": 29852092.0,
      "step": 2420
    },
    {
      "entropy": 0.956595167517662,
      "epoch": 1.2748815165876777,
      "grad_norm": 0.2639332115650177,
      "learning_rate": 0.0001893075681103905,
      "loss": 0.18435338139533997,
      "mean_token_accuracy": 0.9255244880914688,
      "num_tokens": 29864428.0,
      "step": 2421
    },
    {
      "entropy": 0.9653457999229431,
      "epoch": 1.2754081095313323,
      "grad_norm": 0.2683272659778595,
      "learning_rate": 0.0001892971991016378,
      "loss": 0.19316944479942322,
      "mean_token_accuracy": 0.9227836430072784,
      "num_tokens": 29876764.0,
      "step": 2422
    },
    {
      "entropy": 0.9717456251382828,
      "epoch": 1.275934702474987,
      "grad_norm": 0.2988039553165436,
      "learning_rate": 0.00018928682538547252,
      "loss": 0.2070363163948059,
      "mean_token_accuracy": 0.9153457581996918,
      "num_tokens": 29889100.0,
      "step": 2423
    },
    {
      "entropy": 0.968423455953598,
      "epoch": 1.2764612954186414,
      "grad_norm": 0.2866435647010803,
      "learning_rate": 0.0001892764469625106,
      "loss": 0.22353532910346985,
      "mean_token_accuracy": 0.9065069407224655,
      "num_tokens": 29901436.0,
      "step": 2424
    },
    {
      "entropy": 1.0139219462871552,
      "epoch": 1.2769878883622958,
      "grad_norm": 0.3166240155696869,
      "learning_rate": 0.000189266063833368,
      "loss": 0.21094951033592224,
      "mean_token_accuracy": 0.9138725250959396,
      "num_tokens": 29913772.0,
      "step": 2425
    },
    {
      "entropy": 0.9436186701059341,
      "epoch": 1.2775144813059505,
      "grad_norm": 0.266792893409729,
      "learning_rate": 0.00018925567599866123,
      "loss": 0.1921631246805191,
      "mean_token_accuracy": 0.923053190112114,
      "num_tokens": 29926108.0,
      "step": 2426
    },
    {
      "entropy": 0.9432852864265442,
      "epoch": 1.2780410742496051,
      "grad_norm": 0.27570202946662903,
      "learning_rate": 0.00018924528345900689,
      "loss": 0.20491164922714233,
      "mean_token_accuracy": 0.9194526821374893,
      "num_tokens": 29938444.0,
      "step": 2427
    },
    {
      "entropy": 1.0146455764770508,
      "epoch": 1.2785676671932595,
      "grad_norm": 0.3002910614013672,
      "learning_rate": 0.00018923488621502198,
      "loss": 0.21310178935527802,
      "mean_token_accuracy": 0.9097048193216324,
      "num_tokens": 29950780.0,
      "step": 2428
    },
    {
      "entropy": 0.9136994779109955,
      "epoch": 1.2790942601369142,
      "grad_norm": 0.2668123245239258,
      "learning_rate": 0.0001892244842673237,
      "loss": 0.19984757900238037,
      "mean_token_accuracy": 0.9176211655139923,
      "num_tokens": 29963116.0,
      "step": 2429
    },
    {
      "entropy": 0.9539975225925446,
      "epoch": 1.2796208530805688,
      "grad_norm": 0.27499574422836304,
      "learning_rate": 0.0001892140776165295,
      "loss": 0.19691187143325806,
      "mean_token_accuracy": 0.9206452816724777,
      "num_tokens": 29975452.0,
      "step": 2430
    },
    {
      "entropy": 0.9451378732919693,
      "epoch": 1.2801474460242233,
      "grad_norm": 0.26354870200157166,
      "learning_rate": 0.00018920366626325722,
      "loss": 0.19863781332969666,
      "mean_token_accuracy": 0.9141591340303421,
      "num_tokens": 29987788.0,
      "step": 2431
    },
    {
      "entropy": 0.9598711580038071,
      "epoch": 1.280674038967878,
      "grad_norm": 0.27625903487205505,
      "learning_rate": 0.00018919325020812493,
      "loss": 0.18384377658367157,
      "mean_token_accuracy": 0.9277281761169434,
      "num_tokens": 30000124.0,
      "step": 2432
    },
    {
      "entropy": 0.9703495502471924,
      "epoch": 1.2812006319115323,
      "grad_norm": 0.23485948145389557,
      "learning_rate": 0.00018918282945175091,
      "loss": 0.16364209353923798,
      "mean_token_accuracy": 0.9334625601768494,
      "num_tokens": 30012460.0,
      "step": 2433
    },
    {
      "entropy": 0.9682765752077103,
      "epoch": 1.281727224855187,
      "grad_norm": 0.28086769580841064,
      "learning_rate": 0.00018917240399475387,
      "loss": 0.20836037397384644,
      "mean_token_accuracy": 0.9150678068399429,
      "num_tokens": 30024796.0,
      "step": 2434
    },
    {
      "entropy": 0.9923833310604095,
      "epoch": 1.2822538177988414,
      "grad_norm": 0.2846622169017792,
      "learning_rate": 0.00018916197383775263,
      "loss": 0.19592522084712982,
      "mean_token_accuracy": 0.9235945791006088,
      "num_tokens": 30037132.0,
      "step": 2435
    },
    {
      "entropy": 0.9900262653827667,
      "epoch": 1.282780410742496,
      "grad_norm": 0.2638978958129883,
      "learning_rate": 0.00018915153898136638,
      "loss": 0.20760269463062286,
      "mean_token_accuracy": 0.9165667444467545,
      "num_tokens": 30049468.0,
      "step": 2436
    },
    {
      "entropy": 0.9972607046365738,
      "epoch": 1.2833070036861507,
      "grad_norm": 0.2679845988750458,
      "learning_rate": 0.0001891410994262146,
      "loss": 0.21356555819511414,
      "mean_token_accuracy": 0.9128797203302383,
      "num_tokens": 30061804.0,
      "step": 2437
    },
    {
      "entropy": 1.0473797023296356,
      "epoch": 1.2838335966298051,
      "grad_norm": 0.29876357316970825,
      "learning_rate": 0.000189130655172917,
      "loss": 0.1855165660381317,
      "mean_token_accuracy": 0.9234071522951126,
      "num_tokens": 30074140.0,
      "step": 2438
    },
    {
      "entropy": 1.0313687026500702,
      "epoch": 1.2843601895734598,
      "grad_norm": 0.3158549666404724,
      "learning_rate": 0.00018912020622209363,
      "loss": 0.19452282786369324,
      "mean_token_accuracy": 0.9206828027963638,
      "num_tokens": 30086476.0,
      "step": 2439
    },
    {
      "entropy": 1.0512953251600266,
      "epoch": 1.2848867825171142,
      "grad_norm": 0.289561927318573,
      "learning_rate": 0.00018910975257436477,
      "loss": 0.21705621480941772,
      "mean_token_accuracy": 0.9130683243274689,
      "num_tokens": 30098812.0,
      "step": 2440
    },
    {
      "entropy": 1.0579238533973694,
      "epoch": 1.2854133754607688,
      "grad_norm": 0.27795788645744324,
      "learning_rate": 0.00018909929423035097,
      "loss": 0.1992982029914856,
      "mean_token_accuracy": 0.9181873351335526,
      "num_tokens": 30111148.0,
      "step": 2441
    },
    {
      "entropy": 1.053389072418213,
      "epoch": 1.2859399684044233,
      "grad_norm": 0.2999892830848694,
      "learning_rate": 0.00018908883119067314,
      "loss": 0.1986619383096695,
      "mean_token_accuracy": 0.9188312739133835,
      "num_tokens": 30123484.0,
      "step": 2442
    },
    {
      "entropy": 1.055832028388977,
      "epoch": 1.286466561348078,
      "grad_norm": 0.31872841715812683,
      "learning_rate": 0.00018907836345595234,
      "loss": 0.22551697492599487,
      "mean_token_accuracy": 0.9126669466495514,
      "num_tokens": 30135820.0,
      "step": 2443
    },
    {
      "entropy": 1.0761081874370575,
      "epoch": 1.2869931542917326,
      "grad_norm": 0.28309813141822815,
      "learning_rate": 0.00018906789102680996,
      "loss": 0.20621955394744873,
      "mean_token_accuracy": 0.9136619865894318,
      "num_tokens": 30148156.0,
      "step": 2444
    },
    {
      "entropy": 1.0669479370117188,
      "epoch": 1.287519747235387,
      "grad_norm": 0.2921207845211029,
      "learning_rate": 0.0001890574139038678,
      "loss": 0.19345825910568237,
      "mean_token_accuracy": 0.9210665971040726,
      "num_tokens": 30160492.0,
      "step": 2445
    },
    {
      "entropy": 1.082845777273178,
      "epoch": 1.2880463401790416,
      "grad_norm": 0.3050430715084076,
      "learning_rate": 0.00018904693208774773,
      "loss": 0.20412419736385345,
      "mean_token_accuracy": 0.9169301390647888,
      "num_tokens": 30172828.0,
      "step": 2446
    },
    {
      "entropy": 1.037044733762741,
      "epoch": 1.2885729331226963,
      "grad_norm": 0.2966107130050659,
      "learning_rate": 0.00018903644557907202,
      "loss": 0.2055833488702774,
      "mean_token_accuracy": 0.9147552996873856,
      "num_tokens": 30185164.0,
      "step": 2447
    },
    {
      "entropy": 1.0677853226661682,
      "epoch": 1.2890995260663507,
      "grad_norm": 0.2843310534954071,
      "learning_rate": 0.0001890259543784632,
      "loss": 0.21242734789848328,
      "mean_token_accuracy": 0.9103299677371979,
      "num_tokens": 30197500.0,
      "step": 2448
    },
    {
      "entropy": 1.077808290719986,
      "epoch": 1.2896261190100053,
      "grad_norm": 0.3029589056968689,
      "learning_rate": 0.00018901545848654406,
      "loss": 0.21383321285247803,
      "mean_token_accuracy": 0.9146206676959991,
      "num_tokens": 30209836.0,
      "step": 2449
    },
    {
      "entropy": 1.067981243133545,
      "epoch": 1.2901527119536598,
      "grad_norm": 0.26611074805259705,
      "learning_rate": 0.0001890049579039377,
      "loss": 0.2002156376838684,
      "mean_token_accuracy": 0.9168877750635147,
      "num_tokens": 30222172.0,
      "step": 2450
    },
    {
      "entropy": 1.0742313861846924,
      "epoch": 1.2906793048973144,
      "grad_norm": 0.310247540473938,
      "learning_rate": 0.00018899445263126748,
      "loss": 0.20398539304733276,
      "mean_token_accuracy": 0.914374902844429,
      "num_tokens": 30234508.0,
      "step": 2451
    },
    {
      "entropy": 1.0848418474197388,
      "epoch": 1.2912058978409688,
      "grad_norm": 0.2897074818611145,
      "learning_rate": 0.000188983942669157,
      "loss": 0.19188785552978516,
      "mean_token_accuracy": 0.9222336858510971,
      "num_tokens": 30246844.0,
      "step": 2452
    },
    {
      "entropy": 1.1648883521556854,
      "epoch": 1.2917324907846235,
      "grad_norm": 0.3317052125930786,
      "learning_rate": 0.00018897342801823017,
      "loss": 0.21206307411193848,
      "mean_token_accuracy": 0.9162902683019638,
      "num_tokens": 30259180.0,
      "step": 2453
    },
    {
      "entropy": 1.089240550994873,
      "epoch": 1.2922590837282781,
      "grad_norm": 0.29545658826828003,
      "learning_rate": 0.00018896290867911124,
      "loss": 0.224837988615036,
      "mean_token_accuracy": 0.9070383608341217,
      "num_tokens": 30271516.0,
      "step": 2454
    },
    {
      "entropy": 1.0672079622745514,
      "epoch": 1.2927856766719326,
      "grad_norm": 0.28104934096336365,
      "learning_rate": 0.00018895238465242465,
      "loss": 0.18683423101902008,
      "mean_token_accuracy": 0.9193482846021652,
      "num_tokens": 30283852.0,
      "step": 2455
    },
    {
      "entropy": 1.109650194644928,
      "epoch": 1.2933122696155872,
      "grad_norm": 0.30402952432632446,
      "learning_rate": 0.00018894185593879513,
      "loss": 0.21138614416122437,
      "mean_token_accuracy": 0.9192399084568024,
      "num_tokens": 30296188.0,
      "step": 2456
    },
    {
      "entropy": 1.0995548069477081,
      "epoch": 1.2938388625592416,
      "grad_norm": 0.2972696125507355,
      "learning_rate": 0.00018893132253884777,
      "loss": 0.21770839393138885,
      "mean_token_accuracy": 0.9064899533987045,
      "num_tokens": 30308524.0,
      "step": 2457
    },
    {
      "entropy": 1.0900613367557526,
      "epoch": 1.2943654555028963,
      "grad_norm": 0.29551947116851807,
      "learning_rate": 0.00018892078445320785,
      "loss": 0.22550401091575623,
      "mean_token_accuracy": 0.9079246073961258,
      "num_tokens": 30320860.0,
      "step": 2458
    },
    {
      "entropy": 1.0972884893417358,
      "epoch": 1.2948920484465507,
      "grad_norm": 0.29107943177223206,
      "learning_rate": 0.0001889102416825009,
      "loss": 0.20210495591163635,
      "mean_token_accuracy": 0.918797492980957,
      "num_tokens": 30333196.0,
      "step": 2459
    },
    {
      "entropy": 1.1345873475074768,
      "epoch": 1.2954186413902053,
      "grad_norm": 0.2670139968395233,
      "learning_rate": 0.00018889969422735282,
      "loss": 0.2063610702753067,
      "mean_token_accuracy": 0.9150100499391556,
      "num_tokens": 30345532.0,
      "step": 2460
    },
    {
      "entropy": 1.1534136831760406,
      "epoch": 1.29594523433386,
      "grad_norm": 0.2725052535533905,
      "learning_rate": 0.00018888914208838977,
      "loss": 0.20247100293636322,
      "mean_token_accuracy": 0.9132510721683502,
      "num_tokens": 30357868.0,
      "step": 2461
    },
    {
      "entropy": 1.1401869356632233,
      "epoch": 1.2964718272775144,
      "grad_norm": 0.27257704734802246,
      "learning_rate": 0.00018887858526623816,
      "loss": 0.20108935236930847,
      "mean_token_accuracy": 0.9205031096935272,
      "num_tokens": 30370204.0,
      "step": 2462
    },
    {
      "entropy": 1.1363012492656708,
      "epoch": 1.296998420221169,
      "grad_norm": 0.25703343749046326,
      "learning_rate": 0.00018886802376152467,
      "loss": 0.1905408352613449,
      "mean_token_accuracy": 0.9216178208589554,
      "num_tokens": 30382540.0,
      "step": 2463
    },
    {
      "entropy": 1.1291438043117523,
      "epoch": 1.2975250131648237,
      "grad_norm": 0.26567402482032776,
      "learning_rate": 0.00018885745757487633,
      "loss": 0.2002328634262085,
      "mean_token_accuracy": 0.9188759624958038,
      "num_tokens": 30394876.0,
      "step": 2464
    },
    {
      "entropy": 1.1460005939006805,
      "epoch": 1.2980516061084781,
      "grad_norm": 0.2740042805671692,
      "learning_rate": 0.0001888468867069203,
      "loss": 0.2134706676006317,
      "mean_token_accuracy": 0.9186908006668091,
      "num_tokens": 30407212.0,
      "step": 2465
    },
    {
      "entropy": 1.1675595045089722,
      "epoch": 1.2985781990521326,
      "grad_norm": 0.27058151364326477,
      "learning_rate": 0.0001888363111582842,
      "loss": 0.20257370173931122,
      "mean_token_accuracy": 0.9163499921560287,
      "num_tokens": 30419548.0,
      "step": 2466
    },
    {
      "entropy": 1.129315197467804,
      "epoch": 1.2991047919957872,
      "grad_norm": 0.2428874373435974,
      "learning_rate": 0.00018882573092959576,
      "loss": 0.19501042366027832,
      "mean_token_accuracy": 0.9155511409044266,
      "num_tokens": 30431884.0,
      "step": 2467
    },
    {
      "entropy": 1.1659575700759888,
      "epoch": 1.2996313849394419,
      "grad_norm": 0.2721167504787445,
      "learning_rate": 0.00018881514602148313,
      "loss": 0.19770276546478271,
      "mean_token_accuracy": 0.9158978760242462,
      "num_tokens": 30444220.0,
      "step": 2468
    },
    {
      "entropy": 1.112877070903778,
      "epoch": 1.3001579778830963,
      "grad_norm": 0.2472962588071823,
      "learning_rate": 0.00018880455643457465,
      "loss": 0.19223164021968842,
      "mean_token_accuracy": 0.924824669957161,
      "num_tokens": 30456556.0,
      "step": 2469
    },
    {
      "entropy": 1.1800079345703125,
      "epoch": 1.300684570826751,
      "grad_norm": 0.2697281837463379,
      "learning_rate": 0.00018879396216949895,
      "loss": 0.1964273303747177,
      "mean_token_accuracy": 0.9209067970514297,
      "num_tokens": 30468892.0,
      "step": 2470
    },
    {
      "entropy": 1.163857877254486,
      "epoch": 1.3012111637704056,
      "grad_norm": 0.3045770227909088,
      "learning_rate": 0.00018878336322688494,
      "loss": 0.22345393896102905,
      "mean_token_accuracy": 0.9061350077390671,
      "num_tokens": 30481228.0,
      "step": 2471
    },
    {
      "entropy": 1.14396533370018,
      "epoch": 1.30173775671406,
      "grad_norm": 0.28983190655708313,
      "learning_rate": 0.0001887727596073619,
      "loss": 0.22206184267997742,
      "mean_token_accuracy": 0.9082365930080414,
      "num_tokens": 30493564.0,
      "step": 2472
    },
    {
      "entropy": 1.1775724589824677,
      "epoch": 1.3022643496577146,
      "grad_norm": 0.2802092432975769,
      "learning_rate": 0.0001887621513115592,
      "loss": 0.18831028044223785,
      "mean_token_accuracy": 0.9222876578569412,
      "num_tokens": 30505900.0,
      "step": 2473
    },
    {
      "entropy": 1.1594789624214172,
      "epoch": 1.302790942601369,
      "grad_norm": 0.28708985447883606,
      "learning_rate": 0.00018875153834010665,
      "loss": 0.19792331755161285,
      "mean_token_accuracy": 0.9184091091156006,
      "num_tokens": 30518236.0,
      "step": 2474
    },
    {
      "entropy": 1.1835083365440369,
      "epoch": 1.3033175355450237,
      "grad_norm": 0.2943269610404968,
      "learning_rate": 0.00018874092069363427,
      "loss": 0.2158336043357849,
      "mean_token_accuracy": 0.9099272340536118,
      "num_tokens": 30530572.0,
      "step": 2475
    },
    {
      "entropy": 1.1280877888202667,
      "epoch": 1.3038441284886781,
      "grad_norm": 0.29678258299827576,
      "learning_rate": 0.00018873029837277236,
      "loss": 0.2089228481054306,
      "mean_token_accuracy": 0.9140326231718063,
      "num_tokens": 30542908.0,
      "step": 2476
    },
    {
      "entropy": 1.1274373829364777,
      "epoch": 1.3043707214323328,
      "grad_norm": 0.2565828561782837,
      "learning_rate": 0.0001887196713781515,
      "loss": 0.1824226826429367,
      "mean_token_accuracy": 0.9261182844638824,
      "num_tokens": 30555244.0,
      "step": 2477
    },
    {
      "entropy": 1.1345534920692444,
      "epoch": 1.3048973143759874,
      "grad_norm": 0.29827621579170227,
      "learning_rate": 0.0001887090397104026,
      "loss": 0.20668740570545197,
      "mean_token_accuracy": 0.9158112704753876,
      "num_tokens": 30567580.0,
      "step": 2478
    },
    {
      "entropy": 1.1208145916461945,
      "epoch": 1.3054239073196419,
      "grad_norm": 0.31276044249534607,
      "learning_rate": 0.00018869840337015675,
      "loss": 0.2087455838918686,
      "mean_token_accuracy": 0.9144948124885559,
      "num_tokens": 30579916.0,
      "step": 2479
    },
    {
      "entropy": 1.153564065694809,
      "epoch": 1.3059505002632965,
      "grad_norm": 0.29550793766975403,
      "learning_rate": 0.00018868776235804536,
      "loss": 0.213548481464386,
      "mean_token_accuracy": 0.9138739556074142,
      "num_tokens": 30592252.0,
      "step": 2480
    },
    {
      "entropy": 1.1374950110912323,
      "epoch": 1.3064770932069512,
      "grad_norm": 0.27735957503318787,
      "learning_rate": 0.00018867711667470017,
      "loss": 0.19929325580596924,
      "mean_token_accuracy": 0.9164140969514847,
      "num_tokens": 30604588.0,
      "step": 2481
    },
    {
      "entropy": 1.1209405362606049,
      "epoch": 1.3070036861506056,
      "grad_norm": 0.2788814902305603,
      "learning_rate": 0.0001886664663207531,
      "loss": 0.21787486970424652,
      "mean_token_accuracy": 0.9160508364439011,
      "num_tokens": 30616924.0,
      "step": 2482
    },
    {
      "entropy": 1.1751077771186829,
      "epoch": 1.30753027909426,
      "grad_norm": 0.27281054854393005,
      "learning_rate": 0.00018865581129683644,
      "loss": 0.20231208205223083,
      "mean_token_accuracy": 0.9140424728393555,
      "num_tokens": 30629260.0,
      "step": 2483
    },
    {
      "entropy": 1.1451273262500763,
      "epoch": 1.3080568720379147,
      "grad_norm": 0.2548876404762268,
      "learning_rate": 0.00018864515160358272,
      "loss": 0.20404735207557678,
      "mean_token_accuracy": 0.9158653914928436,
      "num_tokens": 30641596.0,
      "step": 2484
    },
    {
      "entropy": 1.223433405160904,
      "epoch": 1.3085834649815693,
      "grad_norm": 0.28999221324920654,
      "learning_rate": 0.0001886344872416247,
      "loss": 0.21280592679977417,
      "mean_token_accuracy": 0.908152237534523,
      "num_tokens": 30653932.0,
      "step": 2485
    },
    {
      "entropy": 1.2113437056541443,
      "epoch": 1.3091100579252237,
      "grad_norm": 0.2526291012763977,
      "learning_rate": 0.00018862381821159552,
      "loss": 0.18035513162612915,
      "mean_token_accuracy": 0.9277927577495575,
      "num_tokens": 30666268.0,
      "step": 2486
    },
    {
      "entropy": 1.2897081971168518,
      "epoch": 1.3096366508688784,
      "grad_norm": 0.2631675899028778,
      "learning_rate": 0.0001886131445141285,
      "loss": 0.2015797197818756,
      "mean_token_accuracy": 0.918338343501091,
      "num_tokens": 30678604.0,
      "step": 2487
    },
    {
      "entropy": 1.2078757286071777,
      "epoch": 1.310163243812533,
      "grad_norm": 0.24873696267604828,
      "learning_rate": 0.00018860246614985725,
      "loss": 0.18720902502536774,
      "mean_token_accuracy": 0.9217247664928436,
      "num_tokens": 30690940.0,
      "step": 2488
    },
    {
      "entropy": 1.2295198142528534,
      "epoch": 1.3106898367561874,
      "grad_norm": 0.27777987718582153,
      "learning_rate": 0.00018859178311941575,
      "loss": 0.21630996465682983,
      "mean_token_accuracy": 0.9144271612167358,
      "num_tokens": 30703276.0,
      "step": 2489
    },
    {
      "entropy": 1.2049448788166046,
      "epoch": 1.311216429699842,
      "grad_norm": 0.26715347170829773,
      "learning_rate": 0.00018858109542343814,
      "loss": 0.19998157024383545,
      "mean_token_accuracy": 0.9232079237699509,
      "num_tokens": 30715612.0,
      "step": 2490
    },
    {
      "entropy": 1.1993949711322784,
      "epoch": 1.3117430226434965,
      "grad_norm": 0.2785743474960327,
      "learning_rate": 0.00018857040306255888,
      "loss": 0.20864132046699524,
      "mean_token_accuracy": 0.9150129109621048,
      "num_tokens": 30727948.0,
      "step": 2491
    },
    {
      "entropy": 1.197268307209015,
      "epoch": 1.3122696155871512,
      "grad_norm": 0.29957106709480286,
      "learning_rate": 0.0001885597060374128,
      "loss": 0.2152678519487381,
      "mean_token_accuracy": 0.9146653860807419,
      "num_tokens": 30740284.0,
      "step": 2492
    },
    {
      "entropy": 1.1922446489334106,
      "epoch": 1.3127962085308056,
      "grad_norm": 0.31052011251449585,
      "learning_rate": 0.00018854900434863477,
      "loss": 0.20205189287662506,
      "mean_token_accuracy": 0.9173094034194946,
      "num_tokens": 30752620.0,
      "step": 2493
    },
    {
      "entropy": 1.1869801878929138,
      "epoch": 1.3133228014744602,
      "grad_norm": 0.2926376760005951,
      "learning_rate": 0.0001885382979968602,
      "loss": 0.21089938282966614,
      "mean_token_accuracy": 0.9158413708209991,
      "num_tokens": 30764956.0,
      "step": 2494
    },
    {
      "entropy": 1.132896363735199,
      "epoch": 1.3138493944181149,
      "grad_norm": 0.27938100695610046,
      "learning_rate": 0.00018852758698272465,
      "loss": 0.1964457780122757,
      "mean_token_accuracy": 0.9246141463518143,
      "num_tokens": 30777292.0,
      "step": 2495
    },
    {
      "entropy": 1.2086142599582672,
      "epoch": 1.3143759873617693,
      "grad_norm": 0.28923818469047546,
      "learning_rate": 0.00018851687130686396,
      "loss": 0.2008797824382782,
      "mean_token_accuracy": 0.916870579123497,
      "num_tokens": 30789628.0,
      "step": 2496
    },
    {
      "entropy": 1.1438094079494476,
      "epoch": 1.314902580305424,
      "grad_norm": 0.2761872708797455,
      "learning_rate": 0.00018850615096991424,
      "loss": 0.19743943214416504,
      "mean_token_accuracy": 0.9220564365386963,
      "num_tokens": 30801964.0,
      "step": 2497
    },
    {
      "entropy": 1.1207115054130554,
      "epoch": 1.3154291732490786,
      "grad_norm": 0.2810716927051544,
      "learning_rate": 0.0001884954259725119,
      "loss": 0.20330996811389923,
      "mean_token_accuracy": 0.9163385629653931,
      "num_tokens": 30814300.0,
      "step": 2498
    },
    {
      "entropy": 1.1683085262775421,
      "epoch": 1.315955766192733,
      "grad_norm": 0.2724354565143585,
      "learning_rate": 0.00018848469631529364,
      "loss": 0.18855315446853638,
      "mean_token_accuracy": 0.9251711666584015,
      "num_tokens": 30826636.0,
      "step": 2499
    },
    {
      "entropy": 1.1003861129283905,
      "epoch": 1.3164823591363874,
      "grad_norm": 0.2779734432697296,
      "learning_rate": 0.00018847396199889638,
      "loss": 0.20665448904037476,
      "mean_token_accuracy": 0.9169312715530396,
      "num_tokens": 30838972.0,
      "step": 2500
    },
    {
      "entropy": 1.1359674334526062,
      "epoch": 1.317008952080042,
      "grad_norm": 0.28787779808044434,
      "learning_rate": 0.0001884632230239574,
      "loss": 0.19744256138801575,
      "mean_token_accuracy": 0.9244921803474426,
      "num_tokens": 30851308.0,
      "step": 2501
    },
    {
      "entropy": 1.1084416806697845,
      "epoch": 1.3175355450236967,
      "grad_norm": 0.2643449306488037,
      "learning_rate": 0.00018845247939111418,
      "loss": 0.19814109802246094,
      "mean_token_accuracy": 0.9191816598176956,
      "num_tokens": 30863644.0,
      "step": 2502
    },
    {
      "entropy": 1.1480841636657715,
      "epoch": 1.3180621379673512,
      "grad_norm": 0.2745157778263092,
      "learning_rate": 0.00018844173110100452,
      "loss": 0.20436102151870728,
      "mean_token_accuracy": 0.9175873845815659,
      "num_tokens": 30875980.0,
      "step": 2503
    },
    {
      "entropy": 1.168861299753189,
      "epoch": 1.3185887309110058,
      "grad_norm": 0.28558439016342163,
      "learning_rate": 0.00018843097815426644,
      "loss": 0.19720155000686646,
      "mean_token_accuracy": 0.921770378947258,
      "num_tokens": 30888316.0,
      "step": 2504
    },
    {
      "entropy": 1.1487717032432556,
      "epoch": 1.3191153238546605,
      "grad_norm": 0.252396821975708,
      "learning_rate": 0.00018842022055153837,
      "loss": 0.18812143802642822,
      "mean_token_accuracy": 0.9198298454284668,
      "num_tokens": 30900652.0,
      "step": 2505
    },
    {
      "entropy": 1.0874556303024292,
      "epoch": 1.3196419167983149,
      "grad_norm": 0.27975741028785706,
      "learning_rate": 0.00018840945829345885,
      "loss": 0.19726543128490448,
      "mean_token_accuracy": 0.9175487458705902,
      "num_tokens": 30912988.0,
      "step": 2506
    },
    {
      "entropy": 1.0931753516197205,
      "epoch": 1.3201685097419695,
      "grad_norm": 0.2753717005252838,
      "learning_rate": 0.0001883986913806668,
      "loss": 0.19207552075386047,
      "mean_token_accuracy": 0.9233367294073105,
      "num_tokens": 30925324.0,
      "step": 2507
    },
    {
      "entropy": 1.0678606629371643,
      "epoch": 1.320695102685624,
      "grad_norm": 0.2772499620914459,
      "learning_rate": 0.00018838791981380136,
      "loss": 0.20343419909477234,
      "mean_token_accuracy": 0.9183335304260254,
      "num_tokens": 30937660.0,
      "step": 2508
    },
    {
      "entropy": 1.0918059051036835,
      "epoch": 1.3212216956292786,
      "grad_norm": 0.29579028487205505,
      "learning_rate": 0.00018837714359350201,
      "loss": 0.2152230143547058,
      "mean_token_accuracy": 0.9208053946495056,
      "num_tokens": 30949996.0,
      "step": 2509
    },
    {
      "entropy": 1.0503469556570053,
      "epoch": 1.321748288572933,
      "grad_norm": 0.2653261423110962,
      "learning_rate": 0.0001883663627204085,
      "loss": 0.1930423229932785,
      "mean_token_accuracy": 0.9191714227199554,
      "num_tokens": 30962332.0,
      "step": 2510
    },
    {
      "entropy": 1.1172581315040588,
      "epoch": 1.3222748815165877,
      "grad_norm": 0.3529946208000183,
      "learning_rate": 0.00018835557719516078,
      "loss": 0.22693800926208496,
      "mean_token_accuracy": 0.9077934175729752,
      "num_tokens": 30974668.0,
      "step": 2511
    },
    {
      "entropy": 1.0955396890640259,
      "epoch": 1.3228014744602423,
      "grad_norm": 0.2981196343898773,
      "learning_rate": 0.0001883447870183991,
      "loss": 0.19838613271713257,
      "mean_token_accuracy": 0.9167172312736511,
      "num_tokens": 30987004.0,
      "step": 2512
    },
    {
      "entropy": 1.0746179521083832,
      "epoch": 1.3233280674038967,
      "grad_norm": 0.30476272106170654,
      "learning_rate": 0.00018833399219076407,
      "loss": 0.21524569392204285,
      "mean_token_accuracy": 0.9110986590385437,
      "num_tokens": 30999340.0,
      "step": 2513
    },
    {
      "entropy": 1.1119599342346191,
      "epoch": 1.3238546603475514,
      "grad_norm": 0.3129141628742218,
      "learning_rate": 0.00018832319271289646,
      "loss": 0.20734190940856934,
      "mean_token_accuracy": 0.9160514175891876,
      "num_tokens": 31011676.0,
      "step": 2514
    },
    {
      "entropy": 1.0783126950263977,
      "epoch": 1.3243812532912058,
      "grad_norm": 0.2936292588710785,
      "learning_rate": 0.00018831238858543742,
      "loss": 0.20680686831474304,
      "mean_token_accuracy": 0.9129305332899094,
      "num_tokens": 31024012.0,
      "step": 2515
    },
    {
      "entropy": 1.0885819494724274,
      "epoch": 1.3249078462348605,
      "grad_norm": 0.29339733719825745,
      "learning_rate": 0.00018830157980902833,
      "loss": 0.20416975021362305,
      "mean_token_accuracy": 0.9174858927726746,
      "num_tokens": 31036348.0,
      "step": 2516
    },
    {
      "entropy": 1.1106742322444916,
      "epoch": 1.325434439178515,
      "grad_norm": 0.3055555522441864,
      "learning_rate": 0.00018829076638431078,
      "loss": 0.2278045117855072,
      "mean_token_accuracy": 0.9134476184844971,
      "num_tokens": 31048684.0,
      "step": 2517
    },
    {
      "entropy": 1.1332001090049744,
      "epoch": 1.3259610321221695,
      "grad_norm": 0.27308133244514465,
      "learning_rate": 0.00018827994831192675,
      "loss": 0.19392089545726776,
      "mean_token_accuracy": 0.9187697917222977,
      "num_tokens": 31061020.0,
      "step": 2518
    },
    {
      "entropy": 1.196645975112915,
      "epoch": 1.3264876250658242,
      "grad_norm": 0.3041642904281616,
      "learning_rate": 0.00018826912559251848,
      "loss": 0.23238268494606018,
      "mean_token_accuracy": 0.9100022912025452,
      "num_tokens": 31073356.0,
      "step": 2519
    },
    {
      "entropy": 1.1595803797245026,
      "epoch": 1.3270142180094786,
      "grad_norm": 0.2854364812374115,
      "learning_rate": 0.00018825829822672837,
      "loss": 0.21763843297958374,
      "mean_token_accuracy": 0.911407396197319,
      "num_tokens": 31085692.0,
      "step": 2520
    },
    {
      "entropy": 1.1426350474357605,
      "epoch": 1.3275408109531333,
      "grad_norm": 0.26290595531463623,
      "learning_rate": 0.00018824746621519923,
      "loss": 0.18908298015594482,
      "mean_token_accuracy": 0.9210202246904373,
      "num_tokens": 31098028.0,
      "step": 2521
    },
    {
      "entropy": 1.1872995793819427,
      "epoch": 1.328067403896788,
      "grad_norm": 0.2785587012767792,
      "learning_rate": 0.00018823662955857407,
      "loss": 0.20560133457183838,
      "mean_token_accuracy": 0.9162561148405075,
      "num_tokens": 31110364.0,
      "step": 2522
    },
    {
      "entropy": 1.1389261186122894,
      "epoch": 1.3285939968404423,
      "grad_norm": 0.2547476291656494,
      "learning_rate": 0.00018822578825749623,
      "loss": 0.20492050051689148,
      "mean_token_accuracy": 0.9174861460924149,
      "num_tokens": 31122700.0,
      "step": 2523
    },
    {
      "entropy": 1.1379226744174957,
      "epoch": 1.329120589784097,
      "grad_norm": 0.2671321630477905,
      "learning_rate": 0.0001882149423126093,
      "loss": 0.21442070603370667,
      "mean_token_accuracy": 0.916014552116394,
      "num_tokens": 31135036.0,
      "step": 2524
    },
    {
      "entropy": 1.1240211725234985,
      "epoch": 1.3296471827277514,
      "grad_norm": 0.28270992636680603,
      "learning_rate": 0.00018820409172455707,
      "loss": 0.20910367369651794,
      "mean_token_accuracy": 0.9152929037809372,
      "num_tokens": 31147372.0,
      "step": 2525
    },
    {
      "entropy": 1.126050055027008,
      "epoch": 1.330173775671406,
      "grad_norm": 0.2661881148815155,
      "learning_rate": 0.00018819323649398373,
      "loss": 0.19098810851573944,
      "mean_token_accuracy": 0.9209559559822083,
      "num_tokens": 31159708.0,
      "step": 2526
    },
    {
      "entropy": 1.1266259253025055,
      "epoch": 1.3307003686150605,
      "grad_norm": 0.2929874658584595,
      "learning_rate": 0.0001881823766215337,
      "loss": 0.20036344230175018,
      "mean_token_accuracy": 0.9238387793302536,
      "num_tokens": 31172044.0,
      "step": 2527
    },
    {
      "entropy": 1.0866254270076752,
      "epoch": 1.3312269615587151,
      "grad_norm": 0.28896623849868774,
      "learning_rate": 0.00018817151210785162,
      "loss": 0.18842336535453796,
      "mean_token_accuracy": 0.9209622591733932,
      "num_tokens": 31184380.0,
      "step": 2528
    },
    {
      "entropy": 1.1249730587005615,
      "epoch": 1.3317535545023698,
      "grad_norm": 0.2872648537158966,
      "learning_rate": 0.0001881606429535825,
      "loss": 0.22201552987098694,
      "mean_token_accuracy": 0.9108192026615143,
      "num_tokens": 31196716.0,
      "step": 2529
    },
    {
      "entropy": 1.0949527323246002,
      "epoch": 1.3322801474460242,
      "grad_norm": 0.2851123511791229,
      "learning_rate": 0.0001881497691593716,
      "loss": 0.1967121958732605,
      "mean_token_accuracy": 0.9210963398218155,
      "num_tokens": 31209052.0,
      "step": 2530
    },
    {
      "entropy": 1.1146687865257263,
      "epoch": 1.3328067403896788,
      "grad_norm": 0.3102349638938904,
      "learning_rate": 0.00018813889072586434,
      "loss": 0.23011735081672668,
      "mean_token_accuracy": 0.9073590189218521,
      "num_tokens": 31221388.0,
      "step": 2531
    },
    {
      "entropy": 1.0698721408843994,
      "epoch": 1.3333333333333333,
      "grad_norm": 0.2832903563976288,
      "learning_rate": 0.0001881280076537066,
      "loss": 0.1880122423171997,
      "mean_token_accuracy": 0.9258202612400055,
      "num_tokens": 31233724.0,
      "step": 2532
    },
    {
      "entropy": 1.1224130392074585,
      "epoch": 1.333859926276988,
      "grad_norm": 0.2848287522792816,
      "learning_rate": 0.0001881171199435444,
      "loss": 0.19834977388381958,
      "mean_token_accuracy": 0.9168813228607178,
      "num_tokens": 31246060.0,
      "step": 2533
    },
    {
      "entropy": 1.1105233430862427,
      "epoch": 1.3343865192206423,
      "grad_norm": 0.3011896014213562,
      "learning_rate": 0.0001881062275960241,
      "loss": 0.22225533425807953,
      "mean_token_accuracy": 0.9092197567224503,
      "num_tokens": 31258396.0,
      "step": 2534
    },
    {
      "entropy": 1.0629687905311584,
      "epoch": 1.334913112164297,
      "grad_norm": 0.2885998785495758,
      "learning_rate": 0.0001880953306117923,
      "loss": 0.2081342339515686,
      "mean_token_accuracy": 0.9201711565256119,
      "num_tokens": 31270732.0,
      "step": 2535
    },
    {
      "entropy": 1.0981874465942383,
      "epoch": 1.3354397051079516,
      "grad_norm": 0.2881809175014496,
      "learning_rate": 0.0001880844289914959,
      "loss": 0.20444408059120178,
      "mean_token_accuracy": 0.9140055626630783,
      "num_tokens": 31283068.0,
      "step": 2536
    },
    {
      "entropy": 1.0878173410892487,
      "epoch": 1.335966298051606,
      "grad_norm": 0.2887677252292633,
      "learning_rate": 0.00018807352273578204,
      "loss": 0.20989957451820374,
      "mean_token_accuracy": 0.9208216220140457,
      "num_tokens": 31295404.0,
      "step": 2537
    },
    {
      "entropy": 1.1007463037967682,
      "epoch": 1.3364928909952607,
      "grad_norm": 0.2962489426136017,
      "learning_rate": 0.0001880626118452982,
      "loss": 0.20988334715366364,
      "mean_token_accuracy": 0.9117045700550079,
      "num_tokens": 31307740.0,
      "step": 2538
    },
    {
      "entropy": 1.1124551892280579,
      "epoch": 1.3370194839389153,
      "grad_norm": 0.2889869809150696,
      "learning_rate": 0.00018805169632069206,
      "loss": 0.22036269307136536,
      "mean_token_accuracy": 0.9166446030139923,
      "num_tokens": 31320076.0,
      "step": 2539
    },
    {
      "entropy": 1.0473781377077103,
      "epoch": 1.3375460768825698,
      "grad_norm": 0.28756895661354065,
      "learning_rate": 0.00018804077616261166,
      "loss": 0.20982427895069122,
      "mean_token_accuracy": 0.9156763106584549,
      "num_tokens": 31332412.0,
      "step": 2540
    },
    {
      "entropy": 1.0588495880365372,
      "epoch": 1.3380726698262242,
      "grad_norm": 0.2706986665725708,
      "learning_rate": 0.00018802985137170522,
      "loss": 0.2128053605556488,
      "mean_token_accuracy": 0.9166304022073746,
      "num_tokens": 31344748.0,
      "step": 2541
    },
    {
      "entropy": 1.075078159570694,
      "epoch": 1.3385992627698788,
      "grad_norm": 0.2772495746612549,
      "learning_rate": 0.0001880189219486213,
      "loss": 0.1886967122554779,
      "mean_token_accuracy": 0.9169599562883377,
      "num_tokens": 31357084.0,
      "step": 2542
    },
    {
      "entropy": 1.0650361776351929,
      "epoch": 1.3391258557135335,
      "grad_norm": 0.3137603998184204,
      "learning_rate": 0.0001880079878940087,
      "loss": 0.20629137754440308,
      "mean_token_accuracy": 0.919654980301857,
      "num_tokens": 31369420.0,
      "step": 2543
    },
    {
      "entropy": 1.0738177597522736,
      "epoch": 1.339652448657188,
      "grad_norm": 0.2894216775894165,
      "learning_rate": 0.00018799704920851652,
      "loss": 0.22449420392513275,
      "mean_token_accuracy": 0.9042561650276184,
      "num_tokens": 31381756.0,
      "step": 2544
    },
    {
      "entropy": 1.0527926087379456,
      "epoch": 1.3401790416008426,
      "grad_norm": 0.2740827202796936,
      "learning_rate": 0.00018798610589279417,
      "loss": 0.203867569565773,
      "mean_token_accuracy": 0.9112063497304916,
      "num_tokens": 31394092.0,
      "step": 2545
    },
    {
      "entropy": 1.0129956007003784,
      "epoch": 1.3407056345444972,
      "grad_norm": 0.2815206050872803,
      "learning_rate": 0.00018797515794749122,
      "loss": 0.18543724715709686,
      "mean_token_accuracy": 0.9268083721399307,
      "num_tokens": 31406428.0,
      "step": 2546
    },
    {
      "entropy": 1.032931998372078,
      "epoch": 1.3412322274881516,
      "grad_norm": 0.2881038784980774,
      "learning_rate": 0.00018796420537325762,
      "loss": 0.21719786524772644,
      "mean_token_accuracy": 0.918227955698967,
      "num_tokens": 31418764.0,
      "step": 2547
    },
    {
      "entropy": 1.0120973587036133,
      "epoch": 1.3417588204318063,
      "grad_norm": 0.27740973234176636,
      "learning_rate": 0.00018795324817074354,
      "loss": 0.18840999901294708,
      "mean_token_accuracy": 0.922258198261261,
      "num_tokens": 31431100.0,
      "step": 2548
    },
    {
      "entropy": 1.0631686747074127,
      "epoch": 1.3422854133754607,
      "grad_norm": 0.3168363869190216,
      "learning_rate": 0.0001879422863405995,
      "loss": 0.21135640144348145,
      "mean_token_accuracy": 0.9179410338401794,
      "num_tokens": 31443436.0,
      "step": 2549
    },
    {
      "entropy": 1.0500916242599487,
      "epoch": 1.3428120063191153,
      "grad_norm": 0.29962658882141113,
      "learning_rate": 0.00018793131988347617,
      "loss": 0.2103479504585266,
      "mean_token_accuracy": 0.9147076457738876,
      "num_tokens": 31455772.0,
      "step": 2550
    },
    {
      "entropy": 1.0350684523582458,
      "epoch": 1.3433385992627698,
      "grad_norm": 0.2993636727333069,
      "learning_rate": 0.0001879203488000246,
      "loss": 0.2143697738647461,
      "mean_token_accuracy": 0.9145818799734116,
      "num_tokens": 31468108.0,
      "step": 2551
    },
    {
      "entropy": 1.0683605670928955,
      "epoch": 1.3438651922064244,
      "grad_norm": 0.31891852617263794,
      "learning_rate": 0.0001879093730908961,
      "loss": 0.2051960527896881,
      "mean_token_accuracy": 0.9145154505968094,
      "num_tokens": 31480444.0,
      "step": 2552
    },
    {
      "entropy": 1.0360463559627533,
      "epoch": 1.344391785150079,
      "grad_norm": 0.29179492592811584,
      "learning_rate": 0.0001878983927567422,
      "loss": 0.2059982270002365,
      "mean_token_accuracy": 0.9194035530090332,
      "num_tokens": 31492780.0,
      "step": 2553
    },
    {
      "entropy": 1.093918263912201,
      "epoch": 1.3449183780937335,
      "grad_norm": 0.3079806864261627,
      "learning_rate": 0.0001878874077982147,
      "loss": 0.21530432999134064,
      "mean_token_accuracy": 0.9130852967500687,
      "num_tokens": 31505116.0,
      "step": 2554
    },
    {
      "entropy": 1.1094603836536407,
      "epoch": 1.3454449710373881,
      "grad_norm": 0.2917912006378174,
      "learning_rate": 0.00018787641821596583,
      "loss": 0.2050892859697342,
      "mean_token_accuracy": 0.9154047220945358,
      "num_tokens": 31517452.0,
      "step": 2555
    },
    {
      "entropy": 1.1357096135616302,
      "epoch": 1.3459715639810428,
      "grad_norm": 0.27194029092788696,
      "learning_rate": 0.00018786542401064789,
      "loss": 0.18924210965633392,
      "mean_token_accuracy": 0.9251724630594254,
      "num_tokens": 31529788.0,
      "step": 2556
    },
    {
      "entropy": 1.1269418597221375,
      "epoch": 1.3464981569246972,
      "grad_norm": 0.26160678267478943,
      "learning_rate": 0.00018785442518291353,
      "loss": 0.20647940039634705,
      "mean_token_accuracy": 0.9157357662916183,
      "num_tokens": 31542124.0,
      "step": 2557
    },
    {
      "entropy": 1.1875733733177185,
      "epoch": 1.3470247498683516,
      "grad_norm": 0.2644807696342468,
      "learning_rate": 0.00018784342173341575,
      "loss": 0.1932739019393921,
      "mean_token_accuracy": 0.9194819778203964,
      "num_tokens": 31554460.0,
      "step": 2558
    },
    {
      "entropy": 1.1785165071487427,
      "epoch": 1.3475513428120063,
      "grad_norm": 0.27935707569122314,
      "learning_rate": 0.00018783241366280773,
      "loss": 0.2223643660545349,
      "mean_token_accuracy": 0.9112606793642044,
      "num_tokens": 31566796.0,
      "step": 2559
    },
    {
      "entropy": 1.1744961738586426,
      "epoch": 1.348077935755661,
      "grad_norm": 0.272098183631897,
      "learning_rate": 0.0001878214009717429,
      "loss": 0.2064208835363388,
      "mean_token_accuracy": 0.9138854891061783,
      "num_tokens": 31579132.0,
      "step": 2560
    },
    {
      "entropy": 1.1525810956954956,
      "epoch": 1.3486045286993154,
      "grad_norm": 0.2795564532279968,
      "learning_rate": 0.00018781038366087513,
      "loss": 0.20882105827331543,
      "mean_token_accuracy": 0.9144481867551804,
      "num_tokens": 31591468.0,
      "step": 2561
    },
    {
      "entropy": 1.104248732328415,
      "epoch": 1.34913112164297,
      "grad_norm": 0.23773279786109924,
      "learning_rate": 0.00018779936173085837,
      "loss": 0.18065296113491058,
      "mean_token_accuracy": 0.9290421605110168,
      "num_tokens": 31603804.0,
      "step": 2562
    },
    {
      "entropy": 1.1595117151737213,
      "epoch": 1.3496577145866246,
      "grad_norm": 0.2747121751308441,
      "learning_rate": 0.00018778833518234698,
      "loss": 0.20582349598407745,
      "mean_token_accuracy": 0.912534311413765,
      "num_tokens": 31616140.0,
      "step": 2563
    },
    {
      "entropy": 1.155326247215271,
      "epoch": 1.350184307530279,
      "grad_norm": 0.2732286751270294,
      "learning_rate": 0.00018777730401599549,
      "loss": 0.19955134391784668,
      "mean_token_accuracy": 0.9210360944271088,
      "num_tokens": 31628476.0,
      "step": 2564
    },
    {
      "entropy": 1.1245983242988586,
      "epoch": 1.3507109004739337,
      "grad_norm": 0.24134601652622223,
      "learning_rate": 0.00018776626823245878,
      "loss": 0.18833664059638977,
      "mean_token_accuracy": 0.9211855530738831,
      "num_tokens": 31640812.0,
      "step": 2565
    },
    {
      "entropy": 1.1322357952594757,
      "epoch": 1.3512374934175881,
      "grad_norm": 0.27595973014831543,
      "learning_rate": 0.00018775522783239198,
      "loss": 0.20659556984901428,
      "mean_token_accuracy": 0.921311542391777,
      "num_tokens": 31653148.0,
      "step": 2566
    },
    {
      "entropy": 1.1415637731552124,
      "epoch": 1.3517640863612428,
      "grad_norm": 0.2799292802810669,
      "learning_rate": 0.00018774418281645053,
      "loss": 0.18609049916267395,
      "mean_token_accuracy": 0.9205182492733002,
      "num_tokens": 31665484.0,
      "step": 2567
    },
    {
      "entropy": 1.099078118801117,
      "epoch": 1.3522906793048972,
      "grad_norm": 0.356923907995224,
      "learning_rate": 0.00018773313318529,
      "loss": 0.20763400197029114,
      "mean_token_accuracy": 0.9158278256654739,
      "num_tokens": 31677820.0,
      "step": 2568
    },
    {
      "entropy": 1.1672882437705994,
      "epoch": 1.3528172722485519,
      "grad_norm": 0.323013573884964,
      "learning_rate": 0.00018772207893956647,
      "loss": 0.229105606675148,
      "mean_token_accuracy": 0.9055341184139252,
      "num_tokens": 31690156.0,
      "step": 2569
    },
    {
      "entropy": 1.129069447517395,
      "epoch": 1.3533438651922065,
      "grad_norm": 0.28814175724983215,
      "learning_rate": 0.00018771102007993613,
      "loss": 0.21873462200164795,
      "mean_token_accuracy": 0.9137783348560333,
      "num_tokens": 31702492.0,
      "step": 2570
    },
    {
      "entropy": 1.1057601869106293,
      "epoch": 1.353870458135861,
      "grad_norm": 0.27318117022514343,
      "learning_rate": 0.00018769995660705542,
      "loss": 0.21677207946777344,
      "mean_token_accuracy": 0.9136002212762833,
      "num_tokens": 31714828.0,
      "step": 2571
    },
    {
      "entropy": 1.1042208969593048,
      "epoch": 1.3543970510795156,
      "grad_norm": 0.28476253151893616,
      "learning_rate": 0.0001876888885215812,
      "loss": 0.20920810103416443,
      "mean_token_accuracy": 0.9151055365800858,
      "num_tokens": 31727164.0,
      "step": 2572
    },
    {
      "entropy": 1.133261263370514,
      "epoch": 1.3549236440231702,
      "grad_norm": 0.2653651237487793,
      "learning_rate": 0.00018767781582417043,
      "loss": 0.1804226189851761,
      "mean_token_accuracy": 0.9289367198944092,
      "num_tokens": 31739500.0,
      "step": 2573
    },
    {
      "entropy": 1.1577132642269135,
      "epoch": 1.3554502369668247,
      "grad_norm": 0.27154141664505005,
      "learning_rate": 0.00018766673851548047,
      "loss": 0.18820011615753174,
      "mean_token_accuracy": 0.9214843064546585,
      "num_tokens": 31751836.0,
      "step": 2574
    },
    {
      "entropy": 1.1380637884140015,
      "epoch": 1.355976829910479,
      "grad_norm": 0.2832103371620178,
      "learning_rate": 0.00018765565659616896,
      "loss": 0.2163006067276001,
      "mean_token_accuracy": 0.9146066755056381,
      "num_tokens": 31764172.0,
      "step": 2575
    },
    {
      "entropy": 1.171122670173645,
      "epoch": 1.3565034228541337,
      "grad_norm": 0.29119494557380676,
      "learning_rate": 0.00018764457006689372,
      "loss": 0.1951453536748886,
      "mean_token_accuracy": 0.916601300239563,
      "num_tokens": 31776508.0,
      "step": 2576
    },
    {
      "entropy": 1.1506705582141876,
      "epoch": 1.3570300157977884,
      "grad_norm": 0.28178977966308594,
      "learning_rate": 0.0001876334789283129,
      "loss": 0.21672406792640686,
      "mean_token_accuracy": 0.9112057536840439,
      "num_tokens": 31788844.0,
      "step": 2577
    },
    {
      "entropy": 1.1319689452648163,
      "epoch": 1.3575566087414428,
      "grad_norm": 0.31361258029937744,
      "learning_rate": 0.0001876223831810849,
      "loss": 0.23558440804481506,
      "mean_token_accuracy": 0.9097436219453812,
      "num_tokens": 31801180.0,
      "step": 2578
    },
    {
      "entropy": 1.1845204532146454,
      "epoch": 1.3580832016850974,
      "grad_norm": 0.2968812882900238,
      "learning_rate": 0.00018761128282586843,
      "loss": 0.20870262384414673,
      "mean_token_accuracy": 0.9157723784446716,
      "num_tokens": 31813516.0,
      "step": 2579
    },
    {
      "entropy": 1.1385733485221863,
      "epoch": 1.358609794628752,
      "grad_norm": 0.283302366733551,
      "learning_rate": 0.00018760017786332247,
      "loss": 0.20222227275371552,
      "mean_token_accuracy": 0.9147961437702179,
      "num_tokens": 31825852.0,
      "step": 2580
    },
    {
      "entropy": 1.1306065320968628,
      "epoch": 1.3591363875724065,
      "grad_norm": 0.3002590537071228,
      "learning_rate": 0.0001875890682941062,
      "loss": 0.23286172747612,
      "mean_token_accuracy": 0.9055652320384979,
      "num_tokens": 31838188.0,
      "step": 2581
    },
    {
      "entropy": 1.1369206607341766,
      "epoch": 1.3596629805160612,
      "grad_norm": 0.2827116847038269,
      "learning_rate": 0.00018757795411887922,
      "loss": 0.20414981245994568,
      "mean_token_accuracy": 0.9247657358646393,
      "num_tokens": 31850524.0,
      "step": 2582
    },
    {
      "entropy": 1.182368516921997,
      "epoch": 1.3601895734597156,
      "grad_norm": 0.2734077572822571,
      "learning_rate": 0.00018756683533830123,
      "loss": 0.18725436925888062,
      "mean_token_accuracy": 0.9269359111785889,
      "num_tokens": 31862860.0,
      "step": 2583
    },
    {
      "entropy": 1.1202270686626434,
      "epoch": 1.3607161664033702,
      "grad_norm": 0.27396515011787415,
      "learning_rate": 0.00018755571195303234,
      "loss": 0.19522534310817719,
      "mean_token_accuracy": 0.9229871332645416,
      "num_tokens": 31875196.0,
      "step": 2584
    },
    {
      "entropy": 1.1103180348873138,
      "epoch": 1.3612427593470247,
      "grad_norm": 0.27679014205932617,
      "learning_rate": 0.00018754458396373286,
      "loss": 0.1996929943561554,
      "mean_token_accuracy": 0.9204212427139282,
      "num_tokens": 31887532.0,
      "step": 2585
    },
    {
      "entropy": 1.1616276800632477,
      "epoch": 1.3617693522906793,
      "grad_norm": 0.29670003056526184,
      "learning_rate": 0.00018753345137106336,
      "loss": 0.23619894683361053,
      "mean_token_accuracy": 0.9007562100887299,
      "num_tokens": 31899868.0,
      "step": 2586
    },
    {
      "entropy": 1.1399752795696259,
      "epoch": 1.362295945234334,
      "grad_norm": 0.2779025435447693,
      "learning_rate": 0.0001875223141756848,
      "loss": 0.20615586638450623,
      "mean_token_accuracy": 0.9151979237794876,
      "num_tokens": 31912204.0,
      "step": 2587
    },
    {
      "entropy": 1.167212337255478,
      "epoch": 1.3628225381779884,
      "grad_norm": 0.2890734076499939,
      "learning_rate": 0.00018751117237825825,
      "loss": 0.20611752569675446,
      "mean_token_accuracy": 0.9171196669340134,
      "num_tokens": 31924540.0,
      "step": 2588
    },
    {
      "entropy": 1.126388281583786,
      "epoch": 1.363349131121643,
      "grad_norm": 0.27933791279792786,
      "learning_rate": 0.00018750002597944516,
      "loss": 0.20168091356754303,
      "mean_token_accuracy": 0.9160471111536026,
      "num_tokens": 31936876.0,
      "step": 2589
    },
    {
      "entropy": 1.1319892406463623,
      "epoch": 1.3638757240652974,
      "grad_norm": 0.28239932656288147,
      "learning_rate": 0.00018748887497990727,
      "loss": 0.20251701772212982,
      "mean_token_accuracy": 0.917927622795105,
      "num_tokens": 31949212.0,
      "step": 2590
    },
    {
      "entropy": 1.1169626712799072,
      "epoch": 1.364402317008952,
      "grad_norm": 0.280332088470459,
      "learning_rate": 0.00018747771938030648,
      "loss": 0.18491880595684052,
      "mean_token_accuracy": 0.9226613193750381,
      "num_tokens": 31961548.0,
      "step": 2591
    },
    {
      "entropy": 1.1182208061218262,
      "epoch": 1.3649289099526065,
      "grad_norm": 0.2557113766670227,
      "learning_rate": 0.00018746655918130506,
      "loss": 0.17858107388019562,
      "mean_token_accuracy": 0.9297724813222885,
      "num_tokens": 31973884.0,
      "step": 2592
    },
    {
      "entropy": 1.1194300055503845,
      "epoch": 1.3654555028962612,
      "grad_norm": 0.29216471314430237,
      "learning_rate": 0.00018745539438356557,
      "loss": 0.20672905445098877,
      "mean_token_accuracy": 0.9149999022483826,
      "num_tokens": 31986220.0,
      "step": 2593
    },
    {
      "entropy": 1.1690471470355988,
      "epoch": 1.3659820958399158,
      "grad_norm": 0.29650187492370605,
      "learning_rate": 0.00018744422498775072,
      "loss": 0.22313809394836426,
      "mean_token_accuracy": 0.9091308563947678,
      "num_tokens": 31998556.0,
      "step": 2594
    },
    {
      "entropy": 1.1421934366226196,
      "epoch": 1.3665086887835702,
      "grad_norm": 0.2731371521949768,
      "learning_rate": 0.0001874330509945236,
      "loss": 0.19944173097610474,
      "mean_token_accuracy": 0.9169053733348846,
      "num_tokens": 32010892.0,
      "step": 2595
    },
    {
      "entropy": 1.143957257270813,
      "epoch": 1.3670352817272249,
      "grad_norm": 0.3047042787075043,
      "learning_rate": 0.00018742187240454762,
      "loss": 0.20404207706451416,
      "mean_token_accuracy": 0.9194851368665695,
      "num_tokens": 32023228.0,
      "step": 2596
    },
    {
      "entropy": 1.1393778026103973,
      "epoch": 1.3675618746708795,
      "grad_norm": 0.30072396993637085,
      "learning_rate": 0.00018741068921848626,
      "loss": 0.2037821114063263,
      "mean_token_accuracy": 0.9212421625852585,
      "num_tokens": 32035564.0,
      "step": 2597
    },
    {
      "entropy": 1.139537751674652,
      "epoch": 1.368088467614534,
      "grad_norm": 0.3391966223716736,
      "learning_rate": 0.00018739950143700348,
      "loss": 0.19885693490505219,
      "mean_token_accuracy": 0.9235148131847382,
      "num_tokens": 32047900.0,
      "step": 2598
    },
    {
      "entropy": 1.1805144548416138,
      "epoch": 1.3686150605581886,
      "grad_norm": 0.2925131022930145,
      "learning_rate": 0.00018738830906076342,
      "loss": 0.20037105679512024,
      "mean_token_accuracy": 0.9239792376756668,
      "num_tokens": 32060236.0,
      "step": 2599
    },
    {
      "entropy": 1.185164213180542,
      "epoch": 1.369141653501843,
      "grad_norm": 0.2835521996021271,
      "learning_rate": 0.00018737711209043054,
      "loss": 0.20223882794380188,
      "mean_token_accuracy": 0.917163297533989,
      "num_tokens": 32072572.0,
      "step": 2600
    },
    {
      "entropy": 1.157685101032257,
      "epoch": 1.3696682464454977,
      "grad_norm": 0.27268749475479126,
      "learning_rate": 0.00018736591052666946,
      "loss": 0.1903955042362213,
      "mean_token_accuracy": 0.9194441586732864,
      "num_tokens": 32084908.0,
      "step": 2601
    },
    {
      "entropy": 1.2693307399749756,
      "epoch": 1.370194839389152,
      "grad_norm": 0.28890320658683777,
      "learning_rate": 0.00018735470437014523,
      "loss": 0.20518316328525543,
      "mean_token_accuracy": 0.9177209436893463,
      "num_tokens": 32097244.0,
      "step": 2602
    },
    {
      "entropy": 1.2482496500015259,
      "epoch": 1.3707214323328067,
      "grad_norm": 0.30512097477912903,
      "learning_rate": 0.00018734349362152305,
      "loss": 0.21335013210773468,
      "mean_token_accuracy": 0.9146852493286133,
      "num_tokens": 32109580.0,
      "step": 2603
    },
    {
      "entropy": 1.2782647907733917,
      "epoch": 1.3712480252764614,
      "grad_norm": 0.29834169149398804,
      "learning_rate": 0.0001873322782814684,
      "loss": 0.2050481140613556,
      "mean_token_accuracy": 0.9144423753023148,
      "num_tokens": 32121916.0,
      "step": 2604
    },
    {
      "entropy": 1.2587519884109497,
      "epoch": 1.3717746182201158,
      "grad_norm": 0.28562498092651367,
      "learning_rate": 0.00018732105835064715,
      "loss": 0.19426429271697998,
      "mean_token_accuracy": 0.920745924115181,
      "num_tokens": 32134252.0,
      "step": 2605
    },
    {
      "entropy": 1.2290140092372894,
      "epoch": 1.3723012111637705,
      "grad_norm": 0.30887603759765625,
      "learning_rate": 0.00018730983382972535,
      "loss": 0.23400862514972687,
      "mean_token_accuracy": 0.9006239920854568,
      "num_tokens": 32146588.0,
      "step": 2606
    },
    {
      "entropy": 1.2600108683109283,
      "epoch": 1.3728278041074249,
      "grad_norm": 0.3018042743206024,
      "learning_rate": 0.00018729860471936927,
      "loss": 0.2075652927160263,
      "mean_token_accuracy": 0.9155648052692413,
      "num_tokens": 32158924.0,
      "step": 2607
    },
    {
      "entropy": 1.2218706607818604,
      "epoch": 1.3733543970510795,
      "grad_norm": 0.29793429374694824,
      "learning_rate": 0.00018728737102024557,
      "loss": 0.219403937458992,
      "mean_token_accuracy": 0.9094530642032623,
      "num_tokens": 32171260.0,
      "step": 2608
    },
    {
      "entropy": 1.3601875305175781,
      "epoch": 1.373880989994734,
      "grad_norm": 0.32878509163856506,
      "learning_rate": 0.00018727613273302114,
      "loss": 0.21100544929504395,
      "mean_token_accuracy": 0.9185794442892075,
      "num_tokens": 32183596.0,
      "step": 2609
    },
    {
      "entropy": 1.2661290764808655,
      "epoch": 1.3744075829383886,
      "grad_norm": 0.27767306566238403,
      "learning_rate": 0.0001872648898583631,
      "loss": 0.22158057987689972,
      "mean_token_accuracy": 0.9131675958633423,
      "num_tokens": 32195932.0,
      "step": 2610
    },
    {
      "entropy": 1.2464967370033264,
      "epoch": 1.3749341758820433,
      "grad_norm": 0.2783961892127991,
      "learning_rate": 0.00018725364239693888,
      "loss": 0.193161278963089,
      "mean_token_accuracy": 0.9226386845111847,
      "num_tokens": 32208268.0,
      "step": 2611
    },
    {
      "entropy": 1.2675247490406036,
      "epoch": 1.3754607688256977,
      "grad_norm": 0.2909803092479706,
      "learning_rate": 0.00018724239034941618,
      "loss": 0.2184634506702423,
      "mean_token_accuracy": 0.9100800156593323,
      "num_tokens": 32220604.0,
      "step": 2612
    },
    {
      "entropy": 1.223971039056778,
      "epoch": 1.3759873617693523,
      "grad_norm": 0.2707139551639557,
      "learning_rate": 0.00018723113371646295,
      "loss": 0.1935388147830963,
      "mean_token_accuracy": 0.9257960468530655,
      "num_tokens": 32232940.0,
      "step": 2613
    },
    {
      "entropy": 1.2640557885169983,
      "epoch": 1.376513954713007,
      "grad_norm": 0.257673978805542,
      "learning_rate": 0.00018721987249874746,
      "loss": 0.17373917996883392,
      "mean_token_accuracy": 0.9268936216831207,
      "num_tokens": 32245276.0,
      "step": 2614
    },
    {
      "entropy": 1.260954111814499,
      "epoch": 1.3770405476566614,
      "grad_norm": 0.2596103847026825,
      "learning_rate": 0.00018720860669693823,
      "loss": 0.19530557096004486,
      "mean_token_accuracy": 0.9233595281839371,
      "num_tokens": 32257612.0,
      "step": 2615
    },
    {
      "entropy": 1.2570489048957825,
      "epoch": 1.3775671406003158,
      "grad_norm": 0.2871362268924713,
      "learning_rate": 0.000187197336311704,
      "loss": 0.21668970584869385,
      "mean_token_accuracy": 0.9106542319059372,
      "num_tokens": 32269948.0,
      "step": 2616
    },
    {
      "entropy": 1.2458524703979492,
      "epoch": 1.3780937335439705,
      "grad_norm": 0.29486218094825745,
      "learning_rate": 0.00018718606134371385,
      "loss": 0.21607458591461182,
      "mean_token_accuracy": 0.9180518984794617,
      "num_tokens": 32282284.0,
      "step": 2617
    },
    {
      "entropy": 1.1982189416885376,
      "epoch": 1.3786203264876251,
      "grad_norm": 0.266152560710907,
      "learning_rate": 0.00018717478179363715,
      "loss": 0.1885453462600708,
      "mean_token_accuracy": 0.920305460691452,
      "num_tokens": 32294620.0,
      "step": 2618
    },
    {
      "entropy": 1.1810212433338165,
      "epoch": 1.3791469194312795,
      "grad_norm": 0.34556058049201965,
      "learning_rate": 0.00018716349766214344,
      "loss": 0.2133154273033142,
      "mean_token_accuracy": 0.9144544005393982,
      "num_tokens": 32306956.0,
      "step": 2619
    },
    {
      "entropy": 1.1619344651699066,
      "epoch": 1.3796735123749342,
      "grad_norm": 0.27238133549690247,
      "learning_rate": 0.0001871522089499026,
      "loss": 0.18257702887058258,
      "mean_token_accuracy": 0.926590159535408,
      "num_tokens": 32319292.0,
      "step": 2620
    },
    {
      "entropy": 1.1952409446239471,
      "epoch": 1.3802001053185888,
      "grad_norm": 0.29954856634140015,
      "learning_rate": 0.0001871409156575848,
      "loss": 0.19620761275291443,
      "mean_token_accuracy": 0.9240281283855438,
      "num_tokens": 32331628.0,
      "step": 2621
    },
    {
      "entropy": 1.1484823524951935,
      "epoch": 1.3807266982622433,
      "grad_norm": 0.2953477203845978,
      "learning_rate": 0.00018712961778586046,
      "loss": 0.1918509304523468,
      "mean_token_accuracy": 0.9219135642051697,
      "num_tokens": 32343964.0,
      "step": 2622
    },
    {
      "entropy": 1.1857792735099792,
      "epoch": 1.381253291205898,
      "grad_norm": 0.29551708698272705,
      "learning_rate": 0.00018711831533540024,
      "loss": 0.1936018168926239,
      "mean_token_accuracy": 0.9209873676300049,
      "num_tokens": 32356300.0,
      "step": 2623
    },
    {
      "entropy": 1.1783287823200226,
      "epoch": 1.3817798841495523,
      "grad_norm": 0.3080747127532959,
      "learning_rate": 0.00018710700830687515,
      "loss": 0.20185130834579468,
      "mean_token_accuracy": 0.9161763191223145,
      "num_tokens": 32368636.0,
      "step": 2624
    },
    {
      "entropy": 1.1695483326911926,
      "epoch": 1.382306477093207,
      "grad_norm": 0.3095219135284424,
      "learning_rate": 0.00018709569670095636,
      "loss": 0.21143102645874023,
      "mean_token_accuracy": 0.9149758964776993,
      "num_tokens": 32380972.0,
      "step": 2625
    },
    {
      "entropy": 1.1955194175243378,
      "epoch": 1.3828330700368614,
      "grad_norm": 0.2797792851924896,
      "learning_rate": 0.00018708438051831544,
      "loss": 0.19557036459445953,
      "mean_token_accuracy": 0.921556681394577,
      "num_tokens": 32393308.0,
      "step": 2626
    },
    {
      "entropy": 1.204679548740387,
      "epoch": 1.383359662980516,
      "grad_norm": 0.2982284128665924,
      "learning_rate": 0.0001870730597596241,
      "loss": 0.19409126043319702,
      "mean_token_accuracy": 0.9241613745689392,
      "num_tokens": 32405644.0,
      "step": 2627
    },
    {
      "entropy": 1.2394556403160095,
      "epoch": 1.3838862559241707,
      "grad_norm": 0.27367693185806274,
      "learning_rate": 0.00018706173442555445,
      "loss": 0.19036509096622467,
      "mean_token_accuracy": 0.9197610020637512,
      "num_tokens": 32417980.0,
      "step": 2628
    },
    {
      "entropy": 1.1927104592323303,
      "epoch": 1.3844128488678251,
      "grad_norm": 0.2765471935272217,
      "learning_rate": 0.00018705040451677874,
      "loss": 0.19922952353954315,
      "mean_token_accuracy": 0.9253708273172379,
      "num_tokens": 32430316.0,
      "step": 2629
    },
    {
      "entropy": 1.192032277584076,
      "epoch": 1.3849394418114798,
      "grad_norm": 0.2959486246109009,
      "learning_rate": 0.0001870390700339696,
      "loss": 0.21350061893463135,
      "mean_token_accuracy": 0.915240153670311,
      "num_tokens": 32442652.0,
      "step": 2630
    },
    {
      "entropy": 1.21938955783844,
      "epoch": 1.3854660347551344,
      "grad_norm": 0.2759414613246918,
      "learning_rate": 0.00018702773097779992,
      "loss": 0.20076552033424377,
      "mean_token_accuracy": 0.9171973913908005,
      "num_tokens": 32454988.0,
      "step": 2631
    },
    {
      "entropy": 1.1680685877799988,
      "epoch": 1.3859926276987888,
      "grad_norm": 0.23211334645748138,
      "learning_rate": 0.00018701638734894277,
      "loss": 0.16931575536727905,
      "mean_token_accuracy": 0.9348405301570892,
      "num_tokens": 32467324.0,
      "step": 2632
    },
    {
      "entropy": 1.2244385778903961,
      "epoch": 1.3865192206424433,
      "grad_norm": 0.3211120367050171,
      "learning_rate": 0.0001870050391480716,
      "loss": 0.22183112800121307,
      "mean_token_accuracy": 0.9124436974525452,
      "num_tokens": 32479660.0,
      "step": 2633
    },
    {
      "entropy": 1.1737190783023834,
      "epoch": 1.387045813586098,
      "grad_norm": 0.29617705941200256,
      "learning_rate": 0.00018699368637586008,
      "loss": 0.22190259397029877,
      "mean_token_accuracy": 0.9131051301956177,
      "num_tokens": 32491996.0,
      "step": 2634
    },
    {
      "entropy": 1.217042326927185,
      "epoch": 1.3875724065297526,
      "grad_norm": 0.32766997814178467,
      "learning_rate": 0.00018698232903298215,
      "loss": 0.19216807186603546,
      "mean_token_accuracy": 0.9217599779367447,
      "num_tokens": 32504332.0,
      "step": 2635
    },
    {
      "entropy": 1.2256495356559753,
      "epoch": 1.388098999473407,
      "grad_norm": 0.286161333322525,
      "learning_rate": 0.00018697096712011203,
      "loss": 0.1934540718793869,
      "mean_token_accuracy": 0.9223566651344299,
      "num_tokens": 32516668.0,
      "step": 2636
    },
    {
      "entropy": 1.1700023114681244,
      "epoch": 1.3886255924170616,
      "grad_norm": 0.2702758014202118,
      "learning_rate": 0.00018695960063792421,
      "loss": 0.2006986141204834,
      "mean_token_accuracy": 0.9158665686845779,
      "num_tokens": 32529004.0,
      "step": 2637
    },
    {
      "entropy": 1.1851938664913177,
      "epoch": 1.3891521853607163,
      "grad_norm": 0.2678104639053345,
      "learning_rate": 0.00018694822958709346,
      "loss": 0.1974492073059082,
      "mean_token_accuracy": 0.9265509694814682,
      "num_tokens": 32541340.0,
      "step": 2638
    },
    {
      "entropy": 1.1814790070056915,
      "epoch": 1.3896787783043707,
      "grad_norm": 0.2913559079170227,
      "learning_rate": 0.00018693685396829483,
      "loss": 0.20490042865276337,
      "mean_token_accuracy": 0.911860391497612,
      "num_tokens": 32553676.0,
      "step": 2639
    },
    {
      "entropy": 1.1520279347896576,
      "epoch": 1.3902053712480253,
      "grad_norm": 0.3147750496864319,
      "learning_rate": 0.00018692547378220358,
      "loss": 0.20494982600212097,
      "mean_token_accuracy": 0.9163292199373245,
      "num_tokens": 32566012.0,
      "step": 2640
    },
    {
      "entropy": 1.1539938151836395,
      "epoch": 1.3907319641916798,
      "grad_norm": 0.27995991706848145,
      "learning_rate": 0.0001869140890294953,
      "loss": 0.19520774483680725,
      "mean_token_accuracy": 0.9176739156246185,
      "num_tokens": 32578348.0,
      "step": 2641
    },
    {
      "entropy": 1.1977797150611877,
      "epoch": 1.3912585571353344,
      "grad_norm": 0.2937062978744507,
      "learning_rate": 0.00018690269971084587,
      "loss": 0.20538851618766785,
      "mean_token_accuracy": 0.9172974228858948,
      "num_tokens": 32590684.0,
      "step": 2642
    },
    {
      "entropy": 1.187753289937973,
      "epoch": 1.3917851500789888,
      "grad_norm": 0.2842217683792114,
      "learning_rate": 0.00018689130582693136,
      "loss": 0.2011350840330124,
      "mean_token_accuracy": 0.9161236584186554,
      "num_tokens": 32603020.0,
      "step": 2643
    },
    {
      "entropy": 1.1970960795879364,
      "epoch": 1.3923117430226435,
      "grad_norm": 0.2834356129169464,
      "learning_rate": 0.00018687990737842818,
      "loss": 0.20202001929283142,
      "mean_token_accuracy": 0.920697346329689,
      "num_tokens": 32615356.0,
      "step": 2644
    },
    {
      "entropy": 1.1325391829013824,
      "epoch": 1.3928383359662981,
      "grad_norm": 0.2595110833644867,
      "learning_rate": 0.00018686850436601298,
      "loss": 0.17119885981082916,
      "mean_token_accuracy": 0.9274217188358307,
      "num_tokens": 32627692.0,
      "step": 2645
    },
    {
      "entropy": 1.203096479177475,
      "epoch": 1.3933649289099526,
      "grad_norm": 0.28552207350730896,
      "learning_rate": 0.00018685709679036272,
      "loss": 0.1984875351190567,
      "mean_token_accuracy": 0.9184497147798538,
      "num_tokens": 32640028.0,
      "step": 2646
    },
    {
      "entropy": 1.2174711227416992,
      "epoch": 1.3938915218536072,
      "grad_norm": 0.3431668281555176,
      "learning_rate": 0.00018684568465215457,
      "loss": 0.23073092103004456,
      "mean_token_accuracy": 0.9064096510410309,
      "num_tokens": 32652364.0,
      "step": 2647
    },
    {
      "entropy": 1.1742651760578156,
      "epoch": 1.3944181147972619,
      "grad_norm": 0.2867361605167389,
      "learning_rate": 0.00018683426795206603,
      "loss": 0.20909307897090912,
      "mean_token_accuracy": 0.9160559475421906,
      "num_tokens": 32664700.0,
      "step": 2648
    },
    {
      "entropy": 1.1931123435497284,
      "epoch": 1.3949447077409163,
      "grad_norm": 0.27413511276245117,
      "learning_rate": 0.0001868228466907748,
      "loss": 0.1944372057914734,
      "mean_token_accuracy": 0.9192412346601486,
      "num_tokens": 32677036.0,
      "step": 2649
    },
    {
      "entropy": 1.200911283493042,
      "epoch": 1.3954713006845707,
      "grad_norm": 0.32713064551353455,
      "learning_rate": 0.0001868114208689589,
      "loss": 0.22633163630962372,
      "mean_token_accuracy": 0.9066626876592636,
      "num_tokens": 32689372.0,
      "step": 2650
    },
    {
      "entropy": 1.173728585243225,
      "epoch": 1.3959978936282254,
      "grad_norm": 0.2703946828842163,
      "learning_rate": 0.00018679999048729667,
      "loss": 0.2105778455734253,
      "mean_token_accuracy": 0.9153214693069458,
      "num_tokens": 32701708.0,
      "step": 2651
    },
    {
      "entropy": 1.2183512151241302,
      "epoch": 1.39652448657188,
      "grad_norm": 0.2865394651889801,
      "learning_rate": 0.00018678855554646665,
      "loss": 0.19708535075187683,
      "mean_token_accuracy": 0.9225267171859741,
      "num_tokens": 32714044.0,
      "step": 2652
    },
    {
      "entropy": 1.2178658246994019,
      "epoch": 1.3970510795155344,
      "grad_norm": 0.2824452519416809,
      "learning_rate": 0.00018677711604714762,
      "loss": 0.22327733039855957,
      "mean_token_accuracy": 0.9088855236768723,
      "num_tokens": 32726380.0,
      "step": 2653
    },
    {
      "entropy": 1.2230781316757202,
      "epoch": 1.397577672459189,
      "grad_norm": 0.2827853560447693,
      "learning_rate": 0.00018676567199001868,
      "loss": 0.19934280216693878,
      "mean_token_accuracy": 0.9177891463041306,
      "num_tokens": 32738716.0,
      "step": 2654
    },
    {
      "entropy": 1.1976493299007416,
      "epoch": 1.3981042654028437,
      "grad_norm": 0.26556023955345154,
      "learning_rate": 0.00018675422337575926,
      "loss": 0.2057933360338211,
      "mean_token_accuracy": 0.918548509478569,
      "num_tokens": 32751052.0,
      "step": 2655
    },
    {
      "entropy": 1.2001135349273682,
      "epoch": 1.3986308583464981,
      "grad_norm": 0.2884269058704376,
      "learning_rate": 0.0001867427702050489,
      "loss": 0.20356959104537964,
      "mean_token_accuracy": 0.9222962260246277,
      "num_tokens": 32763388.0,
      "step": 2656
    },
    {
      "entropy": 1.206467181444168,
      "epoch": 1.3991574512901528,
      "grad_norm": 0.2789088189601898,
      "learning_rate": 0.0001867313124785676,
      "loss": 0.200584277510643,
      "mean_token_accuracy": 0.9238597601652145,
      "num_tokens": 32775724.0,
      "step": 2657
    },
    {
      "entropy": 1.1685158610343933,
      "epoch": 1.3996840442338072,
      "grad_norm": 0.262813001871109,
      "learning_rate": 0.00018671985019699553,
      "loss": 0.19860176742076874,
      "mean_token_accuracy": 0.9217494130134583,
      "num_tokens": 32788060.0,
      "step": 2658
    },
    {
      "entropy": 1.1696327328681946,
      "epoch": 1.4002106371774619,
      "grad_norm": 0.2782774865627289,
      "learning_rate": 0.00018670838336101308,
      "loss": 0.20912325382232666,
      "mean_token_accuracy": 0.9155998378992081,
      "num_tokens": 32800396.0,
      "step": 2659
    },
    {
      "entropy": 1.1536991596221924,
      "epoch": 1.4007372301211163,
      "grad_norm": 0.2786557972431183,
      "learning_rate": 0.00018669691197130097,
      "loss": 0.1923229843378067,
      "mean_token_accuracy": 0.9198986142873764,
      "num_tokens": 32812732.0,
      "step": 2660
    },
    {
      "entropy": 1.1637346744537354,
      "epoch": 1.401263823064771,
      "grad_norm": 0.2913641035556793,
      "learning_rate": 0.00018668543602854028,
      "loss": 0.19935357570648193,
      "mean_token_accuracy": 0.91521255671978,
      "num_tokens": 32825068.0,
      "step": 2661
    },
    {
      "entropy": 1.1952362954616547,
      "epoch": 1.4017904160084256,
      "grad_norm": 0.27064937353134155,
      "learning_rate": 0.00018667395553341213,
      "loss": 0.19324931502342224,
      "mean_token_accuracy": 0.9230464100837708,
      "num_tokens": 32837404.0,
      "step": 2662
    },
    {
      "entropy": 1.172296166419983,
      "epoch": 1.40231700895208,
      "grad_norm": 0.2844633162021637,
      "learning_rate": 0.0001866624704865982,
      "loss": 0.21573956310749054,
      "mean_token_accuracy": 0.9178134202957153,
      "num_tokens": 32849740.0,
      "step": 2663
    },
    {
      "entropy": 1.1426364183425903,
      "epoch": 1.4028436018957346,
      "grad_norm": 0.2857990860939026,
      "learning_rate": 0.0001866509808887802,
      "loss": 0.19915884733200073,
      "mean_token_accuracy": 0.919086143374443,
      "num_tokens": 32862076.0,
      "step": 2664
    },
    {
      "entropy": 1.1875398755073547,
      "epoch": 1.403370194839389,
      "grad_norm": 0.29901546239852905,
      "learning_rate": 0.00018663948674064025,
      "loss": 0.2148965746164322,
      "mean_token_accuracy": 0.9187829047441483,
      "num_tokens": 32874412.0,
      "step": 2665
    },
    {
      "entropy": 1.1802086234092712,
      "epoch": 1.4038967877830437,
      "grad_norm": 0.30026695132255554,
      "learning_rate": 0.0001866279880428606,
      "loss": 0.21468932926654816,
      "mean_token_accuracy": 0.9141033738851547,
      "num_tokens": 32886748.0,
      "step": 2666
    },
    {
      "entropy": 1.231868177652359,
      "epoch": 1.4044233807266981,
      "grad_norm": 0.3457041382789612,
      "learning_rate": 0.00018661648479612397,
      "loss": 0.21766367554664612,
      "mean_token_accuracy": 0.9147387742996216,
      "num_tokens": 32899084.0,
      "step": 2667
    },
    {
      "entropy": 1.1672147512435913,
      "epoch": 1.4049499736703528,
      "grad_norm": 0.29802289605140686,
      "learning_rate": 0.00018660497700111317,
      "loss": 0.1930084526538849,
      "mean_token_accuracy": 0.9234433770179749,
      "num_tokens": 32911420.0,
      "step": 2668
    },
    {
      "entropy": 1.2288798391819,
      "epoch": 1.4054765666140074,
      "grad_norm": 0.2817768454551697,
      "learning_rate": 0.00018659346465851138,
      "loss": 0.2119956761598587,
      "mean_token_accuracy": 0.9163240343332291,
      "num_tokens": 32923756.0,
      "step": 2669
    },
    {
      "entropy": 1.2406794726848602,
      "epoch": 1.4060031595576619,
      "grad_norm": 0.3248741030693054,
      "learning_rate": 0.000186581947769002,
      "loss": 0.21822670102119446,
      "mean_token_accuracy": 0.9133480787277222,
      "num_tokens": 32936092.0,
      "step": 2670
    },
    {
      "entropy": 1.1928761005401611,
      "epoch": 1.4065297525013165,
      "grad_norm": 0.29017162322998047,
      "learning_rate": 0.00018657042633326876,
      "loss": 0.19945287704467773,
      "mean_token_accuracy": 0.921704575419426,
      "num_tokens": 32948428.0,
      "step": 2671
    },
    {
      "entropy": 1.2523100972175598,
      "epoch": 1.4070563454449712,
      "grad_norm": 0.2896433174610138,
      "learning_rate": 0.00018655890035199557,
      "loss": 0.20614881813526154,
      "mean_token_accuracy": 0.9159245640039444,
      "num_tokens": 32960764.0,
      "step": 2672
    },
    {
      "entropy": 1.2614913880825043,
      "epoch": 1.4075829383886256,
      "grad_norm": 0.2920669615268707,
      "learning_rate": 0.0001865473698258667,
      "loss": 0.2224857062101364,
      "mean_token_accuracy": 0.9055542200803757,
      "num_tokens": 32973100.0,
      "step": 2673
    },
    {
      "entropy": 1.2949616014957428,
      "epoch": 1.4081095313322802,
      "grad_norm": 0.3054332435131073,
      "learning_rate": 0.00018653583475556663,
      "loss": 0.20840829610824585,
      "mean_token_accuracy": 0.9158879816532135,
      "num_tokens": 32985436.0,
      "step": 2674
    },
    {
      "entropy": 1.3014479279518127,
      "epoch": 1.4086361242759347,
      "grad_norm": 0.293302059173584,
      "learning_rate": 0.00018652429514178016,
      "loss": 0.2146826684474945,
      "mean_token_accuracy": 0.9126481860876083,
      "num_tokens": 32997772.0,
      "step": 2675
    },
    {
      "entropy": 1.3077419102191925,
      "epoch": 1.4091627172195893,
      "grad_norm": 0.2812763750553131,
      "learning_rate": 0.00018651275098519228,
      "loss": 0.2168470174074173,
      "mean_token_accuracy": 0.9146820157766342,
      "num_tokens": 33010108.0,
      "step": 2676
    },
    {
      "entropy": 1.3591797947883606,
      "epoch": 1.4096893101632437,
      "grad_norm": 0.27339309453964233,
      "learning_rate": 0.0001865012022864883,
      "loss": 0.2067457139492035,
      "mean_token_accuracy": 0.910685271024704,
      "num_tokens": 33022444.0,
      "step": 2677
    },
    {
      "entropy": 1.3165797293186188,
      "epoch": 1.4102159031068984,
      "grad_norm": 0.26994794607162476,
      "learning_rate": 0.00018648964904635388,
      "loss": 0.20063388347625732,
      "mean_token_accuracy": 0.915714293718338,
      "num_tokens": 33034780.0,
      "step": 2678
    },
    {
      "entropy": 1.350212275981903,
      "epoch": 1.410742496050553,
      "grad_norm": 0.26788488030433655,
      "learning_rate": 0.00018647809126547476,
      "loss": 0.2082659900188446,
      "mean_token_accuracy": 0.9163226187229156,
      "num_tokens": 33047116.0,
      "step": 2679
    },
    {
      "entropy": 1.3494655191898346,
      "epoch": 1.4112690889942074,
      "grad_norm": 0.3014771342277527,
      "learning_rate": 0.00018646652894453714,
      "loss": 0.2213258445262909,
      "mean_token_accuracy": 0.9125747829675674,
      "num_tokens": 33059452.0,
      "step": 2680
    },
    {
      "entropy": 1.3585596978664398,
      "epoch": 1.411795681937862,
      "grad_norm": 0.2993260324001312,
      "learning_rate": 0.00018645496208422738,
      "loss": 0.23662728071212769,
      "mean_token_accuracy": 0.9095396250486374,
      "num_tokens": 33071788.0,
      "step": 2681
    },
    {
      "entropy": 1.2951436042785645,
      "epoch": 1.4123222748815165,
      "grad_norm": 0.28097495436668396,
      "learning_rate": 0.00018644339068523208,
      "loss": 0.19883355498313904,
      "mean_token_accuracy": 0.9165204763412476,
      "num_tokens": 33084124.0,
      "step": 2682
    },
    {
      "entropy": 1.2786961197853088,
      "epoch": 1.4128488678251712,
      "grad_norm": 0.31551140546798706,
      "learning_rate": 0.00018643181474823828,
      "loss": 0.22398042678833008,
      "mean_token_accuracy": 0.9099626541137695,
      "num_tokens": 33096460.0,
      "step": 2683
    },
    {
      "entropy": 1.353494018316269,
      "epoch": 1.4133754607688256,
      "grad_norm": 0.27544382214546204,
      "learning_rate": 0.00018642023427393312,
      "loss": 0.19528742134571075,
      "mean_token_accuracy": 0.9221656024456024,
      "num_tokens": 33108796.0,
      "step": 2684
    },
    {
      "entropy": 1.3157804608345032,
      "epoch": 1.4139020537124802,
      "grad_norm": 0.26549386978149414,
      "learning_rate": 0.00018640864926300403,
      "loss": 0.1861998438835144,
      "mean_token_accuracy": 0.924147367477417,
      "num_tokens": 33121132.0,
      "step": 2685
    },
    {
      "entropy": 1.2875000834465027,
      "epoch": 1.4144286466561349,
      "grad_norm": 0.3759959042072296,
      "learning_rate": 0.00018639705971613878,
      "loss": 0.1876804083585739,
      "mean_token_accuracy": 0.9247392117977142,
      "num_tokens": 33133468.0,
      "step": 2686
    },
    {
      "entropy": 1.278031587600708,
      "epoch": 1.4149552395997893,
      "grad_norm": 0.2772931456565857,
      "learning_rate": 0.00018638546563402538,
      "loss": 0.1913738250732422,
      "mean_token_accuracy": 0.9191721677780151,
      "num_tokens": 33145804.0,
      "step": 2687
    },
    {
      "entropy": 1.3772215843200684,
      "epoch": 1.415481832543444,
      "grad_norm": 0.32579630613327026,
      "learning_rate": 0.00018637386701735208,
      "loss": 0.20003564655780792,
      "mean_token_accuracy": 0.9227780699729919,
      "num_tokens": 33158140.0,
      "step": 2688
    },
    {
      "entropy": 1.330332100391388,
      "epoch": 1.4160084254870986,
      "grad_norm": 0.3118603229522705,
      "learning_rate": 0.00018636226386680748,
      "loss": 0.1903379112482071,
      "mean_token_accuracy": 0.9189924746751785,
      "num_tokens": 33170476.0,
      "step": 2689
    },
    {
      "entropy": 1.3054493367671967,
      "epoch": 1.416535018430753,
      "grad_norm": 0.4101371467113495,
      "learning_rate": 0.0001863506561830803,
      "loss": 0.2160140872001648,
      "mean_token_accuracy": 0.9076957255601883,
      "num_tokens": 33182812.0,
      "step": 2690
    },
    {
      "entropy": 1.3156910836696625,
      "epoch": 1.4170616113744074,
      "grad_norm": 0.2924982011318207,
      "learning_rate": 0.0001863390439668597,
      "loss": 0.19913962483406067,
      "mean_token_accuracy": 0.9208202958106995,
      "num_tokens": 33195148.0,
      "step": 2691
    },
    {
      "entropy": 1.3482598662376404,
      "epoch": 1.417588204318062,
      "grad_norm": 0.27372100949287415,
      "learning_rate": 0.00018632742721883495,
      "loss": 0.19036367535591125,
      "mean_token_accuracy": 0.9198184758424759,
      "num_tokens": 33207484.0,
      "step": 2692
    },
    {
      "entropy": 1.3402643203735352,
      "epoch": 1.4181147972617167,
      "grad_norm": 0.2865314781665802,
      "learning_rate": 0.00018631580593969578,
      "loss": 0.20403306186199188,
      "mean_token_accuracy": 0.9174676239490509,
      "num_tokens": 33219820.0,
      "step": 2693
    },
    {
      "entropy": 1.3814849257469177,
      "epoch": 1.4186413902053712,
      "grad_norm": 0.298798531293869,
      "learning_rate": 0.00018630418013013199,
      "loss": 0.19326891005039215,
      "mean_token_accuracy": 0.9204562902450562,
      "num_tokens": 33232156.0,
      "step": 2694
    },
    {
      "entropy": 1.3497715592384338,
      "epoch": 1.4191679831490258,
      "grad_norm": 0.2983977794647217,
      "learning_rate": 0.00018629254979083372,
      "loss": 0.2127034068107605,
      "mean_token_accuracy": 0.9153210371732712,
      "num_tokens": 33244492.0,
      "step": 2695
    },
    {
      "entropy": 1.3387363851070404,
      "epoch": 1.4196945760926805,
      "grad_norm": 0.2673841714859009,
      "learning_rate": 0.00018628091492249149,
      "loss": 0.1976945847272873,
      "mean_token_accuracy": 0.9180579632520676,
      "num_tokens": 33256828.0,
      "step": 2696
    },
    {
      "entropy": 1.2811682224273682,
      "epoch": 1.4202211690363349,
      "grad_norm": 0.2700756788253784,
      "learning_rate": 0.00018626927552579592,
      "loss": 0.19237622618675232,
      "mean_token_accuracy": 0.9196698665618896,
      "num_tokens": 33269164.0,
      "step": 2697
    },
    {
      "entropy": 1.2764002978801727,
      "epoch": 1.4207477619799895,
      "grad_norm": 0.2527097463607788,
      "learning_rate": 0.00018625763160143796,
      "loss": 0.1781824678182602,
      "mean_token_accuracy": 0.9250129163265228,
      "num_tokens": 33281500.0,
      "step": 2698
    },
    {
      "entropy": 1.3061909675598145,
      "epoch": 1.421274354923644,
      "grad_norm": 0.27409103512763977,
      "learning_rate": 0.0001862459831501089,
      "loss": 0.1834535151720047,
      "mean_token_accuracy": 0.9291260540485382,
      "num_tokens": 33293836.0,
      "step": 2699
    },
    {
      "entropy": 1.4255090951919556,
      "epoch": 1.4218009478672986,
      "grad_norm": 0.35492557287216187,
      "learning_rate": 0.00018623433017250025,
      "loss": 0.2331741452217102,
      "mean_token_accuracy": 0.9052500277757645,
      "num_tokens": 33306172.0,
      "step": 2700
    },
    {
      "entropy": 1.3506521880626678,
      "epoch": 1.422327540810953,
      "grad_norm": 0.26301318407058716,
      "learning_rate": 0.0001862226726693037,
      "loss": 0.19686710834503174,
      "mean_token_accuracy": 0.9148916304111481,
      "num_tokens": 33318508.0,
      "step": 2701
    },
    {
      "entropy": 1.3578969836235046,
      "epoch": 1.4228541337546077,
      "grad_norm": 0.30393311381340027,
      "learning_rate": 0.00018621101064121136,
      "loss": 0.21072648465633392,
      "mean_token_accuracy": 0.9144587218761444,
      "num_tokens": 33330844.0,
      "step": 2702
    },
    {
      "entropy": 1.360473245382309,
      "epoch": 1.4233807266982623,
      "grad_norm": 0.27426621317863464,
      "learning_rate": 0.0001861993440889155,
      "loss": 0.1906125247478485,
      "mean_token_accuracy": 0.919151559472084,
      "num_tokens": 33343180.0,
      "step": 2703
    },
    {
      "entropy": 1.3859812319278717,
      "epoch": 1.4239073196419167,
      "grad_norm": 0.3062906861305237,
      "learning_rate": 0.0001861876730131087,
      "loss": 0.2006826102733612,
      "mean_token_accuracy": 0.915339469909668,
      "num_tokens": 33355516.0,
      "step": 2704
    },
    {
      "entropy": 1.3931569159030914,
      "epoch": 1.4244339125855714,
      "grad_norm": 0.27953484654426575,
      "learning_rate": 0.0001861759974144838,
      "loss": 0.20995528995990753,
      "mean_token_accuracy": 0.9106060713529587,
      "num_tokens": 33367852.0,
      "step": 2705
    },
    {
      "entropy": 1.4093142449855804,
      "epoch": 1.424960505529226,
      "grad_norm": 0.28498291969299316,
      "learning_rate": 0.00018616431729373393,
      "loss": 0.1967875361442566,
      "mean_token_accuracy": 0.9221569150686264,
      "num_tokens": 33380188.0,
      "step": 2706
    },
    {
      "entropy": 1.3485685586929321,
      "epoch": 1.4254870984728805,
      "grad_norm": 0.299116849899292,
      "learning_rate": 0.00018615263265155246,
      "loss": 0.20301219820976257,
      "mean_token_accuracy": 0.9173270463943481,
      "num_tokens": 33392524.0,
      "step": 2707
    },
    {
      "entropy": 1.3589564263820648,
      "epoch": 1.426013691416535,
      "grad_norm": 0.2964306175708771,
      "learning_rate": 0.00018614094348863306,
      "loss": 0.19253556430339813,
      "mean_token_accuracy": 0.9231040179729462,
      "num_tokens": 33404860.0,
      "step": 2708
    },
    {
      "entropy": 1.3823421597480774,
      "epoch": 1.4265402843601895,
      "grad_norm": 0.34210389852523804,
      "learning_rate": 0.0001861292498056696,
      "loss": 0.2110666185617447,
      "mean_token_accuracy": 0.9131441861391068,
      "num_tokens": 33417196.0,
      "step": 2709
    },
    {
      "entropy": 1.3455529510974884,
      "epoch": 1.4270668773038442,
      "grad_norm": 0.2749879062175751,
      "learning_rate": 0.00018611755160335633,
      "loss": 0.17992670834064484,
      "mean_token_accuracy": 0.9268961399793625,
      "num_tokens": 33429532.0,
      "step": 2710
    },
    {
      "entropy": 1.3250888884067535,
      "epoch": 1.4275934702474986,
      "grad_norm": 0.30342763662338257,
      "learning_rate": 0.00018610584888238764,
      "loss": 0.21458712220191956,
      "mean_token_accuracy": 0.9091243892908096,
      "num_tokens": 33441868.0,
      "step": 2711
    },
    {
      "entropy": 1.2988539934158325,
      "epoch": 1.4281200631911533,
      "grad_norm": 0.279607355594635,
      "learning_rate": 0.00018609414164345829,
      "loss": 0.1852928102016449,
      "mean_token_accuracy": 0.9220867305994034,
      "num_tokens": 33454204.0,
      "step": 2712
    },
    {
      "entropy": 1.314719408750534,
      "epoch": 1.428646656134808,
      "grad_norm": 0.28591296076774597,
      "learning_rate": 0.00018608242988726325,
      "loss": 0.17573139071464539,
      "mean_token_accuracy": 0.9259963184595108,
      "num_tokens": 33466540.0,
      "step": 2713
    },
    {
      "entropy": 1.2993620932102203,
      "epoch": 1.4291732490784623,
      "grad_norm": 0.2799248695373535,
      "learning_rate": 0.00018607071361449783,
      "loss": 0.21279355883598328,
      "mean_token_accuracy": 0.9184603095054626,
      "num_tokens": 33478876.0,
      "step": 2714
    },
    {
      "entropy": 1.2804417610168457,
      "epoch": 1.429699842022117,
      "grad_norm": 0.27112871408462524,
      "learning_rate": 0.00018605899282585743,
      "loss": 0.2066880762577057,
      "mean_token_accuracy": 0.9143773913383484,
      "num_tokens": 33491212.0,
      "step": 2715
    },
    {
      "entropy": 1.3299663960933685,
      "epoch": 1.4302264349657714,
      "grad_norm": 0.2755028009414673,
      "learning_rate": 0.000186047267522038,
      "loss": 0.20685116946697235,
      "mean_token_accuracy": 0.9201144576072693,
      "num_tokens": 33503548.0,
      "step": 2716
    },
    {
      "entropy": 1.2998407185077667,
      "epoch": 1.430753027909426,
      "grad_norm": 0.2707025408744812,
      "learning_rate": 0.00018603553770373552,
      "loss": 0.19559168815612793,
      "mean_token_accuracy": 0.9165090620517731,
      "num_tokens": 33515884.0,
      "step": 2717
    },
    {
      "entropy": 1.3012645542621613,
      "epoch": 1.4312796208530805,
      "grad_norm": 0.2650182247161865,
      "learning_rate": 0.00018602380337164634,
      "loss": 0.17805761098861694,
      "mean_token_accuracy": 0.9261130839586258,
      "num_tokens": 33528220.0,
      "step": 2718
    },
    {
      "entropy": 1.2708977460861206,
      "epoch": 1.4318062137967351,
      "grad_norm": 0.27006566524505615,
      "learning_rate": 0.00018601206452646706,
      "loss": 0.18752358853816986,
      "mean_token_accuracy": 0.9268027991056442,
      "num_tokens": 33540556.0,
      "step": 2719
    },
    {
      "entropy": 1.2761503458023071,
      "epoch": 1.4323328067403898,
      "grad_norm": 0.2888233959674835,
      "learning_rate": 0.00018600032116889453,
      "loss": 0.2155730128288269,
      "mean_token_accuracy": 0.9141710996627808,
      "num_tokens": 33552892.0,
      "step": 2720
    },
    {
      "entropy": 1.2359469830989838,
      "epoch": 1.4328593996840442,
      "grad_norm": 0.29737192392349243,
      "learning_rate": 0.00018598857329962587,
      "loss": 0.21056918799877167,
      "mean_token_accuracy": 0.9146129190921783,
      "num_tokens": 33565228.0,
      "step": 2721
    },
    {
      "entropy": 1.188820868730545,
      "epoch": 1.4333859926276988,
      "grad_norm": 0.303146630525589,
      "learning_rate": 0.00018597682091935856,
      "loss": 0.202779620885849,
      "mean_token_accuracy": 0.9155432283878326,
      "num_tokens": 33577564.0,
      "step": 2722
    },
    {
      "entropy": 1.2396386861801147,
      "epoch": 1.4339125855713535,
      "grad_norm": 0.2838284969329834,
      "learning_rate": 0.00018596506402879018,
      "loss": 0.2013065367937088,
      "mean_token_accuracy": 0.9150380194187164,
      "num_tokens": 33589900.0,
      "step": 2723
    },
    {
      "entropy": 1.2167082726955414,
      "epoch": 1.434439178515008,
      "grad_norm": 0.28906625509262085,
      "learning_rate": 0.0001859533026286187,
      "loss": 0.21567818522453308,
      "mean_token_accuracy": 0.9125472903251648,
      "num_tokens": 33602236.0,
      "step": 2724
    },
    {
      "entropy": 1.2206608951091766,
      "epoch": 1.4349657714586623,
      "grad_norm": 0.27452901005744934,
      "learning_rate": 0.00018594153671954236,
      "loss": 0.22003887593746185,
      "mean_token_accuracy": 0.9081568866968155,
      "num_tokens": 33614572.0,
      "step": 2725
    },
    {
      "entropy": 1.2153219878673553,
      "epoch": 1.435492364402317,
      "grad_norm": 0.28818845748901367,
      "learning_rate": 0.0001859297663022596,
      "loss": 0.21699745953083038,
      "mean_token_accuracy": 0.9116252958774567,
      "num_tokens": 33626908.0,
      "step": 2726
    },
    {
      "entropy": 1.2302789092063904,
      "epoch": 1.4360189573459716,
      "grad_norm": 0.2688583433628082,
      "learning_rate": 0.00018591799137746915,
      "loss": 0.19513952732086182,
      "mean_token_accuracy": 0.922428548336029,
      "num_tokens": 33639244.0,
      "step": 2727
    },
    {
      "entropy": 1.2083250284194946,
      "epoch": 1.436545550289626,
      "grad_norm": 0.2625478506088257,
      "learning_rate": 0.00018590621194587007,
      "loss": 0.20137974619865417,
      "mean_token_accuracy": 0.9156837463378906,
      "num_tokens": 33651580.0,
      "step": 2728
    },
    {
      "entropy": 1.251519650220871,
      "epoch": 1.4370721432332807,
      "grad_norm": 0.29624244570732117,
      "learning_rate": 0.00018589442800816158,
      "loss": 0.2012304663658142,
      "mean_token_accuracy": 0.9153359830379486,
      "num_tokens": 33663916.0,
      "step": 2729
    },
    {
      "entropy": 1.2322251498699188,
      "epoch": 1.4375987361769353,
      "grad_norm": 0.2729564309120178,
      "learning_rate": 0.00018588263956504327,
      "loss": 0.1936715543270111,
      "mean_token_accuracy": 0.9187065362930298,
      "num_tokens": 33676252.0,
      "step": 2730
    },
    {
      "entropy": 1.263836681842804,
      "epoch": 1.4381253291205898,
      "grad_norm": 0.2771749198436737,
      "learning_rate": 0.00018587084661721487,
      "loss": 0.19794949889183044,
      "mean_token_accuracy": 0.9143384546041489,
      "num_tokens": 33688588.0,
      "step": 2731
    },
    {
      "entropy": 1.2738985121250153,
      "epoch": 1.4386519220642444,
      "grad_norm": 0.30315807461738586,
      "learning_rate": 0.00018585904916537655,
      "loss": 0.20564697682857513,
      "mean_token_accuracy": 0.9139324128627777,
      "num_tokens": 33700924.0,
      "step": 2732
    },
    {
      "entropy": 1.2370990812778473,
      "epoch": 1.4391785150078988,
      "grad_norm": 0.3091336488723755,
      "learning_rate": 0.0001858472472102286,
      "loss": 0.2253596931695938,
      "mean_token_accuracy": 0.9058148860931396,
      "num_tokens": 33713260.0,
      "step": 2733
    },
    {
      "entropy": 1.2049230933189392,
      "epoch": 1.4397051079515535,
      "grad_norm": 0.2731562554836273,
      "learning_rate": 0.0001858354407524717,
      "loss": 0.18417584896087646,
      "mean_token_accuracy": 0.9258820116519928,
      "num_tokens": 33725596.0,
      "step": 2734
    },
    {
      "entropy": 1.187551736831665,
      "epoch": 1.440231700895208,
      "grad_norm": 0.27385076880455017,
      "learning_rate": 0.00018582362979280666,
      "loss": 0.1865832656621933,
      "mean_token_accuracy": 0.9226599186658859,
      "num_tokens": 33737932.0,
      "step": 2735
    },
    {
      "entropy": 1.286464512348175,
      "epoch": 1.4407582938388626,
      "grad_norm": 0.31394457817077637,
      "learning_rate": 0.00018581181433193465,
      "loss": 0.2028738260269165,
      "mean_token_accuracy": 0.923179104924202,
      "num_tokens": 33750268.0,
      "step": 2736
    },
    {
      "entropy": 1.2562143802642822,
      "epoch": 1.4412848867825172,
      "grad_norm": 0.2848510146141052,
      "learning_rate": 0.0001857999943705571,
      "loss": 0.19614958763122559,
      "mean_token_accuracy": 0.9193147718906403,
      "num_tokens": 33762604.0,
      "step": 2737
    },
    {
      "entropy": 1.1915212869644165,
      "epoch": 1.4418114797261716,
      "grad_norm": 0.28245916962623596,
      "learning_rate": 0.00018578816990937568,
      "loss": 0.2020082175731659,
      "mean_token_accuracy": 0.9246124476194382,
      "num_tokens": 33774940.0,
      "step": 2738
    },
    {
      "entropy": 1.1338683068752289,
      "epoch": 1.4423380726698263,
      "grad_norm": 0.2531015872955322,
      "learning_rate": 0.0001857763409490923,
      "loss": 0.19030161201953888,
      "mean_token_accuracy": 0.9210085272789001,
      "num_tokens": 33787276.0,
      "step": 2739
    },
    {
      "entropy": 1.1661230623722076,
      "epoch": 1.4428646656134807,
      "grad_norm": 0.2816557288169861,
      "learning_rate": 0.00018576450749040925,
      "loss": 0.19968748092651367,
      "mean_token_accuracy": 0.9191047251224518,
      "num_tokens": 33799612.0,
      "step": 2740
    },
    {
      "entropy": 1.226789653301239,
      "epoch": 1.4433912585571353,
      "grad_norm": 0.2743048369884491,
      "learning_rate": 0.00018575266953402896,
      "loss": 0.2070472538471222,
      "mean_token_accuracy": 0.9168623089790344,
      "num_tokens": 33811948.0,
      "step": 2741
    },
    {
      "entropy": 1.226866990327835,
      "epoch": 1.4439178515007898,
      "grad_norm": 0.27064448595046997,
      "learning_rate": 0.00018574082708065423,
      "loss": 0.19173914194107056,
      "mean_token_accuracy": 0.9202168434858322,
      "num_tokens": 33824284.0,
      "step": 2742
    },
    {
      "entropy": 1.1833116710186005,
      "epoch": 1.4444444444444444,
      "grad_norm": 0.2748906910419464,
      "learning_rate": 0.00018572898013098803,
      "loss": 0.20692995190620422,
      "mean_token_accuracy": 0.9154257923364639,
      "num_tokens": 33836620.0,
      "step": 2743
    },
    {
      "entropy": 1.2158693671226501,
      "epoch": 1.444971037388099,
      "grad_norm": 0.28100040555000305,
      "learning_rate": 0.00018571712868573368,
      "loss": 0.19965311884880066,
      "mean_token_accuracy": 0.9194239974021912,
      "num_tokens": 33848956.0,
      "step": 2744
    },
    {
      "entropy": 1.1900866031646729,
      "epoch": 1.4454976303317535,
      "grad_norm": 0.24971705675125122,
      "learning_rate": 0.00018570527274559468,
      "loss": 0.19126318395137787,
      "mean_token_accuracy": 0.9166027456521988,
      "num_tokens": 33861292.0,
      "step": 2745
    },
    {
      "entropy": 1.2365811467170715,
      "epoch": 1.4460242232754081,
      "grad_norm": 0.2837916612625122,
      "learning_rate": 0.0001856934123112749,
      "loss": 0.21584585309028625,
      "mean_token_accuracy": 0.9098552912473679,
      "num_tokens": 33873628.0,
      "step": 2746
    },
    {
      "entropy": 1.177088737487793,
      "epoch": 1.4465508162190628,
      "grad_norm": 0.25515440106391907,
      "learning_rate": 0.00018568154738347837,
      "loss": 0.18173626065254211,
      "mean_token_accuracy": 0.9271458983421326,
      "num_tokens": 33885964.0,
      "step": 2747
    },
    {
      "entropy": 1.1429234147071838,
      "epoch": 1.4470774091627172,
      "grad_norm": 0.25770071148872375,
      "learning_rate": 0.00018566967796290954,
      "loss": 0.19852091372013092,
      "mean_token_accuracy": 0.9165006726980209,
      "num_tokens": 33898300.0,
      "step": 2748
    },
    {
      "entropy": 1.1901302635669708,
      "epoch": 1.4476040021063719,
      "grad_norm": 0.2882276177406311,
      "learning_rate": 0.00018565780405027294,
      "loss": 0.2006741464138031,
      "mean_token_accuracy": 0.9199828803539276,
      "num_tokens": 33910636.0,
      "step": 2749
    },
    {
      "entropy": 1.1173346042633057,
      "epoch": 1.4481305950500263,
      "grad_norm": 0.2821039855480194,
      "learning_rate": 0.00018564592564627346,
      "loss": 0.2002512663602829,
      "mean_token_accuracy": 0.9197804033756256,
      "num_tokens": 33922972.0,
      "step": 2750
    },
    {
      "entropy": 1.1957898437976837,
      "epoch": 1.448657187993681,
      "grad_norm": 0.2622511684894562,
      "learning_rate": 0.0001856340427516163,
      "loss": 0.18590982258319855,
      "mean_token_accuracy": 0.9213429242372513,
      "num_tokens": 33935308.0,
      "step": 2751
    },
    {
      "entropy": 1.1354258060455322,
      "epoch": 1.4491837809373354,
      "grad_norm": 0.29836171865463257,
      "learning_rate": 0.00018562215536700684,
      "loss": 0.21873271465301514,
      "mean_token_accuracy": 0.9166739583015442,
      "num_tokens": 33947644.0,
      "step": 2752
    },
    {
      "entropy": 1.1251291930675507,
      "epoch": 1.44971037388099,
      "grad_norm": 0.28312045335769653,
      "learning_rate": 0.00018561026349315075,
      "loss": 0.1926782727241516,
      "mean_token_accuracy": 0.9215989857912064,
      "num_tokens": 33959980.0,
      "step": 2753
    },
    {
      "entropy": 1.1114425361156464,
      "epoch": 1.4502369668246446,
      "grad_norm": 0.2983197569847107,
      "learning_rate": 0.00018559836713075408,
      "loss": 0.20771543681621552,
      "mean_token_accuracy": 0.9185106605291367,
      "num_tokens": 33972316.0,
      "step": 2754
    },
    {
      "entropy": 1.137503057718277,
      "epoch": 1.450763559768299,
      "grad_norm": 0.2714558243751526,
      "learning_rate": 0.0001855864662805229,
      "loss": 0.20167207717895508,
      "mean_token_accuracy": 0.915928527712822,
      "num_tokens": 33984652.0,
      "step": 2755
    },
    {
      "entropy": 1.1457951962947845,
      "epoch": 1.4512901527119537,
      "grad_norm": 0.2937759459018707,
      "learning_rate": 0.00018557456094316378,
      "loss": 0.2275719940662384,
      "mean_token_accuracy": 0.9085268825292587,
      "num_tokens": 33996988.0,
      "step": 2756
    },
    {
      "entropy": 1.1622550189495087,
      "epoch": 1.4518167456556081,
      "grad_norm": 0.28541475534439087,
      "learning_rate": 0.00018556265111938346,
      "loss": 0.2119087278842926,
      "mean_token_accuracy": 0.9126318097114563,
      "num_tokens": 34009324.0,
      "step": 2757
    },
    {
      "entropy": 1.1782312989234924,
      "epoch": 1.4523433385992628,
      "grad_norm": 0.2765743136405945,
      "learning_rate": 0.000185550736809889,
      "loss": 0.2087940275669098,
      "mean_token_accuracy": 0.917047530412674,
      "num_tokens": 34021660.0,
      "step": 2758
    },
    {
      "entropy": 1.1760065853595734,
      "epoch": 1.4528699315429172,
      "grad_norm": 0.3112678527832031,
      "learning_rate": 0.0001855388180153876,
      "loss": 0.22077929973602295,
      "mean_token_accuracy": 0.9056372046470642,
      "num_tokens": 34033996.0,
      "step": 2759
    },
    {
      "entropy": 1.15424644947052,
      "epoch": 1.4533965244865719,
      "grad_norm": 0.2751659154891968,
      "learning_rate": 0.00018552689473658683,
      "loss": 0.1892865151166916,
      "mean_token_accuracy": 0.9226154536008835,
      "num_tokens": 34046332.0,
      "step": 2760
    },
    {
      "entropy": 1.1498244404792786,
      "epoch": 1.4539231174302265,
      "grad_norm": 0.286650687456131,
      "learning_rate": 0.00018551496697419458,
      "loss": 0.20391809940338135,
      "mean_token_accuracy": 0.9146721214056015,
      "num_tokens": 34058668.0,
      "step": 2761
    },
    {
      "entropy": 1.1243232488632202,
      "epoch": 1.454449710373881,
      "grad_norm": 0.27432703971862793,
      "learning_rate": 0.0001855030347289188,
      "loss": 0.20199985802173615,
      "mean_token_accuracy": 0.9169850200414658,
      "num_tokens": 34071004.0,
      "step": 2762
    },
    {
      "entropy": 1.0830981135368347,
      "epoch": 1.4549763033175356,
      "grad_norm": 0.2619624137878418,
      "learning_rate": 0.00018549109800146797,
      "loss": 0.20818959176540375,
      "mean_token_accuracy": 0.9124700576066971,
      "num_tokens": 34083340.0,
      "step": 2763
    },
    {
      "entropy": 1.1054499447345734,
      "epoch": 1.4555028962611902,
      "grad_norm": 0.2768312692642212,
      "learning_rate": 0.00018547915679255063,
      "loss": 0.21730467677116394,
      "mean_token_accuracy": 0.9118637144565582,
      "num_tokens": 34095676.0,
      "step": 2764
    },
    {
      "entropy": 1.0992402136325836,
      "epoch": 1.4560294892048447,
      "grad_norm": 0.27909815311431885,
      "learning_rate": 0.0001854672111028757,
      "loss": 0.21124142408370972,
      "mean_token_accuracy": 0.9161806106567383,
      "num_tokens": 34108012.0,
      "step": 2765
    },
    {
      "entropy": 1.097431093454361,
      "epoch": 1.456556082148499,
      "grad_norm": 0.2552739083766937,
      "learning_rate": 0.0001854552609331523,
      "loss": 0.19730296730995178,
      "mean_token_accuracy": 0.9189020395278931,
      "num_tokens": 34120348.0,
      "step": 2766
    },
    {
      "entropy": 1.0860483348369598,
      "epoch": 1.4570826750921537,
      "grad_norm": 0.25408875942230225,
      "learning_rate": 0.0001854433062840898,
      "loss": 0.1871172934770584,
      "mean_token_accuracy": 0.9257505685091019,
      "num_tokens": 34132684.0,
      "step": 2767
    },
    {
      "entropy": 1.1571218967437744,
      "epoch": 1.4576092680358084,
      "grad_norm": 0.27546441555023193,
      "learning_rate": 0.00018543134715639796,
      "loss": 0.21564674377441406,
      "mean_token_accuracy": 0.9180003553628922,
      "num_tokens": 34145020.0,
      "step": 2768
    },
    {
      "entropy": 1.1057285368442535,
      "epoch": 1.4581358609794628,
      "grad_norm": 0.28240835666656494,
      "learning_rate": 0.00018541938355078668,
      "loss": 0.206221804022789,
      "mean_token_accuracy": 0.9218368530273438,
      "num_tokens": 34157356.0,
      "step": 2769
    },
    {
      "entropy": 1.103806346654892,
      "epoch": 1.4586624539231174,
      "grad_norm": 0.2811862826347351,
      "learning_rate": 0.00018540741546796616,
      "loss": 0.21019676327705383,
      "mean_token_accuracy": 0.9138357639312744,
      "num_tokens": 34169692.0,
      "step": 2770
    },
    {
      "entropy": 1.0941491723060608,
      "epoch": 1.459189046866772,
      "grad_norm": 0.2942333221435547,
      "learning_rate": 0.00018539544290864692,
      "loss": 0.21564032137393951,
      "mean_token_accuracy": 0.9101224541664124,
      "num_tokens": 34182028.0,
      "step": 2771
    },
    {
      "entropy": 1.1355141997337341,
      "epoch": 1.4597156398104265,
      "grad_norm": 0.275642991065979,
      "learning_rate": 0.00018538346587353965,
      "loss": 0.1873530000448227,
      "mean_token_accuracy": 0.9246546924114227,
      "num_tokens": 34194364.0,
      "step": 2772
    },
    {
      "entropy": 1.1305597126483917,
      "epoch": 1.4602422327540812,
      "grad_norm": 0.2881714105606079,
      "learning_rate": 0.0001853714843633554,
      "loss": 0.2074936330318451,
      "mean_token_accuracy": 0.9153118878602982,
      "num_tokens": 34206700.0,
      "step": 2773
    },
    {
      "entropy": 1.1241107285022736,
      "epoch": 1.4607688256977356,
      "grad_norm": 0.2789553999900818,
      "learning_rate": 0.00018535949837880539,
      "loss": 0.2105862945318222,
      "mean_token_accuracy": 0.9097577780485153,
      "num_tokens": 34219036.0,
      "step": 2774
    },
    {
      "entropy": 1.160140573978424,
      "epoch": 1.4612954186413902,
      "grad_norm": 0.28768390417099,
      "learning_rate": 0.00018534750792060123,
      "loss": 0.22017565369606018,
      "mean_token_accuracy": 0.9114931225776672,
      "num_tokens": 34231372.0,
      "step": 2775
    },
    {
      "entropy": 1.1482027173042297,
      "epoch": 1.4618220115850447,
      "grad_norm": 0.2865487039089203,
      "learning_rate": 0.00018533551298945467,
      "loss": 0.2026289850473404,
      "mean_token_accuracy": 0.9202168434858322,
      "num_tokens": 34243708.0,
      "step": 2776
    },
    {
      "entropy": 1.1577872037887573,
      "epoch": 1.4623486045286993,
      "grad_norm": 0.2701823115348816,
      "learning_rate": 0.00018532351358607776,
      "loss": 0.19837477803230286,
      "mean_token_accuracy": 0.9212843328714371,
      "num_tokens": 34256044.0,
      "step": 2777
    },
    {
      "entropy": 1.1805611550807953,
      "epoch": 1.462875197472354,
      "grad_norm": 0.26425090432167053,
      "learning_rate": 0.0001853115097111829,
      "loss": 0.19171418249607086,
      "mean_token_accuracy": 0.9227536916732788,
      "num_tokens": 34268380.0,
      "step": 2778
    },
    {
      "entropy": 1.148007869720459,
      "epoch": 1.4634017904160084,
      "grad_norm": 0.2617400586605072,
      "learning_rate": 0.00018529950136548265,
      "loss": 0.20737731456756592,
      "mean_token_accuracy": 0.9154264628887177,
      "num_tokens": 34280716.0,
      "step": 2779
    },
    {
      "entropy": 1.1435377299785614,
      "epoch": 1.463928383359663,
      "grad_norm": 0.30841219425201416,
      "learning_rate": 0.0001852874885496899,
      "loss": 0.19211293756961823,
      "mean_token_accuracy": 0.9231193363666534,
      "num_tokens": 34293052.0,
      "step": 2780
    },
    {
      "entropy": 1.139043003320694,
      "epoch": 1.4644549763033177,
      "grad_norm": 0.27137625217437744,
      "learning_rate": 0.00018527547126451774,
      "loss": 0.1936057209968567,
      "mean_token_accuracy": 0.9202312529087067,
      "num_tokens": 34305388.0,
      "step": 2781
    },
    {
      "entropy": 1.1803344190120697,
      "epoch": 1.464981569246972,
      "grad_norm": 0.27719569206237793,
      "learning_rate": 0.00018526344951067957,
      "loss": 0.19913798570632935,
      "mean_token_accuracy": 0.9173661470413208,
      "num_tokens": 34317724.0,
      "step": 2782
    },
    {
      "entropy": 1.1878839433193207,
      "epoch": 1.4655081621906265,
      "grad_norm": 0.2626071870326996,
      "learning_rate": 0.00018525142328888915,
      "loss": 0.1828022301197052,
      "mean_token_accuracy": 0.9273342788219452,
      "num_tokens": 34330060.0,
      "step": 2783
    },
    {
      "entropy": 1.1797403395175934,
      "epoch": 1.4660347551342812,
      "grad_norm": 0.2877488434314728,
      "learning_rate": 0.00018523939259986025,
      "loss": 0.20025065541267395,
      "mean_token_accuracy": 0.9134245812892914,
      "num_tokens": 34342396.0,
      "step": 2784
    },
    {
      "entropy": 1.1705401837825775,
      "epoch": 1.4665613480779358,
      "grad_norm": 0.3143531382083893,
      "learning_rate": 0.00018522735744430714,
      "loss": 0.21406927704811096,
      "mean_token_accuracy": 0.9137789458036423,
      "num_tokens": 34354732.0,
      "step": 2785
    },
    {
      "entropy": 1.2104687094688416,
      "epoch": 1.4670879410215902,
      "grad_norm": 0.2813572585582733,
      "learning_rate": 0.00018521531782294433,
      "loss": 0.20670530200004578,
      "mean_token_accuracy": 0.9143806546926498,
      "num_tokens": 34367068.0,
      "step": 2786
    },
    {
      "entropy": 1.097440928220749,
      "epoch": 1.4676145339652449,
      "grad_norm": 0.26660245656967163,
      "learning_rate": 0.00018520327373648644,
      "loss": 0.20290805399417877,
      "mean_token_accuracy": 0.9134158790111542,
      "num_tokens": 34379404.0,
      "step": 2787
    },
    {
      "entropy": 1.1643409132957458,
      "epoch": 1.4681411269088995,
      "grad_norm": 0.30806708335876465,
      "learning_rate": 0.00018519122518564853,
      "loss": 0.23489350080490112,
      "mean_token_accuracy": 0.9051381498575211,
      "num_tokens": 34391740.0,
      "step": 2788
    },
    {
      "entropy": 1.1760393381118774,
      "epoch": 1.468667719852554,
      "grad_norm": 0.30387377738952637,
      "learning_rate": 0.00018517917217114583,
      "loss": 0.20897458493709564,
      "mean_token_accuracy": 0.9153114259243011,
      "num_tokens": 34404076.0,
      "step": 2789
    },
    {
      "entropy": 1.170498788356781,
      "epoch": 1.4691943127962086,
      "grad_norm": 0.3018253445625305,
      "learning_rate": 0.00018516711469369386,
      "loss": 0.22422116994857788,
      "mean_token_accuracy": 0.9047844111919403,
      "num_tokens": 34416412.0,
      "step": 2790
    },
    {
      "entropy": 1.2046408951282501,
      "epoch": 1.469720905739863,
      "grad_norm": 0.2658669650554657,
      "learning_rate": 0.00018515505275400838,
      "loss": 0.18931232392787933,
      "mean_token_accuracy": 0.9234489351511002,
      "num_tokens": 34428748.0,
      "step": 2791
    },
    {
      "entropy": 1.1974002122879028,
      "epoch": 1.4702474986835177,
      "grad_norm": 0.27987316250801086,
      "learning_rate": 0.00018514298635280546,
      "loss": 0.20819401741027832,
      "mean_token_accuracy": 0.9120146036148071,
      "num_tokens": 34441084.0,
      "step": 2792
    },
    {
      "entropy": 1.20514714717865,
      "epoch": 1.470774091627172,
      "grad_norm": 0.26001980900764465,
      "learning_rate": 0.0001851309154908014,
      "loss": 0.19804999232292175,
      "mean_token_accuracy": 0.9171354323625565,
      "num_tokens": 34453420.0,
      "step": 2793
    },
    {
      "entropy": 1.2106068134307861,
      "epoch": 1.4713006845708267,
      "grad_norm": 0.2716507017612457,
      "learning_rate": 0.0001851188401687128,
      "loss": 0.19358912110328674,
      "mean_token_accuracy": 0.9196220189332962,
      "num_tokens": 34465756.0,
      "step": 2794
    },
    {
      "entropy": 1.231193095445633,
      "epoch": 1.4718272775144814,
      "grad_norm": 0.26500481367111206,
      "learning_rate": 0.00018510676038725648,
      "loss": 0.19936080276966095,
      "mean_token_accuracy": 0.9158136695623398,
      "num_tokens": 34478092.0,
      "step": 2795
    },
    {
      "entropy": 1.2479934096336365,
      "epoch": 1.4723538704581358,
      "grad_norm": 0.2973508834838867,
      "learning_rate": 0.00018509467614714954,
      "loss": 0.22026515007019043,
      "mean_token_accuracy": 0.911934033036232,
      "num_tokens": 34490428.0,
      "step": 2796
    },
    {
      "entropy": 1.2747034132480621,
      "epoch": 1.4728804634017905,
      "grad_norm": 0.3173184394836426,
      "learning_rate": 0.0001850825874491094,
      "loss": 0.2203284353017807,
      "mean_token_accuracy": 0.9120416641235352,
      "num_tokens": 34502764.0,
      "step": 2797
    },
    {
      "entropy": 1.1952545940876007,
      "epoch": 1.473407056345445,
      "grad_norm": 0.2990429401397705,
      "learning_rate": 0.0001850704942938536,
      "loss": 0.21640948951244354,
      "mean_token_accuracy": 0.9109156578779221,
      "num_tokens": 34515100.0,
      "step": 2798
    },
    {
      "entropy": 1.2120374739170074,
      "epoch": 1.4739336492890995,
      "grad_norm": 0.2810207009315491,
      "learning_rate": 0.00018505839668210013,
      "loss": 0.2099630981683731,
      "mean_token_accuracy": 0.9159301519393921,
      "num_tokens": 34527436.0,
      "step": 2799
    },
    {
      "entropy": 1.196582019329071,
      "epoch": 1.474460242232754,
      "grad_norm": 0.3099983334541321,
      "learning_rate": 0.00018504629461456716,
      "loss": 0.19392438232898712,
      "mean_token_accuracy": 0.9205189049243927,
      "num_tokens": 34539772.0,
      "step": 2800
    },
    {
      "entropy": 1.161557525396347,
      "epoch": 1.4749868351764086,
      "grad_norm": 0.2738645672798157,
      "learning_rate": 0.000185034188091973,
      "loss": 0.20201784372329712,
      "mean_token_accuracy": 0.9174954891204834,
      "num_tokens": 34552108.0,
      "step": 2801
    },
    {
      "entropy": 1.2087386548519135,
      "epoch": 1.4755134281200633,
      "grad_norm": 0.3033156991004944,
      "learning_rate": 0.00018502207711503646,
      "loss": 0.21341297030448914,
      "mean_token_accuracy": 0.9146996885538101,
      "num_tokens": 34564444.0,
      "step": 2802
    },
    {
      "entropy": 1.2204181849956512,
      "epoch": 1.4760400210637177,
      "grad_norm": 0.29885637760162354,
      "learning_rate": 0.0001850099616844765,
      "loss": 0.23106718063354492,
      "mean_token_accuracy": 0.9073407649993896,
      "num_tokens": 34576780.0,
      "step": 2803
    },
    {
      "entropy": 1.193909078836441,
      "epoch": 1.4765666140073723,
      "grad_norm": 0.29883119463920593,
      "learning_rate": 0.00018499784180101226,
      "loss": 0.20271748304367065,
      "mean_token_accuracy": 0.9174822568893433,
      "num_tokens": 34589116.0,
      "step": 2804
    },
    {
      "entropy": 1.195640653371811,
      "epoch": 1.477093206951027,
      "grad_norm": 0.3074481189250946,
      "learning_rate": 0.0001849857174653633,
      "loss": 0.2207125425338745,
      "mean_token_accuracy": 0.9117613136768341,
      "num_tokens": 34601452.0,
      "step": 2805
    },
    {
      "entropy": 1.1937943994998932,
      "epoch": 1.4776197998946814,
      "grad_norm": 0.30861732363700867,
      "learning_rate": 0.00018497358867824933,
      "loss": 0.23427164554595947,
      "mean_token_accuracy": 0.9043611437082291,
      "num_tokens": 34613788.0,
      "step": 2806
    },
    {
      "entropy": 1.209238052368164,
      "epoch": 1.478146392838336,
      "grad_norm": 0.29167798161506653,
      "learning_rate": 0.00018496145544039038,
      "loss": 0.20825859904289246,
      "mean_token_accuracy": 0.9163528382778168,
      "num_tokens": 34626124.0,
      "step": 2807
    },
    {
      "entropy": 1.2348674535751343,
      "epoch": 1.4786729857819905,
      "grad_norm": 0.30392780900001526,
      "learning_rate": 0.00018494931775250671,
      "loss": 0.213743194937706,
      "mean_token_accuracy": 0.9198024868965149,
      "num_tokens": 34638460.0,
      "step": 2808
    },
    {
      "entropy": 1.2187554240226746,
      "epoch": 1.4791995787256451,
      "grad_norm": 0.26182109117507935,
      "learning_rate": 0.00018493717561531893,
      "loss": 0.20783649384975433,
      "mean_token_accuracy": 0.9161714315414429,
      "num_tokens": 34650796.0,
      "step": 2809
    },
    {
      "entropy": 1.2145146429538727,
      "epoch": 1.4797261716692995,
      "grad_norm": 0.2773604094982147,
      "learning_rate": 0.0001849250290295478,
      "loss": 0.20019277930259705,
      "mean_token_accuracy": 0.9166374951601028,
      "num_tokens": 34663132.0,
      "step": 2810
    },
    {
      "entropy": 1.1926589012145996,
      "epoch": 1.4802527646129542,
      "grad_norm": 0.26068729162216187,
      "learning_rate": 0.00018491287799591436,
      "loss": 0.20261983573436737,
      "mean_token_accuracy": 0.9161369353532791,
      "num_tokens": 34675468.0,
      "step": 2811
    },
    {
      "entropy": 1.2438783645629883,
      "epoch": 1.4807793575566088,
      "grad_norm": 0.28087806701660156,
      "learning_rate": 0.00018490072251513997,
      "loss": 0.20033368468284607,
      "mean_token_accuracy": 0.9154119491577148,
      "num_tokens": 34687804.0,
      "step": 2812
    },
    {
      "entropy": 1.1822409331798553,
      "epoch": 1.4813059505002633,
      "grad_norm": 0.29859158396720886,
      "learning_rate": 0.00018488856258794625,
      "loss": 0.21145831048488617,
      "mean_token_accuracy": 0.9173175245523453,
      "num_tokens": 34700140.0,
      "step": 2813
    },
    {
      "entropy": 1.134483963251114,
      "epoch": 1.481832543443918,
      "grad_norm": 0.27768927812576294,
      "learning_rate": 0.00018487639821505506,
      "loss": 0.19765517115592957,
      "mean_token_accuracy": 0.9187143296003342,
      "num_tokens": 34712476.0,
      "step": 2814
    },
    {
      "entropy": 1.1635059118270874,
      "epoch": 1.4823591363875723,
      "grad_norm": 0.2897265553474426,
      "learning_rate": 0.00018486422939718853,
      "loss": 0.20456260442733765,
      "mean_token_accuracy": 0.9187261015176773,
      "num_tokens": 34724812.0,
      "step": 2815
    },
    {
      "entropy": 1.0995591282844543,
      "epoch": 1.482885729331227,
      "grad_norm": 0.3071244955062866,
      "learning_rate": 0.00018485205613506903,
      "loss": 0.22590550780296326,
      "mean_token_accuracy": 0.9051891267299652,
      "num_tokens": 34737148.0,
      "step": 2816
    },
    {
      "entropy": 1.1542208194732666,
      "epoch": 1.4834123222748814,
      "grad_norm": 0.27810409665107727,
      "learning_rate": 0.00018483987842941922,
      "loss": 0.19803431630134583,
      "mean_token_accuracy": 0.922128900885582,
      "num_tokens": 34749484.0,
      "step": 2817
    },
    {
      "entropy": 1.1647526919841766,
      "epoch": 1.483938915218536,
      "grad_norm": 0.2628480792045593,
      "learning_rate": 0.00018482769628096207,
      "loss": 0.1754264533519745,
      "mean_token_accuracy": 0.9252566993236542,
      "num_tokens": 34761820.0,
      "step": 2818
    },
    {
      "entropy": 1.1333464682102203,
      "epoch": 1.4844655081621907,
      "grad_norm": 0.32637807726860046,
      "learning_rate": 0.00018481550969042069,
      "loss": 0.22181954979896545,
      "mean_token_accuracy": 0.9101926386356354,
      "num_tokens": 34774156.0,
      "step": 2819
    },
    {
      "entropy": 1.125653713941574,
      "epoch": 1.4849921011058451,
      "grad_norm": 0.3150729835033417,
      "learning_rate": 0.00018480331865851853,
      "loss": 0.22693070769309998,
      "mean_token_accuracy": 0.9116131216287613,
      "num_tokens": 34786492.0,
      "step": 2820
    },
    {
      "entropy": 1.1617139875888824,
      "epoch": 1.4855186940494998,
      "grad_norm": 0.3212885856628418,
      "learning_rate": 0.00018479112318597936,
      "loss": 0.22024253010749817,
      "mean_token_accuracy": 0.9142894148826599,
      "num_tokens": 34798828.0,
      "step": 2821
    },
    {
      "entropy": 1.130389004945755,
      "epoch": 1.4860452869931544,
      "grad_norm": 0.26759734749794006,
      "learning_rate": 0.00018477892327352713,
      "loss": 0.18030500411987305,
      "mean_token_accuracy": 0.9289237558841705,
      "num_tokens": 34811164.0,
      "step": 2822
    },
    {
      "entropy": 1.1489178240299225,
      "epoch": 1.4865718799368088,
      "grad_norm": 0.28973206877708435,
      "learning_rate": 0.00018476671892188605,
      "loss": 0.20612166821956635,
      "mean_token_accuracy": 0.9198791682720184,
      "num_tokens": 34823500.0,
      "step": 2823
    },
    {
      "entropy": 1.196854144334793,
      "epoch": 1.4870984728804635,
      "grad_norm": 0.29863879084587097,
      "learning_rate": 0.00018475451013178062,
      "loss": 0.20180033147335052,
      "mean_token_accuracy": 0.9196022301912308,
      "num_tokens": 34835836.0,
      "step": 2824
    },
    {
      "entropy": 1.1838077902793884,
      "epoch": 1.487625065824118,
      "grad_norm": 0.30217835307121277,
      "learning_rate": 0.00018474229690393568,
      "loss": 0.19603033363819122,
      "mean_token_accuracy": 0.9209345132112503,
      "num_tokens": 34848172.0,
      "step": 2825
    },
    {
      "entropy": 1.256701648235321,
      "epoch": 1.4881516587677726,
      "grad_norm": 0.2757579982280731,
      "learning_rate": 0.00018473007923907616,
      "loss": 0.19625365734100342,
      "mean_token_accuracy": 0.9234248697757721,
      "num_tokens": 34860508.0,
      "step": 2826
    },
    {
      "entropy": 1.2282088100910187,
      "epoch": 1.488678251711427,
      "grad_norm": 0.2982552945613861,
      "learning_rate": 0.0001847178571379274,
      "loss": 0.20196665823459625,
      "mean_token_accuracy": 0.9129941761493683,
      "num_tokens": 34872844.0,
      "step": 2827
    },
    {
      "entropy": 1.1983471512794495,
      "epoch": 1.4892048446550816,
      "grad_norm": 0.25875699520111084,
      "learning_rate": 0.00018470563060121498,
      "loss": 0.19371843338012695,
      "mean_token_accuracy": 0.9161286354064941,
      "num_tokens": 34885180.0,
      "step": 2828
    },
    {
      "entropy": 1.2658935189247131,
      "epoch": 1.4897314375987363,
      "grad_norm": 0.28629231452941895,
      "learning_rate": 0.0001846933996296647,
      "loss": 0.20451796054840088,
      "mean_token_accuracy": 0.9187720865011215,
      "num_tokens": 34897516.0,
      "step": 2829
    },
    {
      "entropy": 1.2226274013519287,
      "epoch": 1.4902580305423907,
      "grad_norm": 0.26883167028427124,
      "learning_rate": 0.00018468116422400258,
      "loss": 0.19514040648937225,
      "mean_token_accuracy": 0.9247150421142578,
      "num_tokens": 34909852.0,
      "step": 2830
    },
    {
      "entropy": 1.2325372695922852,
      "epoch": 1.4907846234860453,
      "grad_norm": 0.27920812368392944,
      "learning_rate": 0.00018466892438495503,
      "loss": 0.20849348604679108,
      "mean_token_accuracy": 0.9168450087308884,
      "num_tokens": 34922188.0,
      "step": 2831
    },
    {
      "entropy": 1.2324140071868896,
      "epoch": 1.4913112164296998,
      "grad_norm": 0.29039597511291504,
      "learning_rate": 0.00018465668011324863,
      "loss": 0.21684204041957855,
      "mean_token_accuracy": 0.9118513017892838,
      "num_tokens": 34934524.0,
      "step": 2832
    },
    {
      "entropy": 1.158925324678421,
      "epoch": 1.4918378093733544,
      "grad_norm": 0.2734275162220001,
      "learning_rate": 0.0001846444314096103,
      "loss": 0.1901196390390396,
      "mean_token_accuracy": 0.9206565618515015,
      "num_tokens": 34946860.0,
      "step": 2833
    },
    {
      "entropy": 1.179129183292389,
      "epoch": 1.4923644023170088,
      "grad_norm": 0.24893783032894135,
      "learning_rate": 0.0001846321782747671,
      "loss": 0.1810513734817505,
      "mean_token_accuracy": 0.925764262676239,
      "num_tokens": 34959196.0,
      "step": 2834
    },
    {
      "entropy": 1.1447475254535675,
      "epoch": 1.4928909952606635,
      "grad_norm": 0.2840208411216736,
      "learning_rate": 0.00018461992070944652,
      "loss": 0.20713338255882263,
      "mean_token_accuracy": 0.9199363738298416,
      "num_tokens": 34971532.0,
      "step": 2835
    },
    {
      "entropy": 1.2216734886169434,
      "epoch": 1.4934175882043181,
      "grad_norm": 0.28631988167762756,
      "learning_rate": 0.00018460765871437614,
      "loss": 0.20792193710803986,
      "mean_token_accuracy": 0.9214426577091217,
      "num_tokens": 34983868.0,
      "step": 2836
    },
    {
      "entropy": 1.1670808494091034,
      "epoch": 1.4939441811479726,
      "grad_norm": 0.2994043529033661,
      "learning_rate": 0.00018459539229028388,
      "loss": 0.2228715866804123,
      "mean_token_accuracy": 0.9080910384654999,
      "num_tokens": 34996204.0,
      "step": 2837
    },
    {
      "entropy": 1.1826681196689606,
      "epoch": 1.4944707740916272,
      "grad_norm": 0.2632300853729248,
      "learning_rate": 0.00018458312143789798,
      "loss": 0.19896544516086578,
      "mean_token_accuracy": 0.9158213883638382,
      "num_tokens": 35008540.0,
      "step": 2838
    },
    {
      "entropy": 1.1997667253017426,
      "epoch": 1.4949973670352819,
      "grad_norm": 0.28206589818000793,
      "learning_rate": 0.00018457084615794685,
      "loss": 0.18993441760540009,
      "mean_token_accuracy": 0.9219527989625931,
      "num_tokens": 35020876.0,
      "step": 2839
    },
    {
      "entropy": 1.2673965096473694,
      "epoch": 1.4955239599789363,
      "grad_norm": 0.29593348503112793,
      "learning_rate": 0.00018455856645115923,
      "loss": 0.2120542973279953,
      "mean_token_accuracy": 0.9183967113494873,
      "num_tokens": 35033212.0,
      "step": 2840
    },
    {
      "entropy": 1.2509069740772247,
      "epoch": 1.4960505529225907,
      "grad_norm": 0.4201185405254364,
      "learning_rate": 0.0001845462823182641,
      "loss": 0.21505090594291687,
      "mean_token_accuracy": 0.9150316119194031,
      "num_tokens": 35045548.0,
      "step": 2841
    },
    {
      "entropy": 1.3054153323173523,
      "epoch": 1.4965771458662454,
      "grad_norm": 0.2993493378162384,
      "learning_rate": 0.0001845339937599906,
      "loss": 0.1938711553812027,
      "mean_token_accuracy": 0.91843082010746,
      "num_tokens": 35057884.0,
      "step": 2842
    },
    {
      "entropy": 1.2852679789066315,
      "epoch": 1.4971037388099,
      "grad_norm": 0.28747817873954773,
      "learning_rate": 0.0001845217007770684,
      "loss": 0.18465346097946167,
      "mean_token_accuracy": 0.9223109483718872,
      "num_tokens": 35070220.0,
      "step": 2843
    },
    {
      "entropy": 1.3424168229103088,
      "epoch": 1.4976303317535544,
      "grad_norm": 0.2738924026489258,
      "learning_rate": 0.0001845094033702271,
      "loss": 0.19440293312072754,
      "mean_token_accuracy": 0.9208139926195145,
      "num_tokens": 35082556.0,
      "step": 2844
    },
    {
      "entropy": 1.2790354490280151,
      "epoch": 1.498156924697209,
      "grad_norm": 0.26316049695014954,
      "learning_rate": 0.00018449710154019687,
      "loss": 0.18655042350292206,
      "mean_token_accuracy": 0.9237352013587952,
      "num_tokens": 35094892.0,
      "step": 2845
    },
    {
      "entropy": 1.3539345264434814,
      "epoch": 1.4986835176408637,
      "grad_norm": 0.29045310616493225,
      "learning_rate": 0.00018448479528770784,
      "loss": 0.19067838788032532,
      "mean_token_accuracy": 0.9230899512767792,
      "num_tokens": 35107228.0,
      "step": 2846
    },
    {
      "entropy": 1.4051901996135712,
      "epoch": 1.4992101105845181,
      "grad_norm": 0.256143718957901,
      "learning_rate": 0.0001844724846134907,
      "loss": 0.2035890519618988,
      "mean_token_accuracy": 0.9209445118904114,
      "num_tokens": 35119564.0,
      "step": 2847
    },
    {
      "entropy": 1.4236982464790344,
      "epoch": 1.4997367035281728,
      "grad_norm": 0.28411853313446045,
      "learning_rate": 0.00018446016951827619,
      "loss": 0.20746804773807526,
      "mean_token_accuracy": 0.9163666069507599,
      "num_tokens": 35131900.0,
      "step": 2848
    },
    {
      "entropy": 1.4779279828071594,
      "epoch": 1.5002632964718274,
      "grad_norm": 0.31305235624313354,
      "learning_rate": 0.0001844478500027954,
      "loss": 0.20554354786872864,
      "mean_token_accuracy": 0.9199778288602829,
      "num_tokens": 35144236.0,
      "step": 2849
    },
    {
      "entropy": 1.436387151479721,
      "epoch": 1.5007898894154819,
      "grad_norm": 0.2702447772026062,
      "learning_rate": 0.00018443552606777966,
      "loss": 0.18683519959449768,
      "mean_token_accuracy": 0.9251002669334412,
      "num_tokens": 35156572.0,
      "step": 2850
    },
    {
      "entropy": 1.4672654867172241,
      "epoch": 1.5013164823591363,
      "grad_norm": 0.2640582323074341,
      "learning_rate": 0.0001844231977139606,
      "loss": 0.17821809649467468,
      "mean_token_accuracy": 0.9315766096115112,
      "num_tokens": 35168908.0,
      "step": 2851
    },
    {
      "entropy": 1.4444270133972168,
      "epoch": 1.501843075302791,
      "grad_norm": 0.29034173488616943,
      "learning_rate": 0.00018441086494207004,
      "loss": 0.20143531262874603,
      "mean_token_accuracy": 0.9211236834526062,
      "num_tokens": 35181244.0,
      "step": 2852
    },
    {
      "entropy": 1.447830855846405,
      "epoch": 1.5023696682464456,
      "grad_norm": 0.26047781109809875,
      "learning_rate": 0.00018439852775284016,
      "loss": 0.1932010054588318,
      "mean_token_accuracy": 0.9227436780929565,
      "num_tokens": 35193580.0,
      "step": 2853
    },
    {
      "entropy": 1.5056044459342957,
      "epoch": 1.5028962611901,
      "grad_norm": 0.2798239588737488,
      "learning_rate": 0.0001843861861470033,
      "loss": 0.197336345911026,
      "mean_token_accuracy": 0.9250242561101913,
      "num_tokens": 35205916.0,
      "step": 2854
    },
    {
      "entropy": 1.3938361406326294,
      "epoch": 1.5034228541337546,
      "grad_norm": 0.2706337571144104,
      "learning_rate": 0.00018437384012529213,
      "loss": 0.19167521595954895,
      "mean_token_accuracy": 0.9251234829425812,
      "num_tokens": 35218252.0,
      "step": 2855
    },
    {
      "entropy": 1.41653311252594,
      "epoch": 1.5039494470774093,
      "grad_norm": 0.25845250487327576,
      "learning_rate": 0.00018436148968843956,
      "loss": 0.18029874563217163,
      "mean_token_accuracy": 0.9265570193529129,
      "num_tokens": 35230588.0,
      "step": 2856
    },
    {
      "entropy": 1.449202448129654,
      "epoch": 1.5044760400210637,
      "grad_norm": 0.26958736777305603,
      "learning_rate": 0.00018434913483717871,
      "loss": 0.19115948677062988,
      "mean_token_accuracy": 0.9232787936925888,
      "num_tokens": 35242924.0,
      "step": 2857
    },
    {
      "entropy": 1.4373081922531128,
      "epoch": 1.5050026329647181,
      "grad_norm": 0.2909432649612427,
      "learning_rate": 0.00018433677557224312,
      "loss": 0.2226668894290924,
      "mean_token_accuracy": 0.9091159552335739,
      "num_tokens": 35255260.0,
      "step": 2858
    },
    {
      "entropy": 1.4466613829135895,
      "epoch": 1.5055292259083728,
      "grad_norm": 0.28296786546707153,
      "learning_rate": 0.0001843244118943664,
      "loss": 0.19658687710762024,
      "mean_token_accuracy": 0.9204612225294113,
      "num_tokens": 35267596.0,
      "step": 2859
    },
    {
      "entropy": 1.4727692604064941,
      "epoch": 1.5060558188520274,
      "grad_norm": 0.27067407965660095,
      "learning_rate": 0.00018431204380428258,
      "loss": 0.17784729599952698,
      "mean_token_accuracy": 0.9264029860496521,
      "num_tokens": 35279932.0,
      "step": 2860
    },
    {
      "entropy": 1.4386701583862305,
      "epoch": 1.5065824117956819,
      "grad_norm": 0.28752195835113525,
      "learning_rate": 0.0001842996713027258,
      "loss": 0.20581099390983582,
      "mean_token_accuracy": 0.9142017662525177,
      "num_tokens": 35292268.0,
      "step": 2861
    },
    {
      "entropy": 1.4239437580108643,
      "epoch": 1.5071090047393365,
      "grad_norm": 0.274856835603714,
      "learning_rate": 0.00018428729439043062,
      "loss": 0.18968544900417328,
      "mean_token_accuracy": 0.921679362654686,
      "num_tokens": 35304604.0,
      "step": 2862
    },
    {
      "entropy": 1.4606831967830658,
      "epoch": 1.5076355976829912,
      "grad_norm": 0.2826034426689148,
      "learning_rate": 0.00018427491306813171,
      "loss": 0.19272050261497498,
      "mean_token_accuracy": 0.9195594191551208,
      "num_tokens": 35316940.0,
      "step": 2863
    },
    {
      "entropy": 1.4668301343917847,
      "epoch": 1.5081621906266456,
      "grad_norm": 0.32095038890838623,
      "learning_rate": 0.00018426252733656415,
      "loss": 0.20584404468536377,
      "mean_token_accuracy": 0.915522113442421,
      "num_tokens": 35329276.0,
      "step": 2864
    },
    {
      "entropy": 1.3869444131851196,
      "epoch": 1.5086887835703,
      "grad_norm": 0.28584420680999756,
      "learning_rate": 0.00018425013719646318,
      "loss": 0.21644830703735352,
      "mean_token_accuracy": 0.9124075770378113,
      "num_tokens": 35341612.0,
      "step": 2865
    },
    {
      "entropy": 1.3982872068881989,
      "epoch": 1.5092153765139549,
      "grad_norm": 0.2921147048473358,
      "learning_rate": 0.00018423774264856433,
      "loss": 0.22454163432121277,
      "mean_token_accuracy": 0.9074548035860062,
      "num_tokens": 35353948.0,
      "step": 2866
    },
    {
      "entropy": 1.4254017174243927,
      "epoch": 1.5097419694576093,
      "grad_norm": 0.29195889830589294,
      "learning_rate": 0.00018422534369360335,
      "loss": 0.20947468280792236,
      "mean_token_accuracy": 0.9152242243289948,
      "num_tokens": 35366284.0,
      "step": 2867
    },
    {
      "entropy": 1.420524001121521,
      "epoch": 1.5102685624012637,
      "grad_norm": 0.28271856904029846,
      "learning_rate": 0.00018421294033231638,
      "loss": 0.19233407080173492,
      "mean_token_accuracy": 0.922194093465805,
      "num_tokens": 35378620.0,
      "step": 2868
    },
    {
      "entropy": 1.4296413362026215,
      "epoch": 1.5107951553449184,
      "grad_norm": 0.29727426171302795,
      "learning_rate": 0.00018420053256543967,
      "loss": 0.21647101640701294,
      "mean_token_accuracy": 0.9113524258136749,
      "num_tokens": 35390956.0,
      "step": 2869
    },
    {
      "entropy": 1.383836716413498,
      "epoch": 1.511321748288573,
      "grad_norm": 0.2626645565032959,
      "learning_rate": 0.00018418812039370982,
      "loss": 0.17489789426326752,
      "mean_token_accuracy": 0.9316018223762512,
      "num_tokens": 35403292.0,
      "step": 2870
    },
    {
      "entropy": 1.3209747970104218,
      "epoch": 1.5118483412322274,
      "grad_norm": 0.2824258804321289,
      "learning_rate": 0.00018417570381786365,
      "loss": 0.2016294300556183,
      "mean_token_accuracy": 0.9183167815208435,
      "num_tokens": 35415628.0,
      "step": 2871
    },
    {
      "entropy": 1.3184476792812347,
      "epoch": 1.512374934175882,
      "grad_norm": 0.290884405374527,
      "learning_rate": 0.00018416328283863827,
      "loss": 0.20478402078151703,
      "mean_token_accuracy": 0.9130069315433502,
      "num_tokens": 35427964.0,
      "step": 2872
    },
    {
      "entropy": 1.268851488828659,
      "epoch": 1.5129015271195367,
      "grad_norm": 0.2771494686603546,
      "learning_rate": 0.00018415085745677106,
      "loss": 0.1991674304008484,
      "mean_token_accuracy": 0.9160973429679871,
      "num_tokens": 35440300.0,
      "step": 2873
    },
    {
      "entropy": 1.3079243302345276,
      "epoch": 1.5134281200631912,
      "grad_norm": 0.30500468611717224,
      "learning_rate": 0.0001841384276729996,
      "loss": 0.22135591506958008,
      "mean_token_accuracy": 0.9075585156679153,
      "num_tokens": 35452636.0,
      "step": 2874
    },
    {
      "entropy": 1.2402482628822327,
      "epoch": 1.5139547130068456,
      "grad_norm": 0.2842849791049957,
      "learning_rate": 0.00018412599348806185,
      "loss": 0.2187984585762024,
      "mean_token_accuracy": 0.9123281091451645,
      "num_tokens": 35464972.0,
      "step": 2875
    },
    {
      "entropy": 1.1900142133235931,
      "epoch": 1.5144813059505002,
      "grad_norm": 0.27149152755737305,
      "learning_rate": 0.00018411355490269589,
      "loss": 0.20663440227508545,
      "mean_token_accuracy": 0.9168604612350464,
      "num_tokens": 35477308.0,
      "step": 2876
    },
    {
      "entropy": 1.2160408794879913,
      "epoch": 1.5150078988941549,
      "grad_norm": 0.3139822483062744,
      "learning_rate": 0.00018410111191764013,
      "loss": 0.19389238953590393,
      "mean_token_accuracy": 0.9242724925279617,
      "num_tokens": 35489644.0,
      "step": 2877
    },
    {
      "entropy": 1.2444290518760681,
      "epoch": 1.5155344918378093,
      "grad_norm": 0.2705695331096649,
      "learning_rate": 0.00018408866453363326,
      "loss": 0.19504526257514954,
      "mean_token_accuracy": 0.921023428440094,
      "num_tokens": 35501980.0,
      "step": 2878
    },
    {
      "entropy": 1.2649579048156738,
      "epoch": 1.516061084781464,
      "grad_norm": 0.2965560853481293,
      "learning_rate": 0.00018407621275141417,
      "loss": 0.20819805562496185,
      "mean_token_accuracy": 0.9142050892114639,
      "num_tokens": 35514316.0,
      "step": 2879
    },
    {
      "entropy": 1.270400732755661,
      "epoch": 1.5165876777251186,
      "grad_norm": 0.2897675931453705,
      "learning_rate": 0.0001840637565717221,
      "loss": 0.213098406791687,
      "mean_token_accuracy": 0.910462811589241,
      "num_tokens": 35526652.0,
      "step": 2880
    },
    {
      "entropy": 1.2283774614334106,
      "epoch": 1.517114270668773,
      "grad_norm": 0.29091915488243103,
      "learning_rate": 0.00018405129599529646,
      "loss": 0.19621555507183075,
      "mean_token_accuracy": 0.9233623296022415,
      "num_tokens": 35538988.0,
      "step": 2881
    },
    {
      "entropy": 1.2129921317100525,
      "epoch": 1.5176408636124274,
      "grad_norm": 0.36581671237945557,
      "learning_rate": 0.00018403883102287698,
      "loss": 0.19229567050933838,
      "mean_token_accuracy": 0.9180682301521301,
      "num_tokens": 35551324.0,
      "step": 2882
    },
    {
      "entropy": 1.2191410660743713,
      "epoch": 1.518167456556082,
      "grad_norm": 0.2867785394191742,
      "learning_rate": 0.00018402636165520367,
      "loss": 0.211622953414917,
      "mean_token_accuracy": 0.9161818474531174,
      "num_tokens": 35563660.0,
      "step": 2883
    },
    {
      "entropy": 1.2190195620059967,
      "epoch": 1.5186940494997367,
      "grad_norm": 0.25288647413253784,
      "learning_rate": 0.0001840138878930167,
      "loss": 0.19131462275981903,
      "mean_token_accuracy": 0.9191330075263977,
      "num_tokens": 35575996.0,
      "step": 2884
    },
    {
      "entropy": 1.2443910837173462,
      "epoch": 1.5192206424433912,
      "grad_norm": 0.27648887038230896,
      "learning_rate": 0.00018400140973705658,
      "loss": 0.20976126194000244,
      "mean_token_accuracy": 0.9147668480873108,
      "num_tokens": 35588332.0,
      "step": 2885
    },
    {
      "entropy": 1.2364069521427155,
      "epoch": 1.5197472353870458,
      "grad_norm": 0.2660056948661804,
      "learning_rate": 0.00018398892718806412,
      "loss": 0.18810099363327026,
      "mean_token_accuracy": 0.9223190397024155,
      "num_tokens": 35600668.0,
      "step": 2886
    },
    {
      "entropy": 1.2349788844585419,
      "epoch": 1.5202738283307005,
      "grad_norm": 0.25507408380508423,
      "learning_rate": 0.00018397644024678028,
      "loss": 0.20845112204551697,
      "mean_token_accuracy": 0.9182122647762299,
      "num_tokens": 35613004.0,
      "step": 2887
    },
    {
      "entropy": 1.2452967166900635,
      "epoch": 1.5208004212743549,
      "grad_norm": 0.2838190495967865,
      "learning_rate": 0.0001839639489139463,
      "loss": 0.19937308132648468,
      "mean_token_accuracy": 0.9176253229379654,
      "num_tokens": 35625340.0,
      "step": 2888
    },
    {
      "entropy": 1.214994490146637,
      "epoch": 1.5213270142180095,
      "grad_norm": 0.268673837184906,
      "learning_rate": 0.0001839514531903038,
      "loss": 0.1951359212398529,
      "mean_token_accuracy": 0.9214799553155899,
      "num_tokens": 35637676.0,
      "step": 2889
    },
    {
      "entropy": 1.270435869693756,
      "epoch": 1.5218536071616642,
      "grad_norm": 0.3197346031665802,
      "learning_rate": 0.00018393895307659456,
      "loss": 0.20956359803676605,
      "mean_token_accuracy": 0.9150731861591339,
      "num_tokens": 35650012.0,
      "step": 2890
    },
    {
      "entropy": 1.3132632076740265,
      "epoch": 1.5223802001053186,
      "grad_norm": 0.299365758895874,
      "learning_rate": 0.0001839264485735606,
      "loss": 0.204919695854187,
      "mean_token_accuracy": 0.912919282913208,
      "num_tokens": 35662348.0,
      "step": 2891
    },
    {
      "entropy": 1.306254893541336,
      "epoch": 1.522906793048973,
      "grad_norm": 0.2999991178512573,
      "learning_rate": 0.00018391393968194428,
      "loss": 0.221295565366745,
      "mean_token_accuracy": 0.9113573879003525,
      "num_tokens": 35674684.0,
      "step": 2892
    },
    {
      "entropy": 1.3227771818637848,
      "epoch": 1.5234333859926277,
      "grad_norm": 0.29826298356056213,
      "learning_rate": 0.00018390142640248817,
      "loss": 0.2028276026248932,
      "mean_token_accuracy": 0.9149623513221741,
      "num_tokens": 35687020.0,
      "step": 2893
    },
    {
      "entropy": 1.2494766116142273,
      "epoch": 1.5239599789362823,
      "grad_norm": 0.29955431818962097,
      "learning_rate": 0.0001838889087359351,
      "loss": 0.2077026069164276,
      "mean_token_accuracy": 0.918857529759407,
      "num_tokens": 35699356.0,
      "step": 2894
    },
    {
      "entropy": 1.2577630579471588,
      "epoch": 1.5244865718799367,
      "grad_norm": 0.2640293836593628,
      "learning_rate": 0.00018387638668302818,
      "loss": 0.20387566089630127,
      "mean_token_accuracy": 0.9126008599996567,
      "num_tokens": 35711692.0,
      "step": 2895
    },
    {
      "entropy": 1.3035214841365814,
      "epoch": 1.5250131648235914,
      "grad_norm": 0.27383658289909363,
      "learning_rate": 0.00018386386024451076,
      "loss": 0.19004280865192413,
      "mean_token_accuracy": 0.9235143810510635,
      "num_tokens": 35724028.0,
      "step": 2896
    },
    {
      "entropy": 1.3059624433517456,
      "epoch": 1.525539757767246,
      "grad_norm": 0.3266020715236664,
      "learning_rate": 0.00018385132942112646,
      "loss": 0.21989792585372925,
      "mean_token_accuracy": 0.9094354212284088,
      "num_tokens": 35736364.0,
      "step": 2897
    },
    {
      "entropy": 1.255292147397995,
      "epoch": 1.5260663507109005,
      "grad_norm": 0.29097259044647217,
      "learning_rate": 0.0001838387942136192,
      "loss": 0.20556314289569855,
      "mean_token_accuracy": 0.9147609919309616,
      "num_tokens": 35748700.0,
      "step": 2898
    },
    {
      "entropy": 1.293520450592041,
      "epoch": 1.526592943654555,
      "grad_norm": 0.2748951017856598,
      "learning_rate": 0.00018382625462273305,
      "loss": 0.1994977742433548,
      "mean_token_accuracy": 0.9188051521778107,
      "num_tokens": 35761036.0,
      "step": 2899
    },
    {
      "entropy": 1.2601102590560913,
      "epoch": 1.5271195365982095,
      "grad_norm": 0.25538790225982666,
      "learning_rate": 0.00018381371064921247,
      "loss": 0.19155453145503998,
      "mean_token_accuracy": 0.9221813231706619,
      "num_tokens": 35773372.0,
      "step": 2900
    },
    {
      "entropy": 1.302675575017929,
      "epoch": 1.5276461295418642,
      "grad_norm": 0.27591919898986816,
      "learning_rate": 0.0001838011622938021,
      "loss": 0.19985324144363403,
      "mean_token_accuracy": 0.9170637875795364,
      "num_tokens": 35785708.0,
      "step": 2901
    },
    {
      "entropy": 1.3327484726905823,
      "epoch": 1.5281727224855186,
      "grad_norm": 0.2705720365047455,
      "learning_rate": 0.0001837886095572469,
      "loss": 0.1857602745294571,
      "mean_token_accuracy": 0.9223667979240417,
      "num_tokens": 35798044.0,
      "step": 2902
    },
    {
      "entropy": 1.308729499578476,
      "epoch": 1.5286993154291733,
      "grad_norm": 0.2666257321834564,
      "learning_rate": 0.00018377605244029204,
      "loss": 0.18631060421466827,
      "mean_token_accuracy": 0.9259549677371979,
      "num_tokens": 35810380.0,
      "step": 2903
    },
    {
      "entropy": 1.3280511498451233,
      "epoch": 1.529225908372828,
      "grad_norm": 0.2811123728752136,
      "learning_rate": 0.00018376349094368288,
      "loss": 0.19803892076015472,
      "mean_token_accuracy": 0.9206533879041672,
      "num_tokens": 35822716.0,
      "step": 2904
    },
    {
      "entropy": 1.2749126553535461,
      "epoch": 1.5297525013164823,
      "grad_norm": 0.27242687344551086,
      "learning_rate": 0.00018375092506816524,
      "loss": 0.18691426515579224,
      "mean_token_accuracy": 0.921400859951973,
      "num_tokens": 35835052.0,
      "step": 2905
    },
    {
      "entropy": 1.3258703649044037,
      "epoch": 1.5302790942601368,
      "grad_norm": 0.2818097770214081,
      "learning_rate": 0.000183738354814485,
      "loss": 0.19853971898555756,
      "mean_token_accuracy": 0.9171813875436783,
      "num_tokens": 35847388.0,
      "step": 2906
    },
    {
      "entropy": 1.331975519657135,
      "epoch": 1.5308056872037916,
      "grad_norm": 0.29557204246520996,
      "learning_rate": 0.00018372578018338844,
      "loss": 0.19931554794311523,
      "mean_token_accuracy": 0.9204997718334198,
      "num_tokens": 35859724.0,
      "step": 2907
    },
    {
      "entropy": 1.3371560275554657,
      "epoch": 1.531332280147446,
      "grad_norm": 0.30124926567077637,
      "learning_rate": 0.00018371320117562199,
      "loss": 0.20128431916236877,
      "mean_token_accuracy": 0.9167551845312119,
      "num_tokens": 35872060.0,
      "step": 2908
    },
    {
      "entropy": 1.2963175475597382,
      "epoch": 1.5318588730911005,
      "grad_norm": 0.26844584941864014,
      "learning_rate": 0.00018370061779193243,
      "loss": 0.18807059526443481,
      "mean_token_accuracy": 0.9203794151544571,
      "num_tokens": 35884396.0,
      "step": 2909
    },
    {
      "entropy": 1.2872518599033356,
      "epoch": 1.5323854660347551,
      "grad_norm": 0.2787841260433197,
      "learning_rate": 0.0001836880300330668,
      "loss": 0.18961066007614136,
      "mean_token_accuracy": 0.9234008938074112,
      "num_tokens": 35896732.0,
      "step": 2910
    },
    {
      "entropy": 1.3187111616134644,
      "epoch": 1.5329120589784098,
      "grad_norm": 0.26749753952026367,
      "learning_rate": 0.00018367543789977225,
      "loss": 0.1896454393863678,
      "mean_token_accuracy": 0.919704869389534,
      "num_tokens": 35909068.0,
      "step": 2911
    },
    {
      "entropy": 1.2901692986488342,
      "epoch": 1.5334386519220642,
      "grad_norm": 0.28337034583091736,
      "learning_rate": 0.00018366284139279638,
      "loss": 0.19293320178985596,
      "mean_token_accuracy": 0.9212239682674408,
      "num_tokens": 35921404.0,
      "step": 2912
    },
    {
      "entropy": 1.2707647681236267,
      "epoch": 1.5339652448657188,
      "grad_norm": 0.2982810437679291,
      "learning_rate": 0.00018365024051288694,
      "loss": 0.19825252890586853,
      "mean_token_accuracy": 0.9236468225717545,
      "num_tokens": 35933740.0,
      "step": 2913
    },
    {
      "entropy": 1.3728724122047424,
      "epoch": 1.5344918378093735,
      "grad_norm": 0.2963562309741974,
      "learning_rate": 0.000183637635260792,
      "loss": 0.2167905569076538,
      "mean_token_accuracy": 0.9136679619550705,
      "num_tokens": 35946076.0,
      "step": 2914
    },
    {
      "entropy": 1.3078700304031372,
      "epoch": 1.535018430753028,
      "grad_norm": 0.2905045747756958,
      "learning_rate": 0.00018362502563725987,
      "loss": 0.22399064898490906,
      "mean_token_accuracy": 0.9054947942495346,
      "num_tokens": 35958412.0,
      "step": 2915
    },
    {
      "entropy": 1.292427510023117,
      "epoch": 1.5355450236966823,
      "grad_norm": 0.25560230016708374,
      "learning_rate": 0.0001836124116430391,
      "loss": 0.17694160342216492,
      "mean_token_accuracy": 0.9285813271999359,
      "num_tokens": 35970748.0,
      "step": 2916
    },
    {
      "entropy": 1.3606195151805878,
      "epoch": 1.536071616640337,
      "grad_norm": 0.2602185904979706,
      "learning_rate": 0.00018359979327887846,
      "loss": 0.20653316378593445,
      "mean_token_accuracy": 0.9120610505342484,
      "num_tokens": 35983084.0,
      "step": 2917
    },
    {
      "entropy": 1.340090036392212,
      "epoch": 1.5365982095839916,
      "grad_norm": 0.2481660097837448,
      "learning_rate": 0.00018358717054552706,
      "loss": 0.17147809267044067,
      "mean_token_accuracy": 0.9288907945156097,
      "num_tokens": 35995420.0,
      "step": 2918
    },
    {
      "entropy": 1.317726194858551,
      "epoch": 1.537124802527646,
      "grad_norm": 0.2963104546070099,
      "learning_rate": 0.00018357454344373426,
      "loss": 0.22185461223125458,
      "mean_token_accuracy": 0.9066198915243149,
      "num_tokens": 36007756.0,
      "step": 2919
    },
    {
      "entropy": 1.2521053552627563,
      "epoch": 1.5376513954713007,
      "grad_norm": 0.2623257040977478,
      "learning_rate": 0.00018356191197424964,
      "loss": 0.19398631155490875,
      "mean_token_accuracy": 0.9210677295923233,
      "num_tokens": 36020092.0,
      "step": 2920
    },
    {
      "entropy": 1.3271432518959045,
      "epoch": 1.5381779884149553,
      "grad_norm": 0.29276102781295776,
      "learning_rate": 0.00018354927613782306,
      "loss": 0.21266180276870728,
      "mean_token_accuracy": 0.9133126139640808,
      "num_tokens": 36032428.0,
      "step": 2921
    },
    {
      "entropy": 1.299724519252777,
      "epoch": 1.5387045813586098,
      "grad_norm": 0.27011600136756897,
      "learning_rate": 0.00018353663593520466,
      "loss": 0.20557110011577606,
      "mean_token_accuracy": 0.9162522554397583,
      "num_tokens": 36044764.0,
      "step": 2922
    },
    {
      "entropy": 1.3156613111495972,
      "epoch": 1.5392311743022642,
      "grad_norm": 0.2779374420642853,
      "learning_rate": 0.0001835239913671447,
      "loss": 0.2038457691669464,
      "mean_token_accuracy": 0.913342073559761,
      "num_tokens": 36057100.0,
      "step": 2923
    },
    {
      "entropy": 1.354040265083313,
      "epoch": 1.539757767245919,
      "grad_norm": 0.2893989384174347,
      "learning_rate": 0.000183511342434394,
      "loss": 0.22883382439613342,
      "mean_token_accuracy": 0.9079950302839279,
      "num_tokens": 36069436.0,
      "step": 2924
    },
    {
      "entropy": 1.3318619132041931,
      "epoch": 1.5402843601895735,
      "grad_norm": 0.2893321216106415,
      "learning_rate": 0.0001834986891377033,
      "loss": 0.2252848893404007,
      "mean_token_accuracy": 0.9112117439508438,
      "num_tokens": 36081772.0,
      "step": 2925
    },
    {
      "entropy": 1.4141647219657898,
      "epoch": 1.540810953133228,
      "grad_norm": 0.28723177313804626,
      "learning_rate": 0.0001834860314778238,
      "loss": 0.2135084867477417,
      "mean_token_accuracy": 0.9165365993976593,
      "num_tokens": 36094108.0,
      "step": 2926
    },
    {
      "entropy": 1.3629777133464813,
      "epoch": 1.5413375460768826,
      "grad_norm": 0.2711016833782196,
      "learning_rate": 0.00018347336945550696,
      "loss": 0.20635992288589478,
      "mean_token_accuracy": 0.9175506234169006,
      "num_tokens": 36106444.0,
      "step": 2927
    },
    {
      "entropy": 1.3666974306106567,
      "epoch": 1.5418641390205372,
      "grad_norm": 0.26723483204841614,
      "learning_rate": 0.00018346070307150443,
      "loss": 0.1943853497505188,
      "mean_token_accuracy": 0.9215286374092102,
      "num_tokens": 36118780.0,
      "step": 2928
    },
    {
      "entropy": 1.379920780658722,
      "epoch": 1.5423907319641916,
      "grad_norm": 0.36375778913497925,
      "learning_rate": 0.00018344803232656805,
      "loss": 0.2259494662284851,
      "mean_token_accuracy": 0.9057151675224304,
      "num_tokens": 36131116.0,
      "step": 2929
    },
    {
      "entropy": 1.3832117319107056,
      "epoch": 1.5429173249078463,
      "grad_norm": 0.30295872688293457,
      "learning_rate": 0.00018343535722145011,
      "loss": 0.22572673857212067,
      "mean_token_accuracy": 0.9093717187643051,
      "num_tokens": 36143452.0,
      "step": 2930
    },
    {
      "entropy": 1.3965988755226135,
      "epoch": 1.543443917851501,
      "grad_norm": 0.28036290407180786,
      "learning_rate": 0.00018342267775690302,
      "loss": 0.1899104118347168,
      "mean_token_accuracy": 0.9211853593587875,
      "num_tokens": 36155788.0,
      "step": 2931
    },
    {
      "entropy": 1.3536584973335266,
      "epoch": 1.5439705107951553,
      "grad_norm": 0.28164905309677124,
      "learning_rate": 0.00018340999393367952,
      "loss": 0.20792821049690247,
      "mean_token_accuracy": 0.9128543734550476,
      "num_tokens": 36168124.0,
      "step": 2932
    },
    {
      "entropy": 1.3484612703323364,
      "epoch": 1.5444971037388098,
      "grad_norm": 0.2770152688026428,
      "learning_rate": 0.00018339730575253252,
      "loss": 0.1988334357738495,
      "mean_token_accuracy": 0.9181084334850311,
      "num_tokens": 36180460.0,
      "step": 2933
    },
    {
      "entropy": 1.3058361113071442,
      "epoch": 1.5450236966824644,
      "grad_norm": 0.2750353515148163,
      "learning_rate": 0.0001833846132142153,
      "loss": 0.1986725926399231,
      "mean_token_accuracy": 0.9165878146886826,
      "num_tokens": 36192796.0,
      "step": 2934
    },
    {
      "entropy": 1.3110437989234924,
      "epoch": 1.545550289626119,
      "grad_norm": 0.28275609016418457,
      "learning_rate": 0.00018337191631948127,
      "loss": 0.2109573483467102,
      "mean_token_accuracy": 0.9166303426027298,
      "num_tokens": 36205132.0,
      "step": 2935
    },
    {
      "entropy": 1.3241609632968903,
      "epoch": 1.5460768825697735,
      "grad_norm": 0.3148845136165619,
      "learning_rate": 0.00018335921506908425,
      "loss": 0.23299464583396912,
      "mean_token_accuracy": 0.9048751890659332,
      "num_tokens": 36217468.0,
      "step": 2936
    },
    {
      "entropy": 1.2985778152942657,
      "epoch": 1.5466034755134281,
      "grad_norm": 0.2876144051551819,
      "learning_rate": 0.0001833465094637782,
      "loss": 0.19456106424331665,
      "mean_token_accuracy": 0.9193484932184219,
      "num_tokens": 36229804.0,
      "step": 2937
    },
    {
      "entropy": 1.2285209596157074,
      "epoch": 1.5471300684570828,
      "grad_norm": 0.2816034257411957,
      "learning_rate": 0.00018333379950431734,
      "loss": 0.2107038050889969,
      "mean_token_accuracy": 0.9162741601467133,
      "num_tokens": 36242140.0,
      "step": 2938
    },
    {
      "entropy": 1.270815670490265,
      "epoch": 1.5476566614007372,
      "grad_norm": 0.30017396807670593,
      "learning_rate": 0.00018332108519145623,
      "loss": 0.22914569079875946,
      "mean_token_accuracy": 0.9055030792951584,
      "num_tokens": 36254476.0,
      "step": 2939
    },
    {
      "entropy": 1.251087248325348,
      "epoch": 1.5481832543443916,
      "grad_norm": 0.2832695543766022,
      "learning_rate": 0.00018330836652594967,
      "loss": 0.19750654697418213,
      "mean_token_accuracy": 0.9202122986316681,
      "num_tokens": 36266812.0,
      "step": 2940
    },
    {
      "entropy": 1.2241024374961853,
      "epoch": 1.5487098472880465,
      "grad_norm": 0.2506945729255676,
      "learning_rate": 0.00018329564350855264,
      "loss": 0.19707068800926208,
      "mean_token_accuracy": 0.9187771528959274,
      "num_tokens": 36279148.0,
      "step": 2941
    },
    {
      "entropy": 1.2423125505447388,
      "epoch": 1.549236440231701,
      "grad_norm": 0.2908343970775604,
      "learning_rate": 0.00018328291614002048,
      "loss": 0.2287912517786026,
      "mean_token_accuracy": 0.9047258198261261,
      "num_tokens": 36291484.0,
      "step": 2942
    },
    {
      "entropy": 1.2304923236370087,
      "epoch": 1.5497630331753554,
      "grad_norm": 0.2795937955379486,
      "learning_rate": 0.0001832701844211087,
      "loss": 0.18848249316215515,
      "mean_token_accuracy": 0.9215769469738007,
      "num_tokens": 36303820.0,
      "step": 2943
    },
    {
      "entropy": 1.236438661813736,
      "epoch": 1.55028962611901,
      "grad_norm": 0.26903602480888367,
      "learning_rate": 0.0001832574483525731,
      "loss": 0.20066887140274048,
      "mean_token_accuracy": 0.9172261506319046,
      "num_tokens": 36316156.0,
      "step": 2944
    },
    {
      "entropy": 1.2736486494541168,
      "epoch": 1.5508162190626646,
      "grad_norm": 0.29723483324050903,
      "learning_rate": 0.0001832447079351698,
      "loss": 0.19067001342773438,
      "mean_token_accuracy": 0.9212439507246017,
      "num_tokens": 36328492.0,
      "step": 2945
    },
    {
      "entropy": 1.2396992444992065,
      "epoch": 1.551342812006319,
      "grad_norm": 0.29464191198349,
      "learning_rate": 0.0001832319631696551,
      "loss": 0.23630069196224213,
      "mean_token_accuracy": 0.9037884026765823,
      "num_tokens": 36340828.0,
      "step": 2946
    },
    {
      "entropy": 1.1540693938732147,
      "epoch": 1.5518694049499737,
      "grad_norm": 0.2827843129634857,
      "learning_rate": 0.00018321921405678554,
      "loss": 0.19755500555038452,
      "mean_token_accuracy": 0.9163824766874313,
      "num_tokens": 36353164.0,
      "step": 2947
    },
    {
      "entropy": 1.2334130108356476,
      "epoch": 1.5523959978936284,
      "grad_norm": 0.2816913425922394,
      "learning_rate": 0.000183206460597318,
      "loss": 0.20551520586013794,
      "mean_token_accuracy": 0.9176173806190491,
      "num_tokens": 36365500.0,
      "step": 2948
    },
    {
      "entropy": 1.1994071006774902,
      "epoch": 1.5529225908372828,
      "grad_norm": 0.278768390417099,
      "learning_rate": 0.0001831937027920096,
      "loss": 0.19148175418376923,
      "mean_token_accuracy": 0.9233074188232422,
      "num_tokens": 36377836.0,
      "step": 2949
    },
    {
      "entropy": 1.1344974339008331,
      "epoch": 1.5534491837809372,
      "grad_norm": 0.28436407446861267,
      "learning_rate": 0.00018318094064161765,
      "loss": 0.20278222858905792,
      "mean_token_accuracy": 0.9229292422533035,
      "num_tokens": 36390172.0,
      "step": 2950
    },
    {
      "entropy": 1.1987184286117554,
      "epoch": 1.5539757767245919,
      "grad_norm": 0.29216986894607544,
      "learning_rate": 0.0001831681741468998,
      "loss": 0.21907229721546173,
      "mean_token_accuracy": 0.9171420633792877,
      "num_tokens": 36402508.0,
      "step": 2951
    },
    {
      "entropy": 1.1500971913337708,
      "epoch": 1.5545023696682465,
      "grad_norm": 0.2810475826263428,
      "learning_rate": 0.00018315540330861386,
      "loss": 0.18719711899757385,
      "mean_token_accuracy": 0.9203604012727737,
      "num_tokens": 36414844.0,
      "step": 2952
    },
    {
      "entropy": 1.1963881850242615,
      "epoch": 1.555028962611901,
      "grad_norm": 0.26603206992149353,
      "learning_rate": 0.0001831426281275181,
      "loss": 0.1819818615913391,
      "mean_token_accuracy": 0.9269210547208786,
      "num_tokens": 36427180.0,
      "step": 2953
    },
    {
      "entropy": 1.1963586807250977,
      "epoch": 1.5555555555555556,
      "grad_norm": 0.26459643244743347,
      "learning_rate": 0.00018312984860437072,
      "loss": 0.1856173276901245,
      "mean_token_accuracy": 0.9259962737560272,
      "num_tokens": 36439516.0,
      "step": 2954
    },
    {
      "entropy": 1.2363464832305908,
      "epoch": 1.5560821484992102,
      "grad_norm": 0.2887548804283142,
      "learning_rate": 0.00018311706473993051,
      "loss": 0.21210099756717682,
      "mean_token_accuracy": 0.9151240438222885,
      "num_tokens": 36451852.0,
      "step": 2955
    },
    {
      "entropy": 1.2128595113754272,
      "epoch": 1.5566087414428647,
      "grad_norm": 0.25528791546821594,
      "learning_rate": 0.00018310427653495632,
      "loss": 0.18278326094150543,
      "mean_token_accuracy": 0.9265086054801941,
      "num_tokens": 36464188.0,
      "step": 2956
    },
    {
      "entropy": 1.2223176658153534,
      "epoch": 1.557135334386519,
      "grad_norm": 0.28269702196121216,
      "learning_rate": 0.0001830914839902073,
      "loss": 0.18532417714595795,
      "mean_token_accuracy": 0.9275024235248566,
      "num_tokens": 36476524.0,
      "step": 2957
    },
    {
      "entropy": 1.21820667386055,
      "epoch": 1.5576619273301737,
      "grad_norm": 0.28611740469932556,
      "learning_rate": 0.0001830786871064429,
      "loss": 0.1948256492614746,
      "mean_token_accuracy": 0.918529212474823,
      "num_tokens": 36488860.0,
      "step": 2958
    },
    {
      "entropy": 1.2262729704380035,
      "epoch": 1.5581885202738284,
      "grad_norm": 0.31858813762664795,
      "learning_rate": 0.00018306588588442278,
      "loss": 0.21474799513816833,
      "mean_token_accuracy": 0.9137346744537354,
      "num_tokens": 36501196.0,
      "step": 2959
    },
    {
      "entropy": 1.2295315265655518,
      "epoch": 1.5587151132174828,
      "grad_norm": 0.2632119357585907,
      "learning_rate": 0.00018305308032490688,
      "loss": 0.17983609437942505,
      "mean_token_accuracy": 0.9228373020887375,
      "num_tokens": 36513532.0,
      "step": 2960
    },
    {
      "entropy": 1.1716614663600922,
      "epoch": 1.5592417061611374,
      "grad_norm": 0.29031118750572205,
      "learning_rate": 0.0001830402704286554,
      "loss": 0.2138049602508545,
      "mean_token_accuracy": 0.9134559780359268,
      "num_tokens": 36525868.0,
      "step": 2961
    },
    {
      "entropy": 1.223358392715454,
      "epoch": 1.559768299104792,
      "grad_norm": 0.2957732677459717,
      "learning_rate": 0.00018302745619642874,
      "loss": 0.20625029504299164,
      "mean_token_accuracy": 0.9132367968559265,
      "num_tokens": 36538204.0,
      "step": 2962
    },
    {
      "entropy": 1.2663150429725647,
      "epoch": 1.5602948920484465,
      "grad_norm": 0.28148961067199707,
      "learning_rate": 0.00018301463762898765,
      "loss": 0.21098777651786804,
      "mean_token_accuracy": 0.9172250628471375,
      "num_tokens": 36550540.0,
      "step": 2963
    },
    {
      "entropy": 1.20645871758461,
      "epoch": 1.5608214849921012,
      "grad_norm": 0.25845208764076233,
      "learning_rate": 0.0001830018147270931,
      "loss": 0.18882045149803162,
      "mean_token_accuracy": 0.9250784367322922,
      "num_tokens": 36562876.0,
      "step": 2964
    },
    {
      "entropy": 1.2523989081382751,
      "epoch": 1.5613480779357558,
      "grad_norm": 0.2542496919631958,
      "learning_rate": 0.0001829889874915063,
      "loss": 0.19145433604717255,
      "mean_token_accuracy": 0.9239866137504578,
      "num_tokens": 36575212.0,
      "step": 2965
    },
    {
      "entropy": 1.262599766254425,
      "epoch": 1.5618746708794102,
      "grad_norm": 0.25594037771224976,
      "learning_rate": 0.00018297615592298872,
      "loss": 0.18396718800067902,
      "mean_token_accuracy": 0.9223331660032272,
      "num_tokens": 36587548.0,
      "step": 2966
    },
    {
      "entropy": 1.2905260622501373,
      "epoch": 1.5624012638230647,
      "grad_norm": 0.2649231255054474,
      "learning_rate": 0.00018296332002230208,
      "loss": 0.2021484375,
      "mean_token_accuracy": 0.9192336350679398,
      "num_tokens": 36599884.0,
      "step": 2967
    },
    {
      "entropy": 1.3087227940559387,
      "epoch": 1.5629278567667193,
      "grad_norm": 0.2674988806247711,
      "learning_rate": 0.00018295047979020843,
      "loss": 0.20633465051651,
      "mean_token_accuracy": 0.9197041392326355,
      "num_tokens": 36612220.0,
      "step": 2968
    },
    {
      "entropy": 1.2428779602050781,
      "epoch": 1.563454449710374,
      "grad_norm": 0.3095720112323761,
      "learning_rate": 0.00018293763522746995,
      "loss": 0.21707621216773987,
      "mean_token_accuracy": 0.909159705042839,
      "num_tokens": 36624556.0,
      "step": 2969
    },
    {
      "entropy": 1.2382693886756897,
      "epoch": 1.5639810426540284,
      "grad_norm": 0.2674081325531006,
      "learning_rate": 0.0001829247863348492,
      "loss": 0.19426822662353516,
      "mean_token_accuracy": 0.9186785072088242,
      "num_tokens": 36636892.0,
      "step": 2970
    },
    {
      "entropy": 1.3028753995895386,
      "epoch": 1.564507635597683,
      "grad_norm": 0.32122448086738586,
      "learning_rate": 0.0001829119331131089,
      "loss": 0.20461226999759674,
      "mean_token_accuracy": 0.9157003164291382,
      "num_tokens": 36649228.0,
      "step": 2971
    },
    {
      "entropy": 1.2865008413791656,
      "epoch": 1.5650342285413377,
      "grad_norm": 0.2696692645549774,
      "learning_rate": 0.0001828990755630121,
      "loss": 0.21100306510925293,
      "mean_token_accuracy": 0.9124954044818878,
      "num_tokens": 36661564.0,
      "step": 2972
    },
    {
      "entropy": 1.2516160607337952,
      "epoch": 1.565560821484992,
      "grad_norm": 0.2710763216018677,
      "learning_rate": 0.0001828862136853221,
      "loss": 0.1940108835697174,
      "mean_token_accuracy": 0.9198983609676361,
      "num_tokens": 36673900.0,
      "step": 2973
    },
    {
      "entropy": 1.286393016576767,
      "epoch": 1.5660874144286465,
      "grad_norm": 0.2848055958747864,
      "learning_rate": 0.00018287334748080236,
      "loss": 0.20404544472694397,
      "mean_token_accuracy": 0.9162185341119766,
      "num_tokens": 36686236.0,
      "step": 2974
    },
    {
      "entropy": 1.3426242470741272,
      "epoch": 1.5666140073723012,
      "grad_norm": 0.2882706820964813,
      "learning_rate": 0.00018286047695021678,
      "loss": 0.22008925676345825,
      "mean_token_accuracy": 0.9098223000764847,
      "num_tokens": 36698572.0,
      "step": 2975
    },
    {
      "entropy": 1.3892289400100708,
      "epoch": 1.5671406003159558,
      "grad_norm": 0.3032713532447815,
      "learning_rate": 0.0001828476020943293,
      "loss": 0.20148518681526184,
      "mean_token_accuracy": 0.9154692888259888,
      "num_tokens": 36710908.0,
      "step": 2976
    },
    {
      "entropy": 1.3957372307777405,
      "epoch": 1.5676671932596102,
      "grad_norm": 0.29055866599082947,
      "learning_rate": 0.0001828347229139043,
      "loss": 0.21428599953651428,
      "mean_token_accuracy": 0.9150131195783615,
      "num_tokens": 36723244.0,
      "step": 2977
    },
    {
      "entropy": 1.3814803957939148,
      "epoch": 1.5681937862032649,
      "grad_norm": 0.32633841037750244,
      "learning_rate": 0.00018282183940970627,
      "loss": 0.21884474158287048,
      "mean_token_accuracy": 0.9115358591079712,
      "num_tokens": 36735580.0,
      "step": 2978
    },
    {
      "entropy": 1.4007826447486877,
      "epoch": 1.5687203791469195,
      "grad_norm": 0.25024279952049255,
      "learning_rate": 0.00018280895158250012,
      "loss": 0.1829865425825119,
      "mean_token_accuracy": 0.9268663376569748,
      "num_tokens": 36747916.0,
      "step": 2979
    },
    {
      "entropy": 1.4274977445602417,
      "epoch": 1.569246972090574,
      "grad_norm": 0.3046821057796478,
      "learning_rate": 0.00018279605943305084,
      "loss": 0.2076328843832016,
      "mean_token_accuracy": 0.9161526709794998,
      "num_tokens": 36760252.0,
      "step": 2980
    },
    {
      "entropy": 1.3855017125606537,
      "epoch": 1.5697735650342284,
      "grad_norm": 0.3147059381008148,
      "learning_rate": 0.0001827831629621238,
      "loss": 0.21519812941551208,
      "mean_token_accuracy": 0.911521315574646,
      "num_tokens": 36772588.0,
      "step": 2981
    },
    {
      "entropy": 1.400354415178299,
      "epoch": 1.5703001579778832,
      "grad_norm": 0.264636367559433,
      "learning_rate": 0.00018277026217048464,
      "loss": 0.18982136249542236,
      "mean_token_accuracy": 0.9241060167551041,
      "num_tokens": 36784924.0,
      "step": 2982
    },
    {
      "entropy": 1.3857916295528412,
      "epoch": 1.5708267509215377,
      "grad_norm": 0.2669126093387604,
      "learning_rate": 0.0001827573570588991,
      "loss": 0.18247658014297485,
      "mean_token_accuracy": 0.927479475736618,
      "num_tokens": 36797260.0,
      "step": 2983
    },
    {
      "entropy": 1.3634335398674011,
      "epoch": 1.571353343865192,
      "grad_norm": 0.2893115282058716,
      "learning_rate": 0.00018274444762813333,
      "loss": 0.21217575669288635,
      "mean_token_accuracy": 0.9115361571311951,
      "num_tokens": 36809596.0,
      "step": 2984
    },
    {
      "entropy": 1.4070159196853638,
      "epoch": 1.5718799368088467,
      "grad_norm": 0.28923332691192627,
      "learning_rate": 0.0001827315338789537,
      "loss": 0.20832975208759308,
      "mean_token_accuracy": 0.9146668463945389,
      "num_tokens": 36821932.0,
      "step": 2985
    },
    {
      "entropy": 1.3532110452651978,
      "epoch": 1.5724065297525014,
      "grad_norm": 0.31392931938171387,
      "learning_rate": 0.00018271861581212686,
      "loss": 0.22782333195209503,
      "mean_token_accuracy": 0.9090021848678589,
      "num_tokens": 36834268.0,
      "step": 2986
    },
    {
      "entropy": 1.3118796646595001,
      "epoch": 1.5729331226961558,
      "grad_norm": 0.266999751329422,
      "learning_rate": 0.00018270569342841958,
      "loss": 0.18851986527442932,
      "mean_token_accuracy": 0.9194877445697784,
      "num_tokens": 36846604.0,
      "step": 2987
    },
    {
      "entropy": 1.312853455543518,
      "epoch": 1.5734597156398105,
      "grad_norm": 0.31233277916908264,
      "learning_rate": 0.00018269276672859905,
      "loss": 0.2123054414987564,
      "mean_token_accuracy": 0.913995623588562,
      "num_tokens": 36858940.0,
      "step": 2988
    },
    {
      "entropy": 1.3079098761081696,
      "epoch": 1.573986308583465,
      "grad_norm": 0.32085880637168884,
      "learning_rate": 0.0001826798357134327,
      "loss": 0.20270395278930664,
      "mean_token_accuracy": 0.9202222973108292,
      "num_tokens": 36871276.0,
      "step": 2989
    },
    {
      "entropy": 1.2647546529769897,
      "epoch": 1.5745129015271195,
      "grad_norm": 0.28355729579925537,
      "learning_rate": 0.00018266690038368805,
      "loss": 0.21255768835544586,
      "mean_token_accuracy": 0.9135896861553192,
      "num_tokens": 36883612.0,
      "step": 2990
    },
    {
      "entropy": 1.3451344072818756,
      "epoch": 1.575039494470774,
      "grad_norm": 0.2934354543685913,
      "learning_rate": 0.00018265396074013308,
      "loss": 0.1980549544095993,
      "mean_token_accuracy": 0.9198568016290665,
      "num_tokens": 36895948.0,
      "step": 2991
    },
    {
      "entropy": 1.3011061251163483,
      "epoch": 1.5755660874144286,
      "grad_norm": 0.2980000674724579,
      "learning_rate": 0.00018264101678353592,
      "loss": 0.217035174369812,
      "mean_token_accuracy": 0.9113334864377975,
      "num_tokens": 36908284.0,
      "step": 2992
    },
    {
      "entropy": 1.3149963021278381,
      "epoch": 1.5760926803580833,
      "grad_norm": 0.2599365711212158,
      "learning_rate": 0.000182628068514665,
      "loss": 0.18322350084781647,
      "mean_token_accuracy": 0.9244182854890823,
      "num_tokens": 36920620.0,
      "step": 2993
    },
    {
      "entropy": 1.2898566722869873,
      "epoch": 1.5766192733017377,
      "grad_norm": 0.2668352425098419,
      "learning_rate": 0.0001826151159342889,
      "loss": 0.21018338203430176,
      "mean_token_accuracy": 0.915318638086319,
      "num_tokens": 36932956.0,
      "step": 2994
    },
    {
      "entropy": 1.3840532004833221,
      "epoch": 1.5771458662453923,
      "grad_norm": 0.3147556781768799,
      "learning_rate": 0.00018260215904317666,
      "loss": 0.19453351199626923,
      "mean_token_accuracy": 0.9237265735864639,
      "num_tokens": 36945292.0,
      "step": 2995
    },
    {
      "entropy": 1.3175931572914124,
      "epoch": 1.577672459189047,
      "grad_norm": 0.2718995213508606,
      "learning_rate": 0.00018258919784209737,
      "loss": 0.20508694648742676,
      "mean_token_accuracy": 0.9145184606313705,
      "num_tokens": 36957628.0,
      "step": 2996
    },
    {
      "entropy": 1.270597904920578,
      "epoch": 1.5781990521327014,
      "grad_norm": 0.28836652636528015,
      "learning_rate": 0.0001825762323318205,
      "loss": 0.21483203768730164,
      "mean_token_accuracy": 0.9124789834022522,
      "num_tokens": 36969964.0,
      "step": 2997
    },
    {
      "entropy": 1.295089215040207,
      "epoch": 1.5787256450763558,
      "grad_norm": 0.2733713686466217,
      "learning_rate": 0.00018256326251311572,
      "loss": 0.20500729978084564,
      "mean_token_accuracy": 0.9173796772956848,
      "num_tokens": 36982300.0,
      "step": 2998
    },
    {
      "entropy": 1.337970644235611,
      "epoch": 1.5792522380200107,
      "grad_norm": 0.2778991162776947,
      "learning_rate": 0.00018255028838675297,
      "loss": 0.20565159618854523,
      "mean_token_accuracy": 0.9136320203542709,
      "num_tokens": 36994636.0,
      "step": 2999
    },
    {
      "entropy": 1.2989743649959564,
      "epoch": 1.5797788309636651,
      "grad_norm": 0.25889068841934204,
      "learning_rate": 0.00018253730995350246,
      "loss": 0.20373526215553284,
      "mean_token_accuracy": 0.9146820902824402,
      "num_tokens": 37006972.0,
      "step": 3000
    },
    {
      "entropy": 1.2905791997909546,
      "epoch": 1.5803054239073195,
      "grad_norm": 0.2675490081310272,
      "learning_rate": 0.00018252432721413463,
      "loss": 0.1987248659133911,
      "mean_token_accuracy": 0.9181516915559769,
      "num_tokens": 37019308.0,
      "step": 3001
    },
    {
      "entropy": 1.361492395401001,
      "epoch": 1.5808320168509742,
      "grad_norm": 0.2765367329120636,
      "learning_rate": 0.00018251134016942017,
      "loss": 0.2040756493806839,
      "mean_token_accuracy": 0.9120943993330002,
      "num_tokens": 37031644.0,
      "step": 3002
    },
    {
      "entropy": 1.315775990486145,
      "epoch": 1.5813586097946288,
      "grad_norm": 0.29613080620765686,
      "learning_rate": 0.00018249834882013013,
      "loss": 0.22368191182613373,
      "mean_token_accuracy": 0.9061230272054672,
      "num_tokens": 37043980.0,
      "step": 3003
    },
    {
      "entropy": 1.3352965414524078,
      "epoch": 1.5818852027382833,
      "grad_norm": 0.27554982900619507,
      "learning_rate": 0.0001824853531670356,
      "loss": 0.19793741405010223,
      "mean_token_accuracy": 0.9158347100019455,
      "num_tokens": 37056316.0,
      "step": 3004
    },
    {
      "entropy": 1.316037505865097,
      "epoch": 1.582411795681938,
      "grad_norm": 0.27454736828804016,
      "learning_rate": 0.00018247235321090818,
      "loss": 0.18979346752166748,
      "mean_token_accuracy": 0.9240574240684509,
      "num_tokens": 37068652.0,
      "step": 3005
    },
    {
      "entropy": 1.3708411157131195,
      "epoch": 1.5829383886255926,
      "grad_norm": 0.3125503361225128,
      "learning_rate": 0.00018245934895251954,
      "loss": 0.2330584079027176,
      "mean_token_accuracy": 0.9037808179855347,
      "num_tokens": 37080988.0,
      "step": 3006
    },
    {
      "entropy": 1.3388588428497314,
      "epoch": 1.583464981569247,
      "grad_norm": 0.26093074679374695,
      "learning_rate": 0.00018244634039264168,
      "loss": 0.18843436241149902,
      "mean_token_accuracy": 0.9198068231344223,
      "num_tokens": 37093324.0,
      "step": 3007
    },
    {
      "entropy": 1.329328566789627,
      "epoch": 1.5839915745129014,
      "grad_norm": 0.263508141040802,
      "learning_rate": 0.00018243332753204683,
      "loss": 0.1939357966184616,
      "mean_token_accuracy": 0.9173231720924377,
      "num_tokens": 37105660.0,
      "step": 3008
    },
    {
      "entropy": 1.3546556532382965,
      "epoch": 1.584518167456556,
      "grad_norm": 0.30631089210510254,
      "learning_rate": 0.00018242031037150753,
      "loss": 0.21186725795269012,
      "mean_token_accuracy": 0.9202042520046234,
      "num_tokens": 37117996.0,
      "step": 3009
    },
    {
      "entropy": 1.3740338385105133,
      "epoch": 1.5850447604002107,
      "grad_norm": 0.2766059935092926,
      "learning_rate": 0.00018240728891179647,
      "loss": 0.21025089919567108,
      "mean_token_accuracy": 0.9121541678905487,
      "num_tokens": 37130332.0,
      "step": 3010
    },
    {
      "entropy": 1.3977608382701874,
      "epoch": 1.5855713533438651,
      "grad_norm": 0.2912761867046356,
      "learning_rate": 0.0001823942631536867,
      "loss": 0.19623880088329315,
      "mean_token_accuracy": 0.9209590405225754,
      "num_tokens": 37142668.0,
      "step": 3011
    },
    {
      "entropy": 1.3947394490242004,
      "epoch": 1.5860979462875198,
      "grad_norm": 0.2996085584163666,
      "learning_rate": 0.00018238123309795146,
      "loss": 0.21346694231033325,
      "mean_token_accuracy": 0.913955420255661,
      "num_tokens": 37155004.0,
      "step": 3012
    },
    {
      "entropy": 1.4347988963127136,
      "epoch": 1.5866245392311744,
      "grad_norm": 0.2614869475364685,
      "learning_rate": 0.0001823681987453643,
      "loss": 0.19650410115718842,
      "mean_token_accuracy": 0.9160020351409912,
      "num_tokens": 37167340.0,
      "step": 3013
    },
    {
      "entropy": 1.4444604516029358,
      "epoch": 1.5871511321748288,
      "grad_norm": 0.2813529074192047,
      "learning_rate": 0.00018235516009669898,
      "loss": 0.21237128973007202,
      "mean_token_accuracy": 0.9177996516227722,
      "num_tokens": 37179676.0,
      "step": 3014
    },
    {
      "entropy": 1.4176826477050781,
      "epoch": 1.5876777251184833,
      "grad_norm": 0.26034414768218994,
      "learning_rate": 0.00018234211715272952,
      "loss": 0.19177646934986115,
      "mean_token_accuracy": 0.9230294227600098,
      "num_tokens": 37192012.0,
      "step": 3015
    },
    {
      "entropy": 1.4028598964214325,
      "epoch": 1.5882043180621381,
      "grad_norm": 0.24113664031028748,
      "learning_rate": 0.00018232906991423015,
      "loss": 0.18486344814300537,
      "mean_token_accuracy": 0.9242925494909286,
      "num_tokens": 37204348.0,
      "step": 3016
    },
    {
      "entropy": 1.4833852052688599,
      "epoch": 1.5887309110057926,
      "grad_norm": 0.2903052866458893,
      "learning_rate": 0.00018231601838197551,
      "loss": 0.21897634863853455,
      "mean_token_accuracy": 0.913661926984787,
      "num_tokens": 37216684.0,
      "step": 3017
    },
    {
      "entropy": 1.4880701303482056,
      "epoch": 1.589257503949447,
      "grad_norm": 0.27285730838775635,
      "learning_rate": 0.00018230296255674039,
      "loss": 0.20681525766849518,
      "mean_token_accuracy": 0.9199357628822327,
      "num_tokens": 37229020.0,
      "step": 3018
    },
    {
      "entropy": 1.499198168516159,
      "epoch": 1.5897840968931016,
      "grad_norm": 0.26768195629119873,
      "learning_rate": 0.00018228990243929972,
      "loss": 0.1827520728111267,
      "mean_token_accuracy": 0.9268405586481094,
      "num_tokens": 37241356.0,
      "step": 3019
    },
    {
      "entropy": 1.4143783450126648,
      "epoch": 1.5903106898367563,
      "grad_norm": 0.24908189475536346,
      "learning_rate": 0.00018227683803042895,
      "loss": 0.1842612624168396,
      "mean_token_accuracy": 0.9244772493839264,
      "num_tokens": 37253692.0,
      "step": 3020
    },
    {
      "entropy": 1.4475407600402832,
      "epoch": 1.5908372827804107,
      "grad_norm": 0.2763029634952545,
      "learning_rate": 0.0001822637693309035,
      "loss": 0.20840662717819214,
      "mean_token_accuracy": 0.9138160645961761,
      "num_tokens": 37266028.0,
      "step": 3021
    },
    {
      "entropy": 1.4383432567119598,
      "epoch": 1.5913638757240653,
      "grad_norm": 0.26852625608444214,
      "learning_rate": 0.00018225069634149929,
      "loss": 0.18256303668022156,
      "mean_token_accuracy": 0.9246383905410767,
      "num_tokens": 37278364.0,
      "step": 3022
    },
    {
      "entropy": 1.4704687893390656,
      "epoch": 1.59189046866772,
      "grad_norm": 0.2806725800037384,
      "learning_rate": 0.00018223761906299233,
      "loss": 0.20903661847114563,
      "mean_token_accuracy": 0.9137071222066879,
      "num_tokens": 37290700.0,
      "step": 3023
    },
    {
      "entropy": 1.459957629442215,
      "epoch": 1.5924170616113744,
      "grad_norm": 0.2818518280982971,
      "learning_rate": 0.00018222453749615897,
      "loss": 0.19897016882896423,
      "mean_token_accuracy": 0.9180968701839447,
      "num_tokens": 37303036.0,
      "step": 3024
    },
    {
      "entropy": 1.4267137944698334,
      "epoch": 1.5929436545550288,
      "grad_norm": 0.27753016352653503,
      "learning_rate": 0.0001822114516417758,
      "loss": 0.19683846831321716,
      "mean_token_accuracy": 0.9210103005170822,
      "num_tokens": 37315372.0,
      "step": 3025
    },
    {
      "entropy": 1.4476919174194336,
      "epoch": 1.5934702474986835,
      "grad_norm": 0.31564968824386597,
      "learning_rate": 0.00018219836150061957,
      "loss": 0.2070121020078659,
      "mean_token_accuracy": 0.9174090623855591,
      "num_tokens": 37327708.0,
      "step": 3026
    },
    {
      "entropy": 1.4439688622951508,
      "epoch": 1.5939968404423381,
      "grad_norm": 0.30953073501586914,
      "learning_rate": 0.00018218526707346745,
      "loss": 0.20584158599376678,
      "mean_token_accuracy": 0.9158007055521011,
      "num_tokens": 37340044.0,
      "step": 3027
    },
    {
      "entropy": 1.3989494144916534,
      "epoch": 1.5945234333859926,
      "grad_norm": 0.2637103497982025,
      "learning_rate": 0.0001821721683610968,
      "loss": 0.1927478164434433,
      "mean_token_accuracy": 0.9223356992006302,
      "num_tokens": 37352380.0,
      "step": 3028
    },
    {
      "entropy": 1.4854789972305298,
      "epoch": 1.5950500263296472,
      "grad_norm": 0.36630088090896606,
      "learning_rate": 0.0001821590653642851,
      "loss": 0.20990176498889923,
      "mean_token_accuracy": 0.9147738516330719,
      "num_tokens": 37364716.0,
      "step": 3029
    },
    {
      "entropy": 1.4564665853977203,
      "epoch": 1.5955766192733019,
      "grad_norm": 0.26051509380340576,
      "learning_rate": 0.00018214595808381027,
      "loss": 0.18453973531723022,
      "mean_token_accuracy": 0.9241913110017776,
      "num_tokens": 37377052.0,
      "step": 3030
    },
    {
      "entropy": 1.4754319489002228,
      "epoch": 1.5961032122169563,
      "grad_norm": 0.29275190830230713,
      "learning_rate": 0.00018213284652045044,
      "loss": 0.21073368191719055,
      "mean_token_accuracy": 0.9117372781038284,
      "num_tokens": 37389388.0,
      "step": 3031
    },
    {
      "entropy": 1.4840569198131561,
      "epoch": 1.5966298051606107,
      "grad_norm": 0.3072666525840759,
      "learning_rate": 0.00018211973067498388,
      "loss": 0.21400286257266998,
      "mean_token_accuracy": 0.9100693017244339,
      "num_tokens": 37401724.0,
      "step": 3032
    },
    {
      "entropy": 1.4717089235782623,
      "epoch": 1.5971563981042654,
      "grad_norm": 0.2699028551578522,
      "learning_rate": 0.00018210661054818928,
      "loss": 0.19959458708763123,
      "mean_token_accuracy": 0.9229124337434769,
      "num_tokens": 37414060.0,
      "step": 3033
    },
    {
      "entropy": 1.4640864431858063,
      "epoch": 1.59768299104792,
      "grad_norm": 0.2825060486793518,
      "learning_rate": 0.00018209348614084552,
      "loss": 0.20268788933753967,
      "mean_token_accuracy": 0.9144919216632843,
      "num_tokens": 37426396.0,
      "step": 3034
    },
    {
      "entropy": 1.4468800127506256,
      "epoch": 1.5982095839915744,
      "grad_norm": 0.2595546841621399,
      "learning_rate": 0.00018208035745373164,
      "loss": 0.19899562001228333,
      "mean_token_accuracy": 0.9162727296352386,
      "num_tokens": 37438732.0,
      "step": 3035
    },
    {
      "entropy": 1.350676715373993,
      "epoch": 1.598736176935229,
      "grad_norm": 0.24199891090393066,
      "learning_rate": 0.00018206722448762709,
      "loss": 0.1900622546672821,
      "mean_token_accuracy": 0.9200112372636795,
      "num_tokens": 37451068.0,
      "step": 3036
    },
    {
      "entropy": 1.400856375694275,
      "epoch": 1.5992627698788837,
      "grad_norm": 0.27306118607521057,
      "learning_rate": 0.00018205408724331141,
      "loss": 0.20936831831932068,
      "mean_token_accuracy": 0.914875254034996,
      "num_tokens": 37463404.0,
      "step": 3037
    },
    {
      "entropy": 1.4073627889156342,
      "epoch": 1.5997893628225381,
      "grad_norm": 0.2765679359436035,
      "learning_rate": 0.00018204094572156458,
      "loss": 0.20035390555858612,
      "mean_token_accuracy": 0.9182957410812378,
      "num_tokens": 37475740.0,
      "step": 3038
    },
    {
      "entropy": 1.3531818389892578,
      "epoch": 1.6003159557661928,
      "grad_norm": 0.2836569845676422,
      "learning_rate": 0.00018202779992316668,
      "loss": 0.19704091548919678,
      "mean_token_accuracy": 0.9173637330532074,
      "num_tokens": 37488076.0,
      "step": 3039
    },
    {
      "entropy": 1.3923197388648987,
      "epoch": 1.6008425487098474,
      "grad_norm": 0.27982088923454285,
      "learning_rate": 0.0001820146498488981,
      "loss": 0.1999388188123703,
      "mean_token_accuracy": 0.9147376418113708,
      "num_tokens": 37500412.0,
      "step": 3040
    },
    {
      "entropy": 1.3494668304920197,
      "epoch": 1.6013691416535019,
      "grad_norm": 0.26349878311157227,
      "learning_rate": 0.00018200149549953948,
      "loss": 0.18103967607021332,
      "mean_token_accuracy": 0.9260935634374619,
      "num_tokens": 37512748.0,
      "step": 3041
    },
    {
      "entropy": 1.3204819858074188,
      "epoch": 1.6018957345971563,
      "grad_norm": 0.25867944955825806,
      "learning_rate": 0.00018198833687587174,
      "loss": 0.1823253035545349,
      "mean_token_accuracy": 0.9261195659637451,
      "num_tokens": 37525084.0,
      "step": 3042
    },
    {
      "entropy": 1.3195437788963318,
      "epoch": 1.602422327540811,
      "grad_norm": 0.2801850736141205,
      "learning_rate": 0.000181975173978676,
      "loss": 0.20542439818382263,
      "mean_token_accuracy": 0.923522487282753,
      "num_tokens": 37537420.0,
      "step": 3043
    },
    {
      "entropy": 1.2695345878601074,
      "epoch": 1.6029489204844656,
      "grad_norm": 0.2899746298789978,
      "learning_rate": 0.00018196200680873373,
      "loss": 0.20561306178569794,
      "mean_token_accuracy": 0.9245081543922424,
      "num_tokens": 37549756.0,
      "step": 3044
    },
    {
      "entropy": 1.3507100939750671,
      "epoch": 1.60347551342812,
      "grad_norm": 0.2811649739742279,
      "learning_rate": 0.00018194883536682647,
      "loss": 0.1866132616996765,
      "mean_token_accuracy": 0.9258896708488464,
      "num_tokens": 37562092.0,
      "step": 3045
    },
    {
      "entropy": 1.3393961191177368,
      "epoch": 1.6040021063717746,
      "grad_norm": 0.28140196204185486,
      "learning_rate": 0.0001819356596537363,
      "loss": 0.20368412137031555,
      "mean_token_accuracy": 0.9165063202381134,
      "num_tokens": 37574428.0,
      "step": 3046
    },
    {
      "entropy": 1.359088659286499,
      "epoch": 1.6045286993154293,
      "grad_norm": 0.27477335929870605,
      "learning_rate": 0.0001819224796702452,
      "loss": 0.19396814703941345,
      "mean_token_accuracy": 0.9202080518007278,
      "num_tokens": 37586764.0,
      "step": 3047
    },
    {
      "entropy": 1.3814587593078613,
      "epoch": 1.6050552922590837,
      "grad_norm": 0.29211121797561646,
      "learning_rate": 0.00018190929541713572,
      "loss": 0.20291253924369812,
      "mean_token_accuracy": 0.9152930378913879,
      "num_tokens": 37599100.0,
      "step": 3048
    },
    {
      "entropy": 1.3579456806182861,
      "epoch": 1.6055818852027381,
      "grad_norm": 0.28613388538360596,
      "learning_rate": 0.0001818961068951905,
      "loss": 0.20026324689388275,
      "mean_token_accuracy": 0.9189726412296295,
      "num_tokens": 37611436.0,
      "step": 3049
    },
    {
      "entropy": 1.3749539852142334,
      "epoch": 1.6061084781463928,
      "grad_norm": 0.2683996558189392,
      "learning_rate": 0.00018188291410519244,
      "loss": 0.1952074021100998,
      "mean_token_accuracy": 0.920792356133461,
      "num_tokens": 37623772.0,
      "step": 3050
    },
    {
      "entropy": 1.3418311476707458,
      "epoch": 1.6066350710900474,
      "grad_norm": 0.2496364861726761,
      "learning_rate": 0.00018186971704792473,
      "loss": 0.1910889744758606,
      "mean_token_accuracy": 0.922058030962944,
      "num_tokens": 37636108.0,
      "step": 3051
    },
    {
      "entropy": 1.4323306381702423,
      "epoch": 1.6071616640337019,
      "grad_norm": 0.29833361506462097,
      "learning_rate": 0.00018185651572417082,
      "loss": 0.2218259572982788,
      "mean_token_accuracy": 0.9126611351966858,
      "num_tokens": 37648444.0,
      "step": 3052
    },
    {
      "entropy": 1.4041076600551605,
      "epoch": 1.6076882569773565,
      "grad_norm": 0.268341600894928,
      "learning_rate": 0.00018184331013471435,
      "loss": 0.1920214593410492,
      "mean_token_accuracy": 0.921149805188179,
      "num_tokens": 37660780.0,
      "step": 3053
    },
    {
      "entropy": 1.425819456577301,
      "epoch": 1.6082148499210112,
      "grad_norm": 0.2559536099433899,
      "learning_rate": 0.00018183010028033936,
      "loss": 0.190832257270813,
      "mean_token_accuracy": 0.9199120700359344,
      "num_tokens": 37673116.0,
      "step": 3054
    },
    {
      "entropy": 1.434893399477005,
      "epoch": 1.6087414428646656,
      "grad_norm": 0.2749493420124054,
      "learning_rate": 0.0001818168861618299,
      "loss": 0.2014118880033493,
      "mean_token_accuracy": 0.919294461607933,
      "num_tokens": 37685452.0,
      "step": 3055
    },
    {
      "entropy": 1.438273936510086,
      "epoch": 1.60926803580832,
      "grad_norm": 0.275410532951355,
      "learning_rate": 0.00018180366777997053,
      "loss": 0.19788864254951477,
      "mean_token_accuracy": 0.9188237637281418,
      "num_tokens": 37697788.0,
      "step": 3056
    },
    {
      "entropy": 1.432799518108368,
      "epoch": 1.6097946287519749,
      "grad_norm": 0.2629889249801636,
      "learning_rate": 0.00018179044513554592,
      "loss": 0.19507870078086853,
      "mean_token_accuracy": 0.9229663461446762,
      "num_tokens": 37710124.0,
      "step": 3057
    },
    {
      "entropy": 1.4455545842647552,
      "epoch": 1.6103212216956293,
      "grad_norm": 0.26387691497802734,
      "learning_rate": 0.00018177721822934097,
      "loss": 0.19666314125061035,
      "mean_token_accuracy": 0.9165372550487518,
      "num_tokens": 37722460.0,
      "step": 3058
    },
    {
      "entropy": 1.4410416781902313,
      "epoch": 1.6108478146392837,
      "grad_norm": 0.2629290819168091,
      "learning_rate": 0.00018176398706214094,
      "loss": 0.19509999454021454,
      "mean_token_accuracy": 0.9173319488763809,
      "num_tokens": 37734796.0,
      "step": 3059
    },
    {
      "entropy": 1.4279845058918,
      "epoch": 1.6113744075829384,
      "grad_norm": 0.2658042311668396,
      "learning_rate": 0.0001817507516347313,
      "loss": 0.19231608510017395,
      "mean_token_accuracy": 0.9192933142185211,
      "num_tokens": 37747132.0,
      "step": 3060
    },
    {
      "entropy": 1.4642042517662048,
      "epoch": 1.611901000526593,
      "grad_norm": 0.28102269768714905,
      "learning_rate": 0.0001817375119478977,
      "loss": 0.20499512553215027,
      "mean_token_accuracy": 0.9162133932113647,
      "num_tokens": 37759468.0,
      "step": 3061
    },
    {
      "entropy": 1.4484304189682007,
      "epoch": 1.6124275934702474,
      "grad_norm": 0.27475789189338684,
      "learning_rate": 0.00018172426800242613,
      "loss": 0.20957040786743164,
      "mean_token_accuracy": 0.9155688285827637,
      "num_tokens": 37771804.0,
      "step": 3062
    },
    {
      "entropy": 1.482906699180603,
      "epoch": 1.612954186413902,
      "grad_norm": 0.29586583375930786,
      "learning_rate": 0.00018171101979910282,
      "loss": 0.22400207817554474,
      "mean_token_accuracy": 0.909784346818924,
      "num_tokens": 37784140.0,
      "step": 3063
    },
    {
      "entropy": 1.4169412851333618,
      "epoch": 1.6134807793575567,
      "grad_norm": 0.27837517857551575,
      "learning_rate": 0.00018169776733871422,
      "loss": 0.20341360569000244,
      "mean_token_accuracy": 0.9175834655761719,
      "num_tokens": 37796476.0,
      "step": 3064
    },
    {
      "entropy": 1.4276740849018097,
      "epoch": 1.6140073723012112,
      "grad_norm": 0.2935931086540222,
      "learning_rate": 0.00018168451062204705,
      "loss": 0.2186911404132843,
      "mean_token_accuracy": 0.9123528748750687,
      "num_tokens": 37808812.0,
      "step": 3065
    },
    {
      "entropy": 1.365132212638855,
      "epoch": 1.6145339652448656,
      "grad_norm": 0.28768110275268555,
      "learning_rate": 0.0001816712496498883,
      "loss": 0.200998455286026,
      "mean_token_accuracy": 0.9152180403470993,
      "num_tokens": 37821148.0,
      "step": 3066
    },
    {
      "entropy": 1.3745067417621613,
      "epoch": 1.6150605581885202,
      "grad_norm": 0.27627506852149963,
      "learning_rate": 0.0001816579844230252,
      "loss": 0.18625864386558533,
      "mean_token_accuracy": 0.924001932144165,
      "num_tokens": 37833484.0,
      "step": 3067
    },
    {
      "entropy": 1.4378848373889923,
      "epoch": 1.6155871511321749,
      "grad_norm": 0.2882876992225647,
      "learning_rate": 0.00018164471494224523,
      "loss": 0.2090378701686859,
      "mean_token_accuracy": 0.9168783128261566,
      "num_tokens": 37845820.0,
      "step": 3068
    },
    {
      "entropy": 1.3996251821517944,
      "epoch": 1.6161137440758293,
      "grad_norm": 0.2672761380672455,
      "learning_rate": 0.00018163144120833608,
      "loss": 0.1835557073354721,
      "mean_token_accuracy": 0.9194261729717255,
      "num_tokens": 37858156.0,
      "step": 3069
    },
    {
      "entropy": 1.4170182049274445,
      "epoch": 1.616640337019484,
      "grad_norm": 0.27003446221351624,
      "learning_rate": 0.0001816181632220858,
      "loss": 0.19711357355117798,
      "mean_token_accuracy": 0.9203603714704514,
      "num_tokens": 37870492.0,
      "step": 3070
    },
    {
      "entropy": 1.3691149055957794,
      "epoch": 1.6171669299631386,
      "grad_norm": 0.3474653661251068,
      "learning_rate": 0.00018160488098428257,
      "loss": 0.20167282223701477,
      "mean_token_accuracy": 0.9143191576004028,
      "num_tokens": 37882828.0,
      "step": 3071
    },
    {
      "entropy": 1.4236225485801697,
      "epoch": 1.617693522906793,
      "grad_norm": 0.2758263349533081,
      "learning_rate": 0.00018159159449571488,
      "loss": 0.17881087958812714,
      "mean_token_accuracy": 0.9302650541067123,
      "num_tokens": 37895164.0,
      "step": 3072
    },
    {
      "entropy": 1.3676467835903168,
      "epoch": 1.6182201158504474,
      "grad_norm": 0.2386929839849472,
      "learning_rate": 0.00018157830375717152,
      "loss": 0.17412015795707703,
      "mean_token_accuracy": 0.9283822029829025,
      "num_tokens": 37907500.0,
      "step": 3073
    },
    {
      "entropy": 1.3564641177654266,
      "epoch": 1.6187467087941023,
      "grad_norm": 0.27638673782348633,
      "learning_rate": 0.00018156500876944146,
      "loss": 0.19623751938343048,
      "mean_token_accuracy": 0.9200726747512817,
      "num_tokens": 37919836.0,
      "step": 3074
    },
    {
      "entropy": 1.3610458672046661,
      "epoch": 1.6192733017377567,
      "grad_norm": 0.26332005858421326,
      "learning_rate": 0.00018155170953331395,
      "loss": 0.20123399794101715,
      "mean_token_accuracy": 0.9213415533304214,
      "num_tokens": 37932172.0,
      "step": 3075
    },
    {
      "entropy": 1.3856143057346344,
      "epoch": 1.6197998946814112,
      "grad_norm": 0.271820604801178,
      "learning_rate": 0.00018153840604957845,
      "loss": 0.20177248120307922,
      "mean_token_accuracy": 0.9182474464178085,
      "num_tokens": 37944508.0,
      "step": 3076
    },
    {
      "entropy": 1.4035983979701996,
      "epoch": 1.6203264876250658,
      "grad_norm": 0.2817936837673187,
      "learning_rate": 0.00018152509831902475,
      "loss": 0.20706719160079956,
      "mean_token_accuracy": 0.9193074107170105,
      "num_tokens": 37956844.0,
      "step": 3077
    },
    {
      "entropy": 1.3983607590198517,
      "epoch": 1.6208530805687205,
      "grad_norm": 0.28577920794487,
      "learning_rate": 0.00018151178634244285,
      "loss": 0.20233751833438873,
      "mean_token_accuracy": 0.9215396642684937,
      "num_tokens": 37969180.0,
      "step": 3078
    },
    {
      "entropy": 1.3775334060192108,
      "epoch": 1.6213796735123749,
      "grad_norm": 0.27165067195892334,
      "learning_rate": 0.00018149847012062303,
      "loss": 0.20568159222602844,
      "mean_token_accuracy": 0.9150748550891876,
      "num_tokens": 37981516.0,
      "step": 3079
    },
    {
      "entropy": 1.3893254697322845,
      "epoch": 1.6219062664560295,
      "grad_norm": 0.2963791489601135,
      "learning_rate": 0.0001814851496543557,
      "loss": 0.2330780327320099,
      "mean_token_accuracy": 0.9025420695543289,
      "num_tokens": 37993852.0,
      "step": 3080
    },
    {
      "entropy": 1.413798749446869,
      "epoch": 1.6224328593996842,
      "grad_norm": 0.28095170855522156,
      "learning_rate": 0.00018147182494443175,
      "loss": 0.21211260557174683,
      "mean_token_accuracy": 0.9117957949638367,
      "num_tokens": 38006188.0,
      "step": 3081
    },
    {
      "entropy": 1.3956659436225891,
      "epoch": 1.6229594523433386,
      "grad_norm": 0.24814847111701965,
      "learning_rate": 0.00018145849599164205,
      "loss": 0.19040384888648987,
      "mean_token_accuracy": 0.9219196736812592,
      "num_tokens": 38018524.0,
      "step": 3082
    },
    {
      "entropy": 1.4209897816181183,
      "epoch": 1.623486045286993,
      "grad_norm": 0.2665543854236603,
      "learning_rate": 0.00018144516279677798,
      "loss": 0.17405079305171967,
      "mean_token_accuracy": 0.9293960332870483,
      "num_tokens": 38030860.0,
      "step": 3083
    },
    {
      "entropy": 1.4261830151081085,
      "epoch": 1.6240126382306477,
      "grad_norm": 0.25339123606681824,
      "learning_rate": 0.000181431825360631,
      "loss": 0.19218730926513672,
      "mean_token_accuracy": 0.923881396651268,
      "num_tokens": 38043196.0,
      "step": 3084
    },
    {
      "entropy": 1.4016647636890411,
      "epoch": 1.6245392311743023,
      "grad_norm": 0.2759294807910919,
      "learning_rate": 0.00018141848368399287,
      "loss": 0.2100793719291687,
      "mean_token_accuracy": 0.917973980307579,
      "num_tokens": 38055532.0,
      "step": 3085
    },
    {
      "entropy": 1.4507297277450562,
      "epoch": 1.6250658241179567,
      "grad_norm": 0.2608982026576996,
      "learning_rate": 0.00018140513776765564,
      "loss": 0.19737984240055084,
      "mean_token_accuracy": 0.9216252714395523,
      "num_tokens": 38067868.0,
      "step": 3086
    },
    {
      "entropy": 1.3792197108268738,
      "epoch": 1.6255924170616114,
      "grad_norm": 0.24353928864002228,
      "learning_rate": 0.00018139178761241155,
      "loss": 0.18090927600860596,
      "mean_token_accuracy": 0.9230251908302307,
      "num_tokens": 38080204.0,
      "step": 3087
    },
    {
      "entropy": 1.4379716217517853,
      "epoch": 1.626119010005266,
      "grad_norm": 0.2893258035182953,
      "learning_rate": 0.00018137843321905316,
      "loss": 0.1988956481218338,
      "mean_token_accuracy": 0.9219255745410919,
      "num_tokens": 38092540.0,
      "step": 3088
    },
    {
      "entropy": 1.451730102300644,
      "epoch": 1.6266456029489205,
      "grad_norm": 0.2763747572898865,
      "learning_rate": 0.00018136507458837317,
      "loss": 0.20928485691547394,
      "mean_token_accuracy": 0.9120761156082153,
      "num_tokens": 38104876.0,
      "step": 3089
    },
    {
      "entropy": 1.3144610822200775,
      "epoch": 1.627172195892575,
      "grad_norm": 0.2689838111400604,
      "learning_rate": 0.00018135171172116465,
      "loss": 0.18982107937335968,
      "mean_token_accuracy": 0.9245468080043793,
      "num_tokens": 38117212.0,
      "step": 3090
    },
    {
      "entropy": 1.4066085815429688,
      "epoch": 1.6276987888362298,
      "grad_norm": 0.30367499589920044,
      "learning_rate": 0.0001813383446182209,
      "loss": 0.22956493496894836,
      "mean_token_accuracy": 0.9036817699670792,
      "num_tokens": 38129548.0,
      "step": 3091
    },
    {
      "entropy": 1.4063359797000885,
      "epoch": 1.6282253817798842,
      "grad_norm": 0.28438472747802734,
      "learning_rate": 0.0001813249732803354,
      "loss": 0.20594382286071777,
      "mean_token_accuracy": 0.9164040237665176,
      "num_tokens": 38141884.0,
      "step": 3092
    },
    {
      "entropy": 1.3957843482494354,
      "epoch": 1.6287519747235386,
      "grad_norm": 0.2880820333957672,
      "learning_rate": 0.00018131159770830201,
      "loss": 0.21393732726573944,
      "mean_token_accuracy": 0.9140539467334747,
      "num_tokens": 38154220.0,
      "step": 3093
    },
    {
      "entropy": 1.378790944814682,
      "epoch": 1.6292785676671933,
      "grad_norm": 0.2896707057952881,
      "learning_rate": 0.00018129821790291464,
      "loss": 0.19451484084129333,
      "mean_token_accuracy": 0.9213518798351288,
      "num_tokens": 38166556.0,
      "step": 3094
    },
    {
      "entropy": 1.3447534143924713,
      "epoch": 1.629805160610848,
      "grad_norm": 0.2766105532646179,
      "learning_rate": 0.00018128483386496767,
      "loss": 0.18321573734283447,
      "mean_token_accuracy": 0.9283481389284134,
      "num_tokens": 38178892.0,
      "step": 3095
    },
    {
      "entropy": 1.3944838345050812,
      "epoch": 1.6303317535545023,
      "grad_norm": 0.2913741171360016,
      "learning_rate": 0.00018127144559525554,
      "loss": 0.21014010906219482,
      "mean_token_accuracy": 0.9120883643627167,
      "num_tokens": 38191228.0,
      "step": 3096
    },
    {
      "entropy": 1.368843674659729,
      "epoch": 1.630858346498157,
      "grad_norm": 0.29250389337539673,
      "learning_rate": 0.00018125805309457312,
      "loss": 0.18993695080280304,
      "mean_token_accuracy": 0.9213495254516602,
      "num_tokens": 38203564.0,
      "step": 3097
    },
    {
      "entropy": 1.3401963114738464,
      "epoch": 1.6313849394418116,
      "grad_norm": 0.2760806083679199,
      "learning_rate": 0.00018124465636371541,
      "loss": 0.19008806347846985,
      "mean_token_accuracy": 0.9225458651781082,
      "num_tokens": 38215900.0,
      "step": 3098
    },
    {
      "entropy": 1.2728153765201569,
      "epoch": 1.631911532385466,
      "grad_norm": 0.26979056000709534,
      "learning_rate": 0.0001812312554034777,
      "loss": 0.20473819971084595,
      "mean_token_accuracy": 0.9232187569141388,
      "num_tokens": 38228236.0,
      "step": 3099
    },
    {
      "entropy": 1.3584931790828705,
      "epoch": 1.6324381253291205,
      "grad_norm": 0.2959223687648773,
      "learning_rate": 0.00018121785021465552,
      "loss": 0.2270255982875824,
      "mean_token_accuracy": 0.905184268951416,
      "num_tokens": 38240572.0,
      "step": 3100
    },
    {
      "entropy": 1.3553913831710815,
      "epoch": 1.6329647182727751,
      "grad_norm": 0.26358816027641296,
      "learning_rate": 0.00018120444079804465,
      "loss": 0.20057646930217743,
      "mean_token_accuracy": 0.9184893518686295,
      "num_tokens": 38252908.0,
      "step": 3101
    },
    {
      "entropy": 1.3980708122253418,
      "epoch": 1.6334913112164298,
      "grad_norm": 0.2931104302406311,
      "learning_rate": 0.00018119102715444116,
      "loss": 0.21116019785404205,
      "mean_token_accuracy": 0.9135480672121048,
      "num_tokens": 38265244.0,
      "step": 3102
    },
    {
      "entropy": 1.4063577055931091,
      "epoch": 1.6340179041600842,
      "grad_norm": 0.2717793583869934,
      "learning_rate": 0.00018117760928464133,
      "loss": 0.20886561274528503,
      "mean_token_accuracy": 0.9101425409317017,
      "num_tokens": 38277580.0,
      "step": 3103
    },
    {
      "entropy": 1.4049569368362427,
      "epoch": 1.6345444971037388,
      "grad_norm": 0.2535911500453949,
      "learning_rate": 0.00018116418718944167,
      "loss": 0.1921328455209732,
      "mean_token_accuracy": 0.9207909554243088,
      "num_tokens": 38289916.0,
      "step": 3104
    },
    {
      "entropy": 1.4364557266235352,
      "epoch": 1.6350710900473935,
      "grad_norm": 0.2491796612739563,
      "learning_rate": 0.000181150760869639,
      "loss": 0.19639377295970917,
      "mean_token_accuracy": 0.9185868501663208,
      "num_tokens": 38302252.0,
      "step": 3105
    },
    {
      "entropy": 1.422313004732132,
      "epoch": 1.635597682991048,
      "grad_norm": 0.2487051635980606,
      "learning_rate": 0.00018113733032603036,
      "loss": 0.18390119075775146,
      "mean_token_accuracy": 0.9225007146596909,
      "num_tokens": 38314588.0,
      "step": 3106
    },
    {
      "entropy": 1.4768524467945099,
      "epoch": 1.6361242759347023,
      "grad_norm": 0.26604220271110535,
      "learning_rate": 0.000181123895559413,
      "loss": 0.18640431761741638,
      "mean_token_accuracy": 0.925470381975174,
      "num_tokens": 38326924.0,
      "step": 3107
    },
    {
      "entropy": 1.407800167798996,
      "epoch": 1.636650868878357,
      "grad_norm": 0.2594732642173767,
      "learning_rate": 0.00018111045657058454,
      "loss": 0.20143477618694305,
      "mean_token_accuracy": 0.9186883568763733,
      "num_tokens": 38339260.0,
      "step": 3108
    },
    {
      "entropy": 1.4646252691745758,
      "epoch": 1.6371774618220116,
      "grad_norm": 0.2773849666118622,
      "learning_rate": 0.00018109701336034272,
      "loss": 0.21977382898330688,
      "mean_token_accuracy": 0.9103269428014755,
      "num_tokens": 38351596.0,
      "step": 3109
    },
    {
      "entropy": 1.4419430792331696,
      "epoch": 1.637704054765666,
      "grad_norm": 0.2626135051250458,
      "learning_rate": 0.0001810835659294856,
      "loss": 0.19379013776779175,
      "mean_token_accuracy": 0.9196373075246811,
      "num_tokens": 38363932.0,
      "step": 3110
    },
    {
      "entropy": 1.5183590054512024,
      "epoch": 1.6382306477093207,
      "grad_norm": 0.2934761047363281,
      "learning_rate": 0.0001810701142788115,
      "loss": 0.1913427710533142,
      "mean_token_accuracy": 0.9219005703926086,
      "num_tokens": 38376268.0,
      "step": 3111
    },
    {
      "entropy": 1.4548234045505524,
      "epoch": 1.6387572406529753,
      "grad_norm": 0.2878572344779968,
      "learning_rate": 0.00018105665840911894,
      "loss": 0.19860300421714783,
      "mean_token_accuracy": 0.9160057157278061,
      "num_tokens": 38388604.0,
      "step": 3112
    },
    {
      "entropy": 1.4858933091163635,
      "epoch": 1.6392838335966298,
      "grad_norm": 0.3022295832633972,
      "learning_rate": 0.0001810431983212067,
      "loss": 0.20890842378139496,
      "mean_token_accuracy": 0.9164535999298096,
      "num_tokens": 38400940.0,
      "step": 3113
    },
    {
      "entropy": 1.4954878389835358,
      "epoch": 1.6398104265402842,
      "grad_norm": 0.2701857388019562,
      "learning_rate": 0.00018102973401587386,
      "loss": 0.19486674666404724,
      "mean_token_accuracy": 0.9214330017566681,
      "num_tokens": 38413276.0,
      "step": 3114
    },
    {
      "entropy": 1.4851151406764984,
      "epoch": 1.640337019483939,
      "grad_norm": 0.2806397080421448,
      "learning_rate": 0.0001810162654939197,
      "loss": 0.19537058472633362,
      "mean_token_accuracy": 0.9198283851146698,
      "num_tokens": 38425612.0,
      "step": 3115
    },
    {
      "entropy": 1.5154271125793457,
      "epoch": 1.6408636124275935,
      "grad_norm": 0.30362340807914734,
      "learning_rate": 0.00018100279275614378,
      "loss": 0.20024994015693665,
      "mean_token_accuracy": 0.9208974838256836,
      "num_tokens": 38437948.0,
      "step": 3116
    },
    {
      "entropy": 1.4964538216590881,
      "epoch": 1.641390205371248,
      "grad_norm": 0.27380117774009705,
      "learning_rate": 0.00018098931580334587,
      "loss": 0.19451157748699188,
      "mean_token_accuracy": 0.922481581568718,
      "num_tokens": 38450284.0,
      "step": 3117
    },
    {
      "entropy": 1.4973836839199066,
      "epoch": 1.6419167983149026,
      "grad_norm": 0.2761797606945038,
      "learning_rate": 0.00018097583463632606,
      "loss": 0.20249421894550323,
      "mean_token_accuracy": 0.9215444773435593,
      "num_tokens": 38462620.0,
      "step": 3118
    },
    {
      "entropy": 1.5529511272907257,
      "epoch": 1.6424433912585572,
      "grad_norm": 0.289413183927536,
      "learning_rate": 0.0001809623492558846,
      "loss": 0.2158457338809967,
      "mean_token_accuracy": 0.9128688126802444,
      "num_tokens": 38474956.0,
      "step": 3119
    },
    {
      "entropy": 1.5039259195327759,
      "epoch": 1.6429699842022116,
      "grad_norm": 0.2677723467350006,
      "learning_rate": 0.0001809488596628221,
      "loss": 0.2032872885465622,
      "mean_token_accuracy": 0.9197664707899094,
      "num_tokens": 38487292.0,
      "step": 3120
    },
    {
      "entropy": 1.5852990448474884,
      "epoch": 1.6434965771458663,
      "grad_norm": 0.2714881896972656,
      "learning_rate": 0.0001809353658579393,
      "loss": 0.2129548341035843,
      "mean_token_accuracy": 0.9137803167104721,
      "num_tokens": 38499628.0,
      "step": 3121
    },
    {
      "entropy": 1.5083645284175873,
      "epoch": 1.644023170089521,
      "grad_norm": 0.32005801796913147,
      "learning_rate": 0.00018092186784203729,
      "loss": 0.20612075924873352,
      "mean_token_accuracy": 0.9157820492982864,
      "num_tokens": 38511964.0,
      "step": 3122
    },
    {
      "entropy": 1.563103348016739,
      "epoch": 1.6445497630331753,
      "grad_norm": 0.2607233226299286,
      "learning_rate": 0.00018090836561591734,
      "loss": 0.19675372540950775,
      "mean_token_accuracy": 0.9205005466938019,
      "num_tokens": 38524300.0,
      "step": 3123
    },
    {
      "entropy": 1.559066355228424,
      "epoch": 1.6450763559768298,
      "grad_norm": 0.2431030124425888,
      "learning_rate": 0.000180894859180381,
      "loss": 0.19404737651348114,
      "mean_token_accuracy": 0.9214319586753845,
      "num_tokens": 38536636.0,
      "step": 3124
    },
    {
      "entropy": 1.5827949941158295,
      "epoch": 1.6456029489204844,
      "grad_norm": 0.2936856150627136,
      "learning_rate": 0.0001808813485362301,
      "loss": 0.21814733743667603,
      "mean_token_accuracy": 0.9099378287792206,
      "num_tokens": 38548972.0,
      "step": 3125
    },
    {
      "entropy": 1.578831136226654,
      "epoch": 1.646129541864139,
      "grad_norm": 0.2785643935203552,
      "learning_rate": 0.00018086783368426664,
      "loss": 0.19619745016098022,
      "mean_token_accuracy": 0.9219939261674881,
      "num_tokens": 38561308.0,
      "step": 3126
    },
    {
      "entropy": 1.5287668108940125,
      "epoch": 1.6466561348077935,
      "grad_norm": 0.2629009485244751,
      "learning_rate": 0.00018085431462529295,
      "loss": 0.19254928827285767,
      "mean_token_accuracy": 0.9238256067037582,
      "num_tokens": 38573644.0,
      "step": 3127
    },
    {
      "entropy": 1.5286193788051605,
      "epoch": 1.6471827277514481,
      "grad_norm": 0.2730830907821655,
      "learning_rate": 0.00018084079136011157,
      "loss": 0.21922871470451355,
      "mean_token_accuracy": 0.9088734835386276,
      "num_tokens": 38585980.0,
      "step": 3128
    },
    {
      "entropy": 1.5122800767421722,
      "epoch": 1.6477093206951028,
      "grad_norm": 0.2737482786178589,
      "learning_rate": 0.0001808272638895253,
      "loss": 0.20768514275550842,
      "mean_token_accuracy": 0.913787379860878,
      "num_tokens": 38598316.0,
      "step": 3129
    },
    {
      "entropy": 1.5126305520534515,
      "epoch": 1.6482359136387572,
      "grad_norm": 0.2560437321662903,
      "learning_rate": 0.00018081373221433717,
      "loss": 0.18702422082424164,
      "mean_token_accuracy": 0.9235630929470062,
      "num_tokens": 38610652.0,
      "step": 3130
    },
    {
      "entropy": 1.425947904586792,
      "epoch": 1.6487625065824116,
      "grad_norm": 0.26945164799690247,
      "learning_rate": 0.00018080019633535051,
      "loss": 0.2076614946126938,
      "mean_token_accuracy": 0.914527490735054,
      "num_tokens": 38622988.0,
      "step": 3131
    },
    {
      "entropy": 1.4670809209346771,
      "epoch": 1.6492890995260665,
      "grad_norm": 0.27858737111091614,
      "learning_rate": 0.00018078665625336887,
      "loss": 0.20628421008586884,
      "mean_token_accuracy": 0.9157915860414505,
      "num_tokens": 38635324.0,
      "step": 3132
    },
    {
      "entropy": 1.358038753271103,
      "epoch": 1.649815692469721,
      "grad_norm": 0.24805906414985657,
      "learning_rate": 0.00018077311196919597,
      "loss": 0.18904471397399902,
      "mean_token_accuracy": 0.9245041012763977,
      "num_tokens": 38647660.0,
      "step": 3133
    },
    {
      "entropy": 1.324582278728485,
      "epoch": 1.6503422854133754,
      "grad_norm": 0.2563723027706146,
      "learning_rate": 0.00018075956348363595,
      "loss": 0.1867881715297699,
      "mean_token_accuracy": 0.9226547628641129,
      "num_tokens": 38659996.0,
      "step": 3134
    },
    {
      "entropy": 1.320959597826004,
      "epoch": 1.65086887835703,
      "grad_norm": 0.2722032070159912,
      "learning_rate": 0.00018074601079749307,
      "loss": 0.182020366191864,
      "mean_token_accuracy": 0.9248357862234116,
      "num_tokens": 38672332.0,
      "step": 3135
    },
    {
      "entropy": 1.2622464001178741,
      "epoch": 1.6513954713006846,
      "grad_norm": 0.27287331223487854,
      "learning_rate": 0.00018073245391157184,
      "loss": 0.20617710053920746,
      "mean_token_accuracy": 0.9157890826463699,
      "num_tokens": 38684668.0,
      "step": 3136
    },
    {
      "entropy": 1.2935321927070618,
      "epoch": 1.651922064244339,
      "grad_norm": 0.29414066672325134,
      "learning_rate": 0.00018071889282667708,
      "loss": 0.20646268129348755,
      "mean_token_accuracy": 0.9180812388658524,
      "num_tokens": 38697004.0,
      "step": 3137
    },
    {
      "entropy": 1.2393367290496826,
      "epoch": 1.6524486571879937,
      "grad_norm": 0.27941057085990906,
      "learning_rate": 0.00018070532754361387,
      "loss": 0.19918499886989594,
      "mean_token_accuracy": 0.9213303327560425,
      "num_tokens": 38709340.0,
      "step": 3138
    },
    {
      "entropy": 1.2073590159416199,
      "epoch": 1.6529752501316484,
      "grad_norm": 0.2652994394302368,
      "learning_rate": 0.00018069175806318747,
      "loss": 0.2018781453371048,
      "mean_token_accuracy": 0.9207476526498795,
      "num_tokens": 38721676.0,
      "step": 3139
    },
    {
      "entropy": 1.2771650850772858,
      "epoch": 1.6535018430753028,
      "grad_norm": 0.3261018991470337,
      "learning_rate": 0.00018067818438620336,
      "loss": 0.23345033824443817,
      "mean_token_accuracy": 0.9006090611219406,
      "num_tokens": 38734012.0,
      "step": 3140
    },
    {
      "entropy": 1.2039735317230225,
      "epoch": 1.6540284360189572,
      "grad_norm": 0.265409916639328,
      "learning_rate": 0.0001806646065134674,
      "loss": 0.19166678190231323,
      "mean_token_accuracy": 0.9186943769454956,
      "num_tokens": 38746348.0,
      "step": 3141
    },
    {
      "entropy": 1.1508864760398865,
      "epoch": 1.6545550289626119,
      "grad_norm": 0.2634056806564331,
      "learning_rate": 0.00018065102444578566,
      "loss": 0.20034006237983704,
      "mean_token_accuracy": 0.9211664348840714,
      "num_tokens": 38758684.0,
      "step": 3142
    },
    {
      "entropy": 1.1475915908813477,
      "epoch": 1.6550816219062665,
      "grad_norm": 0.2569848895072937,
      "learning_rate": 0.00018063743818396436,
      "loss": 0.18887409567832947,
      "mean_token_accuracy": 0.9205733388662338,
      "num_tokens": 38771020.0,
      "step": 3143
    },
    {
      "entropy": 1.1361444592475891,
      "epoch": 1.655608214849921,
      "grad_norm": 0.23779945075511932,
      "learning_rate": 0.00018062384772881004,
      "loss": 0.17688196897506714,
      "mean_token_accuracy": 0.9271897524595261,
      "num_tokens": 38783356.0,
      "step": 3144
    },
    {
      "entropy": 1.2171200811862946,
      "epoch": 1.6561348077935756,
      "grad_norm": 0.2639971375465393,
      "learning_rate": 0.00018061025308112953,
      "loss": 0.19022762775421143,
      "mean_token_accuracy": 0.9262509793043137,
      "num_tokens": 38795692.0,
      "step": 3145
    },
    {
      "entropy": 1.1198914051055908,
      "epoch": 1.6566614007372302,
      "grad_norm": 0.27189239859580994,
      "learning_rate": 0.00018059665424172987,
      "loss": 0.19100147485733032,
      "mean_token_accuracy": 0.9232686161994934,
      "num_tokens": 38808028.0,
      "step": 3146
    },
    {
      "entropy": 1.1745660603046417,
      "epoch": 1.6571879936808847,
      "grad_norm": 0.26474079489707947,
      "learning_rate": 0.00018058305121141827,
      "loss": 0.187790185213089,
      "mean_token_accuracy": 0.9250635355710983,
      "num_tokens": 38820364.0,
      "step": 3147
    },
    {
      "entropy": 1.1938625872135162,
      "epoch": 1.657714586624539,
      "grad_norm": 0.26579174399375916,
      "learning_rate": 0.0001805694439910023,
      "loss": 0.19708439707756042,
      "mean_token_accuracy": 0.9244084358215332,
      "num_tokens": 38832700.0,
      "step": 3148
    },
    {
      "entropy": 1.1665718257427216,
      "epoch": 1.658241179568194,
      "grad_norm": 0.28311535716056824,
      "learning_rate": 0.0001805558325812898,
      "loss": 0.2201116979122162,
      "mean_token_accuracy": 0.9134600609540939,
      "num_tokens": 38845036.0,
      "step": 3149
    },
    {
      "entropy": 1.1456275880336761,
      "epoch": 1.6587677725118484,
      "grad_norm": 0.2647797763347626,
      "learning_rate": 0.0001805422169830887,
      "loss": 0.20335599780082703,
      "mean_token_accuracy": 0.9210527390241623,
      "num_tokens": 38857372.0,
      "step": 3150
    },
    {
      "entropy": 1.2212076783180237,
      "epoch": 1.6592943654555028,
      "grad_norm": 0.2934233844280243,
      "learning_rate": 0.00018052859719720735,
      "loss": 0.21342241764068604,
      "mean_token_accuracy": 0.9127740412950516,
      "num_tokens": 38869708.0,
      "step": 3151
    },
    {
      "entropy": 1.1577803790569305,
      "epoch": 1.6598209583991574,
      "grad_norm": 0.28830140829086304,
      "learning_rate": 0.0001805149732244543,
      "loss": 0.22134548425674438,
      "mean_token_accuracy": 0.9122626036405563,
      "num_tokens": 38882044.0,
      "step": 3152
    },
    {
      "entropy": 1.1979770064353943,
      "epoch": 1.660347551342812,
      "grad_norm": 0.29721346497535706,
      "learning_rate": 0.0001805013450656382,
      "loss": 0.2053031623363495,
      "mean_token_accuracy": 0.9165612608194351,
      "num_tokens": 38894380.0,
      "step": 3153
    },
    {
      "entropy": 1.234178513288498,
      "epoch": 1.6608741442864665,
      "grad_norm": 0.25609639286994934,
      "learning_rate": 0.00018048771272156821,
      "loss": 0.18612919747829437,
      "mean_token_accuracy": 0.9224274605512619,
      "num_tokens": 38906716.0,
      "step": 3154
    },
    {
      "entropy": 1.1776945292949677,
      "epoch": 1.6614007372301212,
      "grad_norm": 0.30163243412971497,
      "learning_rate": 0.00018047407619305354,
      "loss": 0.20453007519245148,
      "mean_token_accuracy": 0.9143376648426056,
      "num_tokens": 38919052.0,
      "step": 3155
    },
    {
      "entropy": 1.1364782750606537,
      "epoch": 1.6619273301737758,
      "grad_norm": 0.26865363121032715,
      "learning_rate": 0.00018046043548090371,
      "loss": 0.19255727529525757,
      "mean_token_accuracy": 0.9212214797735214,
      "num_tokens": 38931388.0,
      "step": 3156
    },
    {
      "entropy": 1.1973776817321777,
      "epoch": 1.6624539231174302,
      "grad_norm": 0.290986031293869,
      "learning_rate": 0.0001804467905859285,
      "loss": 0.21269451081752777,
      "mean_token_accuracy": 0.9207015186548233,
      "num_tokens": 38943724.0,
      "step": 3157
    },
    {
      "entropy": 1.2418961226940155,
      "epoch": 1.6629805160610847,
      "grad_norm": 0.2690465748310089,
      "learning_rate": 0.00018043314150893795,
      "loss": 0.18485507369041443,
      "mean_token_accuracy": 0.923929750919342,
      "num_tokens": 38956060.0,
      "step": 3158
    },
    {
      "entropy": 1.2224768996238708,
      "epoch": 1.6635071090047393,
      "grad_norm": 0.2874819040298462,
      "learning_rate": 0.0001804194882507423,
      "loss": 0.20011025667190552,
      "mean_token_accuracy": 0.9204146862030029,
      "num_tokens": 38968396.0,
      "step": 3159
    },
    {
      "entropy": 1.2215831875801086,
      "epoch": 1.664033701948394,
      "grad_norm": 0.303321897983551,
      "learning_rate": 0.00018040583081215206,
      "loss": 0.21922917664051056,
      "mean_token_accuracy": 0.9072663635015488,
      "num_tokens": 38980732.0,
      "step": 3160
    },
    {
      "entropy": 1.3148879706859589,
      "epoch": 1.6645602948920484,
      "grad_norm": 0.3454381227493286,
      "learning_rate": 0.00018039216919397798,
      "loss": 0.23111680150032043,
      "mean_token_accuracy": 0.9131344109773636,
      "num_tokens": 38993068.0,
      "step": 3161
    },
    {
      "entropy": 1.2151626348495483,
      "epoch": 1.665086887835703,
      "grad_norm": 0.2705059349536896,
      "learning_rate": 0.00018037850339703112,
      "loss": 0.2070206105709076,
      "mean_token_accuracy": 0.915043517947197,
      "num_tokens": 39005404.0,
      "step": 3162
    },
    {
      "entropy": 1.2460331618785858,
      "epoch": 1.6656134807793577,
      "grad_norm": 0.28544241189956665,
      "learning_rate": 0.00018036483342212268,
      "loss": 0.2099902629852295,
      "mean_token_accuracy": 0.9136664867401123,
      "num_tokens": 39017740.0,
      "step": 3163
    },
    {
      "entropy": 1.2310877740383148,
      "epoch": 1.666140073723012,
      "grad_norm": 0.2626836895942688,
      "learning_rate": 0.00018035115927006424,
      "loss": 0.2030152976512909,
      "mean_token_accuracy": 0.9205620735883713,
      "num_tokens": 39030076.0,
      "step": 3164
    },
    {
      "entropy": 1.3169174790382385,
      "epoch": 1.6666666666666665,
      "grad_norm": 0.2741585373878479,
      "learning_rate": 0.00018033748094166746,
      "loss": 0.19826257228851318,
      "mean_token_accuracy": 0.9195802211761475,
      "num_tokens": 39042412.0,
      "step": 3165
    },
    {
      "entropy": 1.2823103368282318,
      "epoch": 1.6671932596103214,
      "grad_norm": 0.25388139486312866,
      "learning_rate": 0.00018032379843774442,
      "loss": 0.20759499073028564,
      "mean_token_accuracy": 0.9136326611042023,
      "num_tokens": 39054748.0,
      "step": 3166
    },
    {
      "entropy": 1.2326855063438416,
      "epoch": 1.6677198525539758,
      "grad_norm": 0.2549154758453369,
      "learning_rate": 0.00018031011175910738,
      "loss": 0.20668521523475647,
      "mean_token_accuracy": 0.9171313345432281,
      "num_tokens": 39067084.0,
      "step": 3167
    },
    {
      "entropy": 1.2375471889972687,
      "epoch": 1.6682464454976302,
      "grad_norm": 0.2682606279850006,
      "learning_rate": 0.00018029642090656878,
      "loss": 0.19891971349716187,
      "mean_token_accuracy": 0.9194265604019165,
      "num_tokens": 39079420.0,
      "step": 3168
    },
    {
      "entropy": 1.2905383706092834,
      "epoch": 1.6687730384412849,
      "grad_norm": 0.2665676176548004,
      "learning_rate": 0.00018028272588094138,
      "loss": 0.19952955842018127,
      "mean_token_accuracy": 0.9163491129875183,
      "num_tokens": 39091756.0,
      "step": 3169
    },
    {
      "entropy": 1.2198565304279327,
      "epoch": 1.6692996313849395,
      "grad_norm": 0.28037944436073303,
      "learning_rate": 0.00018026902668303818,
      "loss": 0.19893498718738556,
      "mean_token_accuracy": 0.9220035374164581,
      "num_tokens": 39104092.0,
      "step": 3170
    },
    {
      "entropy": 1.1831121444702148,
      "epoch": 1.669826224328594,
      "grad_norm": 0.2583988904953003,
      "learning_rate": 0.00018025532331367243,
      "loss": 0.17476224899291992,
      "mean_token_accuracy": 0.9299489706754684,
      "num_tokens": 39116428.0,
      "step": 3171
    },
    {
      "entropy": 1.2536409497261047,
      "epoch": 1.6703528172722486,
      "grad_norm": 0.28530681133270264,
      "learning_rate": 0.0001802416157736576,
      "loss": 0.1837761104106903,
      "mean_token_accuracy": 0.9242200702428818,
      "num_tokens": 39128764.0,
      "step": 3172
    },
    {
      "entropy": 1.2332830131053925,
      "epoch": 1.6708794102159032,
      "grad_norm": 0.29961588978767395,
      "learning_rate": 0.00018022790406380747,
      "loss": 0.2195456326007843,
      "mean_token_accuracy": 0.9104225635528564,
      "num_tokens": 39141100.0,
      "step": 3173
    },
    {
      "entropy": 1.1516872346401215,
      "epoch": 1.6714060031595577,
      "grad_norm": 0.2845155596733093,
      "learning_rate": 0.00018021418818493597,
      "loss": 0.20701955258846283,
      "mean_token_accuracy": 0.9150958508253098,
      "num_tokens": 39153436.0,
      "step": 3174
    },
    {
      "entropy": 1.2093952596187592,
      "epoch": 1.671932596103212,
      "grad_norm": 0.29277166724205017,
      "learning_rate": 0.00018020046813785735,
      "loss": 0.1868697553873062,
      "mean_token_accuracy": 0.9202113300561905,
      "num_tokens": 39165772.0,
      "step": 3175
    },
    {
      "entropy": 1.1511423885822296,
      "epoch": 1.6724591890468667,
      "grad_norm": 0.282686710357666,
      "learning_rate": 0.00018018674392338612,
      "loss": 0.18359015882015228,
      "mean_token_accuracy": 0.9301719814538956,
      "num_tokens": 39178108.0,
      "step": 3176
    },
    {
      "entropy": 1.158335953950882,
      "epoch": 1.6729857819905214,
      "grad_norm": 0.2753232419490814,
      "learning_rate": 0.00018017301554233697,
      "loss": 0.2069041132926941,
      "mean_token_accuracy": 0.9147633165121078,
      "num_tokens": 39190444.0,
      "step": 3177
    },
    {
      "entropy": 1.194052815437317,
      "epoch": 1.6735123749341758,
      "grad_norm": 0.2822655141353607,
      "learning_rate": 0.0001801592829955249,
      "loss": 0.1951710730791092,
      "mean_token_accuracy": 0.9215719550848007,
      "num_tokens": 39202780.0,
      "step": 3178
    },
    {
      "entropy": 1.151505172252655,
      "epoch": 1.6740389678778305,
      "grad_norm": 0.28857794404029846,
      "learning_rate": 0.00018014554628376512,
      "loss": 0.20368535816669464,
      "mean_token_accuracy": 0.9183078706264496,
      "num_tokens": 39215116.0,
      "step": 3179
    },
    {
      "entropy": 1.2312898933887482,
      "epoch": 1.674565560821485,
      "grad_norm": 0.306927353143692,
      "learning_rate": 0.00018013180540787309,
      "loss": 0.2206253558397293,
      "mean_token_accuracy": 0.9108968675136566,
      "num_tokens": 39227452.0,
      "step": 3180
    },
    {
      "entropy": 1.2053121328353882,
      "epoch": 1.6750921537651395,
      "grad_norm": 0.27955883741378784,
      "learning_rate": 0.00018011806036866452,
      "loss": 0.19037887454032898,
      "mean_token_accuracy": 0.917056143283844,
      "num_tokens": 39239788.0,
      "step": 3181
    },
    {
      "entropy": 1.1672830879688263,
      "epoch": 1.675618746708794,
      "grad_norm": 0.28999653458595276,
      "learning_rate": 0.0001801043111669554,
      "loss": 0.19952133297920227,
      "mean_token_accuracy": 0.917973592877388,
      "num_tokens": 39252124.0,
      "step": 3182
    },
    {
      "entropy": 1.2045989036560059,
      "epoch": 1.6761453396524486,
      "grad_norm": 0.28392067551612854,
      "learning_rate": 0.0001800905578035619,
      "loss": 0.2129674255847931,
      "mean_token_accuracy": 0.9127481281757355,
      "num_tokens": 39264460.0,
      "step": 3183
    },
    {
      "entropy": 1.19984170794487,
      "epoch": 1.6766719325961033,
      "grad_norm": 0.25406205654144287,
      "learning_rate": 0.00018007680027930053,
      "loss": 0.19678980112075806,
      "mean_token_accuracy": 0.9213206321001053,
      "num_tokens": 39276796.0,
      "step": 3184
    },
    {
      "entropy": 1.1855994760990143,
      "epoch": 1.6771985255397577,
      "grad_norm": 0.2667410373687744,
      "learning_rate": 0.00018006303859498794,
      "loss": 0.19079680740833282,
      "mean_token_accuracy": 0.9214341640472412,
      "num_tokens": 39289132.0,
      "step": 3185
    },
    {
      "entropy": 1.2297048270702362,
      "epoch": 1.6777251184834123,
      "grad_norm": 0.2681465446949005,
      "learning_rate": 0.0001800492727514411,
      "loss": 0.209198996424675,
      "mean_token_accuracy": 0.9140588194131851,
      "num_tokens": 39301468.0,
      "step": 3186
    },
    {
      "entropy": 1.2474170923233032,
      "epoch": 1.678251711427067,
      "grad_norm": 0.29584237933158875,
      "learning_rate": 0.00018003550274947727,
      "loss": 0.2179972529411316,
      "mean_token_accuracy": 0.910770982503891,
      "num_tokens": 39313804.0,
      "step": 3187
    },
    {
      "entropy": 1.2513139247894287,
      "epoch": 1.6787783043707214,
      "grad_norm": 0.26859045028686523,
      "learning_rate": 0.00018002172858991381,
      "loss": 0.19416972994804382,
      "mean_token_accuracy": 0.9244131296873093,
      "num_tokens": 39326140.0,
      "step": 3188
    },
    {
      "entropy": 1.2674452066421509,
      "epoch": 1.6793048973143758,
      "grad_norm": 0.2633540630340576,
      "learning_rate": 0.00018000795027356845,
      "loss": 0.206901416182518,
      "mean_token_accuracy": 0.9144023954868317,
      "num_tokens": 39338476.0,
      "step": 3189
    },
    {
      "entropy": 1.2862504124641418,
      "epoch": 1.6798314902580307,
      "grad_norm": 0.2601005434989929,
      "learning_rate": 0.00017999416780125908,
      "loss": 0.1940033733844757,
      "mean_token_accuracy": 0.9206139743328094,
      "num_tokens": 39350812.0,
      "step": 3190
    },
    {
      "entropy": 1.296876072883606,
      "epoch": 1.6803580832016851,
      "grad_norm": 0.2819749116897583,
      "learning_rate": 0.00017998038117380397,
      "loss": 0.20579656958580017,
      "mean_token_accuracy": 0.914523109793663,
      "num_tokens": 39363148.0,
      "step": 3191
    },
    {
      "entropy": 1.3008518517017365,
      "epoch": 1.6808846761453395,
      "grad_norm": 0.2690666913986206,
      "learning_rate": 0.00017996659039202148,
      "loss": 0.18763180077075958,
      "mean_token_accuracy": 0.9187602549791336,
      "num_tokens": 39375484.0,
      "step": 3192
    },
    {
      "entropy": 1.2868019342422485,
      "epoch": 1.6814112690889942,
      "grad_norm": 0.2677951753139496,
      "learning_rate": 0.00017995279545673034,
      "loss": 0.19632895290851593,
      "mean_token_accuracy": 0.9192786663770676,
      "num_tokens": 39387820.0,
      "step": 3193
    },
    {
      "entropy": 1.268946647644043,
      "epoch": 1.6819378620326488,
      "grad_norm": 0.2508223354816437,
      "learning_rate": 0.00017993899636874943,
      "loss": 0.18477728962898254,
      "mean_token_accuracy": 0.9243815243244171,
      "num_tokens": 39400156.0,
      "step": 3194
    },
    {
      "entropy": 1.322548896074295,
      "epoch": 1.6824644549763033,
      "grad_norm": 0.28498703241348267,
      "learning_rate": 0.00017992519312889796,
      "loss": 0.21436122059822083,
      "mean_token_accuracy": 0.9100509434938431,
      "num_tokens": 39412492.0,
      "step": 3195
    },
    {
      "entropy": 1.3395605385303497,
      "epoch": 1.682991047919958,
      "grad_norm": 0.26486754417419434,
      "learning_rate": 0.0001799113857379953,
      "loss": 0.1872069537639618,
      "mean_token_accuracy": 0.9222505837678909,
      "num_tokens": 39424828.0,
      "step": 3196
    },
    {
      "entropy": 1.308861881494522,
      "epoch": 1.6835176408636126,
      "grad_norm": 0.27471840381622314,
      "learning_rate": 0.00017989757419686116,
      "loss": 0.18717561662197113,
      "mean_token_accuracy": 0.9235621690750122,
      "num_tokens": 39437164.0,
      "step": 3197
    },
    {
      "entropy": 1.2929215729236603,
      "epoch": 1.684044233807267,
      "grad_norm": 0.2788255214691162,
      "learning_rate": 0.00017988375850631541,
      "loss": 0.20904980599880219,
      "mean_token_accuracy": 0.9153354316949844,
      "num_tokens": 39449500.0,
      "step": 3198
    },
    {
      "entropy": 1.2841696441173553,
      "epoch": 1.6845708267509214,
      "grad_norm": 0.30049991607666016,
      "learning_rate": 0.00017986993866717828,
      "loss": 0.2012903392314911,
      "mean_token_accuracy": 0.9142125397920609,
      "num_tokens": 39461836.0,
      "step": 3199
    },
    {
      "entropy": 1.261197030544281,
      "epoch": 1.685097419694576,
      "grad_norm": 0.2596745491027832,
      "learning_rate": 0.0001798561146802701,
      "loss": 0.1863638311624527,
      "mean_token_accuracy": 0.925238847732544,
      "num_tokens": 39474172.0,
      "step": 3200
    },
    {
      "entropy": 1.3221564590930939,
      "epoch": 1.6856240126382307,
      "grad_norm": 0.28279244899749756,
      "learning_rate": 0.00017984228654641153,
      "loss": 0.21060873568058014,
      "mean_token_accuracy": 0.9115581065416336,
      "num_tokens": 39486508.0,
      "step": 3201
    },
    {
      "entropy": 1.2866823077201843,
      "epoch": 1.6861506055818851,
      "grad_norm": 0.24817955493927002,
      "learning_rate": 0.00017982845426642348,
      "loss": 0.18661949038505554,
      "mean_token_accuracy": 0.9202256947755814,
      "num_tokens": 39498844.0,
      "step": 3202
    },
    {
      "entropy": 1.320129543542862,
      "epoch": 1.6866771985255398,
      "grad_norm": 0.2350441813468933,
      "learning_rate": 0.0001798146178411271,
      "loss": 0.17725782096385956,
      "mean_token_accuracy": 0.9308050870895386,
      "num_tokens": 39511180.0,
      "step": 3203
    },
    {
      "entropy": 1.410641998052597,
      "epoch": 1.6872037914691944,
      "grad_norm": 0.2893678545951843,
      "learning_rate": 0.00017980077727134376,
      "loss": 0.19813215732574463,
      "mean_token_accuracy": 0.9261365532875061,
      "num_tokens": 39523516.0,
      "step": 3204
    },
    {
      "entropy": 1.2950620353221893,
      "epoch": 1.6877303844128488,
      "grad_norm": 0.2544447183609009,
      "learning_rate": 0.0001797869325578951,
      "loss": 0.1995326578617096,
      "mean_token_accuracy": 0.9175571799278259,
      "num_tokens": 39535852.0,
      "step": 3205
    },
    {
      "entropy": 1.3695853650569916,
      "epoch": 1.6882569773565033,
      "grad_norm": 0.26028668880462646,
      "learning_rate": 0.00017977308370160304,
      "loss": 0.1951325535774231,
      "mean_token_accuracy": 0.9233993589878082,
      "num_tokens": 39548188.0,
      "step": 3206
    },
    {
      "entropy": 1.3739449977874756,
      "epoch": 1.6887835703001581,
      "grad_norm": 0.24822396039962769,
      "learning_rate": 0.00017975923070328965,
      "loss": 0.1763211488723755,
      "mean_token_accuracy": 0.9289481788873672,
      "num_tokens": 39560524.0,
      "step": 3207
    },
    {
      "entropy": 1.303276151418686,
      "epoch": 1.6893101632438126,
      "grad_norm": 0.234592005610466,
      "learning_rate": 0.00017974537356377733,
      "loss": 0.1826443076133728,
      "mean_token_accuracy": 0.925462156534195,
      "num_tokens": 39572860.0,
      "step": 3208
    },
    {
      "entropy": 1.353267639875412,
      "epoch": 1.689836756187467,
      "grad_norm": 0.25751370191574097,
      "learning_rate": 0.00017973151228388865,
      "loss": 0.18108654022216797,
      "mean_token_accuracy": 0.9255900233983994,
      "num_tokens": 39585196.0,
      "step": 3209
    },
    {
      "entropy": 1.3624829649925232,
      "epoch": 1.6903633491311216,
      "grad_norm": 0.27678364515304565,
      "learning_rate": 0.00017971764686444653,
      "loss": 0.20060954988002777,
      "mean_token_accuracy": 0.9197713881731033,
      "num_tokens": 39597532.0,
      "step": 3210
    },
    {
      "entropy": 1.3278814852237701,
      "epoch": 1.6908899420747763,
      "grad_norm": 0.26017773151397705,
      "learning_rate": 0.00017970377730627409,
      "loss": 0.1876058727502823,
      "mean_token_accuracy": 0.9261418581008911,
      "num_tokens": 39609868.0,
      "step": 3211
    },
    {
      "entropy": 1.335858404636383,
      "epoch": 1.6914165350184307,
      "grad_norm": 0.267092227935791,
      "learning_rate": 0.0001796899036101946,
      "loss": 0.20062491297721863,
      "mean_token_accuracy": 0.916288897395134,
      "num_tokens": 39622204.0,
      "step": 3212
    },
    {
      "entropy": 1.3814048171043396,
      "epoch": 1.6919431279620853,
      "grad_norm": 0.264210969209671,
      "learning_rate": 0.00017967602577703175,
      "loss": 0.19217954576015472,
      "mean_token_accuracy": 0.9183042198419571,
      "num_tokens": 39634540.0,
      "step": 3213
    },
    {
      "entropy": 1.3550595045089722,
      "epoch": 1.69246972090574,
      "grad_norm": 0.28114888072013855,
      "learning_rate": 0.00017966214380760938,
      "loss": 0.20994627475738525,
      "mean_token_accuracy": 0.9140847474336624,
      "num_tokens": 39646876.0,
      "step": 3214
    },
    {
      "entropy": 1.2746717631816864,
      "epoch": 1.6929963138493944,
      "grad_norm": 0.26488783955574036,
      "learning_rate": 0.0001796482577027515,
      "loss": 0.20270992815494537,
      "mean_token_accuracy": 0.9152298867702484,
      "num_tokens": 39659212.0,
      "step": 3215
    },
    {
      "entropy": 1.3254249095916748,
      "epoch": 1.6935229067930488,
      "grad_norm": 0.2747063636779785,
      "learning_rate": 0.00017963436746328253,
      "loss": 0.20579099655151367,
      "mean_token_accuracy": 0.9173053056001663,
      "num_tokens": 39671548.0,
      "step": 3216
    },
    {
      "entropy": 1.3438137471675873,
      "epoch": 1.6940494997367035,
      "grad_norm": 0.2942642569541931,
      "learning_rate": 0.00017962047309002702,
      "loss": 0.19839614629745483,
      "mean_token_accuracy": 0.9128136783838272,
      "num_tokens": 39683884.0,
      "step": 3217
    },
    {
      "entropy": 1.2776517271995544,
      "epoch": 1.6945760926803581,
      "grad_norm": 0.2709120213985443,
      "learning_rate": 0.0001796065745838098,
      "loss": 0.191197469830513,
      "mean_token_accuracy": 0.9194577038288116,
      "num_tokens": 39696220.0,
      "step": 3218
    },
    {
      "entropy": 1.338829666376114,
      "epoch": 1.6951026856240126,
      "grad_norm": 0.29193103313446045,
      "learning_rate": 0.00017959267194545597,
      "loss": 0.21261580288410187,
      "mean_token_accuracy": 0.9161190837621689,
      "num_tokens": 39708556.0,
      "step": 3219
    },
    {
      "entropy": 1.3053942620754242,
      "epoch": 1.6956292785676672,
      "grad_norm": 0.2555198669433594,
      "learning_rate": 0.00017957876517579076,
      "loss": 0.1975700557231903,
      "mean_token_accuracy": 0.9193883389234543,
      "num_tokens": 39720892.0,
      "step": 3220
    },
    {
      "entropy": 1.3086311519145966,
      "epoch": 1.6961558715113219,
      "grad_norm": 0.2704666554927826,
      "learning_rate": 0.00017956485427563984,
      "loss": 0.19425427913665771,
      "mean_token_accuracy": 0.9231333583593369,
      "num_tokens": 39733228.0,
      "step": 3221
    },
    {
      "entropy": 1.3097478151321411,
      "epoch": 1.6966824644549763,
      "grad_norm": 0.2648825943470001,
      "learning_rate": 0.000179550939245829,
      "loss": 0.1881440281867981,
      "mean_token_accuracy": 0.9222024977207184,
      "num_tokens": 39745564.0,
      "step": 3222
    },
    {
      "entropy": 1.341114103794098,
      "epoch": 1.6972090573986307,
      "grad_norm": 0.29292595386505127,
      "learning_rate": 0.00017953702008718426,
      "loss": 0.20591764152050018,
      "mean_token_accuracy": 0.9151119291782379,
      "num_tokens": 39757900.0,
      "step": 3223
    },
    {
      "entropy": 1.3235185742378235,
      "epoch": 1.6977356503422856,
      "grad_norm": 0.2651384472846985,
      "learning_rate": 0.0001795230968005319,
      "loss": 0.183737114071846,
      "mean_token_accuracy": 0.9256300330162048,
      "num_tokens": 39770236.0,
      "step": 3224
    },
    {
      "entropy": 1.3323991000652313,
      "epoch": 1.69826224328594,
      "grad_norm": 0.2805165946483612,
      "learning_rate": 0.0001795091693866985,
      "loss": 0.21993422508239746,
      "mean_token_accuracy": 0.9102575182914734,
      "num_tokens": 39782572.0,
      "step": 3225
    },
    {
      "entropy": 1.33993661403656,
      "epoch": 1.6987888362295944,
      "grad_norm": 0.2455688714981079,
      "learning_rate": 0.00017949523784651085,
      "loss": 0.1799536943435669,
      "mean_token_accuracy": 0.9266008138656616,
      "num_tokens": 39794908.0,
      "step": 3226
    },
    {
      "entropy": 1.3271623253822327,
      "epoch": 1.699315429173249,
      "grad_norm": 0.24277228116989136,
      "learning_rate": 0.00017948130218079598,
      "loss": 0.1912137269973755,
      "mean_token_accuracy": 0.9215309470891953,
      "num_tokens": 39807244.0,
      "step": 3227
    },
    {
      "entropy": 1.3194293975830078,
      "epoch": 1.6998420221169037,
      "grad_norm": 0.2657921612262726,
      "learning_rate": 0.00017946736239038118,
      "loss": 0.20347270369529724,
      "mean_token_accuracy": 0.9160242527723312,
      "num_tokens": 39819580.0,
      "step": 3228
    },
    {
      "entropy": 1.283385008573532,
      "epoch": 1.7003686150605581,
      "grad_norm": 0.2746362090110779,
      "learning_rate": 0.00017945341847609396,
      "loss": 0.19498057663440704,
      "mean_token_accuracy": 0.9214595407247543,
      "num_tokens": 39831916.0,
      "step": 3229
    },
    {
      "entropy": 1.3247983157634735,
      "epoch": 1.7008952080042128,
      "grad_norm": 0.2631048262119293,
      "learning_rate": 0.00017943947043876207,
      "loss": 0.19207647442817688,
      "mean_token_accuracy": 0.9226651191711426,
      "num_tokens": 39844252.0,
      "step": 3230
    },
    {
      "entropy": 1.3293734192848206,
      "epoch": 1.7014218009478674,
      "grad_norm": 0.28094783425331116,
      "learning_rate": 0.00017942551827921356,
      "loss": 0.2019774615764618,
      "mean_token_accuracy": 0.9162713438272476,
      "num_tokens": 39856588.0,
      "step": 3231
    },
    {
      "entropy": 1.338496744632721,
      "epoch": 1.7019483938915219,
      "grad_norm": 0.2915278375148773,
      "learning_rate": 0.00017941156199827664,
      "loss": 0.21198813617229462,
      "mean_token_accuracy": 0.9130686819553375,
      "num_tokens": 39868924.0,
      "step": 3232
    },
    {
      "entropy": 1.3526046872138977,
      "epoch": 1.7024749868351763,
      "grad_norm": 0.2703212797641754,
      "learning_rate": 0.00017939760159677988,
      "loss": 0.18862299621105194,
      "mean_token_accuracy": 0.9240467101335526,
      "num_tokens": 39881260.0,
      "step": 3233
    },
    {
      "entropy": 1.336966633796692,
      "epoch": 1.703001579778831,
      "grad_norm": 0.2505403161048889,
      "learning_rate": 0.00017938363707555198,
      "loss": 0.1721397340297699,
      "mean_token_accuracy": 0.928515300154686,
      "num_tokens": 39893596.0,
      "step": 3234
    },
    {
      "entropy": 1.2895264327526093,
      "epoch": 1.7035281727224856,
      "grad_norm": 0.27174112200737,
      "learning_rate": 0.00017936966843542195,
      "loss": 0.18495216965675354,
      "mean_token_accuracy": 0.9250223636627197,
      "num_tokens": 39905932.0,
      "step": 3235
    },
    {
      "entropy": 1.294345647096634,
      "epoch": 1.70405476566614,
      "grad_norm": 0.31108003854751587,
      "learning_rate": 0.000179355695677219,
      "loss": 0.22918599843978882,
      "mean_token_accuracy": 0.9110808223485947,
      "num_tokens": 39918268.0,
      "step": 3236
    },
    {
      "entropy": 1.2462213635444641,
      "epoch": 1.7045813586097947,
      "grad_norm": 0.2745544910430908,
      "learning_rate": 0.00017934171880177265,
      "loss": 0.20311737060546875,
      "mean_token_accuracy": 0.9179251939058304,
      "num_tokens": 39930604.0,
      "step": 3237
    },
    {
      "entropy": 1.2923784255981445,
      "epoch": 1.7051079515534493,
      "grad_norm": 0.2750037610530853,
      "learning_rate": 0.00017932773780991262,
      "loss": 0.20452114939689636,
      "mean_token_accuracy": 0.9181451350450516,
      "num_tokens": 39942940.0,
      "step": 3238
    },
    {
      "entropy": 1.2803550958633423,
      "epoch": 1.7056345444971037,
      "grad_norm": 0.29063141345977783,
      "learning_rate": 0.00017931375270246889,
      "loss": 0.2145015299320221,
      "mean_token_accuracy": 0.9154239445924759,
      "num_tokens": 39955276.0,
      "step": 3239
    },
    {
      "entropy": 1.3234757483005524,
      "epoch": 1.7061611374407581,
      "grad_norm": 0.2841363549232483,
      "learning_rate": 0.0001792997634802716,
      "loss": 0.2186480611562729,
      "mean_token_accuracy": 0.9111131429672241,
      "num_tokens": 39967612.0,
      "step": 3240
    },
    {
      "entropy": 1.2682130932807922,
      "epoch": 1.706687730384413,
      "grad_norm": 0.25876402854919434,
      "learning_rate": 0.0001792857701441513,
      "loss": 0.20549407601356506,
      "mean_token_accuracy": 0.9134150445461273,
      "num_tokens": 39979948.0,
      "step": 3241
    },
    {
      "entropy": 1.285068303346634,
      "epoch": 1.7072143233280674,
      "grad_norm": 0.2754906415939331,
      "learning_rate": 0.00017927177269493866,
      "loss": 0.2059261053800583,
      "mean_token_accuracy": 0.9146608859300613,
      "num_tokens": 39992284.0,
      "step": 3242
    },
    {
      "entropy": 1.3181243240833282,
      "epoch": 1.7077409162717219,
      "grad_norm": 0.28565770387649536,
      "learning_rate": 0.0001792577711334646,
      "loss": 0.20405897498130798,
      "mean_token_accuracy": 0.9132011830806732,
      "num_tokens": 40004620.0,
      "step": 3243
    },
    {
      "entropy": 1.2542202472686768,
      "epoch": 1.7082675092153765,
      "grad_norm": 0.23145394027233124,
      "learning_rate": 0.00017924376546056035,
      "loss": 0.1774619072675705,
      "mean_token_accuracy": 0.9282086789608002,
      "num_tokens": 40016956.0,
      "step": 3244
    },
    {
      "entropy": 1.3313210010528564,
      "epoch": 1.7087941021590312,
      "grad_norm": 0.2665564715862274,
      "learning_rate": 0.0001792297556770573,
      "loss": 0.19780829548835754,
      "mean_token_accuracy": 0.9161526560783386,
      "num_tokens": 40029292.0,
      "step": 3245
    },
    {
      "entropy": 1.395924985408783,
      "epoch": 1.7093206951026856,
      "grad_norm": 0.28649643063545227,
      "learning_rate": 0.00017921574178378718,
      "loss": 0.19208171963691711,
      "mean_token_accuracy": 0.9216294139623642,
      "num_tokens": 40041628.0,
      "step": 3246
    },
    {
      "entropy": 1.323070913553238,
      "epoch": 1.7098472880463402,
      "grad_norm": 0.28210774064064026,
      "learning_rate": 0.00017920172378158188,
      "loss": 0.1932576298713684,
      "mean_token_accuracy": 0.9173065274953842,
      "num_tokens": 40053964.0,
      "step": 3247
    },
    {
      "entropy": 1.2962166965007782,
      "epoch": 1.7103738809899949,
      "grad_norm": 0.2612176537513733,
      "learning_rate": 0.00017918770167127355,
      "loss": 0.1979486644268036,
      "mean_token_accuracy": 0.9187701344490051,
      "num_tokens": 40066300.0,
      "step": 3248
    },
    {
      "entropy": 1.3086665868759155,
      "epoch": 1.7109004739336493,
      "grad_norm": 0.27552592754364014,
      "learning_rate": 0.00017917367545369469,
      "loss": 0.22310736775398254,
      "mean_token_accuracy": 0.9084995836019516,
      "num_tokens": 40078636.0,
      "step": 3249
    },
    {
      "entropy": 1.281711757183075,
      "epoch": 1.7114270668773037,
      "grad_norm": 0.3259299397468567,
      "learning_rate": 0.00017915964512967784,
      "loss": 0.1875232756137848,
      "mean_token_accuracy": 0.9251815974712372,
      "num_tokens": 40090972.0,
      "step": 3250
    },
    {
      "entropy": 1.3861521780490875,
      "epoch": 1.7119536598209584,
      "grad_norm": 0.313814640045166,
      "learning_rate": 0.00017914561070005598,
      "loss": 0.21126891672611237,
      "mean_token_accuracy": 0.9117094427347183,
      "num_tokens": 40103308.0,
      "step": 3251
    },
    {
      "entropy": 1.3072172403335571,
      "epoch": 1.712480252764613,
      "grad_norm": 0.3379667103290558,
      "learning_rate": 0.0001791315721656622,
      "loss": 0.19064629077911377,
      "mean_token_accuracy": 0.9234190434217453,
      "num_tokens": 40115644.0,
      "step": 3252
    },
    {
      "entropy": 1.2437802255153656,
      "epoch": 1.7130068457082674,
      "grad_norm": 0.26171544194221497,
      "learning_rate": 0.00017911752952732993,
      "loss": 0.19964775443077087,
      "mean_token_accuracy": 0.9152403026819229,
      "num_tokens": 40127980.0,
      "step": 3253
    },
    {
      "entropy": 1.2901654541492462,
      "epoch": 1.713533438651922,
      "grad_norm": 0.26816311478614807,
      "learning_rate": 0.00017910348278589276,
      "loss": 0.20712605118751526,
      "mean_token_accuracy": 0.9133376479148865,
      "num_tokens": 40140316.0,
      "step": 3254
    },
    {
      "entropy": 1.2976833879947662,
      "epoch": 1.7140600315955767,
      "grad_norm": 0.2973898947238922,
      "learning_rate": 0.0001790894319421846,
      "loss": 0.21765637397766113,
      "mean_token_accuracy": 0.9121676534414291,
      "num_tokens": 40152652.0,
      "step": 3255
    },
    {
      "entropy": 1.277085781097412,
      "epoch": 1.7145866245392312,
      "grad_norm": 0.2639175355434418,
      "learning_rate": 0.00017907537699703955,
      "loss": 0.19076862931251526,
      "mean_token_accuracy": 0.9191281646490097,
      "num_tokens": 40164988.0,
      "step": 3256
    },
    {
      "entropy": 1.3531688749790192,
      "epoch": 1.7151132174828856,
      "grad_norm": 0.27909165620803833,
      "learning_rate": 0.00017906131795129198,
      "loss": 0.18622638285160065,
      "mean_token_accuracy": 0.9284979552030563,
      "num_tokens": 40177324.0,
      "step": 3257
    },
    {
      "entropy": 1.242502599954605,
      "epoch": 1.7156398104265402,
      "grad_norm": 0.2541889250278473,
      "learning_rate": 0.00017904725480577646,
      "loss": 0.18720562756061554,
      "mean_token_accuracy": 0.9233348220586777,
      "num_tokens": 40189660.0,
      "step": 3258
    },
    {
      "entropy": 1.2683814764022827,
      "epoch": 1.7161664033701949,
      "grad_norm": 0.29242897033691406,
      "learning_rate": 0.0001790331875613279,
      "loss": 0.214931458234787,
      "mean_token_accuracy": 0.910991907119751,
      "num_tokens": 40201996.0,
      "step": 3259
    },
    {
      "entropy": 1.337379515171051,
      "epoch": 1.7166929963138493,
      "grad_norm": 0.2787255048751831,
      "learning_rate": 0.0001790191162187813,
      "loss": 0.2036406546831131,
      "mean_token_accuracy": 0.9184634983539581,
      "num_tokens": 40214332.0,
      "step": 3260
    },
    {
      "entropy": 1.257769137620926,
      "epoch": 1.717219589257504,
      "grad_norm": 0.25968125462532043,
      "learning_rate": 0.00017900504077897206,
      "loss": 0.19390641152858734,
      "mean_token_accuracy": 0.9182478189468384,
      "num_tokens": 40226668.0,
      "step": 3261
    },
    {
      "entropy": 1.3203734755516052,
      "epoch": 1.7177461822011586,
      "grad_norm": 0.30727487802505493,
      "learning_rate": 0.00017899096124273576,
      "loss": 0.22841309010982513,
      "mean_token_accuracy": 0.9035236537456512,
      "num_tokens": 40239004.0,
      "step": 3262
    },
    {
      "entropy": 1.2281590104103088,
      "epoch": 1.718272775144813,
      "grad_norm": 0.2654935121536255,
      "learning_rate": 0.0001789768776109082,
      "loss": 0.19407188892364502,
      "mean_token_accuracy": 0.9239892512559891,
      "num_tokens": 40251340.0,
      "step": 3263
    },
    {
      "entropy": 1.2642222046852112,
      "epoch": 1.7187993680884674,
      "grad_norm": 0.28257355093955994,
      "learning_rate": 0.0001789627898843254,
      "loss": 0.2133817970752716,
      "mean_token_accuracy": 0.9196331948041916,
      "num_tokens": 40263676.0,
      "step": 3264
    },
    {
      "entropy": 1.3009636104106903,
      "epoch": 1.7193259610321223,
      "grad_norm": 0.2984471917152405,
      "learning_rate": 0.00017894869806382377,
      "loss": 0.21264006197452545,
      "mean_token_accuracy": 0.9153096079826355,
      "num_tokens": 40276012.0,
      "step": 3265
    },
    {
      "entropy": 1.3348345458507538,
      "epoch": 1.7198525539757767,
      "grad_norm": 0.28683388233184814,
      "learning_rate": 0.00017893460215023975,
      "loss": 0.20870405435562134,
      "mean_token_accuracy": 0.9114125370979309,
      "num_tokens": 40288348.0,
      "step": 3266
    },
    {
      "entropy": 1.2550526559352875,
      "epoch": 1.7203791469194312,
      "grad_norm": 0.25430989265441895,
      "learning_rate": 0.0001789205021444102,
      "loss": 0.19396786391735077,
      "mean_token_accuracy": 0.9199714511632919,
      "num_tokens": 40300684.0,
      "step": 3267
    },
    {
      "entropy": 1.303754299879074,
      "epoch": 1.7209057398630858,
      "grad_norm": 0.26650470495224,
      "learning_rate": 0.00017890639804717215,
      "loss": 0.19954167306423187,
      "mean_token_accuracy": 0.9163612723350525,
      "num_tokens": 40313020.0,
      "step": 3268
    },
    {
      "entropy": 1.2093145549297333,
      "epoch": 1.7214323328067405,
      "grad_norm": 0.29942092299461365,
      "learning_rate": 0.00017889228985936285,
      "loss": 0.2275795340538025,
      "mean_token_accuracy": 0.9083501994609833,
      "num_tokens": 40325356.0,
      "step": 3269
    },
    {
      "entropy": 1.2750758230686188,
      "epoch": 1.7219589257503949,
      "grad_norm": 0.2606362998485565,
      "learning_rate": 0.00017887817758181987,
      "loss": 0.19598644971847534,
      "mean_token_accuracy": 0.9228833317756653,
      "num_tokens": 40337692.0,
      "step": 3270
    },
    {
      "entropy": 1.2626889944076538,
      "epoch": 1.7224855186940495,
      "grad_norm": 0.2737748622894287,
      "learning_rate": 0.0001788640612153809,
      "loss": 0.20550641417503357,
      "mean_token_accuracy": 0.9146024286746979,
      "num_tokens": 40350028.0,
      "step": 3271
    },
    {
      "entropy": 1.2671963572502136,
      "epoch": 1.7230121116377042,
      "grad_norm": 0.295658677816391,
      "learning_rate": 0.00017884994076088398,
      "loss": 0.2249743938446045,
      "mean_token_accuracy": 0.9077863395214081,
      "num_tokens": 40362364.0,
      "step": 3272
    },
    {
      "entropy": 1.256992220878601,
      "epoch": 1.7235387045813586,
      "grad_norm": 0.2737193703651428,
      "learning_rate": 0.0001788358162191674,
      "loss": 0.21205906569957733,
      "mean_token_accuracy": 0.915488988161087,
      "num_tokens": 40374700.0,
      "step": 3273
    },
    {
      "entropy": 1.213915228843689,
      "epoch": 1.724065297525013,
      "grad_norm": 0.25337645411491394,
      "learning_rate": 0.00017882168759106957,
      "loss": 0.18594782054424286,
      "mean_token_accuracy": 0.9206974655389786,
      "num_tokens": 40387036.0,
      "step": 3274
    },
    {
      "entropy": 1.2303422391414642,
      "epoch": 1.7245918904686677,
      "grad_norm": 0.28474873304367065,
      "learning_rate": 0.0001788075548774293,
      "loss": 0.2072606086730957,
      "mean_token_accuracy": 0.9152207672595978,
      "num_tokens": 40399372.0,
      "step": 3275
    },
    {
      "entropy": 1.2771183252334595,
      "epoch": 1.7251184834123223,
      "grad_norm": 0.27036532759666443,
      "learning_rate": 0.00017879341807908555,
      "loss": 0.1881149560213089,
      "mean_token_accuracy": 0.92488132417202,
      "num_tokens": 40411708.0,
      "step": 3276
    },
    {
      "entropy": 1.2247534692287445,
      "epoch": 1.7256450763559767,
      "grad_norm": 0.2693551778793335,
      "learning_rate": 0.0001787792771968775,
      "loss": 0.2171371728181839,
      "mean_token_accuracy": 0.91679947078228,
      "num_tokens": 40424044.0,
      "step": 3277
    },
    {
      "entropy": 1.2200176119804382,
      "epoch": 1.7261716692996314,
      "grad_norm": 0.2627072334289551,
      "learning_rate": 0.00017876513223164466,
      "loss": 0.19582206010818481,
      "mean_token_accuracy": 0.9185523092746735,
      "num_tokens": 40436380.0,
      "step": 3278
    },
    {
      "entropy": 1.2262316644191742,
      "epoch": 1.726698262243286,
      "grad_norm": 0.26391029357910156,
      "learning_rate": 0.00017875098318422667,
      "loss": 0.18920688331127167,
      "mean_token_accuracy": 0.9226961433887482,
      "num_tokens": 40448716.0,
      "step": 3279
    },
    {
      "entropy": 1.2009854018688202,
      "epoch": 1.7272248551869405,
      "grad_norm": 0.2786342203617096,
      "learning_rate": 0.00017873683005546358,
      "loss": 0.2078881859779358,
      "mean_token_accuracy": 0.9170061498880386,
      "num_tokens": 40461052.0,
      "step": 3280
    },
    {
      "entropy": 1.219001442193985,
      "epoch": 1.727751448130595,
      "grad_norm": 0.34324315190315247,
      "learning_rate": 0.0001787226728461955,
      "loss": 0.2038731575012207,
      "mean_token_accuracy": 0.9136009812355042,
      "num_tokens": 40473388.0,
      "step": 3281
    },
    {
      "entropy": 1.1982538998126984,
      "epoch": 1.7282780410742498,
      "grad_norm": 0.3113863170146942,
      "learning_rate": 0.00017870851155726286,
      "loss": 0.2176617681980133,
      "mean_token_accuracy": 0.9123759865760803,
      "num_tokens": 40485724.0,
      "step": 3282
    },
    {
      "entropy": 1.2175420820713043,
      "epoch": 1.7288046340179042,
      "grad_norm": 0.27579864859580994,
      "learning_rate": 0.0001786943461895064,
      "loss": 0.2074568271636963,
      "mean_token_accuracy": 0.9171206802129745,
      "num_tokens": 40498060.0,
      "step": 3283
    },
    {
      "entropy": 1.1838345229625702,
      "epoch": 1.7293312269615586,
      "grad_norm": 0.2992401421070099,
      "learning_rate": 0.00017868017674376694,
      "loss": 0.20533637702465057,
      "mean_token_accuracy": 0.9176798313856125,
      "num_tokens": 40510396.0,
      "step": 3284
    },
    {
      "entropy": 1.2316609919071198,
      "epoch": 1.7298578199052133,
      "grad_norm": 0.3072640001773834,
      "learning_rate": 0.00017866600322088568,
      "loss": 0.21974754333496094,
      "mean_token_accuracy": 0.9138150960206985,
      "num_tokens": 40522732.0,
      "step": 3285
    },
    {
      "entropy": 1.2149794399738312,
      "epoch": 1.730384412848868,
      "grad_norm": 0.2651062309741974,
      "learning_rate": 0.00017865182562170403,
      "loss": 0.1996372640132904,
      "mean_token_accuracy": 0.9211166948080063,
      "num_tokens": 40535068.0,
      "step": 3286
    },
    {
      "entropy": 1.2700909972190857,
      "epoch": 1.7309110057925223,
      "grad_norm": 0.2680571377277374,
      "learning_rate": 0.00017863764394706363,
      "loss": 0.1831960678100586,
      "mean_token_accuracy": 0.926160454750061,
      "num_tokens": 40547404.0,
      "step": 3287
    },
    {
      "entropy": 1.2134180068969727,
      "epoch": 1.731437598736177,
      "grad_norm": 0.2633264362812042,
      "learning_rate": 0.0001786234581978064,
      "loss": 0.19375747442245483,
      "mean_token_accuracy": 0.9277003556489944,
      "num_tokens": 40559740.0,
      "step": 3288
    },
    {
      "entropy": 1.2066058814525604,
      "epoch": 1.7319641916798316,
      "grad_norm": 0.2779509723186493,
      "learning_rate": 0.00017860926837477437,
      "loss": 0.2058328092098236,
      "mean_token_accuracy": 0.9140601456165314,
      "num_tokens": 40572076.0,
      "step": 3289
    },
    {
      "entropy": 1.2765550911426544,
      "epoch": 1.732490784623486,
      "grad_norm": 0.26993927359580994,
      "learning_rate": 0.00017859507447880998,
      "loss": 0.19165240228176117,
      "mean_token_accuracy": 0.9281281679868698,
      "num_tokens": 40584412.0,
      "step": 3290
    },
    {
      "entropy": 1.260106086730957,
      "epoch": 1.7330173775671405,
      "grad_norm": 0.2724703848361969,
      "learning_rate": 0.00017858087651075584,
      "loss": 0.18630154430866241,
      "mean_token_accuracy": 0.9287521094083786,
      "num_tokens": 40596748.0,
      "step": 3291
    },
    {
      "entropy": 1.2476516962051392,
      "epoch": 1.7335439705107951,
      "grad_norm": 0.2591682970523834,
      "learning_rate": 0.00017856667447145475,
      "loss": 0.18758217990398407,
      "mean_token_accuracy": 0.9250517785549164,
      "num_tokens": 40609084.0,
      "step": 3292
    },
    {
      "entropy": 1.2707555294036865,
      "epoch": 1.7340705634544498,
      "grad_norm": 0.3217529058456421,
      "learning_rate": 0.00017855246836174986,
      "loss": 0.21787390112876892,
      "mean_token_accuracy": 0.9111398458480835,
      "num_tokens": 40621420.0,
      "step": 3293
    },
    {
      "entropy": 1.2611454129219055,
      "epoch": 1.7345971563981042,
      "grad_norm": 0.26216942071914673,
      "learning_rate": 0.00017853825818248443,
      "loss": 0.19248586893081665,
      "mean_token_accuracy": 0.9217905402183533,
      "num_tokens": 40633756.0,
      "step": 3294
    },
    {
      "entropy": 1.2096803486347198,
      "epoch": 1.7351237493417588,
      "grad_norm": 0.2533584535121918,
      "learning_rate": 0.00017852404393450214,
      "loss": 0.19668050110340118,
      "mean_token_accuracy": 0.9232807159423828,
      "num_tokens": 40646092.0,
      "step": 3295
    },
    {
      "entropy": 1.197802484035492,
      "epoch": 1.7356503422854135,
      "grad_norm": 0.25326570868492126,
      "learning_rate": 0.00017850982561864673,
      "loss": 0.17445367574691772,
      "mean_token_accuracy": 0.9264692068099976,
      "num_tokens": 40658428.0,
      "step": 3296
    },
    {
      "entropy": 1.2270566523075104,
      "epoch": 1.736176935229068,
      "grad_norm": 0.26528990268707275,
      "learning_rate": 0.00017849560323576226,
      "loss": 0.1855545938014984,
      "mean_token_accuracy": 0.9231870025396347,
      "num_tokens": 40670764.0,
      "step": 3297
    },
    {
      "entropy": 1.2130609154701233,
      "epoch": 1.7367035281727223,
      "grad_norm": 0.28013598918914795,
      "learning_rate": 0.00017848137678669307,
      "loss": 0.19601064920425415,
      "mean_token_accuracy": 0.9243181943893433,
      "num_tokens": 40683100.0,
      "step": 3298
    },
    {
      "entropy": 1.226871132850647,
      "epoch": 1.7372301211163772,
      "grad_norm": 0.2821487486362457,
      "learning_rate": 0.0001784671462722837,
      "loss": 0.21541166305541992,
      "mean_token_accuracy": 0.9133672714233398,
      "num_tokens": 40695436.0,
      "step": 3299
    },
    {
      "entropy": 1.2181604504585266,
      "epoch": 1.7377567140600316,
      "grad_norm": 0.2949363589286804,
      "learning_rate": 0.00017845291169337886,
      "loss": 0.2169766128063202,
      "mean_token_accuracy": 0.9101507663726807,
      "num_tokens": 40707772.0,
      "step": 3300
    },
    {
      "entropy": 1.237768977880478,
      "epoch": 1.738283307003686,
      "grad_norm": 0.3046022951602936,
      "learning_rate": 0.00017843867305082368,
      "loss": 0.20753508806228638,
      "mean_token_accuracy": 0.9195269197225571,
      "num_tokens": 40720108.0,
      "step": 3301
    },
    {
      "entropy": 1.206713229417801,
      "epoch": 1.7388098999473407,
      "grad_norm": 0.2679121196269989,
      "learning_rate": 0.0001784244303454633,
      "loss": 0.2043764740228653,
      "mean_token_accuracy": 0.9116311520338058,
      "num_tokens": 40732444.0,
      "step": 3302
    },
    {
      "entropy": 1.2439360618591309,
      "epoch": 1.7393364928909953,
      "grad_norm": 0.25374096632003784,
      "learning_rate": 0.00017841018357814338,
      "loss": 0.18161284923553467,
      "mean_token_accuracy": 0.9243684113025665,
      "num_tokens": 40744780.0,
      "step": 3303
    },
    {
      "entropy": 1.2688191533088684,
      "epoch": 1.7398630858346498,
      "grad_norm": 0.28752121329307556,
      "learning_rate": 0.00017839593274970953,
      "loss": 0.20974797010421753,
      "mean_token_accuracy": 0.9088063091039658,
      "num_tokens": 40757116.0,
      "step": 3304
    },
    {
      "entropy": 1.2342472076416016,
      "epoch": 1.7403896787783044,
      "grad_norm": 0.2605990469455719,
      "learning_rate": 0.00017838167786100786,
      "loss": 0.2021808922290802,
      "mean_token_accuracy": 0.9125561267137527,
      "num_tokens": 40769452.0,
      "step": 3305
    },
    {
      "entropy": 1.2520089447498322,
      "epoch": 1.740916271721959,
      "grad_norm": 0.2876226007938385,
      "learning_rate": 0.00017836741891288447,
      "loss": 0.21958418190479279,
      "mean_token_accuracy": 0.9127276241779327,
      "num_tokens": 40781788.0,
      "step": 3306
    },
    {
      "entropy": 1.2590630948543549,
      "epoch": 1.7414428646656135,
      "grad_norm": 0.267333984375,
      "learning_rate": 0.00017835315590618593,
      "loss": 0.19477830827236176,
      "mean_token_accuracy": 0.9198647290468216,
      "num_tokens": 40794124.0,
      "step": 3307
    },
    {
      "entropy": 1.222892850637436,
      "epoch": 1.741969457609268,
      "grad_norm": 0.2734076976776123,
      "learning_rate": 0.0001783388888417589,
      "loss": 0.1936633288860321,
      "mean_token_accuracy": 0.9192850589752197,
      "num_tokens": 40806460.0,
      "step": 3308
    },
    {
      "entropy": 1.2516216337680817,
      "epoch": 1.7424960505529226,
      "grad_norm": 0.293828547000885,
      "learning_rate": 0.00017832461772045033,
      "loss": 0.20226381719112396,
      "mean_token_accuracy": 0.9191479235887527,
      "num_tokens": 40818796.0,
      "step": 3309
    },
    {
      "entropy": 1.2808569073677063,
      "epoch": 1.7430226434965772,
      "grad_norm": 0.29275745153427124,
      "learning_rate": 0.00017831034254310748,
      "loss": 0.2209526002407074,
      "mean_token_accuracy": 0.9084486216306686,
      "num_tokens": 40831132.0,
      "step": 3310
    },
    {
      "entropy": 1.2265948355197906,
      "epoch": 1.7435492364402316,
      "grad_norm": 0.2859675884246826,
      "learning_rate": 0.0001782960633105777,
      "loss": 0.20897722244262695,
      "mean_token_accuracy": 0.9134756773710251,
      "num_tokens": 40843468.0,
      "step": 3311
    },
    {
      "entropy": 1.2819762527942657,
      "epoch": 1.7440758293838863,
      "grad_norm": 0.2739836275577545,
      "learning_rate": 0.0001782817800237087,
      "loss": 0.20494410395622253,
      "mean_token_accuracy": 0.9149334281682968,
      "num_tokens": 40855804.0,
      "step": 3312
    },
    {
      "entropy": 1.2840421795845032,
      "epoch": 1.744602422327541,
      "grad_norm": 0.27893322706222534,
      "learning_rate": 0.0001782674926833484,
      "loss": 0.20927336812019348,
      "mean_token_accuracy": 0.9126023948192596,
      "num_tokens": 40868140.0,
      "step": 3313
    },
    {
      "entropy": 1.2541827261447906,
      "epoch": 1.7451290152711953,
      "grad_norm": 0.2732979953289032,
      "learning_rate": 0.00017825320129034495,
      "loss": 0.20452941954135895,
      "mean_token_accuracy": 0.9158912748098373,
      "num_tokens": 40880476.0,
      "step": 3314
    },
    {
      "entropy": 1.2899048030376434,
      "epoch": 1.7456556082148498,
      "grad_norm": 0.3004266321659088,
      "learning_rate": 0.0001782389058455468,
      "loss": 0.20099236071109772,
      "mean_token_accuracy": 0.9254409223794937,
      "num_tokens": 40892812.0,
      "step": 3315
    },
    {
      "entropy": 1.287499874830246,
      "epoch": 1.7461822011585044,
      "grad_norm": 0.2748308479785919,
      "learning_rate": 0.00017822460634980245,
      "loss": 0.19413264095783234,
      "mean_token_accuracy": 0.9179486334323883,
      "num_tokens": 40905148.0,
      "step": 3316
    },
    {
      "entropy": 1.2944321632385254,
      "epoch": 1.746708794102159,
      "grad_norm": 0.2908708453178406,
      "learning_rate": 0.00017821030280396093,
      "loss": 0.21369019150733948,
      "mean_token_accuracy": 0.9173711091279984,
      "num_tokens": 40917484.0,
      "step": 3317
    },
    {
      "entropy": 1.262908548116684,
      "epoch": 1.7472353870458135,
      "grad_norm": 0.25998809933662415,
      "learning_rate": 0.00017819599520887126,
      "loss": 0.19156329333782196,
      "mean_token_accuracy": 0.9206810593605042,
      "num_tokens": 40929820.0,
      "step": 3318
    },
    {
      "entropy": 1.3015094697475433,
      "epoch": 1.7477619799894681,
      "grad_norm": 0.30315208435058594,
      "learning_rate": 0.00017818168356538284,
      "loss": 0.2028445154428482,
      "mean_token_accuracy": 0.9121357649564743,
      "num_tokens": 40942156.0,
      "step": 3319
    },
    {
      "entropy": 1.2577976882457733,
      "epoch": 1.7482885729331228,
      "grad_norm": 0.251533180475235,
      "learning_rate": 0.00017816736787434526,
      "loss": 0.17267408967018127,
      "mean_token_accuracy": 0.9274961352348328,
      "num_tokens": 40954492.0,
      "step": 3320
    },
    {
      "entropy": 1.298771619796753,
      "epoch": 1.7488151658767772,
      "grad_norm": 0.26891663670539856,
      "learning_rate": 0.00017815304813660835,
      "loss": 0.20470818877220154,
      "mean_token_accuracy": 0.9156215190887451,
      "num_tokens": 40966828.0,
      "step": 3321
    },
    {
      "entropy": 1.297717809677124,
      "epoch": 1.7493417588204319,
      "grad_norm": 0.277751624584198,
      "learning_rate": 0.00017813872435302222,
      "loss": 0.1922859251499176,
      "mean_token_accuracy": 0.9206653833389282,
      "num_tokens": 40979164.0,
      "step": 3322
    },
    {
      "entropy": 1.3184978663921356,
      "epoch": 1.7498683517640865,
      "grad_norm": 0.28764608502388,
      "learning_rate": 0.00017812439652443716,
      "loss": 0.19598035514354706,
      "mean_token_accuracy": 0.922876164317131,
      "num_tokens": 40991500.0,
      "step": 3323
    },
    {
      "entropy": 1.2648021578788757,
      "epoch": 1.750394944707741,
      "grad_norm": 0.2744153141975403,
      "learning_rate": 0.0001781100646517037,
      "loss": 0.19071339070796967,
      "mean_token_accuracy": 0.9216628968715668,
      "num_tokens": 41003836.0,
      "step": 3324
    },
    {
      "entropy": 1.2510377168655396,
      "epoch": 1.7509215376513954,
      "grad_norm": 0.2512511610984802,
      "learning_rate": 0.00017809572873567274,
      "loss": 0.16837379336357117,
      "mean_token_accuracy": 0.9307546615600586,
      "num_tokens": 41016172.0,
      "step": 3325
    },
    {
      "entropy": 1.282473772764206,
      "epoch": 1.75144813059505,
      "grad_norm": 0.29696282744407654,
      "learning_rate": 0.00017808138877719522,
      "loss": 0.19326961040496826,
      "mean_token_accuracy": 0.9177820086479187,
      "num_tokens": 41028508.0,
      "step": 3326
    },
    {
      "entropy": 1.2485735416412354,
      "epoch": 1.7519747235387046,
      "grad_norm": 0.25719940662384033,
      "learning_rate": 0.00017806704477712245,
      "loss": 0.19185180962085724,
      "mean_token_accuracy": 0.9225496500730515,
      "num_tokens": 41040844.0,
      "step": 3327
    },
    {
      "entropy": 1.254735678434372,
      "epoch": 1.752501316482359,
      "grad_norm": 0.2844921946525574,
      "learning_rate": 0.000178052696736306,
      "loss": 0.20377546548843384,
      "mean_token_accuracy": 0.9177062809467316,
      "num_tokens": 41053180.0,
      "step": 3328
    },
    {
      "entropy": 1.2341787219047546,
      "epoch": 1.7530279094260137,
      "grad_norm": 0.2937590479850769,
      "learning_rate": 0.00017803834465559754,
      "loss": 0.2154744565486908,
      "mean_token_accuracy": 0.9114655256271362,
      "num_tokens": 41065516.0,
      "step": 3329
    },
    {
      "entropy": 1.2713043689727783,
      "epoch": 1.7535545023696684,
      "grad_norm": 0.29252561926841736,
      "learning_rate": 0.00017802398853584915,
      "loss": 0.20079272985458374,
      "mean_token_accuracy": 0.9159099161624908,
      "num_tokens": 41077852.0,
      "step": 3330
    },
    {
      "entropy": 1.2464990019798279,
      "epoch": 1.7540810953133228,
      "grad_norm": 0.2867441177368164,
      "learning_rate": 0.000178009628377913,
      "loss": 0.2051943838596344,
      "mean_token_accuracy": 0.9145941138267517,
      "num_tokens": 41090188.0,
      "step": 3331
    },
    {
      "entropy": 1.2052176296710968,
      "epoch": 1.7546076882569772,
      "grad_norm": 0.257271945476532,
      "learning_rate": 0.0001779952641826416,
      "loss": 0.18216028809547424,
      "mean_token_accuracy": 0.9247520416975021,
      "num_tokens": 41102524.0,
      "step": 3332
    },
    {
      "entropy": 1.2561128735542297,
      "epoch": 1.7551342812006319,
      "grad_norm": 0.2668805718421936,
      "learning_rate": 0.00017798089595088773,
      "loss": 0.17578330636024475,
      "mean_token_accuracy": 0.9274698793888092,
      "num_tokens": 41114860.0,
      "step": 3333
    },
    {
      "entropy": 1.2273137271404266,
      "epoch": 1.7556608741442865,
      "grad_norm": 0.28731396794319153,
      "learning_rate": 0.00017796652368350422,
      "loss": 0.20657789707183838,
      "mean_token_accuracy": 0.9174284040927887,
      "num_tokens": 41127196.0,
      "step": 3334
    },
    {
      "entropy": 1.217975229024887,
      "epoch": 1.756187467087941,
      "grad_norm": 0.24883601069450378,
      "learning_rate": 0.00017795214738134437,
      "loss": 0.1853816658258438,
      "mean_token_accuracy": 0.9215361028909683,
      "num_tokens": 41139532.0,
      "step": 3335
    },
    {
      "entropy": 1.270724892616272,
      "epoch": 1.7567140600315956,
      "grad_norm": 0.3023955523967743,
      "learning_rate": 0.00017793776704526156,
      "loss": 0.21046261489391327,
      "mean_token_accuracy": 0.9148201197385788,
      "num_tokens": 41151868.0,
      "step": 3336
    },
    {
      "entropy": 1.3148931860923767,
      "epoch": 1.7572406529752502,
      "grad_norm": 0.31092697381973267,
      "learning_rate": 0.00017792338267610955,
      "loss": 0.22172124683856964,
      "mean_token_accuracy": 0.9151052385568619,
      "num_tokens": 41164204.0,
      "step": 3337
    },
    {
      "entropy": 1.2632471919059753,
      "epoch": 1.7577672459189047,
      "grad_norm": 0.26783210039138794,
      "learning_rate": 0.00017790899427474216,
      "loss": 0.197731152176857,
      "mean_token_accuracy": 0.9220364391803741,
      "num_tokens": 41176540.0,
      "step": 3338
    },
    {
      "entropy": 1.3053453266620636,
      "epoch": 1.758293838862559,
      "grad_norm": 0.28016164898872375,
      "learning_rate": 0.0001778946018420136,
      "loss": 0.19119513034820557,
      "mean_token_accuracy": 0.9238210469484329,
      "num_tokens": 41188876.0,
      "step": 3339
    },
    {
      "entropy": 1.2755919098854065,
      "epoch": 1.758820431806214,
      "grad_norm": 0.2692145109176636,
      "learning_rate": 0.00017788020537877822,
      "loss": 0.20970848202705383,
      "mean_token_accuracy": 0.9092446118593216,
      "num_tokens": 41201212.0,
      "step": 3340
    },
    {
      "entropy": 1.3126140236854553,
      "epoch": 1.7593470247498684,
      "grad_norm": 0.2891995906829834,
      "learning_rate": 0.00017786580488589072,
      "loss": 0.20718203485012054,
      "mean_token_accuracy": 0.9121265858411789,
      "num_tokens": 41213548.0,
      "step": 3341
    },
    {
      "entropy": 1.2515637278556824,
      "epoch": 1.7598736176935228,
      "grad_norm": 0.2759207487106323,
      "learning_rate": 0.0001778514003642059,
      "loss": 0.2197999209165573,
      "mean_token_accuracy": 0.9112512469291687,
      "num_tokens": 41225884.0,
      "step": 3342
    },
    {
      "entropy": 1.3375881612300873,
      "epoch": 1.7604002106371774,
      "grad_norm": 0.3448629677295685,
      "learning_rate": 0.0001778369918145789,
      "loss": 0.22216130793094635,
      "mean_token_accuracy": 0.9111101776361465,
      "num_tokens": 41238220.0,
      "step": 3343
    },
    {
      "entropy": 1.3089163899421692,
      "epoch": 1.760926803580832,
      "grad_norm": 0.3070039749145508,
      "learning_rate": 0.00017782257923786512,
      "loss": 0.227981299161911,
      "mean_token_accuracy": 0.9092568904161453,
      "num_tokens": 41250556.0,
      "step": 3344
    },
    {
      "entropy": 1.2772272825241089,
      "epoch": 1.7614533965244865,
      "grad_norm": 0.3033801317214966,
      "learning_rate": 0.0001778081626349201,
      "loss": 0.23961672186851501,
      "mean_token_accuracy": 0.9030078947544098,
      "num_tokens": 41262892.0,
      "step": 3345
    },
    {
      "entropy": 1.2329810559749603,
      "epoch": 1.7619799894681412,
      "grad_norm": 0.2866532504558563,
      "learning_rate": 0.00017779374200659967,
      "loss": 0.19439920783042908,
      "mean_token_accuracy": 0.9247799217700958,
      "num_tokens": 41275228.0,
      "step": 3346
    },
    {
      "entropy": 1.3167122602462769,
      "epoch": 1.7625065824117958,
      "grad_norm": 0.2825206518173218,
      "learning_rate": 0.0001777793173537599,
      "loss": 0.22407029569149017,
      "mean_token_accuracy": 0.9099050462245941,
      "num_tokens": 41287564.0,
      "step": 3347
    },
    {
      "entropy": 1.2629252672195435,
      "epoch": 1.7630331753554502,
      "grad_norm": 0.2708073854446411,
      "learning_rate": 0.00017776488867725712,
      "loss": 0.20463934540748596,
      "mean_token_accuracy": 0.9120416194200516,
      "num_tokens": 41299900.0,
      "step": 3348
    },
    {
      "entropy": 1.27415269613266,
      "epoch": 1.7635597682991047,
      "grad_norm": 0.2722845673561096,
      "learning_rate": 0.00017775045597794787,
      "loss": 0.21537022292613983,
      "mean_token_accuracy": 0.9141311943531036,
      "num_tokens": 41312236.0,
      "step": 3349
    },
    {
      "entropy": 1.3064779043197632,
      "epoch": 1.7640863612427593,
      "grad_norm": 0.2894989848136902,
      "learning_rate": 0.00017773601925668892,
      "loss": 0.22586101293563843,
      "mean_token_accuracy": 0.9097247868776321,
      "num_tokens": 41324572.0,
      "step": 3350
    },
    {
      "entropy": 1.2612329721450806,
      "epoch": 1.764612954186414,
      "grad_norm": 0.29836785793304443,
      "learning_rate": 0.00017772157851433728,
      "loss": 0.21758875250816345,
      "mean_token_accuracy": 0.9101037234067917,
      "num_tokens": 41336908.0,
      "step": 3351
    },
    {
      "entropy": 1.2290297746658325,
      "epoch": 1.7651395471300684,
      "grad_norm": 0.25658851861953735,
      "learning_rate": 0.00017770713375175027,
      "loss": 0.2039872407913208,
      "mean_token_accuracy": 0.9155770987272263,
      "num_tokens": 41349244.0,
      "step": 3352
    },
    {
      "entropy": 1.2159041166305542,
      "epoch": 1.765666140073723,
      "grad_norm": 0.25883957743644714,
      "learning_rate": 0.00017769268496978534,
      "loss": 0.2031729817390442,
      "mean_token_accuracy": 0.9135902374982834,
      "num_tokens": 41361580.0,
      "step": 3353
    },
    {
      "entropy": 1.2781166732311249,
      "epoch": 1.7661927330173777,
      "grad_norm": 0.2804759740829468,
      "learning_rate": 0.0001776782321693003,
      "loss": 0.20843571424484253,
      "mean_token_accuracy": 0.9182161241769791,
      "num_tokens": 41373916.0,
      "step": 3354
    },
    {
      "entropy": 1.2583058774471283,
      "epoch": 1.766719325961032,
      "grad_norm": 0.2756879925727844,
      "learning_rate": 0.000177663775351153,
      "loss": 0.20797543227672577,
      "mean_token_accuracy": 0.9124239981174469,
      "num_tokens": 41386252.0,
      "step": 3355
    },
    {
      "entropy": 1.216934710741043,
      "epoch": 1.7672459189046865,
      "grad_norm": 0.24864549934864044,
      "learning_rate": 0.00017764931451620176,
      "loss": 0.19844800233840942,
      "mean_token_accuracy": 0.9174752831459045,
      "num_tokens": 41398588.0,
      "step": 3356
    },
    {
      "entropy": 1.1914585530757904,
      "epoch": 1.7677725118483414,
      "grad_norm": 0.3007160723209381,
      "learning_rate": 0.000177634849665305,
      "loss": 0.19987277686595917,
      "mean_token_accuracy": 0.9170159250497818,
      "num_tokens": 41410924.0,
      "step": 3357
    },
    {
      "entropy": 1.2236018180847168,
      "epoch": 1.7682991047919958,
      "grad_norm": 0.26258760690689087,
      "learning_rate": 0.00017762038079932143,
      "loss": 0.19202816486358643,
      "mean_token_accuracy": 0.9185950011014938,
      "num_tokens": 41423260.0,
      "step": 3358
    },
    {
      "entropy": 1.203150063753128,
      "epoch": 1.7688256977356502,
      "grad_norm": 0.30711668729782104,
      "learning_rate": 0.00017760590791910996,
      "loss": 0.199526846408844,
      "mean_token_accuracy": 0.9178106486797333,
      "num_tokens": 41435596.0,
      "step": 3359
    },
    {
      "entropy": 1.1577493846416473,
      "epoch": 1.7693522906793049,
      "grad_norm": 0.2871035933494568,
      "learning_rate": 0.0001775914310255298,
      "loss": 0.20930145680904388,
      "mean_token_accuracy": 0.9113402664661407,
      "num_tokens": 41447932.0,
      "step": 3360
    },
    {
      "entropy": 1.1829438507556915,
      "epoch": 1.7698788836229595,
      "grad_norm": 0.2672130763530731,
      "learning_rate": 0.0001775769501194403,
      "loss": 0.19939038157463074,
      "mean_token_accuracy": 0.9205279648303986,
      "num_tokens": 41460268.0,
      "step": 3361
    },
    {
      "entropy": 1.1942402124404907,
      "epoch": 1.770405476566614,
      "grad_norm": 0.25566044449806213,
      "learning_rate": 0.00017756246520170114,
      "loss": 0.1758129745721817,
      "mean_token_accuracy": 0.9335868656635284,
      "num_tokens": 41472604.0,
      "step": 3362
    },
    {
      "entropy": 1.1532124280929565,
      "epoch": 1.7709320695102686,
      "grad_norm": 0.27559491991996765,
      "learning_rate": 0.00017754797627317223,
      "loss": 0.21339191496372223,
      "mean_token_accuracy": 0.9085791558027267,
      "num_tokens": 41484940.0,
      "step": 3363
    },
    {
      "entropy": 1.1915392577648163,
      "epoch": 1.7714586624539232,
      "grad_norm": 0.27328014373779297,
      "learning_rate": 0.00017753348333471368,
      "loss": 0.19107362627983093,
      "mean_token_accuracy": 0.919417142868042,
      "num_tokens": 41497276.0,
      "step": 3364
    },
    {
      "entropy": 1.2415129840373993,
      "epoch": 1.7719852553975777,
      "grad_norm": 0.3027941584587097,
      "learning_rate": 0.00017751898638718582,
      "loss": 0.21972572803497314,
      "mean_token_accuracy": 0.9154131710529327,
      "num_tokens": 41509612.0,
      "step": 3365
    },
    {
      "entropy": 1.2410273849964142,
      "epoch": 1.772511848341232,
      "grad_norm": 0.2616625726222992,
      "learning_rate": 0.00017750448543144923,
      "loss": 0.19885270297527313,
      "mean_token_accuracy": 0.9170425534248352,
      "num_tokens": 41521948.0,
      "step": 3366
    },
    {
      "entropy": 1.17729052901268,
      "epoch": 1.7730384412848867,
      "grad_norm": 0.2854722738265991,
      "learning_rate": 0.00017748998046836483,
      "loss": 0.21116960048675537,
      "mean_token_accuracy": 0.9104287326335907,
      "num_tokens": 41534284.0,
      "step": 3367
    },
    {
      "entropy": 1.2824692726135254,
      "epoch": 1.7735650342285414,
      "grad_norm": 0.3191331923007965,
      "learning_rate": 0.00017747547149879364,
      "loss": 0.22000783681869507,
      "mean_token_accuracy": 0.906844973564148,
      "num_tokens": 41546620.0,
      "step": 3368
    },
    {
      "entropy": 1.2594166994094849,
      "epoch": 1.7740916271721958,
      "grad_norm": 0.28365424275398254,
      "learning_rate": 0.000177460958523597,
      "loss": 0.2075151801109314,
      "mean_token_accuracy": 0.913050040602684,
      "num_tokens": 41558956.0,
      "step": 3369
    },
    {
      "entropy": 1.235844612121582,
      "epoch": 1.7746182201158505,
      "grad_norm": 0.2545213997364044,
      "learning_rate": 0.00017744644154363642,
      "loss": 0.19338078796863556,
      "mean_token_accuracy": 0.923160582780838,
      "num_tokens": 41571292.0,
      "step": 3370
    },
    {
      "entropy": 1.3051631450653076,
      "epoch": 1.7751448130595051,
      "grad_norm": 0.2971054017543793,
      "learning_rate": 0.00017743192055977373,
      "loss": 0.20190899074077606,
      "mean_token_accuracy": 0.9218249917030334,
      "num_tokens": 41583628.0,
      "step": 3371
    },
    {
      "entropy": 1.2504850327968597,
      "epoch": 1.7756714060031595,
      "grad_norm": 0.2931920886039734,
      "learning_rate": 0.00017741739557287093,
      "loss": 0.22522364556789398,
      "mean_token_accuracy": 0.9096525609493256,
      "num_tokens": 41595964.0,
      "step": 3372
    },
    {
      "entropy": 1.239783525466919,
      "epoch": 1.776197998946814,
      "grad_norm": 0.2665524482727051,
      "learning_rate": 0.0001774028665837903,
      "loss": 0.20252424478530884,
      "mean_token_accuracy": 0.9206974357366562,
      "num_tokens": 41608300.0,
      "step": 3373
    },
    {
      "entropy": 1.2758963406085968,
      "epoch": 1.7767245918904688,
      "grad_norm": 0.28818976879119873,
      "learning_rate": 0.00017738833359339434,
      "loss": 0.20459452271461487,
      "mean_token_accuracy": 0.9180116057395935,
      "num_tokens": 41620636.0,
      "step": 3374
    },
    {
      "entropy": 1.20317143201828,
      "epoch": 1.7772511848341233,
      "grad_norm": 0.24318887293338776,
      "learning_rate": 0.0001773737966025458,
      "loss": 0.18710657954216003,
      "mean_token_accuracy": 0.9200632572174072,
      "num_tokens": 41632972.0,
      "step": 3375
    },
    {
      "entropy": 1.2684413492679596,
      "epoch": 1.7777777777777777,
      "grad_norm": 0.2672540843486786,
      "learning_rate": 0.0001773592556121076,
      "loss": 0.21147683262825012,
      "mean_token_accuracy": 0.9108813256025314,
      "num_tokens": 41645308.0,
      "step": 3376
    },
    {
      "entropy": 1.2610873579978943,
      "epoch": 1.7783043707214323,
      "grad_norm": 0.2789379060268402,
      "learning_rate": 0.00017734471062294302,
      "loss": 0.19667166471481323,
      "mean_token_accuracy": 0.9191772490739822,
      "num_tokens": 41657644.0,
      "step": 3377
    },
    {
      "entropy": 1.209890991449356,
      "epoch": 1.778830963665087,
      "grad_norm": 0.259087473154068,
      "learning_rate": 0.00017733016163591547,
      "loss": 0.19788247346878052,
      "mean_token_accuracy": 0.9188321679830551,
      "num_tokens": 41669980.0,
      "step": 3378
    },
    {
      "entropy": 1.2839680910110474,
      "epoch": 1.7793575566087414,
      "grad_norm": 0.27051210403442383,
      "learning_rate": 0.00017731560865188865,
      "loss": 0.19803057610988617,
      "mean_token_accuracy": 0.921959638595581,
      "num_tokens": 41682316.0,
      "step": 3379
    },
    {
      "entropy": 1.2436229586601257,
      "epoch": 1.779884149552396,
      "grad_norm": 0.28252285718917847,
      "learning_rate": 0.0001773010516717265,
      "loss": 0.19300350546836853,
      "mean_token_accuracy": 0.9220046401023865,
      "num_tokens": 41694652.0,
      "step": 3380
    },
    {
      "entropy": 1.2022179961204529,
      "epoch": 1.7804107424960507,
      "grad_norm": 0.27916550636291504,
      "learning_rate": 0.00017728649069629316,
      "loss": 0.21073868870735168,
      "mean_token_accuracy": 0.913501039147377,
      "num_tokens": 41706988.0,
      "step": 3381
    },
    {
      "entropy": 1.1777479350566864,
      "epoch": 1.7809373354397051,
      "grad_norm": 0.2819908857345581,
      "learning_rate": 0.00017727192572645307,
      "loss": 0.2172769159078598,
      "mean_token_accuracy": 0.9113239943981171,
      "num_tokens": 41719324.0,
      "step": 3382
    },
    {
      "entropy": 1.2020470798015594,
      "epoch": 1.7814639283833595,
      "grad_norm": 0.2866542637348175,
      "learning_rate": 0.00017725735676307083,
      "loss": 0.20779943466186523,
      "mean_token_accuracy": 0.9181617349386215,
      "num_tokens": 41731660.0,
      "step": 3383
    },
    {
      "entropy": 1.2138996124267578,
      "epoch": 1.7819905213270142,
      "grad_norm": 0.26489347219467163,
      "learning_rate": 0.0001772427838070113,
      "loss": 0.2014807164669037,
      "mean_token_accuracy": 0.917888268828392,
      "num_tokens": 41743996.0,
      "step": 3384
    },
    {
      "entropy": 1.1408116817474365,
      "epoch": 1.7825171142706688,
      "grad_norm": 0.2681819498538971,
      "learning_rate": 0.00017722820685913963,
      "loss": 0.21120908856391907,
      "mean_token_accuracy": 0.9159476608037949,
      "num_tokens": 41756332.0,
      "step": 3385
    },
    {
      "entropy": 1.1952861547470093,
      "epoch": 1.7830437072143233,
      "grad_norm": 0.27355825901031494,
      "learning_rate": 0.00017721362592032117,
      "loss": 0.20519855618476868,
      "mean_token_accuracy": 0.9159957766532898,
      "num_tokens": 41768668.0,
      "step": 3386
    },
    {
      "entropy": 1.2068920731544495,
      "epoch": 1.783570300157978,
      "grad_norm": 0.28034836053848267,
      "learning_rate": 0.00017719904099142147,
      "loss": 0.20763035118579865,
      "mean_token_accuracy": 0.9192133694887161,
      "num_tokens": 41781004.0,
      "step": 3387
    },
    {
      "entropy": 1.1186204552650452,
      "epoch": 1.7840968931016326,
      "grad_norm": 0.26254552602767944,
      "learning_rate": 0.0001771844520733064,
      "loss": 0.20995590090751648,
      "mean_token_accuracy": 0.9153067320585251,
      "num_tokens": 41793340.0,
      "step": 3388
    },
    {
      "entropy": 1.1552630364894867,
      "epoch": 1.784623486045287,
      "grad_norm": 0.2836894392967224,
      "learning_rate": 0.00017716985916684197,
      "loss": 0.21084123849868774,
      "mean_token_accuracy": 0.9113008379936218,
      "num_tokens": 41805676.0,
      "step": 3389
    },
    {
      "entropy": 1.1729071736335754,
      "epoch": 1.7851500789889414,
      "grad_norm": 0.2675871253013611,
      "learning_rate": 0.0001771552622728945,
      "loss": 0.19911614060401917,
      "mean_token_accuracy": 0.9161395877599716,
      "num_tokens": 41818012.0,
      "step": 3390
    },
    {
      "entropy": 1.183312863111496,
      "epoch": 1.785676671932596,
      "grad_norm": 0.29449769854545593,
      "learning_rate": 0.00017714066139233053,
      "loss": 0.19318415224552155,
      "mean_token_accuracy": 0.920307382941246,
      "num_tokens": 41830348.0,
      "step": 3391
    },
    {
      "entropy": 1.1941271126270294,
      "epoch": 1.7862032648762507,
      "grad_norm": 0.29486992955207825,
      "learning_rate": 0.0001771260565260168,
      "loss": 0.20498642325401306,
      "mean_token_accuracy": 0.9168984740972519,
      "num_tokens": 41842684.0,
      "step": 3392
    },
    {
      "entropy": 1.1674256920814514,
      "epoch": 1.7867298578199051,
      "grad_norm": 0.3044561743736267,
      "learning_rate": 0.00017711144767482034,
      "loss": 0.2078413963317871,
      "mean_token_accuracy": 0.9115557223558426,
      "num_tokens": 41855020.0,
      "step": 3393
    },
    {
      "entropy": 1.115517109632492,
      "epoch": 1.7872564507635598,
      "grad_norm": 0.2597251832485199,
      "learning_rate": 0.00017709683483960837,
      "loss": 0.18797630071640015,
      "mean_token_accuracy": 0.9255786091089249,
      "num_tokens": 41867356.0,
      "step": 3394
    },
    {
      "entropy": 1.1435833275318146,
      "epoch": 1.7877830437072144,
      "grad_norm": 0.26411309838294983,
      "learning_rate": 0.0001770822180212484,
      "loss": 0.1814870834350586,
      "mean_token_accuracy": 0.9257442653179169,
      "num_tokens": 41879692.0,
      "step": 3395
    },
    {
      "entropy": 1.1526050865650177,
      "epoch": 1.7883096366508688,
      "grad_norm": 0.27172785997390747,
      "learning_rate": 0.00017706759722060816,
      "loss": 0.19988085329532623,
      "mean_token_accuracy": 0.9178711026906967,
      "num_tokens": 41892028.0,
      "step": 3396
    },
    {
      "entropy": 1.1092619001865387,
      "epoch": 1.7888362295945235,
      "grad_norm": 0.26404666900634766,
      "learning_rate": 0.00017705297243855551,
      "loss": 0.18823406100273132,
      "mean_token_accuracy": 0.9220343977212906,
      "num_tokens": 41904364.0,
      "step": 3397
    },
    {
      "entropy": 1.1228880286216736,
      "epoch": 1.7893628225381781,
      "grad_norm": 0.2574518620967865,
      "learning_rate": 0.00017703834367595878,
      "loss": 0.18857592344284058,
      "mean_token_accuracy": 0.9215908646583557,
      "num_tokens": 41916700.0,
      "step": 3398
    },
    {
      "entropy": 1.1254870891571045,
      "epoch": 1.7898894154818326,
      "grad_norm": 0.27541834115982056,
      "learning_rate": 0.00017702371093368627,
      "loss": 0.18809503316879272,
      "mean_token_accuracy": 0.9209438264369965,
      "num_tokens": 41929036.0,
      "step": 3399
    },
    {
      "entropy": 1.075088083744049,
      "epoch": 1.790416008425487,
      "grad_norm": 0.25025203824043274,
      "learning_rate": 0.00017700907421260668,
      "loss": 0.1794566810131073,
      "mean_token_accuracy": 0.925526887178421,
      "num_tokens": 41941372.0,
      "step": 3400
    },
    {
      "entropy": 1.1571705043315887,
      "epoch": 1.7909426013691416,
      "grad_norm": 0.2782227396965027,
      "learning_rate": 0.0001769944335135889,
      "loss": 0.19144821166992188,
      "mean_token_accuracy": 0.9179984331130981,
      "num_tokens": 41953708.0,
      "step": 3401
    },
    {
      "entropy": 1.0587176382541656,
      "epoch": 1.7914691943127963,
      "grad_norm": 0.2607192397117615,
      "learning_rate": 0.00017697978883750212,
      "loss": 0.19511093199253082,
      "mean_token_accuracy": 0.9198868274688721,
      "num_tokens": 41966044.0,
      "step": 3402
    },
    {
      "entropy": 1.1079656779766083,
      "epoch": 1.7919957872564507,
      "grad_norm": 0.26396170258522034,
      "learning_rate": 0.00017696514018521563,
      "loss": 0.21819618344306946,
      "mean_token_accuracy": 0.910576269030571,
      "num_tokens": 41978380.0,
      "step": 3403
    },
    {
      "entropy": 1.0812654197216034,
      "epoch": 1.7925223802001053,
      "grad_norm": 0.2712666690349579,
      "learning_rate": 0.00017695048755759908,
      "loss": 0.20643654465675354,
      "mean_token_accuracy": 0.9118922054767609,
      "num_tokens": 41990716.0,
      "step": 3404
    },
    {
      "entropy": 1.0948403775691986,
      "epoch": 1.79304897314376,
      "grad_norm": 0.2753002345561981,
      "learning_rate": 0.0001769358309555223,
      "loss": 0.21245992183685303,
      "mean_token_accuracy": 0.9119813740253448,
      "num_tokens": 42003052.0,
      "step": 3405
    },
    {
      "entropy": 1.1134866774082184,
      "epoch": 1.7935755660874144,
      "grad_norm": 0.24785494804382324,
      "learning_rate": 0.00017692117037985538,
      "loss": 0.1743498146533966,
      "mean_token_accuracy": 0.9319243878126144,
      "num_tokens": 42015388.0,
      "step": 3406
    },
    {
      "entropy": 1.03944431245327,
      "epoch": 1.7941021590310688,
      "grad_norm": 0.25819504261016846,
      "learning_rate": 0.0001769065058314686,
      "loss": 0.2109576016664505,
      "mean_token_accuracy": 0.9150231778621674,
      "num_tokens": 42027724.0,
      "step": 3407
    },
    {
      "entropy": 1.079315334558487,
      "epoch": 1.7946287519747235,
      "grad_norm": 0.26192978024482727,
      "learning_rate": 0.00017689183731123257,
      "loss": 0.20323598384857178,
      "mean_token_accuracy": 0.9163467139005661,
      "num_tokens": 42040060.0,
      "step": 3408
    },
    {
      "entropy": 1.1146911978721619,
      "epoch": 1.7951553449183781,
      "grad_norm": 0.2505536377429962,
      "learning_rate": 0.00017687716482001797,
      "loss": 0.18577729165554047,
      "mean_token_accuracy": 0.9239400625228882,
      "num_tokens": 42052396.0,
      "step": 3409
    },
    {
      "entropy": 1.0605261623859406,
      "epoch": 1.7956819378620326,
      "grad_norm": 0.25102049112319946,
      "learning_rate": 0.00017686248835869595,
      "loss": 0.20101284980773926,
      "mean_token_accuracy": 0.9199343025684357,
      "num_tokens": 42064732.0,
      "step": 3410
    },
    {
      "entropy": 1.0440888553857803,
      "epoch": 1.7962085308056872,
      "grad_norm": 0.26397231221199036,
      "learning_rate": 0.00017684780792813764,
      "loss": 0.2093440592288971,
      "mean_token_accuracy": 0.9137531369924545,
      "num_tokens": 42077068.0,
      "step": 3411
    },
    {
      "entropy": 1.0457901358604431,
      "epoch": 1.7967351237493419,
      "grad_norm": 0.28600117564201355,
      "learning_rate": 0.00017683312352921463,
      "loss": 0.21590805053710938,
      "mean_token_accuracy": 0.9133930206298828,
      "num_tokens": 42089404.0,
      "step": 3412
    },
    {
      "entropy": 1.0765321105718613,
      "epoch": 1.7972617166929963,
      "grad_norm": 0.27367517352104187,
      "learning_rate": 0.00017681843516279862,
      "loss": 0.17508497834205627,
      "mean_token_accuracy": 0.9287613779306412,
      "num_tokens": 42101740.0,
      "step": 3413
    },
    {
      "entropy": 1.0435933470726013,
      "epoch": 1.7977883096366507,
      "grad_norm": 0.3039937913417816,
      "learning_rate": 0.00017680374282976152,
      "loss": 0.22740893065929413,
      "mean_token_accuracy": 0.9117265343666077,
      "num_tokens": 42114076.0,
      "step": 3414
    },
    {
      "entropy": 1.0765089988708496,
      "epoch": 1.7983149025803056,
      "grad_norm": 0.29611867666244507,
      "learning_rate": 0.00017678904653097558,
      "loss": 0.2006775140762329,
      "mean_token_accuracy": 0.917385458946228,
      "num_tokens": 42126412.0,
      "step": 3415
    },
    {
      "entropy": 1.0454992651939392,
      "epoch": 1.79884149552396,
      "grad_norm": 0.27161067724227905,
      "learning_rate": 0.00017677434626731323,
      "loss": 0.2067311704158783,
      "mean_token_accuracy": 0.918012872338295,
      "num_tokens": 42138748.0,
      "step": 3416
    },
    {
      "entropy": 1.063028246164322,
      "epoch": 1.7993680884676144,
      "grad_norm": 0.30537426471710205,
      "learning_rate": 0.00017675964203964716,
      "loss": 0.22716304659843445,
      "mean_token_accuracy": 0.911460742354393,
      "num_tokens": 42151084.0,
      "step": 3417
    },
    {
      "entropy": 1.020344227552414,
      "epoch": 1.799894681411269,
      "grad_norm": 0.2623611092567444,
      "learning_rate": 0.00017674493384885022,
      "loss": 0.17709854245185852,
      "mean_token_accuracy": 0.9269069582223892,
      "num_tokens": 42163420.0,
      "step": 3418
    },
    {
      "entropy": 1.056466594338417,
      "epoch": 1.8004212743549237,
      "grad_norm": 0.28182798624038696,
      "learning_rate": 0.0001767302216957956,
      "loss": 0.2123803049325943,
      "mean_token_accuracy": 0.919185072183609,
      "num_tokens": 42175756.0,
      "step": 3419
    },
    {
      "entropy": 1.020190790295601,
      "epoch": 1.8009478672985781,
      "grad_norm": 0.2913738489151001,
      "learning_rate": 0.00017671550558135662,
      "loss": 0.2207547426223755,
      "mean_token_accuracy": 0.9098331481218338,
      "num_tokens": 42188092.0,
      "step": 3420
    },
    {
      "entropy": 1.0069181025028229,
      "epoch": 1.8014744602422328,
      "grad_norm": 0.3414415717124939,
      "learning_rate": 0.00017670078550640692,
      "loss": 0.20877774059772491,
      "mean_token_accuracy": 0.9150732308626175,
      "num_tokens": 42200428.0,
      "step": 3421
    },
    {
      "entropy": 0.9878948777914047,
      "epoch": 1.8020010531858874,
      "grad_norm": 0.2625672221183777,
      "learning_rate": 0.00017668606147182037,
      "loss": 0.18287813663482666,
      "mean_token_accuracy": 0.9247158914804459,
      "num_tokens": 42212764.0,
      "step": 3422
    },
    {
      "entropy": 0.9836813509464264,
      "epoch": 1.8025276461295419,
      "grad_norm": 0.2774287462234497,
      "learning_rate": 0.000176671333478471,
      "loss": 0.20418068766593933,
      "mean_token_accuracy": 0.9173961877822876,
      "num_tokens": 42225100.0,
      "step": 3423
    },
    {
      "entropy": 0.9749710410833359,
      "epoch": 1.8030542390731963,
      "grad_norm": 0.2605920433998108,
      "learning_rate": 0.00017665660152723319,
      "loss": 0.19839239120483398,
      "mean_token_accuracy": 0.9227689206600189,
      "num_tokens": 42237436.0,
      "step": 3424
    },
    {
      "entropy": 0.9625023603439331,
      "epoch": 1.803580832016851,
      "grad_norm": 0.2580849230289459,
      "learning_rate": 0.00017664186561898145,
      "loss": 0.19285142421722412,
      "mean_token_accuracy": 0.9149420410394669,
      "num_tokens": 42249772.0,
      "step": 3425
    },
    {
      "entropy": 0.9385105073451996,
      "epoch": 1.8041074249605056,
      "grad_norm": 0.2744423747062683,
      "learning_rate": 0.0001766271257545905,
      "loss": 0.22436729073524475,
      "mean_token_accuracy": 0.9081188142299652,
      "num_tokens": 42262108.0,
      "step": 3426
    },
    {
      "entropy": 0.9655201435089111,
      "epoch": 1.80463401790416,
      "grad_norm": 0.25290924310684204,
      "learning_rate": 0.00017661238193493548,
      "loss": 0.19769905507564545,
      "mean_token_accuracy": 0.9191560298204422,
      "num_tokens": 42274444.0,
      "step": 3427
    },
    {
      "entropy": 0.9744076132774353,
      "epoch": 1.8051606108478147,
      "grad_norm": 0.36244046688079834,
      "learning_rate": 0.00017659763416089157,
      "loss": 0.20538192987442017,
      "mean_token_accuracy": 0.91567362844944,
      "num_tokens": 42286780.0,
      "step": 3428
    },
    {
      "entropy": 0.9812397658824921,
      "epoch": 1.8056872037914693,
      "grad_norm": 0.2958359718322754,
      "learning_rate": 0.00017658288243333433,
      "loss": 0.22860632836818695,
      "mean_token_accuracy": 0.9086659848690033,
      "num_tokens": 42299116.0,
      "step": 3429
    },
    {
      "entropy": 0.9486215561628342,
      "epoch": 1.8062137967351237,
      "grad_norm": 0.2708651125431061,
      "learning_rate": 0.00017656812675313936,
      "loss": 0.20294052362442017,
      "mean_token_accuracy": 0.9186287671327591,
      "num_tokens": 42311452.0,
      "step": 3430
    },
    {
      "entropy": 1.018494963645935,
      "epoch": 1.8067403896787781,
      "grad_norm": 0.2610625922679901,
      "learning_rate": 0.00017655336712118274,
      "loss": 0.18903633952140808,
      "mean_token_accuracy": 0.9225384891033173,
      "num_tokens": 42323788.0,
      "step": 3431
    },
    {
      "entropy": 0.9701050668954849,
      "epoch": 1.807266982622433,
      "grad_norm": 0.2695980966091156,
      "learning_rate": 0.0001765386035383406,
      "loss": 0.20392455160617828,
      "mean_token_accuracy": 0.913482204079628,
      "num_tokens": 42336124.0,
      "step": 3432
    },
    {
      "entropy": 0.9882481843233109,
      "epoch": 1.8077935755660874,
      "grad_norm": 0.27849382162094116,
      "learning_rate": 0.0001765238360054894,
      "loss": 0.2001408338546753,
      "mean_token_accuracy": 0.9178249388933182,
      "num_tokens": 42348460.0,
      "step": 3433
    },
    {
      "entropy": 0.9755127131938934,
      "epoch": 1.8083201685097419,
      "grad_norm": 0.2871023714542389,
      "learning_rate": 0.00017650906452350577,
      "loss": 0.21643081307411194,
      "mean_token_accuracy": 0.9128262996673584,
      "num_tokens": 42360796.0,
      "step": 3434
    },
    {
      "entropy": 1.0046819299459457,
      "epoch": 1.8088467614533965,
      "grad_norm": 0.26911982893943787,
      "learning_rate": 0.00017649428909326665,
      "loss": 0.20376306772232056,
      "mean_token_accuracy": 0.9150490611791611,
      "num_tokens": 42373132.0,
      "step": 3435
    },
    {
      "entropy": 1.020606204867363,
      "epoch": 1.8093733543970512,
      "grad_norm": 0.2887180745601654,
      "learning_rate": 0.00017647950971564914,
      "loss": 0.2110900580883026,
      "mean_token_accuracy": 0.9153491258621216,
      "num_tokens": 42385468.0,
      "step": 3436
    },
    {
      "entropy": 0.9779709279537201,
      "epoch": 1.8098999473407056,
      "grad_norm": 0.2773571014404297,
      "learning_rate": 0.00017646472639153057,
      "loss": 0.21756364405155182,
      "mean_token_accuracy": 0.915818601846695,
      "num_tokens": 42397804.0,
      "step": 3437
    },
    {
      "entropy": 1.0198432505130768,
      "epoch": 1.8104265402843602,
      "grad_norm": 0.25971168279647827,
      "learning_rate": 0.00017644993912178863,
      "loss": 0.19616732001304626,
      "mean_token_accuracy": 0.9186940789222717,
      "num_tokens": 42410140.0,
      "step": 3438
    },
    {
      "entropy": 1.0019106715917587,
      "epoch": 1.8109531332280149,
      "grad_norm": 0.27781882882118225,
      "learning_rate": 0.00017643514790730106,
      "loss": 0.22385036945343018,
      "mean_token_accuracy": 0.9076634049415588,
      "num_tokens": 42422476.0,
      "step": 3439
    },
    {
      "entropy": 1.0161989629268646,
      "epoch": 1.8114797261716693,
      "grad_norm": 0.2737201452255249,
      "learning_rate": 0.00017642035274894596,
      "loss": 0.20915377140045166,
      "mean_token_accuracy": 0.9130603224039078,
      "num_tokens": 42434812.0,
      "step": 3440
    },
    {
      "entropy": 1.0263062715530396,
      "epoch": 1.8120063191153237,
      "grad_norm": 0.2517765760421753,
      "learning_rate": 0.00017640555364760173,
      "loss": 0.18634292483329773,
      "mean_token_accuracy": 0.9243203103542328,
      "num_tokens": 42447148.0,
      "step": 3441
    },
    {
      "entropy": 1.0274717211723328,
      "epoch": 1.8125329120589784,
      "grad_norm": 0.27562084794044495,
      "learning_rate": 0.00017639075060414675,
      "loss": 0.2285502851009369,
      "mean_token_accuracy": 0.905607059597969,
      "num_tokens": 42459484.0,
      "step": 3442
    },
    {
      "entropy": 1.0182528346776962,
      "epoch": 1.813059505002633,
      "grad_norm": 0.2482491433620453,
      "learning_rate": 0.00017637594361945988,
      "loss": 0.20179533958435059,
      "mean_token_accuracy": 0.9154495447874069,
      "num_tokens": 42471820.0,
      "step": 3443
    },
    {
      "entropy": 1.0020128339529037,
      "epoch": 1.8135860979462874,
      "grad_norm": 0.237760528922081,
      "learning_rate": 0.00017636113269442008,
      "loss": 0.16691860556602478,
      "mean_token_accuracy": 0.9328106641769409,
      "num_tokens": 42484156.0,
      "step": 3444
    },
    {
      "entropy": 0.9634300619363785,
      "epoch": 1.814112690889942,
      "grad_norm": 0.25161245465278625,
      "learning_rate": 0.00017634631782990665,
      "loss": 0.19200566411018372,
      "mean_token_accuracy": 0.9238070696592331,
      "num_tokens": 42496492.0,
      "step": 3445
    },
    {
      "entropy": 0.9657048434019089,
      "epoch": 1.8146392838335967,
      "grad_norm": 0.2663264572620392,
      "learning_rate": 0.00017633149902679903,
      "loss": 0.19708439707756042,
      "mean_token_accuracy": 0.9215977638959885,
      "num_tokens": 42508828.0,
      "step": 3446
    },
    {
      "entropy": 1.0027496963739395,
      "epoch": 1.8151658767772512,
      "grad_norm": 0.257934033870697,
      "learning_rate": 0.0001763166762859769,
      "loss": 0.19249504804611206,
      "mean_token_accuracy": 0.918593168258667,
      "num_tokens": 42521164.0,
      "step": 3447
    },
    {
      "entropy": 0.9626820534467697,
      "epoch": 1.8156924697209056,
      "grad_norm": 0.246200829744339,
      "learning_rate": 0.0001763018496083202,
      "loss": 0.17840000987052917,
      "mean_token_accuracy": 0.9268343597650528,
      "num_tokens": 42533500.0,
      "step": 3448
    },
    {
      "entropy": 0.9367417544126511,
      "epoch": 1.8162190626645605,
      "grad_norm": 0.2481808215379715,
      "learning_rate": 0.0001762870189947092,
      "loss": 0.18670141696929932,
      "mean_token_accuracy": 0.9216200262308121,
      "num_tokens": 42545836.0,
      "step": 3449
    },
    {
      "entropy": 0.9394432306289673,
      "epoch": 1.8167456556082149,
      "grad_norm": 0.26846200227737427,
      "learning_rate": 0.00017627218444602418,
      "loss": 0.19002090394496918,
      "mean_token_accuracy": 0.9229268431663513,
      "num_tokens": 42558172.0,
      "step": 3450
    },
    {
      "entropy": 0.9051215350627899,
      "epoch": 1.8172722485518693,
      "grad_norm": 0.2518659830093384,
      "learning_rate": 0.0001762573459631458,
      "loss": 0.19580325484275818,
      "mean_token_accuracy": 0.9171916991472244,
      "num_tokens": 42570508.0,
      "step": 3451
    },
    {
      "entropy": 0.9344974011182785,
      "epoch": 1.817798841495524,
      "grad_norm": 0.28661391139030457,
      "learning_rate": 0.000176242503546955,
      "loss": 0.2089633345603943,
      "mean_token_accuracy": 0.9145671278238297,
      "num_tokens": 42582844.0,
      "step": 3452
    },
    {
      "entropy": 0.9151428788900375,
      "epoch": 1.8183254344391786,
      "grad_norm": 0.27580344676971436,
      "learning_rate": 0.0001762276571983329,
      "loss": 0.20280398428440094,
      "mean_token_accuracy": 0.9187836647033691,
      "num_tokens": 42595180.0,
      "step": 3453
    },
    {
      "entropy": 0.9216799587011337,
      "epoch": 1.818852027382833,
      "grad_norm": 0.3177351653575897,
      "learning_rate": 0.00017621280691816076,
      "loss": 0.22672337293624878,
      "mean_token_accuracy": 0.9106907099485397,
      "num_tokens": 42607516.0,
      "step": 3454
    },
    {
      "entropy": 0.9563778191804886,
      "epoch": 1.8193786203264877,
      "grad_norm": 0.2692190110683441,
      "learning_rate": 0.0001761979527073202,
      "loss": 0.18519625067710876,
      "mean_token_accuracy": 0.9241679459810257,
      "num_tokens": 42619852.0,
      "step": 3455
    },
    {
      "entropy": 0.9491466730833054,
      "epoch": 1.8199052132701423,
      "grad_norm": 0.30144041776657104,
      "learning_rate": 0.00017618309456669297,
      "loss": 0.20708760619163513,
      "mean_token_accuracy": 0.9165724217891693,
      "num_tokens": 42632188.0,
      "step": 3456
    },
    {
      "entropy": 0.9680854380130768,
      "epoch": 1.8204318062137967,
      "grad_norm": 0.29503685235977173,
      "learning_rate": 0.00017616823249716118,
      "loss": 0.20235475897789001,
      "mean_token_accuracy": 0.9136970639228821,
      "num_tokens": 42644524.0,
      "step": 3457
    },
    {
      "entropy": 0.9400119334459305,
      "epoch": 1.8209583991574512,
      "grad_norm": 0.29825398325920105,
      "learning_rate": 0.00017615336649960713,
      "loss": 0.20762497186660767,
      "mean_token_accuracy": 0.9167003929615021,
      "num_tokens": 42656860.0,
      "step": 3458
    },
    {
      "entropy": 0.9631544351577759,
      "epoch": 1.8214849921011058,
      "grad_norm": 0.28420135378837585,
      "learning_rate": 0.00017613849657491327,
      "loss": 0.21617071330547333,
      "mean_token_accuracy": 0.9141428619623184,
      "num_tokens": 42669196.0,
      "step": 3459
    },
    {
      "entropy": 0.9421249181032181,
      "epoch": 1.8220115850447605,
      "grad_norm": 0.2639506459236145,
      "learning_rate": 0.00017612362272396233,
      "loss": 0.19029353559017181,
      "mean_token_accuracy": 0.9208075702190399,
      "num_tokens": 42681532.0,
      "step": 3460
    },
    {
      "entropy": 0.9671123921871185,
      "epoch": 1.8225381779884149,
      "grad_norm": 0.26940131187438965,
      "learning_rate": 0.00017610874494763728,
      "loss": 0.19277174770832062,
      "mean_token_accuracy": 0.921785444021225,
      "num_tokens": 42693868.0,
      "step": 3461
    },
    {
      "entropy": 0.9888048470020294,
      "epoch": 1.8230647709320695,
      "grad_norm": 0.27126485109329224,
      "learning_rate": 0.0001760938632468214,
      "loss": 0.20671769976615906,
      "mean_token_accuracy": 0.918752908706665,
      "num_tokens": 42706204.0,
      "step": 3462
    },
    {
      "entropy": 1.0084779560565948,
      "epoch": 1.8235913638757242,
      "grad_norm": 0.2975345849990845,
      "learning_rate": 0.00017607897762239808,
      "loss": 0.2272503823041916,
      "mean_token_accuracy": 0.9058665335178375,
      "num_tokens": 42718540.0,
      "step": 3463
    },
    {
      "entropy": 1.0197101086378098,
      "epoch": 1.8241179568193786,
      "grad_norm": 0.28193390369415283,
      "learning_rate": 0.00017606408807525098,
      "loss": 0.20244908332824707,
      "mean_token_accuracy": 0.9168011546134949,
      "num_tokens": 42730876.0,
      "step": 3464
    },
    {
      "entropy": 1.0353141129016876,
      "epoch": 1.824644549763033,
      "grad_norm": 0.2784873843193054,
      "learning_rate": 0.00017604919460626403,
      "loss": 0.21005836129188538,
      "mean_token_accuracy": 0.9177538007497787,
      "num_tokens": 42743212.0,
      "step": 3465
    },
    {
      "entropy": 1.076245129108429,
      "epoch": 1.8251711427066877,
      "grad_norm": 0.27502164244651794,
      "learning_rate": 0.00017603429721632134,
      "loss": 0.1887526512145996,
      "mean_token_accuracy": 0.921863242983818,
      "num_tokens": 42755548.0,
      "step": 3466
    },
    {
      "entropy": 1.062199056148529,
      "epoch": 1.8256977356503423,
      "grad_norm": 0.276706337928772,
      "learning_rate": 0.00017601939590630733,
      "loss": 0.18946623802185059,
      "mean_token_accuracy": 0.921543076634407,
      "num_tokens": 42767884.0,
      "step": 3467
    },
    {
      "entropy": 1.0050623714923859,
      "epoch": 1.8262243285939967,
      "grad_norm": 0.39241406321525574,
      "learning_rate": 0.00017600449067710656,
      "loss": 0.20796050131320953,
      "mean_token_accuracy": 0.918712854385376,
      "num_tokens": 42780220.0,
      "step": 3468
    },
    {
      "entropy": 0.9754211008548737,
      "epoch": 1.8267509215376514,
      "grad_norm": 0.25771889090538025,
      "learning_rate": 0.0001759895815296039,
      "loss": 0.19497676193714142,
      "mean_token_accuracy": 0.9191407561302185,
      "num_tokens": 42792556.0,
      "step": 3469
    },
    {
      "entropy": 1.0407283902168274,
      "epoch": 1.827277514481306,
      "grad_norm": 0.29367595911026,
      "learning_rate": 0.00017597466846468437,
      "loss": 0.19909502565860748,
      "mean_token_accuracy": 0.9213258177042007,
      "num_tokens": 42804892.0,
      "step": 3470
    },
    {
      "entropy": 1.0311505794525146,
      "epoch": 1.8278041074249605,
      "grad_norm": 0.28731125593185425,
      "learning_rate": 0.00017595975148323333,
      "loss": 0.22046415507793427,
      "mean_token_accuracy": 0.9133670032024384,
      "num_tokens": 42817228.0,
      "step": 3471
    },
    {
      "entropy": 1.0354828983545303,
      "epoch": 1.8283307003686151,
      "grad_norm": 0.257552832365036,
      "learning_rate": 0.00017594483058613625,
      "loss": 0.19373366236686707,
      "mean_token_accuracy": 0.915319561958313,
      "num_tokens": 42829564.0,
      "step": 3472
    },
    {
      "entropy": 1.0075701475143433,
      "epoch": 1.8288572933122698,
      "grad_norm": 0.30350086092948914,
      "learning_rate": 0.000175929905774279,
      "loss": 0.21878236532211304,
      "mean_token_accuracy": 0.9075680375099182,
      "num_tokens": 42841900.0,
      "step": 3473
    },
    {
      "entropy": 1.0445679128170013,
      "epoch": 1.8293838862559242,
      "grad_norm": 0.26493775844573975,
      "learning_rate": 0.00017591497704854745,
      "loss": 0.19537273049354553,
      "mean_token_accuracy": 0.9180216193199158,
      "num_tokens": 42854236.0,
      "step": 3474
    },
    {
      "entropy": 1.0321323722600937,
      "epoch": 1.8299104791995786,
      "grad_norm": 0.25514036417007446,
      "learning_rate": 0.0001759000444098279,
      "loss": 0.19023671746253967,
      "mean_token_accuracy": 0.9226201772689819,
      "num_tokens": 42866572.0,
      "step": 3475
    },
    {
      "entropy": 1.0683675110340118,
      "epoch": 1.8304370721432333,
      "grad_norm": 0.24957162141799927,
      "learning_rate": 0.00017588510785900688,
      "loss": 0.19595561921596527,
      "mean_token_accuracy": 0.9198444336652756,
      "num_tokens": 42878908.0,
      "step": 3476
    },
    {
      "entropy": 1.0605555176734924,
      "epoch": 1.830963665086888,
      "grad_norm": 0.26935040950775146,
      "learning_rate": 0.00017587016739697094,
      "loss": 0.207814022898674,
      "mean_token_accuracy": 0.9161375761032104,
      "num_tokens": 42891244.0,
      "step": 3477
    },
    {
      "entropy": 0.979072317481041,
      "epoch": 1.8314902580305423,
      "grad_norm": 0.2588823437690735,
      "learning_rate": 0.00017585522302460707,
      "loss": 0.1919378638267517,
      "mean_token_accuracy": 0.9216379672288895,
      "num_tokens": 42903580.0,
      "step": 3478
    },
    {
      "entropy": 1.034737452864647,
      "epoch": 1.832016850974197,
      "grad_norm": 0.27618345618247986,
      "learning_rate": 0.00017584027474280248,
      "loss": 0.20401082932949066,
      "mean_token_accuracy": 0.9176820069551468,
      "num_tokens": 42915916.0,
      "step": 3479
    },
    {
      "entropy": 1.0233630985021591,
      "epoch": 1.8325434439178516,
      "grad_norm": 0.29381322860717773,
      "learning_rate": 0.00017582532255244456,
      "loss": 0.20440031588077545,
      "mean_token_accuracy": 0.9106693267822266,
      "num_tokens": 42928252.0,
      "step": 3480
    },
    {
      "entropy": 1.0043966621160507,
      "epoch": 1.833070036861506,
      "grad_norm": 0.28479161858558655,
      "learning_rate": 0.00017581036645442083,
      "loss": 0.22450262308120728,
      "mean_token_accuracy": 0.9132276177406311,
      "num_tokens": 42940588.0,
      "step": 3481
    },
    {
      "entropy": 1.0819280743598938,
      "epoch": 1.8335966298051605,
      "grad_norm": 0.2651804983615875,
      "learning_rate": 0.00017579540644961926,
      "loss": 0.2066086381673813,
      "mean_token_accuracy": 0.9166118055582047,
      "num_tokens": 42952924.0,
      "step": 3482
    },
    {
      "entropy": 1.0218794494867325,
      "epoch": 1.8341232227488151,
      "grad_norm": 0.26943325996398926,
      "learning_rate": 0.00017578044253892787,
      "loss": 0.2052810937166214,
      "mean_token_accuracy": 0.918945848941803,
      "num_tokens": 42965260.0,
      "step": 3483
    },
    {
      "entropy": 1.075775384902954,
      "epoch": 1.8346498156924698,
      "grad_norm": 0.28176063299179077,
      "learning_rate": 0.00017576547472323501,
      "loss": 0.20732028782367706,
      "mean_token_accuracy": 0.91796013712883,
      "num_tokens": 42977596.0,
      "step": 3484
    },
    {
      "entropy": 0.9888935834169388,
      "epoch": 1.8351764086361242,
      "grad_norm": 0.24343425035476685,
      "learning_rate": 0.0001757505030034292,
      "loss": 0.1734849065542221,
      "mean_token_accuracy": 0.9293620884418488,
      "num_tokens": 42989932.0,
      "step": 3485
    },
    {
      "entropy": 0.9950584769248962,
      "epoch": 1.8357030015797788,
      "grad_norm": 0.23256437480449677,
      "learning_rate": 0.0001757355273803993,
      "loss": 0.18044212460517883,
      "mean_token_accuracy": 0.927759513258934,
      "num_tokens": 43002268.0,
      "step": 3486
    },
    {
      "entropy": 1.018806740641594,
      "epoch": 1.8362295945234335,
      "grad_norm": 0.28248098492622375,
      "learning_rate": 0.00017572054785503424,
      "loss": 0.22589240968227386,
      "mean_token_accuracy": 0.9108239859342575,
      "num_tokens": 43014604.0,
      "step": 3487
    },
    {
      "entropy": 0.9609373807907104,
      "epoch": 1.836756187467088,
      "grad_norm": 0.24391081929206848,
      "learning_rate": 0.00017570556442822333,
      "loss": 0.18671506643295288,
      "mean_token_accuracy": 0.9225711077451706,
      "num_tokens": 43026940.0,
      "step": 3488
    },
    {
      "entropy": 0.988832950592041,
      "epoch": 1.8372827804107423,
      "grad_norm": 0.2958183288574219,
      "learning_rate": 0.000175690577100856,
      "loss": 0.21036623418331146,
      "mean_token_accuracy": 0.9159839749336243,
      "num_tokens": 43039276.0,
      "step": 3489
    },
    {
      "entropy": 0.9858183264732361,
      "epoch": 1.8378093733543972,
      "grad_norm": 0.26455992460250854,
      "learning_rate": 0.00017567558587382198,
      "loss": 0.19656260311603546,
      "mean_token_accuracy": 0.9173092693090439,
      "num_tokens": 43051612.0,
      "step": 3490
    },
    {
      "entropy": 0.9774465411901474,
      "epoch": 1.8383359662980516,
      "grad_norm": 0.24327997863292694,
      "learning_rate": 0.00017566059074801123,
      "loss": 0.169850692152977,
      "mean_token_accuracy": 0.9312596470117569,
      "num_tokens": 43063948.0,
      "step": 3491
    },
    {
      "entropy": 0.96585513651371,
      "epoch": 1.838862559241706,
      "grad_norm": 0.26748207211494446,
      "learning_rate": 0.0001756455917243139,
      "loss": 0.19751040637493134,
      "mean_token_accuracy": 0.9155432730913162,
      "num_tokens": 43076284.0,
      "step": 3492
    },
    {
      "entropy": 0.985474169254303,
      "epoch": 1.8393891521853607,
      "grad_norm": 0.2670723497867584,
      "learning_rate": 0.0001756305888036204,
      "loss": 0.1883074939250946,
      "mean_token_accuracy": 0.9221811294555664,
      "num_tokens": 43088620.0,
      "step": 3493
    },
    {
      "entropy": 0.97892165184021,
      "epoch": 1.8399157451290153,
      "grad_norm": 0.27449530363082886,
      "learning_rate": 0.00017561558198682134,
      "loss": 0.2017425000667572,
      "mean_token_accuracy": 0.9193790405988693,
      "num_tokens": 43100956.0,
      "step": 3494
    },
    {
      "entropy": 0.92783522605896,
      "epoch": 1.8404423380726698,
      "grad_norm": 0.2700694501399994,
      "learning_rate": 0.00017560057127480764,
      "loss": 0.19871242344379425,
      "mean_token_accuracy": 0.9196399599313736,
      "num_tokens": 43113292.0,
      "step": 3495
    },
    {
      "entropy": 0.9279954582452774,
      "epoch": 1.8409689310163244,
      "grad_norm": 0.26998549699783325,
      "learning_rate": 0.00017558555666847037,
      "loss": 0.17693020403385162,
      "mean_token_accuracy": 0.9300927370786667,
      "num_tokens": 43125628.0,
      "step": 3496
    },
    {
      "entropy": 0.9917383641004562,
      "epoch": 1.841495523959979,
      "grad_norm": 0.31583210825920105,
      "learning_rate": 0.00017557053816870088,
      "loss": 0.227731853723526,
      "mean_token_accuracy": 0.9075176864862442,
      "num_tokens": 43137964.0,
      "step": 3497
    },
    {
      "entropy": 0.9914578646421432,
      "epoch": 1.8420221169036335,
      "grad_norm": 0.29741397500038147,
      "learning_rate": 0.00017555551577639068,
      "loss": 0.21265819668769836,
      "mean_token_accuracy": 0.9156584292650223,
      "num_tokens": 43150300.0,
      "step": 3498
    },
    {
      "entropy": 0.9868685007095337,
      "epoch": 1.842548709847288,
      "grad_norm": 0.3175033628940582,
      "learning_rate": 0.00017554048949243163,
      "loss": 0.20174245536327362,
      "mean_token_accuracy": 0.9204602241516113,
      "num_tokens": 43162636.0,
      "step": 3499
    },
    {
      "entropy": 0.9585036486387253,
      "epoch": 1.8430753027909426,
      "grad_norm": 0.2692413032054901,
      "learning_rate": 0.0001755254593177157,
      "loss": 0.1967068463563919,
      "mean_token_accuracy": 0.9194019138813019,
      "num_tokens": 43174972.0,
      "step": 3500
    },
    {
      "entropy": 1.0175974816083908,
      "epoch": 1.8436018957345972,
      "grad_norm": 0.2601134777069092,
      "learning_rate": 0.00017551042525313515,
      "loss": 0.20446570217609406,
      "mean_token_accuracy": 0.9177961945533752,
      "num_tokens": 43187308.0,
      "step": 3501
    },
    {
      "entropy": 0.9750374257564545,
      "epoch": 1.8441284886782516,
      "grad_norm": 0.2664104998111725,
      "learning_rate": 0.00017549538729958247,
      "loss": 0.20090380311012268,
      "mean_token_accuracy": 0.9167754501104355,
      "num_tokens": 43199644.0,
      "step": 3502
    },
    {
      "entropy": 1.0277975350618362,
      "epoch": 1.8446550816219063,
      "grad_norm": 0.26602786779403687,
      "learning_rate": 0.0001754803454579504,
      "loss": 0.21057289838790894,
      "mean_token_accuracy": 0.9140521287918091,
      "num_tokens": 43211980.0,
      "step": 3503
    },
    {
      "entropy": 0.9721402525901794,
      "epoch": 1.845181674565561,
      "grad_norm": 0.2602728605270386,
      "learning_rate": 0.00017546529972913187,
      "loss": 0.20174983143806458,
      "mean_token_accuracy": 0.9159922003746033,
      "num_tokens": 43224316.0,
      "step": 3504
    },
    {
      "entropy": 1.054141342639923,
      "epoch": 1.8457082675092154,
      "grad_norm": 0.2536306381225586,
      "learning_rate": 0.00017545025011402005,
      "loss": 0.19397109746932983,
      "mean_token_accuracy": 0.9237479120492935,
      "num_tokens": 43236652.0,
      "step": 3505
    },
    {
      "entropy": 1.041559875011444,
      "epoch": 1.8462348604528698,
      "grad_norm": 0.2991665303707123,
      "learning_rate": 0.00017543519661350837,
      "loss": 0.20816954970359802,
      "mean_token_accuracy": 0.9192373007535934,
      "num_tokens": 43248988.0,
      "step": 3506
    },
    {
      "entropy": 1.058956652879715,
      "epoch": 1.8467614533965246,
      "grad_norm": 0.29321545362472534,
      "learning_rate": 0.00017542013922849044,
      "loss": 0.2185562252998352,
      "mean_token_accuracy": 0.9086525142192841,
      "num_tokens": 43261324.0,
      "step": 3507
    },
    {
      "entropy": 1.0227807611227036,
      "epoch": 1.847288046340179,
      "grad_norm": 0.2425878793001175,
      "learning_rate": 0.00017540507795986014,
      "loss": 0.18261486291885376,
      "mean_token_accuracy": 0.9194140434265137,
      "num_tokens": 43273660.0,
      "step": 3508
    },
    {
      "entropy": 1.0270606130361557,
      "epoch": 1.8478146392838335,
      "grad_norm": 0.25291311740875244,
      "learning_rate": 0.00017539001280851156,
      "loss": 0.19907575845718384,
      "mean_token_accuracy": 0.9177414625883102,
      "num_tokens": 43285996.0,
      "step": 3509
    },
    {
      "entropy": 1.0220745652914047,
      "epoch": 1.8483412322274881,
      "grad_norm": 0.23823751509189606,
      "learning_rate": 0.00017537494377533908,
      "loss": 0.18174749612808228,
      "mean_token_accuracy": 0.9222448915243149,
      "num_tokens": 43298332.0,
      "step": 3510
    },
    {
      "entropy": 1.0170291811227798,
      "epoch": 1.8488678251711428,
      "grad_norm": 0.28503599762916565,
      "learning_rate": 0.00017535987086123717,
      "loss": 0.20761212706565857,
      "mean_token_accuracy": 0.9194875508546829,
      "num_tokens": 43310668.0,
      "step": 3511
    },
    {
      "entropy": 0.9842036962509155,
      "epoch": 1.8493944181147972,
      "grad_norm": 0.2714064121246338,
      "learning_rate": 0.00017534479406710072,
      "loss": 0.19920268654823303,
      "mean_token_accuracy": 0.9208008050918579,
      "num_tokens": 43323004.0,
      "step": 3512
    },
    {
      "entropy": 0.9929454177618027,
      "epoch": 1.8499210110584519,
      "grad_norm": 0.30067718029022217,
      "learning_rate": 0.00017532971339382467,
      "loss": 0.2108234167098999,
      "mean_token_accuracy": 0.9151533842086792,
      "num_tokens": 43335340.0,
      "step": 3513
    },
    {
      "entropy": 0.9094464778900146,
      "epoch": 1.8504476040021065,
      "grad_norm": 0.269818514585495,
      "learning_rate": 0.0001753146288423043,
      "loss": 0.19665856659412384,
      "mean_token_accuracy": 0.9184691607952118,
      "num_tokens": 43347676.0,
      "step": 3514
    },
    {
      "entropy": 0.9886156469583511,
      "epoch": 1.850974196945761,
      "grad_norm": 0.28764402866363525,
      "learning_rate": 0.00017529954041343516,
      "loss": 0.20603036880493164,
      "mean_token_accuracy": 0.920213133096695,
      "num_tokens": 43360012.0,
      "step": 3515
    },
    {
      "entropy": 0.9759783744812012,
      "epoch": 1.8515007898894154,
      "grad_norm": 0.2881333529949188,
      "learning_rate": 0.00017528444810811284,
      "loss": 0.19697514176368713,
      "mean_token_accuracy": 0.9233171492815018,
      "num_tokens": 43372348.0,
      "step": 3516
    },
    {
      "entropy": 0.956908330321312,
      "epoch": 1.85202738283307,
      "grad_norm": 0.2677208185195923,
      "learning_rate": 0.00017526935192723335,
      "loss": 0.19714459776878357,
      "mean_token_accuracy": 0.9196808934211731,
      "num_tokens": 43384684.0,
      "step": 3517
    },
    {
      "entropy": 0.9836724400520325,
      "epoch": 1.8525539757767246,
      "grad_norm": 0.30804702639579773,
      "learning_rate": 0.00017525425187169287,
      "loss": 0.21788114309310913,
      "mean_token_accuracy": 0.9175039529800415,
      "num_tokens": 43397020.0,
      "step": 3518
    },
    {
      "entropy": 1.002650871872902,
      "epoch": 1.853080568720379,
      "grad_norm": 0.2799338102340698,
      "learning_rate": 0.00017523914794238777,
      "loss": 0.1947844922542572,
      "mean_token_accuracy": 0.9189659506082535,
      "num_tokens": 43409356.0,
      "step": 3519
    },
    {
      "entropy": 0.9216767400503159,
      "epoch": 1.8536071616640337,
      "grad_norm": 0.24135830998420715,
      "learning_rate": 0.00017522404014021472,
      "loss": 0.18965280055999756,
      "mean_token_accuracy": 0.9230446666479111,
      "num_tokens": 43421692.0,
      "step": 3520
    },
    {
      "entropy": 0.9524090141057968,
      "epoch": 1.8541337546076884,
      "grad_norm": 0.26251375675201416,
      "learning_rate": 0.00017520892846607057,
      "loss": 0.20700010657310486,
      "mean_token_accuracy": 0.9134894460439682,
      "num_tokens": 43434028.0,
      "step": 3521
    },
    {
      "entropy": 1.011262759566307,
      "epoch": 1.8546603475513428,
      "grad_norm": 0.30192479491233826,
      "learning_rate": 0.00017519381292085238,
      "loss": 0.22291095554828644,
      "mean_token_accuracy": 0.9098920524120331,
      "num_tokens": 43446364.0,
      "step": 3522
    },
    {
      "entropy": 0.9867189973592758,
      "epoch": 1.8551869404949972,
      "grad_norm": 0.2610762119293213,
      "learning_rate": 0.00017517869350545747,
      "loss": 0.20300611853599548,
      "mean_token_accuracy": 0.918722614645958,
      "num_tokens": 43458700.0,
      "step": 3523
    },
    {
      "entropy": 0.9844164401292801,
      "epoch": 1.855713533438652,
      "grad_norm": 0.2718338072299957,
      "learning_rate": 0.00017516357022078348,
      "loss": 0.21439701318740845,
      "mean_token_accuracy": 0.9139817208051682,
      "num_tokens": 43471036.0,
      "step": 3524
    },
    {
      "entropy": 1.026760846376419,
      "epoch": 1.8562401263823065,
      "grad_norm": 0.28833097219467163,
      "learning_rate": 0.0001751484430677281,
      "loss": 0.1966688632965088,
      "mean_token_accuracy": 0.9210084974765778,
      "num_tokens": 43483372.0,
      "step": 3525
    },
    {
      "entropy": 0.9831033647060394,
      "epoch": 1.856766719325961,
      "grad_norm": 0.24818982183933258,
      "learning_rate": 0.00017513331204718934,
      "loss": 0.18625837564468384,
      "mean_token_accuracy": 0.9207738190889359,
      "num_tokens": 43495708.0,
      "step": 3526
    },
    {
      "entropy": 1.0132111757993698,
      "epoch": 1.8572933122696156,
      "grad_norm": 0.24483777582645416,
      "learning_rate": 0.00017511817716006552,
      "loss": 0.18312522768974304,
      "mean_token_accuracy": 0.9218669086694717,
      "num_tokens": 43508044.0,
      "step": 3527
    },
    {
      "entropy": 1.0282989740371704,
      "epoch": 1.8578199052132702,
      "grad_norm": 0.2627302408218384,
      "learning_rate": 0.00017510303840725504,
      "loss": 0.19312997162342072,
      "mean_token_accuracy": 0.9195411801338196,
      "num_tokens": 43520380.0,
      "step": 3528
    },
    {
      "entropy": 1.0256311148405075,
      "epoch": 1.8583464981569247,
      "grad_norm": 0.2712998390197754,
      "learning_rate": 0.0001750878957896566,
      "loss": 0.20464766025543213,
      "mean_token_accuracy": 0.9199627637863159,
      "num_tokens": 43532716.0,
      "step": 3529
    },
    {
      "entropy": 1.0233651250600815,
      "epoch": 1.8588730911005793,
      "grad_norm": 0.25352466106414795,
      "learning_rate": 0.00017507274930816916,
      "loss": 0.19999416172504425,
      "mean_token_accuracy": 0.9188231825828552,
      "num_tokens": 43545052.0,
      "step": 3530
    },
    {
      "entropy": 1.0036712139844894,
      "epoch": 1.859399684044234,
      "grad_norm": 0.25642892718315125,
      "learning_rate": 0.00017505759896369188,
      "loss": 0.17751988768577576,
      "mean_token_accuracy": 0.9266696274280548,
      "num_tokens": 43557388.0,
      "step": 3531
    },
    {
      "entropy": 1.0442322492599487,
      "epoch": 1.8599262769878884,
      "grad_norm": 0.27309077978134155,
      "learning_rate": 0.0001750424447571241,
      "loss": 0.20416924357414246,
      "mean_token_accuracy": 0.9193257242441177,
      "num_tokens": 43569724.0,
      "step": 3532
    },
    {
      "entropy": 1.0406121015548706,
      "epoch": 1.8604528699315428,
      "grad_norm": 0.2676863372325897,
      "learning_rate": 0.0001750272866893655,
      "loss": 0.2130739539861679,
      "mean_token_accuracy": 0.9103453308343887,
      "num_tokens": 43582060.0,
      "step": 3533
    },
    {
      "entropy": 1.0540429800748825,
      "epoch": 1.8609794628751974,
      "grad_norm": 0.26795145869255066,
      "learning_rate": 0.00017501212476131589,
      "loss": 0.19300617277622223,
      "mean_token_accuracy": 0.9267665147781372,
      "num_tokens": 43594396.0,
      "step": 3534
    },
    {
      "entropy": 1.0778743624687195,
      "epoch": 1.861506055818852,
      "grad_norm": 0.2814680337905884,
      "learning_rate": 0.00017499695897387534,
      "loss": 0.21657180786132812,
      "mean_token_accuracy": 0.913690060377121,
      "num_tokens": 43606732.0,
      "step": 3535
    },
    {
      "entropy": 1.0144944339990616,
      "epoch": 1.8620326487625065,
      "grad_norm": 0.25453317165374756,
      "learning_rate": 0.00017498178932794418,
      "loss": 0.19731523096561432,
      "mean_token_accuracy": 0.9211816489696503,
      "num_tokens": 43619068.0,
      "step": 3536
    },
    {
      "entropy": 1.0487922728061676,
      "epoch": 1.8625592417061612,
      "grad_norm": 0.29628288745880127,
      "learning_rate": 0.00017496661582442293,
      "loss": 0.20683038234710693,
      "mean_token_accuracy": 0.9144911020994186,
      "num_tokens": 43631404.0,
      "step": 3537
    },
    {
      "entropy": 1.0728088915348053,
      "epoch": 1.8630858346498158,
      "grad_norm": 0.2915079593658447,
      "learning_rate": 0.00017495143846421235,
      "loss": 0.20339436829090118,
      "mean_token_accuracy": 0.920454278588295,
      "num_tokens": 43643740.0,
      "step": 3538
    },
    {
      "entropy": 1.04716295003891,
      "epoch": 1.8636124275934702,
      "grad_norm": 0.23968999087810516,
      "learning_rate": 0.00017493625724821345,
      "loss": 0.17932488024234772,
      "mean_token_accuracy": 0.9295078665018082,
      "num_tokens": 43656076.0,
      "step": 3539
    },
    {
      "entropy": 1.1039165556430817,
      "epoch": 1.8641390205371247,
      "grad_norm": 0.2852647006511688,
      "learning_rate": 0.0001749210721773274,
      "loss": 0.20938575267791748,
      "mean_token_accuracy": 0.9154763072729111,
      "num_tokens": 43668412.0,
      "step": 3540
    },
    {
      "entropy": 1.062223643064499,
      "epoch": 1.8646656134807793,
      "grad_norm": 0.2730938494205475,
      "learning_rate": 0.00017490588325245573,
      "loss": 0.20177365839481354,
      "mean_token_accuracy": 0.920450896024704,
      "num_tokens": 43680748.0,
      "step": 3541
    },
    {
      "entropy": 1.0817009508609772,
      "epoch": 1.865192206424434,
      "grad_norm": 0.25735530257225037,
      "learning_rate": 0.00017489069047450003,
      "loss": 0.1912008374929428,
      "mean_token_accuracy": 0.9246111661195755,
      "num_tokens": 43693084.0,
      "step": 3542
    },
    {
      "entropy": 1.0735929310321808,
      "epoch": 1.8657187993680884,
      "grad_norm": 0.26666948199272156,
      "learning_rate": 0.00017487549384436228,
      "loss": 0.19975747168064117,
      "mean_token_accuracy": 0.9182255566120148,
      "num_tokens": 43705420.0,
      "step": 3543
    },
    {
      "entropy": 1.0770608186721802,
      "epoch": 1.866245392311743,
      "grad_norm": 0.2631293833255768,
      "learning_rate": 0.00017486029336294455,
      "loss": 0.21597373485565186,
      "mean_token_accuracy": 0.9157756865024567,
      "num_tokens": 43717756.0,
      "step": 3544
    },
    {
      "entropy": 1.0684048980474472,
      "epoch": 1.8667719852553977,
      "grad_norm": 0.2667410969734192,
      "learning_rate": 0.00017484508903114926,
      "loss": 0.19906820356845856,
      "mean_token_accuracy": 0.9196314960718155,
      "num_tokens": 43730092.0,
      "step": 3545
    },
    {
      "entropy": 1.1074407696723938,
      "epoch": 1.867298578199052,
      "grad_norm": 0.2910633981227875,
      "learning_rate": 0.00017482988084987897,
      "loss": 0.20462411642074585,
      "mean_token_accuracy": 0.9201758056879044,
      "num_tokens": 43742428.0,
      "step": 3546
    },
    {
      "entropy": 1.1205481886863708,
      "epoch": 1.8678251711427067,
      "grad_norm": 0.2661513388156891,
      "learning_rate": 0.00017481466882003651,
      "loss": 0.19449472427368164,
      "mean_token_accuracy": 0.9201523959636688,
      "num_tokens": 43754764.0,
      "step": 3547
    },
    {
      "entropy": 1.058198630809784,
      "epoch": 1.8683517640863614,
      "grad_norm": 0.26210594177246094,
      "learning_rate": 0.00017479945294252496,
      "loss": 0.21112874150276184,
      "mean_token_accuracy": 0.913567453622818,
      "num_tokens": 43767100.0,
      "step": 3548
    },
    {
      "entropy": 1.0497896373271942,
      "epoch": 1.8688783570300158,
      "grad_norm": 0.24254575371742249,
      "learning_rate": 0.00017478423321824752,
      "loss": 0.17959487438201904,
      "mean_token_accuracy": 0.9248331785202026,
      "num_tokens": 43779436.0,
      "step": 3549
    },
    {
      "entropy": 1.0271850526332855,
      "epoch": 1.8694049499736702,
      "grad_norm": 0.25798940658569336,
      "learning_rate": 0.00017476900964810777,
      "loss": 0.20330272614955902,
      "mean_token_accuracy": 0.9145640432834625,
      "num_tokens": 43791772.0,
      "step": 3550
    },
    {
      "entropy": 1.0344137400388718,
      "epoch": 1.8699315429173249,
      "grad_norm": 0.2523990869522095,
      "learning_rate": 0.00017475378223300944,
      "loss": 0.18353012204170227,
      "mean_token_accuracy": 0.9226906001567841,
      "num_tokens": 43804108.0,
      "step": 3551
    },
    {
      "entropy": 1.0250817984342575,
      "epoch": 1.8704581358609795,
      "grad_norm": 0.2655175030231476,
      "learning_rate": 0.00017473855097385646,
      "loss": 0.1963675618171692,
      "mean_token_accuracy": 0.9198670238256454,
      "num_tokens": 43816444.0,
      "step": 3552
    },
    {
      "entropy": 1.0592041909694672,
      "epoch": 1.870984728804634,
      "grad_norm": 0.2882557511329651,
      "learning_rate": 0.000174723315871553,
      "loss": 0.222836434841156,
      "mean_token_accuracy": 0.913982018828392,
      "num_tokens": 43828780.0,
      "step": 3553
    },
    {
      "entropy": 1.0100958496332169,
      "epoch": 1.8715113217482886,
      "grad_norm": 0.2879355847835541,
      "learning_rate": 0.00017470807692700354,
      "loss": 0.1921842247247696,
      "mean_token_accuracy": 0.9170998483896255,
      "num_tokens": 43841116.0,
      "step": 3554
    },
    {
      "entropy": 1.0180272459983826,
      "epoch": 1.8720379146919433,
      "grad_norm": 0.26359114050865173,
      "learning_rate": 0.0001746928341411127,
      "loss": 0.20603729784488678,
      "mean_token_accuracy": 0.9175525456666946,
      "num_tokens": 43853452.0,
      "step": 3555
    },
    {
      "entropy": 1.07178395986557,
      "epoch": 1.8725645076355977,
      "grad_norm": 0.27470099925994873,
      "learning_rate": 0.00017467758751478537,
      "loss": 0.20463289320468903,
      "mean_token_accuracy": 0.914940133690834,
      "num_tokens": 43865788.0,
      "step": 3556
    },
    {
      "entropy": 0.9943696707487106,
      "epoch": 1.873091100579252,
      "grad_norm": 0.24466237425804138,
      "learning_rate": 0.0001746623370489266,
      "loss": 0.17426809668540955,
      "mean_token_accuracy": 0.9257047772407532,
      "num_tokens": 43878124.0,
      "step": 3557
    },
    {
      "entropy": 1.0398575961589813,
      "epoch": 1.8736176935229067,
      "grad_norm": 0.2756820619106293,
      "learning_rate": 0.0001746470827444418,
      "loss": 0.21039924025535583,
      "mean_token_accuracy": 0.9163792431354523,
      "num_tokens": 43890460.0,
      "step": 3558
    },
    {
      "entropy": 1.113322228193283,
      "epoch": 1.8741442864665614,
      "grad_norm": 0.29872143268585205,
      "learning_rate": 0.00017463182460223647,
      "loss": 0.19940727949142456,
      "mean_token_accuracy": 0.9192183166742325,
      "num_tokens": 43902796.0,
      "step": 3559
    },
    {
      "entropy": 1.1167441308498383,
      "epoch": 1.8746708794102158,
      "grad_norm": 0.2722742557525635,
      "learning_rate": 0.00017461656262321643,
      "loss": 0.187882661819458,
      "mean_token_accuracy": 0.9232105165719986,
      "num_tokens": 43915132.0,
      "step": 3560
    },
    {
      "entropy": 1.0867882072925568,
      "epoch": 1.8751974723538705,
      "grad_norm": 0.2816410958766937,
      "learning_rate": 0.00017460129680828767,
      "loss": 0.20031344890594482,
      "mean_token_accuracy": 0.9189879596233368,
      "num_tokens": 43927468.0,
      "step": 3561
    },
    {
      "entropy": 1.1575427651405334,
      "epoch": 1.8757240652975251,
      "grad_norm": 0.29825645685195923,
      "learning_rate": 0.00017458602715835644,
      "loss": 0.2101304531097412,
      "mean_token_accuracy": 0.9124222993850708,
      "num_tokens": 43939804.0,
      "step": 3562
    },
    {
      "entropy": 1.0926944315433502,
      "epoch": 1.8762506582411795,
      "grad_norm": 0.2628743350505829,
      "learning_rate": 0.00017457075367432922,
      "loss": 0.19311033189296722,
      "mean_token_accuracy": 0.922664999961853,
      "num_tokens": 43952140.0,
      "step": 3563
    },
    {
      "entropy": 1.0990844368934631,
      "epoch": 1.876777251184834,
      "grad_norm": 0.27810606360435486,
      "learning_rate": 0.0001745554763571127,
      "loss": 0.19149093329906464,
      "mean_token_accuracy": 0.9224977344274521,
      "num_tokens": 43964476.0,
      "step": 3564
    },
    {
      "entropy": 1.0753406584262848,
      "epoch": 1.8773038441284888,
      "grad_norm": 0.2753046452999115,
      "learning_rate": 0.0001745401952076138,
      "loss": 0.20910421013832092,
      "mean_token_accuracy": 0.9132356494665146,
      "num_tokens": 43976812.0,
      "step": 3565
    },
    {
      "entropy": 1.114586502313614,
      "epoch": 1.8778304370721433,
      "grad_norm": 0.2780378758907318,
      "learning_rate": 0.00017452491022673967,
      "loss": 0.20378395915031433,
      "mean_token_accuracy": 0.9182945340871811,
      "num_tokens": 43989148.0,
      "step": 3566
    },
    {
      "entropy": 1.1071798503398895,
      "epoch": 1.8783570300157977,
      "grad_norm": 0.2734641432762146,
      "learning_rate": 0.00017450962141539773,
      "loss": 0.2131178081035614,
      "mean_token_accuracy": 0.9139434695243835,
      "num_tokens": 44001484.0,
      "step": 3567
    },
    {
      "entropy": 1.0694519877433777,
      "epoch": 1.8788836229594523,
      "grad_norm": 0.26163414120674133,
      "learning_rate": 0.00017449432877449553,
      "loss": 0.19922369718551636,
      "mean_token_accuracy": 0.9218622297048569,
      "num_tokens": 44013820.0,
      "step": 3568
    },
    {
      "entropy": 1.0787781774997711,
      "epoch": 1.879410215903107,
      "grad_norm": 0.2676631212234497,
      "learning_rate": 0.00017447903230494096,
      "loss": 0.18829788267612457,
      "mean_token_accuracy": 0.9208130836486816,
      "num_tokens": 44026156.0,
      "step": 3569
    },
    {
      "entropy": 1.106066107749939,
      "epoch": 1.8799368088467614,
      "grad_norm": 0.26856163144111633,
      "learning_rate": 0.00017446373200764203,
      "loss": 0.1804138720035553,
      "mean_token_accuracy": 0.919698104262352,
      "num_tokens": 44038492.0,
      "step": 3570
    },
    {
      "entropy": 1.141218662261963,
      "epoch": 1.880463401790416,
      "grad_norm": 0.2687745988368988,
      "learning_rate": 0.00017444842788350706,
      "loss": 0.1945033073425293,
      "mean_token_accuracy": 0.9200787246227264,
      "num_tokens": 44050828.0,
      "step": 3571
    },
    {
      "entropy": 1.1304888129234314,
      "epoch": 1.8809899947340707,
      "grad_norm": 0.25851964950561523,
      "learning_rate": 0.00017443311993344457,
      "loss": 0.19336003065109253,
      "mean_token_accuracy": 0.9188802093267441,
      "num_tokens": 44063164.0,
      "step": 3572
    },
    {
      "entropy": 1.1865220665931702,
      "epoch": 1.8815165876777251,
      "grad_norm": 0.29074305295944214,
      "learning_rate": 0.00017441780815836326,
      "loss": 0.1957467496395111,
      "mean_token_accuracy": 0.9202824831008911,
      "num_tokens": 44075500.0,
      "step": 3573
    },
    {
      "entropy": 1.1729064583778381,
      "epoch": 1.8820431806213795,
      "grad_norm": 0.2514488399028778,
      "learning_rate": 0.00017440249255917218,
      "loss": 0.17970673739910126,
      "mean_token_accuracy": 0.9232833385467529,
      "num_tokens": 44087836.0,
      "step": 3574
    },
    {
      "entropy": 1.1274474561214447,
      "epoch": 1.8825697735650342,
      "grad_norm": 0.27165448665618896,
      "learning_rate": 0.0001743871731367805,
      "loss": 0.18584352731704712,
      "mean_token_accuracy": 0.925078958272934,
      "num_tokens": 44100172.0,
      "step": 3575
    },
    {
      "entropy": 1.1284977793693542,
      "epoch": 1.8830963665086888,
      "grad_norm": 0.31029313802719116,
      "learning_rate": 0.00017437184989209755,
      "loss": 0.22961436212062836,
      "mean_token_accuracy": 0.9089037775993347,
      "num_tokens": 44112508.0,
      "step": 3576
    },
    {
      "entropy": 1.1539038717746735,
      "epoch": 1.8836229594523433,
      "grad_norm": 0.2871949076652527,
      "learning_rate": 0.00017435652282603309,
      "loss": 0.2109770029783249,
      "mean_token_accuracy": 0.9152273237705231,
      "num_tokens": 44124844.0,
      "step": 3577
    },
    {
      "entropy": 1.2086017429828644,
      "epoch": 1.884149552395998,
      "grad_norm": 0.2972370684146881,
      "learning_rate": 0.000174341191939497,
      "loss": 0.22491230070590973,
      "mean_token_accuracy": 0.9096596390008926,
      "num_tokens": 44137180.0,
      "step": 3578
    },
    {
      "entropy": 1.1673455834388733,
      "epoch": 1.8846761453396526,
      "grad_norm": 0.29084253311157227,
      "learning_rate": 0.00017432585723339933,
      "loss": 0.21230041980743408,
      "mean_token_accuracy": 0.9149164706468582,
      "num_tokens": 44149516.0,
      "step": 3579
    },
    {
      "entropy": 1.1442771553993225,
      "epoch": 1.885202738283307,
      "grad_norm": 0.2587001621723175,
      "learning_rate": 0.00017431051870865044,
      "loss": 0.19252054393291473,
      "mean_token_accuracy": 0.9200115948915482,
      "num_tokens": 44161852.0,
      "step": 3580
    },
    {
      "entropy": 1.185366839170456,
      "epoch": 1.8857293312269614,
      "grad_norm": 0.26025739312171936,
      "learning_rate": 0.00017429517636616086,
      "loss": 0.1987098902463913,
      "mean_token_accuracy": 0.9187203794717789,
      "num_tokens": 44174188.0,
      "step": 3581
    },
    {
      "entropy": 1.1812830567359924,
      "epoch": 1.8862559241706163,
      "grad_norm": 0.26062434911727905,
      "learning_rate": 0.00017427983020684145,
      "loss": 0.19222667813301086,
      "mean_token_accuracy": 0.9221339523792267,
      "num_tokens": 44186524.0,
      "step": 3582
    },
    {
      "entropy": 1.188157171010971,
      "epoch": 1.8867825171142707,
      "grad_norm": 0.27801626920700073,
      "learning_rate": 0.0001742644802316031,
      "loss": 0.19216661155223846,
      "mean_token_accuracy": 0.9170040041208267,
      "num_tokens": 44198860.0,
      "step": 3583
    },
    {
      "entropy": 1.1949248611927032,
      "epoch": 1.8873091100579251,
      "grad_norm": 0.24728041887283325,
      "learning_rate": 0.0001742491264413572,
      "loss": 0.187982976436615,
      "mean_token_accuracy": 0.9224367737770081,
      "num_tokens": 44211196.0,
      "step": 3584
    },
    {
      "entropy": 1.208105355501175,
      "epoch": 1.8878357030015798,
      "grad_norm": 0.32599449157714844,
      "learning_rate": 0.00017423376883701509,
      "loss": 0.1942829042673111,
      "mean_token_accuracy": 0.9238618314266205,
      "num_tokens": 44223532.0,
      "step": 3585
    },
    {
      "entropy": 1.2586550116539001,
      "epoch": 1.8883622959452344,
      "grad_norm": 0.2833736836910248,
      "learning_rate": 0.00017421840741948852,
      "loss": 0.1979784369468689,
      "mean_token_accuracy": 0.9184264540672302,
      "num_tokens": 44235868.0,
      "step": 3586
    },
    {
      "entropy": 1.2161791920661926,
      "epoch": 1.8888888888888888,
      "grad_norm": 0.27175453305244446,
      "learning_rate": 0.0001742030421896894,
      "loss": 0.18781299889087677,
      "mean_token_accuracy": 0.923054501414299,
      "num_tokens": 44248204.0,
      "step": 3587
    },
    {
      "entropy": 1.274359554052353,
      "epoch": 1.8894154818325435,
      "grad_norm": 0.300649493932724,
      "learning_rate": 0.00017418767314852986,
      "loss": 0.21826910972595215,
      "mean_token_accuracy": 0.906836673617363,
      "num_tokens": 44260540.0,
      "step": 3588
    },
    {
      "entropy": 1.2163540720939636,
      "epoch": 1.8899420747761981,
      "grad_norm": 0.25215044617652893,
      "learning_rate": 0.00017417230029692228,
      "loss": 0.20363149046897888,
      "mean_token_accuracy": 0.918889969587326,
      "num_tokens": 44272876.0,
      "step": 3589
    },
    {
      "entropy": 1.2891521751880646,
      "epoch": 1.8904686677198526,
      "grad_norm": 0.2741580903530121,
      "learning_rate": 0.00017415692363577926,
      "loss": 0.19672846794128418,
      "mean_token_accuracy": 0.9205434173345566,
      "num_tokens": 44285212.0,
      "step": 3590
    },
    {
      "entropy": 1.2554100453853607,
      "epoch": 1.890995260663507,
      "grad_norm": 0.2708304226398468,
      "learning_rate": 0.00017414154316601363,
      "loss": 0.19891998171806335,
      "mean_token_accuracy": 0.9197966605424881,
      "num_tokens": 44297548.0,
      "step": 3591
    },
    {
      "entropy": 1.2006763517856598,
      "epoch": 1.8915218536071616,
      "grad_norm": 0.2572955787181854,
      "learning_rate": 0.00017412615888853837,
      "loss": 0.196050226688385,
      "mean_token_accuracy": 0.923739418387413,
      "num_tokens": 44309884.0,
      "step": 3592
    },
    {
      "entropy": 1.2014833986759186,
      "epoch": 1.8920484465508163,
      "grad_norm": 0.2789122760295868,
      "learning_rate": 0.00017411077080426688,
      "loss": 0.19379383325576782,
      "mean_token_accuracy": 0.9179646968841553,
      "num_tokens": 44322220.0,
      "step": 3593
    },
    {
      "entropy": 1.2050150334835052,
      "epoch": 1.8925750394944707,
      "grad_norm": 0.2656964659690857,
      "learning_rate": 0.00017409537891411255,
      "loss": 0.2036164104938507,
      "mean_token_accuracy": 0.9147919863462448,
      "num_tokens": 44334556.0,
      "step": 3594
    },
    {
      "entropy": 1.1853137016296387,
      "epoch": 1.8931016324381253,
      "grad_norm": 0.2627456784248352,
      "learning_rate": 0.00017407998321898916,
      "loss": 0.2011202722787857,
      "mean_token_accuracy": 0.9194943606853485,
      "num_tokens": 44346892.0,
      "step": 3595
    },
    {
      "entropy": 1.1744475960731506,
      "epoch": 1.89362822538178,
      "grad_norm": 0.2604866027832031,
      "learning_rate": 0.0001740645837198107,
      "loss": 0.2077919840812683,
      "mean_token_accuracy": 0.9139293134212494,
      "num_tokens": 44359228.0,
      "step": 3596
    },
    {
      "entropy": 1.173338770866394,
      "epoch": 1.8941548183254344,
      "grad_norm": 0.2635132968425751,
      "learning_rate": 0.00017404918041749126,
      "loss": 0.20279425382614136,
      "mean_token_accuracy": 0.9182106554508209,
      "num_tokens": 44371564.0,
      "step": 3597
    },
    {
      "entropy": 1.1710366010665894,
      "epoch": 1.8946814112690888,
      "grad_norm": 0.2733679711818695,
      "learning_rate": 0.0001740337733129453,
      "loss": 0.19158853590488434,
      "mean_token_accuracy": 0.9208205044269562,
      "num_tokens": 44383900.0,
      "step": 3598
    },
    {
      "entropy": 1.113152414560318,
      "epoch": 1.8952080042127437,
      "grad_norm": 0.24724428355693817,
      "learning_rate": 0.0001740183624070874,
      "loss": 0.18724195659160614,
      "mean_token_accuracy": 0.922689363360405,
      "num_tokens": 44396236.0,
      "step": 3599
    },
    {
      "entropy": 1.1231700778007507,
      "epoch": 1.8957345971563981,
      "grad_norm": 0.2390860617160797,
      "learning_rate": 0.00017400294770083247,
      "loss": 0.16968569159507751,
      "mean_token_accuracy": 0.9303603619337082,
      "num_tokens": 44408572.0,
      "step": 3600
    },
    {
      "entropy": 1.2249704897403717,
      "epoch": 1.8962611901000526,
      "grad_norm": 0.2889772951602936,
      "learning_rate": 0.0001739875291950956,
      "loss": 0.20824474096298218,
      "mean_token_accuracy": 0.9169190227985382,
      "num_tokens": 44420908.0,
      "step": 3601
    },
    {
      "entropy": 1.1704488396644592,
      "epoch": 1.8967877830437072,
      "grad_norm": 0.2680244445800781,
      "learning_rate": 0.00017397210689079203,
      "loss": 0.20673055946826935,
      "mean_token_accuracy": 0.9180959910154343,
      "num_tokens": 44433244.0,
      "step": 3602
    },
    {
      "entropy": 1.1878012716770172,
      "epoch": 1.8973143759873619,
      "grad_norm": 0.2975465953350067,
      "learning_rate": 0.00017395668078883734,
      "loss": 0.20991596579551697,
      "mean_token_accuracy": 0.9183283597230911,
      "num_tokens": 44445580.0,
      "step": 3603
    },
    {
      "entropy": 1.1773861050605774,
      "epoch": 1.8978409689310163,
      "grad_norm": 0.2718547284603119,
      "learning_rate": 0.0001739412508901473,
      "loss": 0.20017877221107483,
      "mean_token_accuracy": 0.9203770905733109,
      "num_tokens": 44457916.0,
      "step": 3604
    },
    {
      "entropy": 1.1368711590766907,
      "epoch": 1.898367561874671,
      "grad_norm": 0.2801942527294159,
      "learning_rate": 0.00017392581719563786,
      "loss": 0.20202471315860748,
      "mean_token_accuracy": 0.9181565046310425,
      "num_tokens": 44470252.0,
      "step": 3605
    },
    {
      "entropy": 1.1801301538944244,
      "epoch": 1.8988941548183256,
      "grad_norm": 0.30454739928245544,
      "learning_rate": 0.00017391037970622525,
      "loss": 0.2132726013660431,
      "mean_token_accuracy": 0.9120954275131226,
      "num_tokens": 44482588.0,
      "step": 3606
    },
    {
      "entropy": 1.1741478145122528,
      "epoch": 1.89942074776198,
      "grad_norm": 0.26906171441078186,
      "learning_rate": 0.00017389493842282587,
      "loss": 0.19884657859802246,
      "mean_token_accuracy": 0.917621061205864,
      "num_tokens": 44494924.0,
      "step": 3607
    },
    {
      "entropy": 1.1842193901538849,
      "epoch": 1.8999473407056344,
      "grad_norm": 0.25997108221054077,
      "learning_rate": 0.00017387949334635644,
      "loss": 0.19672074913978577,
      "mean_token_accuracy": 0.9219666868448257,
      "num_tokens": 44507260.0,
      "step": 3608
    },
    {
      "entropy": 1.1503727734088898,
      "epoch": 1.900473933649289,
      "grad_norm": 0.27212950587272644,
      "learning_rate": 0.00017386404447773378,
      "loss": 0.19361799955368042,
      "mean_token_accuracy": 0.9216768741607666,
      "num_tokens": 44519596.0,
      "step": 3609
    },
    {
      "entropy": 1.2314816415309906,
      "epoch": 1.9010005265929437,
      "grad_norm": 0.27919554710388184,
      "learning_rate": 0.00017384859181787503,
      "loss": 0.21054218709468842,
      "mean_token_accuracy": 0.9163404256105423,
      "num_tokens": 44531932.0,
      "step": 3610
    },
    {
      "entropy": 1.1658604443073273,
      "epoch": 1.9015271195365981,
      "grad_norm": 0.2740512192249298,
      "learning_rate": 0.00017383313536769755,
      "loss": 0.18958908319473267,
      "mean_token_accuracy": 0.9197452515363693,
      "num_tokens": 44544268.0,
      "step": 3611
    },
    {
      "entropy": 1.2027749419212341,
      "epoch": 1.9020537124802528,
      "grad_norm": 0.264543741941452,
      "learning_rate": 0.00017381767512811885,
      "loss": 0.20404848456382751,
      "mean_token_accuracy": 0.916832685470581,
      "num_tokens": 44556604.0,
      "step": 3612
    },
    {
      "entropy": 1.1814589202404022,
      "epoch": 1.9025803054239074,
      "grad_norm": 0.2642267048358917,
      "learning_rate": 0.00017380221110005673,
      "loss": 0.19606810808181763,
      "mean_token_accuracy": 0.918126568198204,
      "num_tokens": 44568940.0,
      "step": 3613
    },
    {
      "entropy": 1.1763567328453064,
      "epoch": 1.9031068983675619,
      "grad_norm": 0.2508313059806824,
      "learning_rate": 0.0001737867432844292,
      "loss": 0.1951427459716797,
      "mean_token_accuracy": 0.9166415929794312,
      "num_tokens": 44581276.0,
      "step": 3614
    },
    {
      "entropy": 1.2736837267875671,
      "epoch": 1.9036334913112163,
      "grad_norm": 0.2942288815975189,
      "learning_rate": 0.0001737712716821545,
      "loss": 0.21287119388580322,
      "mean_token_accuracy": 0.9135104715824127,
      "num_tokens": 44593612.0,
      "step": 3615
    },
    {
      "entropy": 1.2576221525669098,
      "epoch": 1.904160084254871,
      "grad_norm": 0.255913108587265,
      "learning_rate": 0.00017375579629415105,
      "loss": 0.1842191517353058,
      "mean_token_accuracy": 0.9250049591064453,
      "num_tokens": 44605948.0,
      "step": 3616
    },
    {
      "entropy": 1.251773089170456,
      "epoch": 1.9046866771985256,
      "grad_norm": 0.2758917212486267,
      "learning_rate": 0.00017374031712133765,
      "loss": 0.21524254977703094,
      "mean_token_accuracy": 0.9172472357749939,
      "num_tokens": 44618284.0,
      "step": 3617
    },
    {
      "entropy": 1.2298138439655304,
      "epoch": 1.90521327014218,
      "grad_norm": 0.2682817876338959,
      "learning_rate": 0.0001737248341646331,
      "loss": 0.19541317224502563,
      "mean_token_accuracy": 0.9179702699184418,
      "num_tokens": 44630620.0,
      "step": 3618
    },
    {
      "entropy": 1.2236283421516418,
      "epoch": 1.9057398630858347,
      "grad_norm": 0.26307037472724915,
      "learning_rate": 0.00017370934742495656,
      "loss": 0.20737697184085846,
      "mean_token_accuracy": 0.9137110263109207,
      "num_tokens": 44642956.0,
      "step": 3619
    },
    {
      "entropy": 1.243878036737442,
      "epoch": 1.9062664560294893,
      "grad_norm": 0.2502284049987793,
      "learning_rate": 0.0001736938569032274,
      "loss": 0.20081773400306702,
      "mean_token_accuracy": 0.9174046218395233,
      "num_tokens": 44655292.0,
      "step": 3620
    },
    {
      "entropy": 1.2514996528625488,
      "epoch": 1.9067930489731437,
      "grad_norm": 0.250877320766449,
      "learning_rate": 0.00017367836260036515,
      "loss": 0.18941593170166016,
      "mean_token_accuracy": 0.9203214943408966,
      "num_tokens": 44667628.0,
      "step": 3621
    },
    {
      "entropy": 1.2207768857479095,
      "epoch": 1.9073196419167984,
      "grad_norm": 0.29659348726272583,
      "learning_rate": 0.00017366286451728967,
      "loss": 0.18773111701011658,
      "mean_token_accuracy": 0.9253512769937515,
      "num_tokens": 44679964.0,
      "step": 3622
    },
    {
      "entropy": 1.1811819076538086,
      "epoch": 1.907846234860453,
      "grad_norm": 0.25071752071380615,
      "learning_rate": 0.000173647362654921,
      "loss": 0.1999988704919815,
      "mean_token_accuracy": 0.9158650636672974,
      "num_tokens": 44692300.0,
      "step": 3623
    },
    {
      "entropy": 1.2793206870555878,
      "epoch": 1.9083728278041074,
      "grad_norm": 0.28312739729881287,
      "learning_rate": 0.0001736318570141794,
      "loss": 0.20251990854740143,
      "mean_token_accuracy": 0.9155915677547455,
      "num_tokens": 44704636.0,
      "step": 3624
    },
    {
      "entropy": 1.2060322165489197,
      "epoch": 1.9088994207477619,
      "grad_norm": 0.24651455879211426,
      "learning_rate": 0.00017361634759598525,
      "loss": 0.19613385200500488,
      "mean_token_accuracy": 0.9165442436933517,
      "num_tokens": 44716972.0,
      "step": 3625
    },
    {
      "entropy": 1.1942104697227478,
      "epoch": 1.9094260136914165,
      "grad_norm": 0.2771541178226471,
      "learning_rate": 0.0001736008344012594,
      "loss": 0.21619367599487305,
      "mean_token_accuracy": 0.9139316976070404,
      "num_tokens": 44729308.0,
      "step": 3626
    },
    {
      "entropy": 1.1322157979011536,
      "epoch": 1.9099526066350712,
      "grad_norm": 0.25058284401893616,
      "learning_rate": 0.00017358531743092266,
      "loss": 0.18762576580047607,
      "mean_token_accuracy": 0.9215554296970367,
      "num_tokens": 44741644.0,
      "step": 3627
    },
    {
      "entropy": 1.2009500563144684,
      "epoch": 1.9104791995787256,
      "grad_norm": 0.27951785922050476,
      "learning_rate": 0.00017356979668589625,
      "loss": 0.2000935822725296,
      "mean_token_accuracy": 0.9177307933568954,
      "num_tokens": 44753980.0,
      "step": 3628
    },
    {
      "entropy": 1.185960978269577,
      "epoch": 1.9110057925223802,
      "grad_norm": 0.2635090947151184,
      "learning_rate": 0.00017355427216710152,
      "loss": 0.19260722398757935,
      "mean_token_accuracy": 0.9208453297615051,
      "num_tokens": 44766316.0,
      "step": 3629
    },
    {
      "entropy": 1.179717481136322,
      "epoch": 1.9115323854660349,
      "grad_norm": 0.291584312915802,
      "learning_rate": 0.00017353874387546005,
      "loss": 0.20042484998703003,
      "mean_token_accuracy": 0.91756771504879,
      "num_tokens": 44778652.0,
      "step": 3630
    },
    {
      "entropy": 1.1897325813770294,
      "epoch": 1.9120589784096893,
      "grad_norm": 0.2626997232437134,
      "learning_rate": 0.00017352321181189373,
      "loss": 0.19658340513706207,
      "mean_token_accuracy": 0.9180559068918228,
      "num_tokens": 44790988.0,
      "step": 3631
    },
    {
      "entropy": 1.1979995667934418,
      "epoch": 1.9125855713533437,
      "grad_norm": 0.28412285447120667,
      "learning_rate": 0.00017350767597732453,
      "loss": 0.21948397159576416,
      "mean_token_accuracy": 0.9098146557807922,
      "num_tokens": 44803324.0,
      "step": 3632
    },
    {
      "entropy": 1.1823217272758484,
      "epoch": 1.9131121642969984,
      "grad_norm": 0.27307748794555664,
      "learning_rate": 0.00017349213637267476,
      "loss": 0.21455425024032593,
      "mean_token_accuracy": 0.9142604321241379,
      "num_tokens": 44815660.0,
      "step": 3633
    },
    {
      "entropy": 1.2042896449565887,
      "epoch": 1.913638757240653,
      "grad_norm": 0.31086379289627075,
      "learning_rate": 0.00017347659299886693,
      "loss": 0.21350350975990295,
      "mean_token_accuracy": 0.9092680513858795,
      "num_tokens": 44827996.0,
      "step": 3634
    },
    {
      "entropy": 1.2283751964569092,
      "epoch": 1.9141653501843074,
      "grad_norm": 0.2663927376270294,
      "learning_rate": 0.00017346104585682373,
      "loss": 0.19472280144691467,
      "mean_token_accuracy": 0.9193729162216187,
      "num_tokens": 44840332.0,
      "step": 3635
    },
    {
      "entropy": 1.183545082807541,
      "epoch": 1.914691943127962,
      "grad_norm": 0.2771006226539612,
      "learning_rate": 0.00017344549494746815,
      "loss": 0.19897666573524475,
      "mean_token_accuracy": 0.9188507944345474,
      "num_tokens": 44852668.0,
      "step": 3636
    },
    {
      "entropy": 1.2296278774738312,
      "epoch": 1.9152185360716167,
      "grad_norm": 0.27564653754234314,
      "learning_rate": 0.0001734299402717233,
      "loss": 0.20981676876544952,
      "mean_token_accuracy": 0.9140519946813583,
      "num_tokens": 44865004.0,
      "step": 3637
    },
    {
      "entropy": 1.2284190952777863,
      "epoch": 1.9157451290152712,
      "grad_norm": 0.26339784264564514,
      "learning_rate": 0.0001734143818305126,
      "loss": 0.19718164205551147,
      "mean_token_accuracy": 0.9198196828365326,
      "num_tokens": 44877340.0,
      "step": 3638
    },
    {
      "entropy": 1.1841385960578918,
      "epoch": 1.9162717219589256,
      "grad_norm": 0.28311020135879517,
      "learning_rate": 0.00017339881962475965,
      "loss": 0.21129263937473297,
      "mean_token_accuracy": 0.912701204419136,
      "num_tokens": 44889676.0,
      "step": 3639
    },
    {
      "entropy": 1.1703853607177734,
      "epoch": 1.9167983149025805,
      "grad_norm": 0.24933747947216034,
      "learning_rate": 0.00017338325365538827,
      "loss": 0.2056872695684433,
      "mean_token_accuracy": 0.9159862697124481,
      "num_tokens": 44902012.0,
      "step": 3640
    },
    {
      "entropy": 1.2645052373409271,
      "epoch": 1.9173249078462349,
      "grad_norm": 0.26389506459236145,
      "learning_rate": 0.00017336768392332258,
      "loss": 0.19748938083648682,
      "mean_token_accuracy": 0.9188157171010971,
      "num_tokens": 44914348.0,
      "step": 3641
    },
    {
      "entropy": 1.2652300000190735,
      "epoch": 1.9178515007898893,
      "grad_norm": 0.255624383687973,
      "learning_rate": 0.00017335211042948683,
      "loss": 0.18445834517478943,
      "mean_token_accuracy": 0.9235481023788452,
      "num_tokens": 44926684.0,
      "step": 3642
    },
    {
      "entropy": 1.2511534094810486,
      "epoch": 1.918378093733544,
      "grad_norm": 0.26123735308647156,
      "learning_rate": 0.0001733365331748055,
      "loss": 0.19365046918392181,
      "mean_token_accuracy": 0.921040266752243,
      "num_tokens": 44939020.0,
      "step": 3643
    },
    {
      "entropy": 1.2718633115291595,
      "epoch": 1.9189046866771986,
      "grad_norm": 0.28643572330474854,
      "learning_rate": 0.00017332095216020338,
      "loss": 0.2148754596710205,
      "mean_token_accuracy": 0.9159789234399796,
      "num_tokens": 44951356.0,
      "step": 3644
    },
    {
      "entropy": 1.3365231156349182,
      "epoch": 1.919431279620853,
      "grad_norm": 0.2991255223751068,
      "learning_rate": 0.0001733053673866054,
      "loss": 0.2158765196800232,
      "mean_token_accuracy": 0.9090882688760757,
      "num_tokens": 44963692.0,
      "step": 3645
    },
    {
      "entropy": 1.2829530537128448,
      "epoch": 1.9199578725645077,
      "grad_norm": 0.31265580654144287,
      "learning_rate": 0.0001732897788549367,
      "loss": 0.22893324494361877,
      "mean_token_accuracy": 0.9042454361915588,
      "num_tokens": 44976028.0,
      "step": 3646
    },
    {
      "entropy": 1.2604970932006836,
      "epoch": 1.9204844655081623,
      "grad_norm": 0.28193265199661255,
      "learning_rate": 0.00017327418656612272,
      "loss": 0.20636524260044098,
      "mean_token_accuracy": 0.9135946780443192,
      "num_tokens": 44988364.0,
      "step": 3647
    },
    {
      "entropy": 1.3023660778999329,
      "epoch": 1.9210110584518167,
      "grad_norm": 0.28154146671295166,
      "learning_rate": 0.00017325859052108906,
      "loss": 0.2026195228099823,
      "mean_token_accuracy": 0.9120919853448868,
      "num_tokens": 45000700.0,
      "step": 3648
    },
    {
      "entropy": 1.265262395143509,
      "epoch": 1.9215376513954712,
      "grad_norm": 0.24685363471508026,
      "learning_rate": 0.0001732429907207616,
      "loss": 0.19602523744106293,
      "mean_token_accuracy": 0.9172793030738831,
      "num_tokens": 45013036.0,
      "step": 3649
    },
    {
      "entropy": 1.3104206323623657,
      "epoch": 1.9220642443391258,
      "grad_norm": 0.31668615341186523,
      "learning_rate": 0.00017322738716606638,
      "loss": 0.22926807403564453,
      "mean_token_accuracy": 0.9078969061374664,
      "num_tokens": 45025372.0,
      "step": 3650
    },
    {
      "entropy": 1.2556992173194885,
      "epoch": 1.9225908372827805,
      "grad_norm": 0.2678292691707611,
      "learning_rate": 0.00017321177985792972,
      "loss": 0.19617974758148193,
      "mean_token_accuracy": 0.9211210906505585,
      "num_tokens": 45037708.0,
      "step": 3651
    },
    {
      "entropy": 1.3135022222995758,
      "epoch": 1.9231174302264349,
      "grad_norm": 0.26799413561820984,
      "learning_rate": 0.0001731961687972781,
      "loss": 0.20180177688598633,
      "mean_token_accuracy": 0.916679710149765,
      "num_tokens": 45050044.0,
      "step": 3652
    },
    {
      "entropy": 1.2593377232551575,
      "epoch": 1.9236440231700895,
      "grad_norm": 0.26281964778900146,
      "learning_rate": 0.0001731805539850383,
      "loss": 0.1954004168510437,
      "mean_token_accuracy": 0.9243785589933395,
      "num_tokens": 45062380.0,
      "step": 3653
    },
    {
      "entropy": 1.2636898756027222,
      "epoch": 1.9241706161137442,
      "grad_norm": 0.26839685440063477,
      "learning_rate": 0.00017316493542213722,
      "loss": 0.19961456954479218,
      "mean_token_accuracy": 0.9217253774404526,
      "num_tokens": 45074716.0,
      "step": 3654
    },
    {
      "entropy": 1.2462226152420044,
      "epoch": 1.9246972090573986,
      "grad_norm": 0.2672293186187744,
      "learning_rate": 0.00017314931310950208,
      "loss": 0.191788911819458,
      "mean_token_accuracy": 0.9207832664251328,
      "num_tokens": 45087052.0,
      "step": 3655
    },
    {
      "entropy": 1.24080029129982,
      "epoch": 1.925223802001053,
      "grad_norm": 0.27368423342704773,
      "learning_rate": 0.0001731336870480603,
      "loss": 0.19978970289230347,
      "mean_token_accuracy": 0.9169470369815826,
      "num_tokens": 45099388.0,
      "step": 3656
    },
    {
      "entropy": 1.1998139321804047,
      "epoch": 1.925750394944708,
      "grad_norm": 0.2728565037250519,
      "learning_rate": 0.00017311805723873946,
      "loss": 0.20359551906585693,
      "mean_token_accuracy": 0.9137473702430725,
      "num_tokens": 45111724.0,
      "step": 3657
    },
    {
      "entropy": 1.2402350306510925,
      "epoch": 1.9262769878883623,
      "grad_norm": 0.28210633993148804,
      "learning_rate": 0.00017310242368246746,
      "loss": 0.19972732663154602,
      "mean_token_accuracy": 0.9203901588916779,
      "num_tokens": 45124060.0,
      "step": 3658
    },
    {
      "entropy": 1.2057775259017944,
      "epoch": 1.9268035808320167,
      "grad_norm": 0.26620033383369446,
      "learning_rate": 0.00017308678638017238,
      "loss": 0.19069263339042664,
      "mean_token_accuracy": 0.9222879111766815,
      "num_tokens": 45136396.0,
      "step": 3659
    },
    {
      "entropy": 1.2476109266281128,
      "epoch": 1.9273301737756714,
      "grad_norm": 0.24337823688983917,
      "learning_rate": 0.00017307114533278247,
      "loss": 0.1780460923910141,
      "mean_token_accuracy": 0.9250701814889908,
      "num_tokens": 45148732.0,
      "step": 3660
    },
    {
      "entropy": 1.1989704370498657,
      "epoch": 1.927856766719326,
      "grad_norm": 0.25690510869026184,
      "learning_rate": 0.00017305550054122625,
      "loss": 0.19341933727264404,
      "mean_token_accuracy": 0.9228748381137848,
      "num_tokens": 45161068.0,
      "step": 3661
    },
    {
      "entropy": 1.2263188660144806,
      "epoch": 1.9283833596629805,
      "grad_norm": 0.26640671491622925,
      "learning_rate": 0.0001730398520064325,
      "loss": 0.18614225089550018,
      "mean_token_accuracy": 0.9255273640155792,
      "num_tokens": 45173404.0,
      "step": 3662
    },
    {
      "entropy": 1.272974044084549,
      "epoch": 1.9289099526066351,
      "grad_norm": 0.3008297383785248,
      "learning_rate": 0.00017302419972933014,
      "loss": 0.2073616087436676,
      "mean_token_accuracy": 0.9149825423955917,
      "num_tokens": 45185740.0,
      "step": 3663
    },
    {
      "entropy": 1.2914739549160004,
      "epoch": 1.9294365455502898,
      "grad_norm": 0.27512502670288086,
      "learning_rate": 0.0001730085437108484,
      "loss": 0.20276200771331787,
      "mean_token_accuracy": 0.9159463495016098,
      "num_tokens": 45198076.0,
      "step": 3664
    },
    {
      "entropy": 1.3106342256069183,
      "epoch": 1.9299631384939442,
      "grad_norm": 0.29478466510772705,
      "learning_rate": 0.00017299288395191664,
      "loss": 0.20114478468894958,
      "mean_token_accuracy": 0.9164300113916397,
      "num_tokens": 45210412.0,
      "step": 3665
    },
    {
      "entropy": 1.3369048535823822,
      "epoch": 1.9304897314375986,
      "grad_norm": 0.27624809741973877,
      "learning_rate": 0.0001729772204534645,
      "loss": 0.20260167121887207,
      "mean_token_accuracy": 0.9126608073711395,
      "num_tokens": 45222748.0,
      "step": 3666
    },
    {
      "entropy": 1.3504050076007843,
      "epoch": 1.9310163243812533,
      "grad_norm": 0.31120526790618896,
      "learning_rate": 0.00017296155321642188,
      "loss": 0.2171817421913147,
      "mean_token_accuracy": 0.9078565239906311,
      "num_tokens": 45235084.0,
      "step": 3667
    },
    {
      "entropy": 1.3360699117183685,
      "epoch": 1.931542917324908,
      "grad_norm": 0.29817092418670654,
      "learning_rate": 0.00017294588224171879,
      "loss": 0.1921902298927307,
      "mean_token_accuracy": 0.9241172969341278,
      "num_tokens": 45247420.0,
      "step": 3668
    },
    {
      "entropy": 1.3836988508701324,
      "epoch": 1.9320695102685623,
      "grad_norm": 0.2684880793094635,
      "learning_rate": 0.00017293020753028556,
      "loss": 0.20802155137062073,
      "mean_token_accuracy": 0.9166839867830276,
      "num_tokens": 45259756.0,
      "step": 3669
    },
    {
      "entropy": 1.4190702736377716,
      "epoch": 1.932596103212217,
      "grad_norm": 0.25781887769699097,
      "learning_rate": 0.00017291452908305268,
      "loss": 0.18944363296031952,
      "mean_token_accuracy": 0.9194059371948242,
      "num_tokens": 45272092.0,
      "step": 3670
    },
    {
      "entropy": 1.4297086596488953,
      "epoch": 1.9331226961558716,
      "grad_norm": 0.25668513774871826,
      "learning_rate": 0.00017289884690095092,
      "loss": 0.18865437805652618,
      "mean_token_accuracy": 0.9221978783607483,
      "num_tokens": 45284428.0,
      "step": 3671
    },
    {
      "entropy": 1.4811699092388153,
      "epoch": 1.933649289099526,
      "grad_norm": 0.25223368406295776,
      "learning_rate": 0.00017288316098491122,
      "loss": 0.1869308352470398,
      "mean_token_accuracy": 0.926489993929863,
      "num_tokens": 45296764.0,
      "step": 3672
    },
    {
      "entropy": 1.4673098027706146,
      "epoch": 1.9341758820431805,
      "grad_norm": 0.2823149263858795,
      "learning_rate": 0.00017286747133586473,
      "loss": 0.21887610852718353,
      "mean_token_accuracy": 0.9101450145244598,
      "num_tokens": 45309100.0,
      "step": 3673
    },
    {
      "entropy": 1.4374184310436249,
      "epoch": 1.9347024749868353,
      "grad_norm": 0.2951841652393341,
      "learning_rate": 0.00017285177795474293,
      "loss": 0.21811825037002563,
      "mean_token_accuracy": 0.9097328335046768,
      "num_tokens": 45321436.0,
      "step": 3674
    },
    {
      "entropy": 1.474625825881958,
      "epoch": 1.9352290679304898,
      "grad_norm": 0.2604941725730896,
      "learning_rate": 0.00017283608084247738,
      "loss": 0.18887674808502197,
      "mean_token_accuracy": 0.9238714724779129,
      "num_tokens": 45333772.0,
      "step": 3675
    },
    {
      "entropy": 1.4799661338329315,
      "epoch": 1.9357556608741442,
      "grad_norm": 0.2468399554491043,
      "learning_rate": 0.00017282037999999996,
      "loss": 0.20713761448860168,
      "mean_token_accuracy": 0.9197785556316376,
      "num_tokens": 45346108.0,
      "step": 3676
    },
    {
      "entropy": 1.4618735611438751,
      "epoch": 1.9362822538177988,
      "grad_norm": 0.24083919823169708,
      "learning_rate": 0.0001728046754282427,
      "loss": 0.18771159648895264,
      "mean_token_accuracy": 0.9197709262371063,
      "num_tokens": 45358444.0,
      "step": 3677
    },
    {
      "entropy": 1.5067411363124847,
      "epoch": 1.9368088467614535,
      "grad_norm": 0.2688843607902527,
      "learning_rate": 0.00017278896712813794,
      "loss": 0.19874204695224762,
      "mean_token_accuracy": 0.9193889200687408,
      "num_tokens": 45370780.0,
      "step": 3678
    },
    {
      "entropy": 1.4961673319339752,
      "epoch": 1.937335439705108,
      "grad_norm": 0.28911784291267395,
      "learning_rate": 0.00017277325510061813,
      "loss": 0.20285126566886902,
      "mean_token_accuracy": 0.9150000810623169,
      "num_tokens": 45383116.0,
      "step": 3679
    },
    {
      "entropy": 1.5026283860206604,
      "epoch": 1.9378620326487626,
      "grad_norm": 0.26418668031692505,
      "learning_rate": 0.00017275753934661606,
      "loss": 0.2061035931110382,
      "mean_token_accuracy": 0.9167807698249817,
      "num_tokens": 45395452.0,
      "step": 3680
    },
    {
      "entropy": 1.4821618497371674,
      "epoch": 1.9383886255924172,
      "grad_norm": 0.2582075893878937,
      "learning_rate": 0.00017274181986706464,
      "loss": 0.19469109177589417,
      "mean_token_accuracy": 0.9196734130382538,
      "num_tokens": 45407788.0,
      "step": 3681
    },
    {
      "entropy": 1.4628108739852905,
      "epoch": 1.9389152185360716,
      "grad_norm": 0.27692705392837524,
      "learning_rate": 0.0001727260966628971,
      "loss": 0.20121949911117554,
      "mean_token_accuracy": 0.9166281670331955,
      "num_tokens": 45420124.0,
      "step": 3682
    },
    {
      "entropy": 1.4591149389743805,
      "epoch": 1.939441811479726,
      "grad_norm": 0.28909337520599365,
      "learning_rate": 0.00017271036973504674,
      "loss": 0.19279229640960693,
      "mean_token_accuracy": 0.9220520555973053,
      "num_tokens": 45432460.0,
      "step": 3683
    },
    {
      "entropy": 1.5424366295337677,
      "epoch": 1.9399684044233807,
      "grad_norm": 0.3063618242740631,
      "learning_rate": 0.00017269463908444724,
      "loss": 0.2191891372203827,
      "mean_token_accuracy": 0.9127875715494156,
      "num_tokens": 45444796.0,
      "step": 3684
    },
    {
      "entropy": 1.4530845880508423,
      "epoch": 1.9404949973670353,
      "grad_norm": 0.2661772072315216,
      "learning_rate": 0.00017267890471203248,
      "loss": 0.19795681536197662,
      "mean_token_accuracy": 0.9154502600431442,
      "num_tokens": 45457132.0,
      "step": 3685
    },
    {
      "entropy": 1.4750907719135284,
      "epoch": 1.9410215903106898,
      "grad_norm": 0.2790045440196991,
      "learning_rate": 0.00017266316661873642,
      "loss": 0.22019743919372559,
      "mean_token_accuracy": 0.9140889048576355,
      "num_tokens": 45469468.0,
      "step": 3686
    },
    {
      "entropy": 1.4428746104240417,
      "epoch": 1.9415481832543444,
      "grad_norm": 0.26486507058143616,
      "learning_rate": 0.0001726474248054934,
      "loss": 0.1950511783361435,
      "mean_token_accuracy": 0.9183469414710999,
      "num_tokens": 45481804.0,
      "step": 3687
    },
    {
      "entropy": 1.5106236040592194,
      "epoch": 1.942074776197999,
      "grad_norm": 0.2506488263607025,
      "learning_rate": 0.00017263167927323794,
      "loss": 0.1819375455379486,
      "mean_token_accuracy": 0.9284966439008713,
      "num_tokens": 45494140.0,
      "step": 3688
    },
    {
      "entropy": 1.5388555824756622,
      "epoch": 1.9426013691416535,
      "grad_norm": 0.300800085067749,
      "learning_rate": 0.00017261593002290468,
      "loss": 0.19233667850494385,
      "mean_token_accuracy": 0.9199601858854294,
      "num_tokens": 45506476.0,
      "step": 3689
    },
    {
      "entropy": 1.48341503739357,
      "epoch": 1.943127962085308,
      "grad_norm": 0.32261741161346436,
      "learning_rate": 0.00017260017705542864,
      "loss": 0.21268944442272186,
      "mean_token_accuracy": 0.9111090153455734,
      "num_tokens": 45518812.0,
      "step": 3690
    },
    {
      "entropy": 1.5718600451946259,
      "epoch": 1.9436545550289626,
      "grad_norm": 0.29194772243499756,
      "learning_rate": 0.00017258442037174497,
      "loss": 0.19810812175273895,
      "mean_token_accuracy": 0.9194205850362778,
      "num_tokens": 45531148.0,
      "step": 3691
    },
    {
      "entropy": 1.4727455377578735,
      "epoch": 1.9441811479726172,
      "grad_norm": 0.263203501701355,
      "learning_rate": 0.00017256865997278896,
      "loss": 0.21588562428951263,
      "mean_token_accuracy": 0.9114230424165726,
      "num_tokens": 45543484.0,
      "step": 3692
    },
    {
      "entropy": 1.5082348585128784,
      "epoch": 1.9447077409162716,
      "grad_norm": 0.2874946594238281,
      "learning_rate": 0.0001725528958594963,
      "loss": 0.1873784065246582,
      "mean_token_accuracy": 0.9259777516126633,
      "num_tokens": 45555820.0,
      "step": 3693
    },
    {
      "entropy": 1.5053280889987946,
      "epoch": 1.9452343338599263,
      "grad_norm": 0.2794869542121887,
      "learning_rate": 0.00017253712803280284,
      "loss": 0.20504489541053772,
      "mean_token_accuracy": 0.9110683053731918,
      "num_tokens": 45568156.0,
      "step": 3694
    },
    {
      "entropy": 1.5468563735485077,
      "epoch": 1.945760926803581,
      "grad_norm": 0.27621355652809143,
      "learning_rate": 0.00017252135649364455,
      "loss": 0.21246497333049774,
      "mean_token_accuracy": 0.9150105863809586,
      "num_tokens": 45580492.0,
      "step": 3695
    },
    {
      "entropy": 1.5053722560405731,
      "epoch": 1.9462875197472354,
      "grad_norm": 0.2672613561153412,
      "learning_rate": 0.00017250558124295778,
      "loss": 0.20558930933475494,
      "mean_token_accuracy": 0.9144411534070969,
      "num_tokens": 45592828.0,
      "step": 3696
    },
    {
      "entropy": 1.4701599478721619,
      "epoch": 1.94681411269089,
      "grad_norm": 0.25279685854911804,
      "learning_rate": 0.0001724898022816789,
      "loss": 0.19829775393009186,
      "mean_token_accuracy": 0.921133354306221,
      "num_tokens": 45605164.0,
      "step": 3697
    },
    {
      "entropy": 1.5011445581912994,
      "epoch": 1.9473407056345446,
      "grad_norm": 0.2716440260410309,
      "learning_rate": 0.0001724740196107447,
      "loss": 0.20478171110153198,
      "mean_token_accuracy": 0.9158901125192642,
      "num_tokens": 45617500.0,
      "step": 3698
    },
    {
      "entropy": 1.480711042881012,
      "epoch": 1.947867298578199,
      "grad_norm": 0.2888793349266052,
      "learning_rate": 0.00017245823323109208,
      "loss": 0.19861237704753876,
      "mean_token_accuracy": 0.9149367064237595,
      "num_tokens": 45629836.0,
      "step": 3699
    },
    {
      "entropy": 1.426710307598114,
      "epoch": 1.9483938915218535,
      "grad_norm": 0.28729501366615295,
      "learning_rate": 0.00017244244314365822,
      "loss": 0.20462435483932495,
      "mean_token_accuracy": 0.9161471426486969,
      "num_tokens": 45642172.0,
      "step": 3700
    },
    {
      "entropy": 1.4438749551773071,
      "epoch": 1.9489204844655081,
      "grad_norm": 0.28520667552948,
      "learning_rate": 0.0001724266493493804,
      "loss": 0.20798423886299133,
      "mean_token_accuracy": 0.9113110899925232,
      "num_tokens": 45654508.0,
      "step": 3701
    },
    {
      "entropy": 1.4134547710418701,
      "epoch": 1.9494470774091628,
      "grad_norm": 0.26647916436195374,
      "learning_rate": 0.0001724108518491963,
      "loss": 0.19534523785114288,
      "mean_token_accuracy": 0.9230214208364487,
      "num_tokens": 45666844.0,
      "step": 3702
    },
    {
      "entropy": 1.4336603581905365,
      "epoch": 1.9499736703528172,
      "grad_norm": 0.27049604058265686,
      "learning_rate": 0.00017239505064404367,
      "loss": 0.18309858441352844,
      "mean_token_accuracy": 0.922065794467926,
      "num_tokens": 45679180.0,
      "step": 3703
    },
    {
      "entropy": 1.4100782871246338,
      "epoch": 1.9505002632964719,
      "grad_norm": 0.2777698338031769,
      "learning_rate": 0.00017237924573486053,
      "loss": 0.21323612332344055,
      "mean_token_accuracy": 0.9146802127361298,
      "num_tokens": 45691516.0,
      "step": 3704
    },
    {
      "entropy": 1.3497129678726196,
      "epoch": 1.9510268562401265,
      "grad_norm": 0.26056769490242004,
      "learning_rate": 0.00017236343712258515,
      "loss": 0.18909797072410583,
      "mean_token_accuracy": 0.9163524806499481,
      "num_tokens": 45703852.0,
      "step": 3705
    },
    {
      "entropy": 1.442967563867569,
      "epoch": 1.951553449183781,
      "grad_norm": 0.3076518774032593,
      "learning_rate": 0.000172347624808156,
      "loss": 0.20429672300815582,
      "mean_token_accuracy": 0.9184731394052505,
      "num_tokens": 45716188.0,
      "step": 3706
    },
    {
      "entropy": 1.3970259726047516,
      "epoch": 1.9520800421274354,
      "grad_norm": 0.2776000201702118,
      "learning_rate": 0.00017233180879251176,
      "loss": 0.18460848927497864,
      "mean_token_accuracy": 0.9276632517576218,
      "num_tokens": 45728524.0,
      "step": 3707
    },
    {
      "entropy": 1.3923786878585815,
      "epoch": 1.95260663507109,
      "grad_norm": 0.2807813286781311,
      "learning_rate": 0.00017231598907659133,
      "loss": 0.19556443393230438,
      "mean_token_accuracy": 0.9221478253602982,
      "num_tokens": 45740860.0,
      "step": 3708
    },
    {
      "entropy": 1.352414458990097,
      "epoch": 1.9531332280147446,
      "grad_norm": 0.27617254853248596,
      "learning_rate": 0.0001723001656613338,
      "loss": 0.20182929933071136,
      "mean_token_accuracy": 0.9149250984191895,
      "num_tokens": 45753196.0,
      "step": 3709
    },
    {
      "entropy": 1.3327330350875854,
      "epoch": 1.953659820958399,
      "grad_norm": 0.34774214029312134,
      "learning_rate": 0.00017228433854767859,
      "loss": 0.20589327812194824,
      "mean_token_accuracy": 0.9095378816127777,
      "num_tokens": 45765532.0,
      "step": 3710
    },
    {
      "entropy": 1.3821765780448914,
      "epoch": 1.9541864139020537,
      "grad_norm": 0.27642711997032166,
      "learning_rate": 0.00017226850773656522,
      "loss": 0.18724891543388367,
      "mean_token_accuracy": 0.9221665114164352,
      "num_tokens": 45777868.0,
      "step": 3711
    },
    {
      "entropy": 1.3774185180664062,
      "epoch": 1.9547130068457084,
      "grad_norm": 0.2937214970588684,
      "learning_rate": 0.00017225267322893345,
      "loss": 0.2190551459789276,
      "mean_token_accuracy": 0.9116322845220566,
      "num_tokens": 45790204.0,
      "step": 3712
    },
    {
      "entropy": 1.4014697074890137,
      "epoch": 1.9552395997893628,
      "grad_norm": 0.28920748829841614,
      "learning_rate": 0.0001722368350257233,
      "loss": 0.21408043801784515,
      "mean_token_accuracy": 0.9153937697410583,
      "num_tokens": 45802540.0,
      "step": 3713
    },
    {
      "entropy": 1.3347688913345337,
      "epoch": 1.9557661927330172,
      "grad_norm": 0.2642535865306854,
      "learning_rate": 0.000172220993127875,
      "loss": 0.18664056062698364,
      "mean_token_accuracy": 0.9244186133146286,
      "num_tokens": 45814876.0,
      "step": 3714
    },
    {
      "entropy": 1.4463875889778137,
      "epoch": 1.956292785676672,
      "grad_norm": 0.29163017868995667,
      "learning_rate": 0.000172205147536329,
      "loss": 0.1997067928314209,
      "mean_token_accuracy": 0.9191901385784149,
      "num_tokens": 45827212.0,
      "step": 3715
    },
    {
      "entropy": 1.356296420097351,
      "epoch": 1.9568193786203265,
      "grad_norm": 0.2423374354839325,
      "learning_rate": 0.000172189298252026,
      "loss": 0.17346829175949097,
      "mean_token_accuracy": 0.9300062209367752,
      "num_tokens": 45839548.0,
      "step": 3716
    },
    {
      "entropy": 1.4623576700687408,
      "epoch": 1.957345971563981,
      "grad_norm": 0.29987430572509766,
      "learning_rate": 0.00017217344527590674,
      "loss": 0.21269282698631287,
      "mean_token_accuracy": 0.910034641623497,
      "num_tokens": 45851884.0,
      "step": 3717
    },
    {
      "entropy": 1.3989585936069489,
      "epoch": 1.9578725645076356,
      "grad_norm": 0.2627890408039093,
      "learning_rate": 0.00017215758860891246,
      "loss": 0.19222907721996307,
      "mean_token_accuracy": 0.9272477477788925,
      "num_tokens": 45864220.0,
      "step": 3718
    },
    {
      "entropy": 1.3934078514575958,
      "epoch": 1.9583991574512902,
      "grad_norm": 0.29076218605041504,
      "learning_rate": 0.00017214172825198444,
      "loss": 0.19307014346122742,
      "mean_token_accuracy": 0.9236427992582321,
      "num_tokens": 45876556.0,
      "step": 3719
    },
    {
      "entropy": 1.3593478798866272,
      "epoch": 1.9589257503949447,
      "grad_norm": 0.2587772309780121,
      "learning_rate": 0.00017212586420606416,
      "loss": 0.19868242740631104,
      "mean_token_accuracy": 0.916473776102066,
      "num_tokens": 45888892.0,
      "step": 3720
    },
    {
      "entropy": 1.3617135286331177,
      "epoch": 1.9594523433385993,
      "grad_norm": 0.26174452900886536,
      "learning_rate": 0.00017210999647209344,
      "loss": 0.19469386339187622,
      "mean_token_accuracy": 0.9216127842664719,
      "num_tokens": 45901228.0,
      "step": 3721
    },
    {
      "entropy": 1.3824553787708282,
      "epoch": 1.959978936282254,
      "grad_norm": 0.2771162688732147,
      "learning_rate": 0.00017209412505101424,
      "loss": 0.19940043985843658,
      "mean_token_accuracy": 0.9135098457336426,
      "num_tokens": 45913564.0,
      "step": 3722
    },
    {
      "entropy": 1.396307110786438,
      "epoch": 1.9605055292259084,
      "grad_norm": 0.2760707139968872,
      "learning_rate": 0.00017207824994376874,
      "loss": 0.20229698717594147,
      "mean_token_accuracy": 0.9171424061059952,
      "num_tokens": 45925900.0,
      "step": 3723
    },
    {
      "entropy": 1.386707216501236,
      "epoch": 1.9610321221695628,
      "grad_norm": 0.290345698595047,
      "learning_rate": 0.00017206237115129937,
      "loss": 0.17667615413665771,
      "mean_token_accuracy": 0.9289303123950958,
      "num_tokens": 45938236.0,
      "step": 3724
    },
    {
      "entropy": 1.4016615748405457,
      "epoch": 1.9615587151132174,
      "grad_norm": 0.2932867407798767,
      "learning_rate": 0.00017204648867454874,
      "loss": 0.2111460119485855,
      "mean_token_accuracy": 0.915149986743927,
      "num_tokens": 45950572.0,
      "step": 3725
    },
    {
      "entropy": 1.3835887610912323,
      "epoch": 1.962085308056872,
      "grad_norm": 0.27471598982810974,
      "learning_rate": 0.0001720306025144597,
      "loss": 0.19538769125938416,
      "mean_token_accuracy": 0.9221490770578384,
      "num_tokens": 45962908.0,
      "step": 3726
    },
    {
      "entropy": 1.361358791589737,
      "epoch": 1.9626119010005265,
      "grad_norm": 0.3096008002758026,
      "learning_rate": 0.00017201471267197534,
      "loss": 0.22924600541591644,
      "mean_token_accuracy": 0.912015825510025,
      "num_tokens": 45975244.0,
      "step": 3727
    },
    {
      "entropy": 1.351409524679184,
      "epoch": 1.9631384939441812,
      "grad_norm": 0.27216240763664246,
      "learning_rate": 0.00017199881914803895,
      "loss": 0.19787034392356873,
      "mean_token_accuracy": 0.9182685315608978,
      "num_tokens": 45987580.0,
      "step": 3728
    },
    {
      "entropy": 1.4106302857398987,
      "epoch": 1.9636650868878358,
      "grad_norm": 0.28636589646339417,
      "learning_rate": 0.00017198292194359403,
      "loss": 0.22176605463027954,
      "mean_token_accuracy": 0.9129254519939423,
      "num_tokens": 45999916.0,
      "step": 3729
    },
    {
      "entropy": 1.351713240146637,
      "epoch": 1.9641916798314902,
      "grad_norm": 0.2533683180809021,
      "learning_rate": 0.00017196702105958428,
      "loss": 0.19331756234169006,
      "mean_token_accuracy": 0.9180753231048584,
      "num_tokens": 46012252.0,
      "step": 3730
    },
    {
      "entropy": 1.4381624460220337,
      "epoch": 1.9647182727751447,
      "grad_norm": 0.2788030207157135,
      "learning_rate": 0.00017195111649695368,
      "loss": 0.21007753908634186,
      "mean_token_accuracy": 0.9156930446624756,
      "num_tokens": 46024588.0,
      "step": 3731
    },
    {
      "entropy": 1.412168174982071,
      "epoch": 1.9652448657187995,
      "grad_norm": 0.31602713465690613,
      "learning_rate": 0.00017193520825664632,
      "loss": 0.17869442701339722,
      "mean_token_accuracy": 0.9241944849491119,
      "num_tokens": 46036924.0,
      "step": 3732
    },
    {
      "entropy": 1.3544166386127472,
      "epoch": 1.965771458662454,
      "grad_norm": 0.27904483675956726,
      "learning_rate": 0.0001719192963396067,
      "loss": 0.17792828381061554,
      "mean_token_accuracy": 0.9248538911342621,
      "num_tokens": 46049260.0,
      "step": 3733
    },
    {
      "entropy": 1.380949467420578,
      "epoch": 1.9662980516061084,
      "grad_norm": 0.266156405210495,
      "learning_rate": 0.0001719033807467793,
      "loss": 0.19987784326076508,
      "mean_token_accuracy": 0.9200113117694855,
      "num_tokens": 46061596.0,
      "step": 3734
    },
    {
      "entropy": 1.4363450706005096,
      "epoch": 1.966824644549763,
      "grad_norm": 0.25485169887542725,
      "learning_rate": 0.000171887461479109,
      "loss": 0.18918852508068085,
      "mean_token_accuracy": 0.926123857498169,
      "num_tokens": 46073932.0,
      "step": 3735
    },
    {
      "entropy": 1.5534420013427734,
      "epoch": 1.9673512374934177,
      "grad_norm": 0.3038926422595978,
      "learning_rate": 0.00017187153853754082,
      "loss": 0.1933830827474594,
      "mean_token_accuracy": 0.9212097078561783,
      "num_tokens": 46086268.0,
      "step": 3736
    },
    {
      "entropy": 1.4520435631275177,
      "epoch": 1.967877830437072,
      "grad_norm": 0.2602609395980835,
      "learning_rate": 0.00017185561192302004,
      "loss": 0.20847854018211365,
      "mean_token_accuracy": 0.914483830332756,
      "num_tokens": 46098604.0,
      "step": 3737
    },
    {
      "entropy": 1.4301496744155884,
      "epoch": 1.9684044233807267,
      "grad_norm": 0.28433549404144287,
      "learning_rate": 0.00017183968163649206,
      "loss": 0.20116156339645386,
      "mean_token_accuracy": 0.9151373505592346,
      "num_tokens": 46110940.0,
      "step": 3738
    },
    {
      "entropy": 1.4304525554180145,
      "epoch": 1.9689310163243814,
      "grad_norm": 0.24822618067264557,
      "learning_rate": 0.00017182374767890263,
      "loss": 0.20083917677402496,
      "mean_token_accuracy": 0.9184573292732239,
      "num_tokens": 46123276.0,
      "step": 3739
    },
    {
      "entropy": 1.4638838469982147,
      "epoch": 1.9694576092680358,
      "grad_norm": 0.2700670659542084,
      "learning_rate": 0.00017180781005119765,
      "loss": 0.20267093181610107,
      "mean_token_accuracy": 0.9162935614585876,
      "num_tokens": 46135612.0,
      "step": 3740
    },
    {
      "entropy": 1.3683336675167084,
      "epoch": 1.9699842022116902,
      "grad_norm": 0.26389580965042114,
      "learning_rate": 0.00017179186875432322,
      "loss": 0.1996886432170868,
      "mean_token_accuracy": 0.9172192066907883,
      "num_tokens": 46147948.0,
      "step": 3741
    },
    {
      "entropy": 1.411024570465088,
      "epoch": 1.9705107951553449,
      "grad_norm": 0.26272377371788025,
      "learning_rate": 0.00017177592378922566,
      "loss": 0.20570415258407593,
      "mean_token_accuracy": 0.9192255139350891,
      "num_tokens": 46160284.0,
      "step": 3742
    },
    {
      "entropy": 1.3969583213329315,
      "epoch": 1.9710373880989995,
      "grad_norm": 0.2648279666900635,
      "learning_rate": 0.0001717599751568516,
      "loss": 0.19804665446281433,
      "mean_token_accuracy": 0.9161637872457504,
      "num_tokens": 46172620.0,
      "step": 3743
    },
    {
      "entropy": 1.430682510137558,
      "epoch": 1.971563981042654,
      "grad_norm": 0.2636677920818329,
      "learning_rate": 0.00017174402285814776,
      "loss": 0.20058637857437134,
      "mean_token_accuracy": 0.9166569113731384,
      "num_tokens": 46184956.0,
      "step": 3744
    },
    {
      "entropy": 1.4314651489257812,
      "epoch": 1.9720905739863086,
      "grad_norm": 0.27471134066581726,
      "learning_rate": 0.00017172806689406118,
      "loss": 0.19406890869140625,
      "mean_token_accuracy": 0.9242938160896301,
      "num_tokens": 46197292.0,
      "step": 3745
    },
    {
      "entropy": 1.4192952811717987,
      "epoch": 1.9726171669299633,
      "grad_norm": 0.30075007677078247,
      "learning_rate": 0.00017171210726553903,
      "loss": 0.21153703331947327,
      "mean_token_accuracy": 0.91255022585392,
      "num_tokens": 46209628.0,
      "step": 3746
    },
    {
      "entropy": 1.4196560978889465,
      "epoch": 1.9731437598736177,
      "grad_norm": 0.2734670341014862,
      "learning_rate": 0.00017169614397352875,
      "loss": 0.19642943143844604,
      "mean_token_accuracy": 0.9212541729211807,
      "num_tokens": 46221964.0,
      "step": 3747
    },
    {
      "entropy": 1.3763989508152008,
      "epoch": 1.973670352817272,
      "grad_norm": 0.24970701336860657,
      "learning_rate": 0.00017168017701897802,
      "loss": 0.19371524453163147,
      "mean_token_accuracy": 0.9200150370597839,
      "num_tokens": 46234300.0,
      "step": 3748
    },
    {
      "entropy": 1.417275309562683,
      "epoch": 1.974196945760927,
      "grad_norm": 0.2494390606880188,
      "learning_rate": 0.00017166420640283466,
      "loss": 0.19952630996704102,
      "mean_token_accuracy": 0.9210073202848434,
      "num_tokens": 46246636.0,
      "step": 3749
    },
    {
      "entropy": 1.380375474691391,
      "epoch": 1.9747235387045814,
      "grad_norm": 0.2673463225364685,
      "learning_rate": 0.00017164823212604676,
      "loss": 0.19341374933719635,
      "mean_token_accuracy": 0.9220623522996902,
      "num_tokens": 46258972.0,
      "step": 3750
    },
    {
      "entropy": 1.3973598778247833,
      "epoch": 1.9752501316482358,
      "grad_norm": 0.28491443395614624,
      "learning_rate": 0.00017163225418956267,
      "loss": 0.21687844395637512,
      "mean_token_accuracy": 0.913015678524971,
      "num_tokens": 46271308.0,
      "step": 3751
    },
    {
      "entropy": 1.4152703285217285,
      "epoch": 1.9757767245918905,
      "grad_norm": 0.2640281021595001,
      "learning_rate": 0.00017161627259433082,
      "loss": 0.19280694425106049,
      "mean_token_accuracy": 0.9230195134878159,
      "num_tokens": 46283644.0,
      "step": 3752
    },
    {
      "entropy": 1.381377637386322,
      "epoch": 1.9763033175355451,
      "grad_norm": 0.2697671949863434,
      "learning_rate": 0.00017160028734130003,
      "loss": 0.18636345863342285,
      "mean_token_accuracy": 0.9252691715955734,
      "num_tokens": 46295980.0,
      "step": 3753
    },
    {
      "entropy": 1.3591782450675964,
      "epoch": 1.9768299104791995,
      "grad_norm": 0.2595213055610657,
      "learning_rate": 0.0001715842984314192,
      "loss": 0.19123241305351257,
      "mean_token_accuracy": 0.9189948439598083,
      "num_tokens": 46308316.0,
      "step": 3754
    },
    {
      "entropy": 1.3821331858634949,
      "epoch": 1.9773565034228542,
      "grad_norm": 0.27835193276405334,
      "learning_rate": 0.0001715683058656375,
      "loss": 0.2042919397354126,
      "mean_token_accuracy": 0.9147767573595047,
      "num_tokens": 46320652.0,
      "step": 3755
    },
    {
      "entropy": 1.3725225925445557,
      "epoch": 1.9778830963665088,
      "grad_norm": 0.2779839336872101,
      "learning_rate": 0.0001715523096449043,
      "loss": 0.20555046200752258,
      "mean_token_accuracy": 0.9161995947360992,
      "num_tokens": 46332988.0,
      "step": 3756
    },
    {
      "entropy": 1.3390932083129883,
      "epoch": 1.9784096893101633,
      "grad_norm": 0.2884056270122528,
      "learning_rate": 0.00017153630977016933,
      "loss": 0.21466374397277832,
      "mean_token_accuracy": 0.9098951071500778,
      "num_tokens": 46345324.0,
      "step": 3757
    },
    {
      "entropy": 1.304327368736267,
      "epoch": 1.9789362822538177,
      "grad_norm": 0.28233397006988525,
      "learning_rate": 0.00017152030624238224,
      "loss": 0.2030801624059677,
      "mean_token_accuracy": 0.9174643605947495,
      "num_tokens": 46357660.0,
      "step": 3758
    },
    {
      "entropy": 1.3582344055175781,
      "epoch": 1.9794628751974723,
      "grad_norm": 0.2734123170375824,
      "learning_rate": 0.00017150429906249314,
      "loss": 0.20131902396678925,
      "mean_token_accuracy": 0.9201581329107285,
      "num_tokens": 46369996.0,
      "step": 3759
    },
    {
      "entropy": 1.3707594871520996,
      "epoch": 1.979989468141127,
      "grad_norm": 0.28123629093170166,
      "learning_rate": 0.0001714882882314523,
      "loss": 0.22594565153121948,
      "mean_token_accuracy": 0.9095589369535446,
      "num_tokens": 46382332.0,
      "step": 3760
    },
    {
      "entropy": 1.325995683670044,
      "epoch": 1.9805160610847814,
      "grad_norm": 0.2784866690635681,
      "learning_rate": 0.00017147227375021015,
      "loss": 0.21171274781227112,
      "mean_token_accuracy": 0.9164798855781555,
      "num_tokens": 46394668.0,
      "step": 3761
    },
    {
      "entropy": 1.3826212286949158,
      "epoch": 1.981042654028436,
      "grad_norm": 0.27320340275764465,
      "learning_rate": 0.0001714562556197174,
      "loss": 0.18267717957496643,
      "mean_token_accuracy": 0.9259230494499207,
      "num_tokens": 46407004.0,
      "step": 3762
    },
    {
      "entropy": 1.2950341999530792,
      "epoch": 1.9815692469720907,
      "grad_norm": 0.28982457518577576,
      "learning_rate": 0.00017144023384092495,
      "loss": 0.20509836077690125,
      "mean_token_accuracy": 0.9163513332605362,
      "num_tokens": 46419340.0,
      "step": 3763
    },
    {
      "entropy": 1.3356437385082245,
      "epoch": 1.9820958399157451,
      "grad_norm": 0.2567256689071655,
      "learning_rate": 0.00017142420841478393,
      "loss": 0.18219546973705292,
      "mean_token_accuracy": 0.9224260747432709,
      "num_tokens": 46431676.0,
      "step": 3764
    },
    {
      "entropy": 1.3211258351802826,
      "epoch": 1.9826224328593995,
      "grad_norm": 0.25619199872016907,
      "learning_rate": 0.00017140817934224565,
      "loss": 0.20122160017490387,
      "mean_token_accuracy": 0.9165990948677063,
      "num_tokens": 46444012.0,
      "step": 3765
    },
    {
      "entropy": 1.2932257950305939,
      "epoch": 1.9831490258030542,
      "grad_norm": 0.2684885561466217,
      "learning_rate": 0.00017139214662426167,
      "loss": 0.1878378689289093,
      "mean_token_accuracy": 0.9228431284427643,
      "num_tokens": 46456348.0,
      "step": 3766
    },
    {
      "entropy": 1.2775112092494965,
      "epoch": 1.9836756187467088,
      "grad_norm": 0.2664264738559723,
      "learning_rate": 0.00017137611026178375,
      "loss": 0.19962744414806366,
      "mean_token_accuracy": 0.9182207137346268,
      "num_tokens": 46468684.0,
      "step": 3767
    },
    {
      "entropy": 1.2961704730987549,
      "epoch": 1.9842022116903633,
      "grad_norm": 0.27163180708885193,
      "learning_rate": 0.00017136007025576392,
      "loss": 0.2212834358215332,
      "mean_token_accuracy": 0.9135020226240158,
      "num_tokens": 46481020.0,
      "step": 3768
    },
    {
      "entropy": 1.3635116219520569,
      "epoch": 1.984728804634018,
      "grad_norm": 0.2555164396762848,
      "learning_rate": 0.00017134402660715436,
      "loss": 0.19105497002601624,
      "mean_token_accuracy": 0.92401522397995,
      "num_tokens": 46493356.0,
      "step": 3769
    },
    {
      "entropy": 1.2995677888393402,
      "epoch": 1.9852553975776726,
      "grad_norm": 0.2641841173171997,
      "learning_rate": 0.00017132797931690747,
      "loss": 0.19776709377765656,
      "mean_token_accuracy": 0.9210506677627563,
      "num_tokens": 46505692.0,
      "step": 3770
    },
    {
      "entropy": 1.2360694408416748,
      "epoch": 1.985781990521327,
      "grad_norm": 0.2550431489944458,
      "learning_rate": 0.0001713119283859759,
      "loss": 0.18957817554473877,
      "mean_token_accuracy": 0.9171388000249863,
      "num_tokens": 46518028.0,
      "step": 3771
    },
    {
      "entropy": 1.273363709449768,
      "epoch": 1.9863085834649814,
      "grad_norm": 0.25279700756073,
      "learning_rate": 0.00017129587381531247,
      "loss": 0.19003725051879883,
      "mean_token_accuracy": 0.9217302948236465,
      "num_tokens": 46530364.0,
      "step": 3772
    },
    {
      "entropy": 1.2529352307319641,
      "epoch": 1.9868351764086363,
      "grad_norm": 0.2749396860599518,
      "learning_rate": 0.0001712798156058703,
      "loss": 0.1858321577310562,
      "mean_token_accuracy": 0.9264263361692429,
      "num_tokens": 46542700.0,
      "step": 3773
    },
    {
      "entropy": 1.2869006991386414,
      "epoch": 1.9873617693522907,
      "grad_norm": 0.28312820196151733,
      "learning_rate": 0.00017126375375860263,
      "loss": 0.21077433228492737,
      "mean_token_accuracy": 0.9111466556787491,
      "num_tokens": 46555036.0,
      "step": 3774
    },
    {
      "entropy": 1.2937982976436615,
      "epoch": 1.9878883622959451,
      "grad_norm": 0.29525041580200195,
      "learning_rate": 0.00017124768827446297,
      "loss": 0.21481990814208984,
      "mean_token_accuracy": 0.9122337698936462,
      "num_tokens": 46567372.0,
      "step": 3775
    },
    {
      "entropy": 1.287131816148758,
      "epoch": 1.9884149552395998,
      "grad_norm": 0.2926379442214966,
      "learning_rate": 0.00017123161915440503,
      "loss": 0.20220476388931274,
      "mean_token_accuracy": 0.9196414649486542,
      "num_tokens": 46579708.0,
      "step": 3776
    },
    {
      "entropy": 1.2510655224323273,
      "epoch": 1.9889415481832544,
      "grad_norm": 0.307608962059021,
      "learning_rate": 0.00017121554639938272,
      "loss": 0.1835639774799347,
      "mean_token_accuracy": 0.9256336688995361,
      "num_tokens": 46592044.0,
      "step": 3777
    },
    {
      "entropy": 1.198952555656433,
      "epoch": 1.9894681411269088,
      "grad_norm": 0.2576758563518524,
      "learning_rate": 0.00017119947001035027,
      "loss": 0.20444577932357788,
      "mean_token_accuracy": 0.9202164262533188,
      "num_tokens": 46604380.0,
      "step": 3778
    },
    {
      "entropy": 1.30607271194458,
      "epoch": 1.9899947340705635,
      "grad_norm": 0.2850700914859772,
      "learning_rate": 0.00017118338998826197,
      "loss": 0.1958879828453064,
      "mean_token_accuracy": 0.9212403148412704,
      "num_tokens": 46616716.0,
      "step": 3779
    },
    {
      "entropy": 1.1765548586845398,
      "epoch": 1.9905213270142181,
      "grad_norm": 0.28321370482444763,
      "learning_rate": 0.00017116730633407238,
      "loss": 0.21571975946426392,
      "mean_token_accuracy": 0.9140463322401047,
      "num_tokens": 46629052.0,
      "step": 3780
    },
    {
      "entropy": 1.2251535058021545,
      "epoch": 1.9910479199578726,
      "grad_norm": 0.2587471604347229,
      "learning_rate": 0.00017115121904873637,
      "loss": 0.19872674345970154,
      "mean_token_accuracy": 0.9179710447788239,
      "num_tokens": 46641388.0,
      "step": 3781
    },
    {
      "entropy": 1.2262464761734009,
      "epoch": 1.991574512901527,
      "grad_norm": 0.26432153582572937,
      "learning_rate": 0.00017113512813320887,
      "loss": 0.18742287158966064,
      "mean_token_accuracy": 0.9234320670366287,
      "num_tokens": 46653724.0,
      "step": 3782
    },
    {
      "entropy": 1.2304887175559998,
      "epoch": 1.9921011058451816,
      "grad_norm": 0.3122429847717285,
      "learning_rate": 0.00017111903358844513,
      "loss": 0.20218738913536072,
      "mean_token_accuracy": 0.9151853322982788,
      "num_tokens": 46666060.0,
      "step": 3783
    },
    {
      "entropy": 1.254802256822586,
      "epoch": 1.9926276987888363,
      "grad_norm": 0.27966201305389404,
      "learning_rate": 0.0001711029354154006,
      "loss": 0.2041388899087906,
      "mean_token_accuracy": 0.9164295345544815,
      "num_tokens": 46678396.0,
      "step": 3784
    },
    {
      "entropy": 1.2461448013782501,
      "epoch": 1.9931542917324907,
      "grad_norm": 0.2727980315685272,
      "learning_rate": 0.00017108683361503093,
      "loss": 0.2023455798625946,
      "mean_token_accuracy": 0.9165138304233551,
      "num_tokens": 46690732.0,
      "step": 3785
    },
    {
      "entropy": 1.2635038495063782,
      "epoch": 1.9936808846761453,
      "grad_norm": 0.27869847416877747,
      "learning_rate": 0.000171070728188292,
      "loss": 0.21800480782985687,
      "mean_token_accuracy": 0.9098168015480042,
      "num_tokens": 46703068.0,
      "step": 3786
    },
    {
      "entropy": 1.3022080957889557,
      "epoch": 1.9942074776198,
      "grad_norm": 0.2651629149913788,
      "learning_rate": 0.00017105461913613986,
      "loss": 0.2050429880619049,
      "mean_token_accuracy": 0.9191136509180069,
      "num_tokens": 46715404.0,
      "step": 3787
    },
    {
      "entropy": 1.316438913345337,
      "epoch": 1.9947340705634544,
      "grad_norm": 0.2717432379722595,
      "learning_rate": 0.00017103850645953085,
      "loss": 0.20991484820842743,
      "mean_token_accuracy": 0.9131201207637787,
      "num_tokens": 46727740.0,
      "step": 3788
    },
    {
      "entropy": 1.3070935606956482,
      "epoch": 1.9952606635071088,
      "grad_norm": 0.28713029623031616,
      "learning_rate": 0.00017102239015942148,
      "loss": 0.22580859065055847,
      "mean_token_accuracy": 0.9063994288444519,
      "num_tokens": 46740076.0,
      "step": 3789
    },
    {
      "entropy": 1.2703860700130463,
      "epoch": 1.9957872564507637,
      "grad_norm": 0.2608626186847687,
      "learning_rate": 0.00017100627023676848,
      "loss": 0.20214328169822693,
      "mean_token_accuracy": 0.9189681559801102,
      "num_tokens": 46752412.0,
      "step": 3790
    },
    {
      "entropy": 1.248716801404953,
      "epoch": 1.9963138493944181,
      "grad_norm": 0.2430690973997116,
      "learning_rate": 0.00017099014669252877,
      "loss": 0.18712979555130005,
      "mean_token_accuracy": 0.9251755028963089,
      "num_tokens": 46764748.0,
      "step": 3791
    },
    {
      "entropy": 1.222788542509079,
      "epoch": 1.9968404423380726,
      "grad_norm": 0.277614027261734,
      "learning_rate": 0.0001709740195276595,
      "loss": 0.21778146922588348,
      "mean_token_accuracy": 0.9087530225515366,
      "num_tokens": 46777084.0,
      "step": 3792
    },
    {
      "entropy": 1.204005628824234,
      "epoch": 1.9973670352817272,
      "grad_norm": 0.2973788380622864,
      "learning_rate": 0.00017095788874311814,
      "loss": 0.19019125401973724,
      "mean_token_accuracy": 0.9220655113458633,
      "num_tokens": 46789420.0,
      "step": 3793
    },
    {
      "entropy": 1.2420277893543243,
      "epoch": 1.9978936282253819,
      "grad_norm": 0.25097668170928955,
      "learning_rate": 0.00017094175433986214,
      "loss": 0.197739377617836,
      "mean_token_accuracy": 0.9215272665023804,
      "num_tokens": 46801756.0,
      "step": 3794
    },
    {
      "entropy": 1.2404221594333649,
      "epoch": 1.9984202211690363,
      "grad_norm": 0.2478831261396408,
      "learning_rate": 0.0001709256163188494,
      "loss": 0.19220499694347382,
      "mean_token_accuracy": 0.9218998998403549,
      "num_tokens": 46814092.0,
      "step": 3795
    },
    {
      "entropy": 1.2480852603912354,
      "epoch": 1.998946814112691,
      "grad_norm": 0.25873619318008423,
      "learning_rate": 0.000170909474681038,
      "loss": 0.18990661203861237,
      "mean_token_accuracy": 0.9263609051704407,
      "num_tokens": 46826428.0,
      "step": 3796
    },
    {
      "entropy": 1.1696776449680328,
      "epoch": 1.9994734070563456,
      "grad_norm": 0.2425045371055603,
      "learning_rate": 0.00017089332942738604,
      "loss": 0.18430982530117035,
      "mean_token_accuracy": 0.9185814708471298,
      "num_tokens": 46838764.0,
      "step": 3797
    },
    {
      "entropy": 1.269264966249466,
      "epoch": 2.0,
      "grad_norm": 0.29400986433029175,
      "learning_rate": 0.000170877180558852,
      "loss": 0.1966848373413086,
      "mean_token_accuracy": 0.9257184416055679,
      "num_tokens": 46850072.0,
      "step": 3798
    },
    {
      "epoch": 2.0,
      "eval_entropy": 1.244152147623035,
      "eval_loss": 0.20280839502811432,
      "eval_mean_token_accuracy": 0.9176469020557253,
      "eval_num_tokens": 46850072.0,
      "eval_runtime": 665.3153,
      "eval_samples_per_second": 8.563,
      "eval_steps_per_second": 2.854,
      "step": 3798
    },
    {
      "entropy": 1.2631646990776062,
      "epoch": 2.0005265929436544,
      "grad_norm": 0.25198251008987427,
      "learning_rate": 0.00017086102807639458,
      "loss": 0.1870889663696289,
      "mean_token_accuracy": 0.9224293082952499,
      "num_tokens": 46862408.0,
      "step": 3799
    },
    {
      "entropy": 1.1980132162570953,
      "epoch": 2.0010531858873093,
      "grad_norm": 0.2404579371213913,
      "learning_rate": 0.00017084487198097266,
      "loss": 0.18207380175590515,
      "mean_token_accuracy": 0.9251429438591003,
      "num_tokens": 46874744.0,
      "step": 3800
    },
    {
      "entropy": 1.237736165523529,
      "epoch": 2.0015797788309637,
      "grad_norm": 0.27258867025375366,
      "learning_rate": 0.00017082871227354533,
      "loss": 0.18785671889781952,
      "mean_token_accuracy": 0.9198215752840042,
      "num_tokens": 46887080.0,
      "step": 3801
    },
    {
      "entropy": 1.245652198791504,
      "epoch": 2.002106371774618,
      "grad_norm": 0.2836478352546692,
      "learning_rate": 0.0001708125489550719,
      "loss": 0.19735261797904968,
      "mean_token_accuracy": 0.9185144305229187,
      "num_tokens": 46899416.0,
      "step": 3802
    },
    {
      "entropy": 1.2350683510303497,
      "epoch": 2.0026329647182726,
      "grad_norm": 0.28520843386650085,
      "learning_rate": 0.00017079638202651185,
      "loss": 0.181529700756073,
      "mean_token_accuracy": 0.9275423139333725,
      "num_tokens": 46911752.0,
      "step": 3803
    },
    {
      "entropy": 1.254747599363327,
      "epoch": 2.0031595576619274,
      "grad_norm": 0.2920467257499695,
      "learning_rate": 0.00017078021148882498,
      "loss": 0.2086395025253296,
      "mean_token_accuracy": 0.9195475280284882,
      "num_tokens": 46924088.0,
      "step": 3804
    },
    {
      "entropy": 1.1598127484321594,
      "epoch": 2.003686150605582,
      "grad_norm": 0.2568848431110382,
      "learning_rate": 0.00017076403734297112,
      "loss": 0.1833052933216095,
      "mean_token_accuracy": 0.9211831390857697,
      "num_tokens": 46936424.0,
      "step": 3805
    },
    {
      "entropy": 1.2467805445194244,
      "epoch": 2.0042127435492363,
      "grad_norm": 0.29362115263938904,
      "learning_rate": 0.00017074785958991063,
      "loss": 0.20824174582958221,
      "mean_token_accuracy": 0.9118938595056534,
      "num_tokens": 46948760.0,
      "step": 3806
    },
    {
      "entropy": 1.231805145740509,
      "epoch": 2.004739336492891,
      "grad_norm": 0.29164865612983704,
      "learning_rate": 0.0001707316782306037,
      "loss": 0.2124311774969101,
      "mean_token_accuracy": 0.9105638265609741,
      "num_tokens": 46961096.0,
      "step": 3807
    },
    {
      "entropy": 1.1256294250488281,
      "epoch": 2.0052659294365456,
      "grad_norm": 0.2622910141944885,
      "learning_rate": 0.00017071549326601107,
      "loss": 0.1925264447927475,
      "mean_token_accuracy": 0.9230650365352631,
      "num_tokens": 46973432.0,
      "step": 3808
    },
    {
      "entropy": 1.2006166875362396,
      "epoch": 2.0057925223802,
      "grad_norm": 0.28640827536582947,
      "learning_rate": 0.00017069930469709344,
      "loss": 0.20924372971057892,
      "mean_token_accuracy": 0.9144559353590012,
      "num_tokens": 46985768.0,
      "step": 3809
    },
    {
      "entropy": 1.2369438111782074,
      "epoch": 2.006319115323855,
      "grad_norm": 0.25838255882263184,
      "learning_rate": 0.0001706831125248119,
      "loss": 0.18743355572223663,
      "mean_token_accuracy": 0.9234079271554947,
      "num_tokens": 46998104.0,
      "step": 3810
    },
    {
      "entropy": 1.17817023396492,
      "epoch": 2.0068457082675093,
      "grad_norm": 0.2704300880432129,
      "learning_rate": 0.00017066691675012764,
      "loss": 0.18841452896595,
      "mean_token_accuracy": 0.9241382032632828,
      "num_tokens": 47010440.0,
      "step": 3811
    },
    {
      "entropy": 1.2373998761177063,
      "epoch": 2.0073723012111637,
      "grad_norm": 0.2672933042049408,
      "learning_rate": 0.00017065071737400213,
      "loss": 0.1963019222021103,
      "mean_token_accuracy": 0.9170713722705841,
      "num_tokens": 47022776.0,
      "step": 3812
    },
    {
      "entropy": 1.140851616859436,
      "epoch": 2.007898894154818,
      "grad_norm": 0.25465431809425354,
      "learning_rate": 0.00017063451439739704,
      "loss": 0.18710798025131226,
      "mean_token_accuracy": 0.9227855503559113,
      "num_tokens": 47035112.0,
      "step": 3813
    },
    {
      "entropy": 1.1974302530288696,
      "epoch": 2.008425487098473,
      "grad_norm": 0.2557191550731659,
      "learning_rate": 0.0001706183078212742,
      "loss": 0.17992272973060608,
      "mean_token_accuracy": 0.9254202246665955,
      "num_tokens": 47047448.0,
      "step": 3814
    },
    {
      "entropy": 1.2048839926719666,
      "epoch": 2.0089520800421274,
      "grad_norm": 0.27587470412254333,
      "learning_rate": 0.00017060209764659576,
      "loss": 0.18894129991531372,
      "mean_token_accuracy": 0.9178142994642258,
      "num_tokens": 47059784.0,
      "step": 3815
    },
    {
      "entropy": 1.2109751403331757,
      "epoch": 2.009478672985782,
      "grad_norm": 0.2797783613204956,
      "learning_rate": 0.000170585883874324,
      "loss": 0.19585882127285004,
      "mean_token_accuracy": 0.9170439839363098,
      "num_tokens": 47072120.0,
      "step": 3816
    },
    {
      "entropy": 1.1850585043430328,
      "epoch": 2.0100052659294367,
      "grad_norm": 0.24356874823570251,
      "learning_rate": 0.00017056966650542146,
      "loss": 0.1750049591064453,
      "mean_token_accuracy": 0.9231248944997787,
      "num_tokens": 47084456.0,
      "step": 3817
    },
    {
      "entropy": 1.2129537165164948,
      "epoch": 2.010531858873091,
      "grad_norm": 0.2753424644470215,
      "learning_rate": 0.00017055344554085081,
      "loss": 0.1915561556816101,
      "mean_token_accuracy": 0.9132361859083176,
      "num_tokens": 47096792.0,
      "step": 3818
    },
    {
      "entropy": 1.1568632423877716,
      "epoch": 2.0110584518167456,
      "grad_norm": 0.25440388917922974,
      "learning_rate": 0.00017053722098157504,
      "loss": 0.18248328566551208,
      "mean_token_accuracy": 0.9241259098052979,
      "num_tokens": 47109128.0,
      "step": 3819
    },
    {
      "entropy": 1.2225323617458344,
      "epoch": 2.0115850447604,
      "grad_norm": 0.2667352259159088,
      "learning_rate": 0.00017052099282855728,
      "loss": 0.1801188588142395,
      "mean_token_accuracy": 0.9245119541883469,
      "num_tokens": 47121464.0,
      "step": 3820
    },
    {
      "entropy": 1.2486798465251923,
      "epoch": 2.012111637704055,
      "grad_norm": 0.2779337465763092,
      "learning_rate": 0.00017050476108276095,
      "loss": 0.1984904706478119,
      "mean_token_accuracy": 0.9179847091436386,
      "num_tokens": 47133800.0,
      "step": 3821
    },
    {
      "entropy": 1.2059212625026703,
      "epoch": 2.0126382306477093,
      "grad_norm": 0.2828848659992218,
      "learning_rate": 0.0001704885257451496,
      "loss": 0.19344468414783478,
      "mean_token_accuracy": 0.9199190735816956,
      "num_tokens": 47146136.0,
      "step": 3822
    },
    {
      "entropy": 1.2063279151916504,
      "epoch": 2.0131648235913637,
      "grad_norm": 0.24573519825935364,
      "learning_rate": 0.000170472286816687,
      "loss": 0.17926183342933655,
      "mean_token_accuracy": 0.9242213070392609,
      "num_tokens": 47158472.0,
      "step": 3823
    },
    {
      "entropy": 1.2062674760818481,
      "epoch": 2.0136914165350186,
      "grad_norm": 0.2900843620300293,
      "learning_rate": 0.00017045604429833722,
      "loss": 0.18898425996303558,
      "mean_token_accuracy": 0.9269550293684006,
      "num_tokens": 47170808.0,
      "step": 3824
    },
    {
      "entropy": 1.2762680351734161,
      "epoch": 2.014218009478673,
      "grad_norm": 0.2975054979324341,
      "learning_rate": 0.00017043979819106447,
      "loss": 0.20459939539432526,
      "mean_token_accuracy": 0.918785035610199,
      "num_tokens": 47183144.0,
      "step": 3825
    },
    {
      "entropy": 1.257956087589264,
      "epoch": 2.0147446024223274,
      "grad_norm": 0.27851611375808716,
      "learning_rate": 0.00017042354849583312,
      "loss": 0.17341488599777222,
      "mean_token_accuracy": 0.9255774170160294,
      "num_tokens": 47195480.0,
      "step": 3826
    },
    {
      "entropy": 1.2822742462158203,
      "epoch": 2.0152711953659823,
      "grad_norm": 0.2700733244419098,
      "learning_rate": 0.0001704072952136079,
      "loss": 0.18918979167938232,
      "mean_token_accuracy": 0.921903595328331,
      "num_tokens": 47207816.0,
      "step": 3827
    },
    {
      "entropy": 1.26444673538208,
      "epoch": 2.0157977883096367,
      "grad_norm": 0.28335216641426086,
      "learning_rate": 0.00017039103834535366,
      "loss": 0.1854114681482315,
      "mean_token_accuracy": 0.9210407435894012,
      "num_tokens": 47220152.0,
      "step": 3828
    },
    {
      "entropy": 1.2844519019126892,
      "epoch": 2.016324381253291,
      "grad_norm": 0.291987806558609,
      "learning_rate": 0.00017037477789203545,
      "loss": 0.172478586435318,
      "mean_token_accuracy": 0.9290066659450531,
      "num_tokens": 47232488.0,
      "step": 3829
    },
    {
      "entropy": 1.1888793110847473,
      "epoch": 2.0168509741969456,
      "grad_norm": 0.2763528823852539,
      "learning_rate": 0.00017035851385461856,
      "loss": 0.1754818558692932,
      "mean_token_accuracy": 0.9284273386001587,
      "num_tokens": 47244824.0,
      "step": 3830
    },
    {
      "entropy": 1.2850163877010345,
      "epoch": 2.0173775671406005,
      "grad_norm": 0.2939310371875763,
      "learning_rate": 0.00017034224623406848,
      "loss": 0.1882777065038681,
      "mean_token_accuracy": 0.9187758266925812,
      "num_tokens": 47257160.0,
      "step": 3831
    },
    {
      "entropy": 1.278912514448166,
      "epoch": 2.017904160084255,
      "grad_norm": 0.26976412534713745,
      "learning_rate": 0.00017032597503135097,
      "loss": 0.18176186084747314,
      "mean_token_accuracy": 0.9222287982702255,
      "num_tokens": 47269496.0,
      "step": 3832
    },
    {
      "entropy": 1.2571650743484497,
      "epoch": 2.0184307530279093,
      "grad_norm": 0.2843485474586487,
      "learning_rate": 0.0001703097002474319,
      "loss": 0.1895454078912735,
      "mean_token_accuracy": 0.9216977059841156,
      "num_tokens": 47281832.0,
      "step": 3833
    },
    {
      "entropy": 1.239231675863266,
      "epoch": 2.018957345971564,
      "grad_norm": 0.2837580442428589,
      "learning_rate": 0.00017029342188327746,
      "loss": 0.18688803911209106,
      "mean_token_accuracy": 0.9240525960922241,
      "num_tokens": 47294168.0,
      "step": 3834
    },
    {
      "entropy": 1.199841469526291,
      "epoch": 2.0194839389152186,
      "grad_norm": 0.2625201344490051,
      "learning_rate": 0.00017027713993985399,
      "loss": 0.1796603798866272,
      "mean_token_accuracy": 0.9272690117359161,
      "num_tokens": 47306504.0,
      "step": 3835
    },
    {
      "entropy": 1.266678363084793,
      "epoch": 2.020010531858873,
      "grad_norm": 0.2874893844127655,
      "learning_rate": 0.00017026085441812803,
      "loss": 0.20230376720428467,
      "mean_token_accuracy": 0.9180373549461365,
      "num_tokens": 47318840.0,
      "step": 3836
    },
    {
      "entropy": 1.3089313209056854,
      "epoch": 2.0205371248025275,
      "grad_norm": 0.27973634004592896,
      "learning_rate": 0.00017024456531906636,
      "loss": 0.19552314281463623,
      "mean_token_accuracy": 0.9171228557825089,
      "num_tokens": 47331176.0,
      "step": 3837
    },
    {
      "entropy": 1.313205361366272,
      "epoch": 2.0210637177461823,
      "grad_norm": 0.2686075270175934,
      "learning_rate": 0.000170228272643636,
      "loss": 0.18762610852718353,
      "mean_token_accuracy": 0.9212076663970947,
      "num_tokens": 47343512.0,
      "step": 3838
    },
    {
      "entropy": 1.3670341670513153,
      "epoch": 2.0215903106898367,
      "grad_norm": 0.30073225498199463,
      "learning_rate": 0.00017021197639280412,
      "loss": 0.19591176509857178,
      "mean_token_accuracy": 0.9202536344528198,
      "num_tokens": 47355848.0,
      "step": 3839
    },
    {
      "entropy": 1.3341005444526672,
      "epoch": 2.022116903633491,
      "grad_norm": 0.26529407501220703,
      "learning_rate": 0.00017019567656753812,
      "loss": 0.1831834614276886,
      "mean_token_accuracy": 0.9253407269716263,
      "num_tokens": 47368184.0,
      "step": 3840
    },
    {
      "entropy": 1.3194755613803864,
      "epoch": 2.022643496577146,
      "grad_norm": 0.282293438911438,
      "learning_rate": 0.00017017937316880568,
      "loss": 0.20243722200393677,
      "mean_token_accuracy": 0.9165725558996201,
      "num_tokens": 47380520.0,
      "step": 3841
    },
    {
      "entropy": 1.4211511313915253,
      "epoch": 2.0231700895208005,
      "grad_norm": 0.29805392026901245,
      "learning_rate": 0.0001701630661975746,
      "loss": 0.18482814729213715,
      "mean_token_accuracy": 0.9222871661186218,
      "num_tokens": 47392856.0,
      "step": 3842
    },
    {
      "entropy": 1.3491427600383759,
      "epoch": 2.023696682464455,
      "grad_norm": 0.2905563414096832,
      "learning_rate": 0.00017014675565481293,
      "loss": 0.20305126905441284,
      "mean_token_accuracy": 0.9208891540765762,
      "num_tokens": 47405192.0,
      "step": 3843
    },
    {
      "entropy": 1.3559309840202332,
      "epoch": 2.0242232754081093,
      "grad_norm": 0.2943587005138397,
      "learning_rate": 0.00017013044154148894,
      "loss": 0.1753818392753601,
      "mean_token_accuracy": 0.9277485013008118,
      "num_tokens": 47417528.0,
      "step": 3844
    },
    {
      "entropy": 1.3276709914207458,
      "epoch": 2.024749868351764,
      "grad_norm": 0.29700222611427307,
      "learning_rate": 0.00017011412385857112,
      "loss": 0.2055238038301468,
      "mean_token_accuracy": 0.9166817218065262,
      "num_tokens": 47429864.0,
      "step": 3845
    },
    {
      "entropy": 1.298693597316742,
      "epoch": 2.0252764612954186,
      "grad_norm": 0.2802084982395172,
      "learning_rate": 0.00017009780260702812,
      "loss": 0.19370201230049133,
      "mean_token_accuracy": 0.9228375554084778,
      "num_tokens": 47442200.0,
      "step": 3846
    },
    {
      "entropy": 1.3408599495887756,
      "epoch": 2.025803054239073,
      "grad_norm": 0.28054699301719666,
      "learning_rate": 0.00017008147778782885,
      "loss": 0.18260030448436737,
      "mean_token_accuracy": 0.9232022166252136,
      "num_tokens": 47454536.0,
      "step": 3847
    },
    {
      "entropy": 1.2663397789001465,
      "epoch": 2.026329647182728,
      "grad_norm": 0.25765085220336914,
      "learning_rate": 0.00017006514940194246,
      "loss": 0.18359926342964172,
      "mean_token_accuracy": 0.9257479310035706,
      "num_tokens": 47466872.0,
      "step": 3848
    },
    {
      "entropy": 1.316208690404892,
      "epoch": 2.0268562401263823,
      "grad_norm": 0.2987428605556488,
      "learning_rate": 0.0001700488174503382,
      "loss": 0.19594340026378632,
      "mean_token_accuracy": 0.9216845035552979,
      "num_tokens": 47479208.0,
      "step": 3849
    },
    {
      "entropy": 1.3389540016651154,
      "epoch": 2.0273828330700367,
      "grad_norm": 0.27578049898147583,
      "learning_rate": 0.00017003248193398564,
      "loss": 0.19140341877937317,
      "mean_token_accuracy": 0.9192010909318924,
      "num_tokens": 47491544.0,
      "step": 3850
    },
    {
      "entropy": 1.3227673172950745,
      "epoch": 2.0279094260136916,
      "grad_norm": 0.2912873327732086,
      "learning_rate": 0.00017001614285385455,
      "loss": 0.20706209540367126,
      "mean_token_accuracy": 0.9152367115020752,
      "num_tokens": 47503880.0,
      "step": 3851
    },
    {
      "entropy": 1.2799009382724762,
      "epoch": 2.028436018957346,
      "grad_norm": 0.2770322561264038,
      "learning_rate": 0.00016999980021091483,
      "loss": 0.19841603934764862,
      "mean_token_accuracy": 0.9199351519346237,
      "num_tokens": 47516216.0,
      "step": 3852
    },
    {
      "entropy": 1.3126654028892517,
      "epoch": 2.0289626119010005,
      "grad_norm": 0.3020882308483124,
      "learning_rate": 0.0001699834540061367,
      "loss": 0.20801658928394318,
      "mean_token_accuracy": 0.9128816872835159,
      "num_tokens": 47528552.0,
      "step": 3853
    },
    {
      "entropy": 1.2508117258548737,
      "epoch": 2.029489204844655,
      "grad_norm": 0.24238452315330505,
      "learning_rate": 0.00016996710424049051,
      "loss": 0.1554844081401825,
      "mean_token_accuracy": 0.9381654858589172,
      "num_tokens": 47540888.0,
      "step": 3854
    },
    {
      "entropy": 1.278584361076355,
      "epoch": 2.0300157977883098,
      "grad_norm": 0.26651400327682495,
      "learning_rate": 0.00016995075091494685,
      "loss": 0.18668919801712036,
      "mean_token_accuracy": 0.9253730177879333,
      "num_tokens": 47553224.0,
      "step": 3855
    },
    {
      "entropy": 1.2722339630126953,
      "epoch": 2.030542390731964,
      "grad_norm": 0.285241037607193,
      "learning_rate": 0.00016993439403047652,
      "loss": 0.21573114395141602,
      "mean_token_accuracy": 0.9168124496936798,
      "num_tokens": 47565560.0,
      "step": 3856
    },
    {
      "entropy": 1.2240016758441925,
      "epoch": 2.0310689836756186,
      "grad_norm": 0.2548234760761261,
      "learning_rate": 0.00016991803358805058,
      "loss": 0.1847495436668396,
      "mean_token_accuracy": 0.9236398935317993,
      "num_tokens": 47577896.0,
      "step": 3857
    },
    {
      "entropy": 1.2316696643829346,
      "epoch": 2.0315955766192735,
      "grad_norm": 0.28398963809013367,
      "learning_rate": 0.00016990166958864015,
      "loss": 0.19505375623703003,
      "mean_token_accuracy": 0.9168769717216492,
      "num_tokens": 47590232.0,
      "step": 3858
    },
    {
      "entropy": 1.1961296796798706,
      "epoch": 2.032122169562928,
      "grad_norm": 0.27527958154678345,
      "learning_rate": 0.0001698853020332168,
      "loss": 0.17813697457313538,
      "mean_token_accuracy": 0.9261271953582764,
      "num_tokens": 47602568.0,
      "step": 3859
    },
    {
      "entropy": 1.273693025112152,
      "epoch": 2.0326487625065823,
      "grad_norm": 0.3149412274360657,
      "learning_rate": 0.00016986893092275205,
      "loss": 0.1986011564731598,
      "mean_token_accuracy": 0.922387108206749,
      "num_tokens": 47614904.0,
      "step": 3860
    },
    {
      "entropy": 1.2812358438968658,
      "epoch": 2.0331753554502368,
      "grad_norm": 0.2717421352863312,
      "learning_rate": 0.00016985255625821783,
      "loss": 0.1849009394645691,
      "mean_token_accuracy": 0.9245787113904953,
      "num_tokens": 47627240.0,
      "step": 3861
    },
    {
      "entropy": 1.2583646178245544,
      "epoch": 2.0337019483938916,
      "grad_norm": 0.26972222328186035,
      "learning_rate": 0.0001698361780405862,
      "loss": 0.18644441664218903,
      "mean_token_accuracy": 0.920393168926239,
      "num_tokens": 47639576.0,
      "step": 3862
    },
    {
      "entropy": 1.2793927788734436,
      "epoch": 2.034228541337546,
      "grad_norm": 0.2913520932197571,
      "learning_rate": 0.00016981979627082945,
      "loss": 0.17899954319000244,
      "mean_token_accuracy": 0.9248150438070297,
      "num_tokens": 47651912.0,
      "step": 3863
    },
    {
      "entropy": 1.2246437966823578,
      "epoch": 2.0347551342812005,
      "grad_norm": 0.2871529757976532,
      "learning_rate": 0.00016980341094992004,
      "loss": 0.19482392072677612,
      "mean_token_accuracy": 0.9184932112693787,
      "num_tokens": 47664248.0,
      "step": 3864
    },
    {
      "entropy": 1.1911917328834534,
      "epoch": 2.0352817272248553,
      "grad_norm": 0.2914022207260132,
      "learning_rate": 0.00016978702207883065,
      "loss": 0.19326715171337128,
      "mean_token_accuracy": 0.9233204275369644,
      "num_tokens": 47676584.0,
      "step": 3865
    },
    {
      "entropy": 1.2744377553462982,
      "epoch": 2.0358083201685098,
      "grad_norm": 0.26885589957237244,
      "learning_rate": 0.00016977062965853427,
      "loss": 0.19482609629631042,
      "mean_token_accuracy": 0.9205097407102585,
      "num_tokens": 47688920.0,
      "step": 3866
    },
    {
      "entropy": 1.2833932042121887,
      "epoch": 2.036334913112164,
      "grad_norm": 0.2556898295879364,
      "learning_rate": 0.00016975423369000394,
      "loss": 0.17208848893642426,
      "mean_token_accuracy": 0.929677814245224,
      "num_tokens": 47701256.0,
      "step": 3867
    },
    {
      "entropy": 1.297087401151657,
      "epoch": 2.036861506055819,
      "grad_norm": 0.2650216519832611,
      "learning_rate": 0.00016973783417421304,
      "loss": 0.18926797807216644,
      "mean_token_accuracy": 0.9191234409809113,
      "num_tokens": 47713592.0,
      "step": 3868
    },
    {
      "entropy": 1.2456922829151154,
      "epoch": 2.0373880989994735,
      "grad_norm": 0.2747442424297333,
      "learning_rate": 0.00016972143111213512,
      "loss": 0.19703906774520874,
      "mean_token_accuracy": 0.9198992401361465,
      "num_tokens": 47725928.0,
      "step": 3869
    },
    {
      "entropy": 1.330556869506836,
      "epoch": 2.037914691943128,
      "grad_norm": 0.2835814356803894,
      "learning_rate": 0.0001697050245047439,
      "loss": 0.18732693791389465,
      "mean_token_accuracy": 0.9204771220684052,
      "num_tokens": 47738264.0,
      "step": 3870
    },
    {
      "entropy": 1.304270088672638,
      "epoch": 2.0384412848867823,
      "grad_norm": 0.27201247215270996,
      "learning_rate": 0.00016968861435301337,
      "loss": 0.18724404275417328,
      "mean_token_accuracy": 0.921107143163681,
      "num_tokens": 47750600.0,
      "step": 3871
    },
    {
      "entropy": 1.230264812707901,
      "epoch": 2.038967877830437,
      "grad_norm": 0.2530989646911621,
      "learning_rate": 0.0001696722006579177,
      "loss": 0.17592748999595642,
      "mean_token_accuracy": 0.9254534393548965,
      "num_tokens": 47762936.0,
      "step": 3872
    },
    {
      "entropy": 1.2308352589607239,
      "epoch": 2.0394944707740916,
      "grad_norm": 0.2694772779941559,
      "learning_rate": 0.00016965578342043126,
      "loss": 0.19067507982254028,
      "mean_token_accuracy": 0.9236868768930435,
      "num_tokens": 47775272.0,
      "step": 3873
    },
    {
      "entropy": 1.291840374469757,
      "epoch": 2.040021063717746,
      "grad_norm": 0.2618374228477478,
      "learning_rate": 0.00016963936264152867,
      "loss": 0.18578636646270752,
      "mean_token_accuracy": 0.9224181473255157,
      "num_tokens": 47787608.0,
      "step": 3874
    },
    {
      "entropy": 1.2792740166187286,
      "epoch": 2.040547656661401,
      "grad_norm": 0.2911253571510315,
      "learning_rate": 0.0001696229383221847,
      "loss": 0.20944932103157043,
      "mean_token_accuracy": 0.9135683476924896,
      "num_tokens": 47799944.0,
      "step": 3875
    },
    {
      "entropy": 1.3030252158641815,
      "epoch": 2.0410742496050553,
      "grad_norm": 0.2904371917247772,
      "learning_rate": 0.0001696065104633744,
      "loss": 0.20830057561397552,
      "mean_token_accuracy": 0.9120756387710571,
      "num_tokens": 47812280.0,
      "step": 3876
    },
    {
      "entropy": 1.2803774774074554,
      "epoch": 2.0416008425487098,
      "grad_norm": 0.2760698199272156,
      "learning_rate": 0.00016959007906607303,
      "loss": 0.19493544101715088,
      "mean_token_accuracy": 0.9187896698713303,
      "num_tokens": 47824616.0,
      "step": 3877
    },
    {
      "entropy": 1.2554866671562195,
      "epoch": 2.042127435492364,
      "grad_norm": 0.4330475926399231,
      "learning_rate": 0.00016957364413125594,
      "loss": 0.18488185107707977,
      "mean_token_accuracy": 0.9197061508893967,
      "num_tokens": 47836952.0,
      "step": 3878
    },
    {
      "entropy": 1.3076776266098022,
      "epoch": 2.042654028436019,
      "grad_norm": 0.2841300964355469,
      "learning_rate": 0.00016955720565989884,
      "loss": 0.19156453013420105,
      "mean_token_accuracy": 0.9239335805177689,
      "num_tokens": 47849288.0,
      "step": 3879
    },
    {
      "entropy": 1.3151583969593048,
      "epoch": 2.0431806213796735,
      "grad_norm": 0.2921697199344635,
      "learning_rate": 0.00016954076365297758,
      "loss": 0.2065403163433075,
      "mean_token_accuracy": 0.9110155552625656,
      "num_tokens": 47861624.0,
      "step": 3880
    },
    {
      "entropy": 1.2858489751815796,
      "epoch": 2.043707214323328,
      "grad_norm": 0.27680060267448425,
      "learning_rate": 0.00016952431811146818,
      "loss": 0.19812069833278656,
      "mean_token_accuracy": 0.9194805026054382,
      "num_tokens": 47873960.0,
      "step": 3881
    },
    {
      "entropy": 1.3400939404964447,
      "epoch": 2.044233807266983,
      "grad_norm": 0.29581987857818604,
      "learning_rate": 0.00016950786903634696,
      "loss": 0.19464778900146484,
      "mean_token_accuracy": 0.920210674405098,
      "num_tokens": 47886296.0,
      "step": 3882
    },
    {
      "entropy": 1.3481502830982208,
      "epoch": 2.044760400210637,
      "grad_norm": 0.3192988634109497,
      "learning_rate": 0.0001694914164285904,
      "loss": 0.20431135594844818,
      "mean_token_accuracy": 0.918129950761795,
      "num_tokens": 47898632.0,
      "step": 3883
    },
    {
      "entropy": 1.212340623140335,
      "epoch": 2.0452869931542916,
      "grad_norm": 0.23157627880573273,
      "learning_rate": 0.00016947496028917516,
      "loss": 0.17617271840572357,
      "mean_token_accuracy": 0.9243680238723755,
      "num_tokens": 47910968.0,
      "step": 3884
    },
    {
      "entropy": 1.2161346971988678,
      "epoch": 2.0458135860979465,
      "grad_norm": 0.2513905167579651,
      "learning_rate": 0.0001694585006190782,
      "loss": 0.1802637130022049,
      "mean_token_accuracy": 0.924398884177208,
      "num_tokens": 47923304.0,
      "step": 3885
    },
    {
      "entropy": 1.3086143136024475,
      "epoch": 2.046340179041601,
      "grad_norm": 0.2897863984107971,
      "learning_rate": 0.00016944203741927662,
      "loss": 0.20457464456558228,
      "mean_token_accuracy": 0.9144063889980316,
      "num_tokens": 47935640.0,
      "step": 3886
    },
    {
      "entropy": 1.2356386482715607,
      "epoch": 2.0468667719852554,
      "grad_norm": 0.26771265268325806,
      "learning_rate": 0.00016942557069074768,
      "loss": 0.17813019454479218,
      "mean_token_accuracy": 0.9241292774677277,
      "num_tokens": 47947976.0,
      "step": 3887
    },
    {
      "entropy": 1.3301630318164825,
      "epoch": 2.0473933649289098,
      "grad_norm": 0.27502498030662537,
      "learning_rate": 0.000169409100434469,
      "loss": 0.20527851581573486,
      "mean_token_accuracy": 0.9159713387489319,
      "num_tokens": 47960312.0,
      "step": 3888
    },
    {
      "entropy": 1.2828874289989471,
      "epoch": 2.0479199578725646,
      "grad_norm": 0.261263370513916,
      "learning_rate": 0.00016939262665141832,
      "loss": 0.19007109105587006,
      "mean_token_accuracy": 0.9201304614543915,
      "num_tokens": 47972648.0,
      "step": 3889
    },
    {
      "entropy": 1.3016537725925446,
      "epoch": 2.048446550816219,
      "grad_norm": 0.2682885527610779,
      "learning_rate": 0.00016937614934257354,
      "loss": 0.1853100210428238,
      "mean_token_accuracy": 0.9268095791339874,
      "num_tokens": 47984984.0,
      "step": 3890
    },
    {
      "entropy": 1.2769020199775696,
      "epoch": 2.0489731437598735,
      "grad_norm": 0.2722259759902954,
      "learning_rate": 0.00016935966850891283,
      "loss": 0.20796817541122437,
      "mean_token_accuracy": 0.9139941036701202,
      "num_tokens": 47997320.0,
      "step": 3891
    },
    {
      "entropy": 1.258107304573059,
      "epoch": 2.0494997367035284,
      "grad_norm": 0.25843435525894165,
      "learning_rate": 0.00016934318415141457,
      "loss": 0.16999708116054535,
      "mean_token_accuracy": 0.9257606714963913,
      "num_tokens": 48009656.0,
      "step": 3892
    },
    {
      "entropy": 1.1906676590442657,
      "epoch": 2.050026329647183,
      "grad_norm": 0.2513735890388489,
      "learning_rate": 0.00016932669627105738,
      "loss": 0.1697864979505539,
      "mean_token_accuracy": 0.925465926527977,
      "num_tokens": 48021992.0,
      "step": 3893
    },
    {
      "entropy": 1.2723312675952911,
      "epoch": 2.050552922590837,
      "grad_norm": 0.2743954062461853,
      "learning_rate": 0.00016931020486881998,
      "loss": 0.18033276498317719,
      "mean_token_accuracy": 0.9286290258169174,
      "num_tokens": 48034328.0,
      "step": 3894
    },
    {
      "entropy": 1.2819771468639374,
      "epoch": 2.0510795155344916,
      "grad_norm": 0.29130327701568604,
      "learning_rate": 0.00016929370994568142,
      "loss": 0.1982630491256714,
      "mean_token_accuracy": 0.9166873097419739,
      "num_tokens": 48046664.0,
      "step": 3895
    },
    {
      "entropy": 1.2571477890014648,
      "epoch": 2.0516061084781465,
      "grad_norm": 0.27202513813972473,
      "learning_rate": 0.00016927721150262088,
      "loss": 0.18936115503311157,
      "mean_token_accuracy": 0.9197106510400772,
      "num_tokens": 48059000.0,
      "step": 3896
    },
    {
      "entropy": 1.3048891425132751,
      "epoch": 2.052132701421801,
      "grad_norm": 0.27362993359565735,
      "learning_rate": 0.00016926070954061782,
      "loss": 0.19029444456100464,
      "mean_token_accuracy": 0.9229496270418167,
      "num_tokens": 48071336.0,
      "step": 3897
    },
    {
      "entropy": 1.2617215812206268,
      "epoch": 2.0526592943654554,
      "grad_norm": 0.28144410252571106,
      "learning_rate": 0.00016924420406065177,
      "loss": 0.18220680952072144,
      "mean_token_accuracy": 0.9252916872501373,
      "num_tokens": 48083672.0,
      "step": 3898
    },
    {
      "entropy": 1.2789317965507507,
      "epoch": 2.0531858873091102,
      "grad_norm": 0.2608755826950073,
      "learning_rate": 0.00016922769506370268,
      "loss": 0.18006981909275055,
      "mean_token_accuracy": 0.9222934991121292,
      "num_tokens": 48096008.0,
      "step": 3899
    },
    {
      "entropy": 1.22543865442276,
      "epoch": 2.0537124802527646,
      "grad_norm": 0.27741050720214844,
      "learning_rate": 0.00016921118255075053,
      "loss": 0.1895115077495575,
      "mean_token_accuracy": 0.9215471893548965,
      "num_tokens": 48108344.0,
      "step": 3900
    },
    {
      "entropy": 1.2530005872249603,
      "epoch": 2.054239073196419,
      "grad_norm": 0.2714971899986267,
      "learning_rate": 0.00016919466652277556,
      "loss": 0.1776505410671234,
      "mean_token_accuracy": 0.9273561537265778,
      "num_tokens": 48120680.0,
      "step": 3901
    },
    {
      "entropy": 1.2720350325107574,
      "epoch": 2.0547656661400735,
      "grad_norm": 0.26950451731681824,
      "learning_rate": 0.00016917814698075827,
      "loss": 0.19353142380714417,
      "mean_token_accuracy": 0.9203890562057495,
      "num_tokens": 48133016.0,
      "step": 3902
    },
    {
      "entropy": 1.2247913479804993,
      "epoch": 2.0552922590837284,
      "grad_norm": 0.2836177945137024,
      "learning_rate": 0.0001691616239256793,
      "loss": 0.19141322374343872,
      "mean_token_accuracy": 0.9200913459062576,
      "num_tokens": 48145352.0,
      "step": 3903
    },
    {
      "entropy": 1.2887300550937653,
      "epoch": 2.055818852027383,
      "grad_norm": 0.2868979275226593,
      "learning_rate": 0.00016914509735851954,
      "loss": 0.20681561529636383,
      "mean_token_accuracy": 0.9199369102716446,
      "num_tokens": 48157688.0,
      "step": 3904
    },
    {
      "entropy": 1.2575041055679321,
      "epoch": 2.056345444971037,
      "grad_norm": 0.26332929730415344,
      "learning_rate": 0.00016912856728026006,
      "loss": 0.18368637561798096,
      "mean_token_accuracy": 0.9240716248750687,
      "num_tokens": 48170024.0,
      "step": 3905
    },
    {
      "entropy": 1.2432773113250732,
      "epoch": 2.056872037914692,
      "grad_norm": 0.29000067710876465,
      "learning_rate": 0.0001691120336918822,
      "loss": 0.20695710182189941,
      "mean_token_accuracy": 0.9119783192873001,
      "num_tokens": 48182360.0,
      "step": 3906
    },
    {
      "entropy": 1.2723545730113983,
      "epoch": 2.0573986308583465,
      "grad_norm": 0.2832140624523163,
      "learning_rate": 0.0001690954965943674,
      "loss": 0.19655488431453705,
      "mean_token_accuracy": 0.9173429161310196,
      "num_tokens": 48194696.0,
      "step": 3907
    },
    {
      "entropy": 1.3194984793663025,
      "epoch": 2.057925223802001,
      "grad_norm": 0.280369371175766,
      "learning_rate": 0.00016907895598869743,
      "loss": 0.18848003447055817,
      "mean_token_accuracy": 0.9231791347265244,
      "num_tokens": 48207032.0,
      "step": 3908
    },
    {
      "entropy": 1.2152039706707,
      "epoch": 2.058451816745656,
      "grad_norm": 0.2655313014984131,
      "learning_rate": 0.00016906241187585416,
      "loss": 0.19103647768497467,
      "mean_token_accuracy": 0.9197648912668228,
      "num_tokens": 48219368.0,
      "step": 3909
    },
    {
      "entropy": 1.2087748050689697,
      "epoch": 2.0589784096893102,
      "grad_norm": 0.26991021633148193,
      "learning_rate": 0.00016904586425681975,
      "loss": 0.18405042588710785,
      "mean_token_accuracy": 0.9230748862028122,
      "num_tokens": 48231704.0,
      "step": 3910
    },
    {
      "entropy": 1.2276612520217896,
      "epoch": 2.0595050026329647,
      "grad_norm": 0.2636072337627411,
      "learning_rate": 0.00016902931313257652,
      "loss": 0.1714887022972107,
      "mean_token_accuracy": 0.928344801068306,
      "num_tokens": 48244040.0,
      "step": 3911
    },
    {
      "entropy": 1.2997864484786987,
      "epoch": 2.060031595576619,
      "grad_norm": 0.29149720072746277,
      "learning_rate": 0.00016901275850410703,
      "loss": 0.19735336303710938,
      "mean_token_accuracy": 0.9174995422363281,
      "num_tokens": 48256376.0,
      "step": 3912
    },
    {
      "entropy": 1.2492857873439789,
      "epoch": 2.060558188520274,
      "grad_norm": 0.2696395218372345,
      "learning_rate": 0.00016899620037239398,
      "loss": 0.18676462769508362,
      "mean_token_accuracy": 0.9226437360048294,
      "num_tokens": 48268712.0,
      "step": 3913
    },
    {
      "entropy": 1.2569607198238373,
      "epoch": 2.0610847814639284,
      "grad_norm": 0.28245505690574646,
      "learning_rate": 0.00016897963873842043,
      "loss": 0.20712406933307648,
      "mean_token_accuracy": 0.9177358150482178,
      "num_tokens": 48281048.0,
      "step": 3914
    },
    {
      "entropy": 1.2320314943790436,
      "epoch": 2.061611374407583,
      "grad_norm": 0.28270742297172546,
      "learning_rate": 0.00016896307360316943,
      "loss": 0.19076888263225555,
      "mean_token_accuracy": 0.9220731556415558,
      "num_tokens": 48293384.0,
      "step": 3915
    },
    {
      "entropy": 1.3260917365550995,
      "epoch": 2.0621379673512377,
      "grad_norm": 0.2971607744693756,
      "learning_rate": 0.00016894650496762444,
      "loss": 0.19012241065502167,
      "mean_token_accuracy": 0.9222848564386368,
      "num_tokens": 48305720.0,
      "step": 3916
    },
    {
      "entropy": 1.2801413834095001,
      "epoch": 2.062664560294892,
      "grad_norm": 0.30607089400291443,
      "learning_rate": 0.00016892993283276902,
      "loss": 0.20878365635871887,
      "mean_token_accuracy": 0.9175769686698914,
      "num_tokens": 48318056.0,
      "step": 3917
    },
    {
      "entropy": 1.3265896141529083,
      "epoch": 2.0631911532385465,
      "grad_norm": 0.30750954151153564,
      "learning_rate": 0.00016891335719958697,
      "loss": 0.18873047828674316,
      "mean_token_accuracy": 0.920640304684639,
      "num_tokens": 48330392.0,
      "step": 3918
    },
    {
      "entropy": 1.350228101015091,
      "epoch": 2.063717746182201,
      "grad_norm": 0.2853001654148102,
      "learning_rate": 0.00016889677806906225,
      "loss": 0.18502716720104218,
      "mean_token_accuracy": 0.9253284931182861,
      "num_tokens": 48342728.0,
      "step": 3919
    },
    {
      "entropy": 1.2163171470165253,
      "epoch": 2.064244339125856,
      "grad_norm": 0.24412177503108978,
      "learning_rate": 0.0001688801954421791,
      "loss": 0.16768863797187805,
      "mean_token_accuracy": 0.9300560504198074,
      "num_tokens": 48355064.0,
      "step": 3920
    },
    {
      "entropy": 1.3360876441001892,
      "epoch": 2.0647709320695102,
      "grad_norm": 0.27784374356269836,
      "learning_rate": 0.00016886360931992194,
      "loss": 0.1872931867837906,
      "mean_token_accuracy": 0.9237985759973526,
      "num_tokens": 48367400.0,
      "step": 3921
    },
    {
      "entropy": 1.243847370147705,
      "epoch": 2.0652975250131647,
      "grad_norm": 0.28875452280044556,
      "learning_rate": 0.00016884701970327538,
      "loss": 0.19071172177791595,
      "mean_token_accuracy": 0.9176552146673203,
      "num_tokens": 48379736.0,
      "step": 3922
    },
    {
      "entropy": 1.3042690753936768,
      "epoch": 2.0658241179568195,
      "grad_norm": 0.289671927690506,
      "learning_rate": 0.0001688304265932242,
      "loss": 0.20444759726524353,
      "mean_token_accuracy": 0.9158494174480438,
      "num_tokens": 48392072.0,
      "step": 3923
    },
    {
      "entropy": 1.271407037973404,
      "epoch": 2.066350710900474,
      "grad_norm": 0.296796977519989,
      "learning_rate": 0.00016881382999075353,
      "loss": 0.1858009397983551,
      "mean_token_accuracy": 0.920338898897171,
      "num_tokens": 48404408.0,
      "step": 3924
    },
    {
      "entropy": 1.282002031803131,
      "epoch": 2.0668773038441284,
      "grad_norm": 0.27760758996009827,
      "learning_rate": 0.00016879722989684854,
      "loss": 0.198472797870636,
      "mean_token_accuracy": 0.9221088290214539,
      "num_tokens": 48416744.0,
      "step": 3925
    },
    {
      "entropy": 1.2682355344295502,
      "epoch": 2.0674038967877832,
      "grad_norm": 0.25768402218818665,
      "learning_rate": 0.00016878062631249473,
      "loss": 0.1786102056503296,
      "mean_token_accuracy": 0.9221510291099548,
      "num_tokens": 48429080.0,
      "step": 3926
    },
    {
      "entropy": 1.2786115109920502,
      "epoch": 2.0679304897314377,
      "grad_norm": 0.27482444047927856,
      "learning_rate": 0.0001687640192386777,
      "loss": 0.1859729140996933,
      "mean_token_accuracy": 0.9208211004734039,
      "num_tokens": 48441416.0,
      "step": 3927
    },
    {
      "entropy": 1.258926510810852,
      "epoch": 2.068457082675092,
      "grad_norm": 0.28567978739738464,
      "learning_rate": 0.00016874740867638339,
      "loss": 0.18969936668872833,
      "mean_token_accuracy": 0.9214884340763092,
      "num_tokens": 48453752.0,
      "step": 3928
    },
    {
      "entropy": 1.2364517748355865,
      "epoch": 2.0689836756187465,
      "grad_norm": 0.26859763264656067,
      "learning_rate": 0.00016873079462659783,
      "loss": 0.19034039974212646,
      "mean_token_accuracy": 0.9214574694633484,
      "num_tokens": 48466088.0,
      "step": 3929
    },
    {
      "entropy": 1.2525107264518738,
      "epoch": 2.0695102685624014,
      "grad_norm": 0.28058695793151855,
      "learning_rate": 0.00016871417709030727,
      "loss": 0.18607176840305328,
      "mean_token_accuracy": 0.9215060919523239,
      "num_tokens": 48478424.0,
      "step": 3930
    },
    {
      "entropy": 1.2619339525699615,
      "epoch": 2.070036861506056,
      "grad_norm": 0.2828121781349182,
      "learning_rate": 0.00016869755606849826,
      "loss": 0.18340539932250977,
      "mean_token_accuracy": 0.9251066595315933,
      "num_tokens": 48490760.0,
      "step": 3931
    },
    {
      "entropy": 1.2884092330932617,
      "epoch": 2.0705634544497102,
      "grad_norm": 0.26471489667892456,
      "learning_rate": 0.00016868093156215743,
      "loss": 0.17937590181827545,
      "mean_token_accuracy": 0.9243820607662201,
      "num_tokens": 48503096.0,
      "step": 3932
    },
    {
      "entropy": 1.2739530205726624,
      "epoch": 2.071090047393365,
      "grad_norm": 0.26046693325042725,
      "learning_rate": 0.00016866430357227177,
      "loss": 0.1955185979604721,
      "mean_token_accuracy": 0.9210577309131622,
      "num_tokens": 48515432.0,
      "step": 3933
    },
    {
      "entropy": 1.2842492163181305,
      "epoch": 2.0716166403370195,
      "grad_norm": 0.29395169019699097,
      "learning_rate": 0.00016864767209982825,
      "loss": 0.21026445925235748,
      "mean_token_accuracy": 0.9173175990581512,
      "num_tokens": 48527768.0,
      "step": 3934
    },
    {
      "entropy": 1.2743287682533264,
      "epoch": 2.072143233280674,
      "grad_norm": 0.272370845079422,
      "learning_rate": 0.00016863103714581432,
      "loss": 0.1986452043056488,
      "mean_token_accuracy": 0.9205825477838516,
      "num_tokens": 48540104.0,
      "step": 3935
    },
    {
      "entropy": 1.2706570327281952,
      "epoch": 2.0726698262243284,
      "grad_norm": 0.2720491588115692,
      "learning_rate": 0.00016861439871121743,
      "loss": 0.17972148954868317,
      "mean_token_accuracy": 0.9210758358240128,
      "num_tokens": 48552440.0,
      "step": 3936
    },
    {
      "entropy": 1.2344097197055817,
      "epoch": 2.0731964191679833,
      "grad_norm": 0.261137992143631,
      "learning_rate": 0.0001685977567970253,
      "loss": 0.18583646416664124,
      "mean_token_accuracy": 0.9242785274982452,
      "num_tokens": 48564776.0,
      "step": 3937
    },
    {
      "entropy": 1.2211202681064606,
      "epoch": 2.0737230121116377,
      "grad_norm": 0.2840760350227356,
      "learning_rate": 0.0001685811114042259,
      "loss": 0.18876339495182037,
      "mean_token_accuracy": 0.9212382435798645,
      "num_tokens": 48577112.0,
      "step": 3938
    },
    {
      "entropy": 1.2502583861351013,
      "epoch": 2.074249605055292,
      "grad_norm": 0.27275213599205017,
      "learning_rate": 0.00016856446253380735,
      "loss": 0.1872887909412384,
      "mean_token_accuracy": 0.923475593328476,
      "num_tokens": 48589448.0,
      "step": 3939
    },
    {
      "entropy": 1.2701627314090729,
      "epoch": 2.074776197998947,
      "grad_norm": 0.2745870053768158,
      "learning_rate": 0.00016854781018675797,
      "loss": 0.18436497449874878,
      "mean_token_accuracy": 0.9266326576471329,
      "num_tokens": 48601784.0,
      "step": 3940
    },
    {
      "entropy": 1.2546738684177399,
      "epoch": 2.0753027909426014,
      "grad_norm": 0.2822677195072174,
      "learning_rate": 0.0001685311543640664,
      "loss": 0.1884290874004364,
      "mean_token_accuracy": 0.9195769876241684,
      "num_tokens": 48614120.0,
      "step": 3941
    },
    {
      "entropy": 1.2545583546161652,
      "epoch": 2.075829383886256,
      "grad_norm": 0.2970985174179077,
      "learning_rate": 0.0001685144950667213,
      "loss": 0.19396939873695374,
      "mean_token_accuracy": 0.9214707016944885,
      "num_tokens": 48626456.0,
      "step": 3942
    },
    {
      "entropy": 1.243100881576538,
      "epoch": 2.0763559768299107,
      "grad_norm": 0.25157052278518677,
      "learning_rate": 0.0001684978322957117,
      "loss": 0.17563554644584656,
      "mean_token_accuracy": 0.9291206002235413,
      "num_tokens": 48638792.0,
      "step": 3943
    },
    {
      "entropy": 1.2628869712352753,
      "epoch": 2.076882569773565,
      "grad_norm": 0.2984946072101593,
      "learning_rate": 0.00016848116605202673,
      "loss": 0.20061740279197693,
      "mean_token_accuracy": 0.9180862903594971,
      "num_tokens": 48651128.0,
      "step": 3944
    },
    {
      "entropy": 1.3347080945968628,
      "epoch": 2.0774091627172195,
      "grad_norm": 0.3006715476512909,
      "learning_rate": 0.00016846449633665578,
      "loss": 0.18334859609603882,
      "mean_token_accuracy": 0.9282584637403488,
      "num_tokens": 48663464.0,
      "step": 3945
    },
    {
      "entropy": 1.2688239216804504,
      "epoch": 2.077935755660874,
      "grad_norm": 0.26901793479919434,
      "learning_rate": 0.00016844782315058847,
      "loss": 0.1989820897579193,
      "mean_token_accuracy": 0.9150257855653763,
      "num_tokens": 48675800.0,
      "step": 3946
    },
    {
      "entropy": 1.265367865562439,
      "epoch": 2.078462348604529,
      "grad_norm": 0.29596784710884094,
      "learning_rate": 0.00016843114649481452,
      "loss": 0.20104429125785828,
      "mean_token_accuracy": 0.9207267761230469,
      "num_tokens": 48688136.0,
      "step": 3947
    },
    {
      "entropy": 1.267877608537674,
      "epoch": 2.0789889415481833,
      "grad_norm": 0.29306116700172424,
      "learning_rate": 0.00016841446637032396,
      "loss": 0.20623625814914703,
      "mean_token_accuracy": 0.9133555591106415,
      "num_tokens": 48700472.0,
      "step": 3948
    },
    {
      "entropy": 1.2642518281936646,
      "epoch": 2.0795155344918377,
      "grad_norm": 0.2690291404724121,
      "learning_rate": 0.00016839778277810702,
      "loss": 0.181110218167305,
      "mean_token_accuracy": 0.9257129430770874,
      "num_tokens": 48712808.0,
      "step": 3949
    },
    {
      "entropy": 1.3663403391838074,
      "epoch": 2.0800421274354925,
      "grad_norm": 0.29609712958335876,
      "learning_rate": 0.0001683810957191541,
      "loss": 0.19982632994651794,
      "mean_token_accuracy": 0.9155548959970474,
      "num_tokens": 48725144.0,
      "step": 3950
    },
    {
      "entropy": 1.3063276708126068,
      "epoch": 2.080568720379147,
      "grad_norm": 0.2549259066581726,
      "learning_rate": 0.0001683644051944558,
      "loss": 0.16801197826862335,
      "mean_token_accuracy": 0.928315594792366,
      "num_tokens": 48737480.0,
      "step": 3951
    },
    {
      "entropy": 1.3074188232421875,
      "epoch": 2.0810953133228014,
      "grad_norm": 0.2865307331085205,
      "learning_rate": 0.0001683477112050029,
      "loss": 0.1887558102607727,
      "mean_token_accuracy": 0.9195426404476166,
      "num_tokens": 48749816.0,
      "step": 3952
    },
    {
      "entropy": 1.3835014998912811,
      "epoch": 2.081621906266456,
      "grad_norm": 0.27939164638519287,
      "learning_rate": 0.0001683310137517865,
      "loss": 0.1879327893257141,
      "mean_token_accuracy": 0.9196098446846008,
      "num_tokens": 48762152.0,
      "step": 3953
    },
    {
      "entropy": 1.4062438309192657,
      "epoch": 2.0821484992101107,
      "grad_norm": 0.3003182113170624,
      "learning_rate": 0.00016831431283579777,
      "loss": 0.20012448728084564,
      "mean_token_accuracy": 0.9185905754566193,
      "num_tokens": 48774488.0,
      "step": 3954
    },
    {
      "entropy": 1.3671795129776,
      "epoch": 2.082675092153765,
      "grad_norm": 0.2660031318664551,
      "learning_rate": 0.0001682976084580282,
      "loss": 0.19102323055267334,
      "mean_token_accuracy": 0.919410839676857,
      "num_tokens": 48786824.0,
      "step": 3955
    },
    {
      "entropy": 1.325892060995102,
      "epoch": 2.0832016850974195,
      "grad_norm": 0.2605104148387909,
      "learning_rate": 0.00016828090061946934,
      "loss": 0.18359193205833435,
      "mean_token_accuracy": 0.9214528203010559,
      "num_tokens": 48799160.0,
      "step": 3956
    },
    {
      "entropy": 1.3484356999397278,
      "epoch": 2.0837282780410744,
      "grad_norm": 0.2751688063144684,
      "learning_rate": 0.00016826418932111315,
      "loss": 0.18124142289161682,
      "mean_token_accuracy": 0.9217920154333115,
      "num_tokens": 48811496.0,
      "step": 3957
    },
    {
      "entropy": 1.3338854312896729,
      "epoch": 2.084254870984729,
      "grad_norm": 0.26002514362335205,
      "learning_rate": 0.0001682474745639516,
      "loss": 0.166397824883461,
      "mean_token_accuracy": 0.9286703169345856,
      "num_tokens": 48823832.0,
      "step": 3958
    },
    {
      "entropy": 1.3630346655845642,
      "epoch": 2.0847814639283833,
      "grad_norm": 0.28807926177978516,
      "learning_rate": 0.00016823075634897697,
      "loss": 0.1959732472896576,
      "mean_token_accuracy": 0.9200254678726196,
      "num_tokens": 48836168.0,
      "step": 3959
    },
    {
      "entropy": 1.3884229063987732,
      "epoch": 2.085308056872038,
      "grad_norm": 0.2943662405014038,
      "learning_rate": 0.00016821403467718178,
      "loss": 0.18066152930259705,
      "mean_token_accuracy": 0.9316907674074173,
      "num_tokens": 48848504.0,
      "step": 3960
    },
    {
      "entropy": 1.3363731503486633,
      "epoch": 2.0858346498156926,
      "grad_norm": 0.3065935969352722,
      "learning_rate": 0.0001681973095495586,
      "loss": 0.2028815746307373,
      "mean_token_accuracy": 0.9201863408088684,
      "num_tokens": 48860840.0,
      "step": 3961
    },
    {
      "entropy": 1.3781720399856567,
      "epoch": 2.086361242759347,
      "grad_norm": 0.3044969141483307,
      "learning_rate": 0.00016818058096710036,
      "loss": 0.2023879587650299,
      "mean_token_accuracy": 0.9209664463996887,
      "num_tokens": 48873176.0,
      "step": 3962
    },
    {
      "entropy": 1.3339049518108368,
      "epoch": 2.0868878357030014,
      "grad_norm": 0.3165450692176819,
      "learning_rate": 0.00016816384893080012,
      "loss": 0.2043043076992035,
      "mean_token_accuracy": 0.9149027913808823,
      "num_tokens": 48885512.0,
      "step": 3963
    },
    {
      "entropy": 1.3235175609588623,
      "epoch": 2.0874144286466563,
      "grad_norm": 0.2820136845111847,
      "learning_rate": 0.0001681471134416512,
      "loss": 0.18209639191627502,
      "mean_token_accuracy": 0.9246355444192886,
      "num_tokens": 48897848.0,
      "step": 3964
    },
    {
      "entropy": 1.3395156562328339,
      "epoch": 2.0879410215903107,
      "grad_norm": 0.45867112278938293,
      "learning_rate": 0.00016813037450064705,
      "loss": 0.19053226709365845,
      "mean_token_accuracy": 0.9275738596916199,
      "num_tokens": 48910184.0,
      "step": 3965
    },
    {
      "entropy": 1.3038859367370605,
      "epoch": 2.088467614533965,
      "grad_norm": 0.28860345482826233,
      "learning_rate": 0.00016811363210878137,
      "loss": 0.21587130427360535,
      "mean_token_accuracy": 0.9093978554010391,
      "num_tokens": 48922520.0,
      "step": 3966
    },
    {
      "entropy": 1.2706226110458374,
      "epoch": 2.08899420747762,
      "grad_norm": 0.281698077917099,
      "learning_rate": 0.0001680968862670481,
      "loss": 0.1824536919593811,
      "mean_token_accuracy": 0.9227961152791977,
      "num_tokens": 48934856.0,
      "step": 3967
    },
    {
      "entropy": 1.30519899725914,
      "epoch": 2.0895208004212744,
      "grad_norm": 0.28888365626335144,
      "learning_rate": 0.00016808013697644126,
      "loss": 0.19047529995441437,
      "mean_token_accuracy": 0.9213353842496872,
      "num_tokens": 48947192.0,
      "step": 3968
    },
    {
      "entropy": 1.3129478991031647,
      "epoch": 2.090047393364929,
      "grad_norm": 0.30279576778411865,
      "learning_rate": 0.00016806338423795524,
      "loss": 0.197226881980896,
      "mean_token_accuracy": 0.9145682752132416,
      "num_tokens": 48959528.0,
      "step": 3969
    },
    {
      "entropy": 1.3123193085193634,
      "epoch": 2.0905739863085833,
      "grad_norm": 0.2858417332172394,
      "learning_rate": 0.00016804662805258448,
      "loss": 0.1826488971710205,
      "mean_token_accuracy": 0.9254605323076248,
      "num_tokens": 48971864.0,
      "step": 3970
    },
    {
      "entropy": 1.347455382347107,
      "epoch": 2.091100579252238,
      "grad_norm": 0.2954428791999817,
      "learning_rate": 0.00016802986842132374,
      "loss": 0.20556935667991638,
      "mean_token_accuracy": 0.9166451990604401,
      "num_tokens": 48984200.0,
      "step": 3971
    },
    {
      "entropy": 1.319230854511261,
      "epoch": 2.0916271721958926,
      "grad_norm": 0.32010331749916077,
      "learning_rate": 0.00016801310534516798,
      "loss": 0.19692543148994446,
      "mean_token_accuracy": 0.9220879971981049,
      "num_tokens": 48996536.0,
      "step": 3972
    },
    {
      "entropy": 1.3301867544651031,
      "epoch": 2.092153765139547,
      "grad_norm": 0.24198579788208008,
      "learning_rate": 0.00016799633882511224,
      "loss": 0.18499736487865448,
      "mean_token_accuracy": 0.9226900786161423,
      "num_tokens": 49008872.0,
      "step": 3973
    },
    {
      "entropy": 1.3645350635051727,
      "epoch": 2.092680358083202,
      "grad_norm": 0.26797109842300415,
      "learning_rate": 0.00016797956886215195,
      "loss": 0.19249004125595093,
      "mean_token_accuracy": 0.9230862706899643,
      "num_tokens": 49021208.0,
      "step": 3974
    },
    {
      "entropy": 1.340773344039917,
      "epoch": 2.0932069510268563,
      "grad_norm": 0.26169508695602417,
      "learning_rate": 0.0001679627954572825,
      "loss": 0.18936491012573242,
      "mean_token_accuracy": 0.9241386204957962,
      "num_tokens": 49033544.0,
      "step": 3975
    },
    {
      "entropy": 1.3765994012355804,
      "epoch": 2.0937335439705107,
      "grad_norm": 0.2780802249908447,
      "learning_rate": 0.00016794601861149977,
      "loss": 0.1884055733680725,
      "mean_token_accuracy": 0.9221274852752686,
      "num_tokens": 49045880.0,
      "step": 3976
    },
    {
      "entropy": 1.4070016741752625,
      "epoch": 2.0942601369141656,
      "grad_norm": 0.26833170652389526,
      "learning_rate": 0.00016792923832579964,
      "loss": 0.18630680441856384,
      "mean_token_accuracy": 0.9243040978908539,
      "num_tokens": 49058216.0,
      "step": 3977
    },
    {
      "entropy": 1.3762468099594116,
      "epoch": 2.09478672985782,
      "grad_norm": 0.26797711849212646,
      "learning_rate": 0.00016791245460117825,
      "loss": 0.18355493247509003,
      "mean_token_accuracy": 0.9276077747344971,
      "num_tokens": 49070552.0,
      "step": 3978
    },
    {
      "entropy": 1.3690291047096252,
      "epoch": 2.0953133228014744,
      "grad_norm": 0.2826116383075714,
      "learning_rate": 0.000167895667438632,
      "loss": 0.19171208143234253,
      "mean_token_accuracy": 0.9223132878541946,
      "num_tokens": 49082888.0,
      "step": 3979
    },
    {
      "entropy": 1.395127385854721,
      "epoch": 2.095839915745129,
      "grad_norm": 0.29053691029548645,
      "learning_rate": 0.00016787887683915737,
      "loss": 0.1943792700767517,
      "mean_token_accuracy": 0.9202833324670792,
      "num_tokens": 49095224.0,
      "step": 3980
    },
    {
      "entropy": 1.4068172872066498,
      "epoch": 2.0963665086887837,
      "grad_norm": 0.296621710062027,
      "learning_rate": 0.00016786208280375123,
      "loss": 0.18952558934688568,
      "mean_token_accuracy": 0.9180808961391449,
      "num_tokens": 49107560.0,
      "step": 3981
    },
    {
      "entropy": 1.3650657832622528,
      "epoch": 2.096893101632438,
      "grad_norm": 0.3004305064678192,
      "learning_rate": 0.00016784528533341045,
      "loss": 0.19340358674526215,
      "mean_token_accuracy": 0.9233474135398865,
      "num_tokens": 49119896.0,
      "step": 3982
    },
    {
      "entropy": 1.3545269072055817,
      "epoch": 2.0974196945760926,
      "grad_norm": 0.2610722482204437,
      "learning_rate": 0.0001678284844291322,
      "loss": 0.17823578417301178,
      "mean_token_accuracy": 0.927034541964531,
      "num_tokens": 49132232.0,
      "step": 3983
    },
    {
      "entropy": 1.36451455950737,
      "epoch": 2.0979462875197474,
      "grad_norm": 0.29160958528518677,
      "learning_rate": 0.0001678116800919139,
      "loss": 0.19818106293678284,
      "mean_token_accuracy": 0.9195860922336578,
      "num_tokens": 49144568.0,
      "step": 3984
    },
    {
      "entropy": 1.339730054140091,
      "epoch": 2.098472880463402,
      "grad_norm": 0.27172547578811646,
      "learning_rate": 0.00016779487232275312,
      "loss": 0.19754473865032196,
      "mean_token_accuracy": 0.9192177653312683,
      "num_tokens": 49156904.0,
      "step": 3985
    },
    {
      "entropy": 1.3566035628318787,
      "epoch": 2.0989994734070563,
      "grad_norm": 0.2742775082588196,
      "learning_rate": 0.0001677780611226476,
      "loss": 0.18728569149971008,
      "mean_token_accuracy": 0.9219802916049957,
      "num_tokens": 49169240.0,
      "step": 3986
    },
    {
      "entropy": 1.3031164109706879,
      "epoch": 2.0995260663507107,
      "grad_norm": 0.2691558301448822,
      "learning_rate": 0.00016776124649259537,
      "loss": 0.1954994797706604,
      "mean_token_accuracy": 0.9175993353128433,
      "num_tokens": 49181576.0,
      "step": 3987
    },
    {
      "entropy": 1.3594340980052948,
      "epoch": 2.1000526592943656,
      "grad_norm": 0.3043460547924042,
      "learning_rate": 0.0001677444284335946,
      "loss": 0.18774178624153137,
      "mean_token_accuracy": 0.9245233535766602,
      "num_tokens": 49193912.0,
      "step": 3988
    },
    {
      "entropy": 1.3227914571762085,
      "epoch": 2.10057925223802,
      "grad_norm": 0.28486767411231995,
      "learning_rate": 0.00016772760694664364,
      "loss": 0.18533597886562347,
      "mean_token_accuracy": 0.9241953045129776,
      "num_tokens": 49206248.0,
      "step": 3989
    },
    {
      "entropy": 1.34839728474617,
      "epoch": 2.1011058451816744,
      "grad_norm": 0.31815943121910095,
      "learning_rate": 0.00016771078203274114,
      "loss": 0.20659512281417847,
      "mean_token_accuracy": 0.9135080128908157,
      "num_tokens": 49218584.0,
      "step": 3990
    },
    {
      "entropy": 1.3131659030914307,
      "epoch": 2.1016324381253293,
      "grad_norm": 0.2616286873817444,
      "learning_rate": 0.00016769395369288588,
      "loss": 0.16908803582191467,
      "mean_token_accuracy": 0.9313696920871735,
      "num_tokens": 49230920.0,
      "step": 3991
    },
    {
      "entropy": 1.313031405210495,
      "epoch": 2.1021590310689837,
      "grad_norm": 0.27754518389701843,
      "learning_rate": 0.00016767712192807682,
      "loss": 0.19504284858703613,
      "mean_token_accuracy": 0.916403278708458,
      "num_tokens": 49243256.0,
      "step": 3992
    },
    {
      "entropy": 1.304496705532074,
      "epoch": 2.102685624012638,
      "grad_norm": 0.28971877694129944,
      "learning_rate": 0.00016766028673931323,
      "loss": 0.1822929084300995,
      "mean_token_accuracy": 0.9260184317827225,
      "num_tokens": 49255592.0,
      "step": 3993
    },
    {
      "entropy": 1.293164074420929,
      "epoch": 2.103212216956293,
      "grad_norm": 0.2651752233505249,
      "learning_rate": 0.0001676434481275945,
      "loss": 0.18674691021442413,
      "mean_token_accuracy": 0.9198154211044312,
      "num_tokens": 49267928.0,
      "step": 3994
    },
    {
      "entropy": 1.3709676265716553,
      "epoch": 2.1037388098999474,
      "grad_norm": 0.2977954149246216,
      "learning_rate": 0.0001676266060939202,
      "loss": 0.21413739025592804,
      "mean_token_accuracy": 0.9135608524084091,
      "num_tokens": 49280264.0,
      "step": 3995
    },
    {
      "entropy": 1.4219199419021606,
      "epoch": 2.104265402843602,
      "grad_norm": 0.30583176016807556,
      "learning_rate": 0.00016760976063929017,
      "loss": 0.18949593603610992,
      "mean_token_accuracy": 0.9219523072242737,
      "num_tokens": 49292600.0,
      "step": 3996
    },
    {
      "entropy": 1.4119015336036682,
      "epoch": 2.1047919957872563,
      "grad_norm": 0.29364195466041565,
      "learning_rate": 0.00016759291176470444,
      "loss": 0.19350691139698029,
      "mean_token_accuracy": 0.9228742122650146,
      "num_tokens": 49304936.0,
      "step": 3997
    },
    {
      "entropy": 1.362772524356842,
      "epoch": 2.105318588730911,
      "grad_norm": 0.2556858956813812,
      "learning_rate": 0.0001675760594711632,
      "loss": 0.183720663189888,
      "mean_token_accuracy": 0.9237812012434006,
      "num_tokens": 49317272.0,
      "step": 3998
    },
    {
      "entropy": 1.3637798428535461,
      "epoch": 2.1058451816745656,
      "grad_norm": 0.2728027403354645,
      "learning_rate": 0.0001675592037596669,
      "loss": 0.19628110527992249,
      "mean_token_accuracy": 0.9191394597291946,
      "num_tokens": 49329608.0,
      "step": 3999
    },
    {
      "entropy": 1.3810409605503082,
      "epoch": 2.10637177461822,
      "grad_norm": 0.26390331983566284,
      "learning_rate": 0.00016754234463121613,
      "loss": 0.1863417774438858,
      "mean_token_accuracy": 0.9180260598659515,
      "num_tokens": 49341944.0,
      "step": 4000
    },
    {
      "entropy": 1.3575378060340881,
      "epoch": 2.106898367561875,
      "grad_norm": 0.26593732833862305,
      "learning_rate": 0.00016752548208681176,
      "loss": 0.1913304477930069,
      "mean_token_accuracy": 0.9183992445468903,
      "num_tokens": 49354280.0,
      "step": 4001
    },
    {
      "entropy": 1.3803537786006927,
      "epoch": 2.1074249605055293,
      "grad_norm": 0.2804014980792999,
      "learning_rate": 0.0001675086161274548,
      "loss": 0.17945896089076996,
      "mean_token_accuracy": 0.9222467094659805,
      "num_tokens": 49366616.0,
      "step": 4002
    },
    {
      "entropy": 1.3277397155761719,
      "epoch": 2.1079515534491837,
      "grad_norm": 0.27084991335868835,
      "learning_rate": 0.0001674917467541465,
      "loss": 0.2013774961233139,
      "mean_token_accuracy": 0.9130637645721436,
      "num_tokens": 49378952.0,
      "step": 4003
    },
    {
      "entropy": 1.3295770585536957,
      "epoch": 2.108478146392838,
      "grad_norm": 0.2723524868488312,
      "learning_rate": 0.0001674748739678883,
      "loss": 0.18388348817825317,
      "mean_token_accuracy": 0.9248556792736053,
      "num_tokens": 49391288.0,
      "step": 4004
    },
    {
      "entropy": 1.3474342226982117,
      "epoch": 2.109004739336493,
      "grad_norm": 0.3119790554046631,
      "learning_rate": 0.00016745799776968177,
      "loss": 0.21229058504104614,
      "mean_token_accuracy": 0.9167481958866119,
      "num_tokens": 49403624.0,
      "step": 4005
    },
    {
      "entropy": 1.3330671191215515,
      "epoch": 2.1095313322801474,
      "grad_norm": 0.28843197226524353,
      "learning_rate": 0.0001674411181605288,
      "loss": 0.2047140747308731,
      "mean_token_accuracy": 0.9146516472101212,
      "num_tokens": 49415960.0,
      "step": 4006
    },
    {
      "entropy": 1.4152400493621826,
      "epoch": 2.110057925223802,
      "grad_norm": 0.30813169479370117,
      "learning_rate": 0.00016742423514143147,
      "loss": 0.1919257640838623,
      "mean_token_accuracy": 0.9214138835668564,
      "num_tokens": 49428296.0,
      "step": 4007
    },
    {
      "entropy": 1.3549103736877441,
      "epoch": 2.1105845181674567,
      "grad_norm": 0.2713034451007843,
      "learning_rate": 0.000167407348713392,
      "loss": 0.199094757437706,
      "mean_token_accuracy": 0.9188733994960785,
      "num_tokens": 49440632.0,
      "step": 4008
    },
    {
      "entropy": 1.35169118642807,
      "epoch": 2.111111111111111,
      "grad_norm": 0.26357969641685486,
      "learning_rate": 0.00016739045887741283,
      "loss": 0.18052592873573303,
      "mean_token_accuracy": 0.9184356480836868,
      "num_tokens": 49452968.0,
      "step": 4009
    },
    {
      "entropy": 1.3346559405326843,
      "epoch": 2.1116377040547656,
      "grad_norm": 0.27484050393104553,
      "learning_rate": 0.0001673735656344966,
      "loss": 0.18684698641300201,
      "mean_token_accuracy": 0.9233911037445068,
      "num_tokens": 49465304.0,
      "step": 4010
    },
    {
      "entropy": 1.288124918937683,
      "epoch": 2.11216429699842,
      "grad_norm": 0.2778557240962982,
      "learning_rate": 0.00016735666898564616,
      "loss": 0.20437434315681458,
      "mean_token_accuracy": 0.9145655333995819,
      "num_tokens": 49477640.0,
      "step": 4011
    },
    {
      "entropy": 1.4040893018245697,
      "epoch": 2.112690889942075,
      "grad_norm": 0.2719804644584656,
      "learning_rate": 0.0001673397689318646,
      "loss": 0.17971079051494598,
      "mean_token_accuracy": 0.9259855449199677,
      "num_tokens": 49489976.0,
      "step": 4012
    },
    {
      "entropy": 1.3603813648223877,
      "epoch": 2.1132174828857293,
      "grad_norm": 0.28585880994796753,
      "learning_rate": 0.0001673228654741552,
      "loss": 0.18481558561325073,
      "mean_token_accuracy": 0.9266578257083893,
      "num_tokens": 49502312.0,
      "step": 4013
    },
    {
      "entropy": 1.2925344407558441,
      "epoch": 2.1137440758293837,
      "grad_norm": 0.26038700342178345,
      "learning_rate": 0.00016730595861352132,
      "loss": 0.18866756558418274,
      "mean_token_accuracy": 0.9256009310483932,
      "num_tokens": 49514648.0,
      "step": 4014
    },
    {
      "entropy": 1.2987899482250214,
      "epoch": 2.1142706687730386,
      "grad_norm": 0.28457188606262207,
      "learning_rate": 0.0001672890483509667,
      "loss": 0.20553961396217346,
      "mean_token_accuracy": 0.9197067618370056,
      "num_tokens": 49526984.0,
      "step": 4015
    },
    {
      "entropy": 1.3735082149505615,
      "epoch": 2.114797261716693,
      "grad_norm": 0.2766115367412567,
      "learning_rate": 0.00016727213468749517,
      "loss": 0.1972998082637787,
      "mean_token_accuracy": 0.9184480607509613,
      "num_tokens": 49539320.0,
      "step": 4016
    },
    {
      "entropy": 1.395142674446106,
      "epoch": 2.1153238546603474,
      "grad_norm": 0.2885947823524475,
      "learning_rate": 0.00016725521762411084,
      "loss": 0.18944555521011353,
      "mean_token_accuracy": 0.9218645691871643,
      "num_tokens": 49551656.0,
      "step": 4017
    },
    {
      "entropy": 1.397442787885666,
      "epoch": 2.1158504476040023,
      "grad_norm": 0.31787705421447754,
      "learning_rate": 0.00016723829716181797,
      "loss": 0.2024865299463272,
      "mean_token_accuracy": 0.9169978648424149,
      "num_tokens": 49563992.0,
      "step": 4018
    },
    {
      "entropy": 1.3703459203243256,
      "epoch": 2.1163770405476567,
      "grad_norm": 0.27631449699401855,
      "learning_rate": 0.00016722137330162096,
      "loss": 0.1807280331850052,
      "mean_token_accuracy": 0.9232948869466782,
      "num_tokens": 49576328.0,
      "step": 4019
    },
    {
      "entropy": 1.34648397564888,
      "epoch": 2.116903633491311,
      "grad_norm": 0.2967493534088135,
      "learning_rate": 0.00016720444604452454,
      "loss": 0.20159119367599487,
      "mean_token_accuracy": 0.9131289422512054,
      "num_tokens": 49588664.0,
      "step": 4020
    },
    {
      "entropy": 1.3790269792079926,
      "epoch": 2.1174302264349656,
      "grad_norm": 0.28901422023773193,
      "learning_rate": 0.00016718751539153358,
      "loss": 0.19655507802963257,
      "mean_token_accuracy": 0.9193400889635086,
      "num_tokens": 49601000.0,
      "step": 4021
    },
    {
      "entropy": 1.3603746592998505,
      "epoch": 2.1179568193786205,
      "grad_norm": 0.28060466051101685,
      "learning_rate": 0.0001671705813436531,
      "loss": 0.1773560345172882,
      "mean_token_accuracy": 0.9246945679187775,
      "num_tokens": 49613336.0,
      "step": 4022
    },
    {
      "entropy": 1.3825352787971497,
      "epoch": 2.118483412322275,
      "grad_norm": 0.30686402320861816,
      "learning_rate": 0.0001671536439018885,
      "loss": 0.20738644897937775,
      "mean_token_accuracy": 0.9165382385253906,
      "num_tokens": 49625672.0,
      "step": 4023
    },
    {
      "entropy": 1.353473424911499,
      "epoch": 2.1190100052659293,
      "grad_norm": 0.2645687460899353,
      "learning_rate": 0.00016713670306724512,
      "loss": 0.18006260693073273,
      "mean_token_accuracy": 0.9287989288568497,
      "num_tokens": 49638008.0,
      "step": 4024
    },
    {
      "entropy": 1.329434186220169,
      "epoch": 2.119536598209584,
      "grad_norm": 0.23572540283203125,
      "learning_rate": 0.00016711975884072872,
      "loss": 0.1625441312789917,
      "mean_token_accuracy": 0.9347212463617325,
      "num_tokens": 49650344.0,
      "step": 4025
    },
    {
      "entropy": 1.37681046128273,
      "epoch": 2.1200631911532386,
      "grad_norm": 0.25571563839912415,
      "learning_rate": 0.00016710281122334514,
      "loss": 0.17739693820476532,
      "mean_token_accuracy": 0.9257268309593201,
      "num_tokens": 49662680.0,
      "step": 4026
    },
    {
      "entropy": 1.366820365190506,
      "epoch": 2.120589784096893,
      "grad_norm": 0.2715073227882385,
      "learning_rate": 0.0001670858602161005,
      "loss": 0.200881227850914,
      "mean_token_accuracy": 0.9174215346574783,
      "num_tokens": 49675016.0,
      "step": 4027
    },
    {
      "entropy": 1.3869880437850952,
      "epoch": 2.1211163770405475,
      "grad_norm": 0.27762705087661743,
      "learning_rate": 0.00016706890582000105,
      "loss": 0.1989072561264038,
      "mean_token_accuracy": 0.9190517961978912,
      "num_tokens": 49687352.0,
      "step": 4028
    },
    {
      "entropy": 1.3485141694545746,
      "epoch": 2.1216429699842023,
      "grad_norm": 0.28710541129112244,
      "learning_rate": 0.0001670519480360533,
      "loss": 0.20247332751750946,
      "mean_token_accuracy": 0.9163581281900406,
      "num_tokens": 49699688.0,
      "step": 4029
    },
    {
      "entropy": 1.3902212977409363,
      "epoch": 2.1221695629278567,
      "grad_norm": 0.2807939350605011,
      "learning_rate": 0.0001670349868652639,
      "loss": 0.18549507856369019,
      "mean_token_accuracy": 0.9232971370220184,
      "num_tokens": 49712024.0,
      "step": 4030
    },
    {
      "entropy": 1.2936671376228333,
      "epoch": 2.122696155871511,
      "grad_norm": 0.277076780796051,
      "learning_rate": 0.00016701802230863978,
      "loss": 0.18881529569625854,
      "mean_token_accuracy": 0.9208351224660873,
      "num_tokens": 49724360.0,
      "step": 4031
    },
    {
      "entropy": 1.2952979803085327,
      "epoch": 2.123222748815166,
      "grad_norm": 0.28867289423942566,
      "learning_rate": 0.00016700105436718795,
      "loss": 0.18886399269104004,
      "mean_token_accuracy": 0.9231124371290207,
      "num_tokens": 49736696.0,
      "step": 4032
    },
    {
      "entropy": 1.35812708735466,
      "epoch": 2.1237493417588205,
      "grad_norm": 0.30915045738220215,
      "learning_rate": 0.00016698408304191578,
      "loss": 0.21172593533992767,
      "mean_token_accuracy": 0.918636828660965,
      "num_tokens": 49749032.0,
      "step": 4033
    },
    {
      "entropy": 1.29549902677536,
      "epoch": 2.124275934702475,
      "grad_norm": 0.2771642804145813,
      "learning_rate": 0.00016696710833383075,
      "loss": 0.17263811826705933,
      "mean_token_accuracy": 0.9258449673652649,
      "num_tokens": 49761368.0,
      "step": 4034
    },
    {
      "entropy": 1.2696373462677002,
      "epoch": 2.1248025276461293,
      "grad_norm": 0.2755196690559387,
      "learning_rate": 0.00016695013024394052,
      "loss": 0.18899832665920258,
      "mean_token_accuracy": 0.9189085364341736,
      "num_tokens": 49773704.0,
      "step": 4035
    },
    {
      "entropy": 1.253807783126831,
      "epoch": 2.125329120589784,
      "grad_norm": 0.28885576128959656,
      "learning_rate": 0.00016693314877325294,
      "loss": 0.19073623418807983,
      "mean_token_accuracy": 0.9253068566322327,
      "num_tokens": 49786040.0,
      "step": 4036
    },
    {
      "entropy": 1.2731792032718658,
      "epoch": 2.1258557135334386,
      "grad_norm": 0.3004090189933777,
      "learning_rate": 0.00016691616392277618,
      "loss": 0.19338059425354004,
      "mean_token_accuracy": 0.9188876748085022,
      "num_tokens": 49798376.0,
      "step": 4037
    },
    {
      "entropy": 1.210568755865097,
      "epoch": 2.126382306477093,
      "grad_norm": 0.2848544418811798,
      "learning_rate": 0.00016689917569351852,
      "loss": 0.20599129796028137,
      "mean_token_accuracy": 0.9142592400312424,
      "num_tokens": 49810712.0,
      "step": 4038
    },
    {
      "entropy": 1.2268668115139008,
      "epoch": 2.126908899420748,
      "grad_norm": 0.2891899645328522,
      "learning_rate": 0.0001668821840864884,
      "loss": 0.2129664123058319,
      "mean_token_accuracy": 0.9123159050941467,
      "num_tokens": 49823048.0,
      "step": 4039
    },
    {
      "entropy": 1.1677944362163544,
      "epoch": 2.1274354923644023,
      "grad_norm": 0.280416876077652,
      "learning_rate": 0.00016686518910269459,
      "loss": 0.19316363334655762,
      "mean_token_accuracy": 0.9191536903381348,
      "num_tokens": 49835384.0,
      "step": 4040
    },
    {
      "entropy": 1.1698110401630402,
      "epoch": 2.1279620853080567,
      "grad_norm": 0.2729722857475281,
      "learning_rate": 0.00016684819074314587,
      "loss": 0.17371591925621033,
      "mean_token_accuracy": 0.9292128086090088,
      "num_tokens": 49847720.0,
      "step": 4041
    },
    {
      "entropy": 1.1880428791046143,
      "epoch": 2.1284886782517116,
      "grad_norm": 0.2968471050262451,
      "learning_rate": 0.00016683118900885147,
      "loss": 0.19274401664733887,
      "mean_token_accuracy": 0.9200466126203537,
      "num_tokens": 49860056.0,
      "step": 4042
    },
    {
      "entropy": 1.1668127179145813,
      "epoch": 2.129015271195366,
      "grad_norm": 0.24909040331840515,
      "learning_rate": 0.0001668141839008206,
      "loss": 0.1683623492717743,
      "mean_token_accuracy": 0.9272440522909164,
      "num_tokens": 49872392.0,
      "step": 4043
    },
    {
      "entropy": 1.2406319081783295,
      "epoch": 2.1295418641390205,
      "grad_norm": 0.32238873839378357,
      "learning_rate": 0.00016679717542006275,
      "loss": 0.19797292351722717,
      "mean_token_accuracy": 0.9142682403326035,
      "num_tokens": 49884728.0,
      "step": 4044
    },
    {
      "entropy": 1.1888574957847595,
      "epoch": 2.130068457082675,
      "grad_norm": 0.2867093086242676,
      "learning_rate": 0.00016678016356758764,
      "loss": 0.19565987586975098,
      "mean_token_accuracy": 0.921043261885643,
      "num_tokens": 49897064.0,
      "step": 4045
    },
    {
      "entropy": 1.185506522655487,
      "epoch": 2.1305950500263298,
      "grad_norm": 0.29846811294555664,
      "learning_rate": 0.00016676314834440518,
      "loss": 0.20893250405788422,
      "mean_token_accuracy": 0.9120955020189285,
      "num_tokens": 49909400.0,
      "step": 4046
    },
    {
      "entropy": 1.1691254079341888,
      "epoch": 2.131121642969984,
      "grad_norm": 0.2745888829231262,
      "learning_rate": 0.00016674612975152542,
      "loss": 0.18490786850452423,
      "mean_token_accuracy": 0.9252546727657318,
      "num_tokens": 49921736.0,
      "step": 4047
    },
    {
      "entropy": 1.1454492807388306,
      "epoch": 2.1316482359136386,
      "grad_norm": 0.2868290841579437,
      "learning_rate": 0.00016672910778995866,
      "loss": 0.20495843887329102,
      "mean_token_accuracy": 0.9156783670186996,
      "num_tokens": 49934072.0,
      "step": 4048
    },
    {
      "entropy": 1.1920296549797058,
      "epoch": 2.1321748288572935,
      "grad_norm": 0.2875182628631592,
      "learning_rate": 0.00016671208246071546,
      "loss": 0.20269200205802917,
      "mean_token_accuracy": 0.9209336042404175,
      "num_tokens": 49946408.0,
      "step": 4049
    },
    {
      "entropy": 1.194561243057251,
      "epoch": 2.132701421800948,
      "grad_norm": 0.28611719608306885,
      "learning_rate": 0.00016669505376480642,
      "loss": 0.18004143238067627,
      "mean_token_accuracy": 0.9242940843105316,
      "num_tokens": 49958744.0,
      "step": 4050
    },
    {
      "entropy": 1.1752946376800537,
      "epoch": 2.1332280147446023,
      "grad_norm": 0.2773791253566742,
      "learning_rate": 0.00016667802170324252,
      "loss": 0.17548750340938568,
      "mean_token_accuracy": 0.9260925948619843,
      "num_tokens": 49971080.0,
      "step": 4051
    },
    {
      "entropy": 1.1404151320457458,
      "epoch": 2.1337546076882568,
      "grad_norm": 0.24352863430976868,
      "learning_rate": 0.0001666609862770348,
      "loss": 0.17417335510253906,
      "mean_token_accuracy": 0.9301496148109436,
      "num_tokens": 49983416.0,
      "step": 4052
    },
    {
      "entropy": 1.1148101389408112,
      "epoch": 2.1342812006319116,
      "grad_norm": 0.27176782488822937,
      "learning_rate": 0.0001666439474871946,
      "loss": 0.1892414093017578,
      "mean_token_accuracy": 0.9190666973590851,
      "num_tokens": 49995752.0,
      "step": 4053
    },
    {
      "entropy": 1.121658980846405,
      "epoch": 2.134807793575566,
      "grad_norm": 0.2826327979564667,
      "learning_rate": 0.00016662690533473333,
      "loss": 0.2030116468667984,
      "mean_token_accuracy": 0.921414390206337,
      "num_tokens": 50008088.0,
      "step": 4054
    },
    {
      "entropy": 1.1173063218593597,
      "epoch": 2.1353343865192205,
      "grad_norm": 0.2585664689540863,
      "learning_rate": 0.00016660985982066277,
      "loss": 0.177621528506279,
      "mean_token_accuracy": 0.9258757680654526,
      "num_tokens": 50020424.0,
      "step": 4055
    },
    {
      "entropy": 1.1216989159584045,
      "epoch": 2.1358609794628753,
      "grad_norm": 0.27996382117271423,
      "learning_rate": 0.00016659281094599478,
      "loss": 0.18978063762187958,
      "mean_token_accuracy": 0.9182009845972061,
      "num_tokens": 50032760.0,
      "step": 4056
    },
    {
      "entropy": 1.1314687132835388,
      "epoch": 2.1363875724065298,
      "grad_norm": 0.30307766795158386,
      "learning_rate": 0.00016657575871174143,
      "loss": 0.20396795868873596,
      "mean_token_accuracy": 0.9122677892446518,
      "num_tokens": 50045096.0,
      "step": 4057
    },
    {
      "entropy": 1.0608826130628586,
      "epoch": 2.136914165350184,
      "grad_norm": 0.27314090728759766,
      "learning_rate": 0.00016655870311891503,
      "loss": 0.19006496667861938,
      "mean_token_accuracy": 0.9235405623912811,
      "num_tokens": 50057432.0,
      "step": 4058
    },
    {
      "entropy": 1.0740149021148682,
      "epoch": 2.137440758293839,
      "grad_norm": 0.273417204618454,
      "learning_rate": 0.00016654164416852812,
      "loss": 0.1910788118839264,
      "mean_token_accuracy": 0.9203030318021774,
      "num_tokens": 50069768.0,
      "step": 4059
    },
    {
      "entropy": 1.0903103351593018,
      "epoch": 2.1379673512374935,
      "grad_norm": 0.2700668275356293,
      "learning_rate": 0.0001665245818615933,
      "loss": 0.1629205048084259,
      "mean_token_accuracy": 0.9328511208295822,
      "num_tokens": 50082104.0,
      "step": 4060
    },
    {
      "entropy": 1.1036777794361115,
      "epoch": 2.138493944181148,
      "grad_norm": 0.285442054271698,
      "learning_rate": 0.00016650751619912355,
      "loss": 0.1832047700881958,
      "mean_token_accuracy": 0.920674666762352,
      "num_tokens": 50094440.0,
      "step": 4061
    },
    {
      "entropy": 1.0578563511371613,
      "epoch": 2.1390205371248023,
      "grad_norm": 0.2827666103839874,
      "learning_rate": 0.00016649044718213186,
      "loss": 0.17583194375038147,
      "mean_token_accuracy": 0.9268711805343628,
      "num_tokens": 50106776.0,
      "step": 4062
    },
    {
      "entropy": 1.040594071149826,
      "epoch": 2.139547130068457,
      "grad_norm": 0.289329469203949,
      "learning_rate": 0.00016647337481163162,
      "loss": 0.18696846067905426,
      "mean_token_accuracy": 0.9217635691165924,
      "num_tokens": 50119112.0,
      "step": 4063
    },
    {
      "entropy": 1.076176255941391,
      "epoch": 2.1400737230121116,
      "grad_norm": 0.2821875810623169,
      "learning_rate": 0.00016645629908863623,
      "loss": 0.18235281109809875,
      "mean_token_accuracy": 0.927731066942215,
      "num_tokens": 50131448.0,
      "step": 4064
    },
    {
      "entropy": 1.0617471039295197,
      "epoch": 2.140600315955766,
      "grad_norm": 0.29055875539779663,
      "learning_rate": 0.00016643922001415945,
      "loss": 0.19152817130088806,
      "mean_token_accuracy": 0.921459436416626,
      "num_tokens": 50143784.0,
      "step": 4065
    },
    {
      "entropy": 1.0784454494714737,
      "epoch": 2.141126908899421,
      "grad_norm": 0.28317174315452576,
      "learning_rate": 0.0001664221375892151,
      "loss": 0.1880769580602646,
      "mean_token_accuracy": 0.9215621948242188,
      "num_tokens": 50156120.0,
      "step": 4066
    },
    {
      "entropy": 1.1484303772449493,
      "epoch": 2.1416535018430753,
      "grad_norm": 0.29933735728263855,
      "learning_rate": 0.0001664050518148173,
      "loss": 0.19272185862064362,
      "mean_token_accuracy": 0.9227567911148071,
      "num_tokens": 50168456.0,
      "step": 4067
    },
    {
      "entropy": 1.1309453547000885,
      "epoch": 2.1421800947867298,
      "grad_norm": 0.3005389869213104,
      "learning_rate": 0.00016638796269198035,
      "loss": 0.20710280537605286,
      "mean_token_accuracy": 0.9150481820106506,
      "num_tokens": 50180792.0,
      "step": 4068
    },
    {
      "entropy": 1.1589435040950775,
      "epoch": 2.142706687730384,
      "grad_norm": 0.3221890926361084,
      "learning_rate": 0.00016637087022171868,
      "loss": 0.20620915293693542,
      "mean_token_accuracy": 0.9166271984577179,
      "num_tokens": 50193128.0,
      "step": 4069
    },
    {
      "entropy": 1.106980949640274,
      "epoch": 2.143233280674039,
      "grad_norm": 0.2595953047275543,
      "learning_rate": 0.00016635377440504706,
      "loss": 0.1830139011144638,
      "mean_token_accuracy": 0.9249969273805618,
      "num_tokens": 50205464.0,
      "step": 4070
    },
    {
      "entropy": 1.1292275488376617,
      "epoch": 2.1437598736176935,
      "grad_norm": 0.28772836923599243,
      "learning_rate": 0.0001663366752429802,
      "loss": 0.2001112997531891,
      "mean_token_accuracy": 0.9196556061506271,
      "num_tokens": 50217800.0,
      "step": 4071
    },
    {
      "entropy": 1.149866372346878,
      "epoch": 2.144286466561348,
      "grad_norm": 0.25645655393600464,
      "learning_rate": 0.0001663195727365334,
      "loss": 0.17879578471183777,
      "mean_token_accuracy": 0.9276049882173538,
      "num_tokens": 50230136.0,
      "step": 4072
    },
    {
      "entropy": 1.1402757167816162,
      "epoch": 2.144813059505003,
      "grad_norm": 0.24784688651561737,
      "learning_rate": 0.00016630246688672176,
      "loss": 0.16923773288726807,
      "mean_token_accuracy": 0.9288965910673141,
      "num_tokens": 50242472.0,
      "step": 4073
    },
    {
      "entropy": 1.1733649671077728,
      "epoch": 2.145339652448657,
      "grad_norm": 0.2653653025627136,
      "learning_rate": 0.00016628535769456083,
      "loss": 0.19141364097595215,
      "mean_token_accuracy": 0.9215678870677948,
      "num_tokens": 50254808.0,
      "step": 4074
    },
    {
      "entropy": 1.1664075255393982,
      "epoch": 2.1458662453923116,
      "grad_norm": 0.2801195979118347,
      "learning_rate": 0.00016626824516106624,
      "loss": 0.2021307796239853,
      "mean_token_accuracy": 0.918086051940918,
      "num_tokens": 50267144.0,
      "step": 4075
    },
    {
      "entropy": 1.1810933351516724,
      "epoch": 2.1463928383359665,
      "grad_norm": 0.2909698486328125,
      "learning_rate": 0.00016625112928725394,
      "loss": 0.2095974087715149,
      "mean_token_accuracy": 0.9153055250644684,
      "num_tokens": 50279480.0,
      "step": 4076
    },
    {
      "entropy": 1.1292328536510468,
      "epoch": 2.146919431279621,
      "grad_norm": 0.24777865409851074,
      "learning_rate": 0.00016623401007413995,
      "loss": 0.18146950006484985,
      "mean_token_accuracy": 0.9228164702653885,
      "num_tokens": 50291816.0,
      "step": 4077
    },
    {
      "entropy": 1.2031064331531525,
      "epoch": 2.1474460242232754,
      "grad_norm": 0.29414162039756775,
      "learning_rate": 0.0001662168875227405,
      "loss": 0.19175127148628235,
      "mean_token_accuracy": 0.9232201874256134,
      "num_tokens": 50304152.0,
      "step": 4078
    },
    {
      "entropy": 1.2237733602523804,
      "epoch": 2.1479726171669298,
      "grad_norm": 0.2574436366558075,
      "learning_rate": 0.00016619976163407212,
      "loss": 0.18326136469841003,
      "mean_token_accuracy": 0.927291065454483,
      "num_tokens": 50316488.0,
      "step": 4079
    },
    {
      "entropy": 1.1478267908096313,
      "epoch": 2.1484992101105846,
      "grad_norm": 0.25556036829948425,
      "learning_rate": 0.00016618263240915143,
      "loss": 0.17815645039081573,
      "mean_token_accuracy": 0.9254791438579559,
      "num_tokens": 50328824.0,
      "step": 4080
    },
    {
      "entropy": 1.1542611420154572,
      "epoch": 2.149025803054239,
      "grad_norm": 0.2943160831928253,
      "learning_rate": 0.00016616549984899533,
      "loss": 0.2042732536792755,
      "mean_token_accuracy": 0.9145019352436066,
      "num_tokens": 50341160.0,
      "step": 4081
    },
    {
      "entropy": 1.1703786253929138,
      "epoch": 2.1495523959978935,
      "grad_norm": 0.29997357726097107,
      "learning_rate": 0.00016614836395462086,
      "loss": 0.18360967934131622,
      "mean_token_accuracy": 0.922757014632225,
      "num_tokens": 50353496.0,
      "step": 4082
    },
    {
      "entropy": 1.14296355843544,
      "epoch": 2.1500789889415484,
      "grad_norm": 0.4659420847892761,
      "learning_rate": 0.00016613122472704527,
      "loss": 0.19003994762897491,
      "mean_token_accuracy": 0.9201502650976181,
      "num_tokens": 50365832.0,
      "step": 4083
    },
    {
      "entropy": 1.2191529273986816,
      "epoch": 2.150605581885203,
      "grad_norm": 0.2996703088283539,
      "learning_rate": 0.00016611408216728603,
      "loss": 0.2007676362991333,
      "mean_token_accuracy": 0.9181395620107651,
      "num_tokens": 50378168.0,
      "step": 4084
    },
    {
      "entropy": 1.2132177352905273,
      "epoch": 2.151132174828857,
      "grad_norm": 0.25647181272506714,
      "learning_rate": 0.0001660969362763608,
      "loss": 0.1759507954120636,
      "mean_token_accuracy": 0.9269961714744568,
      "num_tokens": 50390504.0,
      "step": 4085
    },
    {
      "entropy": 1.220710039138794,
      "epoch": 2.1516587677725116,
      "grad_norm": 0.26862823963165283,
      "learning_rate": 0.0001660797870552874,
      "loss": 0.20389479398727417,
      "mean_token_accuracy": 0.9192573726177216,
      "num_tokens": 50402840.0,
      "step": 4086
    },
    {
      "entropy": 1.3179146647453308,
      "epoch": 2.1521853607161665,
      "grad_norm": 0.32815021276474,
      "learning_rate": 0.00016606263450508391,
      "loss": 0.20844385027885437,
      "mean_token_accuracy": 0.9123887866735458,
      "num_tokens": 50415176.0,
      "step": 4087
    },
    {
      "entropy": 1.270151823759079,
      "epoch": 2.152711953659821,
      "grad_norm": 0.28897103667259216,
      "learning_rate": 0.00016604547862676856,
      "loss": 0.19628524780273438,
      "mean_token_accuracy": 0.9163263142108917,
      "num_tokens": 50427512.0,
      "step": 4088
    },
    {
      "entropy": 1.297392338514328,
      "epoch": 2.1532385466034754,
      "grad_norm": 0.29361578822135925,
      "learning_rate": 0.0001660283194213598,
      "loss": 0.1971634477376938,
      "mean_token_accuracy": 0.9190499484539032,
      "num_tokens": 50439848.0,
      "step": 4089
    },
    {
      "entropy": 1.2932831943035126,
      "epoch": 2.1537651395471302,
      "grad_norm": 0.25311169028282166,
      "learning_rate": 0.0001660111568898763,
      "loss": 0.18418122828006744,
      "mean_token_accuracy": 0.9244278967380524,
      "num_tokens": 50452184.0,
      "step": 4090
    },
    {
      "entropy": 1.3569526970386505,
      "epoch": 2.1542917324907846,
      "grad_norm": 0.2899273633956909,
      "learning_rate": 0.00016599399103333685,
      "loss": 0.21381260454654694,
      "mean_token_accuracy": 0.9070511162281036,
      "num_tokens": 50464520.0,
      "step": 4091
    },
    {
      "entropy": 1.378174602985382,
      "epoch": 2.154818325434439,
      "grad_norm": 0.2990644574165344,
      "learning_rate": 0.00016597682185276049,
      "loss": 0.2194102704524994,
      "mean_token_accuracy": 0.9084511697292328,
      "num_tokens": 50476856.0,
      "step": 4092
    },
    {
      "entropy": 1.399436503648758,
      "epoch": 2.155344918378094,
      "grad_norm": 0.2860548198223114,
      "learning_rate": 0.00016595964934916653,
      "loss": 0.19893941283226013,
      "mean_token_accuracy": 0.9167734533548355,
      "num_tokens": 50489192.0,
      "step": 4093
    },
    {
      "entropy": 1.370680034160614,
      "epoch": 2.1558715113217484,
      "grad_norm": 0.28559643030166626,
      "learning_rate": 0.00016594247352357433,
      "loss": 0.19687819480895996,
      "mean_token_accuracy": 0.9142584204673767,
      "num_tokens": 50501528.0,
      "step": 4094
    },
    {
      "entropy": 1.359871655702591,
      "epoch": 2.156398104265403,
      "grad_norm": 0.2813403010368347,
      "learning_rate": 0.0001659252943770035,
      "loss": 0.21558703482151031,
      "mean_token_accuracy": 0.9130292385816574,
      "num_tokens": 50513864.0,
      "step": 4095
    },
    {
      "entropy": 1.4364489018917084,
      "epoch": 2.156924697209057,
      "grad_norm": 0.2761105000972748,
      "learning_rate": 0.00016590811191047393,
      "loss": 0.19772423803806305,
      "mean_token_accuracy": 0.916876494884491,
      "num_tokens": 50526200.0,
      "step": 4096
    },
    {
      "entropy": 1.4437676966190338,
      "epoch": 2.157451290152712,
      "grad_norm": 0.26975753903388977,
      "learning_rate": 0.0001658909261250056,
      "loss": 0.18524104356765747,
      "mean_token_accuracy": 0.921722486615181,
      "num_tokens": 50538536.0,
      "step": 4097
    },
    {
      "entropy": 1.4069486558437347,
      "epoch": 2.1579778830963665,
      "grad_norm": 0.2700808346271515,
      "learning_rate": 0.00016587373702161876,
      "loss": 0.1872127503156662,
      "mean_token_accuracy": 0.9228634834289551,
      "num_tokens": 50550872.0,
      "step": 4098
    },
    {
      "entropy": 1.4102502763271332,
      "epoch": 2.158504476040021,
      "grad_norm": 0.28084951639175415,
      "learning_rate": 0.00016585654460133376,
      "loss": 0.1960507333278656,
      "mean_token_accuracy": 0.9219342321157455,
      "num_tokens": 50563208.0,
      "step": 4099
    },
    {
      "entropy": 1.4461718797683716,
      "epoch": 2.159031068983676,
      "grad_norm": 0.2957639992237091,
      "learning_rate": 0.0001658393488651713,
      "loss": 0.18612340092658997,
      "mean_token_accuracy": 0.9263795763254166,
      "num_tokens": 50575544.0,
      "step": 4100
    },
    {
      "entropy": 1.4473920464515686,
      "epoch": 2.1595576619273302,
      "grad_norm": 0.261575311422348,
      "learning_rate": 0.00016582214981415217,
      "loss": 0.1952032893896103,
      "mean_token_accuracy": 0.9171668738126755,
      "num_tokens": 50587880.0,
      "step": 4101
    },
    {
      "entropy": 1.4263883233070374,
      "epoch": 2.1600842548709847,
      "grad_norm": 0.33648961782455444,
      "learning_rate": 0.00016580494744929735,
      "loss": 0.1953122764825821,
      "mean_token_accuracy": 0.9167916476726532,
      "num_tokens": 50600216.0,
      "step": 4102
    },
    {
      "entropy": 1.4109848737716675,
      "epoch": 2.160610847814639,
      "grad_norm": 0.2898204028606415,
      "learning_rate": 0.00016578774177162807,
      "loss": 0.19459910690784454,
      "mean_token_accuracy": 0.918088361620903,
      "num_tokens": 50612552.0,
      "step": 4103
    },
    {
      "entropy": 1.4371325075626373,
      "epoch": 2.161137440758294,
      "grad_norm": 0.2821919918060303,
      "learning_rate": 0.00016577053278216567,
      "loss": 0.20267830789089203,
      "mean_token_accuracy": 0.9202786535024643,
      "num_tokens": 50624888.0,
      "step": 4104
    },
    {
      "entropy": 1.4190351366996765,
      "epoch": 2.1616640337019484,
      "grad_norm": 0.28132855892181396,
      "learning_rate": 0.00016575332048193183,
      "loss": 0.1873740255832672,
      "mean_token_accuracy": 0.9211129248142242,
      "num_tokens": 50637224.0,
      "step": 4105
    },
    {
      "entropy": 1.3971222043037415,
      "epoch": 2.162190626645603,
      "grad_norm": 0.2738102674484253,
      "learning_rate": 0.0001657361048719483,
      "loss": 0.20248481631278992,
      "mean_token_accuracy": 0.9121163636445999,
      "num_tokens": 50649560.0,
      "step": 4106
    },
    {
      "entropy": 1.482112854719162,
      "epoch": 2.1627172195892577,
      "grad_norm": 0.2849135994911194,
      "learning_rate": 0.0001657188859532371,
      "loss": 0.2112928032875061,
      "mean_token_accuracy": 0.9163548201322556,
      "num_tokens": 50661896.0,
      "step": 4107
    },
    {
      "entropy": 1.4393364191055298,
      "epoch": 2.163243812532912,
      "grad_norm": 0.27296361327171326,
      "learning_rate": 0.00016570166372682034,
      "loss": 0.1941811442375183,
      "mean_token_accuracy": 0.9172745048999786,
      "num_tokens": 50674232.0,
      "step": 4108
    },
    {
      "entropy": 1.3936603665351868,
      "epoch": 2.1637704054765665,
      "grad_norm": 0.2626595199108124,
      "learning_rate": 0.00016568443819372052,
      "loss": 0.18277618288993835,
      "mean_token_accuracy": 0.9309801757335663,
      "num_tokens": 50686568.0,
      "step": 4109
    },
    {
      "entropy": 1.369593769311905,
      "epoch": 2.1642969984202214,
      "grad_norm": 0.2655249834060669,
      "learning_rate": 0.0001656672093549601,
      "loss": 0.19196127355098724,
      "mean_token_accuracy": 0.9230923056602478,
      "num_tokens": 50698904.0,
      "step": 4110
    },
    {
      "entropy": 1.4366631209850311,
      "epoch": 2.164823591363876,
      "grad_norm": 0.29604408144950867,
      "learning_rate": 0.00016564997721156197,
      "loss": 0.1939295530319214,
      "mean_token_accuracy": 0.9189808517694473,
      "num_tokens": 50711240.0,
      "step": 4111
    },
    {
      "entropy": 1.4341607093811035,
      "epoch": 2.1653501843075302,
      "grad_norm": 0.28088974952697754,
      "learning_rate": 0.000165632741764549,
      "loss": 0.2065155804157257,
      "mean_token_accuracy": 0.9164612293243408,
      "num_tokens": 50723576.0,
      "step": 4112
    },
    {
      "entropy": 1.4529509842395782,
      "epoch": 2.1658767772511847,
      "grad_norm": 0.2820344567298889,
      "learning_rate": 0.00016561550301494443,
      "loss": 0.18620197474956512,
      "mean_token_accuracy": 0.9235643148422241,
      "num_tokens": 50735912.0,
      "step": 4113
    },
    {
      "entropy": 1.4021332263946533,
      "epoch": 2.1664033701948395,
      "grad_norm": 0.2705285847187042,
      "learning_rate": 0.0001655982609637716,
      "loss": 0.19482757151126862,
      "mean_token_accuracy": 0.917400673031807,
      "num_tokens": 50748248.0,
      "step": 4114
    },
    {
      "entropy": 1.3551163375377655,
      "epoch": 2.166929963138494,
      "grad_norm": 0.2499101608991623,
      "learning_rate": 0.0001655810156120541,
      "loss": 0.17850345373153687,
      "mean_token_accuracy": 0.9285051375627518,
      "num_tokens": 50760584.0,
      "step": 4115
    },
    {
      "entropy": 1.3864115178585052,
      "epoch": 2.1674565560821484,
      "grad_norm": 0.283306360244751,
      "learning_rate": 0.00016556376696081558,
      "loss": 0.1874375194311142,
      "mean_token_accuracy": 0.9245508462190628,
      "num_tokens": 50772920.0,
      "step": 4116
    },
    {
      "entropy": 1.3825859427452087,
      "epoch": 2.1679831490258032,
      "grad_norm": 0.25388672947883606,
      "learning_rate": 0.0001655465150110801,
      "loss": 0.17021594941616058,
      "mean_token_accuracy": 0.9311476945877075,
      "num_tokens": 50785256.0,
      "step": 4117
    },
    {
      "entropy": 1.3711049556732178,
      "epoch": 2.1685097419694577,
      "grad_norm": 0.2609105408191681,
      "learning_rate": 0.00016552925976387177,
      "loss": 0.17833492159843445,
      "mean_token_accuracy": 0.9264799654483795,
      "num_tokens": 50797592.0,
      "step": 4118
    },
    {
      "entropy": 1.3915640711784363,
      "epoch": 2.169036334913112,
      "grad_norm": 0.30802109837532043,
      "learning_rate": 0.00016551200122021494,
      "loss": 0.20191773772239685,
      "mean_token_accuracy": 0.9182671159505844,
      "num_tokens": 50809928.0,
      "step": 4119
    },
    {
      "entropy": 1.2875527441501617,
      "epoch": 2.1695629278567665,
      "grad_norm": 0.2597365975379944,
      "learning_rate": 0.00016549473938113414,
      "loss": 0.18255290389060974,
      "mean_token_accuracy": 0.9238605499267578,
      "num_tokens": 50822264.0,
      "step": 4120
    },
    {
      "entropy": 1.2972512543201447,
      "epoch": 2.1700895208004214,
      "grad_norm": 0.2637297511100769,
      "learning_rate": 0.00016547747424765412,
      "loss": 0.17539075016975403,
      "mean_token_accuracy": 0.9288503229618073,
      "num_tokens": 50834600.0,
      "step": 4121
    },
    {
      "entropy": 1.3471184074878693,
      "epoch": 2.170616113744076,
      "grad_norm": 0.29643920063972473,
      "learning_rate": 0.0001654602058207998,
      "loss": 0.19519950449466705,
      "mean_token_accuracy": 0.9189782440662384,
      "num_tokens": 50846936.0,
      "step": 4122
    },
    {
      "entropy": 1.3144559860229492,
      "epoch": 2.1711427066877302,
      "grad_norm": 0.30460768938064575,
      "learning_rate": 0.0001654429341015963,
      "loss": 0.1963726282119751,
      "mean_token_accuracy": 0.9202246516942978,
      "num_tokens": 50859272.0,
      "step": 4123
    },
    {
      "entropy": 1.2872081398963928,
      "epoch": 2.171669299631385,
      "grad_norm": 0.27515682578086853,
      "learning_rate": 0.00016542565909106896,
      "loss": 0.18613222241401672,
      "mean_token_accuracy": 0.922202005982399,
      "num_tokens": 50871608.0,
      "step": 4124
    },
    {
      "entropy": 1.3294596076011658,
      "epoch": 2.1721958925750395,
      "grad_norm": 0.2827659547328949,
      "learning_rate": 0.00016540838079024325,
      "loss": 0.1930113434791565,
      "mean_token_accuracy": 0.926962748169899,
      "num_tokens": 50883944.0,
      "step": 4125
    },
    {
      "entropy": 1.3070390224456787,
      "epoch": 2.172722485518694,
      "grad_norm": 0.29022613167762756,
      "learning_rate": 0.00016539109920014498,
      "loss": 0.18820495903491974,
      "mean_token_accuracy": 0.9214640110731125,
      "num_tokens": 50896280.0,
      "step": 4126
    },
    {
      "entropy": 1.2923785746097565,
      "epoch": 2.173249078462349,
      "grad_norm": 0.30840960144996643,
      "learning_rate": 0.00016537381432179992,
      "loss": 0.19509851932525635,
      "mean_token_accuracy": 0.9167175590991974,
      "num_tokens": 50908616.0,
      "step": 4127
    },
    {
      "entropy": 1.2731912434101105,
      "epoch": 2.1737756714060033,
      "grad_norm": 0.262724906206131,
      "learning_rate": 0.00016535652615623433,
      "loss": 0.17678092420101166,
      "mean_token_accuracy": 0.924164742231369,
      "num_tokens": 50920952.0,
      "step": 4128
    },
    {
      "entropy": 1.2588714957237244,
      "epoch": 2.1743022643496577,
      "grad_norm": 0.2767009139060974,
      "learning_rate": 0.00016533923470447436,
      "loss": 0.1992766261100769,
      "mean_token_accuracy": 0.9205213189125061,
      "num_tokens": 50933288.0,
      "step": 4129
    },
    {
      "entropy": 1.2620333433151245,
      "epoch": 2.174828857293312,
      "grad_norm": 0.27543267607688904,
      "learning_rate": 0.0001653219399675466,
      "loss": 0.1907360702753067,
      "mean_token_accuracy": 0.9243224263191223,
      "num_tokens": 50945624.0,
      "step": 4130
    },
    {
      "entropy": 1.251461535692215,
      "epoch": 2.175355450236967,
      "grad_norm": 0.2669181823730469,
      "learning_rate": 0.00016530464194647768,
      "loss": 0.19080904126167297,
      "mean_token_accuracy": 0.9203681647777557,
      "num_tokens": 50957960.0,
      "step": 4131
    },
    {
      "entropy": 1.2553662955760956,
      "epoch": 2.1758820431806214,
      "grad_norm": 0.2973247468471527,
      "learning_rate": 0.0001652873406422945,
      "loss": 0.2005629688501358,
      "mean_token_accuracy": 0.9205776154994965,
      "num_tokens": 50970296.0,
      "step": 4132
    },
    {
      "entropy": 1.2845535278320312,
      "epoch": 2.176408636124276,
      "grad_norm": 0.2855272591114044,
      "learning_rate": 0.00016527003605602418,
      "loss": 0.19210195541381836,
      "mean_token_accuracy": 0.9204275608062744,
      "num_tokens": 50982632.0,
      "step": 4133
    },
    {
      "entropy": 1.2553153038024902,
      "epoch": 2.1769352290679307,
      "grad_norm": 0.2743854224681854,
      "learning_rate": 0.00016525272818869392,
      "loss": 0.19348418712615967,
      "mean_token_accuracy": 0.9160150438547134,
      "num_tokens": 50994968.0,
      "step": 4134
    },
    {
      "entropy": 1.2861834466457367,
      "epoch": 2.177461822011585,
      "grad_norm": 0.3126857578754425,
      "learning_rate": 0.00016523541704133128,
      "loss": 0.2036605328321457,
      "mean_token_accuracy": 0.91865274310112,
      "num_tokens": 51007304.0,
      "step": 4135
    },
    {
      "entropy": 1.2617907226085663,
      "epoch": 2.1779884149552395,
      "grad_norm": 0.2782585620880127,
      "learning_rate": 0.00016521810261496385,
      "loss": 0.1961769163608551,
      "mean_token_accuracy": 0.9191503077745438,
      "num_tokens": 51019640.0,
      "step": 4136
    },
    {
      "entropy": 1.2762066721916199,
      "epoch": 2.178515007898894,
      "grad_norm": 0.2730770707130432,
      "learning_rate": 0.00016520078491061944,
      "loss": 0.20689794421195984,
      "mean_token_accuracy": 0.9149703979492188,
      "num_tokens": 51031976.0,
      "step": 4137
    },
    {
      "entropy": 1.272611141204834,
      "epoch": 2.179041600842549,
      "grad_norm": 0.2779129445552826,
      "learning_rate": 0.00016518346392932625,
      "loss": 0.18954318761825562,
      "mean_token_accuracy": 0.9203405678272247,
      "num_tokens": 51044312.0,
      "step": 4138
    },
    {
      "entropy": 1.2742144763469696,
      "epoch": 2.1795681937862033,
      "grad_norm": 0.26842811703681946,
      "learning_rate": 0.00016516613967211237,
      "loss": 0.19438216090202332,
      "mean_token_accuracy": 0.9190889149904251,
      "num_tokens": 51056648.0,
      "step": 4139
    },
    {
      "entropy": 1.2541859447956085,
      "epoch": 2.1800947867298577,
      "grad_norm": 0.25953954458236694,
      "learning_rate": 0.00016514881214000637,
      "loss": 0.16562017798423767,
      "mean_token_accuracy": 0.9315860420465469,
      "num_tokens": 51068984.0,
      "step": 4140
    },
    {
      "entropy": 1.2067217528820038,
      "epoch": 2.1806213796735125,
      "grad_norm": 0.26448673009872437,
      "learning_rate": 0.00016513148133403678,
      "loss": 0.19268856942653656,
      "mean_token_accuracy": 0.9189769923686981,
      "num_tokens": 51081320.0,
      "step": 4141
    },
    {
      "entropy": 1.2760578393936157,
      "epoch": 2.181147972617167,
      "grad_norm": 0.28249260783195496,
      "learning_rate": 0.00016511414725523248,
      "loss": 0.19554588198661804,
      "mean_token_accuracy": 0.9178217202425003,
      "num_tokens": 51093656.0,
      "step": 4142
    },
    {
      "entropy": 1.225668877363205,
      "epoch": 2.1816745655608214,
      "grad_norm": 0.29703161120414734,
      "learning_rate": 0.00016509680990462253,
      "loss": 0.19665035605430603,
      "mean_token_accuracy": 0.9156013280153275,
      "num_tokens": 51105992.0,
      "step": 4143
    },
    {
      "entropy": 1.2031238079071045,
      "epoch": 2.1822011585044763,
      "grad_norm": 0.2988886535167694,
      "learning_rate": 0.00016507946928323607,
      "loss": 0.21606066823005676,
      "mean_token_accuracy": 0.912405326962471,
      "num_tokens": 51118328.0,
      "step": 4144
    },
    {
      "entropy": 1.2246394753456116,
      "epoch": 2.1827277514481307,
      "grad_norm": 0.3065650165081024,
      "learning_rate": 0.00016506212539210252,
      "loss": 0.19041046500205994,
      "mean_token_accuracy": 0.9208615720272064,
      "num_tokens": 51130664.0,
      "step": 4145
    },
    {
      "entropy": 1.2080211341381073,
      "epoch": 2.183254344391785,
      "grad_norm": 0.2621246576309204,
      "learning_rate": 0.00016504477823225155,
      "loss": 0.18758952617645264,
      "mean_token_accuracy": 0.9203426092863083,
      "num_tokens": 51143000.0,
      "step": 4146
    },
    {
      "entropy": 1.2131967544555664,
      "epoch": 2.1837809373354395,
      "grad_norm": 0.279575377702713,
      "learning_rate": 0.00016502742780471292,
      "loss": 0.19576191902160645,
      "mean_token_accuracy": 0.9203895181417465,
      "num_tokens": 51155336.0,
      "step": 4147
    },
    {
      "entropy": 1.2280302047729492,
      "epoch": 2.1843075302790944,
      "grad_norm": 0.30413857102394104,
      "learning_rate": 0.0001650100741105166,
      "loss": 0.23050157725811005,
      "mean_token_accuracy": 0.9074624329805374,
      "num_tokens": 51167672.0,
      "step": 4148
    },
    {
      "entropy": 1.134459227323532,
      "epoch": 2.184834123222749,
      "grad_norm": 0.2748672366142273,
      "learning_rate": 0.00016499271715069282,
      "loss": 0.20069140195846558,
      "mean_token_accuracy": 0.9178691804409027,
      "num_tokens": 51180008.0,
      "step": 4149
    },
    {
      "entropy": 1.1993885934352875,
      "epoch": 2.1853607161664033,
      "grad_norm": 0.2918178141117096,
      "learning_rate": 0.000164975356926272,
      "loss": 0.1967269331216812,
      "mean_token_accuracy": 0.9205320626497269,
      "num_tokens": 51192344.0,
      "step": 4150
    },
    {
      "entropy": 1.169199138879776,
      "epoch": 2.185887309110058,
      "grad_norm": 0.2743321657180786,
      "learning_rate": 0.00016495799343828457,
      "loss": 0.1919926255941391,
      "mean_token_accuracy": 0.923371896147728,
      "num_tokens": 51204680.0,
      "step": 4151
    },
    {
      "entropy": 1.18020898103714,
      "epoch": 2.1864139020537126,
      "grad_norm": 0.2571604549884796,
      "learning_rate": 0.00016494062668776142,
      "loss": 0.18856392800807953,
      "mean_token_accuracy": 0.9222598224878311,
      "num_tokens": 51217016.0,
      "step": 4152
    },
    {
      "entropy": 1.238032579421997,
      "epoch": 2.186940494997367,
      "grad_norm": 0.31084808707237244,
      "learning_rate": 0.0001649232566757335,
      "loss": 0.2140955924987793,
      "mean_token_accuracy": 0.9109927117824554,
      "num_tokens": 51229352.0,
      "step": 4153
    },
    {
      "entropy": 1.165948748588562,
      "epoch": 2.1874670879410214,
      "grad_norm": 0.25719159841537476,
      "learning_rate": 0.00016490588340323193,
      "loss": 0.17952802777290344,
      "mean_token_accuracy": 0.925691083073616,
      "num_tokens": 51241688.0,
      "step": 4154
    },
    {
      "entropy": 1.1968891322612762,
      "epoch": 2.1879936808846763,
      "grad_norm": 0.31568825244903564,
      "learning_rate": 0.00016488850687128808,
      "loss": 0.200781911611557,
      "mean_token_accuracy": 0.9183004051446915,
      "num_tokens": 51254024.0,
      "step": 4155
    },
    {
      "entropy": 1.1447340846061707,
      "epoch": 2.1885202738283307,
      "grad_norm": 0.27546897530555725,
      "learning_rate": 0.0001648711270809335,
      "loss": 0.18839329481124878,
      "mean_token_accuracy": 0.9184723347425461,
      "num_tokens": 51266360.0,
      "step": 4156
    },
    {
      "entropy": 1.171207994222641,
      "epoch": 2.189046866771985,
      "grad_norm": 0.274943083524704,
      "learning_rate": 0.00016485374403319992,
      "loss": 0.18055374920368195,
      "mean_token_accuracy": 0.9275996685028076,
      "num_tokens": 51278696.0,
      "step": 4157
    },
    {
      "entropy": 1.1181433200836182,
      "epoch": 2.18957345971564,
      "grad_norm": 0.26391127705574036,
      "learning_rate": 0.00016483635772911927,
      "loss": 0.18997561931610107,
      "mean_token_accuracy": 0.9185864478349686,
      "num_tokens": 51291032.0,
      "step": 4158
    },
    {
      "entropy": 1.179248869419098,
      "epoch": 2.1901000526592944,
      "grad_norm": 0.3057918846607208,
      "learning_rate": 0.00016481896816972365,
      "loss": 0.20668058097362518,
      "mean_token_accuracy": 0.9130419045686722,
      "num_tokens": 51303368.0,
      "step": 4159
    },
    {
      "entropy": 1.14694344997406,
      "epoch": 2.190626645602949,
      "grad_norm": 0.29680994153022766,
      "learning_rate": 0.00016480157535604542,
      "loss": 0.20385394990444183,
      "mean_token_accuracy": 0.9219126552343369,
      "num_tokens": 51315704.0,
      "step": 4160
    },
    {
      "entropy": 1.1223139762878418,
      "epoch": 2.1911532385466037,
      "grad_norm": 0.28084614872932434,
      "learning_rate": 0.00016478417928911707,
      "loss": 0.1831260770559311,
      "mean_token_accuracy": 0.921630397439003,
      "num_tokens": 51328040.0,
      "step": 4161
    },
    {
      "entropy": 1.1934473514556885,
      "epoch": 2.191679831490258,
      "grad_norm": 0.2605876922607422,
      "learning_rate": 0.0001647667799699713,
      "loss": 0.17136146128177643,
      "mean_token_accuracy": 0.9267126470804214,
      "num_tokens": 51340376.0,
      "step": 4162
    },
    {
      "entropy": 1.2115715444087982,
      "epoch": 2.1922064244339126,
      "grad_norm": 0.2804056704044342,
      "learning_rate": 0.00016474937739964098,
      "loss": 0.19194582104682922,
      "mean_token_accuracy": 0.9201535731554031,
      "num_tokens": 51352712.0,
      "step": 4163
    },
    {
      "entropy": 1.2309587597846985,
      "epoch": 2.192733017377567,
      "grad_norm": 0.29135453701019287,
      "learning_rate": 0.00016473197157915924,
      "loss": 0.193227156996727,
      "mean_token_accuracy": 0.9226816892623901,
      "num_tokens": 51365048.0,
      "step": 4164
    },
    {
      "entropy": 1.199971079826355,
      "epoch": 2.193259610321222,
      "grad_norm": 0.2691158950328827,
      "learning_rate": 0.00016471456250955935,
      "loss": 0.18729914724826813,
      "mean_token_accuracy": 0.9220250248908997,
      "num_tokens": 51377384.0,
      "step": 4165
    },
    {
      "entropy": 1.2236689925193787,
      "epoch": 2.1937862032648763,
      "grad_norm": 0.28157827258110046,
      "learning_rate": 0.0001646971501918748,
      "loss": 0.1987837255001068,
      "mean_token_accuracy": 0.9200732856988907,
      "num_tokens": 51389720.0,
      "step": 4166
    },
    {
      "entropy": 1.2567576467990875,
      "epoch": 2.1943127962085307,
      "grad_norm": 0.2924027442932129,
      "learning_rate": 0.00016467973462713924,
      "loss": 0.19592998921871185,
      "mean_token_accuracy": 0.9199083000421524,
      "num_tokens": 51402056.0,
      "step": 4167
    },
    {
      "entropy": 1.237093836069107,
      "epoch": 2.194839389152185,
      "grad_norm": 0.24912655353546143,
      "learning_rate": 0.00016466231581638654,
      "loss": 0.18574857711791992,
      "mean_token_accuracy": 0.9237374365329742,
      "num_tokens": 51414392.0,
      "step": 4168
    },
    {
      "entropy": 1.303921639919281,
      "epoch": 2.19536598209584,
      "grad_norm": 0.2773839235305786,
      "learning_rate": 0.00016464489376065073,
      "loss": 0.178799569606781,
      "mean_token_accuracy": 0.9240204840898514,
      "num_tokens": 51426728.0,
      "step": 4169
    },
    {
      "entropy": 1.3061772286891937,
      "epoch": 2.1958925750394944,
      "grad_norm": 0.2626015841960907,
      "learning_rate": 0.00016462746846096607,
      "loss": 0.18042266368865967,
      "mean_token_accuracy": 0.9253515899181366,
      "num_tokens": 51439064.0,
      "step": 4170
    },
    {
      "entropy": 1.2534684240818024,
      "epoch": 2.196419167983149,
      "grad_norm": 0.26793691515922546,
      "learning_rate": 0.00016461003991836702,
      "loss": 0.1747230887413025,
      "mean_token_accuracy": 0.9280012398958206,
      "num_tokens": 51451400.0,
      "step": 4171
    },
    {
      "entropy": 1.2703869342803955,
      "epoch": 2.1969457609268037,
      "grad_norm": 0.2812594473361969,
      "learning_rate": 0.0001645926081338882,
      "loss": 0.1805751770734787,
      "mean_token_accuracy": 0.9254391044378281,
      "num_tokens": 51463736.0,
      "step": 4172
    },
    {
      "entropy": 1.2749820053577423,
      "epoch": 2.197472353870458,
      "grad_norm": 0.28039759397506714,
      "learning_rate": 0.00016457517310856435,
      "loss": 0.1838250607252121,
      "mean_token_accuracy": 0.9279988557100296,
      "num_tokens": 51476072.0,
      "step": 4173
    },
    {
      "entropy": 1.2618244588375092,
      "epoch": 2.1979989468141126,
      "grad_norm": 0.27489322423934937,
      "learning_rate": 0.00016455773484343062,
      "loss": 0.19052180647850037,
      "mean_token_accuracy": 0.9222322404384613,
      "num_tokens": 51488408.0,
      "step": 4174
    },
    {
      "entropy": 1.2714160978794098,
      "epoch": 2.1985255397577674,
      "grad_norm": 0.2729232907295227,
      "learning_rate": 0.00016454029333952218,
      "loss": 0.18227756023406982,
      "mean_token_accuracy": 0.926309272646904,
      "num_tokens": 51500744.0,
      "step": 4175
    },
    {
      "entropy": 1.3097633719444275,
      "epoch": 2.199052132701422,
      "grad_norm": 0.2813631594181061,
      "learning_rate": 0.00016452284859787438,
      "loss": 0.19379457831382751,
      "mean_token_accuracy": 0.9247638136148453,
      "num_tokens": 51513080.0,
      "step": 4176
    },
    {
      "entropy": 1.2534567713737488,
      "epoch": 2.1995787256450763,
      "grad_norm": 0.29952695965766907,
      "learning_rate": 0.00016450540061952284,
      "loss": 0.2159310132265091,
      "mean_token_accuracy": 0.9128957092761993,
      "num_tokens": 51525416.0,
      "step": 4177
    },
    {
      "entropy": 1.2288260757923126,
      "epoch": 2.2001053185887307,
      "grad_norm": 0.28645092248916626,
      "learning_rate": 0.00016448794940550335,
      "loss": 0.18216633796691895,
      "mean_token_accuracy": 0.9225532412528992,
      "num_tokens": 51537752.0,
      "step": 4178
    },
    {
      "entropy": 1.2748102247714996,
      "epoch": 2.2006319115323856,
      "grad_norm": 0.27064216136932373,
      "learning_rate": 0.00016447049495685187,
      "loss": 0.17641235888004303,
      "mean_token_accuracy": 0.9300948679447174,
      "num_tokens": 51550088.0,
      "step": 4179
    },
    {
      "entropy": 1.2677447497844696,
      "epoch": 2.20115850447604,
      "grad_norm": 0.2583957314491272,
      "learning_rate": 0.0001644530372746046,
      "loss": 0.17626498639583588,
      "mean_token_accuracy": 0.9239745885133743,
      "num_tokens": 51562424.0,
      "step": 4180
    },
    {
      "entropy": 1.2885291874408722,
      "epoch": 2.2016850974196944,
      "grad_norm": 0.2887992262840271,
      "learning_rate": 0.00016443557635979788,
      "loss": 0.19101789593696594,
      "mean_token_accuracy": 0.9209521859884262,
      "num_tokens": 51574760.0,
      "step": 4181
    },
    {
      "entropy": 1.2113700211048126,
      "epoch": 2.2022116903633493,
      "grad_norm": 0.2987286150455475,
      "learning_rate": 0.00016441811221346828,
      "loss": 0.18627510964870453,
      "mean_token_accuracy": 0.9210187345743179,
      "num_tokens": 51587096.0,
      "step": 4182
    },
    {
      "entropy": 1.2981871664524078,
      "epoch": 2.2027382833070037,
      "grad_norm": 0.27984488010406494,
      "learning_rate": 0.0001644006448366525,
      "loss": 0.18086984753608704,
      "mean_token_accuracy": 0.9228075742721558,
      "num_tokens": 51599432.0,
      "step": 4183
    },
    {
      "entropy": 1.3021030724048615,
      "epoch": 2.203264876250658,
      "grad_norm": 0.29861992597579956,
      "learning_rate": 0.00016438317423038754,
      "loss": 0.19581389427185059,
      "mean_token_accuracy": 0.9156288206577301,
      "num_tokens": 51611768.0,
      "step": 4184
    },
    {
      "entropy": 1.2574114203453064,
      "epoch": 2.2037914691943126,
      "grad_norm": 0.28540849685668945,
      "learning_rate": 0.00016436570039571048,
      "loss": 0.1952468752861023,
      "mean_token_accuracy": 0.9155928790569305,
      "num_tokens": 51624104.0,
      "step": 4185
    },
    {
      "entropy": 1.3478785157203674,
      "epoch": 2.2043180621379674,
      "grad_norm": 0.46443215012550354,
      "learning_rate": 0.00016434822333365867,
      "loss": 0.17215551435947418,
      "mean_token_accuracy": 0.9300452023744583,
      "num_tokens": 51636440.0,
      "step": 4186
    },
    {
      "entropy": 1.2662345170974731,
      "epoch": 2.204844655081622,
      "grad_norm": 0.29072144627571106,
      "learning_rate": 0.00016433074304526957,
      "loss": 0.19199174642562866,
      "mean_token_accuracy": 0.9260234832763672,
      "num_tokens": 51648776.0,
      "step": 4187
    },
    {
      "entropy": 1.232847660779953,
      "epoch": 2.2053712480252763,
      "grad_norm": 0.25679078698158264,
      "learning_rate": 0.00016431325953158096,
      "loss": 0.17870289087295532,
      "mean_token_accuracy": 0.9293755143880844,
      "num_tokens": 51661112.0,
      "step": 4188
    },
    {
      "entropy": 1.2633108794689178,
      "epoch": 2.205897840968931,
      "grad_norm": 0.266966849565506,
      "learning_rate": 0.00016429577279363067,
      "loss": 0.1845407336950302,
      "mean_token_accuracy": 0.9210513830184937,
      "num_tokens": 51673448.0,
      "step": 4189
    },
    {
      "entropy": 1.2357351183891296,
      "epoch": 2.2064244339125856,
      "grad_norm": 0.27318739891052246,
      "learning_rate": 0.0001642782828324568,
      "loss": 0.18381115794181824,
      "mean_token_accuracy": 0.926314502954483,
      "num_tokens": 51685784.0,
      "step": 4190
    },
    {
      "entropy": 1.2260805070400238,
      "epoch": 2.20695102685624,
      "grad_norm": 0.3009386360645294,
      "learning_rate": 0.00016426078964909766,
      "loss": 0.19563618302345276,
      "mean_token_accuracy": 0.9195328205823898,
      "num_tokens": 51698120.0,
      "step": 4191
    },
    {
      "entropy": 1.2364404797554016,
      "epoch": 2.207477619799895,
      "grad_norm": 0.29123032093048096,
      "learning_rate": 0.00016424329324459167,
      "loss": 0.1951371729373932,
      "mean_token_accuracy": 0.9174755364656448,
      "num_tokens": 51710456.0,
      "step": 4192
    },
    {
      "entropy": 1.2290646433830261,
      "epoch": 2.2080042127435493,
      "grad_norm": 0.27491459250450134,
      "learning_rate": 0.00016422579361997754,
      "loss": 0.18299022316932678,
      "mean_token_accuracy": 0.9223412722349167,
      "num_tokens": 51722792.0,
      "step": 4193
    },
    {
      "entropy": 1.1536138653755188,
      "epoch": 2.2085308056872037,
      "grad_norm": 0.2737578749656677,
      "learning_rate": 0.00016420829077629407,
      "loss": 0.18880535662174225,
      "mean_token_accuracy": 0.9208870977163315,
      "num_tokens": 51735128.0,
      "step": 4194
    },
    {
      "entropy": 1.1930584907531738,
      "epoch": 2.209057398630858,
      "grad_norm": 0.2856224775314331,
      "learning_rate": 0.00016419078471458033,
      "loss": 0.1827823668718338,
      "mean_token_accuracy": 0.9252646416425705,
      "num_tokens": 51747464.0,
      "step": 4195
    },
    {
      "entropy": 1.1551392376422882,
      "epoch": 2.209583991574513,
      "grad_norm": 0.29800665378570557,
      "learning_rate": 0.00016417327543587557,
      "loss": 0.19618409872055054,
      "mean_token_accuracy": 0.9184785932302475,
      "num_tokens": 51759800.0,
      "step": 4196
    },
    {
      "entropy": 1.1179493069648743,
      "epoch": 2.2101105845181674,
      "grad_norm": 0.36465853452682495,
      "learning_rate": 0.00016415576294121918,
      "loss": 0.1905837059020996,
      "mean_token_accuracy": 0.9215804487466812,
      "num_tokens": 51772136.0,
      "step": 4197
    },
    {
      "entropy": 1.1246863901615143,
      "epoch": 2.210637177461822,
      "grad_norm": 0.2792600691318512,
      "learning_rate": 0.0001641382472316508,
      "loss": 0.2024848610162735,
      "mean_token_accuracy": 0.9161612391471863,
      "num_tokens": 51784472.0,
      "step": 4198
    },
    {
      "entropy": 1.1439398527145386,
      "epoch": 2.2111637704054767,
      "grad_norm": 0.2936895787715912,
      "learning_rate": 0.0001641207283082102,
      "loss": 0.18412424623966217,
      "mean_token_accuracy": 0.9253138452768326,
      "num_tokens": 51796808.0,
      "step": 4199
    },
    {
      "entropy": 1.1501588225364685,
      "epoch": 2.211690363349131,
      "grad_norm": 0.2793161869049072,
      "learning_rate": 0.00016410320617193743,
      "loss": 0.192224383354187,
      "mean_token_accuracy": 0.924894392490387,
      "num_tokens": 51809144.0,
      "step": 4200
    },
    {
      "entropy": 1.1433355212211609,
      "epoch": 2.2122169562927856,
      "grad_norm": 0.28388532996177673,
      "learning_rate": 0.0001640856808238726,
      "loss": 0.181789368391037,
      "mean_token_accuracy": 0.9237640649080276,
      "num_tokens": 51821480.0,
      "step": 4201
    },
    {
      "entropy": 1.141024112701416,
      "epoch": 2.21274354923644,
      "grad_norm": 0.275046706199646,
      "learning_rate": 0.00016406815226505618,
      "loss": 0.1791418045759201,
      "mean_token_accuracy": 0.9245376735925674,
      "num_tokens": 51833816.0,
      "step": 4202
    },
    {
      "entropy": 1.1009389460086823,
      "epoch": 2.213270142180095,
      "grad_norm": 0.2538962960243225,
      "learning_rate": 0.00016405062049652867,
      "loss": 0.1699582189321518,
      "mean_token_accuracy": 0.9293850362300873,
      "num_tokens": 51846152.0,
      "step": 4203
    },
    {
      "entropy": 1.1123566925525665,
      "epoch": 2.2137967351237493,
      "grad_norm": 0.2865713834762573,
      "learning_rate": 0.00016403308551933085,
      "loss": 0.19568300247192383,
      "mean_token_accuracy": 0.9213463515043259,
      "num_tokens": 51858488.0,
      "step": 4204
    },
    {
      "entropy": 1.1162798702716827,
      "epoch": 2.2143233280674037,
      "grad_norm": 0.2596893906593323,
      "learning_rate": 0.0001640155473345037,
      "loss": 0.18270273506641388,
      "mean_token_accuracy": 0.9247992038726807,
      "num_tokens": 51870824.0,
      "step": 4205
    },
    {
      "entropy": 1.1294128894805908,
      "epoch": 2.2148499210110586,
      "grad_norm": 0.2684587240219116,
      "learning_rate": 0.0001639980059430883,
      "loss": 0.17550542950630188,
      "mean_token_accuracy": 0.9285458475351334,
      "num_tokens": 51883160.0,
      "step": 4206
    },
    {
      "entropy": 1.1598798334598541,
      "epoch": 2.215376513954713,
      "grad_norm": 0.2630522847175598,
      "learning_rate": 0.00016398046134612603,
      "loss": 0.1818559169769287,
      "mean_token_accuracy": 0.9171760529279709,
      "num_tokens": 51895496.0,
      "step": 4207
    },
    {
      "entropy": 1.1745121777057648,
      "epoch": 2.2159031068983674,
      "grad_norm": 0.2528412640094757,
      "learning_rate": 0.00016396291354465835,
      "loss": 0.1682814657688141,
      "mean_token_accuracy": 0.9309436827898026,
      "num_tokens": 51907832.0,
      "step": 4208
    },
    {
      "entropy": 1.127808690071106,
      "epoch": 2.2164296998420223,
      "grad_norm": 0.27972131967544556,
      "learning_rate": 0.0001639453625397271,
      "loss": 0.1828276664018631,
      "mean_token_accuracy": 0.9239791631698608,
      "num_tokens": 51920168.0,
      "step": 4209
    },
    {
      "entropy": 1.2046663463115692,
      "epoch": 2.2169562927856767,
      "grad_norm": 0.28617292642593384,
      "learning_rate": 0.000163927808332374,
      "loss": 0.1929018497467041,
      "mean_token_accuracy": 0.9188937544822693,
      "num_tokens": 51932504.0,
      "step": 4210
    },
    {
      "entropy": 1.2402802407741547,
      "epoch": 2.217482885729331,
      "grad_norm": 0.30822160840034485,
      "learning_rate": 0.0001639102509236413,
      "loss": 0.22689157724380493,
      "mean_token_accuracy": 0.9049774557352066,
      "num_tokens": 51944840.0,
      "step": 4211
    },
    {
      "entropy": 1.2726169526576996,
      "epoch": 2.2180094786729856,
      "grad_norm": 0.31506627798080444,
      "learning_rate": 0.00016389269031457117,
      "loss": 0.18111148476600647,
      "mean_token_accuracy": 0.922523245215416,
      "num_tokens": 51957176.0,
      "step": 4212
    },
    {
      "entropy": 1.2062140107154846,
      "epoch": 2.2185360716166405,
      "grad_norm": 0.2667774558067322,
      "learning_rate": 0.00016387512650620616,
      "loss": 0.18210844695568085,
      "mean_token_accuracy": 0.9245853424072266,
      "num_tokens": 51969512.0,
      "step": 4213
    },
    {
      "entropy": 1.289791852235794,
      "epoch": 2.219062664560295,
      "grad_norm": 0.3121054470539093,
      "learning_rate": 0.00016385755949958888,
      "loss": 0.22761571407318115,
      "mean_token_accuracy": 0.9057991802692413,
      "num_tokens": 51981848.0,
      "step": 4214
    },
    {
      "entropy": 1.2653508186340332,
      "epoch": 2.2195892575039493,
      "grad_norm": 0.27332237362861633,
      "learning_rate": 0.0001638399892957622,
      "loss": 0.19882501661777496,
      "mean_token_accuracy": 0.9142513871192932,
      "num_tokens": 51994184.0,
      "step": 4215
    },
    {
      "entropy": 1.275252342224121,
      "epoch": 2.220115850447604,
      "grad_norm": 0.2822205126285553,
      "learning_rate": 0.00016382241589576918,
      "loss": 0.19739031791687012,
      "mean_token_accuracy": 0.9224513620138168,
      "num_tokens": 52006520.0,
      "step": 4216
    },
    {
      "entropy": 1.3314436376094818,
      "epoch": 2.2206424433912586,
      "grad_norm": 0.2844606041908264,
      "learning_rate": 0.000163804839300653,
      "loss": 0.18049627542495728,
      "mean_token_accuracy": 0.9332050979137421,
      "num_tokens": 52018856.0,
      "step": 4217
    },
    {
      "entropy": 1.311014175415039,
      "epoch": 2.221169036334913,
      "grad_norm": 0.28489887714385986,
      "learning_rate": 0.00016378725951145712,
      "loss": 0.17989040911197662,
      "mean_token_accuracy": 0.9219749569892883,
      "num_tokens": 52031192.0,
      "step": 4218
    },
    {
      "entropy": 1.2845270931720734,
      "epoch": 2.2216956292785675,
      "grad_norm": 0.27442827820777893,
      "learning_rate": 0.00016376967652922513,
      "loss": 0.19630296528339386,
      "mean_token_accuracy": 0.9183860868215561,
      "num_tokens": 52043528.0,
      "step": 4219
    },
    {
      "entropy": 1.3556245565414429,
      "epoch": 2.2222222222222223,
      "grad_norm": 0.3226070702075958,
      "learning_rate": 0.00016375209035500088,
      "loss": 0.21448081731796265,
      "mean_token_accuracy": 0.912571057677269,
      "num_tokens": 52055864.0,
      "step": 4220
    },
    {
      "entropy": 1.3001513183116913,
      "epoch": 2.2227488151658767,
      "grad_norm": 0.27793556451797485,
      "learning_rate": 0.00016373450098982825,
      "loss": 0.1880507916212082,
      "mean_token_accuracy": 0.9209636598825455,
      "num_tokens": 52068200.0,
      "step": 4221
    },
    {
      "entropy": 1.3293032348155975,
      "epoch": 2.223275408109531,
      "grad_norm": 0.2972993552684784,
      "learning_rate": 0.00016371690843475153,
      "loss": 0.21140480041503906,
      "mean_token_accuracy": 0.9138139635324478,
      "num_tokens": 52080536.0,
      "step": 4222
    },
    {
      "entropy": 1.3074055314064026,
      "epoch": 2.223802001053186,
      "grad_norm": 0.3054673671722412,
      "learning_rate": 0.00016369931269081503,
      "loss": 0.2108297497034073,
      "mean_token_accuracy": 0.9098422229290009,
      "num_tokens": 52092872.0,
      "step": 4223
    },
    {
      "entropy": 1.3391023874282837,
      "epoch": 2.2243285939968405,
      "grad_norm": 0.3014651834964752,
      "learning_rate": 0.00016368171375906332,
      "loss": 0.17931891977787018,
      "mean_token_accuracy": 0.92972631752491,
      "num_tokens": 52105208.0,
      "step": 4224
    },
    {
      "entropy": 1.33422189950943,
      "epoch": 2.224855186940495,
      "grad_norm": 0.2674652636051178,
      "learning_rate": 0.00016366411164054113,
      "loss": 0.19167745113372803,
      "mean_token_accuracy": 0.9219842851161957,
      "num_tokens": 52117544.0,
      "step": 4225
    },
    {
      "entropy": 1.3092358112335205,
      "epoch": 2.2253817798841498,
      "grad_norm": 0.27045342326164246,
      "learning_rate": 0.00016364650633629343,
      "loss": 0.1807311475276947,
      "mean_token_accuracy": 0.9232537895441055,
      "num_tokens": 52129880.0,
      "step": 4226
    },
    {
      "entropy": 1.2968615889549255,
      "epoch": 2.225908372827804,
      "grad_norm": 0.2784993052482605,
      "learning_rate": 0.0001636288978473653,
      "loss": 0.1864447295665741,
      "mean_token_accuracy": 0.9230673313140869,
      "num_tokens": 52142216.0,
      "step": 4227
    },
    {
      "entropy": 1.329050749540329,
      "epoch": 2.2264349657714586,
      "grad_norm": 0.27790436148643494,
      "learning_rate": 0.00016361128617480212,
      "loss": 0.18269293010234833,
      "mean_token_accuracy": 0.9271285682916641,
      "num_tokens": 52154552.0,
      "step": 4228
    },
    {
      "entropy": 1.239962786436081,
      "epoch": 2.226961558715113,
      "grad_norm": 0.29057615995407104,
      "learning_rate": 0.0001635936713196493,
      "loss": 0.2014170140028,
      "mean_token_accuracy": 0.9164643883705139,
      "num_tokens": 52166888.0,
      "step": 4229
    },
    {
      "entropy": 1.3514109253883362,
      "epoch": 2.227488151658768,
      "grad_norm": 0.27092015743255615,
      "learning_rate": 0.00016357605328295262,
      "loss": 0.1870952844619751,
      "mean_token_accuracy": 0.9230411946773529,
      "num_tokens": 52179224.0,
      "step": 4230
    },
    {
      "entropy": 1.3073534667491913,
      "epoch": 2.2280147446024223,
      "grad_norm": 0.2956768870353699,
      "learning_rate": 0.00016355843206575794,
      "loss": 0.19634544849395752,
      "mean_token_accuracy": 0.9185549914836884,
      "num_tokens": 52191560.0,
      "step": 4231
    },
    {
      "entropy": 1.251263678073883,
      "epoch": 2.2285413375460768,
      "grad_norm": 0.309213787317276,
      "learning_rate": 0.00016354080766911127,
      "loss": 0.20773160457611084,
      "mean_token_accuracy": 0.9145363122224808,
      "num_tokens": 52203896.0,
      "step": 4232
    },
    {
      "entropy": 1.2692146599292755,
      "epoch": 2.2290679304897316,
      "grad_norm": 0.2713463008403778,
      "learning_rate": 0.00016352318009405891,
      "loss": 0.16297750174999237,
      "mean_token_accuracy": 0.9313487559556961,
      "num_tokens": 52216232.0,
      "step": 4233
    },
    {
      "entropy": 1.2983099818229675,
      "epoch": 2.229594523433386,
      "grad_norm": 0.4321221709251404,
      "learning_rate": 0.0001635055493416473,
      "loss": 0.18140633404254913,
      "mean_token_accuracy": 0.9271613657474518,
      "num_tokens": 52228568.0,
      "step": 4234
    },
    {
      "entropy": 1.2863817512989044,
      "epoch": 2.2301211163770405,
      "grad_norm": 0.2778743505477905,
      "learning_rate": 0.00016348791541292312,
      "loss": 0.1831590235233307,
      "mean_token_accuracy": 0.9236334264278412,
      "num_tokens": 52240904.0,
      "step": 4235
    },
    {
      "entropy": 1.188327044248581,
      "epoch": 2.230647709320695,
      "grad_norm": 0.27315250039100647,
      "learning_rate": 0.0001634702783089331,
      "loss": 0.17779839038848877,
      "mean_token_accuracy": 0.9250446408987045,
      "num_tokens": 52253240.0,
      "step": 4236
    },
    {
      "entropy": 1.262020230293274,
      "epoch": 2.2311743022643498,
      "grad_norm": 0.3055287003517151,
      "learning_rate": 0.00016345263803072433,
      "loss": 0.20651106536388397,
      "mean_token_accuracy": 0.9169783294200897,
      "num_tokens": 52265576.0,
      "step": 4237
    },
    {
      "entropy": 1.3503897488117218,
      "epoch": 2.231700895208004,
      "grad_norm": 0.29245615005493164,
      "learning_rate": 0.000163434994579344,
      "loss": 0.19317273795604706,
      "mean_token_accuracy": 0.9234201461076736,
      "num_tokens": 52277912.0,
      "step": 4238
    },
    {
      "entropy": 1.2586136758327484,
      "epoch": 2.2322274881516586,
      "grad_norm": 0.2817426025867462,
      "learning_rate": 0.00016341734795583947,
      "loss": 0.2006417214870453,
      "mean_token_accuracy": 0.9167672246694565,
      "num_tokens": 52290248.0,
      "step": 4239
    },
    {
      "entropy": 1.2308941781520844,
      "epoch": 2.2327540810953135,
      "grad_norm": 0.3034634590148926,
      "learning_rate": 0.00016339969816125832,
      "loss": 0.19974562525749207,
      "mean_token_accuracy": 0.9195841699838638,
      "num_tokens": 52302584.0,
      "step": 4240
    },
    {
      "entropy": 1.305918425321579,
      "epoch": 2.233280674038968,
      "grad_norm": 0.2764703035354614,
      "learning_rate": 0.00016338204519664835,
      "loss": 0.19095191359519958,
      "mean_token_accuracy": 0.922985702753067,
      "num_tokens": 52314920.0,
      "step": 4241
    },
    {
      "entropy": 1.2824455499649048,
      "epoch": 2.2338072669826223,
      "grad_norm": 0.25803378224372864,
      "learning_rate": 0.00016336438906305748,
      "loss": 0.18053492903709412,
      "mean_token_accuracy": 0.9281091690063477,
      "num_tokens": 52327256.0,
      "step": 4242
    },
    {
      "entropy": 1.3713833093643188,
      "epoch": 2.234333859926277,
      "grad_norm": 0.26806026697158813,
      "learning_rate": 0.00016334672976153384,
      "loss": 0.1849786639213562,
      "mean_token_accuracy": 0.929167240858078,
      "num_tokens": 52339592.0,
      "step": 4243
    },
    {
      "entropy": 1.3161489963531494,
      "epoch": 2.2348604528699316,
      "grad_norm": 0.25630515813827515,
      "learning_rate": 0.0001633290672931258,
      "loss": 0.17108337581157684,
      "mean_token_accuracy": 0.9340009987354279,
      "num_tokens": 52351928.0,
      "step": 4244
    },
    {
      "entropy": 1.3286859691143036,
      "epoch": 2.235387045813586,
      "grad_norm": 0.29029756784439087,
      "learning_rate": 0.0001633114016588819,
      "loss": 0.20633910596370697,
      "mean_token_accuracy": 0.909833550453186,
      "num_tokens": 52364264.0,
      "step": 4245
    },
    {
      "entropy": 1.329633116722107,
      "epoch": 2.2359136387572405,
      "grad_norm": 0.2533878684043884,
      "learning_rate": 0.00016329373285985078,
      "loss": 0.19495350122451782,
      "mean_token_accuracy": 0.9151149392127991,
      "num_tokens": 52376600.0,
      "step": 4246
    },
    {
      "entropy": 1.3563131988048553,
      "epoch": 2.2364402317008953,
      "grad_norm": 0.26852771639823914,
      "learning_rate": 0.0001632760608970813,
      "loss": 0.18576224148273468,
      "mean_token_accuracy": 0.9227636158466339,
      "num_tokens": 52388936.0,
      "step": 4247
    },
    {
      "entropy": 1.358976662158966,
      "epoch": 2.2369668246445498,
      "grad_norm": 0.2841165065765381,
      "learning_rate": 0.00016325838577162266,
      "loss": 0.2020607888698578,
      "mean_token_accuracy": 0.917872816324234,
      "num_tokens": 52401272.0,
      "step": 4248
    },
    {
      "entropy": 1.4256496131420135,
      "epoch": 2.237493417588204,
      "grad_norm": 0.2953294515609741,
      "learning_rate": 0.00016324070748452408,
      "loss": 0.20347987115383148,
      "mean_token_accuracy": 0.9172584414482117,
      "num_tokens": 52413608.0,
      "step": 4249
    },
    {
      "entropy": 1.382269710302353,
      "epoch": 2.238020010531859,
      "grad_norm": 0.2767752707004547,
      "learning_rate": 0.000163223026036835,
      "loss": 0.19072210788726807,
      "mean_token_accuracy": 0.9170867800712585,
      "num_tokens": 52425944.0,
      "step": 4250
    },
    {
      "entropy": 1.3237777650356293,
      "epoch": 2.2385466034755135,
      "grad_norm": 0.26159167289733887,
      "learning_rate": 0.00016320534142960503,
      "loss": 0.1843932569026947,
      "mean_token_accuracy": 0.921222910284996,
      "num_tokens": 52438280.0,
      "step": 4251
    },
    {
      "entropy": 1.3900092542171478,
      "epoch": 2.239073196419168,
      "grad_norm": 0.29133832454681396,
      "learning_rate": 0.0001631876536638841,
      "loss": 0.18629711866378784,
      "mean_token_accuracy": 0.927318811416626,
      "num_tokens": 52450616.0,
      "step": 4252
    },
    {
      "entropy": 1.3216579258441925,
      "epoch": 2.2395997893628223,
      "grad_norm": 0.28946828842163086,
      "learning_rate": 0.00016316996274072217,
      "loss": 0.19785767793655396,
      "mean_token_accuracy": 0.9227815121412277,
      "num_tokens": 52462952.0,
      "step": 4253
    },
    {
      "entropy": 1.3586134910583496,
      "epoch": 2.240126382306477,
      "grad_norm": 0.27580100297927856,
      "learning_rate": 0.00016315226866116941,
      "loss": 0.195715993642807,
      "mean_token_accuracy": 0.9197490960359573,
      "num_tokens": 52475288.0,
      "step": 4254
    },
    {
      "entropy": 1.3740904033184052,
      "epoch": 2.2406529752501316,
      "grad_norm": 0.28149449825286865,
      "learning_rate": 0.00016313457142627628,
      "loss": 0.1904170960187912,
      "mean_token_accuracy": 0.9227342903614044,
      "num_tokens": 52487624.0,
      "step": 4255
    },
    {
      "entropy": 1.3365713357925415,
      "epoch": 2.241179568193786,
      "grad_norm": 0.2665024697780609,
      "learning_rate": 0.00016311687103709333,
      "loss": 0.1871342957019806,
      "mean_token_accuracy": 0.9221449494361877,
      "num_tokens": 52499960.0,
      "step": 4256
    },
    {
      "entropy": 1.3210458755493164,
      "epoch": 2.241706161137441,
      "grad_norm": 0.29255783557891846,
      "learning_rate": 0.0001630991674946713,
      "loss": 0.20028424263000488,
      "mean_token_accuracy": 0.9167197197675705,
      "num_tokens": 52512296.0,
      "step": 4257
    },
    {
      "entropy": 1.3026900589466095,
      "epoch": 2.2422327540810953,
      "grad_norm": 0.2644364833831787,
      "learning_rate": 0.00016308146080006123,
      "loss": 0.18610155582427979,
      "mean_token_accuracy": 0.9260168969631195,
      "num_tokens": 52524632.0,
      "step": 4258
    },
    {
      "entropy": 1.334023267030716,
      "epoch": 2.2427593470247498,
      "grad_norm": 0.29489240050315857,
      "learning_rate": 0.00016306375095431422,
      "loss": 0.19872379302978516,
      "mean_token_accuracy": 0.9185984432697296,
      "num_tokens": 52536968.0,
      "step": 4259
    },
    {
      "entropy": 1.3495826721191406,
      "epoch": 2.2432859399684046,
      "grad_norm": 0.2851937413215637,
      "learning_rate": 0.00016304603795848157,
      "loss": 0.19495224952697754,
      "mean_token_accuracy": 0.9217026382684708,
      "num_tokens": 52549304.0,
      "step": 4260
    },
    {
      "entropy": 1.305820494890213,
      "epoch": 2.243812532912059,
      "grad_norm": 0.28069403767585754,
      "learning_rate": 0.0001630283218136148,
      "loss": 0.19976091384887695,
      "mean_token_accuracy": 0.9190419614315033,
      "num_tokens": 52561640.0,
      "step": 4261
    },
    {
      "entropy": 1.33967986702919,
      "epoch": 2.2443391258557135,
      "grad_norm": 0.3040577471256256,
      "learning_rate": 0.00016301060252076567,
      "loss": 0.207487553358078,
      "mean_token_accuracy": 0.916191965341568,
      "num_tokens": 52573976.0,
      "step": 4262
    },
    {
      "entropy": 1.29051473736763,
      "epoch": 2.244865718799368,
      "grad_norm": 0.27407845854759216,
      "learning_rate": 0.00016299288008098605,
      "loss": 0.19776585698127747,
      "mean_token_accuracy": 0.9183499664068222,
      "num_tokens": 52586312.0,
      "step": 4263
    },
    {
      "entropy": 1.3070797622203827,
      "epoch": 2.245392311743023,
      "grad_norm": 0.27223050594329834,
      "learning_rate": 0.00016297515449532795,
      "loss": 0.18903937935829163,
      "mean_token_accuracy": 0.9230516403913498,
      "num_tokens": 52598648.0,
      "step": 4264
    },
    {
      "entropy": 1.2732833623886108,
      "epoch": 2.245918904686677,
      "grad_norm": 0.25695350766181946,
      "learning_rate": 0.00016295742576484377,
      "loss": 0.16693276166915894,
      "mean_token_accuracy": 0.9326575249433517,
      "num_tokens": 52610984.0,
      "step": 4265
    },
    {
      "entropy": 1.3194738030433655,
      "epoch": 2.2464454976303316,
      "grad_norm": 0.25403544306755066,
      "learning_rate": 0.0001629396938905858,
      "loss": 0.16951501369476318,
      "mean_token_accuracy": 0.9296520501375198,
      "num_tokens": 52623320.0,
      "step": 4266
    },
    {
      "entropy": 1.315958708524704,
      "epoch": 2.2469720905739865,
      "grad_norm": 0.3142984211444855,
      "learning_rate": 0.00016292195887360683,
      "loss": 0.21278263628482819,
      "mean_token_accuracy": 0.9144498258829117,
      "num_tokens": 52635656.0,
      "step": 4267
    },
    {
      "entropy": 1.3089948296546936,
      "epoch": 2.247498683517641,
      "grad_norm": 0.31691795587539673,
      "learning_rate": 0.00016290422071495958,
      "loss": 0.2112875133752823,
      "mean_token_accuracy": 0.9154753535985947,
      "num_tokens": 52647992.0,
      "step": 4268
    },
    {
      "entropy": 1.3680811524391174,
      "epoch": 2.2480252764612954,
      "grad_norm": 0.31187358498573303,
      "learning_rate": 0.0001628864794156971,
      "loss": 0.21332678198814392,
      "mean_token_accuracy": 0.9160029590129852,
      "num_tokens": 52660328.0,
      "step": 4269
    },
    {
      "entropy": 1.262885570526123,
      "epoch": 2.2485518694049498,
      "grad_norm": 0.2790033519268036,
      "learning_rate": 0.0001628687349768726,
      "loss": 0.18170538544654846,
      "mean_token_accuracy": 0.931568443775177,
      "num_tokens": 52672664.0,
      "step": 4270
    },
    {
      "entropy": 1.265404999256134,
      "epoch": 2.2490784623486046,
      "grad_norm": 0.25181683897972107,
      "learning_rate": 0.00016285098739953938,
      "loss": 0.17339858412742615,
      "mean_token_accuracy": 0.9286882877349854,
      "num_tokens": 52685000.0,
      "step": 4271
    },
    {
      "entropy": 1.3293351829051971,
      "epoch": 2.249605055292259,
      "grad_norm": 0.2925328314304352,
      "learning_rate": 0.00016283323668475115,
      "loss": 0.19423630833625793,
      "mean_token_accuracy": 0.9217019081115723,
      "num_tokens": 52697336.0,
      "step": 4272
    },
    {
      "entropy": 1.3089188933372498,
      "epoch": 2.2501316482359135,
      "grad_norm": 0.31548264622688293,
      "learning_rate": 0.00016281548283356156,
      "loss": 0.18943582475185394,
      "mean_token_accuracy": 0.9233786463737488,
      "num_tokens": 52709672.0,
      "step": 4273
    },
    {
      "entropy": 1.2492122054100037,
      "epoch": 2.2506582411795684,
      "grad_norm": 0.27606451511383057,
      "learning_rate": 0.0001627977258470246,
      "loss": 0.18954810500144958,
      "mean_token_accuracy": 0.9246101230382919,
      "num_tokens": 52722008.0,
      "step": 4274
    },
    {
      "entropy": 1.2629761397838593,
      "epoch": 2.251184834123223,
      "grad_norm": 0.28441867232322693,
      "learning_rate": 0.00016277996572619437,
      "loss": 0.19405168294906616,
      "mean_token_accuracy": 0.923010841012001,
      "num_tokens": 52734344.0,
      "step": 4275
    },
    {
      "entropy": 1.2314798533916473,
      "epoch": 2.251711427066877,
      "grad_norm": 0.2783687710762024,
      "learning_rate": 0.00016276220247212522,
      "loss": 0.18629150092601776,
      "mean_token_accuracy": 0.923210546374321,
      "num_tokens": 52746680.0,
      "step": 4276
    },
    {
      "entropy": 1.2616170942783356,
      "epoch": 2.252238020010532,
      "grad_norm": 0.27606093883514404,
      "learning_rate": 0.0001627444360858716,
      "loss": 0.170989528298378,
      "mean_token_accuracy": 0.9309239387512207,
      "num_tokens": 52759016.0,
      "step": 4277
    },
    {
      "entropy": 1.197077065706253,
      "epoch": 2.2527646129541865,
      "grad_norm": 0.2800334692001343,
      "learning_rate": 0.00016272666656848826,
      "loss": 0.2011934369802475,
      "mean_token_accuracy": 0.9129274040460587,
      "num_tokens": 52771352.0,
      "step": 4278
    },
    {
      "entropy": 1.2784244120121002,
      "epoch": 2.253291205897841,
      "grad_norm": 0.308682382106781,
      "learning_rate": 0.00016270889392103001,
      "loss": 0.20107175409793854,
      "mean_token_accuracy": 0.919810488820076,
      "num_tokens": 52783688.0,
      "step": 4279
    },
    {
      "entropy": 1.2287653982639313,
      "epoch": 2.2538177988414954,
      "grad_norm": 0.28182491660118103,
      "learning_rate": 0.00016269111814455197,
      "loss": 0.18231971561908722,
      "mean_token_accuracy": 0.923517182469368,
      "num_tokens": 52796024.0,
      "step": 4280
    },
    {
      "entropy": 1.2548049986362457,
      "epoch": 2.2543443917851502,
      "grad_norm": 0.3218842148780823,
      "learning_rate": 0.00016267333924010934,
      "loss": 0.207222118973732,
      "mean_token_accuracy": 0.9176555424928665,
      "num_tokens": 52808360.0,
      "step": 4281
    },
    {
      "entropy": 1.2320858240127563,
      "epoch": 2.2548709847288047,
      "grad_norm": 0.3021528124809265,
      "learning_rate": 0.00016265555720875756,
      "loss": 0.19539406895637512,
      "mean_token_accuracy": 0.9212926179170609,
      "num_tokens": 52820696.0,
      "step": 4282
    },
    {
      "entropy": 1.2006807029247284,
      "epoch": 2.255397577672459,
      "grad_norm": 0.25874409079551697,
      "learning_rate": 0.00016263777205155223,
      "loss": 0.17395153641700745,
      "mean_token_accuracy": 0.9307205528020859,
      "num_tokens": 52833032.0,
      "step": 4283
    },
    {
      "entropy": 1.2096614241600037,
      "epoch": 2.2559241706161135,
      "grad_norm": 0.28012752532958984,
      "learning_rate": 0.00016261998376954922,
      "loss": 0.19147738814353943,
      "mean_token_accuracy": 0.9207902103662491,
      "num_tokens": 52845368.0,
      "step": 4284
    },
    {
      "entropy": 1.1712572574615479,
      "epoch": 2.2564507635597684,
      "grad_norm": 0.2657926082611084,
      "learning_rate": 0.0001626021923638044,
      "loss": 0.1850869357585907,
      "mean_token_accuracy": 0.9235047101974487,
      "num_tokens": 52857704.0,
      "step": 4285
    },
    {
      "entropy": 1.2101311385631561,
      "epoch": 2.256977356503423,
      "grad_norm": 0.2975207269191742,
      "learning_rate": 0.00016258439783537406,
      "loss": 0.20111596584320068,
      "mean_token_accuracy": 0.919817328453064,
      "num_tokens": 52870040.0,
      "step": 4286
    },
    {
      "entropy": 1.1905730962753296,
      "epoch": 2.257503949447077,
      "grad_norm": 0.2743525505065918,
      "learning_rate": 0.0001625666001853145,
      "loss": 0.1912555694580078,
      "mean_token_accuracy": 0.9214645624160767,
      "num_tokens": 52882376.0,
      "step": 4287
    },
    {
      "entropy": 1.1946870684623718,
      "epoch": 2.258030542390732,
      "grad_norm": 0.29676419496536255,
      "learning_rate": 0.00016254879941468223,
      "loss": 0.2016541063785553,
      "mean_token_accuracy": 0.9167199581861496,
      "num_tokens": 52894712.0,
      "step": 4288
    },
    {
      "entropy": 1.2338870167732239,
      "epoch": 2.2585571353343865,
      "grad_norm": 0.28669846057891846,
      "learning_rate": 0.00016253099552453407,
      "loss": 0.18671299517154694,
      "mean_token_accuracy": 0.9237444400787354,
      "num_tokens": 52907048.0,
      "step": 4289
    },
    {
      "entropy": 1.2231320142745972,
      "epoch": 2.259083728278041,
      "grad_norm": 0.29122886061668396,
      "learning_rate": 0.00016251318851592686,
      "loss": 0.19480589032173157,
      "mean_token_accuracy": 0.9223777800798416,
      "num_tokens": 52919384.0,
      "step": 4290
    },
    {
      "entropy": 1.2313807010650635,
      "epoch": 2.259610321221696,
      "grad_norm": 0.2886621057987213,
      "learning_rate": 0.0001624953783899177,
      "loss": 0.20268778502941132,
      "mean_token_accuracy": 0.9181878566741943,
      "num_tokens": 52931720.0,
      "step": 4291
    },
    {
      "entropy": 1.2376523911952972,
      "epoch": 2.2601369141653502,
      "grad_norm": 0.26574602723121643,
      "learning_rate": 0.0001624775651475639,
      "loss": 0.1816636472940445,
      "mean_token_accuracy": 0.9279394000768661,
      "num_tokens": 52944056.0,
      "step": 4292
    },
    {
      "entropy": 1.2119337618350983,
      "epoch": 2.2606635071090047,
      "grad_norm": 0.27368560433387756,
      "learning_rate": 0.0001624597487899229,
      "loss": 0.18821629881858826,
      "mean_token_accuracy": 0.9214064329862595,
      "num_tokens": 52956392.0,
      "step": 4293
    },
    {
      "entropy": 1.2653162479400635,
      "epoch": 2.2611901000526595,
      "grad_norm": 0.2919362187385559,
      "learning_rate": 0.0001624419293180524,
      "loss": 0.20762929320335388,
      "mean_token_accuracy": 0.912530779838562,
      "num_tokens": 52968728.0,
      "step": 4294
    },
    {
      "entropy": 1.273656576871872,
      "epoch": 2.261716692996314,
      "grad_norm": 0.2898019552230835,
      "learning_rate": 0.00016242410673301022,
      "loss": 0.1861550211906433,
      "mean_token_accuracy": 0.9242700487375259,
      "num_tokens": 52981064.0,
      "step": 4295
    },
    {
      "entropy": 1.2827538847923279,
      "epoch": 2.2622432859399684,
      "grad_norm": 0.2659165859222412,
      "learning_rate": 0.00016240628103585437,
      "loss": 0.19199910759925842,
      "mean_token_accuracy": 0.920208215713501,
      "num_tokens": 52993400.0,
      "step": 4296
    },
    {
      "entropy": 1.3939678966999054,
      "epoch": 2.262769878883623,
      "grad_norm": 0.2849959135055542,
      "learning_rate": 0.00016238845222764302,
      "loss": 0.17641255259513855,
      "mean_token_accuracy": 0.9284879267215729,
      "num_tokens": 53005736.0,
      "step": 4297
    },
    {
      "entropy": 1.3991989195346832,
      "epoch": 2.2632964718272777,
      "grad_norm": 0.2852117419242859,
      "learning_rate": 0.00016237062030943466,
      "loss": 0.19910377264022827,
      "mean_token_accuracy": 0.9214527308940887,
      "num_tokens": 53018072.0,
      "step": 4298
    },
    {
      "entropy": 1.3499420583248138,
      "epoch": 2.263823064770932,
      "grad_norm": 0.29614681005477905,
      "learning_rate": 0.00016235278528228778,
      "loss": 0.22492265701293945,
      "mean_token_accuracy": 0.9076694399118423,
      "num_tokens": 53030408.0,
      "step": 4299
    },
    {
      "entropy": 1.3215619325637817,
      "epoch": 2.2643496577145865,
      "grad_norm": 0.30598577857017517,
      "learning_rate": 0.00016233494714726118,
      "loss": 0.19633576273918152,
      "mean_token_accuracy": 0.9187863171100616,
      "num_tokens": 53042744.0,
      "step": 4300
    },
    {
      "entropy": 1.3638521432876587,
      "epoch": 2.264876250658241,
      "grad_norm": 0.2770838737487793,
      "learning_rate": 0.00016231710590541382,
      "loss": 0.18263986706733704,
      "mean_token_accuracy": 0.9262545853853226,
      "num_tokens": 53055080.0,
      "step": 4301
    },
    {
      "entropy": 1.319113403558731,
      "epoch": 2.265402843601896,
      "grad_norm": 0.28063175082206726,
      "learning_rate": 0.0001622992615578048,
      "loss": 0.200247123837471,
      "mean_token_accuracy": 0.917742520570755,
      "num_tokens": 53067416.0,
      "step": 4302
    },
    {
      "entropy": 1.2884471416473389,
      "epoch": 2.2659294365455502,
      "grad_norm": 0.2803884446620941,
      "learning_rate": 0.0001622814141054934,
      "loss": 0.19679561257362366,
      "mean_token_accuracy": 0.9207372665405273,
      "num_tokens": 53079752.0,
      "step": 4303
    },
    {
      "entropy": 1.3329062163829803,
      "epoch": 2.2664560294892047,
      "grad_norm": 0.2911641299724579,
      "learning_rate": 0.00016226356354953923,
      "loss": 0.19328062236309052,
      "mean_token_accuracy": 0.9168116897344589,
      "num_tokens": 53092088.0,
      "step": 4304
    },
    {
      "entropy": 1.3833136558532715,
      "epoch": 2.2669826224328595,
      "grad_norm": 0.33080819249153137,
      "learning_rate": 0.00016224570989100185,
      "loss": 0.19918012619018555,
      "mean_token_accuracy": 0.9184055626392365,
      "num_tokens": 53104424.0,
      "step": 4305
    },
    {
      "entropy": 1.3158639967441559,
      "epoch": 2.267509215376514,
      "grad_norm": 0.26879438757896423,
      "learning_rate": 0.0001622278531309412,
      "loss": 0.18872933089733124,
      "mean_token_accuracy": 0.9242192357778549,
      "num_tokens": 53116760.0,
      "step": 4306
    },
    {
      "entropy": 1.388542801141739,
      "epoch": 2.2680358083201684,
      "grad_norm": 0.3074987530708313,
      "learning_rate": 0.00016220999327041735,
      "loss": 0.2129671573638916,
      "mean_token_accuracy": 0.9109326601028442,
      "num_tokens": 53129096.0,
      "step": 4307
    },
    {
      "entropy": 1.3813231885433197,
      "epoch": 2.2685624012638232,
      "grad_norm": 0.2779155373573303,
      "learning_rate": 0.00016219213031049047,
      "loss": 0.18388351798057556,
      "mean_token_accuracy": 0.924717515707016,
      "num_tokens": 53141432.0,
      "step": 4308
    },
    {
      "entropy": 1.3881699442863464,
      "epoch": 2.2690889942074777,
      "grad_norm": 0.27398285269737244,
      "learning_rate": 0.00016217426425222103,
      "loss": 0.19520778954029083,
      "mean_token_accuracy": 0.925410807132721,
      "num_tokens": 53153768.0,
      "step": 4309
    },
    {
      "entropy": 1.3659266531467438,
      "epoch": 2.269615587151132,
      "grad_norm": 0.27490484714508057,
      "learning_rate": 0.00016215639509666957,
      "loss": 0.17789322137832642,
      "mean_token_accuracy": 0.9270453304052353,
      "num_tokens": 53166104.0,
      "step": 4310
    },
    {
      "entropy": 1.3776750266551971,
      "epoch": 2.270142180094787,
      "grad_norm": 0.3111220598220825,
      "learning_rate": 0.000162138522844897,
      "loss": 0.1980695277452469,
      "mean_token_accuracy": 0.9179234504699707,
      "num_tokens": 53178440.0,
      "step": 4311
    },
    {
      "entropy": 1.3968066573143005,
      "epoch": 2.2706687730384414,
      "grad_norm": 0.30460214614868164,
      "learning_rate": 0.00016212064749796418,
      "loss": 0.20272967219352722,
      "mean_token_accuracy": 0.9162765592336655,
      "num_tokens": 53190776.0,
      "step": 4312
    },
    {
      "entropy": 1.3836086690425873,
      "epoch": 2.271195365982096,
      "grad_norm": 0.2826038897037506,
      "learning_rate": 0.00016210276905693227,
      "loss": 0.19780540466308594,
      "mean_token_accuracy": 0.9176635593175888,
      "num_tokens": 53203112.0,
      "step": 4313
    },
    {
      "entropy": 1.39346045255661,
      "epoch": 2.2717219589257502,
      "grad_norm": 0.3027603030204773,
      "learning_rate": 0.00016208488752286269,
      "loss": 0.21689122915267944,
      "mean_token_accuracy": 0.90862837433815,
      "num_tokens": 53215448.0,
      "step": 4314
    },
    {
      "entropy": 1.3023641109466553,
      "epoch": 2.272248551869405,
      "grad_norm": 0.2757156491279602,
      "learning_rate": 0.00016206700289681688,
      "loss": 0.18307238817214966,
      "mean_token_accuracy": 0.9282919764518738,
      "num_tokens": 53227784.0,
      "step": 4315
    },
    {
      "entropy": 1.35664564371109,
      "epoch": 2.2727751448130595,
      "grad_norm": 0.28664660453796387,
      "learning_rate": 0.00016204911517985659,
      "loss": 0.19184020161628723,
      "mean_token_accuracy": 0.920403391122818,
      "num_tokens": 53240120.0,
      "step": 4316
    },
    {
      "entropy": 1.3529805839061737,
      "epoch": 2.273301737756714,
      "grad_norm": 0.28960809111595154,
      "learning_rate": 0.0001620312243730437,
      "loss": 0.18796728551387787,
      "mean_token_accuracy": 0.9186430126428604,
      "num_tokens": 53252456.0,
      "step": 4317
    },
    {
      "entropy": 1.2922759354114532,
      "epoch": 2.2738283307003684,
      "grad_norm": 0.2737516760826111,
      "learning_rate": 0.00016201333047744025,
      "loss": 0.17952796816825867,
      "mean_token_accuracy": 0.9263361543416977,
      "num_tokens": 53264792.0,
      "step": 4318
    },
    {
      "entropy": 1.2446196675300598,
      "epoch": 2.2743549236440233,
      "grad_norm": 0.2577621042728424,
      "learning_rate": 0.00016199543349410854,
      "loss": 0.18877744674682617,
      "mean_token_accuracy": 0.92601378262043,
      "num_tokens": 53277128.0,
      "step": 4319
    },
    {
      "entropy": 1.2977658212184906,
      "epoch": 2.2748815165876777,
      "grad_norm": 0.27885574102401733,
      "learning_rate": 0.000161977533424111,
      "loss": 0.1956649273633957,
      "mean_token_accuracy": 0.9176055639982224,
      "num_tokens": 53289464.0,
      "step": 4320
    },
    {
      "entropy": 1.312001883983612,
      "epoch": 2.275408109531332,
      "grad_norm": 0.2542676031589508,
      "learning_rate": 0.00016195963026851027,
      "loss": 0.16451983153820038,
      "mean_token_accuracy": 0.9343025386333466,
      "num_tokens": 53301800.0,
      "step": 4321
    },
    {
      "entropy": 1.3300804793834686,
      "epoch": 2.275934702474987,
      "grad_norm": 0.2855078876018524,
      "learning_rate": 0.0001619417240283691,
      "loss": 0.18589822947978973,
      "mean_token_accuracy": 0.9211694002151489,
      "num_tokens": 53314136.0,
      "step": 4322
    },
    {
      "entropy": 1.3246194124221802,
      "epoch": 2.2764612954186414,
      "grad_norm": 0.2893201410770416,
      "learning_rate": 0.00016192381470475054,
      "loss": 0.18851175904273987,
      "mean_token_accuracy": 0.9267619699239731,
      "num_tokens": 53326472.0,
      "step": 4323
    },
    {
      "entropy": 1.2888253331184387,
      "epoch": 2.276987888362296,
      "grad_norm": 0.2698841989040375,
      "learning_rate": 0.00016190590229871773,
      "loss": 0.19300393760204315,
      "mean_token_accuracy": 0.9214081466197968,
      "num_tokens": 53338808.0,
      "step": 4324
    },
    {
      "entropy": 1.285881131887436,
      "epoch": 2.2775144813059507,
      "grad_norm": 0.27817195653915405,
      "learning_rate": 0.00016188798681133403,
      "loss": 0.1776391863822937,
      "mean_token_accuracy": 0.9255973249673843,
      "num_tokens": 53351144.0,
      "step": 4325
    },
    {
      "entropy": 1.330023616552353,
      "epoch": 2.278041074249605,
      "grad_norm": 0.2973087728023529,
      "learning_rate": 0.00016187006824366297,
      "loss": 0.21736116707324982,
      "mean_token_accuracy": 0.9165065437555313,
      "num_tokens": 53363480.0,
      "step": 4326
    },
    {
      "entropy": 1.3090458810329437,
      "epoch": 2.2785676671932595,
      "grad_norm": 0.2864411473274231,
      "learning_rate": 0.00016185214659676832,
      "loss": 0.18242886662483215,
      "mean_token_accuracy": 0.9262679815292358,
      "num_tokens": 53375816.0,
      "step": 4327
    },
    {
      "entropy": 1.3205254673957825,
      "epoch": 2.2790942601369144,
      "grad_norm": 0.2983349859714508,
      "learning_rate": 0.0001618342218717139,
      "loss": 0.18693310022354126,
      "mean_token_accuracy": 0.9253579378128052,
      "num_tokens": 53388152.0,
      "step": 4328
    },
    {
      "entropy": 1.3446447551250458,
      "epoch": 2.279620853080569,
      "grad_norm": 0.28103673458099365,
      "learning_rate": 0.00016181629406956385,
      "loss": 0.19234877824783325,
      "mean_token_accuracy": 0.9203990995883942,
      "num_tokens": 53400488.0,
      "step": 4329
    },
    {
      "entropy": 1.3458137810230255,
      "epoch": 2.2801474460242233,
      "grad_norm": 0.30117067694664,
      "learning_rate": 0.00016179836319138243,
      "loss": 0.19072972238063812,
      "mean_token_accuracy": 0.9225273281335831,
      "num_tokens": 53412824.0,
      "step": 4330
    },
    {
      "entropy": 1.3791090846061707,
      "epoch": 2.2806740389678777,
      "grad_norm": 0.3069910705089569,
      "learning_rate": 0.00016178042923823405,
      "loss": 0.20394977927207947,
      "mean_token_accuracy": 0.9209769368171692,
      "num_tokens": 53425160.0,
      "step": 4331
    },
    {
      "entropy": 1.3324054777622223,
      "epoch": 2.2812006319115326,
      "grad_norm": 0.25391244888305664,
      "learning_rate": 0.00016176249221118346,
      "loss": 0.1717671900987625,
      "mean_token_accuracy": 0.9276586025953293,
      "num_tokens": 53437496.0,
      "step": 4332
    },
    {
      "entropy": 1.3155920505523682,
      "epoch": 2.281727224855187,
      "grad_norm": 0.2734282314777374,
      "learning_rate": 0.00016174455211129536,
      "loss": 0.18031081557273865,
      "mean_token_accuracy": 0.9291572421789169,
      "num_tokens": 53449832.0,
      "step": 4333
    },
    {
      "entropy": 1.349910020828247,
      "epoch": 2.2822538177988414,
      "grad_norm": 0.2482135146856308,
      "learning_rate": 0.0001617266089396348,
      "loss": 0.1759587526321411,
      "mean_token_accuracy": 0.9239980727434158,
      "num_tokens": 53462168.0,
      "step": 4334
    },
    {
      "entropy": 1.378309428691864,
      "epoch": 2.282780410742496,
      "grad_norm": 0.27560997009277344,
      "learning_rate": 0.00016170866269726695,
      "loss": 0.19269959628582,
      "mean_token_accuracy": 0.9190654456615448,
      "num_tokens": 53474504.0,
      "step": 4335
    },
    {
      "entropy": 1.3972489535808563,
      "epoch": 2.2833070036861507,
      "grad_norm": 0.27502357959747314,
      "learning_rate": 0.00016169071338525718,
      "loss": 0.19034190475940704,
      "mean_token_accuracy": 0.9181709289550781,
      "num_tokens": 53486840.0,
      "step": 4336
    },
    {
      "entropy": 1.3337211310863495,
      "epoch": 2.283833596629805,
      "grad_norm": 0.2534601092338562,
      "learning_rate": 0.00016167276100467103,
      "loss": 0.17808862030506134,
      "mean_token_accuracy": 0.9231080412864685,
      "num_tokens": 53499176.0,
      "step": 4337
    },
    {
      "entropy": 1.4127607941627502,
      "epoch": 2.2843601895734595,
      "grad_norm": 0.3043602406978607,
      "learning_rate": 0.00016165480555657425,
      "loss": 0.21522895991802216,
      "mean_token_accuracy": 0.9097288846969604,
      "num_tokens": 53511512.0,
      "step": 4338
    },
    {
      "entropy": 1.316486895084381,
      "epoch": 2.2848867825171144,
      "grad_norm": 0.2641633152961731,
      "learning_rate": 0.0001616368470420327,
      "loss": 0.18304666876792908,
      "mean_token_accuracy": 0.9187512695789337,
      "num_tokens": 53523848.0,
      "step": 4339
    },
    {
      "entropy": 1.3706423938274384,
      "epoch": 2.285413375460769,
      "grad_norm": 0.2619921565055847,
      "learning_rate": 0.00016161888546211252,
      "loss": 0.20312339067459106,
      "mean_token_accuracy": 0.913113921880722,
      "num_tokens": 53536184.0,
      "step": 4340
    },
    {
      "entropy": 1.2946105301380157,
      "epoch": 2.2859399684044233,
      "grad_norm": 0.2881142497062683,
      "learning_rate": 0.00016160092081787995,
      "loss": 0.1995781809091568,
      "mean_token_accuracy": 0.9192468374967575,
      "num_tokens": 53548520.0,
      "step": 4341
    },
    {
      "entropy": 1.41688871383667,
      "epoch": 2.286466561348078,
      "grad_norm": 0.2905746102333069,
      "learning_rate": 0.0001615829531104015,
      "loss": 0.21028542518615723,
      "mean_token_accuracy": 0.9094423651695251,
      "num_tokens": 53560856.0,
      "step": 4342
    },
    {
      "entropy": 1.2906908690929413,
      "epoch": 2.2869931542917326,
      "grad_norm": 0.24352654814720154,
      "learning_rate": 0.00016156498234074374,
      "loss": 0.1764492690563202,
      "mean_token_accuracy": 0.9228947311639786,
      "num_tokens": 53573192.0,
      "step": 4343
    },
    {
      "entropy": 1.33392733335495,
      "epoch": 2.287519747235387,
      "grad_norm": 0.26100727915763855,
      "learning_rate": 0.00016154700850997355,
      "loss": 0.19272339344024658,
      "mean_token_accuracy": 0.9170558750629425,
      "num_tokens": 53585528.0,
      "step": 4344
    },
    {
      "entropy": 1.2810238003730774,
      "epoch": 2.288046340179042,
      "grad_norm": 0.2741535007953644,
      "learning_rate": 0.00016152903161915789,
      "loss": 0.1874852180480957,
      "mean_token_accuracy": 0.9228169918060303,
      "num_tokens": 53597864.0,
      "step": 4345
    },
    {
      "entropy": 1.2848420143127441,
      "epoch": 2.2885729331226963,
      "grad_norm": 0.2634083926677704,
      "learning_rate": 0.00016151105166936394,
      "loss": 0.19013947248458862,
      "mean_token_accuracy": 0.9219299107789993,
      "num_tokens": 53610200.0,
      "step": 4346
    },
    {
      "entropy": 1.2308179140090942,
      "epoch": 2.2890995260663507,
      "grad_norm": 0.28782540559768677,
      "learning_rate": 0.0001614930686616591,
      "loss": 0.19066664576530457,
      "mean_token_accuracy": 0.9219117313623428,
      "num_tokens": 53622536.0,
      "step": 4347
    },
    {
      "entropy": 1.2573314607143402,
      "epoch": 2.289626119010005,
      "grad_norm": 0.2767718434333801,
      "learning_rate": 0.00016147508259711088,
      "loss": 0.18270361423492432,
      "mean_token_accuracy": 0.9219647794961929,
      "num_tokens": 53634872.0,
      "step": 4348
    },
    {
      "entropy": 1.299607902765274,
      "epoch": 2.29015271195366,
      "grad_norm": 0.2846214175224304,
      "learning_rate": 0.00016145709347678707,
      "loss": 0.19855788350105286,
      "mean_token_accuracy": 0.9172389805316925,
      "num_tokens": 53647208.0,
      "step": 4349
    },
    {
      "entropy": 1.2599292993545532,
      "epoch": 2.2906793048973144,
      "grad_norm": 0.2899573743343353,
      "learning_rate": 0.0001614391013017555,
      "loss": 0.19316795468330383,
      "mean_token_accuracy": 0.9196108430624008,
      "num_tokens": 53659544.0,
      "step": 4350
    },
    {
      "entropy": 1.2123220264911652,
      "epoch": 2.291205897840969,
      "grad_norm": 0.272175669670105,
      "learning_rate": 0.0001614211060730843,
      "loss": 0.19963154196739197,
      "mean_token_accuracy": 0.9213588237762451,
      "num_tokens": 53671880.0,
      "step": 4351
    },
    {
      "entropy": 1.2599426805973053,
      "epoch": 2.2917324907846233,
      "grad_norm": 0.2871687114238739,
      "learning_rate": 0.0001614031077918417,
      "loss": 0.1960330456495285,
      "mean_token_accuracy": 0.9175105392932892,
      "num_tokens": 53684216.0,
      "step": 4352
    },
    {
      "entropy": 1.2873346507549286,
      "epoch": 2.292259083728278,
      "grad_norm": 0.2764008343219757,
      "learning_rate": 0.00016138510645909626,
      "loss": 0.1853741705417633,
      "mean_token_accuracy": 0.9236430674791336,
      "num_tokens": 53696552.0,
      "step": 4353
    },
    {
      "entropy": 1.2654350399971008,
      "epoch": 2.2927856766719326,
      "grad_norm": 0.3239395320415497,
      "learning_rate": 0.00016136710207591653,
      "loss": 0.18854394555091858,
      "mean_token_accuracy": 0.926001563668251,
      "num_tokens": 53708888.0,
      "step": 4354
    },
    {
      "entropy": 1.2028826177120209,
      "epoch": 2.293312269615587,
      "grad_norm": 0.2788753807544708,
      "learning_rate": 0.0001613490946433713,
      "loss": 0.19624781608581543,
      "mean_token_accuracy": 0.9118239730596542,
      "num_tokens": 53721224.0,
      "step": 4355
    },
    {
      "entropy": 1.2995253801345825,
      "epoch": 2.293838862559242,
      "grad_norm": 0.2672927975654602,
      "learning_rate": 0.0001613310841625296,
      "loss": 0.1733776032924652,
      "mean_token_accuracy": 0.9293322563171387,
      "num_tokens": 53733560.0,
      "step": 4356
    },
    {
      "entropy": 1.2344202101230621,
      "epoch": 2.2943654555028963,
      "grad_norm": 0.2676299512386322,
      "learning_rate": 0.0001613130706344607,
      "loss": 0.18073591589927673,
      "mean_token_accuracy": 0.9267222881317139,
      "num_tokens": 53745896.0,
      "step": 4357
    },
    {
      "entropy": 1.213013082742691,
      "epoch": 2.2948920484465507,
      "grad_norm": 0.29037851095199585,
      "learning_rate": 0.00016129505406023382,
      "loss": 0.20425663888454437,
      "mean_token_accuracy": 0.911371037364006,
      "num_tokens": 53758232.0,
      "step": 4358
    },
    {
      "entropy": 1.283359169960022,
      "epoch": 2.2954186413902056,
      "grad_norm": 0.2981805205345154,
      "learning_rate": 0.00016127703444091852,
      "loss": 0.18576066195964813,
      "mean_token_accuracy": 0.9234738498926163,
      "num_tokens": 53770568.0,
      "step": 4359
    },
    {
      "entropy": 1.3199339807033539,
      "epoch": 2.29594523433386,
      "grad_norm": 0.33105799555778503,
      "learning_rate": 0.00016125901177758457,
      "loss": 0.21435634791851044,
      "mean_token_accuracy": 0.9107922315597534,
      "num_tokens": 53782904.0,
      "step": 4360
    },
    {
      "entropy": 1.299638718366623,
      "epoch": 2.2964718272775144,
      "grad_norm": 0.2832479774951935,
      "learning_rate": 0.00016124098607130184,
      "loss": 0.19559845328330994,
      "mean_token_accuracy": 0.9197534024715424,
      "num_tokens": 53795240.0,
      "step": 4361
    },
    {
      "entropy": 1.2708514034748077,
      "epoch": 2.296998420221169,
      "grad_norm": 0.2999652624130249,
      "learning_rate": 0.00016122295732314042,
      "loss": 0.20792636275291443,
      "mean_token_accuracy": 0.9146489202976227,
      "num_tokens": 53807576.0,
      "step": 4362
    },
    {
      "entropy": 1.2924802005290985,
      "epoch": 2.2975250131648237,
      "grad_norm": 0.27054089307785034,
      "learning_rate": 0.00016120492553417065,
      "loss": 0.1762731522321701,
      "mean_token_accuracy": 0.9278860241174698,
      "num_tokens": 53819912.0,
      "step": 4363
    },
    {
      "entropy": 1.3254149854183197,
      "epoch": 2.298051606108478,
      "grad_norm": 0.2686094045639038,
      "learning_rate": 0.0001611868907054628,
      "loss": 0.17815278470516205,
      "mean_token_accuracy": 0.9264986962080002,
      "num_tokens": 53832248.0,
      "step": 4364
    },
    {
      "entropy": 1.3951729238033295,
      "epoch": 2.2985781990521326,
      "grad_norm": 0.28512006998062134,
      "learning_rate": 0.00016116885283808764,
      "loss": 0.201333150267601,
      "mean_token_accuracy": 0.9201938062906265,
      "num_tokens": 53844584.0,
      "step": 4365
    },
    {
      "entropy": 1.312125563621521,
      "epoch": 2.2991047919957874,
      "grad_norm": 0.27145853638648987,
      "learning_rate": 0.00016115081193311592,
      "loss": 0.17296220362186432,
      "mean_token_accuracy": 0.9303802251815796,
      "num_tokens": 53856920.0,
      "step": 4366
    },
    {
      "entropy": 1.3520390391349792,
      "epoch": 2.299631384939442,
      "grad_norm": 0.2710401713848114,
      "learning_rate": 0.00016113276799161857,
      "loss": 0.1911015808582306,
      "mean_token_accuracy": 0.918425902724266,
      "num_tokens": 53869256.0,
      "step": 4367
    },
    {
      "entropy": 1.4310398697853088,
      "epoch": 2.3001579778830963,
      "grad_norm": 0.30252817273139954,
      "learning_rate": 0.00016111472101466688,
      "loss": 0.20369470119476318,
      "mean_token_accuracy": 0.9167941361665726,
      "num_tokens": 53881592.0,
      "step": 4368
    },
    {
      "entropy": 1.357957899570465,
      "epoch": 2.3006845708267507,
      "grad_norm": 0.26683077216148376,
      "learning_rate": 0.00016109667100333208,
      "loss": 0.1931741088628769,
      "mean_token_accuracy": 0.9199745804071426,
      "num_tokens": 53893928.0,
      "step": 4369
    },
    {
      "entropy": 1.4150426983833313,
      "epoch": 2.3012111637704056,
      "grad_norm": 0.2798272669315338,
      "learning_rate": 0.00016107861795868576,
      "loss": 0.1891252100467682,
      "mean_token_accuracy": 0.9210709631443024,
      "num_tokens": 53906264.0,
      "step": 4370
    },
    {
      "entropy": 1.3921004831790924,
      "epoch": 2.30173775671406,
      "grad_norm": 0.27035799622535706,
      "learning_rate": 0.00016106056188179958,
      "loss": 0.18214790523052216,
      "mean_token_accuracy": 0.9235829561948776,
      "num_tokens": 53918600.0,
      "step": 4371
    },
    {
      "entropy": 1.3772303760051727,
      "epoch": 2.3022643496577144,
      "grad_norm": 0.2842561900615692,
      "learning_rate": 0.00016104250277374548,
      "loss": 0.20173102617263794,
      "mean_token_accuracy": 0.9209082275629044,
      "num_tokens": 53930936.0,
      "step": 4372
    },
    {
      "entropy": 1.3774393200874329,
      "epoch": 2.3027909426013693,
      "grad_norm": 0.27162718772888184,
      "learning_rate": 0.00016102444063559546,
      "loss": 0.19384866952896118,
      "mean_token_accuracy": 0.9170814156532288,
      "num_tokens": 53943272.0,
      "step": 4373
    },
    {
      "entropy": 1.336693435907364,
      "epoch": 2.3033175355450237,
      "grad_norm": 0.25225090980529785,
      "learning_rate": 0.00016100637546842179,
      "loss": 0.1860221028327942,
      "mean_token_accuracy": 0.9214341193437576,
      "num_tokens": 53955608.0,
      "step": 4374
    },
    {
      "entropy": 1.330688863992691,
      "epoch": 2.303844128488678,
      "grad_norm": 0.2605641484260559,
      "learning_rate": 0.00016098830727329693,
      "loss": 0.17431269586086273,
      "mean_token_accuracy": 0.9278644174337387,
      "num_tokens": 53967944.0,
      "step": 4375
    },
    {
      "entropy": 1.344216674566269,
      "epoch": 2.304370721432333,
      "grad_norm": 0.2637542486190796,
      "learning_rate": 0.0001609702360512934,
      "loss": 0.17134076356887817,
      "mean_token_accuracy": 0.929853156208992,
      "num_tokens": 53980280.0,
      "step": 4376
    },
    {
      "entropy": 1.3215482532978058,
      "epoch": 2.3048973143759874,
      "grad_norm": 0.28528934717178345,
      "learning_rate": 0.00016095216180348406,
      "loss": 0.19538789987564087,
      "mean_token_accuracy": 0.9190406650304794,
      "num_tokens": 53992616.0,
      "step": 4377
    },
    {
      "entropy": 1.3346219658851624,
      "epoch": 2.305423907319642,
      "grad_norm": 0.2910776734352112,
      "learning_rate": 0.00016093408453094182,
      "loss": 0.19502344727516174,
      "mean_token_accuracy": 0.9186424911022186,
      "num_tokens": 54004952.0,
      "step": 4378
    },
    {
      "entropy": 1.303836166858673,
      "epoch": 2.3059505002632963,
      "grad_norm": 0.27647870779037476,
      "learning_rate": 0.00016091600423473988,
      "loss": 0.18278121948242188,
      "mean_token_accuracy": 0.9247827529907227,
      "num_tokens": 54017288.0,
      "step": 4379
    },
    {
      "entropy": 1.3144015967845917,
      "epoch": 2.306477093206951,
      "grad_norm": 0.27541449666023254,
      "learning_rate": 0.00016089792091595153,
      "loss": 0.17720472812652588,
      "mean_token_accuracy": 0.9278306365013123,
      "num_tokens": 54029624.0,
      "step": 4380
    },
    {
      "entropy": 1.3188709318637848,
      "epoch": 2.3070036861506056,
      "grad_norm": 0.27120038866996765,
      "learning_rate": 0.00016087983457565026,
      "loss": 0.18134379386901855,
      "mean_token_accuracy": 0.9238206744194031,
      "num_tokens": 54041960.0,
      "step": 4381
    },
    {
      "entropy": 1.2987216711044312,
      "epoch": 2.30753027909426,
      "grad_norm": 0.2692391276359558,
      "learning_rate": 0.00016086174521490976,
      "loss": 0.18199361860752106,
      "mean_token_accuracy": 0.9220229238271713,
      "num_tokens": 54054296.0,
      "step": 4382
    },
    {
      "entropy": 1.3269225060939789,
      "epoch": 2.308056872037915,
      "grad_norm": 0.30096539855003357,
      "learning_rate": 0.00016084365283480388,
      "loss": 0.1970018595457077,
      "mean_token_accuracy": 0.9222443252801895,
      "num_tokens": 54066632.0,
      "step": 4383
    },
    {
      "entropy": 1.2876869142055511,
      "epoch": 2.3085834649815693,
      "grad_norm": 0.278378963470459,
      "learning_rate": 0.00016082555743640668,
      "loss": 0.19296762347221375,
      "mean_token_accuracy": 0.9184974581003189,
      "num_tokens": 54078968.0,
      "step": 4384
    },
    {
      "entropy": 1.3124735057353973,
      "epoch": 2.3091100579252237,
      "grad_norm": 0.2990483045578003,
      "learning_rate": 0.00016080745902079238,
      "loss": 0.1947772204875946,
      "mean_token_accuracy": 0.9188289642333984,
      "num_tokens": 54091304.0,
      "step": 4385
    },
    {
      "entropy": 1.2949390411376953,
      "epoch": 2.309636650868878,
      "grad_norm": 0.276991069316864,
      "learning_rate": 0.0001607893575890354,
      "loss": 0.17927928268909454,
      "mean_token_accuracy": 0.9291679114103317,
      "num_tokens": 54103640.0,
      "step": 4386
    },
    {
      "entropy": 1.2690950632095337,
      "epoch": 2.310163243812533,
      "grad_norm": 0.29863035678863525,
      "learning_rate": 0.0001607712531422102,
      "loss": 0.18910539150238037,
      "mean_token_accuracy": 0.9157494306564331,
      "num_tokens": 54115976.0,
      "step": 4387
    },
    {
      "entropy": 1.3071636855602264,
      "epoch": 2.3106898367561874,
      "grad_norm": 0.2578350603580475,
      "learning_rate": 0.00016075314568139167,
      "loss": 0.1770387887954712,
      "mean_token_accuracy": 0.9272283613681793,
      "num_tokens": 54128312.0,
      "step": 4388
    },
    {
      "entropy": 1.2916749119758606,
      "epoch": 2.311216429699842,
      "grad_norm": 0.2677556872367859,
      "learning_rate": 0.00016073503520765467,
      "loss": 0.19181080162525177,
      "mean_token_accuracy": 0.9190926253795624,
      "num_tokens": 54140648.0,
      "step": 4389
    },
    {
      "entropy": 1.309247374534607,
      "epoch": 2.3117430226434967,
      "grad_norm": 0.2915293872356415,
      "learning_rate": 0.00016071692172207435,
      "loss": 0.1867821216583252,
      "mean_token_accuracy": 0.9195849001407623,
      "num_tokens": 54152984.0,
      "step": 4390
    },
    {
      "entropy": 1.29865163564682,
      "epoch": 2.312269615587151,
      "grad_norm": 0.2806514501571655,
      "learning_rate": 0.00016069880522572597,
      "loss": 0.1817903220653534,
      "mean_token_accuracy": 0.9249827116727829,
      "num_tokens": 54165320.0,
      "step": 4391
    },
    {
      "entropy": 1.2931040525436401,
      "epoch": 2.3127962085308056,
      "grad_norm": 0.29710087180137634,
      "learning_rate": 0.00016068068571968507,
      "loss": 0.20901834964752197,
      "mean_token_accuracy": 0.9162816405296326,
      "num_tokens": 54177656.0,
      "step": 4392
    },
    {
      "entropy": 1.2990557551383972,
      "epoch": 2.3133228014744605,
      "grad_norm": 0.2834272086620331,
      "learning_rate": 0.00016066256320502719,
      "loss": 0.1743437647819519,
      "mean_token_accuracy": 0.9270213842391968,
      "num_tokens": 54189992.0,
      "step": 4393
    },
    {
      "entropy": 1.2775674760341644,
      "epoch": 2.313849394418115,
      "grad_norm": 0.2701371908187866,
      "learning_rate": 0.00016064443768282823,
      "loss": 0.17299538850784302,
      "mean_token_accuracy": 0.9240054935216904,
      "num_tokens": 54202328.0,
      "step": 4394
    },
    {
      "entropy": 1.312081515789032,
      "epoch": 2.3143759873617693,
      "grad_norm": 0.3133409321308136,
      "learning_rate": 0.0001606263091541642,
      "loss": 0.193025603890419,
      "mean_token_accuracy": 0.9182188510894775,
      "num_tokens": 54214664.0,
      "step": 4395
    },
    {
      "entropy": 1.3071553409099579,
      "epoch": 2.3149025803054237,
      "grad_norm": 0.2669501304626465,
      "learning_rate": 0.00016060817762011126,
      "loss": 0.17285308241844177,
      "mean_token_accuracy": 0.9284144788980484,
      "num_tokens": 54227000.0,
      "step": 4396
    },
    {
      "entropy": 1.3454262018203735,
      "epoch": 2.3154291732490786,
      "grad_norm": 0.29012081027030945,
      "learning_rate": 0.00016059004308174578,
      "loss": 0.20970484614372253,
      "mean_token_accuracy": 0.9146772921085358,
      "num_tokens": 54239336.0,
      "step": 4397
    },
    {
      "entropy": 1.3304231464862823,
      "epoch": 2.315955766192733,
      "grad_norm": 0.34615376591682434,
      "learning_rate": 0.0001605719055401443,
      "loss": 0.20989423990249634,
      "mean_token_accuracy": 0.912458136677742,
      "num_tokens": 54251672.0,
      "step": 4398
    },
    {
      "entropy": 1.3078122437000275,
      "epoch": 2.3164823591363874,
      "grad_norm": 0.3172536790370941,
      "learning_rate": 0.00016055376499638356,
      "loss": 0.21419790387153625,
      "mean_token_accuracy": 0.9104336500167847,
      "num_tokens": 54264008.0,
      "step": 4399
    },
    {
      "entropy": 1.3196004033088684,
      "epoch": 2.3170089520800423,
      "grad_norm": 0.26406171917915344,
      "learning_rate": 0.00016053562145154044,
      "loss": 0.1729414165019989,
      "mean_token_accuracy": 0.9300494939088821,
      "num_tokens": 54276344.0,
      "step": 4400
    },
    {
      "entropy": 1.3104143142700195,
      "epoch": 2.3175355450236967,
      "grad_norm": 0.27086126804351807,
      "learning_rate": 0.000160517474906692,
      "loss": 0.19196972250938416,
      "mean_token_accuracy": 0.9190964996814728,
      "num_tokens": 54288680.0,
      "step": 4401
    },
    {
      "entropy": 1.334696739912033,
      "epoch": 2.318062137967351,
      "grad_norm": 0.27888646721839905,
      "learning_rate": 0.00016049932536291552,
      "loss": 0.1965310424566269,
      "mean_token_accuracy": 0.9213904738426208,
      "num_tokens": 54301016.0,
      "step": 4402
    },
    {
      "entropy": 1.3062738478183746,
      "epoch": 2.3185887309110056,
      "grad_norm": 0.274829238653183,
      "learning_rate": 0.00016048117282128842,
      "loss": 0.1787204146385193,
      "mean_token_accuracy": 0.9259192049503326,
      "num_tokens": 54313352.0,
      "step": 4403
    },
    {
      "entropy": 1.3323825299739838,
      "epoch": 2.3191153238546605,
      "grad_norm": 0.3422197699546814,
      "learning_rate": 0.00016046301728288835,
      "loss": 0.2093147188425064,
      "mean_token_accuracy": 0.91293103992939,
      "num_tokens": 54325688.0,
      "step": 4404
    },
    {
      "entropy": 1.3408651053905487,
      "epoch": 2.319641916798315,
      "grad_norm": 0.27782759070396423,
      "learning_rate": 0.00016044485874879303,
      "loss": 0.18714700639247894,
      "mean_token_accuracy": 0.9252272844314575,
      "num_tokens": 54338024.0,
      "step": 4405
    },
    {
      "entropy": 1.2752661406993866,
      "epoch": 2.3201685097419693,
      "grad_norm": 0.2649306356906891,
      "learning_rate": 0.00016042669722008047,
      "loss": 0.18565596640110016,
      "mean_token_accuracy": 0.92513607442379,
      "num_tokens": 54350360.0,
      "step": 4406
    },
    {
      "entropy": 1.3408364355564117,
      "epoch": 2.320695102685624,
      "grad_norm": 0.2725090980529785,
      "learning_rate": 0.00016040853269782881,
      "loss": 0.1821533590555191,
      "mean_token_accuracy": 0.9206462651491165,
      "num_tokens": 54362696.0,
      "step": 4407
    },
    {
      "entropy": 1.3324918746948242,
      "epoch": 2.3212216956292786,
      "grad_norm": 0.2690848708152771,
      "learning_rate": 0.00016039036518311633,
      "loss": 0.19127275049686432,
      "mean_token_accuracy": 0.9218493103981018,
      "num_tokens": 54375032.0,
      "step": 4408
    },
    {
      "entropy": 1.387494683265686,
      "epoch": 2.321748288572933,
      "grad_norm": 0.29377731680870056,
      "learning_rate": 0.00016037219467702162,
      "loss": 0.19303129613399506,
      "mean_token_accuracy": 0.9185487180948257,
      "num_tokens": 54387368.0,
      "step": 4409
    },
    {
      "entropy": 1.3866168558597565,
      "epoch": 2.322274881516588,
      "grad_norm": 0.30252406001091003,
      "learning_rate": 0.00016035402118062329,
      "loss": 0.18900904059410095,
      "mean_token_accuracy": 0.9236882477998734,
      "num_tokens": 54399704.0,
      "step": 4410
    },
    {
      "entropy": 1.3696815967559814,
      "epoch": 2.3228014744602423,
      "grad_norm": 0.3095356225967407,
      "learning_rate": 0.00016033584469500017,
      "loss": 0.21661263704299927,
      "mean_token_accuracy": 0.9109258502721786,
      "num_tokens": 54412040.0,
      "step": 4411
    },
    {
      "entropy": 1.3689590990543365,
      "epoch": 2.3233280674038967,
      "grad_norm": 0.30804696679115295,
      "learning_rate": 0.00016031766522123136,
      "loss": 0.1792442947626114,
      "mean_token_accuracy": 0.9260575175285339,
      "num_tokens": 54424376.0,
      "step": 4412
    },
    {
      "entropy": 1.3878648579120636,
      "epoch": 2.323854660347551,
      "grad_norm": 0.26045045256614685,
      "learning_rate": 0.00016029948276039606,
      "loss": 0.1889743059873581,
      "mean_token_accuracy": 0.9185473471879959,
      "num_tokens": 54436712.0,
      "step": 4413
    },
    {
      "entropy": 1.3724315762519836,
      "epoch": 2.324381253291206,
      "grad_norm": 0.2750447392463684,
      "learning_rate": 0.00016028129731357366,
      "loss": 0.1990983784198761,
      "mean_token_accuracy": 0.9145029634237289,
      "num_tokens": 54449048.0,
      "step": 4414
    },
    {
      "entropy": 1.3819120228290558,
      "epoch": 2.3249078462348605,
      "grad_norm": 0.2784717082977295,
      "learning_rate": 0.00016026310888184363,
      "loss": 0.1772555112838745,
      "mean_token_accuracy": 0.9232773780822754,
      "num_tokens": 54461384.0,
      "step": 4415
    },
    {
      "entropy": 1.4001644849777222,
      "epoch": 2.325434439178515,
      "grad_norm": 0.2536607086658478,
      "learning_rate": 0.00016024491746628578,
      "loss": 0.17961403727531433,
      "mean_token_accuracy": 0.9264721125364304,
      "num_tokens": 54473720.0,
      "step": 4416
    },
    {
      "entropy": 1.4117590487003326,
      "epoch": 2.3259610321221693,
      "grad_norm": 0.2814858853816986,
      "learning_rate": 0.00016022672306798005,
      "loss": 0.2062462568283081,
      "mean_token_accuracy": 0.9142315089702606,
      "num_tokens": 54486056.0,
      "step": 4417
    },
    {
      "entropy": 1.4305457472801208,
      "epoch": 2.326487625065824,
      "grad_norm": 0.2592325508594513,
      "learning_rate": 0.00016020852568800653,
      "loss": 0.1808975487947464,
      "mean_token_accuracy": 0.9286819398403168,
      "num_tokens": 54498392.0,
      "step": 4418
    },
    {
      "entropy": 1.3532997369766235,
      "epoch": 2.3270142180094786,
      "grad_norm": 0.28283217549324036,
      "learning_rate": 0.00016019032532744548,
      "loss": 0.18879854679107666,
      "mean_token_accuracy": 0.9208284020423889,
      "num_tokens": 54510728.0,
      "step": 4419
    },
    {
      "entropy": 1.3316220939159393,
      "epoch": 2.327540810953133,
      "grad_norm": 0.2630079686641693,
      "learning_rate": 0.00016017212198737732,
      "loss": 0.17229636013507843,
      "mean_token_accuracy": 0.9275263994932175,
      "num_tokens": 54523064.0,
      "step": 4420
    },
    {
      "entropy": 1.388702541589737,
      "epoch": 2.328067403896788,
      "grad_norm": 0.2867870032787323,
      "learning_rate": 0.00016015391566888271,
      "loss": 0.20351353287696838,
      "mean_token_accuracy": 0.9170709401369095,
      "num_tokens": 54535400.0,
      "step": 4421
    },
    {
      "entropy": 1.3147936463356018,
      "epoch": 2.3285939968404423,
      "grad_norm": 0.2731653153896332,
      "learning_rate": 0.00016013570637304244,
      "loss": 0.19753791391849518,
      "mean_token_accuracy": 0.9179530590772629,
      "num_tokens": 54547736.0,
      "step": 4422
    },
    {
      "entropy": 1.3937702775001526,
      "epoch": 2.3291205897840968,
      "grad_norm": 0.28775882720947266,
      "learning_rate": 0.00016011749410093753,
      "loss": 0.18566149473190308,
      "mean_token_accuracy": 0.9243120849132538,
      "num_tokens": 54560072.0,
      "step": 4423
    },
    {
      "entropy": 1.3537203073501587,
      "epoch": 2.3296471827277516,
      "grad_norm": 0.2538524866104126,
      "learning_rate": 0.00016009927885364907,
      "loss": 0.18132588267326355,
      "mean_token_accuracy": 0.9244391024112701,
      "num_tokens": 54572408.0,
      "step": 4424
    },
    {
      "entropy": 1.328016459941864,
      "epoch": 2.330173775671406,
      "grad_norm": 0.30183935165405273,
      "learning_rate": 0.00016008106063225845,
      "loss": 0.21323493123054504,
      "mean_token_accuracy": 0.9161577224731445,
      "num_tokens": 54584744.0,
      "step": 4425
    },
    {
      "entropy": 1.3058584034442902,
      "epoch": 2.3307003686150605,
      "grad_norm": 0.28315648436546326,
      "learning_rate": 0.00016006283943784715,
      "loss": 0.20466047525405884,
      "mean_token_accuracy": 0.9173526763916016,
      "num_tokens": 54597080.0,
      "step": 4426
    },
    {
      "entropy": 1.371433824300766,
      "epoch": 2.3312269615587153,
      "grad_norm": 0.2796994149684906,
      "learning_rate": 0.00016004461527149687,
      "loss": 0.19173529744148254,
      "mean_token_accuracy": 0.9192550033330917,
      "num_tokens": 54609416.0,
      "step": 4427
    },
    {
      "entropy": 1.2901140749454498,
      "epoch": 2.3317535545023698,
      "grad_norm": 0.2696458399295807,
      "learning_rate": 0.00016002638813428943,
      "loss": 0.18851207196712494,
      "mean_token_accuracy": 0.9239564538002014,
      "num_tokens": 54621752.0,
      "step": 4428
    },
    {
      "entropy": 1.317545861005783,
      "epoch": 2.332280147446024,
      "grad_norm": 0.3081064522266388,
      "learning_rate": 0.00016000815802730693,
      "loss": 0.20056316256523132,
      "mean_token_accuracy": 0.9122645407915115,
      "num_tokens": 54634088.0,
      "step": 4429
    },
    {
      "entropy": 1.2482907176017761,
      "epoch": 2.3328067403896786,
      "grad_norm": 0.2816668152809143,
      "learning_rate": 0.0001599899249516315,
      "loss": 0.18329255282878876,
      "mean_token_accuracy": 0.9248232841491699,
      "num_tokens": 54646424.0,
      "step": 4430
    },
    {
      "entropy": 1.2596892416477203,
      "epoch": 2.3333333333333335,
      "grad_norm": 0.27750101685523987,
      "learning_rate": 0.00015997168890834567,
      "loss": 0.2004767507314682,
      "mean_token_accuracy": 0.9160079509019852,
      "num_tokens": 54658760.0,
      "step": 4431
    },
    {
      "entropy": 1.2721357941627502,
      "epoch": 2.333859926276988,
      "grad_norm": 0.28640201687812805,
      "learning_rate": 0.00015995344989853193,
      "loss": 0.18829813599586487,
      "mean_token_accuracy": 0.9209965318441391,
      "num_tokens": 54671096.0,
      "step": 4432
    },
    {
      "entropy": 1.281269371509552,
      "epoch": 2.3343865192206423,
      "grad_norm": 0.2667178809642792,
      "learning_rate": 0.00015993520792327294,
      "loss": 0.18476279079914093,
      "mean_token_accuracy": 0.9250704199075699,
      "num_tokens": 54683432.0,
      "step": 4433
    },
    {
      "entropy": 1.2215364575386047,
      "epoch": 2.3349131121642968,
      "grad_norm": 0.29263362288475037,
      "learning_rate": 0.00015991696298365174,
      "loss": 0.21507254242897034,
      "mean_token_accuracy": 0.9111363440752029,
      "num_tokens": 54695768.0,
      "step": 4434
    },
    {
      "entropy": 1.1838223040103912,
      "epoch": 2.3354397051079516,
      "grad_norm": 0.25401586294174194,
      "learning_rate": 0.00015989871508075137,
      "loss": 0.18073657155036926,
      "mean_token_accuracy": 0.9278956204652786,
      "num_tokens": 54708104.0,
      "step": 4435
    },
    {
      "entropy": 1.2792393565177917,
      "epoch": 2.335966298051606,
      "grad_norm": 0.2642403542995453,
      "learning_rate": 0.00015988046421565517,
      "loss": 0.1823883056640625,
      "mean_token_accuracy": 0.9233395755290985,
      "num_tokens": 54720440.0,
      "step": 4436
    },
    {
      "entropy": 1.281785398721695,
      "epoch": 2.3364928909952605,
      "grad_norm": 0.30070072412490845,
      "learning_rate": 0.0001598622103894465,
      "loss": 0.19212502241134644,
      "mean_token_accuracy": 0.9186575263738632,
      "num_tokens": 54732776.0,
      "step": 4437
    },
    {
      "entropy": 1.2644932866096497,
      "epoch": 2.3370194839389153,
      "grad_norm": 0.3136497437953949,
      "learning_rate": 0.00015984395360320902,
      "loss": 0.19951491057872772,
      "mean_token_accuracy": 0.9213179051876068,
      "num_tokens": 54745112.0,
      "step": 4438
    },
    {
      "entropy": 1.2794144749641418,
      "epoch": 2.3375460768825698,
      "grad_norm": 0.27775052189826965,
      "learning_rate": 0.00015982569385802653,
      "loss": 0.18891870975494385,
      "mean_token_accuracy": 0.9204751700162888,
      "num_tokens": 54757448.0,
      "step": 4439
    },
    {
      "entropy": 1.2809376120567322,
      "epoch": 2.338072669826224,
      "grad_norm": 0.28730466961860657,
      "learning_rate": 0.00015980743115498296,
      "loss": 0.20424111187458038,
      "mean_token_accuracy": 0.9203447997570038,
      "num_tokens": 54769784.0,
      "step": 4440
    },
    {
      "entropy": 1.2839619815349579,
      "epoch": 2.338599262769879,
      "grad_norm": 0.3166145384311676,
      "learning_rate": 0.00015978916549516257,
      "loss": 0.2077278047800064,
      "mean_token_accuracy": 0.9169203490018845,
      "num_tokens": 54782120.0,
      "step": 4441
    },
    {
      "entropy": 1.2435754537582397,
      "epoch": 2.3391258557135335,
      "grad_norm": 0.26010408997535706,
      "learning_rate": 0.0001597708968796496,
      "loss": 0.1755719631910324,
      "mean_token_accuracy": 0.9279258698225021,
      "num_tokens": 54794456.0,
      "step": 4442
    },
    {
      "entropy": 1.210321307182312,
      "epoch": 2.339652448657188,
      "grad_norm": 0.26333627104759216,
      "learning_rate": 0.00015975262530952853,
      "loss": 0.17494051158428192,
      "mean_token_accuracy": 0.9276866167783737,
      "num_tokens": 54806792.0,
      "step": 4443
    },
    {
      "entropy": 1.282522976398468,
      "epoch": 2.340179041600843,
      "grad_norm": 0.2993547022342682,
      "learning_rate": 0.0001597343507858841,
      "loss": 0.1840740144252777,
      "mean_token_accuracy": 0.9280155897140503,
      "num_tokens": 54819128.0,
      "step": 4444
    },
    {
      "entropy": 1.2530391216278076,
      "epoch": 2.340705634544497,
      "grad_norm": 0.28764256834983826,
      "learning_rate": 0.00015971607330980117,
      "loss": 0.19909349083900452,
      "mean_token_accuracy": 0.9169367849826813,
      "num_tokens": 54831464.0,
      "step": 4445
    },
    {
      "entropy": 1.3018465340137482,
      "epoch": 2.3412322274881516,
      "grad_norm": 0.2826826870441437,
      "learning_rate": 0.00015969779288236468,
      "loss": 0.19262577593326569,
      "mean_token_accuracy": 0.920672133564949,
      "num_tokens": 54843800.0,
      "step": 4446
    },
    {
      "entropy": 1.2566682994365692,
      "epoch": 2.341758820431806,
      "grad_norm": 0.29000788927078247,
      "learning_rate": 0.0001596795095046599,
      "loss": 0.21003186702728271,
      "mean_token_accuracy": 0.9126958400011063,
      "num_tokens": 54856136.0,
      "step": 4447
    },
    {
      "entropy": 1.2816908955574036,
      "epoch": 2.342285413375461,
      "grad_norm": 0.284256249666214,
      "learning_rate": 0.00015966122317777223,
      "loss": 0.20872412621974945,
      "mean_token_accuracy": 0.9170348197221756,
      "num_tokens": 54868472.0,
      "step": 4448
    },
    {
      "entropy": 1.2890898287296295,
      "epoch": 2.3428120063191153,
      "grad_norm": 0.2809978723526001,
      "learning_rate": 0.00015964293390278718,
      "loss": 0.20213517546653748,
      "mean_token_accuracy": 0.9191296994686127,
      "num_tokens": 54880808.0,
      "step": 4449
    },
    {
      "entropy": 1.3205406069755554,
      "epoch": 2.3433385992627698,
      "grad_norm": 0.27979543805122375,
      "learning_rate": 0.00015962464168079045,
      "loss": 0.195961132645607,
      "mean_token_accuracy": 0.9169862270355225,
      "num_tokens": 54893144.0,
      "step": 4450
    },
    {
      "entropy": 1.334695279598236,
      "epoch": 2.343865192206424,
      "grad_norm": 0.26562100648880005,
      "learning_rate": 0.000159606346512868,
      "loss": 0.1865580528974533,
      "mean_token_accuracy": 0.923822671175003,
      "num_tokens": 54905480.0,
      "step": 4451
    },
    {
      "entropy": 1.3165518641471863,
      "epoch": 2.344391785150079,
      "grad_norm": 0.2917470335960388,
      "learning_rate": 0.0001595880484001059,
      "loss": 0.20150445401668549,
      "mean_token_accuracy": 0.9178837388753891,
      "num_tokens": 54917816.0,
      "step": 4452
    },
    {
      "entropy": 1.3582655489444733,
      "epoch": 2.3449183780937335,
      "grad_norm": 0.26764020323753357,
      "learning_rate": 0.00015956974734359034,
      "loss": 0.18483439087867737,
      "mean_token_accuracy": 0.920818954706192,
      "num_tokens": 54930152.0,
      "step": 4453
    },
    {
      "entropy": 1.3119309544563293,
      "epoch": 2.345444971037388,
      "grad_norm": 0.2685227692127228,
      "learning_rate": 0.00015955144334440783,
      "loss": 0.20485417544841766,
      "mean_token_accuracy": 0.9134972244501114,
      "num_tokens": 54942488.0,
      "step": 4454
    },
    {
      "entropy": 1.3058526515960693,
      "epoch": 2.345971563981043,
      "grad_norm": 0.26575425267219543,
      "learning_rate": 0.00015953313640364492,
      "loss": 0.20995725691318512,
      "mean_token_accuracy": 0.912079855799675,
      "num_tokens": 54954824.0,
      "step": 4455
    },
    {
      "entropy": 1.3167513608932495,
      "epoch": 2.346498156924697,
      "grad_norm": 0.2694823145866394,
      "learning_rate": 0.00015951482652238843,
      "loss": 0.19169099628925323,
      "mean_token_accuracy": 0.9215113818645477,
      "num_tokens": 54967160.0,
      "step": 4456
    },
    {
      "entropy": 1.3394544422626495,
      "epoch": 2.3470247498683516,
      "grad_norm": 0.2890017032623291,
      "learning_rate": 0.00015949651370172525,
      "loss": 0.20341987907886505,
      "mean_token_accuracy": 0.9149002283811569,
      "num_tokens": 54979496.0,
      "step": 4457
    },
    {
      "entropy": 1.344273865222931,
      "epoch": 2.3475513428120065,
      "grad_norm": 0.27608951926231384,
      "learning_rate": 0.00015947819794274258,
      "loss": 0.1875733733177185,
      "mean_token_accuracy": 0.9171324670314789,
      "num_tokens": 54991832.0,
      "step": 4458
    },
    {
      "entropy": 1.3028335869312286,
      "epoch": 2.348077935755661,
      "grad_norm": 0.2981983423233032,
      "learning_rate": 0.00015945987924652764,
      "loss": 0.20515939593315125,
      "mean_token_accuracy": 0.91654272377491,
      "num_tokens": 55004168.0,
      "step": 4459
    },
    {
      "entropy": 1.3539035320281982,
      "epoch": 2.3486045286993154,
      "grad_norm": 0.27947360277175903,
      "learning_rate": 0.00015944155761416797,
      "loss": 0.1902858167886734,
      "mean_token_accuracy": 0.9198180288076401,
      "num_tokens": 55016504.0,
      "step": 4460
    },
    {
      "entropy": 1.317354530096054,
      "epoch": 2.34913112164297,
      "grad_norm": 0.2992118299007416,
      "learning_rate": 0.00015942323304675118,
      "loss": 0.1987239122390747,
      "mean_token_accuracy": 0.9191636443138123,
      "num_tokens": 55028840.0,
      "step": 4461
    },
    {
      "entropy": 1.310655266046524,
      "epoch": 2.3496577145866246,
      "grad_norm": 0.26667365431785583,
      "learning_rate": 0.0001594049055453651,
      "loss": 0.1835993528366089,
      "mean_token_accuracy": 0.922356128692627,
      "num_tokens": 55041176.0,
      "step": 4462
    },
    {
      "entropy": 1.254290759563446,
      "epoch": 2.350184307530279,
      "grad_norm": 0.2572006285190582,
      "learning_rate": 0.00015938657511109776,
      "loss": 0.18122154474258423,
      "mean_token_accuracy": 0.9250338971614838,
      "num_tokens": 55053512.0,
      "step": 4463
    },
    {
      "entropy": 1.330580860376358,
      "epoch": 2.3507109004739335,
      "grad_norm": 0.2904733717441559,
      "learning_rate": 0.00015936824174503727,
      "loss": 0.18564851582050323,
      "mean_token_accuracy": 0.9225763231515884,
      "num_tokens": 55065848.0,
      "step": 4464
    },
    {
      "entropy": 1.285946011543274,
      "epoch": 2.3512374934175884,
      "grad_norm": 0.28024446964263916,
      "learning_rate": 0.000159349905448272,
      "loss": 0.1889662742614746,
      "mean_token_accuracy": 0.9220243245363235,
      "num_tokens": 55078184.0,
      "step": 4465
    },
    {
      "entropy": 1.3022855818271637,
      "epoch": 2.351764086361243,
      "grad_norm": 0.2870224118232727,
      "learning_rate": 0.0001593315662218905,
      "loss": 0.18644535541534424,
      "mean_token_accuracy": 0.9205256253480911,
      "num_tokens": 55090520.0,
      "step": 4466
    },
    {
      "entropy": 1.2169810831546783,
      "epoch": 2.352290679304897,
      "grad_norm": 0.24071332812309265,
      "learning_rate": 0.00015931322406698142,
      "loss": 0.16642692685127258,
      "mean_token_accuracy": 0.930572435259819,
      "num_tokens": 55102856.0,
      "step": 4467
    },
    {
      "entropy": 1.2590057253837585,
      "epoch": 2.3528172722485516,
      "grad_norm": 0.2916509807109833,
      "learning_rate": 0.00015929487898463368,
      "loss": 0.1792268007993698,
      "mean_token_accuracy": 0.9278564751148224,
      "num_tokens": 55115192.0,
      "step": 4468
    },
    {
      "entropy": 1.3424552083015442,
      "epoch": 2.3533438651922065,
      "grad_norm": 0.3079618513584137,
      "learning_rate": 0.00015927653097593626,
      "loss": 0.20497339963912964,
      "mean_token_accuracy": 0.9212962538003922,
      "num_tokens": 55127528.0,
      "step": 4469
    },
    {
      "entropy": 1.3330655097961426,
      "epoch": 2.353870458135861,
      "grad_norm": 0.2870141863822937,
      "learning_rate": 0.0001592581800419784,
      "loss": 0.19076713919639587,
      "mean_token_accuracy": 0.9209161102771759,
      "num_tokens": 55139864.0,
      "step": 4470
    },
    {
      "entropy": 1.3257507979869843,
      "epoch": 2.3543970510795154,
      "grad_norm": 0.28117015957832336,
      "learning_rate": 0.00015923982618384948,
      "loss": 0.20129868388175964,
      "mean_token_accuracy": 0.9233514815568924,
      "num_tokens": 55152200.0,
      "step": 4471
    },
    {
      "entropy": 1.3009706437587738,
      "epoch": 2.3549236440231702,
      "grad_norm": 0.3033812940120697,
      "learning_rate": 0.0001592214694026391,
      "loss": 0.18937650322914124,
      "mean_token_accuracy": 0.9224123954772949,
      "num_tokens": 55164536.0,
      "step": 4472
    },
    {
      "entropy": 1.2963763773441315,
      "epoch": 2.3554502369668247,
      "grad_norm": 0.2638135552406311,
      "learning_rate": 0.00015920310969943697,
      "loss": 0.1769077479839325,
      "mean_token_accuracy": 0.9286084026098251,
      "num_tokens": 55176872.0,
      "step": 4473
    },
    {
      "entropy": 1.3231896758079529,
      "epoch": 2.355976829910479,
      "grad_norm": 0.26197150349617004,
      "learning_rate": 0.00015918474707533298,
      "loss": 0.17836906015872955,
      "mean_token_accuracy": 0.9256467223167419,
      "num_tokens": 55189208.0,
      "step": 4474
    },
    {
      "entropy": 1.3913252353668213,
      "epoch": 2.356503422854134,
      "grad_norm": 0.2851298451423645,
      "learning_rate": 0.00015916638153141722,
      "loss": 0.18936462700366974,
      "mean_token_accuracy": 0.9204428941011429,
      "num_tokens": 55201544.0,
      "step": 4475
    },
    {
      "entropy": 1.3488848209381104,
      "epoch": 2.3570300157977884,
      "grad_norm": 0.30388936400413513,
      "learning_rate": 0.00015914801306877993,
      "loss": 0.21345588564872742,
      "mean_token_accuracy": 0.9136727303266525,
      "num_tokens": 55213880.0,
      "step": 4476
    },
    {
      "entropy": 1.3502046167850494,
      "epoch": 2.357556608741443,
      "grad_norm": 0.28946250677108765,
      "learning_rate": 0.0001591296416885116,
      "loss": 0.20996083319187164,
      "mean_token_accuracy": 0.9157253056764603,
      "num_tokens": 55226216.0,
      "step": 4477
    },
    {
      "entropy": 1.3586644530296326,
      "epoch": 2.3580832016850977,
      "grad_norm": 0.27651745080947876,
      "learning_rate": 0.0001591112673917028,
      "loss": 0.18405741453170776,
      "mean_token_accuracy": 0.924022987484932,
      "num_tokens": 55238552.0,
      "step": 4478
    },
    {
      "entropy": 1.430792659521103,
      "epoch": 2.358609794628752,
      "grad_norm": 0.2895481288433075,
      "learning_rate": 0.00015909289017944426,
      "loss": 0.18251873552799225,
      "mean_token_accuracy": 0.923512801527977,
      "num_tokens": 55250888.0,
      "step": 4479
    },
    {
      "entropy": 1.3671636581420898,
      "epoch": 2.3591363875724065,
      "grad_norm": 0.25113236904144287,
      "learning_rate": 0.00015907451005282698,
      "loss": 0.16419190168380737,
      "mean_token_accuracy": 0.9323639422655106,
      "num_tokens": 55263224.0,
      "step": 4480
    },
    {
      "entropy": 1.354794681072235,
      "epoch": 2.359662980516061,
      "grad_norm": 0.296430766582489,
      "learning_rate": 0.00015905612701294208,
      "loss": 0.20118844509124756,
      "mean_token_accuracy": 0.9132822453975677,
      "num_tokens": 55275560.0,
      "step": 4481
    },
    {
      "entropy": 1.3990836441516876,
      "epoch": 2.360189573459716,
      "grad_norm": 0.28303229808807373,
      "learning_rate": 0.0001590377410608808,
      "loss": 0.20889045298099518,
      "mean_token_accuracy": 0.9137723445892334,
      "num_tokens": 55287896.0,
      "step": 4482
    },
    {
      "entropy": 1.3733917474746704,
      "epoch": 2.3607161664033702,
      "grad_norm": 0.27674078941345215,
      "learning_rate": 0.0001590193521977347,
      "loss": 0.1934237778186798,
      "mean_token_accuracy": 0.921452522277832,
      "num_tokens": 55300232.0,
      "step": 4483
    },
    {
      "entropy": 1.355101853609085,
      "epoch": 2.3612427593470247,
      "grad_norm": 0.24528248608112335,
      "learning_rate": 0.00015900096042459534,
      "loss": 0.17529264092445374,
      "mean_token_accuracy": 0.925555557012558,
      "num_tokens": 55312568.0,
      "step": 4484
    },
    {
      "entropy": 1.4111731946468353,
      "epoch": 2.361769352290679,
      "grad_norm": 0.28693532943725586,
      "learning_rate": 0.00015898256574255456,
      "loss": 0.1895090639591217,
      "mean_token_accuracy": 0.9167554080486298,
      "num_tokens": 55324904.0,
      "step": 4485
    },
    {
      "entropy": 1.3618127703666687,
      "epoch": 2.362295945234334,
      "grad_norm": 0.2586255669593811,
      "learning_rate": 0.00015896416815270437,
      "loss": 0.17887786030769348,
      "mean_token_accuracy": 0.9279444813728333,
      "num_tokens": 55337240.0,
      "step": 4486
    },
    {
      "entropy": 1.4009781777858734,
      "epoch": 2.3628225381779884,
      "grad_norm": 0.2798784673213959,
      "learning_rate": 0.00015894576765613684,
      "loss": 0.2002013474702835,
      "mean_token_accuracy": 0.9198189824819565,
      "num_tokens": 55349576.0,
      "step": 4487
    },
    {
      "entropy": 1.4672830998897552,
      "epoch": 2.363349131121643,
      "grad_norm": 0.30033984780311584,
      "learning_rate": 0.00015892736425394442,
      "loss": 0.19836139678955078,
      "mean_token_accuracy": 0.9234947264194489,
      "num_tokens": 55361912.0,
      "step": 4488
    },
    {
      "entropy": 1.4400977194309235,
      "epoch": 2.3638757240652977,
      "grad_norm": 0.2626227140426636,
      "learning_rate": 0.00015890895794721955,
      "loss": 0.1710968166589737,
      "mean_token_accuracy": 0.9296213835477829,
      "num_tokens": 55374248.0,
      "step": 4489
    },
    {
      "entropy": 1.4096495509147644,
      "epoch": 2.364402317008952,
      "grad_norm": 0.282968670129776,
      "learning_rate": 0.00015889054873705486,
      "loss": 0.1910400092601776,
      "mean_token_accuracy": 0.9231160432100296,
      "num_tokens": 55386584.0,
      "step": 4490
    },
    {
      "entropy": 1.407538741827011,
      "epoch": 2.3649289099526065,
      "grad_norm": 0.2817977964878082,
      "learning_rate": 0.0001588721366245433,
      "loss": 0.17246387898921967,
      "mean_token_accuracy": 0.9259490519762039,
      "num_tokens": 55398920.0,
      "step": 4491
    },
    {
      "entropy": 1.3910426497459412,
      "epoch": 2.3654555028962614,
      "grad_norm": 0.28504490852355957,
      "learning_rate": 0.0001588537216107778,
      "loss": 0.1968700885772705,
      "mean_token_accuracy": 0.9202799946069717,
      "num_tokens": 55411256.0,
      "step": 4492
    },
    {
      "entropy": 1.4219735264778137,
      "epoch": 2.365982095839916,
      "grad_norm": 0.2912696301937103,
      "learning_rate": 0.00015883530369685162,
      "loss": 0.2083374410867691,
      "mean_token_accuracy": 0.9184867888689041,
      "num_tokens": 55423592.0,
      "step": 4493
    },
    {
      "entropy": 1.343996375799179,
      "epoch": 2.3665086887835702,
      "grad_norm": 0.25932249426841736,
      "learning_rate": 0.00015881688288385808,
      "loss": 0.169301837682724,
      "mean_token_accuracy": 0.9262981861829758,
      "num_tokens": 55435928.0,
      "step": 4494
    },
    {
      "entropy": 1.4267469346523285,
      "epoch": 2.367035281727225,
      "grad_norm": 0.2699030339717865,
      "learning_rate": 0.00015879845917289074,
      "loss": 0.18538378179073334,
      "mean_token_accuracy": 0.9267265200614929,
      "num_tokens": 55448264.0,
      "step": 4495
    },
    {
      "entropy": 1.4803634285926819,
      "epoch": 2.3675618746708795,
      "grad_norm": 0.3319495618343353,
      "learning_rate": 0.00015878003256504327,
      "loss": 0.20425397157669067,
      "mean_token_accuracy": 0.9164634346961975,
      "num_tokens": 55460600.0,
      "step": 4496
    },
    {
      "entropy": 1.385725349187851,
      "epoch": 2.368088467614534,
      "grad_norm": 0.25805506110191345,
      "learning_rate": 0.0001587616030614096,
      "loss": 0.18095749616622925,
      "mean_token_accuracy": 0.922220304608345,
      "num_tokens": 55472936.0,
      "step": 4497
    },
    {
      "entropy": 1.4062891602516174,
      "epoch": 2.3686150605581884,
      "grad_norm": 0.25933024287223816,
      "learning_rate": 0.00015874317066308372,
      "loss": 0.18179509043693542,
      "mean_token_accuracy": 0.9226569682359695,
      "num_tokens": 55485272.0,
      "step": 4498
    },
    {
      "entropy": 1.4053870141506195,
      "epoch": 2.3691416535018432,
      "grad_norm": 0.2785022556781769,
      "learning_rate": 0.00015872473537115994,
      "loss": 0.19450442492961884,
      "mean_token_accuracy": 0.9210056364536285,
      "num_tokens": 55497608.0,
      "step": 4499
    },
    {
      "entropy": 1.346811443567276,
      "epoch": 2.3696682464454977,
      "grad_norm": 0.2664346992969513,
      "learning_rate": 0.0001587062971867326,
      "loss": 0.17768292129039764,
      "mean_token_accuracy": 0.926102340221405,
      "num_tokens": 55509944.0,
      "step": 4500
    },
    {
      "entropy": 1.528866469860077,
      "epoch": 2.370194839389152,
      "grad_norm": 0.3126022517681122,
      "learning_rate": 0.00015868785611089628,
      "loss": 0.2166975736618042,
      "mean_token_accuracy": 0.9128769040107727,
      "num_tokens": 55522280.0,
      "step": 4501
    },
    {
      "entropy": 1.4011040925979614,
      "epoch": 2.3707214323328065,
      "grad_norm": 0.261837363243103,
      "learning_rate": 0.00015866941214474572,
      "loss": 0.18179045617580414,
      "mean_token_accuracy": 0.923045888543129,
      "num_tokens": 55534616.0,
      "step": 4502
    },
    {
      "entropy": 1.3975766003131866,
      "epoch": 2.3712480252764614,
      "grad_norm": 0.2936818301677704,
      "learning_rate": 0.0001586509652893758,
      "loss": 0.22525236010551453,
      "mean_token_accuracy": 0.9070886373519897,
      "num_tokens": 55546952.0,
      "step": 4503
    },
    {
      "entropy": 1.379688709974289,
      "epoch": 2.371774618220116,
      "grad_norm": 0.27700334787368774,
      "learning_rate": 0.00015863251554588167,
      "loss": 0.21290606260299683,
      "mean_token_accuracy": 0.910546600818634,
      "num_tokens": 55559288.0,
      "step": 4504
    },
    {
      "entropy": 1.3719074726104736,
      "epoch": 2.3723012111637702,
      "grad_norm": 0.28597623109817505,
      "learning_rate": 0.00015861406291535853,
      "loss": 0.20922252535820007,
      "mean_token_accuracy": 0.9105053544044495,
      "num_tokens": 55571624.0,
      "step": 4505
    },
    {
      "entropy": 1.3768506050109863,
      "epoch": 2.372827804107425,
      "grad_norm": 0.2684311866760254,
      "learning_rate": 0.0001585956073989018,
      "loss": 0.20058728754520416,
      "mean_token_accuracy": 0.9121564775705338,
      "num_tokens": 55583960.0,
      "step": 4506
    },
    {
      "entropy": 1.3842982351779938,
      "epoch": 2.3733543970510795,
      "grad_norm": 0.2636588215827942,
      "learning_rate": 0.00015857714899760714,
      "loss": 0.19942373037338257,
      "mean_token_accuracy": 0.9159530699253082,
      "num_tokens": 55596296.0,
      "step": 4507
    },
    {
      "entropy": 1.406856209039688,
      "epoch": 2.373880989994734,
      "grad_norm": 0.27418437600135803,
      "learning_rate": 0.00015855868771257024,
      "loss": 0.20501425862312317,
      "mean_token_accuracy": 0.9155614525079727,
      "num_tokens": 55608632.0,
      "step": 4508
    },
    {
      "entropy": 1.4277390837669373,
      "epoch": 2.374407582938389,
      "grad_norm": 0.2765462398529053,
      "learning_rate": 0.00015854022354488707,
      "loss": 0.19365204870700836,
      "mean_token_accuracy": 0.9140615314245224,
      "num_tokens": 55620968.0,
      "step": 4509
    },
    {
      "entropy": 1.4029557406902313,
      "epoch": 2.3749341758820433,
      "grad_norm": 0.27889713644981384,
      "learning_rate": 0.00015852175649565375,
      "loss": 0.18732376396656036,
      "mean_token_accuracy": 0.9244106113910675,
      "num_tokens": 55633304.0,
      "step": 4510
    },
    {
      "entropy": 1.3472702205181122,
      "epoch": 2.3754607688256977,
      "grad_norm": 0.24661894142627716,
      "learning_rate": 0.00015850328656596656,
      "loss": 0.17409662902355194,
      "mean_token_accuracy": 0.9250094890594482,
      "num_tokens": 55645640.0,
      "step": 4511
    },
    {
      "entropy": 1.395214170217514,
      "epoch": 2.375987361769352,
      "grad_norm": 0.27474910020828247,
      "learning_rate": 0.00015848481375692195,
      "loss": 0.19719195365905762,
      "mean_token_accuracy": 0.9223782122135162,
      "num_tokens": 55657976.0,
      "step": 4512
    },
    {
      "entropy": 1.3260896503925323,
      "epoch": 2.376513954713007,
      "grad_norm": 0.26417186856269836,
      "learning_rate": 0.00015846633806961647,
      "loss": 0.19815877079963684,
      "mean_token_accuracy": 0.9204468429088593,
      "num_tokens": 55670312.0,
      "step": 4513
    },
    {
      "entropy": 1.3533766269683838,
      "epoch": 2.3770405476566614,
      "grad_norm": 0.2891894578933716,
      "learning_rate": 0.00015844785950514708,
      "loss": 0.1904796063899994,
      "mean_token_accuracy": 0.9149835109710693,
      "num_tokens": 55682648.0,
      "step": 4514
    },
    {
      "entropy": 1.3211008608341217,
      "epoch": 2.377567140600316,
      "grad_norm": 0.26906976103782654,
      "learning_rate": 0.00015842937806461057,
      "loss": 0.1900303214788437,
      "mean_token_accuracy": 0.9222564399242401,
      "num_tokens": 55694984.0,
      "step": 4515
    },
    {
      "entropy": 1.3689544200897217,
      "epoch": 2.3780937335439707,
      "grad_norm": 0.2991812229156494,
      "learning_rate": 0.0001584108937491042,
      "loss": 0.20970046520233154,
      "mean_token_accuracy": 0.9113328009843826,
      "num_tokens": 55707320.0,
      "step": 4516
    },
    {
      "entropy": 1.3300418257713318,
      "epoch": 2.378620326487625,
      "grad_norm": 0.26500844955444336,
      "learning_rate": 0.00015839240655972521,
      "loss": 0.19760391116142273,
      "mean_token_accuracy": 0.9200556874275208,
      "num_tokens": 55719656.0,
      "step": 4517
    },
    {
      "entropy": 1.3151434361934662,
      "epoch": 2.3791469194312795,
      "grad_norm": 0.2675932049751282,
      "learning_rate": 0.00015837391649757112,
      "loss": 0.1939803957939148,
      "mean_token_accuracy": 0.9216755032539368,
      "num_tokens": 55731992.0,
      "step": 4518
    },
    {
      "entropy": 1.3362997472286224,
      "epoch": 2.379673512374934,
      "grad_norm": 0.28336596488952637,
      "learning_rate": 0.00015835542356373953,
      "loss": 0.1987985372543335,
      "mean_token_accuracy": 0.9186425060033798,
      "num_tokens": 55744328.0,
      "step": 4519
    },
    {
      "entropy": 1.3139479160308838,
      "epoch": 2.380200105318589,
      "grad_norm": 0.2725384831428528,
      "learning_rate": 0.0001583369277593283,
      "loss": 0.19263529777526855,
      "mean_token_accuracy": 0.9191529750823975,
      "num_tokens": 55756664.0,
      "step": 4520
    },
    {
      "entropy": 1.3443782925605774,
      "epoch": 2.3807266982622433,
      "grad_norm": 0.2858370840549469,
      "learning_rate": 0.00015831842908543535,
      "loss": 0.18691514432430267,
      "mean_token_accuracy": 0.9257304817438126,
      "num_tokens": 55769000.0,
      "step": 4521
    },
    {
      "entropy": 1.3328638076782227,
      "epoch": 2.3812532912058977,
      "grad_norm": 0.29020607471466064,
      "learning_rate": 0.00015829992754315893,
      "loss": 0.1900632083415985,
      "mean_token_accuracy": 0.9172062873840332,
      "num_tokens": 55781336.0,
      "step": 4522
    },
    {
      "entropy": 1.2869420647621155,
      "epoch": 2.3817798841495526,
      "grad_norm": 0.25875672698020935,
      "learning_rate": 0.0001582814231335973,
      "loss": 0.17436552047729492,
      "mean_token_accuracy": 0.9253624975681305,
      "num_tokens": 55793672.0,
      "step": 4523
    },
    {
      "entropy": 1.2536731362342834,
      "epoch": 2.382306477093207,
      "grad_norm": 0.28048229217529297,
      "learning_rate": 0.00015826291585784898,
      "loss": 0.1865769326686859,
      "mean_token_accuracy": 0.9238808900117874,
      "num_tokens": 55806008.0,
      "step": 4524
    },
    {
      "entropy": 1.23280930519104,
      "epoch": 2.3828330700368614,
      "grad_norm": 0.2543039321899414,
      "learning_rate": 0.00015824440571701268,
      "loss": 0.17272458970546722,
      "mean_token_accuracy": 0.9268594980239868,
      "num_tokens": 55818344.0,
      "step": 4525
    },
    {
      "entropy": 1.2371512353420258,
      "epoch": 2.3833596629805163,
      "grad_norm": 0.26792868971824646,
      "learning_rate": 0.00015822589271218717,
      "loss": 0.19100236892700195,
      "mean_token_accuracy": 0.9224984645843506,
      "num_tokens": 55830680.0,
      "step": 4526
    },
    {
      "entropy": 1.2913587987422943,
      "epoch": 2.3838862559241707,
      "grad_norm": 0.29319706559181213,
      "learning_rate": 0.00015820737684447148,
      "loss": 0.18937891721725464,
      "mean_token_accuracy": 0.9151134341955185,
      "num_tokens": 55843016.0,
      "step": 4527
    },
    {
      "entropy": 1.2287176549434662,
      "epoch": 2.384412848867825,
      "grad_norm": 0.28001704812049866,
      "learning_rate": 0.00015818885811496485,
      "loss": 0.18166252970695496,
      "mean_token_accuracy": 0.9257358908653259,
      "num_tokens": 55855352.0,
      "step": 4528
    },
    {
      "entropy": 1.2606195509433746,
      "epoch": 2.3849394418114795,
      "grad_norm": 0.3030807375907898,
      "learning_rate": 0.0001581703365247665,
      "loss": 0.2063525915145874,
      "mean_token_accuracy": 0.9173136651515961,
      "num_tokens": 55867688.0,
      "step": 4529
    },
    {
      "entropy": 1.2028226852416992,
      "epoch": 2.3854660347551344,
      "grad_norm": 0.26009899377822876,
      "learning_rate": 0.00015815181207497612,
      "loss": 0.17096900939941406,
      "mean_token_accuracy": 0.929774135351181,
      "num_tokens": 55880024.0,
      "step": 4530
    },
    {
      "entropy": 1.2134665250778198,
      "epoch": 2.385992627698789,
      "grad_norm": 0.2998177111148834,
      "learning_rate": 0.00015813328476669324,
      "loss": 0.20142899453639984,
      "mean_token_accuracy": 0.9195566177368164,
      "num_tokens": 55892360.0,
      "step": 4531
    },
    {
      "entropy": 1.1590828597545624,
      "epoch": 2.3865192206424433,
      "grad_norm": 0.28007999062538147,
      "learning_rate": 0.00015811475460101785,
      "loss": 0.18869996070861816,
      "mean_token_accuracy": 0.9209212511777878,
      "num_tokens": 55904696.0,
      "step": 4532
    },
    {
      "entropy": 1.1737896800041199,
      "epoch": 2.387045813586098,
      "grad_norm": 0.2900781035423279,
      "learning_rate": 0.00015809622157904985,
      "loss": 0.18893763422966003,
      "mean_token_accuracy": 0.9233887642621994,
      "num_tokens": 55917032.0,
      "step": 4533
    },
    {
      "entropy": 1.2201267778873444,
      "epoch": 2.3875724065297526,
      "grad_norm": 0.3050040900707245,
      "learning_rate": 0.0001580776857018895,
      "loss": 0.20816883444786072,
      "mean_token_accuracy": 0.9114056080579758,
      "num_tokens": 55929368.0,
      "step": 4534
    },
    {
      "entropy": 1.1619028151035309,
      "epoch": 2.388098999473407,
      "grad_norm": 0.24892818927764893,
      "learning_rate": 0.00015805914697063724,
      "loss": 0.17240609228610992,
      "mean_token_accuracy": 0.9284298419952393,
      "num_tokens": 55941704.0,
      "step": 4535
    },
    {
      "entropy": 1.1720469892024994,
      "epoch": 2.3886255924170614,
      "grad_norm": 0.2732456624507904,
      "learning_rate": 0.00015804060538639347,
      "loss": 0.18014155328273773,
      "mean_token_accuracy": 0.9247136563062668,
      "num_tokens": 55954040.0,
      "step": 4536
    },
    {
      "entropy": 1.213779777288437,
      "epoch": 2.3891521853607163,
      "grad_norm": 0.289726585149765,
      "learning_rate": 0.00015802206095025899,
      "loss": 0.19931016862392426,
      "mean_token_accuracy": 0.9159621000289917,
      "num_tokens": 55966376.0,
      "step": 4537
    },
    {
      "entropy": 1.2300103604793549,
      "epoch": 2.3896787783043707,
      "grad_norm": 0.2903289794921875,
      "learning_rate": 0.00015800351366333462,
      "loss": 0.2093479335308075,
      "mean_token_accuracy": 0.9123352319002151,
      "num_tokens": 55978712.0,
      "step": 4538
    },
    {
      "entropy": 1.2362595200538635,
      "epoch": 2.390205371248025,
      "grad_norm": 0.276406854391098,
      "learning_rate": 0.00015798496352672145,
      "loss": 0.1859022080898285,
      "mean_token_accuracy": 0.9236205220222473,
      "num_tokens": 55991048.0,
      "step": 4539
    },
    {
      "entropy": 1.159458488225937,
      "epoch": 2.39073196419168,
      "grad_norm": 0.2462838739156723,
      "learning_rate": 0.00015796641054152067,
      "loss": 0.1730072796344757,
      "mean_token_accuracy": 0.9274507462978363,
      "num_tokens": 56003384.0,
      "step": 4540
    },
    {
      "entropy": 1.2364217042922974,
      "epoch": 2.3912585571353344,
      "grad_norm": 0.2972487509250641,
      "learning_rate": 0.00015794785470883368,
      "loss": 0.19434306025505066,
      "mean_token_accuracy": 0.9204827547073364,
      "num_tokens": 56015720.0,
      "step": 4541
    },
    {
      "entropy": 1.2326800525188446,
      "epoch": 2.391785150078989,
      "grad_norm": 0.26879581809043884,
      "learning_rate": 0.000157929296029762,
      "loss": 0.1797732710838318,
      "mean_token_accuracy": 0.9242551773786545,
      "num_tokens": 56028056.0,
      "step": 4542
    },
    {
      "entropy": 1.170352816581726,
      "epoch": 2.3923117430226437,
      "grad_norm": 0.26433271169662476,
      "learning_rate": 0.00015791073450540735,
      "loss": 0.18631601333618164,
      "mean_token_accuracy": 0.9216004759073257,
      "num_tokens": 56040392.0,
      "step": 4543
    },
    {
      "entropy": 1.1886663138866425,
      "epoch": 2.392838335966298,
      "grad_norm": 0.26627394556999207,
      "learning_rate": 0.00015789217013687166,
      "loss": 0.18448396027088165,
      "mean_token_accuracy": 0.9213402569293976,
      "num_tokens": 56052728.0,
      "step": 4544
    },
    {
      "entropy": 1.1851184666156769,
      "epoch": 2.3933649289099526,
      "grad_norm": 0.2651765048503876,
      "learning_rate": 0.00015787360292525697,
      "loss": 0.18697291612625122,
      "mean_token_accuracy": 0.9274520725011826,
      "num_tokens": 56065064.0,
      "step": 4545
    },
    {
      "entropy": 1.2112329304218292,
      "epoch": 2.393891521853607,
      "grad_norm": 0.28112831711769104,
      "learning_rate": 0.00015785503287166547,
      "loss": 0.19521315395832062,
      "mean_token_accuracy": 0.9178619235754013,
      "num_tokens": 56077400.0,
      "step": 4546
    },
    {
      "entropy": 1.263805478811264,
      "epoch": 2.394418114797262,
      "grad_norm": 0.29383382201194763,
      "learning_rate": 0.00015783645997719962,
      "loss": 0.2024495154619217,
      "mean_token_accuracy": 0.9234823882579803,
      "num_tokens": 56089736.0,
      "step": 4547
    },
    {
      "entropy": 1.2429666221141815,
      "epoch": 2.3949447077409163,
      "grad_norm": 0.27420487999916077,
      "learning_rate": 0.00015781788424296193,
      "loss": 0.19035124778747559,
      "mean_token_accuracy": 0.9219522327184677,
      "num_tokens": 56102072.0,
      "step": 4548
    },
    {
      "entropy": 1.2490891516208649,
      "epoch": 2.3954713006845707,
      "grad_norm": 0.27929991483688354,
      "learning_rate": 0.00015779930567005518,
      "loss": 0.2075713574886322,
      "mean_token_accuracy": 0.9149742126464844,
      "num_tokens": 56114408.0,
      "step": 4549
    },
    {
      "entropy": 1.2860136330127716,
      "epoch": 2.3959978936282256,
      "grad_norm": 0.27160221338272095,
      "learning_rate": 0.00015778072425958226,
      "loss": 0.17518314719200134,
      "mean_token_accuracy": 0.9276988953351974,
      "num_tokens": 56126744.0,
      "step": 4550
    },
    {
      "entropy": 1.2763722836971283,
      "epoch": 2.39652448657188,
      "grad_norm": 0.29390084743499756,
      "learning_rate": 0.0001577621400126462,
      "loss": 0.19787636399269104,
      "mean_token_accuracy": 0.9196744412183762,
      "num_tokens": 56139080.0,
      "step": 4551
    },
    {
      "entropy": 1.3100577592849731,
      "epoch": 2.3970510795155344,
      "grad_norm": 0.2876659333705902,
      "learning_rate": 0.00015774355293035025,
      "loss": 0.1936154067516327,
      "mean_token_accuracy": 0.9236638695001602,
      "num_tokens": 56151416.0,
      "step": 4552
    },
    {
      "entropy": 1.3262332379817963,
      "epoch": 2.397577672459189,
      "grad_norm": 0.26849564909935,
      "learning_rate": 0.00015772496301379787,
      "loss": 0.1760268211364746,
      "mean_token_accuracy": 0.923967570066452,
      "num_tokens": 56163752.0,
      "step": 4553
    },
    {
      "entropy": 1.3029709458351135,
      "epoch": 2.3981042654028437,
      "grad_norm": 0.29166650772094727,
      "learning_rate": 0.00015770637026409258,
      "loss": 0.18877920508384705,
      "mean_token_accuracy": 0.9225371479988098,
      "num_tokens": 56176088.0,
      "step": 4554
    },
    {
      "entropy": 1.3118659257888794,
      "epoch": 2.398630858346498,
      "grad_norm": 0.28362366557121277,
      "learning_rate": 0.00015768777468233816,
      "loss": 0.19103732705116272,
      "mean_token_accuracy": 0.9223171025514603,
      "num_tokens": 56188424.0,
      "step": 4555
    },
    {
      "entropy": 1.3041376173496246,
      "epoch": 2.3991574512901526,
      "grad_norm": 0.3170718550682068,
      "learning_rate": 0.0001576691762696385,
      "loss": 0.208947092294693,
      "mean_token_accuracy": 0.914543017745018,
      "num_tokens": 56200760.0,
      "step": 4556
    },
    {
      "entropy": 1.291472166776657,
      "epoch": 2.3996840442338074,
      "grad_norm": 0.2711356282234192,
      "learning_rate": 0.00015765057502709767,
      "loss": 0.18828198313713074,
      "mean_token_accuracy": 0.9219943135976791,
      "num_tokens": 56213096.0,
      "step": 4557
    },
    {
      "entropy": 1.2951588928699493,
      "epoch": 2.400210637177462,
      "grad_norm": 0.29720601439476013,
      "learning_rate": 0.0001576319709558199,
      "loss": 0.21101555228233337,
      "mean_token_accuracy": 0.9118272066116333,
      "num_tokens": 56225432.0,
      "step": 4558
    },
    {
      "entropy": 1.3469326496124268,
      "epoch": 2.4007372301211163,
      "grad_norm": 0.2986636757850647,
      "learning_rate": 0.00015761336405690967,
      "loss": 0.18926507234573364,
      "mean_token_accuracy": 0.9249073565006256,
      "num_tokens": 56237768.0,
      "step": 4559
    },
    {
      "entropy": 1.308281421661377,
      "epoch": 2.401263823064771,
      "grad_norm": 0.27620309591293335,
      "learning_rate": 0.00015759475433147155,
      "loss": 0.1896907240152359,
      "mean_token_accuracy": 0.9253196120262146,
      "num_tokens": 56250104.0,
      "step": 4560
    },
    {
      "entropy": 1.3006483912467957,
      "epoch": 2.4017904160084256,
      "grad_norm": 0.28648555278778076,
      "learning_rate": 0.0001575761417806102,
      "loss": 0.1915302276611328,
      "mean_token_accuracy": 0.921184629201889,
      "num_tokens": 56262440.0,
      "step": 4561
    },
    {
      "entropy": 1.3277715742588043,
      "epoch": 2.40231700895208,
      "grad_norm": 0.26527178287506104,
      "learning_rate": 0.00015755752640543068,
      "loss": 0.17917628586292267,
      "mean_token_accuracy": 0.9229338020086288,
      "num_tokens": 56274776.0,
      "step": 4562
    },
    {
      "entropy": 1.3829319775104523,
      "epoch": 2.4028436018957344,
      "grad_norm": 0.2874324321746826,
      "learning_rate": 0.00015753890820703796,
      "loss": 0.1934787929058075,
      "mean_token_accuracy": 0.9214570224285126,
      "num_tokens": 56287112.0,
      "step": 4563
    },
    {
      "entropy": 1.3280232846736908,
      "epoch": 2.4033701948393893,
      "grad_norm": 0.24855974316596985,
      "learning_rate": 0.00015752028718653735,
      "loss": 0.19005198776721954,
      "mean_token_accuracy": 0.9208149313926697,
      "num_tokens": 56299448.0,
      "step": 4564
    },
    {
      "entropy": 1.3538621664047241,
      "epoch": 2.4038967877830437,
      "grad_norm": 0.2776485085487366,
      "learning_rate": 0.00015750166334503428,
      "loss": 0.19672241806983948,
      "mean_token_accuracy": 0.9181719124317169,
      "num_tokens": 56311784.0,
      "step": 4565
    },
    {
      "entropy": 1.368317574262619,
      "epoch": 2.404423380726698,
      "grad_norm": 0.30098727345466614,
      "learning_rate": 0.00015748303668363428,
      "loss": 0.20025476813316345,
      "mean_token_accuracy": 0.916120171546936,
      "num_tokens": 56324120.0,
      "step": 4566
    },
    {
      "entropy": 1.3438273072242737,
      "epoch": 2.4049499736703526,
      "grad_norm": 0.25104859471321106,
      "learning_rate": 0.00015746440720344317,
      "loss": 0.18177089095115662,
      "mean_token_accuracy": 0.9239592254161835,
      "num_tokens": 56336456.0,
      "step": 4567
    },
    {
      "entropy": 1.3766308426856995,
      "epoch": 2.4054765666140074,
      "grad_norm": 0.27121055126190186,
      "learning_rate": 0.00015744577490556683,
      "loss": 0.2015809267759323,
      "mean_token_accuracy": 0.9173483699560165,
      "num_tokens": 56348792.0,
      "step": 4568
    },
    {
      "entropy": 1.4611666798591614,
      "epoch": 2.406003159557662,
      "grad_norm": 0.3079581558704376,
      "learning_rate": 0.00015742713979111135,
      "loss": 0.19990572333335876,
      "mean_token_accuracy": 0.9156835675239563,
      "num_tokens": 56361128.0,
      "step": 4569
    },
    {
      "entropy": 1.457626760005951,
      "epoch": 2.4065297525013163,
      "grad_norm": 0.2969047725200653,
      "learning_rate": 0.00015740850186118306,
      "loss": 0.2033241093158722,
      "mean_token_accuracy": 0.92415751516819,
      "num_tokens": 56373464.0,
      "step": 4570
    },
    {
      "entropy": 1.4298239052295685,
      "epoch": 2.407056345444971,
      "grad_norm": 0.31620898842811584,
      "learning_rate": 0.00015738986111688832,
      "loss": 0.22344595193862915,
      "mean_token_accuracy": 0.9087575227022171,
      "num_tokens": 56385800.0,
      "step": 4571
    },
    {
      "entropy": 1.4056819081306458,
      "epoch": 2.4075829383886256,
      "grad_norm": 0.27822446823120117,
      "learning_rate": 0.0001573712175593337,
      "loss": 0.19859078526496887,
      "mean_token_accuracy": 0.923060268163681,
      "num_tokens": 56398136.0,
      "step": 4572
    },
    {
      "entropy": 1.3622363805770874,
      "epoch": 2.40810953133228,
      "grad_norm": 0.2607809603214264,
      "learning_rate": 0.00015735257118962604,
      "loss": 0.1764621138572693,
      "mean_token_accuracy": 0.9253451228141785,
      "num_tokens": 56410472.0,
      "step": 4573
    },
    {
      "entropy": 1.4228957891464233,
      "epoch": 2.408636124275935,
      "grad_norm": 0.2639370560646057,
      "learning_rate": 0.0001573339220088722,
      "loss": 0.19621466100215912,
      "mean_token_accuracy": 0.920293316245079,
      "num_tokens": 56422808.0,
      "step": 4574
    },
    {
      "entropy": 1.353748857975006,
      "epoch": 2.4091627172195893,
      "grad_norm": 0.2694760859012604,
      "learning_rate": 0.00015731527001817928,
      "loss": 0.18804813921451569,
      "mean_token_accuracy": 0.9210447818040848,
      "num_tokens": 56435144.0,
      "step": 4575
    },
    {
      "entropy": 1.3881179988384247,
      "epoch": 2.4096893101632437,
      "grad_norm": 0.2707253694534302,
      "learning_rate": 0.00015729661521865452,
      "loss": 0.18320326507091522,
      "mean_token_accuracy": 0.9270624965429306,
      "num_tokens": 56447480.0,
      "step": 4576
    },
    {
      "entropy": 1.3343448638916016,
      "epoch": 2.4102159031068986,
      "grad_norm": 0.27668213844299316,
      "learning_rate": 0.00015727795761140545,
      "loss": 0.1835266351699829,
      "mean_token_accuracy": 0.9257790744304657,
      "num_tokens": 56459816.0,
      "step": 4577
    },
    {
      "entropy": 1.321186900138855,
      "epoch": 2.410742496050553,
      "grad_norm": 0.27669277787208557,
      "learning_rate": 0.00015725929719753956,
      "loss": 0.18208348751068115,
      "mean_token_accuracy": 0.9222432971000671,
      "num_tokens": 56472152.0,
      "step": 4578
    },
    {
      "entropy": 1.3364296555519104,
      "epoch": 2.4112690889942074,
      "grad_norm": 0.28166496753692627,
      "learning_rate": 0.00015724063397816468,
      "loss": 0.1879100501537323,
      "mean_token_accuracy": 0.919008269906044,
      "num_tokens": 56484488.0,
      "step": 4579
    },
    {
      "entropy": 1.3278461694717407,
      "epoch": 2.411795681937862,
      "grad_norm": 0.27643710374832153,
      "learning_rate": 0.00015722196795438866,
      "loss": 0.18816009163856506,
      "mean_token_accuracy": 0.9239133149385452,
      "num_tokens": 56496824.0,
      "step": 4580
    },
    {
      "entropy": 1.2958170771598816,
      "epoch": 2.4123222748815167,
      "grad_norm": 0.27953317761421204,
      "learning_rate": 0.0001572032991273196,
      "loss": 0.17952555418014526,
      "mean_token_accuracy": 0.9238559603691101,
      "num_tokens": 56509160.0,
      "step": 4581
    },
    {
      "entropy": 1.3013927638530731,
      "epoch": 2.412848867825171,
      "grad_norm": 0.2748692035675049,
      "learning_rate": 0.00015718462749806587,
      "loss": 0.19492384791374207,
      "mean_token_accuracy": 0.9214185774326324,
      "num_tokens": 56521496.0,
      "step": 4582
    },
    {
      "entropy": 1.3403103947639465,
      "epoch": 2.4133754607688256,
      "grad_norm": 0.30991989374160767,
      "learning_rate": 0.00015716595306773578,
      "loss": 0.19997966289520264,
      "mean_token_accuracy": 0.9153329730033875,
      "num_tokens": 56533832.0,
      "step": 4583
    },
    {
      "entropy": 1.3335986733436584,
      "epoch": 2.41390205371248,
      "grad_norm": 0.29624128341674805,
      "learning_rate": 0.00015714727583743797,
      "loss": 0.2043485939502716,
      "mean_token_accuracy": 0.9172103554010391,
      "num_tokens": 56546168.0,
      "step": 4584
    },
    {
      "entropy": 1.310105711221695,
      "epoch": 2.414428646656135,
      "grad_norm": 0.2946057915687561,
      "learning_rate": 0.00015712859580828116,
      "loss": 0.20172706246376038,
      "mean_token_accuracy": 0.9106387346982956,
      "num_tokens": 56558504.0,
      "step": 4585
    },
    {
      "entropy": 1.3054699003696442,
      "epoch": 2.4149552395997893,
      "grad_norm": 0.26526352763175964,
      "learning_rate": 0.00015710991298137434,
      "loss": 0.1864761859178543,
      "mean_token_accuracy": 0.9250095337629318,
      "num_tokens": 56570840.0,
      "step": 4586
    },
    {
      "entropy": 1.3284521996974945,
      "epoch": 2.4154818325434437,
      "grad_norm": 0.26133716106414795,
      "learning_rate": 0.00015709122735782654,
      "loss": 0.17758479714393616,
      "mean_token_accuracy": 0.9264996349811554,
      "num_tokens": 56583176.0,
      "step": 4587
    },
    {
      "entropy": 1.2835633158683777,
      "epoch": 2.4160084254870986,
      "grad_norm": 0.2573343515396118,
      "learning_rate": 0.00015707253893874705,
      "loss": 0.17470473051071167,
      "mean_token_accuracy": 0.9288149029016495,
      "num_tokens": 56595512.0,
      "step": 4588
    },
    {
      "entropy": 1.341663807630539,
      "epoch": 2.416535018430753,
      "grad_norm": 0.28183892369270325,
      "learning_rate": 0.0001570538477252453,
      "loss": 0.18871092796325684,
      "mean_token_accuracy": 0.923402264714241,
      "num_tokens": 56607848.0,
      "step": 4589
    },
    {
      "entropy": 1.3551491498947144,
      "epoch": 2.4170616113744074,
      "grad_norm": 0.2750586271286011,
      "learning_rate": 0.0001570351537184308,
      "loss": 0.19159851968288422,
      "mean_token_accuracy": 0.9253100901842117,
      "num_tokens": 56620184.0,
      "step": 4590
    },
    {
      "entropy": 1.361526906490326,
      "epoch": 2.4175882043180623,
      "grad_norm": 0.28902900218963623,
      "learning_rate": 0.0001570164569194134,
      "loss": 0.20224696397781372,
      "mean_token_accuracy": 0.9204160273075104,
      "num_tokens": 56632520.0,
      "step": 4591
    },
    {
      "entropy": 1.3157592415809631,
      "epoch": 2.4181147972617167,
      "grad_norm": 0.2721328139305115,
      "learning_rate": 0.00015699775732930303,
      "loss": 0.20679977536201477,
      "mean_token_accuracy": 0.9149771928787231,
      "num_tokens": 56644856.0,
      "step": 4592
    },
    {
      "entropy": 1.3480108678340912,
      "epoch": 2.418641390205371,
      "grad_norm": 0.24639494717121124,
      "learning_rate": 0.00015697905494920967,
      "loss": 0.18107910454273224,
      "mean_token_accuracy": 0.9280508905649185,
      "num_tokens": 56657192.0,
      "step": 4593
    },
    {
      "entropy": 1.340236097574234,
      "epoch": 2.419167983149026,
      "grad_norm": 0.2795374393463135,
      "learning_rate": 0.00015696034978024368,
      "loss": 0.20761287212371826,
      "mean_token_accuracy": 0.9111906141042709,
      "num_tokens": 56669528.0,
      "step": 4594
    },
    {
      "entropy": 1.3402322232723236,
      "epoch": 2.4196945760926805,
      "grad_norm": 0.2711165249347687,
      "learning_rate": 0.00015694164182351538,
      "loss": 0.1791497766971588,
      "mean_token_accuracy": 0.9271443635225296,
      "num_tokens": 56681864.0,
      "step": 4595
    },
    {
      "entropy": 1.3855849206447601,
      "epoch": 2.420221169036335,
      "grad_norm": 0.2867280840873718,
      "learning_rate": 0.00015692293108013545,
      "loss": 0.19872815907001495,
      "mean_token_accuracy": 0.9179540723562241,
      "num_tokens": 56694200.0,
      "step": 4596
    },
    {
      "entropy": 1.3175525963306427,
      "epoch": 2.4207477619799893,
      "grad_norm": 0.27587011456489563,
      "learning_rate": 0.00015690421755121457,
      "loss": 0.18959006667137146,
      "mean_token_accuracy": 0.9211609661579132,
      "num_tokens": 56706536.0,
      "step": 4597
    },
    {
      "entropy": 1.3644810616970062,
      "epoch": 2.421274354923644,
      "grad_norm": 0.2764270305633545,
      "learning_rate": 0.00015688550123786366,
      "loss": 0.1940041482448578,
      "mean_token_accuracy": 0.9237237423658371,
      "num_tokens": 56718872.0,
      "step": 4598
    },
    {
      "entropy": 1.3711154758930206,
      "epoch": 2.4218009478672986,
      "grad_norm": 0.2801766097545624,
      "learning_rate": 0.00015686678214119382,
      "loss": 0.1970905065536499,
      "mean_token_accuracy": 0.9176339656114578,
      "num_tokens": 56731208.0,
      "step": 4599
    },
    {
      "entropy": 1.3100129067897797,
      "epoch": 2.422327540810953,
      "grad_norm": 0.2844012975692749,
      "learning_rate": 0.0001568480602623163,
      "loss": 0.1999197155237198,
      "mean_token_accuracy": 0.9149242490530014,
      "num_tokens": 56743544.0,
      "step": 4600
    },
    {
      "entropy": 1.3215259611606598,
      "epoch": 2.4228541337546075,
      "grad_norm": 0.2659996747970581,
      "learning_rate": 0.00015682933560234248,
      "loss": 0.1908705085515976,
      "mean_token_accuracy": 0.9223310500383377,
      "num_tokens": 56755880.0,
      "step": 4601
    },
    {
      "entropy": 1.3315446376800537,
      "epoch": 2.4233807266982623,
      "grad_norm": 0.28083324432373047,
      "learning_rate": 0.00015681060816238394,
      "loss": 0.1834309995174408,
      "mean_token_accuracy": 0.9284676909446716,
      "num_tokens": 56768216.0,
      "step": 4602
    },
    {
      "entropy": 1.2885978817939758,
      "epoch": 2.4239073196419167,
      "grad_norm": 0.2809126675128937,
      "learning_rate": 0.00015679187794355242,
      "loss": 0.1866403967142105,
      "mean_token_accuracy": 0.9212284684181213,
      "num_tokens": 56780552.0,
      "step": 4603
    },
    {
      "entropy": 1.2973859310150146,
      "epoch": 2.424433912585571,
      "grad_norm": 0.2610211670398712,
      "learning_rate": 0.00015677314494695984,
      "loss": 0.17383259534835815,
      "mean_token_accuracy": 0.927782341837883,
      "num_tokens": 56792888.0,
      "step": 4604
    },
    {
      "entropy": 1.3476527631282806,
      "epoch": 2.424960505529226,
      "grad_norm": 0.3418804705142975,
      "learning_rate": 0.00015675440917371826,
      "loss": 0.19662517309188843,
      "mean_token_accuracy": 0.9165080189704895,
      "num_tokens": 56805224.0,
      "step": 4605
    },
    {
      "entropy": 1.3011980652809143,
      "epoch": 2.4254870984728805,
      "grad_norm": 0.2649618983268738,
      "learning_rate": 0.00015673567062493993,
      "loss": 0.1870417296886444,
      "mean_token_accuracy": 0.9290430843830109,
      "num_tokens": 56817560.0,
      "step": 4606
    },
    {
      "entropy": 1.3310189843177795,
      "epoch": 2.426013691416535,
      "grad_norm": 0.2768966257572174,
      "learning_rate": 0.00015671692930173723,
      "loss": 0.2032088190317154,
      "mean_token_accuracy": 0.9112291485071182,
      "num_tokens": 56829896.0,
      "step": 4607
    },
    {
      "entropy": 1.3227407336235046,
      "epoch": 2.4265402843601898,
      "grad_norm": 0.2667352259159088,
      "learning_rate": 0.0001566981852052227,
      "loss": 0.17368647456169128,
      "mean_token_accuracy": 0.9290298819541931,
      "num_tokens": 56842232.0,
      "step": 4608
    },
    {
      "entropy": 1.3284336924552917,
      "epoch": 2.427066877303844,
      "grad_norm": 0.2627871632575989,
      "learning_rate": 0.0001566794383365091,
      "loss": 0.1836368292570114,
      "mean_token_accuracy": 0.9237141758203506,
      "num_tokens": 56854568.0,
      "step": 4609
    },
    {
      "entropy": 1.3730064928531647,
      "epoch": 2.4275934702474986,
      "grad_norm": 0.27761170268058777,
      "learning_rate": 0.00015666068869670932,
      "loss": 0.20169535279273987,
      "mean_token_accuracy": 0.9138482213020325,
      "num_tokens": 56866904.0,
      "step": 4610
    },
    {
      "entropy": 1.3395865857601166,
      "epoch": 2.4281200631911535,
      "grad_norm": 0.3002219796180725,
      "learning_rate": 0.0001566419362869364,
      "loss": 0.2141709327697754,
      "mean_token_accuracy": 0.9107076972723007,
      "num_tokens": 56879240.0,
      "step": 4611
    },
    {
      "entropy": 1.387347549200058,
      "epoch": 2.428646656134808,
      "grad_norm": 0.28495123982429504,
      "learning_rate": 0.00015662318110830356,
      "loss": 0.19449102878570557,
      "mean_token_accuracy": 0.9209498912096024,
      "num_tokens": 56891576.0,
      "step": 4612
    },
    {
      "entropy": 1.3817225992679596,
      "epoch": 2.4291732490784623,
      "grad_norm": 0.2706870138645172,
      "learning_rate": 0.00015660442316192425,
      "loss": 0.18900935351848602,
      "mean_token_accuracy": 0.9174584001302719,
      "num_tokens": 56903912.0,
      "step": 4613
    },
    {
      "entropy": 1.352656215429306,
      "epoch": 2.4296998420221168,
      "grad_norm": 0.2922493815422058,
      "learning_rate": 0.00015658566244891193,
      "loss": 0.19300967454910278,
      "mean_token_accuracy": 0.9183215498924255,
      "num_tokens": 56916248.0,
      "step": 4614
    },
    {
      "entropy": 1.3354830145835876,
      "epoch": 2.4302264349657716,
      "grad_norm": 0.24328167736530304,
      "learning_rate": 0.00015656689897038036,
      "loss": 0.17530909180641174,
      "mean_token_accuracy": 0.9271158427000046,
      "num_tokens": 56928584.0,
      "step": 4615
    },
    {
      "entropy": 1.387416660785675,
      "epoch": 2.430753027909426,
      "grad_norm": 0.2972657382488251,
      "learning_rate": 0.0001565481327274434,
      "loss": 0.1807890385389328,
      "mean_token_accuracy": 0.9254136830568314,
      "num_tokens": 56940920.0,
      "step": 4616
    },
    {
      "entropy": 1.4199165403842926,
      "epoch": 2.4312796208530805,
      "grad_norm": 0.2991303503513336,
      "learning_rate": 0.0001565293637212151,
      "loss": 0.20695003867149353,
      "mean_token_accuracy": 0.9172242283821106,
      "num_tokens": 56953256.0,
      "step": 4617
    },
    {
      "entropy": 1.3269080817699432,
      "epoch": 2.431806213796735,
      "grad_norm": 0.26729339361190796,
      "learning_rate": 0.00015651059195280972,
      "loss": 0.19468078017234802,
      "mean_token_accuracy": 0.9203730225563049,
      "num_tokens": 56965592.0,
      "step": 4618
    },
    {
      "entropy": 1.3127374053001404,
      "epoch": 2.4323328067403898,
      "grad_norm": 0.26924213767051697,
      "learning_rate": 0.0001564918174233415,
      "loss": 0.19689175486564636,
      "mean_token_accuracy": 0.9202268719673157,
      "num_tokens": 56977928.0,
      "step": 4619
    },
    {
      "entropy": 1.3146660923957825,
      "epoch": 2.432859399684044,
      "grad_norm": 0.2669108510017395,
      "learning_rate": 0.00015647304013392512,
      "loss": 0.17507515847682953,
      "mean_token_accuracy": 0.9252168834209442,
      "num_tokens": 56990264.0,
      "step": 4620
    },
    {
      "entropy": 1.348682701587677,
      "epoch": 2.4333859926276986,
      "grad_norm": 0.29818832874298096,
      "learning_rate": 0.00015645426008567518,
      "loss": 0.19781912863254547,
      "mean_token_accuracy": 0.9172385036945343,
      "num_tokens": 57002600.0,
      "step": 4621
    },
    {
      "entropy": 1.3332387804985046,
      "epoch": 2.4339125855713535,
      "grad_norm": 0.26270851492881775,
      "learning_rate": 0.0001564354772797066,
      "loss": 0.18425442278385162,
      "mean_token_accuracy": 0.9268904775381088,
      "num_tokens": 57014936.0,
      "step": 4622
    },
    {
      "entropy": 1.342206448316574,
      "epoch": 2.434439178515008,
      "grad_norm": 0.2695198059082031,
      "learning_rate": 0.00015641669171713432,
      "loss": 0.18070070445537567,
      "mean_token_accuracy": 0.9249268621206284,
      "num_tokens": 57027272.0,
      "step": 4623
    },
    {
      "entropy": 1.3906873166561127,
      "epoch": 2.4349657714586623,
      "grad_norm": 0.3107193410396576,
      "learning_rate": 0.0001563979033990737,
      "loss": 0.1972734034061432,
      "mean_token_accuracy": 0.9214301854372025,
      "num_tokens": 57039608.0,
      "step": 4624
    },
    {
      "entropy": 1.4046538770198822,
      "epoch": 2.435492364402317,
      "grad_norm": 0.285004585981369,
      "learning_rate": 0.00015637911232663988,
      "loss": 0.18564899265766144,
      "mean_token_accuracy": 0.9255692511796951,
      "num_tokens": 57051944.0,
      "step": 4625
    },
    {
      "entropy": 1.3951005041599274,
      "epoch": 2.4360189573459716,
      "grad_norm": 0.27796947956085205,
      "learning_rate": 0.0001563603185009485,
      "loss": 0.19178025424480438,
      "mean_token_accuracy": 0.9215326011180878,
      "num_tokens": 57064280.0,
      "step": 4626
    },
    {
      "entropy": 1.3433052897453308,
      "epoch": 2.436545550289626,
      "grad_norm": 0.2718515992164612,
      "learning_rate": 0.00015634152192311524,
      "loss": 0.17776957154273987,
      "mean_token_accuracy": 0.9261139184236526,
      "num_tokens": 57076616.0,
      "step": 4627
    },
    {
      "entropy": 1.4135821163654327,
      "epoch": 2.437072143233281,
      "grad_norm": 0.288688987493515,
      "learning_rate": 0.00015632272259425597,
      "loss": 0.18565590679645538,
      "mean_token_accuracy": 0.9266521334648132,
      "num_tokens": 57088952.0,
      "step": 4628
    },
    {
      "entropy": 1.384842336177826,
      "epoch": 2.4375987361769353,
      "grad_norm": 0.2708379030227661,
      "learning_rate": 0.00015630392051548653,
      "loss": 0.19071045517921448,
      "mean_token_accuracy": 0.9206132292747498,
      "num_tokens": 57101288.0,
      "step": 4629
    },
    {
      "entropy": 1.3972707092761993,
      "epoch": 2.4381253291205898,
      "grad_norm": 0.29080888628959656,
      "learning_rate": 0.0001562851156879233,
      "loss": 0.1984952986240387,
      "mean_token_accuracy": 0.913146585226059,
      "num_tokens": 57113624.0,
      "step": 4630
    },
    {
      "entropy": 1.3927036225795746,
      "epoch": 2.438651922064244,
      "grad_norm": 0.2511928379535675,
      "learning_rate": 0.00015626630811268251,
      "loss": 0.16358304023742676,
      "mean_token_accuracy": 0.9305144250392914,
      "num_tokens": 57125960.0,
      "step": 4631
    },
    {
      "entropy": 1.3519012033939362,
      "epoch": 2.439178515007899,
      "grad_norm": 0.27660346031188965,
      "learning_rate": 0.00015624749779088063,
      "loss": 0.19364917278289795,
      "mean_token_accuracy": 0.9169550985097885,
      "num_tokens": 57138296.0,
      "step": 4632
    },
    {
      "entropy": 1.364226907491684,
      "epoch": 2.4397051079515535,
      "grad_norm": 0.28648045659065247,
      "learning_rate": 0.00015622868472363438,
      "loss": 0.19593121111392975,
      "mean_token_accuracy": 0.9160383343696594,
      "num_tokens": 57150632.0,
      "step": 4633
    },
    {
      "entropy": 1.35571950674057,
      "epoch": 2.440231700895208,
      "grad_norm": 0.2771262526512146,
      "learning_rate": 0.00015620986891206058,
      "loss": 0.19743886590003967,
      "mean_token_accuracy": 0.9158811867237091,
      "num_tokens": 57162968.0,
      "step": 4634
    },
    {
      "entropy": 1.3340860903263092,
      "epoch": 2.4407582938388623,
      "grad_norm": 0.26926276087760925,
      "learning_rate": 0.00015619105035727618,
      "loss": 0.1867595613002777,
      "mean_token_accuracy": 0.9221925735473633,
      "num_tokens": 57175304.0,
      "step": 4635
    },
    {
      "entropy": 1.3448993861675262,
      "epoch": 2.441284886782517,
      "grad_norm": 0.24788318574428558,
      "learning_rate": 0.0001561722290603983,
      "loss": 0.1681167185306549,
      "mean_token_accuracy": 0.9328633397817612,
      "num_tokens": 57187640.0,
      "step": 4636
    },
    {
      "entropy": 1.3575890362262726,
      "epoch": 2.4418114797261716,
      "grad_norm": 0.251751571893692,
      "learning_rate": 0.00015615340502254433,
      "loss": 0.17021462321281433,
      "mean_token_accuracy": 0.9230402410030365,
      "num_tokens": 57199976.0,
      "step": 4637
    },
    {
      "entropy": 1.3393411338329315,
      "epoch": 2.442338072669826,
      "grad_norm": 0.27187666296958923,
      "learning_rate": 0.00015613457824483173,
      "loss": 0.1871224194765091,
      "mean_token_accuracy": 0.9203449487686157,
      "num_tokens": 57212312.0,
      "step": 4638
    },
    {
      "entropy": 1.3779464364051819,
      "epoch": 2.442864665613481,
      "grad_norm": 0.29460328817367554,
      "learning_rate": 0.0001561157487283781,
      "loss": 0.19687990844249725,
      "mean_token_accuracy": 0.9119131118059158,
      "num_tokens": 57224648.0,
      "step": 4639
    },
    {
      "entropy": 1.3972814083099365,
      "epoch": 2.4433912585571353,
      "grad_norm": 0.27819696068763733,
      "learning_rate": 0.00015609691647430123,
      "loss": 0.19143208861351013,
      "mean_token_accuracy": 0.9203731119632721,
      "num_tokens": 57236984.0,
      "step": 4640
    },
    {
      "entropy": 1.325174868106842,
      "epoch": 2.4439178515007898,
      "grad_norm": 0.2797207236289978,
      "learning_rate": 0.00015607808148371908,
      "loss": 0.17681381106376648,
      "mean_token_accuracy": 0.9269442856311798,
      "num_tokens": 57249320.0,
      "step": 4641
    },
    {
      "entropy": 1.3612459897994995,
      "epoch": 2.4444444444444446,
      "grad_norm": 0.3185685873031616,
      "learning_rate": 0.00015605924375774986,
      "loss": 0.22030019760131836,
      "mean_token_accuracy": 0.9069929718971252,
      "num_tokens": 57261656.0,
      "step": 4642
    },
    {
      "entropy": 1.3378702402114868,
      "epoch": 2.444971037388099,
      "grad_norm": 0.2868621349334717,
      "learning_rate": 0.00015604040329751176,
      "loss": 0.1965884119272232,
      "mean_token_accuracy": 0.9148354679346085,
      "num_tokens": 57273992.0,
      "step": 4643
    },
    {
      "entropy": 1.3073699176311493,
      "epoch": 2.4454976303317535,
      "grad_norm": 0.29875674843788147,
      "learning_rate": 0.00015602156010412327,
      "loss": 0.20474374294281006,
      "mean_token_accuracy": 0.9165049344301224,
      "num_tokens": 57286328.0,
      "step": 4644
    },
    {
      "entropy": 1.2536769807338715,
      "epoch": 2.446024223275408,
      "grad_norm": 0.29766038060188293,
      "learning_rate": 0.00015600271417870297,
      "loss": 0.20265424251556396,
      "mean_token_accuracy": 0.9158102571964264,
      "num_tokens": 57298664.0,
      "step": 4645
    },
    {
      "entropy": 1.3608869314193726,
      "epoch": 2.446550816219063,
      "grad_norm": 0.29727551341056824,
      "learning_rate": 0.00015598386552236965,
      "loss": 0.20152640342712402,
      "mean_token_accuracy": 0.9140726178884506,
      "num_tokens": 57311000.0,
      "step": 4646
    },
    {
      "entropy": 1.3182392120361328,
      "epoch": 2.447077409162717,
      "grad_norm": 0.26966792345046997,
      "learning_rate": 0.00015596501413624227,
      "loss": 0.18247759342193604,
      "mean_token_accuracy": 0.9275117516517639,
      "num_tokens": 57323336.0,
      "step": 4647
    },
    {
      "entropy": 1.3680989146232605,
      "epoch": 2.4476040021063716,
      "grad_norm": 0.2620154321193695,
      "learning_rate": 0.0001559461600214399,
      "loss": 0.17146539688110352,
      "mean_token_accuracy": 0.928724393248558,
      "num_tokens": 57335672.0,
      "step": 4648
    },
    {
      "entropy": 1.362658977508545,
      "epoch": 2.4481305950500265,
      "grad_norm": 0.25905415415763855,
      "learning_rate": 0.0001559273031790818,
      "loss": 0.19432124495506287,
      "mean_token_accuracy": 0.9210942685604095,
      "num_tokens": 57348008.0,
      "step": 4649
    },
    {
      "entropy": 1.3837052881717682,
      "epoch": 2.448657187993681,
      "grad_norm": 0.26196369528770447,
      "learning_rate": 0.00015590844361028738,
      "loss": 0.18118348717689514,
      "mean_token_accuracy": 0.9250486940145493,
      "num_tokens": 57360344.0,
      "step": 4650
    },
    {
      "entropy": 1.3374846279621124,
      "epoch": 2.4491837809373354,
      "grad_norm": 0.25008776783943176,
      "learning_rate": 0.00015588958131617628,
      "loss": 0.17867028713226318,
      "mean_token_accuracy": 0.9267919510602951,
      "num_tokens": 57372680.0,
      "step": 4651
    },
    {
      "entropy": 1.4585446417331696,
      "epoch": 2.4497103738809898,
      "grad_norm": 0.30302000045776367,
      "learning_rate": 0.00015587071629786817,
      "loss": 0.19897569715976715,
      "mean_token_accuracy": 0.9165971130132675,
      "num_tokens": 57385016.0,
      "step": 4652
    },
    {
      "entropy": 1.3716599345207214,
      "epoch": 2.4502369668246446,
      "grad_norm": 0.2903025150299072,
      "learning_rate": 0.00015585184855648298,
      "loss": 0.19135521352291107,
      "mean_token_accuracy": 0.9185032695531845,
      "num_tokens": 57397352.0,
      "step": 4653
    },
    {
      "entropy": 1.3391970098018646,
      "epoch": 2.450763559768299,
      "grad_norm": 0.2981782853603363,
      "learning_rate": 0.0001558329780931408,
      "loss": 0.19125044345855713,
      "mean_token_accuracy": 0.9198358207941055,
      "num_tokens": 57409688.0,
      "step": 4654
    },
    {
      "entropy": 1.330566942691803,
      "epoch": 2.4512901527119535,
      "grad_norm": 0.28080058097839355,
      "learning_rate": 0.00015581410490896186,
      "loss": 0.19602055847644806,
      "mean_token_accuracy": 0.9168371856212616,
      "num_tokens": 57422024.0,
      "step": 4655
    },
    {
      "entropy": 1.3519669771194458,
      "epoch": 2.4518167456556084,
      "grad_norm": 0.29043126106262207,
      "learning_rate": 0.0001557952290050665,
      "loss": 0.20128756761550903,
      "mean_token_accuracy": 0.9184352457523346,
      "num_tokens": 57434360.0,
      "step": 4656
    },
    {
      "entropy": 1.3798860013484955,
      "epoch": 2.452343338599263,
      "grad_norm": 0.31429749727249146,
      "learning_rate": 0.00015577635038257532,
      "loss": 0.20553100109100342,
      "mean_token_accuracy": 0.9182298481464386,
      "num_tokens": 57446696.0,
      "step": 4657
    },
    {
      "entropy": 1.3261335492134094,
      "epoch": 2.452869931542917,
      "grad_norm": 0.28215980529785156,
      "learning_rate": 0.00015575746904260903,
      "loss": 0.1838231086730957,
      "mean_token_accuracy": 0.9184658974409103,
      "num_tokens": 57459032.0,
      "step": 4658
    },
    {
      "entropy": 1.3846520483493805,
      "epoch": 2.453396524486572,
      "grad_norm": 0.27965161204338074,
      "learning_rate": 0.00015573858498628851,
      "loss": 0.18266066908836365,
      "mean_token_accuracy": 0.9234927743673325,
      "num_tokens": 57471368.0,
      "step": 4659
    },
    {
      "entropy": 1.356047660112381,
      "epoch": 2.4539231174302265,
      "grad_norm": 0.2756958305835724,
      "learning_rate": 0.0001557196982147348,
      "loss": 0.1809377521276474,
      "mean_token_accuracy": 0.9223103225231171,
      "num_tokens": 57483704.0,
      "step": 4660
    },
    {
      "entropy": 1.3236099183559418,
      "epoch": 2.454449710373881,
      "grad_norm": 0.24765440821647644,
      "learning_rate": 0.00015570080872906903,
      "loss": 0.17653913795948029,
      "mean_token_accuracy": 0.926618218421936,
      "num_tokens": 57496040.0,
      "step": 4661
    },
    {
      "entropy": 1.3622512817382812,
      "epoch": 2.4549763033175354,
      "grad_norm": 0.2710879147052765,
      "learning_rate": 0.00015568191653041267,
      "loss": 0.1775762438774109,
      "mean_token_accuracy": 0.9290056526660919,
      "num_tokens": 57508376.0,
      "step": 4662
    },
    {
      "entropy": 1.4271577596664429,
      "epoch": 2.4555028962611902,
      "grad_norm": 0.3400026261806488,
      "learning_rate": 0.00015566302161988712,
      "loss": 0.21410952508449554,
      "mean_token_accuracy": 0.9081162363290787,
      "num_tokens": 57520712.0,
      "step": 4663
    },
    {
      "entropy": 1.3488382995128632,
      "epoch": 2.4560294892048447,
      "grad_norm": 0.2925370931625366,
      "learning_rate": 0.00015564412399861414,
      "loss": 0.19027042388916016,
      "mean_token_accuracy": 0.9205520153045654,
      "num_tokens": 57533048.0,
      "step": 4664
    },
    {
      "entropy": 1.37377467751503,
      "epoch": 2.456556082148499,
      "grad_norm": 0.29143059253692627,
      "learning_rate": 0.00015562522366771556,
      "loss": 0.2077019214630127,
      "mean_token_accuracy": 0.914614737033844,
      "num_tokens": 57545384.0,
      "step": 4665
    },
    {
      "entropy": 1.3595414757728577,
      "epoch": 2.457082675092154,
      "grad_norm": 0.2915399372577667,
      "learning_rate": 0.00015560632062831337,
      "loss": 0.19326642155647278,
      "mean_token_accuracy": 0.9180510342121124,
      "num_tokens": 57557720.0,
      "step": 4666
    },
    {
      "entropy": 1.3504066467285156,
      "epoch": 2.4576092680358084,
      "grad_norm": 0.2623111605644226,
      "learning_rate": 0.00015558741488152977,
      "loss": 0.19019611179828644,
      "mean_token_accuracy": 0.9178232103586197,
      "num_tokens": 57570056.0,
      "step": 4667
    },
    {
      "entropy": 1.4193410575389862,
      "epoch": 2.458135860979463,
      "grad_norm": 0.28917214274406433,
      "learning_rate": 0.00015556850642848698,
      "loss": 0.1902349442243576,
      "mean_token_accuracy": 0.9237756580114365,
      "num_tokens": 57582392.0,
      "step": 4668
    },
    {
      "entropy": 1.458041101694107,
      "epoch": 2.458662453923117,
      "grad_norm": 0.29611340165138245,
      "learning_rate": 0.00015554959527030763,
      "loss": 0.208420529961586,
      "mean_token_accuracy": 0.917582169175148,
      "num_tokens": 57594728.0,
      "step": 4669
    },
    {
      "entropy": 1.4654216468334198,
      "epoch": 2.459189046866772,
      "grad_norm": 0.3103972375392914,
      "learning_rate": 0.00015553068140811426,
      "loss": 0.19962163269519806,
      "mean_token_accuracy": 0.9165554493665695,
      "num_tokens": 57607064.0,
      "step": 4670
    },
    {
      "entropy": 1.4365287721157074,
      "epoch": 2.4597156398104265,
      "grad_norm": 0.2904442548751831,
      "learning_rate": 0.00015551176484302973,
      "loss": 0.2000875622034073,
      "mean_token_accuracy": 0.9224870204925537,
      "num_tokens": 57619400.0,
      "step": 4671
    },
    {
      "entropy": 1.4503169655799866,
      "epoch": 2.460242232754081,
      "grad_norm": 0.3007279634475708,
      "learning_rate": 0.00015549284557617697,
      "loss": 0.20989835262298584,
      "mean_token_accuracy": 0.9129480719566345,
      "num_tokens": 57631736.0,
      "step": 4672
    },
    {
      "entropy": 1.4487575888633728,
      "epoch": 2.460768825697736,
      "grad_norm": 0.2559841573238373,
      "learning_rate": 0.0001554739236086791,
      "loss": 0.17120632529258728,
      "mean_token_accuracy": 0.9269149452447891,
      "num_tokens": 57644072.0,
      "step": 4673
    },
    {
      "entropy": 1.5271873474121094,
      "epoch": 2.4612954186413902,
      "grad_norm": 0.29322513937950134,
      "learning_rate": 0.00015545499894165945,
      "loss": 0.20046186447143555,
      "mean_token_accuracy": 0.9165253788232803,
      "num_tokens": 57656408.0,
      "step": 4674
    },
    {
      "entropy": 1.4375866651535034,
      "epoch": 2.4618220115850447,
      "grad_norm": 0.24410809576511383,
      "learning_rate": 0.00015543607157624141,
      "loss": 0.18259111046791077,
      "mean_token_accuracy": 0.9227998405694962,
      "num_tokens": 57668744.0,
      "step": 4675
    },
    {
      "entropy": 1.431043803691864,
      "epoch": 2.4623486045286995,
      "grad_norm": 0.2558969259262085,
      "learning_rate": 0.00015541714151354866,
      "loss": 0.1708768606185913,
      "mean_token_accuracy": 0.9271507263183594,
      "num_tokens": 57681080.0,
      "step": 4676
    },
    {
      "entropy": 1.4392517507076263,
      "epoch": 2.462875197472354,
      "grad_norm": 0.27898111939430237,
      "learning_rate": 0.00015539820875470493,
      "loss": 0.18238875269889832,
      "mean_token_accuracy": 0.9190804958343506,
      "num_tokens": 57693416.0,
      "step": 4677
    },
    {
      "entropy": 1.4586999714374542,
      "epoch": 2.4634017904160084,
      "grad_norm": 0.2722666561603546,
      "learning_rate": 0.00015537927330083412,
      "loss": 0.1785045862197876,
      "mean_token_accuracy": 0.9265280216932297,
      "num_tokens": 57705752.0,
      "step": 4678
    },
    {
      "entropy": 1.4363029897212982,
      "epoch": 2.463928383359663,
      "grad_norm": 0.2756902277469635,
      "learning_rate": 0.00015536033515306036,
      "loss": 0.19934740662574768,
      "mean_token_accuracy": 0.9203079342842102,
      "num_tokens": 57718088.0,
      "step": 4679
    },
    {
      "entropy": 1.3940038681030273,
      "epoch": 2.4644549763033177,
      "grad_norm": 0.2740958034992218,
      "learning_rate": 0.00015534139431250789,
      "loss": 0.2017807960510254,
      "mean_token_accuracy": 0.9153221100568771,
      "num_tokens": 57730424.0,
      "step": 4680
    },
    {
      "entropy": 1.4366330802440643,
      "epoch": 2.464981569246972,
      "grad_norm": 0.2849027216434479,
      "learning_rate": 0.0001553224507803011,
      "loss": 0.20108060538768768,
      "mean_token_accuracy": 0.9228730350732803,
      "num_tokens": 57742760.0,
      "step": 4681
    },
    {
      "entropy": 1.4022051692008972,
      "epoch": 2.4655081621906265,
      "grad_norm": 0.25662413239479065,
      "learning_rate": 0.00015530350455756457,
      "loss": 0.1900719702243805,
      "mean_token_accuracy": 0.9239797294139862,
      "num_tokens": 57755096.0,
      "step": 4682
    },
    {
      "entropy": 1.3925521671772003,
      "epoch": 2.4660347551342814,
      "grad_norm": 0.27127277851104736,
      "learning_rate": 0.000155284555645423,
      "loss": 0.19370390474796295,
      "mean_token_accuracy": 0.9192579835653305,
      "num_tokens": 57767432.0,
      "step": 4683
    },
    {
      "entropy": 1.4185691177845001,
      "epoch": 2.466561348077936,
      "grad_norm": 0.25780409574508667,
      "learning_rate": 0.00015526560404500138,
      "loss": 0.1686680167913437,
      "mean_token_accuracy": 0.92674720287323,
      "num_tokens": 57779768.0,
      "step": 4684
    },
    {
      "entropy": 1.4209790229797363,
      "epoch": 2.4670879410215902,
      "grad_norm": 0.2740281820297241,
      "learning_rate": 0.00015524664975742457,
      "loss": 0.17835459113121033,
      "mean_token_accuracy": 0.9256129860877991,
      "num_tokens": 57792104.0,
      "step": 4685
    },
    {
      "entropy": 1.400107204914093,
      "epoch": 2.4676145339652447,
      "grad_norm": 0.2728610634803772,
      "learning_rate": 0.00015522769278381797,
      "loss": 0.19086140394210815,
      "mean_token_accuracy": 0.9191389083862305,
      "num_tokens": 57804440.0,
      "step": 4686
    },
    {
      "entropy": 1.3877354562282562,
      "epoch": 2.4681411269088995,
      "grad_norm": 0.2909792959690094,
      "learning_rate": 0.0001552087331253068,
      "loss": 0.19899430871009827,
      "mean_token_accuracy": 0.9197440594434738,
      "num_tokens": 57816776.0,
      "step": 4687
    },
    {
      "entropy": 1.379683405160904,
      "epoch": 2.468667719852554,
      "grad_norm": 0.2930801808834076,
      "learning_rate": 0.00015518977078301667,
      "loss": 0.19332917034626007,
      "mean_token_accuracy": 0.9176100790500641,
      "num_tokens": 57829112.0,
      "step": 4688
    },
    {
      "entropy": 1.384777694940567,
      "epoch": 2.4691943127962084,
      "grad_norm": 0.24330542981624603,
      "learning_rate": 0.00015517080575807319,
      "loss": 0.1706233024597168,
      "mean_token_accuracy": 0.9322277456521988,
      "num_tokens": 57841448.0,
      "step": 4689
    },
    {
      "entropy": 1.354981243610382,
      "epoch": 2.4697209057398632,
      "grad_norm": 0.2700625956058502,
      "learning_rate": 0.00015515183805160228,
      "loss": 0.17830899357795715,
      "mean_token_accuracy": 0.9246155321598053,
      "num_tokens": 57853784.0,
      "step": 4690
    },
    {
      "entropy": 1.348482221364975,
      "epoch": 2.4702474986835177,
      "grad_norm": 0.29389628767967224,
      "learning_rate": 0.00015513286766472993,
      "loss": 0.20552103221416473,
      "mean_token_accuracy": 0.9085098803043365,
      "num_tokens": 57866120.0,
      "step": 4691
    },
    {
      "entropy": 1.3473348915576935,
      "epoch": 2.470774091627172,
      "grad_norm": 0.291768878698349,
      "learning_rate": 0.00015511389459858224,
      "loss": 0.1977330446243286,
      "mean_token_accuracy": 0.91968734562397,
      "num_tokens": 57878456.0,
      "step": 4692
    },
    {
      "entropy": 1.3101034462451935,
      "epoch": 2.471300684570827,
      "grad_norm": 0.26433873176574707,
      "learning_rate": 0.0001550949188542856,
      "loss": 0.18745142221450806,
      "mean_token_accuracy": 0.9202346354722977,
      "num_tokens": 57890792.0,
      "step": 4693
    },
    {
      "entropy": 1.3141293823719025,
      "epoch": 2.4718272775144814,
      "grad_norm": 0.2735169231891632,
      "learning_rate": 0.00015507594043296646,
      "loss": 0.18631723523139954,
      "mean_token_accuracy": 0.9204375594854355,
      "num_tokens": 57903128.0,
      "step": 4694
    },
    {
      "entropy": 1.3004171252250671,
      "epoch": 2.472353870458136,
      "grad_norm": 0.2703790068626404,
      "learning_rate": 0.00015505695933575145,
      "loss": 0.2004447877407074,
      "mean_token_accuracy": 0.9170452207326889,
      "num_tokens": 57915464.0,
      "step": 4695
    },
    {
      "entropy": 1.3433512449264526,
      "epoch": 2.4728804634017902,
      "grad_norm": 0.31277215480804443,
      "learning_rate": 0.00015503797556376737,
      "loss": 0.21629300713539124,
      "mean_token_accuracy": 0.9137300401926041,
      "num_tokens": 57927800.0,
      "step": 4696
    },
    {
      "entropy": 1.3618537187576294,
      "epoch": 2.473407056345445,
      "grad_norm": 0.27481234073638916,
      "learning_rate": 0.0001550189891181412,
      "loss": 0.19332623481750488,
      "mean_token_accuracy": 0.9189375638961792,
      "num_tokens": 57940136.0,
      "step": 4697
    },
    {
      "entropy": 1.3607601523399353,
      "epoch": 2.4739336492890995,
      "grad_norm": 0.27814021706581116,
      "learning_rate": 0.000155,
      "loss": 0.20003022253513336,
      "mean_token_accuracy": 0.919358566403389,
      "num_tokens": 57952472.0,
      "step": 4698
    },
    {
      "entropy": 1.341829001903534,
      "epoch": 2.474460242232754,
      "grad_norm": 0.29070931673049927,
      "learning_rate": 0.0001549810082104711,
      "loss": 0.20077884197235107,
      "mean_token_accuracy": 0.9115552604198456,
      "num_tokens": 57964808.0,
      "step": 4699
    },
    {
      "entropy": 1.3571709096431732,
      "epoch": 2.474986835176409,
      "grad_norm": 0.2642231285572052,
      "learning_rate": 0.00015496201375068193,
      "loss": 0.18121525645256042,
      "mean_token_accuracy": 0.9253357946872711,
      "num_tokens": 57977144.0,
      "step": 4700
    },
    {
      "entropy": 1.3776450157165527,
      "epoch": 2.4755134281200633,
      "grad_norm": 0.2865309715270996,
      "learning_rate": 0.00015494301662176006,
      "loss": 0.18393152952194214,
      "mean_token_accuracy": 0.9276510328054428,
      "num_tokens": 57989480.0,
      "step": 4701
    },
    {
      "entropy": 1.3750910758972168,
      "epoch": 2.4760400210637177,
      "grad_norm": 0.2893163561820984,
      "learning_rate": 0.00015492401682483324,
      "loss": 0.19169265031814575,
      "mean_token_accuracy": 0.9269598424434662,
      "num_tokens": 58001816.0,
      "step": 4702
    },
    {
      "entropy": 1.343192994594574,
      "epoch": 2.476566614007372,
      "grad_norm": 0.2709070146083832,
      "learning_rate": 0.00015490501436102934,
      "loss": 0.1982916295528412,
      "mean_token_accuracy": 0.9179317057132721,
      "num_tokens": 58014152.0,
      "step": 4703
    },
    {
      "entropy": 1.283618688583374,
      "epoch": 2.477093206951027,
      "grad_norm": 0.26297813653945923,
      "learning_rate": 0.0001548860092314765,
      "loss": 0.18507333099842072,
      "mean_token_accuracy": 0.9239300638437271,
      "num_tokens": 58026488.0,
      "step": 4704
    },
    {
      "entropy": 1.3479060232639313,
      "epoch": 2.4776197998946814,
      "grad_norm": 0.2526850998401642,
      "learning_rate": 0.00015486700143730288,
      "loss": 0.16395074129104614,
      "mean_token_accuracy": 0.9379178881645203,
      "num_tokens": 58038824.0,
      "step": 4705
    },
    {
      "entropy": 1.3700522780418396,
      "epoch": 2.478146392838336,
      "grad_norm": 0.2784835696220398,
      "learning_rate": 0.00015484799097963695,
      "loss": 0.19287003576755524,
      "mean_token_accuracy": 0.918778121471405,
      "num_tokens": 58051160.0,
      "step": 4706
    },
    {
      "entropy": 1.3807547688484192,
      "epoch": 2.4786729857819907,
      "grad_norm": 0.2824748158454895,
      "learning_rate": 0.00015482897785960712,
      "loss": 0.18871180713176727,
      "mean_token_accuracy": 0.9191993176937103,
      "num_tokens": 58063496.0,
      "step": 4707
    },
    {
      "entropy": 1.4022415578365326,
      "epoch": 2.479199578725645,
      "grad_norm": 0.28391286730766296,
      "learning_rate": 0.0001548099620783422,
      "loss": 0.200555682182312,
      "mean_token_accuracy": 0.9131074696779251,
      "num_tokens": 58075832.0,
      "step": 4708
    },
    {
      "entropy": 1.428839921951294,
      "epoch": 2.4797261716692995,
      "grad_norm": 0.2970784604549408,
      "learning_rate": 0.000154790943636971,
      "loss": 0.1906449794769287,
      "mean_token_accuracy": 0.9224237501621246,
      "num_tokens": 58088168.0,
      "step": 4709
    },
    {
      "entropy": 1.3637046813964844,
      "epoch": 2.4802527646129544,
      "grad_norm": 0.25533929467201233,
      "learning_rate": 0.0001547719225366225,
      "loss": 0.1766054928302765,
      "mean_token_accuracy": 0.9246645867824554,
      "num_tokens": 58100504.0,
      "step": 4710
    },
    {
      "entropy": 1.4180608987808228,
      "epoch": 2.480779357556609,
      "grad_norm": 0.2714919447898865,
      "learning_rate": 0.00015475289877842595,
      "loss": 0.183756485581398,
      "mean_token_accuracy": 0.924714669585228,
      "num_tokens": 58112840.0,
      "step": 4711
    },
    {
      "entropy": 1.4052789807319641,
      "epoch": 2.4813059505002633,
      "grad_norm": 0.2869175374507904,
      "learning_rate": 0.00015473387236351064,
      "loss": 0.19138438999652863,
      "mean_token_accuracy": 0.9229259192943573,
      "num_tokens": 58125176.0,
      "step": 4712
    },
    {
      "entropy": 1.495551347732544,
      "epoch": 2.4818325434439177,
      "grad_norm": 0.29526039958000183,
      "learning_rate": 0.000154714843293006,
      "loss": 0.20986592769622803,
      "mean_token_accuracy": 0.9124022424221039,
      "num_tokens": 58137512.0,
      "step": 4713
    },
    {
      "entropy": 1.4451051354408264,
      "epoch": 2.4823591363875726,
      "grad_norm": 0.2789779007434845,
      "learning_rate": 0.0001546958115680418,
      "loss": 0.18474294245243073,
      "mean_token_accuracy": 0.9247586131095886,
      "num_tokens": 58149848.0,
      "step": 4714
    },
    {
      "entropy": 1.3909856081008911,
      "epoch": 2.482885729331227,
      "grad_norm": 0.24263574182987213,
      "learning_rate": 0.00015467677718974775,
      "loss": 0.1780616194009781,
      "mean_token_accuracy": 0.9270612746477127,
      "num_tokens": 58162184.0,
      "step": 4715
    },
    {
      "entropy": 1.4340462684631348,
      "epoch": 2.4834123222748814,
      "grad_norm": 0.2651912569999695,
      "learning_rate": 0.00015465774015925384,
      "loss": 0.17408399283885956,
      "mean_token_accuracy": 0.926845371723175,
      "num_tokens": 58174520.0,
      "step": 4716
    },
    {
      "entropy": 1.4597930312156677,
      "epoch": 2.483938915218536,
      "grad_norm": 0.28243258595466614,
      "learning_rate": 0.00015463870047769017,
      "loss": 0.18010583519935608,
      "mean_token_accuracy": 0.9231376051902771,
      "num_tokens": 58186856.0,
      "step": 4717
    },
    {
      "entropy": 1.4875306189060211,
      "epoch": 2.4844655081621907,
      "grad_norm": 0.29506340622901917,
      "learning_rate": 0.000154619658146187,
      "loss": 0.18998458981513977,
      "mean_token_accuracy": 0.9228047430515289,
      "num_tokens": 58199192.0,
      "step": 4718
    },
    {
      "entropy": 1.4901301264762878,
      "epoch": 2.484992101105845,
      "grad_norm": 0.2923277020454407,
      "learning_rate": 0.00015460061316587484,
      "loss": 0.2033998817205429,
      "mean_token_accuracy": 0.9161169081926346,
      "num_tokens": 58211528.0,
      "step": 4719
    },
    {
      "entropy": 1.498829424381256,
      "epoch": 2.4855186940494995,
      "grad_norm": 0.3002236783504486,
      "learning_rate": 0.00015458156553788423,
      "loss": 0.21175146102905273,
      "mean_token_accuracy": 0.9184870421886444,
      "num_tokens": 58223864.0,
      "step": 4720
    },
    {
      "entropy": 1.413221299648285,
      "epoch": 2.4860452869931544,
      "grad_norm": 0.28396493196487427,
      "learning_rate": 0.0001545625152633459,
      "loss": 0.18868695199489594,
      "mean_token_accuracy": 0.9188051074743271,
      "num_tokens": 58236200.0,
      "step": 4721
    },
    {
      "entropy": 1.4876404404640198,
      "epoch": 2.486571879936809,
      "grad_norm": 0.27693501114845276,
      "learning_rate": 0.00015454346234339075,
      "loss": 0.18164798617362976,
      "mean_token_accuracy": 0.9216231107711792,
      "num_tokens": 58248536.0,
      "step": 4722
    },
    {
      "entropy": 1.4857076406478882,
      "epoch": 2.4870984728804633,
      "grad_norm": 0.28282952308654785,
      "learning_rate": 0.00015452440677914987,
      "loss": 0.19472423195838928,
      "mean_token_accuracy": 0.9207596182823181,
      "num_tokens": 58260872.0,
      "step": 4723
    },
    {
      "entropy": 1.5304179787635803,
      "epoch": 2.487625065824118,
      "grad_norm": 0.2851126790046692,
      "learning_rate": 0.00015450534857175448,
      "loss": 0.1930224597454071,
      "mean_token_accuracy": 0.9197488129138947,
      "num_tokens": 58273208.0,
      "step": 4724
    },
    {
      "entropy": 1.5518591701984406,
      "epoch": 2.4881516587677726,
      "grad_norm": 0.33592379093170166,
      "learning_rate": 0.0001544862877223359,
      "loss": 0.19371357560157776,
      "mean_token_accuracy": 0.9196053594350815,
      "num_tokens": 58285544.0,
      "step": 4725
    },
    {
      "entropy": 1.4344633221626282,
      "epoch": 2.488678251711427,
      "grad_norm": 0.2809014618396759,
      "learning_rate": 0.00015446722423202575,
      "loss": 0.1901252120733261,
      "mean_token_accuracy": 0.9172061532735825,
      "num_tokens": 58297880.0,
      "step": 4726
    },
    {
      "entropy": 1.4716311693191528,
      "epoch": 2.489204844655082,
      "grad_norm": 0.2870580554008484,
      "learning_rate": 0.00015444815810195566,
      "loss": 0.19343703985214233,
      "mean_token_accuracy": 0.9194012731313705,
      "num_tokens": 58310216.0,
      "step": 4727
    },
    {
      "entropy": 1.4924903213977814,
      "epoch": 2.4897314375987363,
      "grad_norm": 0.2923792004585266,
      "learning_rate": 0.0001544290893332575,
      "loss": 0.1825990229845047,
      "mean_token_accuracy": 0.9255896955728531,
      "num_tokens": 58322552.0,
      "step": 4728
    },
    {
      "entropy": 1.4978864192962646,
      "epoch": 2.4902580305423907,
      "grad_norm": 0.2822360098361969,
      "learning_rate": 0.00015441001792706322,
      "loss": 0.1829996407032013,
      "mean_token_accuracy": 0.923396423459053,
      "num_tokens": 58334888.0,
      "step": 4729
    },
    {
      "entropy": 1.4193677008152008,
      "epoch": 2.490784623486045,
      "grad_norm": 0.2578467130661011,
      "learning_rate": 0.00015439094388450502,
      "loss": 0.170360267162323,
      "mean_token_accuracy": 0.9318064004182816,
      "num_tokens": 58347224.0,
      "step": 4730
    },
    {
      "entropy": 1.4174853563308716,
      "epoch": 2.4913112164297,
      "grad_norm": 0.26211613416671753,
      "learning_rate": 0.00015437186720671523,
      "loss": 0.18280640244483948,
      "mean_token_accuracy": 0.9214925467967987,
      "num_tokens": 58359560.0,
      "step": 4731
    },
    {
      "entropy": 1.4184439182281494,
      "epoch": 2.4918378093733544,
      "grad_norm": 0.27308353781700134,
      "learning_rate": 0.00015435278789482636,
      "loss": 0.18611527979373932,
      "mean_token_accuracy": 0.9204042702913284,
      "num_tokens": 58371896.0,
      "step": 4732
    },
    {
      "entropy": 1.442617654800415,
      "epoch": 2.492364402317009,
      "grad_norm": 0.2780870199203491,
      "learning_rate": 0.00015433370594997089,
      "loss": 0.1830921471118927,
      "mean_token_accuracy": 0.9229092448949814,
      "num_tokens": 58384232.0,
      "step": 4733
    },
    {
      "entropy": 1.484441488981247,
      "epoch": 2.4928909952606633,
      "grad_norm": 0.27742764353752136,
      "learning_rate": 0.00015431462137328173,
      "loss": 0.1889549344778061,
      "mean_token_accuracy": 0.9233603626489639,
      "num_tokens": 58396568.0,
      "step": 4734
    },
    {
      "entropy": 1.4144461452960968,
      "epoch": 2.493417588204318,
      "grad_norm": 0.29658958315849304,
      "learning_rate": 0.00015429553416589176,
      "loss": 0.19394293427467346,
      "mean_token_accuracy": 0.9221424907445908,
      "num_tokens": 58408904.0,
      "step": 4735
    },
    {
      "entropy": 1.3806745409965515,
      "epoch": 2.4939441811479726,
      "grad_norm": 0.27006301283836365,
      "learning_rate": 0.00015427644432893414,
      "loss": 0.19147875905036926,
      "mean_token_accuracy": 0.9202659428119659,
      "num_tokens": 58421240.0,
      "step": 4736
    },
    {
      "entropy": 1.4355372488498688,
      "epoch": 2.494470774091627,
      "grad_norm": 0.2959778308868408,
      "learning_rate": 0.00015425735186354208,
      "loss": 0.18910780549049377,
      "mean_token_accuracy": 0.920879676938057,
      "num_tokens": 58433576.0,
      "step": 4737
    },
    {
      "entropy": 1.4152382910251617,
      "epoch": 2.494997367035282,
      "grad_norm": 0.2920873165130615,
      "learning_rate": 0.00015423825677084895,
      "loss": 0.196088045835495,
      "mean_token_accuracy": 0.9160876125097275,
      "num_tokens": 58445912.0,
      "step": 4738
    },
    {
      "entropy": 1.4021598100662231,
      "epoch": 2.4955239599789363,
      "grad_norm": 0.2944489121437073,
      "learning_rate": 0.0001542191590519884,
      "loss": 0.19860148429870605,
      "mean_token_accuracy": 0.9189711213111877,
      "num_tokens": 58458248.0,
      "step": 4739
    },
    {
      "entropy": 1.3974989652633667,
      "epoch": 2.4960505529225907,
      "grad_norm": 0.2927993834018707,
      "learning_rate": 0.0001542000587080941,
      "loss": 0.20249119400978088,
      "mean_token_accuracy": 0.9131207466125488,
      "num_tokens": 58470584.0,
      "step": 4740
    },
    {
      "entropy": 1.528170496225357,
      "epoch": 2.4965771458662456,
      "grad_norm": 0.3085220456123352,
      "learning_rate": 0.0001541809557402999,
      "loss": 0.1807674765586853,
      "mean_token_accuracy": 0.9273618459701538,
      "num_tokens": 58482920.0,
      "step": 4741
    },
    {
      "entropy": 1.367598444223404,
      "epoch": 2.4971037388099,
      "grad_norm": 0.2692739963531494,
      "learning_rate": 0.00015416185014973993,
      "loss": 0.19078630208969116,
      "mean_token_accuracy": 0.919693797826767,
      "num_tokens": 58495256.0,
      "step": 4742
    },
    {
      "entropy": 1.339257150888443,
      "epoch": 2.4976303317535544,
      "grad_norm": 0.24560058116912842,
      "learning_rate": 0.00015414274193754823,
      "loss": 0.1724538803100586,
      "mean_token_accuracy": 0.9302109032869339,
      "num_tokens": 58507592.0,
      "step": 4743
    },
    {
      "entropy": 1.411946326494217,
      "epoch": 2.4981569246972093,
      "grad_norm": 0.27875998616218567,
      "learning_rate": 0.00015412363110485928,
      "loss": 0.19445690512657166,
      "mean_token_accuracy": 0.9199328869581223,
      "num_tokens": 58519928.0,
      "step": 4744
    },
    {
      "entropy": 1.4017539620399475,
      "epoch": 2.4986835176408637,
      "grad_norm": 0.2869739234447479,
      "learning_rate": 0.0001541045176528075,
      "loss": 0.18097423017024994,
      "mean_token_accuracy": 0.9258893132209778,
      "num_tokens": 58532264.0,
      "step": 4745
    },
    {
      "entropy": 1.3831762373447418,
      "epoch": 2.499210110584518,
      "grad_norm": 0.30489039421081543,
      "learning_rate": 0.00015408540158252755,
      "loss": 0.1928139328956604,
      "mean_token_accuracy": 0.9191577881574631,
      "num_tokens": 58544600.0,
      "step": 4746
    },
    {
      "entropy": 1.4570479691028595,
      "epoch": 2.4997367035281726,
      "grad_norm": 0.2680588960647583,
      "learning_rate": 0.0001540662828951543,
      "loss": 0.17343565821647644,
      "mean_token_accuracy": 0.9286295920610428,
      "num_tokens": 58556936.0,
      "step": 4747
    },
    {
      "entropy": 1.482690542936325,
      "epoch": 2.5002632964718274,
      "grad_norm": 0.27270379662513733,
      "learning_rate": 0.00015404716159182264,
      "loss": 0.1857714056968689,
      "mean_token_accuracy": 0.9244716465473175,
      "num_tokens": 58569272.0,
      "step": 4748
    },
    {
      "entropy": 1.4173051118850708,
      "epoch": 2.500789889415482,
      "grad_norm": 0.26899290084838867,
      "learning_rate": 0.00015402803767366774,
      "loss": 0.19327764213085175,
      "mean_token_accuracy": 0.9222336858510971,
      "num_tokens": 58581608.0,
      "step": 4749
    },
    {
      "entropy": 1.3975452482700348,
      "epoch": 2.5013164823591363,
      "grad_norm": 0.2648375332355499,
      "learning_rate": 0.00015400891114182488,
      "loss": 0.1719164401292801,
      "mean_token_accuracy": 0.9263694137334824,
      "num_tokens": 58593944.0,
      "step": 4750
    },
    {
      "entropy": 1.4633611142635345,
      "epoch": 2.5018430753027907,
      "grad_norm": 0.27619731426239014,
      "learning_rate": 0.00015398978199742945,
      "loss": 0.19618642330169678,
      "mean_token_accuracy": 0.9181656688451767,
      "num_tokens": 58606280.0,
      "step": 4751
    },
    {
      "entropy": 1.4815970957279205,
      "epoch": 2.5023696682464456,
      "grad_norm": 0.2575641870498657,
      "learning_rate": 0.000153970650241617,
      "loss": 0.1765824407339096,
      "mean_token_accuracy": 0.9256032109260559,
      "num_tokens": 58618616.0,
      "step": 4752
    },
    {
      "entropy": 1.4368881583213806,
      "epoch": 2.5028962611901,
      "grad_norm": 0.260724276304245,
      "learning_rate": 0.0001539515158755234,
      "loss": 0.1661488264799118,
      "mean_token_accuracy": 0.9317044764757156,
      "num_tokens": 58630952.0,
      "step": 4753
    },
    {
      "entropy": 1.537843406200409,
      "epoch": 2.5034228541337544,
      "grad_norm": 0.27667954564094543,
      "learning_rate": 0.00015393237890028443,
      "loss": 0.16878753900527954,
      "mean_token_accuracy": 0.929436519742012,
      "num_tokens": 58643288.0,
      "step": 4754
    },
    {
      "entropy": 1.4854961335659027,
      "epoch": 2.5039494470774093,
      "grad_norm": 0.2952048182487488,
      "learning_rate": 0.00015391323931703617,
      "loss": 0.17188966274261475,
      "mean_token_accuracy": 0.9301230907440186,
      "num_tokens": 58655624.0,
      "step": 4755
    },
    {
      "entropy": 1.5180222392082214,
      "epoch": 2.5044760400210637,
      "grad_norm": 0.3058265745639801,
      "learning_rate": 0.0001538940971269149,
      "loss": 0.19406625628471375,
      "mean_token_accuracy": 0.91983662545681,
      "num_tokens": 58667960.0,
      "step": 4756
    },
    {
      "entropy": 1.5174492299556732,
      "epoch": 2.505002632964718,
      "grad_norm": 0.2982122004032135,
      "learning_rate": 0.0001538749523310569,
      "loss": 0.19521798193454742,
      "mean_token_accuracy": 0.9169194102287292,
      "num_tokens": 58680296.0,
      "step": 4757
    },
    {
      "entropy": 1.5078306794166565,
      "epoch": 2.505529225908373,
      "grad_norm": 0.33034297823905945,
      "learning_rate": 0.00015385580493059865,
      "loss": 0.17905089259147644,
      "mean_token_accuracy": 0.9300853908061981,
      "num_tokens": 58692632.0,
      "step": 4758
    },
    {
      "entropy": 1.5533065497875214,
      "epoch": 2.5060558188520274,
      "grad_norm": 0.3211785554885864,
      "learning_rate": 0.0001538366549266769,
      "loss": 0.2184678316116333,
      "mean_token_accuracy": 0.912133514881134,
      "num_tokens": 58704968.0,
      "step": 4759
    },
    {
      "entropy": 1.5568303167819977,
      "epoch": 2.506582411795682,
      "grad_norm": 0.2776426374912262,
      "learning_rate": 0.00015381750232042847,
      "loss": 0.18281778693199158,
      "mean_token_accuracy": 0.925473615527153,
      "num_tokens": 58717304.0,
      "step": 4760
    },
    {
      "entropy": 1.5951158702373505,
      "epoch": 2.5071090047393367,
      "grad_norm": 0.28495267033576965,
      "learning_rate": 0.0001537983471129903,
      "loss": 0.19968701899051666,
      "mean_token_accuracy": 0.9174495190382004,
      "num_tokens": 58729640.0,
      "step": 4761
    },
    {
      "entropy": 1.6484262645244598,
      "epoch": 2.507635597682991,
      "grad_norm": 0.278410404920578,
      "learning_rate": 0.00015377918930549952,
      "loss": 0.1968478113412857,
      "mean_token_accuracy": 0.9191997349262238,
      "num_tokens": 58741976.0,
      "step": 4762
    },
    {
      "entropy": 1.5548788011074066,
      "epoch": 2.5081621906266456,
      "grad_norm": 0.3318310081958771,
      "learning_rate": 0.00015376002889909345,
      "loss": 0.19419671595096588,
      "mean_token_accuracy": 0.9201514571905136,
      "num_tokens": 58754312.0,
      "step": 4763
    },
    {
      "entropy": 1.5360266864299774,
      "epoch": 2.5086887835703,
      "grad_norm": 0.27847015857696533,
      "learning_rate": 0.00015374086589490951,
      "loss": 0.19744856655597687,
      "mean_token_accuracy": 0.9216972440481186,
      "num_tokens": 58766648.0,
      "step": 4764
    },
    {
      "entropy": 1.5686508417129517,
      "epoch": 2.509215376513955,
      "grad_norm": 0.28824716806411743,
      "learning_rate": 0.00015372170029408534,
      "loss": 0.1984092742204666,
      "mean_token_accuracy": 0.9193175435066223,
      "num_tokens": 58778984.0,
      "step": 4765
    },
    {
      "entropy": 1.5758169889450073,
      "epoch": 2.5097419694576093,
      "grad_norm": 0.2656327486038208,
      "learning_rate": 0.0001537025320977586,
      "loss": 0.1845746636390686,
      "mean_token_accuracy": 0.9232941418886185,
      "num_tokens": 58791320.0,
      "step": 4766
    },
    {
      "entropy": 1.543961524963379,
      "epoch": 2.5102685624012637,
      "grad_norm": 0.27334338426589966,
      "learning_rate": 0.0001536833613070673,
      "loss": 0.18064028024673462,
      "mean_token_accuracy": 0.9252053499221802,
      "num_tokens": 58803656.0,
      "step": 4767
    },
    {
      "entropy": 1.595755398273468,
      "epoch": 2.510795155344918,
      "grad_norm": 0.28184929490089417,
      "learning_rate": 0.00015366418792314937,
      "loss": 0.18395768105983734,
      "mean_token_accuracy": 0.9240316599607468,
      "num_tokens": 58815992.0,
      "step": 4768
    },
    {
      "entropy": 1.529136598110199,
      "epoch": 2.511321748288573,
      "grad_norm": 0.2473924607038498,
      "learning_rate": 0.00015364501194714317,
      "loss": 0.17464369535446167,
      "mean_token_accuracy": 0.9280594140291214,
      "num_tokens": 58828328.0,
      "step": 4769
    },
    {
      "entropy": 1.5212011635303497,
      "epoch": 2.5118483412322274,
      "grad_norm": 0.25740960240364075,
      "learning_rate": 0.00015362583338018694,
      "loss": 0.19115827977657318,
      "mean_token_accuracy": 0.9235165566205978,
      "num_tokens": 58840664.0,
      "step": 4770
    },
    {
      "entropy": 1.5040476322174072,
      "epoch": 2.512374934175882,
      "grad_norm": 0.2637852728366852,
      "learning_rate": 0.0001536066522234193,
      "loss": 0.18358321487903595,
      "mean_token_accuracy": 0.9271368533372879,
      "num_tokens": 58853000.0,
      "step": 4771
    },
    {
      "entropy": 1.4970723390579224,
      "epoch": 2.5129015271195367,
      "grad_norm": 0.26975688338279724,
      "learning_rate": 0.00015358746847797885,
      "loss": 0.19123801589012146,
      "mean_token_accuracy": 0.9218833893537521,
      "num_tokens": 58865336.0,
      "step": 4772
    },
    {
      "entropy": 1.4308944344520569,
      "epoch": 2.513428120063191,
      "grad_norm": 0.26151081919670105,
      "learning_rate": 0.00015356828214500444,
      "loss": 0.18660297989845276,
      "mean_token_accuracy": 0.9219815135002136,
      "num_tokens": 58877672.0,
      "step": 4773
    },
    {
      "entropy": 1.541033297777176,
      "epoch": 2.5139547130068456,
      "grad_norm": 0.30042555928230286,
      "learning_rate": 0.0001535490932256351,
      "loss": 0.20421351492404938,
      "mean_token_accuracy": 0.9185314625501633,
      "num_tokens": 58890008.0,
      "step": 4774
    },
    {
      "entropy": 1.4968733489513397,
      "epoch": 2.5144813059505005,
      "grad_norm": 0.30032971501350403,
      "learning_rate": 0.00015352990172100989,
      "loss": 0.19920584559440613,
      "mean_token_accuracy": 0.9175475388765335,
      "num_tokens": 58902344.0,
      "step": 4775
    },
    {
      "entropy": 1.4867732226848602,
      "epoch": 2.515007898894155,
      "grad_norm": 0.26996806263923645,
      "learning_rate": 0.00015351070763226806,
      "loss": 0.19374677538871765,
      "mean_token_accuracy": 0.9205781370401382,
      "num_tokens": 58914680.0,
      "step": 4776
    },
    {
      "entropy": 1.4824062585830688,
      "epoch": 2.5155344918378093,
      "grad_norm": 0.2542126178741455,
      "learning_rate": 0.0001534915109605492,
      "loss": 0.17349721491336823,
      "mean_token_accuracy": 0.9280759692192078,
      "num_tokens": 58927016.0,
      "step": 4777
    },
    {
      "entropy": 1.4385600686073303,
      "epoch": 2.516061084781464,
      "grad_norm": 0.2770039141178131,
      "learning_rate": 0.0001534723117069928,
      "loss": 0.18407964706420898,
      "mean_token_accuracy": 0.9262531846761703,
      "num_tokens": 58939352.0,
      "step": 4778
    },
    {
      "entropy": 1.3186810910701752,
      "epoch": 2.5165876777251186,
      "grad_norm": 0.2739590108394623,
      "learning_rate": 0.00015345310987273863,
      "loss": 0.18699848651885986,
      "mean_token_accuracy": 0.9226303249597549,
      "num_tokens": 58951688.0,
      "step": 4779
    },
    {
      "entropy": 1.3342479169368744,
      "epoch": 2.517114270668773,
      "grad_norm": 0.2480226755142212,
      "learning_rate": 0.00015343390545892658,
      "loss": 0.17880557477474213,
      "mean_token_accuracy": 0.9291187077760696,
      "num_tokens": 58964024.0,
      "step": 4780
    },
    {
      "entropy": 1.4129751324653625,
      "epoch": 2.5176408636124274,
      "grad_norm": 0.28868386149406433,
      "learning_rate": 0.00015341469846669674,
      "loss": 0.18943078815937042,
      "mean_token_accuracy": 0.9252970963716507,
      "num_tokens": 58976360.0,
      "step": 4781
    },
    {
      "entropy": 1.4371739029884338,
      "epoch": 2.518167456556082,
      "grad_norm": 0.2903843820095062,
      "learning_rate": 0.00015339548889718927,
      "loss": 0.19646281003952026,
      "mean_token_accuracy": 0.9155204445123672,
      "num_tokens": 58988696.0,
      "step": 4782
    },
    {
      "entropy": 1.3534334003925323,
      "epoch": 2.5186940494997367,
      "grad_norm": 0.2689908742904663,
      "learning_rate": 0.00015337627675154452,
      "loss": 0.19126330316066742,
      "mean_token_accuracy": 0.9226963371038437,
      "num_tokens": 59001032.0,
      "step": 4783
    },
    {
      "entropy": 1.403167486190796,
      "epoch": 2.519220642443391,
      "grad_norm": 0.2755464017391205,
      "learning_rate": 0.0001533570620309031,
      "loss": 0.1715218722820282,
      "mean_token_accuracy": 0.9286895543336868,
      "num_tokens": 59013368.0,
      "step": 4784
    },
    {
      "entropy": 1.342715084552765,
      "epoch": 2.5197472353870456,
      "grad_norm": 0.28502821922302246,
      "learning_rate": 0.00015333784473640558,
      "loss": 0.18913885951042175,
      "mean_token_accuracy": 0.9184420257806778,
      "num_tokens": 59025704.0,
      "step": 4785
    },
    {
      "entropy": 1.381771981716156,
      "epoch": 2.5202738283307005,
      "grad_norm": 0.300022691488266,
      "learning_rate": 0.00015331862486919282,
      "loss": 0.20708000659942627,
      "mean_token_accuracy": 0.9179199188947678,
      "num_tokens": 59038040.0,
      "step": 4786
    },
    {
      "entropy": 1.3914389312267303,
      "epoch": 2.520800421274355,
      "grad_norm": 0.32590633630752563,
      "learning_rate": 0.00015329940243040578,
      "loss": 0.20750224590301514,
      "mean_token_accuracy": 0.9141969978809357,
      "num_tokens": 59050376.0,
      "step": 4787
    },
    {
      "entropy": 1.425366759300232,
      "epoch": 2.5213270142180093,
      "grad_norm": 0.2782382071018219,
      "learning_rate": 0.00015328017742118557,
      "loss": 0.19896437227725983,
      "mean_token_accuracy": 0.9168428629636765,
      "num_tokens": 59062712.0,
      "step": 4788
    },
    {
      "entropy": 1.400715321302414,
      "epoch": 2.521853607161664,
      "grad_norm": 0.28630152344703674,
      "learning_rate": 0.00015326094984267352,
      "loss": 0.22256657481193542,
      "mean_token_accuracy": 0.9108149707317352,
      "num_tokens": 59075048.0,
      "step": 4789
    },
    {
      "entropy": 1.4121185541152954,
      "epoch": 2.5223802001053186,
      "grad_norm": 0.29156050086021423,
      "learning_rate": 0.000153241719696011,
      "loss": 0.21156355738639832,
      "mean_token_accuracy": 0.9142804145812988,
      "num_tokens": 59087384.0,
      "step": 4790
    },
    {
      "entropy": 1.4587819874286652,
      "epoch": 2.522906793048973,
      "grad_norm": 0.2809777557849884,
      "learning_rate": 0.00015322248698233958,
      "loss": 0.18381905555725098,
      "mean_token_accuracy": 0.9187909066677094,
      "num_tokens": 59099720.0,
      "step": 4791
    },
    {
      "entropy": 1.4324629604816437,
      "epoch": 2.523433385992628,
      "grad_norm": 0.24410581588745117,
      "learning_rate": 0.00015320325170280107,
      "loss": 0.17199113965034485,
      "mean_token_accuracy": 0.9289192855358124,
      "num_tokens": 59112056.0,
      "step": 4792
    },
    {
      "entropy": 1.457740843296051,
      "epoch": 2.5239599789362823,
      "grad_norm": 0.27388423681259155,
      "learning_rate": 0.00015318401385853729,
      "loss": 0.1924448013305664,
      "mean_token_accuracy": 0.9235370606184006,
      "num_tokens": 59124392.0,
      "step": 4793
    },
    {
      "entropy": 1.4426178336143494,
      "epoch": 2.5244865718799367,
      "grad_norm": 0.2799040973186493,
      "learning_rate": 0.0001531647734506903,
      "loss": 0.20240657031536102,
      "mean_token_accuracy": 0.917314887046814,
      "num_tokens": 59136728.0,
      "step": 4794
    },
    {
      "entropy": 1.4201898574829102,
      "epoch": 2.5250131648235916,
      "grad_norm": 0.29636329412460327,
      "learning_rate": 0.00015314553048040232,
      "loss": 0.20854808390140533,
      "mean_token_accuracy": 0.9121217727661133,
      "num_tokens": 59149064.0,
      "step": 4795
    },
    {
      "entropy": 1.4267604053020477,
      "epoch": 2.525539757767246,
      "grad_norm": 0.26949402689933777,
      "learning_rate": 0.00015312628494881567,
      "loss": 0.18854168057441711,
      "mean_token_accuracy": 0.9252474457025528,
      "num_tokens": 59161400.0,
      "step": 4796
    },
    {
      "entropy": 1.4317746460437775,
      "epoch": 2.5260663507109005,
      "grad_norm": 0.2725363075733185,
      "learning_rate": 0.0001531070368570728,
      "loss": 0.1946963369846344,
      "mean_token_accuracy": 0.9178613424301147,
      "num_tokens": 59173736.0,
      "step": 4797
    },
    {
      "entropy": 1.4510256946086884,
      "epoch": 2.526592943654555,
      "grad_norm": 0.25904926657676697,
      "learning_rate": 0.00015308778620631643,
      "loss": 0.17193448543548584,
      "mean_token_accuracy": 0.9315731972455978,
      "num_tokens": 59186072.0,
      "step": 4798
    },
    {
      "entropy": 1.4164240956306458,
      "epoch": 2.5271195365982093,
      "grad_norm": 0.25664958357810974,
      "learning_rate": 0.00015306853299768935,
      "loss": 0.1800077110528946,
      "mean_token_accuracy": 0.9235340803861618,
      "num_tokens": 59198408.0,
      "step": 4799
    },
    {
      "entropy": 1.4181910455226898,
      "epoch": 2.527646129541864,
      "grad_norm": 0.267265260219574,
      "learning_rate": 0.00015304927723233444,
      "loss": 0.1814480721950531,
      "mean_token_accuracy": 0.926010325551033,
      "num_tokens": 59210744.0,
      "step": 4800
    },
    {
      "entropy": 1.403830498456955,
      "epoch": 2.5281727224855186,
      "grad_norm": 0.2695743143558502,
      "learning_rate": 0.00015303001891139485,
      "loss": 0.18746566772460938,
      "mean_token_accuracy": 0.922560304403305,
      "num_tokens": 59223080.0,
      "step": 4801
    },
    {
      "entropy": 1.4577591121196747,
      "epoch": 2.528699315429173,
      "grad_norm": 0.33329254388809204,
      "learning_rate": 0.00015301075803601387,
      "loss": 0.21829648315906525,
      "mean_token_accuracy": 0.9158814549446106,
      "num_tokens": 59235416.0,
      "step": 4802
    },
    {
      "entropy": 1.4333778023719788,
      "epoch": 2.529225908372828,
      "grad_norm": 0.2889585793018341,
      "learning_rate": 0.00015299149460733485,
      "loss": 0.19638335704803467,
      "mean_token_accuracy": 0.9257490932941437,
      "num_tokens": 59247752.0,
      "step": 4803
    },
    {
      "entropy": 1.375129371881485,
      "epoch": 2.5297525013164823,
      "grad_norm": 0.27963852882385254,
      "learning_rate": 0.0001529722286265014,
      "loss": 0.19143079221248627,
      "mean_token_accuracy": 0.9210483729839325,
      "num_tokens": 59260088.0,
      "step": 4804
    },
    {
      "entropy": 1.4024271965026855,
      "epoch": 2.5302790942601368,
      "grad_norm": 0.28402864933013916,
      "learning_rate": 0.00015295296009465716,
      "loss": 0.19460776448249817,
      "mean_token_accuracy": 0.9193561524152756,
      "num_tokens": 59272424.0,
      "step": 4805
    },
    {
      "entropy": 1.3990450501441956,
      "epoch": 2.5308056872037916,
      "grad_norm": 0.26199251413345337,
      "learning_rate": 0.00015293368901294604,
      "loss": 0.18480481207370758,
      "mean_token_accuracy": 0.9206849783658981,
      "num_tokens": 59284760.0,
      "step": 4806
    },
    {
      "entropy": 1.4049314558506012,
      "epoch": 2.531332280147446,
      "grad_norm": 0.2740122377872467,
      "learning_rate": 0.00015291441538251206,
      "loss": 0.1854100227355957,
      "mean_token_accuracy": 0.9238648563623428,
      "num_tokens": 59297096.0,
      "step": 4807
    },
    {
      "entropy": 1.4527244567871094,
      "epoch": 2.5318588730911005,
      "grad_norm": 0.28150686621665955,
      "learning_rate": 0.00015289513920449933,
      "loss": 0.19049590826034546,
      "mean_token_accuracy": 0.9227531105279922,
      "num_tokens": 59309432.0,
      "step": 4808
    },
    {
      "entropy": 1.429905265569687,
      "epoch": 2.5323854660347553,
      "grad_norm": 0.2997985780239105,
      "learning_rate": 0.00015287586048005222,
      "loss": 0.2127116471529007,
      "mean_token_accuracy": 0.910615935921669,
      "num_tokens": 59321768.0,
      "step": 4809
    },
    {
      "entropy": 1.361231505870819,
      "epoch": 2.5329120589784098,
      "grad_norm": 0.24818584322929382,
      "learning_rate": 0.00015285657921031514,
      "loss": 0.1716461330652237,
      "mean_token_accuracy": 0.9304056763648987,
      "num_tokens": 59334104.0,
      "step": 4810
    },
    {
      "entropy": 1.4534060060977936,
      "epoch": 2.533438651922064,
      "grad_norm": 0.25663110613822937,
      "learning_rate": 0.00015283729539643278,
      "loss": 0.1782129853963852,
      "mean_token_accuracy": 0.9268085807561874,
      "num_tokens": 59346440.0,
      "step": 4811
    },
    {
      "entropy": 1.453408420085907,
      "epoch": 2.533965244865719,
      "grad_norm": 0.2804957330226898,
      "learning_rate": 0.00015281800903954986,
      "loss": 0.19750183820724487,
      "mean_token_accuracy": 0.9193479269742966,
      "num_tokens": 59358776.0,
      "step": 4812
    },
    {
      "entropy": 1.5167824625968933,
      "epoch": 2.5344918378093735,
      "grad_norm": 0.3135470747947693,
      "learning_rate": 0.00015279872014081132,
      "loss": 0.21170248091220856,
      "mean_token_accuracy": 0.9191084504127502,
      "num_tokens": 59371112.0,
      "step": 4813
    },
    {
      "entropy": 1.4512979388237,
      "epoch": 2.535018430753028,
      "grad_norm": 0.27713891863822937,
      "learning_rate": 0.00015277942870136217,
      "loss": 0.18782149255275726,
      "mean_token_accuracy": 0.9250089675188065,
      "num_tokens": 59383448.0,
      "step": 4814
    },
    {
      "entropy": 1.387693852186203,
      "epoch": 2.5355450236966823,
      "grad_norm": 0.29428794980049133,
      "learning_rate": 0.00015276013472234767,
      "loss": 0.19037064909934998,
      "mean_token_accuracy": 0.9212422370910645,
      "num_tokens": 59395784.0,
      "step": 4815
    },
    {
      "entropy": 1.4471594989299774,
      "epoch": 2.5360716166403368,
      "grad_norm": 0.29530593752861023,
      "learning_rate": 0.00015274083820491325,
      "loss": 0.20669062435626984,
      "mean_token_accuracy": 0.9147179573774338,
      "num_tokens": 59408120.0,
      "step": 4816
    },
    {
      "entropy": 1.4373044073581696,
      "epoch": 2.5365982095839916,
      "grad_norm": 0.28317534923553467,
      "learning_rate": 0.00015272153915020436,
      "loss": 0.20463967323303223,
      "mean_token_accuracy": 0.9127626866102219,
      "num_tokens": 59420456.0,
      "step": 4817
    },
    {
      "entropy": 1.4383168518543243,
      "epoch": 2.537124802527646,
      "grad_norm": 0.26127228140830994,
      "learning_rate": 0.00015270223755936668,
      "loss": 0.20316585898399353,
      "mean_token_accuracy": 0.918964073061943,
      "num_tokens": 59432792.0,
      "step": 4818
    },
    {
      "entropy": 1.441445916891098,
      "epoch": 2.5376513954713005,
      "grad_norm": 0.26680976152420044,
      "learning_rate": 0.00015268293343354602,
      "loss": 0.1995728462934494,
      "mean_token_accuracy": 0.916199654340744,
      "num_tokens": 59445128.0,
      "step": 4819
    },
    {
      "entropy": 1.4771413207054138,
      "epoch": 2.5381779884149553,
      "grad_norm": 0.27789434790611267,
      "learning_rate": 0.00015266362677388844,
      "loss": 0.20266149938106537,
      "mean_token_accuracy": 0.9153758585453033,
      "num_tokens": 59457464.0,
      "step": 4820
    },
    {
      "entropy": 1.4663703739643097,
      "epoch": 2.5387045813586098,
      "grad_norm": 0.2747149169445038,
      "learning_rate": 0.00015264431758153993,
      "loss": 0.19435656070709229,
      "mean_token_accuracy": 0.9212976843118668,
      "num_tokens": 59469800.0,
      "step": 4821
    },
    {
      "entropy": 1.4226221442222595,
      "epoch": 2.539231174302264,
      "grad_norm": 0.2740062177181244,
      "learning_rate": 0.00015262500585764687,
      "loss": 0.18962638080120087,
      "mean_token_accuracy": 0.9208030998706818,
      "num_tokens": 59482136.0,
      "step": 4822
    },
    {
      "entropy": 1.4463150799274445,
      "epoch": 2.539757767245919,
      "grad_norm": 0.26949939131736755,
      "learning_rate": 0.00015260569160335566,
      "loss": 0.18399539589881897,
      "mean_token_accuracy": 0.9234223514795303,
      "num_tokens": 59494472.0,
      "step": 4823
    },
    {
      "entropy": 1.4091249704360962,
      "epoch": 2.5402843601895735,
      "grad_norm": 0.30532780289649963,
      "learning_rate": 0.00015258637481981285,
      "loss": 0.19731716811656952,
      "mean_token_accuracy": 0.9179344773292542,
      "num_tokens": 59506808.0,
      "step": 4824
    },
    {
      "entropy": 1.4171604216098785,
      "epoch": 2.540810953133228,
      "grad_norm": 0.29440897703170776,
      "learning_rate": 0.00015256705550816517,
      "loss": 0.195593923330307,
      "mean_token_accuracy": 0.9183009415864944,
      "num_tokens": 59519144.0,
      "step": 4825
    },
    {
      "entropy": 1.4110483527183533,
      "epoch": 2.541337546076883,
      "grad_norm": 0.2890157103538513,
      "learning_rate": 0.00015254773366955951,
      "loss": 0.1955215185880661,
      "mean_token_accuracy": 0.9175896495580673,
      "num_tokens": 59531480.0,
      "step": 4826
    },
    {
      "entropy": 1.430400162935257,
      "epoch": 2.541864139020537,
      "grad_norm": 0.2826859951019287,
      "learning_rate": 0.00015252840930514292,
      "loss": 0.176889106631279,
      "mean_token_accuracy": 0.9274906814098358,
      "num_tokens": 59543816.0,
      "step": 4827
    },
    {
      "entropy": 1.374840259552002,
      "epoch": 2.5423907319641916,
      "grad_norm": 0.2729499638080597,
      "learning_rate": 0.00015250908241606253,
      "loss": 0.18692106008529663,
      "mean_token_accuracy": 0.9284721314907074,
      "num_tokens": 59556152.0,
      "step": 4828
    },
    {
      "entropy": 1.382160872220993,
      "epoch": 2.5429173249078465,
      "grad_norm": 0.2653583288192749,
      "learning_rate": 0.00015248975300346566,
      "loss": 0.18451890349388123,
      "mean_token_accuracy": 0.9224373549222946,
      "num_tokens": 59568488.0,
      "step": 4829
    },
    {
      "entropy": 1.3657977879047394,
      "epoch": 2.543443917851501,
      "grad_norm": 0.24867606163024902,
      "learning_rate": 0.0001524704210684998,
      "loss": 0.1776690036058426,
      "mean_token_accuracy": 0.9281052350997925,
      "num_tokens": 59580824.0,
      "step": 4830
    },
    {
      "entropy": 1.3682537972927094,
      "epoch": 2.5439705107951553,
      "grad_norm": 0.27566197514533997,
      "learning_rate": 0.0001524510866123126,
      "loss": 0.19942131638526917,
      "mean_token_accuracy": 0.9162020832300186,
      "num_tokens": 59593160.0,
      "step": 4831
    },
    {
      "entropy": 1.3574018776416779,
      "epoch": 2.5444971037388098,
      "grad_norm": 0.2771880328655243,
      "learning_rate": 0.0001524317496360518,
      "loss": 0.19043181836605072,
      "mean_token_accuracy": 0.9205653816461563,
      "num_tokens": 59605496.0,
      "step": 4832
    },
    {
      "entropy": 1.36737522482872,
      "epoch": 2.545023696682464,
      "grad_norm": 0.2834945023059845,
      "learning_rate": 0.00015241241014086535,
      "loss": 0.19852997362613678,
      "mean_token_accuracy": 0.9176250100135803,
      "num_tokens": 59617832.0,
      "step": 4833
    },
    {
      "entropy": 1.358602523803711,
      "epoch": 2.545550289626119,
      "grad_norm": 0.26471948623657227,
      "learning_rate": 0.00015239306812790129,
      "loss": 0.18846866488456726,
      "mean_token_accuracy": 0.9247909784317017,
      "num_tokens": 59630168.0,
      "step": 4834
    },
    {
      "entropy": 1.3623977303504944,
      "epoch": 2.5460768825697735,
      "grad_norm": 0.27146539092063904,
      "learning_rate": 0.00015237372359830787,
      "loss": 0.19878418743610382,
      "mean_token_accuracy": 0.9201604127883911,
      "num_tokens": 59642504.0,
      "step": 4835
    },
    {
      "entropy": 1.4069927632808685,
      "epoch": 2.546603475513428,
      "grad_norm": 0.2823067903518677,
      "learning_rate": 0.00015235437655323344,
      "loss": 0.20891442894935608,
      "mean_token_accuracy": 0.9150956869125366,
      "num_tokens": 59654840.0,
      "step": 4836
    },
    {
      "entropy": 1.3921907544136047,
      "epoch": 2.547130068457083,
      "grad_norm": 0.2729669511318207,
      "learning_rate": 0.00015233502699382652,
      "loss": 0.20379957556724548,
      "mean_token_accuracy": 0.9180693328380585,
      "num_tokens": 59667176.0,
      "step": 4837
    },
    {
      "entropy": 1.3551888167858124,
      "epoch": 2.547656661400737,
      "grad_norm": 0.263839989900589,
      "learning_rate": 0.00015231567492123585,
      "loss": 0.1928175985813141,
      "mean_token_accuracy": 0.9204489141702652,
      "num_tokens": 59679512.0,
      "step": 4838
    },
    {
      "entropy": 1.4220162332057953,
      "epoch": 2.5481832543443916,
      "grad_norm": 0.28904277086257935,
      "learning_rate": 0.00015229632033661014,
      "loss": 0.1956554502248764,
      "mean_token_accuracy": 0.9192171394824982,
      "num_tokens": 59691848.0,
      "step": 4839
    },
    {
      "entropy": 1.384028673171997,
      "epoch": 2.5487098472880465,
      "grad_norm": 0.2730739414691925,
      "learning_rate": 0.00015227696324109845,
      "loss": 0.1821494847536087,
      "mean_token_accuracy": 0.9268871694803238,
      "num_tokens": 59704184.0,
      "step": 4840
    },
    {
      "entropy": 1.3748473227024078,
      "epoch": 2.549236440231701,
      "grad_norm": 0.2834646999835968,
      "learning_rate": 0.00015225760363584982,
      "loss": 0.2006215900182724,
      "mean_token_accuracy": 0.9206278324127197,
      "num_tokens": 59716520.0,
      "step": 4841
    },
    {
      "entropy": 1.3900345861911774,
      "epoch": 2.5497630331753554,
      "grad_norm": 0.26609018445014954,
      "learning_rate": 0.0001522382415220136,
      "loss": 0.1950165331363678,
      "mean_token_accuracy": 0.9225917160511017,
      "num_tokens": 59728856.0,
      "step": 4842
    },
    {
      "entropy": 1.3536244630813599,
      "epoch": 2.5502896261190102,
      "grad_norm": 0.273309588432312,
      "learning_rate": 0.00015221887690073914,
      "loss": 0.19759991765022278,
      "mean_token_accuracy": 0.922698363661766,
      "num_tokens": 59741192.0,
      "step": 4843
    },
    {
      "entropy": 1.33824023604393,
      "epoch": 2.5508162190626646,
      "grad_norm": 0.2553465664386749,
      "learning_rate": 0.00015219950977317603,
      "loss": 0.175911545753479,
      "mean_token_accuracy": 0.9264104813337326,
      "num_tokens": 59753528.0,
      "step": 4844
    },
    {
      "entropy": 1.402260184288025,
      "epoch": 2.551342812006319,
      "grad_norm": 0.3060164749622345,
      "learning_rate": 0.00015218014014047398,
      "loss": 0.20706036686897278,
      "mean_token_accuracy": 0.912144809961319,
      "num_tokens": 59765864.0,
      "step": 4845
    },
    {
      "entropy": 1.3535306751728058,
      "epoch": 2.551869404949974,
      "grad_norm": 0.28057345747947693,
      "learning_rate": 0.00015216076800378286,
      "loss": 0.18130075931549072,
      "mean_token_accuracy": 0.9202318787574768,
      "num_tokens": 59778200.0,
      "step": 4846
    },
    {
      "entropy": 1.3663321137428284,
      "epoch": 2.5523959978936284,
      "grad_norm": 0.2771863639354706,
      "learning_rate": 0.00015214139336425272,
      "loss": 0.19252638518810272,
      "mean_token_accuracy": 0.9169972240924835,
      "num_tokens": 59790536.0,
      "step": 4847
    },
    {
      "entropy": 1.380006194114685,
      "epoch": 2.552922590837283,
      "grad_norm": 0.2713053822517395,
      "learning_rate": 0.0001521220162230336,
      "loss": 0.17066720128059387,
      "mean_token_accuracy": 0.9294691681861877,
      "num_tokens": 59802872.0,
      "step": 4848
    },
    {
      "entropy": 1.3988358080387115,
      "epoch": 2.553449183780937,
      "grad_norm": 0.3080752491950989,
      "learning_rate": 0.00015210263658127593,
      "loss": 0.19805192947387695,
      "mean_token_accuracy": 0.9194646626710892,
      "num_tokens": 59815208.0,
      "step": 4849
    },
    {
      "entropy": 1.3025390207767487,
      "epoch": 2.5539757767245916,
      "grad_norm": 0.26315048336982727,
      "learning_rate": 0.00015208325444013009,
      "loss": 0.18763092160224915,
      "mean_token_accuracy": 0.921664372086525,
      "num_tokens": 59827544.0,
      "step": 4850
    },
    {
      "entropy": 1.363812655210495,
      "epoch": 2.5545023696682465,
      "grad_norm": 0.28132131695747375,
      "learning_rate": 0.00015206386980074675,
      "loss": 0.20107382535934448,
      "mean_token_accuracy": 0.9173763394355774,
      "num_tokens": 59839880.0,
      "step": 4851
    },
    {
      "entropy": 1.3637959957122803,
      "epoch": 2.555028962611901,
      "grad_norm": 0.2793247699737549,
      "learning_rate": 0.0001520444826642766,
      "loss": 0.18369603157043457,
      "mean_token_accuracy": 0.9214000850915909,
      "num_tokens": 59852216.0,
      "step": 4852
    },
    {
      "entropy": 1.407531052827835,
      "epoch": 2.5555555555555554,
      "grad_norm": 0.27685731649398804,
      "learning_rate": 0.00015202509303187056,
      "loss": 0.19428351521492004,
      "mean_token_accuracy": 0.9237850308418274,
      "num_tokens": 59864552.0,
      "step": 4853
    },
    {
      "entropy": 1.4391858875751495,
      "epoch": 2.5560821484992102,
      "grad_norm": 0.30099570751190186,
      "learning_rate": 0.0001520057009046797,
      "loss": 0.20823752880096436,
      "mean_token_accuracy": 0.9157876074314117,
      "num_tokens": 59876888.0,
      "step": 4854
    },
    {
      "entropy": 1.4529668688774109,
      "epoch": 2.5566087414428647,
      "grad_norm": 0.3118361234664917,
      "learning_rate": 0.0001519863062838552,
      "loss": 0.2140086144208908,
      "mean_token_accuracy": 0.9144342094659805,
      "num_tokens": 59889224.0,
      "step": 4855
    },
    {
      "entropy": 1.4054121375083923,
      "epoch": 2.557135334386519,
      "grad_norm": 0.2828349471092224,
      "learning_rate": 0.00015196690917054843,
      "loss": 0.18467508256435394,
      "mean_token_accuracy": 0.9245531558990479,
      "num_tokens": 59901560.0,
      "step": 4856
    },
    {
      "entropy": 1.370518833398819,
      "epoch": 2.557661927330174,
      "grad_norm": 0.3127945363521576,
      "learning_rate": 0.00015194750956591086,
      "loss": 0.1978442370891571,
      "mean_token_accuracy": 0.9195757210254669,
      "num_tokens": 59913896.0,
      "step": 4857
    },
    {
      "entropy": 1.3630948066711426,
      "epoch": 2.5581885202738284,
      "grad_norm": 0.238580584526062,
      "learning_rate": 0.00015192810747109413,
      "loss": 0.15261900424957275,
      "mean_token_accuracy": 0.9331966936588287,
      "num_tokens": 59926232.0,
      "step": 4858
    },
    {
      "entropy": 1.411979854106903,
      "epoch": 2.558715113217483,
      "grad_norm": 0.2636284828186035,
      "learning_rate": 0.00015190870288725,
      "loss": 0.17438194155693054,
      "mean_token_accuracy": 0.9271792620420456,
      "num_tokens": 59938568.0,
      "step": 4859
    },
    {
      "entropy": 1.4212656915187836,
      "epoch": 2.5592417061611377,
      "grad_norm": 0.3053138554096222,
      "learning_rate": 0.0001518892958155305,
      "loss": 0.19817118346691132,
      "mean_token_accuracy": 0.9181718081235886,
      "num_tokens": 59950904.0,
      "step": 4860
    },
    {
      "entropy": 1.349739819765091,
      "epoch": 2.559768299104792,
      "grad_norm": 0.24464385211467743,
      "learning_rate": 0.00015186988625708763,
      "loss": 0.1783566176891327,
      "mean_token_accuracy": 0.9276772141456604,
      "num_tokens": 59963240.0,
      "step": 4861
    },
    {
      "entropy": 1.3938109576702118,
      "epoch": 2.5602948920484465,
      "grad_norm": 0.2769966423511505,
      "learning_rate": 0.00015185047421307363,
      "loss": 0.2047450840473175,
      "mean_token_accuracy": 0.9157468527555466,
      "num_tokens": 59975576.0,
      "step": 4862
    },
    {
      "entropy": 1.3291508555412292,
      "epoch": 2.5608214849921014,
      "grad_norm": 0.26295456290245056,
      "learning_rate": 0.00015183105968464094,
      "loss": 0.18852469325065613,
      "mean_token_accuracy": 0.9189134687185287,
      "num_tokens": 59987912.0,
      "step": 4863
    },
    {
      "entropy": 1.3406771123409271,
      "epoch": 2.561348077935756,
      "grad_norm": 0.30215656757354736,
      "learning_rate": 0.000151811642672942,
      "loss": 0.2007867395877838,
      "mean_token_accuracy": 0.9217905700206757,
      "num_tokens": 60000248.0,
      "step": 4864
    },
    {
      "entropy": 1.332020252943039,
      "epoch": 2.5618746708794102,
      "grad_norm": 0.3041894733905792,
      "learning_rate": 0.00015179222317912957,
      "loss": 0.1980377435684204,
      "mean_token_accuracy": 0.9134826958179474,
      "num_tokens": 60012584.0,
      "step": 4865
    },
    {
      "entropy": 1.363236427307129,
      "epoch": 2.5624012638230647,
      "grad_norm": 0.2855970859527588,
      "learning_rate": 0.0001517728012043564,
      "loss": 0.18393242359161377,
      "mean_token_accuracy": 0.920888140797615,
      "num_tokens": 60024920.0,
      "step": 4866
    },
    {
      "entropy": 1.3401431441307068,
      "epoch": 2.562927856766719,
      "grad_norm": 0.294961154460907,
      "learning_rate": 0.00015175337674977555,
      "loss": 0.18346351385116577,
      "mean_token_accuracy": 0.9255680441856384,
      "num_tokens": 60037256.0,
      "step": 4867
    },
    {
      "entropy": 1.3249768912792206,
      "epoch": 2.563454449710374,
      "grad_norm": 0.2849559187889099,
      "learning_rate": 0.00015173394981654004,
      "loss": 0.1919642984867096,
      "mean_token_accuracy": 0.9208333492279053,
      "num_tokens": 60049592.0,
      "step": 4868
    },
    {
      "entropy": 1.2834919691085815,
      "epoch": 2.5639810426540284,
      "grad_norm": 0.2871542274951935,
      "learning_rate": 0.00015171452040580317,
      "loss": 0.1890535205602646,
      "mean_token_accuracy": 0.9206035882234573,
      "num_tokens": 60061928.0,
      "step": 4869
    },
    {
      "entropy": 1.3021434247493744,
      "epoch": 2.564507635597683,
      "grad_norm": 0.30245718359947205,
      "learning_rate": 0.00015169508851871835,
      "loss": 0.20414495468139648,
      "mean_token_accuracy": 0.9124367386102676,
      "num_tokens": 60074264.0,
      "step": 4870
    },
    {
      "entropy": 1.2731509506702423,
      "epoch": 2.5650342285413377,
      "grad_norm": 0.2902166247367859,
      "learning_rate": 0.00015167565415643913,
      "loss": 0.20379520952701569,
      "mean_token_accuracy": 0.9124204516410828,
      "num_tokens": 60086600.0,
      "step": 4871
    },
    {
      "entropy": 1.3015616834163666,
      "epoch": 2.565560821484992,
      "grad_norm": 0.28301534056663513,
      "learning_rate": 0.00015165621732011924,
      "loss": 0.19441409409046173,
      "mean_token_accuracy": 0.9212340712547302,
      "num_tokens": 60098936.0,
      "step": 4872
    },
    {
      "entropy": 1.2701435387134552,
      "epoch": 2.5660874144286465,
      "grad_norm": 0.2904719114303589,
      "learning_rate": 0.00015163677801091255,
      "loss": 0.19993525743484497,
      "mean_token_accuracy": 0.9170213043689728,
      "num_tokens": 60111272.0,
      "step": 4873
    },
    {
      "entropy": 1.2749419212341309,
      "epoch": 2.5666140073723014,
      "grad_norm": 0.2766469419002533,
      "learning_rate": 0.00015161733622997298,
      "loss": 0.18929515779018402,
      "mean_token_accuracy": 0.9223896861076355,
      "num_tokens": 60123608.0,
      "step": 4874
    },
    {
      "entropy": 1.2422993183135986,
      "epoch": 2.567140600315956,
      "grad_norm": 0.27800917625427246,
      "learning_rate": 0.00015159789197845475,
      "loss": 0.1969931721687317,
      "mean_token_accuracy": 0.9148046523332596,
      "num_tokens": 60135944.0,
      "step": 4875
    },
    {
      "entropy": 1.3092963099479675,
      "epoch": 2.5676671932596102,
      "grad_norm": 0.29029369354248047,
      "learning_rate": 0.00015157844525751213,
      "loss": 0.1947655975818634,
      "mean_token_accuracy": 0.9203184247016907,
      "num_tokens": 60148280.0,
      "step": 4876
    },
    {
      "entropy": 1.2354808747768402,
      "epoch": 2.568193786203265,
      "grad_norm": 0.2631555497646332,
      "learning_rate": 0.00015155899606829955,
      "loss": 0.19616174697875977,
      "mean_token_accuracy": 0.9179852455854416,
      "num_tokens": 60160616.0,
      "step": 4877
    },
    {
      "entropy": 1.193561613559723,
      "epoch": 2.5687203791469195,
      "grad_norm": 0.2713335156440735,
      "learning_rate": 0.00015153954441197157,
      "loss": 0.18860536813735962,
      "mean_token_accuracy": 0.9259924739599228,
      "num_tokens": 60172952.0,
      "step": 4878
    },
    {
      "entropy": 1.221862941980362,
      "epoch": 2.569246972090574,
      "grad_norm": 0.2639184594154358,
      "learning_rate": 0.000151520090289683,
      "loss": 0.18559566140174866,
      "mean_token_accuracy": 0.9201700389385223,
      "num_tokens": 60185288.0,
      "step": 4879
    },
    {
      "entropy": 1.2744978070259094,
      "epoch": 2.5697735650342284,
      "grad_norm": 0.2902701497077942,
      "learning_rate": 0.00015150063370258864,
      "loss": 0.19379426538944244,
      "mean_token_accuracy": 0.9194291532039642,
      "num_tokens": 60197624.0,
      "step": 4880
    },
    {
      "entropy": 1.2000911235809326,
      "epoch": 2.5703001579778832,
      "grad_norm": 0.25564298033714294,
      "learning_rate": 0.00015148117465184352,
      "loss": 0.17944049835205078,
      "mean_token_accuracy": 0.9247357249259949,
      "num_tokens": 60209960.0,
      "step": 4881
    },
    {
      "entropy": 1.210292786359787,
      "epoch": 2.5708267509215377,
      "grad_norm": 0.26203998923301697,
      "learning_rate": 0.00015146171313860284,
      "loss": 0.18620462715625763,
      "mean_token_accuracy": 0.9194813221693039,
      "num_tokens": 60222296.0,
      "step": 4882
    },
    {
      "entropy": 1.2274521887302399,
      "epoch": 2.571353343865192,
      "grad_norm": 0.27803394198417664,
      "learning_rate": 0.00015144224916402194,
      "loss": 0.19611826539039612,
      "mean_token_accuracy": 0.9189580529928207,
      "num_tokens": 60234632.0,
      "step": 4883
    },
    {
      "entropy": 1.285475254058838,
      "epoch": 2.5718799368088465,
      "grad_norm": 0.3042823374271393,
      "learning_rate": 0.0001514227827292562,
      "loss": 0.20818758010864258,
      "mean_token_accuracy": 0.9117998331785202,
      "num_tokens": 60246968.0,
      "step": 4884
    },
    {
      "entropy": 1.2546075582504272,
      "epoch": 2.5724065297525014,
      "grad_norm": 0.27190738916397095,
      "learning_rate": 0.00015140331383546132,
      "loss": 0.1968504935503006,
      "mean_token_accuracy": 0.917721763253212,
      "num_tokens": 60259304.0,
      "step": 4885
    },
    {
      "entropy": 1.3057755529880524,
      "epoch": 2.572933122696156,
      "grad_norm": 0.27564769983291626,
      "learning_rate": 0.000151383842483793,
      "loss": 0.19694563746452332,
      "mean_token_accuracy": 0.9217658787965775,
      "num_tokens": 60271640.0,
      "step": 4886
    },
    {
      "entropy": 1.2702763080596924,
      "epoch": 2.5734597156398102,
      "grad_norm": 0.24457713961601257,
      "learning_rate": 0.00015136436867540718,
      "loss": 0.1783868968486786,
      "mean_token_accuracy": 0.9248514473438263,
      "num_tokens": 60283976.0,
      "step": 4887
    },
    {
      "entropy": 1.2917823791503906,
      "epoch": 2.573986308583465,
      "grad_norm": 0.30155834555625916,
      "learning_rate": 0.00015134489241145984,
      "loss": 0.211724653840065,
      "mean_token_accuracy": 0.9160776287317276,
      "num_tokens": 60296312.0,
      "step": 4888
    },
    {
      "entropy": 1.2568625509738922,
      "epoch": 2.5745129015271195,
      "grad_norm": 0.26823991537094116,
      "learning_rate": 0.00015132541369310726,
      "loss": 0.18977221846580505,
      "mean_token_accuracy": 0.9210988581180573,
      "num_tokens": 60308648.0,
      "step": 4889
    },
    {
      "entropy": 1.2876782417297363,
      "epoch": 2.575039494470774,
      "grad_norm": 0.26660630106925964,
      "learning_rate": 0.00015130593252150567,
      "loss": 0.1923052966594696,
      "mean_token_accuracy": 0.9208141714334488,
      "num_tokens": 60320984.0,
      "step": 4890
    },
    {
      "entropy": 1.2527808248996735,
      "epoch": 2.575566087414429,
      "grad_norm": 0.2633249759674072,
      "learning_rate": 0.00015128644889781165,
      "loss": 0.17889824509620667,
      "mean_token_accuracy": 0.9229971915483475,
      "num_tokens": 60333320.0,
      "step": 4891
    },
    {
      "entropy": 1.2528367638587952,
      "epoch": 2.5760926803580833,
      "grad_norm": 0.26531174778938293,
      "learning_rate": 0.0001512669628231818,
      "loss": 0.17675469815731049,
      "mean_token_accuracy": 0.9215357750654221,
      "num_tokens": 60345656.0,
      "step": 4892
    },
    {
      "entropy": 1.2803305685520172,
      "epoch": 2.5766192733017377,
      "grad_norm": 0.2765933573246002,
      "learning_rate": 0.00015124747429877292,
      "loss": 0.1918942928314209,
      "mean_token_accuracy": 0.9219794422388077,
      "num_tokens": 60357992.0,
      "step": 4893
    },
    {
      "entropy": 1.335657685995102,
      "epoch": 2.5771458662453925,
      "grad_norm": 0.31621235609054565,
      "learning_rate": 0.00015122798332574183,
      "loss": 0.1919841319322586,
      "mean_token_accuracy": 0.9153444766998291,
      "num_tokens": 60370328.0,
      "step": 4894
    },
    {
      "entropy": 1.2984404861927032,
      "epoch": 2.577672459189047,
      "grad_norm": 0.29850491881370544,
      "learning_rate": 0.00015120848990524572,
      "loss": 0.1919029951095581,
      "mean_token_accuracy": 0.9215899854898453,
      "num_tokens": 60382664.0,
      "step": 4895
    },
    {
      "entropy": 1.2599246203899384,
      "epoch": 2.5781990521327014,
      "grad_norm": 0.2824483811855316,
      "learning_rate": 0.00015118899403844172,
      "loss": 0.20817939937114716,
      "mean_token_accuracy": 0.9098999500274658,
      "num_tokens": 60395000.0,
      "step": 4896
    },
    {
      "entropy": 1.2227539122104645,
      "epoch": 2.578725645076356,
      "grad_norm": 0.2849332094192505,
      "learning_rate": 0.0001511694957264872,
      "loss": 0.1831064522266388,
      "mean_token_accuracy": 0.9280066043138504,
      "num_tokens": 60407336.0,
      "step": 4897
    },
    {
      "entropy": 1.27097687125206,
      "epoch": 2.5792522380200107,
      "grad_norm": 0.28264176845550537,
      "learning_rate": 0.0001511499949705397,
      "loss": 0.17965328693389893,
      "mean_token_accuracy": 0.9261346906423569,
      "num_tokens": 60419672.0,
      "step": 4898
    },
    {
      "entropy": 1.1849463284015656,
      "epoch": 2.579778830963665,
      "grad_norm": 0.27329471707344055,
      "learning_rate": 0.0001511304917717568,
      "loss": 0.19811423122882843,
      "mean_token_accuracy": 0.9193136245012283,
      "num_tokens": 60432008.0,
      "step": 4899
    },
    {
      "entropy": 1.2668118476867676,
      "epoch": 2.5803054239073195,
      "grad_norm": 0.2986825406551361,
      "learning_rate": 0.00015111098613129637,
      "loss": 0.1940465271472931,
      "mean_token_accuracy": 0.9172029942274094,
      "num_tokens": 60444344.0,
      "step": 4900
    },
    {
      "entropy": 1.2361448407173157,
      "epoch": 2.580832016850974,
      "grad_norm": 0.2798372805118561,
      "learning_rate": 0.00015109147805031633,
      "loss": 0.1894315630197525,
      "mean_token_accuracy": 0.9171324968338013,
      "num_tokens": 60456680.0,
      "step": 4901
    },
    {
      "entropy": 1.2954889237880707,
      "epoch": 2.581358609794629,
      "grad_norm": 0.3161953091621399,
      "learning_rate": 0.00015107196752997467,
      "loss": 0.18676374852657318,
      "mean_token_accuracy": 0.9234345406293869,
      "num_tokens": 60469016.0,
      "step": 4902
    },
    {
      "entropy": 1.2941784262657166,
      "epoch": 2.5818852027382833,
      "grad_norm": 0.31460681557655334,
      "learning_rate": 0.0001510524545714297,
      "loss": 0.19941675662994385,
      "mean_token_accuracy": 0.9116266816854477,
      "num_tokens": 60481352.0,
      "step": 4903
    },
    {
      "entropy": 1.2890789806842804,
      "epoch": 2.5824117956819377,
      "grad_norm": 0.2842191159725189,
      "learning_rate": 0.0001510329391758398,
      "loss": 0.17705805599689484,
      "mean_token_accuracy": 0.9248210489749908,
      "num_tokens": 60493688.0,
      "step": 4904
    },
    {
      "entropy": 1.3004145622253418,
      "epoch": 2.5829383886255926,
      "grad_norm": 0.30338117480278015,
      "learning_rate": 0.00015101342134436345,
      "loss": 0.19950935244560242,
      "mean_token_accuracy": 0.9210843443870544,
      "num_tokens": 60506024.0,
      "step": 4905
    },
    {
      "entropy": 1.3092130422592163,
      "epoch": 2.583464981569247,
      "grad_norm": 0.27796757221221924,
      "learning_rate": 0.0001509939010781593,
      "loss": 0.19549565017223358,
      "mean_token_accuracy": 0.9185791313648224,
      "num_tokens": 60518360.0,
      "step": 4906
    },
    {
      "entropy": 1.3043208718299866,
      "epoch": 2.5839915745129014,
      "grad_norm": 0.2806245684623718,
      "learning_rate": 0.00015097437837838622,
      "loss": 0.18059489130973816,
      "mean_token_accuracy": 0.9228035360574722,
      "num_tokens": 60530696.0,
      "step": 4907
    },
    {
      "entropy": 1.255238801240921,
      "epoch": 2.5845181674565563,
      "grad_norm": 0.25664472579956055,
      "learning_rate": 0.00015095485324620307,
      "loss": 0.17865927517414093,
      "mean_token_accuracy": 0.9228197932243347,
      "num_tokens": 60543032.0,
      "step": 4908
    },
    {
      "entropy": 1.3067887425422668,
      "epoch": 2.5850447604002107,
      "grad_norm": 0.27396947145462036,
      "learning_rate": 0.000150935325682769,
      "loss": 0.17701086401939392,
      "mean_token_accuracy": 0.9229120314121246,
      "num_tokens": 60555368.0,
      "step": 4909
    },
    {
      "entropy": 1.281879723072052,
      "epoch": 2.585571353343865,
      "grad_norm": 0.29621782898902893,
      "learning_rate": 0.0001509157956892432,
      "loss": 0.2053222805261612,
      "mean_token_accuracy": 0.9124819934368134,
      "num_tokens": 60567704.0,
      "step": 4910
    },
    {
      "entropy": 1.2755568325519562,
      "epoch": 2.58609794628752,
      "grad_norm": 0.27881988883018494,
      "learning_rate": 0.00015089626326678514,
      "loss": 0.1889788657426834,
      "mean_token_accuracy": 0.9183150380849838,
      "num_tokens": 60580040.0,
      "step": 4911
    },
    {
      "entropy": 1.2072003185749054,
      "epoch": 2.5866245392311744,
      "grad_norm": 0.2607763111591339,
      "learning_rate": 0.0001508767284165543,
      "loss": 0.16778123378753662,
      "mean_token_accuracy": 0.9284379333257675,
      "num_tokens": 60592376.0,
      "step": 4912
    },
    {
      "entropy": 1.249082773923874,
      "epoch": 2.587151132174829,
      "grad_norm": 0.2747190594673157,
      "learning_rate": 0.0001508571911397103,
      "loss": 0.19317784905433655,
      "mean_token_accuracy": 0.9194312244653702,
      "num_tokens": 60604712.0,
      "step": 4913
    },
    {
      "entropy": 1.2764122486114502,
      "epoch": 2.5876777251184833,
      "grad_norm": 0.28617385029792786,
      "learning_rate": 0.00015083765143741298,
      "loss": 0.1782130002975464,
      "mean_token_accuracy": 0.9292741566896439,
      "num_tokens": 60617048.0,
      "step": 4914
    },
    {
      "entropy": 1.2097508311271667,
      "epoch": 2.588204318062138,
      "grad_norm": 0.2889769971370697,
      "learning_rate": 0.00015081810931082235,
      "loss": 0.18960776925086975,
      "mean_token_accuracy": 0.9243573546409607,
      "num_tokens": 60629384.0,
      "step": 4915
    },
    {
      "entropy": 1.2399142682552338,
      "epoch": 2.5887309110057926,
      "grad_norm": 0.2919212281703949,
      "learning_rate": 0.00015079856476109848,
      "loss": 0.19247275590896606,
      "mean_token_accuracy": 0.9195180237293243,
      "num_tokens": 60641720.0,
      "step": 4916
    },
    {
      "entropy": 1.2030165493488312,
      "epoch": 2.589257503949447,
      "grad_norm": 0.29695242643356323,
      "learning_rate": 0.00015077901778940157,
      "loss": 0.21190334856510162,
      "mean_token_accuracy": 0.9165085405111313,
      "num_tokens": 60654056.0,
      "step": 4917
    },
    {
      "entropy": 1.1943929195404053,
      "epoch": 2.5897840968931014,
      "grad_norm": 0.2831406593322754,
      "learning_rate": 0.0001507594683968921,
      "loss": 0.18078364431858063,
      "mean_token_accuracy": 0.926941767334938,
      "num_tokens": 60666392.0,
      "step": 4918
    },
    {
      "entropy": 1.2222073674201965,
      "epoch": 2.5903106898367563,
      "grad_norm": 0.2951219379901886,
      "learning_rate": 0.00015073991658473055,
      "loss": 0.18491344153881073,
      "mean_token_accuracy": 0.9229899197816849,
      "num_tokens": 60678728.0,
      "step": 4919
    },
    {
      "entropy": 1.2402024269104004,
      "epoch": 2.5908372827804107,
      "grad_norm": 0.2913973927497864,
      "learning_rate": 0.00015072036235407756,
      "loss": 0.20621611177921295,
      "mean_token_accuracy": 0.9144409149885178,
      "num_tokens": 60691064.0,
      "step": 4920
    },
    {
      "entropy": 1.2773911952972412,
      "epoch": 2.591363875724065,
      "grad_norm": 0.27006563544273376,
      "learning_rate": 0.00015070080570609405,
      "loss": 0.18121947348117828,
      "mean_token_accuracy": 0.9238732904195786,
      "num_tokens": 60703400.0,
      "step": 4921
    },
    {
      "entropy": 1.2690989077091217,
      "epoch": 2.59189046866772,
      "grad_norm": 0.268216609954834,
      "learning_rate": 0.00015068124664194087,
      "loss": 0.18139807879924774,
      "mean_token_accuracy": 0.9222427159547806,
      "num_tokens": 60715736.0,
      "step": 4922
    },
    {
      "entropy": 1.200697124004364,
      "epoch": 2.5924170616113744,
      "grad_norm": 0.26179397106170654,
      "learning_rate": 0.00015066168516277923,
      "loss": 0.17864662408828735,
      "mean_token_accuracy": 0.9289314448833466,
      "num_tokens": 60728072.0,
      "step": 4923
    },
    {
      "entropy": 1.2993067800998688,
      "epoch": 2.592943654555029,
      "grad_norm": 0.2966409921646118,
      "learning_rate": 0.0001506421212697703,
      "loss": 0.19692149758338928,
      "mean_token_accuracy": 0.9166679531335831,
      "num_tokens": 60740408.0,
      "step": 4924
    },
    {
      "entropy": 1.2320555746555328,
      "epoch": 2.5934702474986837,
      "grad_norm": 0.2706405520439148,
      "learning_rate": 0.00015062255496407553,
      "loss": 0.17119428515434265,
      "mean_token_accuracy": 0.934698686003685,
      "num_tokens": 60752744.0,
      "step": 4925
    },
    {
      "entropy": 1.3128075301647186,
      "epoch": 2.593996840442338,
      "grad_norm": 0.29050150513648987,
      "learning_rate": 0.00015060298624685646,
      "loss": 0.20369823276996613,
      "mean_token_accuracy": 0.9178123623132706,
      "num_tokens": 60765080.0,
      "step": 4926
    },
    {
      "entropy": 1.3408489227294922,
      "epoch": 2.5945234333859926,
      "grad_norm": 0.28869307041168213,
      "learning_rate": 0.0001505834151192747,
      "loss": 0.19420656561851501,
      "mean_token_accuracy": 0.9170711934566498,
      "num_tokens": 60777416.0,
      "step": 4927
    },
    {
      "entropy": 1.3311878740787506,
      "epoch": 2.5950500263296474,
      "grad_norm": 0.3024950623512268,
      "learning_rate": 0.00015056384158249216,
      "loss": 0.2221667766571045,
      "mean_token_accuracy": 0.912192165851593,
      "num_tokens": 60789752.0,
      "step": 4928
    },
    {
      "entropy": 1.3265336155891418,
      "epoch": 2.595576619273302,
      "grad_norm": 0.2662777602672577,
      "learning_rate": 0.00015054426563767076,
      "loss": 0.1869158297777176,
      "mean_token_accuracy": 0.9230349361896515,
      "num_tokens": 60802088.0,
      "step": 4929
    },
    {
      "entropy": 1.3300181329250336,
      "epoch": 2.5961032122169563,
      "grad_norm": 0.2740636467933655,
      "learning_rate": 0.00015052468728597265,
      "loss": 0.18870602548122406,
      "mean_token_accuracy": 0.9215754866600037,
      "num_tokens": 60814424.0,
      "step": 4930
    },
    {
      "entropy": 1.2919826209545135,
      "epoch": 2.5966298051606107,
      "grad_norm": 0.2929941415786743,
      "learning_rate": 0.00015050510652856,
      "loss": 0.20022767782211304,
      "mean_token_accuracy": 0.9151937812566757,
      "num_tokens": 60826760.0,
      "step": 4931
    },
    {
      "entropy": 1.3107154071331024,
      "epoch": 2.597156398104265,
      "grad_norm": 0.2871747612953186,
      "learning_rate": 0.0001504855233665953,
      "loss": 0.20724672079086304,
      "mean_token_accuracy": 0.9125708937644958,
      "num_tokens": 60839096.0,
      "step": 4932
    },
    {
      "entropy": 1.3196457624435425,
      "epoch": 2.59768299104792,
      "grad_norm": 0.29808494448661804,
      "learning_rate": 0.000150465937801241,
      "loss": 0.20627810060977936,
      "mean_token_accuracy": 0.912663921713829,
      "num_tokens": 60851432.0,
      "step": 4933
    },
    {
      "entropy": 1.3223419189453125,
      "epoch": 2.5982095839915744,
      "grad_norm": 0.29379090666770935,
      "learning_rate": 0.00015044634983365983,
      "loss": 0.20280207693576813,
      "mean_token_accuracy": 0.9133302718400955,
      "num_tokens": 60863768.0,
      "step": 4934
    },
    {
      "entropy": 1.3727116882801056,
      "epoch": 2.598736176935229,
      "grad_norm": 0.2798406779766083,
      "learning_rate": 0.00015042675946501467,
      "loss": 0.1975862979888916,
      "mean_token_accuracy": 0.9198077917098999,
      "num_tokens": 60876104.0,
      "step": 4935
    },
    {
      "entropy": 1.3120767176151276,
      "epoch": 2.5992627698788837,
      "grad_norm": 0.27912437915802,
      "learning_rate": 0.00015040716669646837,
      "loss": 0.22149375081062317,
      "mean_token_accuracy": 0.9102857857942581,
      "num_tokens": 60888440.0,
      "step": 4936
    },
    {
      "entropy": 1.2911328673362732,
      "epoch": 2.599789362822538,
      "grad_norm": 0.2802145183086395,
      "learning_rate": 0.00015038757152918412,
      "loss": 0.19417962431907654,
      "mean_token_accuracy": 0.9157440662384033,
      "num_tokens": 60900776.0,
      "step": 4937
    },
    {
      "entropy": 1.3934805989265442,
      "epoch": 2.6003159557661926,
      "grad_norm": 0.28115054965019226,
      "learning_rate": 0.0001503679739643251,
      "loss": 0.19231252372264862,
      "mean_token_accuracy": 0.9217715561389923,
      "num_tokens": 60913112.0,
      "step": 4938
    },
    {
      "entropy": 1.2947366535663605,
      "epoch": 2.6008425487098474,
      "grad_norm": 0.24634382128715515,
      "learning_rate": 0.00015034837400305478,
      "loss": 0.18823671340942383,
      "mean_token_accuracy": 0.9209298342466354,
      "num_tokens": 60925448.0,
      "step": 4939
    },
    {
      "entropy": 1.2998045682907104,
      "epoch": 2.601369141653502,
      "grad_norm": 0.2468367964029312,
      "learning_rate": 0.00015032877164653668,
      "loss": 0.1801864206790924,
      "mean_token_accuracy": 0.924280971288681,
      "num_tokens": 60937784.0,
      "step": 4940
    },
    {
      "entropy": 1.294558823108673,
      "epoch": 2.6018957345971563,
      "grad_norm": 0.2462131679058075,
      "learning_rate": 0.00015030916689593444,
      "loss": 0.18104076385498047,
      "mean_token_accuracy": 0.923799991607666,
      "num_tokens": 60950120.0,
      "step": 4941
    },
    {
      "entropy": 1.3445029258728027,
      "epoch": 2.602422327540811,
      "grad_norm": 0.25693264603614807,
      "learning_rate": 0.0001502895597524119,
      "loss": 0.17664121091365814,
      "mean_token_accuracy": 0.926972508430481,
      "num_tokens": 60962456.0,
      "step": 4942
    },
    {
      "entropy": 1.3062118887901306,
      "epoch": 2.6029489204844656,
      "grad_norm": 0.2562658488750458,
      "learning_rate": 0.00015026995021713303,
      "loss": 0.17592185735702515,
      "mean_token_accuracy": 0.9261811375617981,
      "num_tokens": 60974792.0,
      "step": 4943
    },
    {
      "entropy": 1.3300414681434631,
      "epoch": 2.60347551342812,
      "grad_norm": 0.2567176818847656,
      "learning_rate": 0.00015025033829126194,
      "loss": 0.17641589045524597,
      "mean_token_accuracy": 0.9250547587871552,
      "num_tokens": 60987128.0,
      "step": 4944
    },
    {
      "entropy": 1.3550803363323212,
      "epoch": 2.604002106371775,
      "grad_norm": 0.30004099011421204,
      "learning_rate": 0.00015023072397596284,
      "loss": 0.19975590705871582,
      "mean_token_accuracy": 0.9152816534042358,
      "num_tokens": 60999464.0,
      "step": 4945
    },
    {
      "entropy": 1.3452115654945374,
      "epoch": 2.6045286993154293,
      "grad_norm": 0.2929515242576599,
      "learning_rate": 0.00015021110727240017,
      "loss": 0.20405176281929016,
      "mean_token_accuracy": 0.9165933579206467,
      "num_tokens": 61011800.0,
      "step": 4946
    },
    {
      "entropy": 1.3382624983787537,
      "epoch": 2.6050552922590837,
      "grad_norm": 0.293535977602005,
      "learning_rate": 0.0001501914881817384,
      "loss": 0.1919645369052887,
      "mean_token_accuracy": 0.9165096879005432,
      "num_tokens": 61024136.0,
      "step": 4947
    },
    {
      "entropy": 1.3008095026016235,
      "epoch": 2.605581885202738,
      "grad_norm": 0.271883100271225,
      "learning_rate": 0.00015017186670514225,
      "loss": 0.1803811490535736,
      "mean_token_accuracy": 0.9252985715866089,
      "num_tokens": 61036472.0,
      "step": 4948
    },
    {
      "entropy": 1.3121626377105713,
      "epoch": 2.6061084781463926,
      "grad_norm": 0.2665010392665863,
      "learning_rate": 0.0001501522428437765,
      "loss": 0.19021756947040558,
      "mean_token_accuracy": 0.917598158121109,
      "num_tokens": 61048808.0,
      "step": 4949
    },
    {
      "entropy": 1.3839079439640045,
      "epoch": 2.6066350710900474,
      "grad_norm": 0.33104774355888367,
      "learning_rate": 0.00015013261659880612,
      "loss": 0.1986754983663559,
      "mean_token_accuracy": 0.9182071089744568,
      "num_tokens": 61061144.0,
      "step": 4950
    },
    {
      "entropy": 1.394430249929428,
      "epoch": 2.607161664033702,
      "grad_norm": 0.30506664514541626,
      "learning_rate": 0.00015011298797139622,
      "loss": 0.20398959517478943,
      "mean_token_accuracy": 0.9175256192684174,
      "num_tokens": 61073480.0,
      "step": 4951
    },
    {
      "entropy": 1.3545320332050323,
      "epoch": 2.6076882569773563,
      "grad_norm": 0.26357078552246094,
      "learning_rate": 0.00015009335696271198,
      "loss": 0.18614837527275085,
      "mean_token_accuracy": 0.9249047487974167,
      "num_tokens": 61085816.0,
      "step": 4952
    },
    {
      "entropy": 1.3907168209552765,
      "epoch": 2.608214849921011,
      "grad_norm": 0.2741307020187378,
      "learning_rate": 0.00015007372357391885,
      "loss": 0.2001461386680603,
      "mean_token_accuracy": 0.9191935360431671,
      "num_tokens": 61098152.0,
      "step": 4953
    },
    {
      "entropy": 1.3543334007263184,
      "epoch": 2.6087414428646656,
      "grad_norm": 0.2835255265235901,
      "learning_rate": 0.0001500540878061823,
      "loss": 0.19529372453689575,
      "mean_token_accuracy": 0.9176384061574936,
      "num_tokens": 61110488.0,
      "step": 4954
    },
    {
      "entropy": 1.4248532056808472,
      "epoch": 2.60926803580832,
      "grad_norm": 0.30024221539497375,
      "learning_rate": 0.00015003444966066803,
      "loss": 0.19290918111801147,
      "mean_token_accuracy": 0.9208212941884995,
      "num_tokens": 61122824.0,
      "step": 4955
    },
    {
      "entropy": 1.4034450054168701,
      "epoch": 2.609794628751975,
      "grad_norm": 0.26515907049179077,
      "learning_rate": 0.0001500148091385418,
      "loss": 0.1898597776889801,
      "mean_token_accuracy": 0.9218143671751022,
      "num_tokens": 61135160.0,
      "step": 4956
    },
    {
      "entropy": 1.3745439052581787,
      "epoch": 2.6103212216956293,
      "grad_norm": 0.2902405858039856,
      "learning_rate": 0.0001499951662409696,
      "loss": 0.1814654916524887,
      "mean_token_accuracy": 0.9276015907526016,
      "num_tokens": 61147496.0,
      "step": 4957
    },
    {
      "entropy": 1.3754574656486511,
      "epoch": 2.6108478146392837,
      "grad_norm": 0.25714364647865295,
      "learning_rate": 0.00014997552096911745,
      "loss": 0.16097311675548553,
      "mean_token_accuracy": 0.9338142424821854,
      "num_tokens": 61159832.0,
      "step": 4958
    },
    {
      "entropy": 1.3770808279514313,
      "epoch": 2.6113744075829386,
      "grad_norm": 0.2730887234210968,
      "learning_rate": 0.00014995587332415164,
      "loss": 0.19565999507904053,
      "mean_token_accuracy": 0.9208303242921829,
      "num_tokens": 61172168.0,
      "step": 4959
    },
    {
      "entropy": 1.419737309217453,
      "epoch": 2.611901000526593,
      "grad_norm": 0.29134243726730347,
      "learning_rate": 0.00014993622330723857,
      "loss": 0.21380697190761566,
      "mean_token_accuracy": 0.9145271182060242,
      "num_tokens": 61184504.0,
      "step": 4960
    },
    {
      "entropy": 1.4257086515426636,
      "epoch": 2.6124275934702474,
      "grad_norm": 0.3023415207862854,
      "learning_rate": 0.0001499165709195446,
      "loss": 0.22057440876960754,
      "mean_token_accuracy": 0.911708801984787,
      "num_tokens": 61196840.0,
      "step": 4961
    },
    {
      "entropy": 1.3803713023662567,
      "epoch": 2.6129541864139023,
      "grad_norm": 0.26708221435546875,
      "learning_rate": 0.0001498969161622365,
      "loss": 0.18439361453056335,
      "mean_token_accuracy": 0.9259890019893646,
      "num_tokens": 61209176.0,
      "step": 4962
    },
    {
      "entropy": 1.3897138237953186,
      "epoch": 2.6134807793575567,
      "grad_norm": 0.2693830132484436,
      "learning_rate": 0.000149877259036481,
      "loss": 0.18453623354434967,
      "mean_token_accuracy": 0.9199992418289185,
      "num_tokens": 61221512.0,
      "step": 4963
    },
    {
      "entropy": 1.3420685827732086,
      "epoch": 2.614007372301211,
      "grad_norm": 0.26749715209007263,
      "learning_rate": 0.0001498575995434451,
      "loss": 0.18146294355392456,
      "mean_token_accuracy": 0.9221750348806381,
      "num_tokens": 61233848.0,
      "step": 4964
    },
    {
      "entropy": 1.424078345298767,
      "epoch": 2.6145339652448656,
      "grad_norm": 0.2738480269908905,
      "learning_rate": 0.00014983793768429582,
      "loss": 0.190167635679245,
      "mean_token_accuracy": 0.9162452816963196,
      "num_tokens": 61246184.0,
      "step": 4965
    },
    {
      "entropy": 1.424032837152481,
      "epoch": 2.61506055818852,
      "grad_norm": 0.3406761586666107,
      "learning_rate": 0.00014981827346020033,
      "loss": 0.21565118432044983,
      "mean_token_accuracy": 0.9105579107999802,
      "num_tokens": 61258520.0,
      "step": 4966
    },
    {
      "entropy": 1.3945941925048828,
      "epoch": 2.615587151132175,
      "grad_norm": 0.28860965371131897,
      "learning_rate": 0.00014979860687232605,
      "loss": 0.19807329773902893,
      "mean_token_accuracy": 0.918640673160553,
      "num_tokens": 61270856.0,
      "step": 4967
    },
    {
      "entropy": 1.4615474939346313,
      "epoch": 2.6161137440758293,
      "grad_norm": 0.2449038326740265,
      "learning_rate": 0.00014977893792184044,
      "loss": 0.17040449380874634,
      "mean_token_accuracy": 0.9294768422842026,
      "num_tokens": 61283192.0,
      "step": 4968
    },
    {
      "entropy": 1.387730598449707,
      "epoch": 2.6166403370194837,
      "grad_norm": 0.24215394258499146,
      "learning_rate": 0.00014975926660991114,
      "loss": 0.1671278029680252,
      "mean_token_accuracy": 0.9315755814313889,
      "num_tokens": 61295528.0,
      "step": 4969
    },
    {
      "entropy": 1.4137276411056519,
      "epoch": 2.6171669299631386,
      "grad_norm": 0.2633579969406128,
      "learning_rate": 0.0001497395929377059,
      "loss": 0.17472606897354126,
      "mean_token_accuracy": 0.9256842881441116,
      "num_tokens": 61307864.0,
      "step": 4970
    },
    {
      "entropy": 1.5112013518810272,
      "epoch": 2.617693522906793,
      "grad_norm": 0.30275458097457886,
      "learning_rate": 0.00014971991690639264,
      "loss": 0.18638665974140167,
      "mean_token_accuracy": 0.9230543673038483,
      "num_tokens": 61320200.0,
      "step": 4971
    },
    {
      "entropy": 1.4545276463031769,
      "epoch": 2.6182201158504474,
      "grad_norm": 0.2607262134552002,
      "learning_rate": 0.00014970023851713945,
      "loss": 0.18338130414485931,
      "mean_token_accuracy": 0.9261559545993805,
      "num_tokens": 61332536.0,
      "step": 4972
    },
    {
      "entropy": 1.5353376269340515,
      "epoch": 2.6187467087941023,
      "grad_norm": 0.2775002717971802,
      "learning_rate": 0.00014968055777111447,
      "loss": 0.19712965190410614,
      "mean_token_accuracy": 0.9215246587991714,
      "num_tokens": 61344872.0,
      "step": 4973
    },
    {
      "entropy": 1.492030531167984,
      "epoch": 2.6192733017377567,
      "grad_norm": 0.29135212302207947,
      "learning_rate": 0.0001496608746694861,
      "loss": 0.19520610570907593,
      "mean_token_accuracy": 0.9200300872325897,
      "num_tokens": 61357208.0,
      "step": 4974
    },
    {
      "entropy": 1.4847548604011536,
      "epoch": 2.619799894681411,
      "grad_norm": 0.2899549603462219,
      "learning_rate": 0.00014964118921342268,
      "loss": 0.20027291774749756,
      "mean_token_accuracy": 0.9164431840181351,
      "num_tokens": 61369544.0,
      "step": 4975
    },
    {
      "entropy": 1.545228362083435,
      "epoch": 2.620326487625066,
      "grad_norm": 0.32349705696105957,
      "learning_rate": 0.00014962150140409292,
      "loss": 0.202370285987854,
      "mean_token_accuracy": 0.9114632904529572,
      "num_tokens": 61381880.0,
      "step": 4976
    },
    {
      "entropy": 1.5073691308498383,
      "epoch": 2.6208530805687205,
      "grad_norm": 0.2822265923023224,
      "learning_rate": 0.0001496018112426656,
      "loss": 0.18207140266895294,
      "mean_token_accuracy": 0.9235720336437225,
      "num_tokens": 61394216.0,
      "step": 4977
    },
    {
      "entropy": 1.4734999239444733,
      "epoch": 2.621379673512375,
      "grad_norm": 0.26013490557670593,
      "learning_rate": 0.0001495821187303095,
      "loss": 0.18135610222816467,
      "mean_token_accuracy": 0.9248005449771881,
      "num_tokens": 61406552.0,
      "step": 4978
    },
    {
      "entropy": 1.4858646392822266,
      "epoch": 2.6219062664560298,
      "grad_norm": 0.26286378502845764,
      "learning_rate": 0.00014956242386819377,
      "loss": 0.17394036054611206,
      "mean_token_accuracy": 0.9256371855735779,
      "num_tokens": 61418888.0,
      "step": 4979
    },
    {
      "entropy": 1.5628124177455902,
      "epoch": 2.622432859399684,
      "grad_norm": 0.29492512345314026,
      "learning_rate": 0.00014954272665748753,
      "loss": 0.20093891024589539,
      "mean_token_accuracy": 0.9158370047807693,
      "num_tokens": 61431224.0,
      "step": 4980
    },
    {
      "entropy": 1.4645267724990845,
      "epoch": 2.6229594523433386,
      "grad_norm": 0.2536724805831909,
      "learning_rate": 0.00014952302709936004,
      "loss": 0.19026266038417816,
      "mean_token_accuracy": 0.9191191345453262,
      "num_tokens": 61443560.0,
      "step": 4981
    },
    {
      "entropy": 1.523823469877243,
      "epoch": 2.623486045286993,
      "grad_norm": 0.2790801227092743,
      "learning_rate": 0.00014950332519498077,
      "loss": 0.18979769945144653,
      "mean_token_accuracy": 0.9234752207994461,
      "num_tokens": 61455896.0,
      "step": 4982
    },
    {
      "entropy": 1.4641594588756561,
      "epoch": 2.6240126382306475,
      "grad_norm": 0.2721916735172272,
      "learning_rate": 0.00014948362094551937,
      "loss": 0.1872018724679947,
      "mean_token_accuracy": 0.9233260899782181,
      "num_tokens": 61468232.0,
      "step": 4983
    },
    {
      "entropy": 1.4742111563682556,
      "epoch": 2.6245392311743023,
      "grad_norm": 0.27885833382606506,
      "learning_rate": 0.00014946391435214555,
      "loss": 0.18875938653945923,
      "mean_token_accuracy": 0.9190693944692612,
      "num_tokens": 61480568.0,
      "step": 4984
    },
    {
      "entropy": 1.4826852679252625,
      "epoch": 2.6250658241179567,
      "grad_norm": 0.2697864770889282,
      "learning_rate": 0.0001494442054160291,
      "loss": 0.18166761100292206,
      "mean_token_accuracy": 0.9245208501815796,
      "num_tokens": 61492904.0,
      "step": 4985
    },
    {
      "entropy": 1.484268456697464,
      "epoch": 2.625592417061611,
      "grad_norm": 0.28331586718559265,
      "learning_rate": 0.00014942449413834008,
      "loss": 0.20358870923519135,
      "mean_token_accuracy": 0.9150479584932327,
      "num_tokens": 61505240.0,
      "step": 4986
    },
    {
      "entropy": 1.5757165551185608,
      "epoch": 2.626119010005266,
      "grad_norm": 0.29690924286842346,
      "learning_rate": 0.0001494047805202486,
      "loss": 0.2068706899881363,
      "mean_token_accuracy": 0.9173736572265625,
      "num_tokens": 61517576.0,
      "step": 4987
    },
    {
      "entropy": 1.533332884311676,
      "epoch": 2.6266456029489205,
      "grad_norm": 0.2705383002758026,
      "learning_rate": 0.000149385064562925,
      "loss": 0.17962002754211426,
      "mean_token_accuracy": 0.9280455708503723,
      "num_tokens": 61529912.0,
      "step": 4988
    },
    {
      "entropy": 1.4875019192695618,
      "epoch": 2.627172195892575,
      "grad_norm": 0.25805172324180603,
      "learning_rate": 0.0001493653462675397,
      "loss": 0.1781042516231537,
      "mean_token_accuracy": 0.9260552823543549,
      "num_tokens": 61542248.0,
      "step": 4989
    },
    {
      "entropy": 1.5069396495819092,
      "epoch": 2.6276987888362298,
      "grad_norm": 0.2651374340057373,
      "learning_rate": 0.0001493456256352632,
      "loss": 0.16291144490242004,
      "mean_token_accuracy": 0.9381222277879715,
      "num_tokens": 61554584.0,
      "step": 4990
    },
    {
      "entropy": 1.512205809354782,
      "epoch": 2.628225381779884,
      "grad_norm": 0.2627486288547516,
      "learning_rate": 0.0001493259026672662,
      "loss": 0.17600515484809875,
      "mean_token_accuracy": 0.9252037554979324,
      "num_tokens": 61566920.0,
      "step": 4991
    },
    {
      "entropy": 1.5533555746078491,
      "epoch": 2.6287519747235386,
      "grad_norm": 0.2759650945663452,
      "learning_rate": 0.0001493061773647196,
      "loss": 0.19801002740859985,
      "mean_token_accuracy": 0.916521742939949,
      "num_tokens": 61579256.0,
      "step": 4992
    },
    {
      "entropy": 1.5064502358436584,
      "epoch": 2.6292785676671935,
      "grad_norm": 0.2789139747619629,
      "learning_rate": 0.00014928644972879435,
      "loss": 0.1943892240524292,
      "mean_token_accuracy": 0.9217264503240585,
      "num_tokens": 61591592.0,
      "step": 4993
    },
    {
      "entropy": 1.4866106510162354,
      "epoch": 2.629805160610848,
      "grad_norm": 0.2809959352016449,
      "learning_rate": 0.00014926671976066153,
      "loss": 0.1923881471157074,
      "mean_token_accuracy": 0.920316144824028,
      "num_tokens": 61603928.0,
      "step": 4994
    },
    {
      "entropy": 1.452727496623993,
      "epoch": 2.6303317535545023,
      "grad_norm": 0.26779183745384216,
      "learning_rate": 0.00014924698746149247,
      "loss": 0.19007578492164612,
      "mean_token_accuracy": 0.9183843284845352,
      "num_tokens": 61616264.0,
      "step": 4995
    },
    {
      "entropy": 1.5179031789302826,
      "epoch": 2.630858346498157,
      "grad_norm": 0.31994399428367615,
      "learning_rate": 0.00014922725283245846,
      "loss": 0.20158439874649048,
      "mean_token_accuracy": 0.9133966416120529,
      "num_tokens": 61628600.0,
      "step": 4996
    },
    {
      "entropy": 1.4967687129974365,
      "epoch": 2.6313849394418116,
      "grad_norm": 0.27653440833091736,
      "learning_rate": 0.00014920751587473109,
      "loss": 0.18471458554267883,
      "mean_token_accuracy": 0.9204710125923157,
      "num_tokens": 61640936.0,
      "step": 4997
    },
    {
      "entropy": 1.4336793720722198,
      "epoch": 2.631911532385466,
      "grad_norm": 0.2625034749507904,
      "learning_rate": 0.00014918777658948207,
      "loss": 0.19384607672691345,
      "mean_token_accuracy": 0.9205842167139053,
      "num_tokens": 61653272.0,
      "step": 4998
    },
    {
      "entropy": 1.457231193780899,
      "epoch": 2.6324381253291205,
      "grad_norm": 0.27422916889190674,
      "learning_rate": 0.00014916803497788312,
      "loss": 0.19413116574287415,
      "mean_token_accuracy": 0.918518990278244,
      "num_tokens": 61665608.0,
      "step": 4999
    },
    {
      "entropy": 1.5142923891544342,
      "epoch": 2.632964718272775,
      "grad_norm": 0.297994464635849,
      "learning_rate": 0.00014914829104110624,
      "loss": 0.18459801375865936,
      "mean_token_accuracy": 0.9225414544343948,
      "num_tokens": 61677944.0,
      "step": 5000
    },
    {
      "entropy": 1.465256243944168,
      "epoch": 2.6334913112164298,
      "grad_norm": 0.26568931341171265,
      "learning_rate": 0.00014912854478032342,
      "loss": 0.16957612335681915,
      "mean_token_accuracy": 0.9300836622714996,
      "num_tokens": 61690280.0,
      "step": 5001
    },
    {
      "entropy": 1.4762811958789825,
      "epoch": 2.634017904160084,
      "grad_norm": 0.2986241579055786,
      "learning_rate": 0.00014910879619670704,
      "loss": 0.2125251293182373,
      "mean_token_accuracy": 0.9084591418504715,
      "num_tokens": 61702616.0,
      "step": 5002
    },
    {
      "entropy": 1.5015441477298737,
      "epoch": 2.6345444971037386,
      "grad_norm": 0.28843849897384644,
      "learning_rate": 0.00014908904529142936,
      "loss": 0.20061375200748444,
      "mean_token_accuracy": 0.9132287502288818,
      "num_tokens": 61714952.0,
      "step": 5003
    },
    {
      "entropy": 1.4222372174263,
      "epoch": 2.6350710900473935,
      "grad_norm": 0.26220911741256714,
      "learning_rate": 0.00014906929206566288,
      "loss": 0.1835784912109375,
      "mean_token_accuracy": 0.9224424511194229,
      "num_tokens": 61727288.0,
      "step": 5004
    },
    {
      "entropy": 1.503590077161789,
      "epoch": 2.635597682991048,
      "grad_norm": 0.2724861204624176,
      "learning_rate": 0.00014904953652058022,
      "loss": 0.18589074909687042,
      "mean_token_accuracy": 0.9205165356397629,
      "num_tokens": 61739624.0,
      "step": 5005
    },
    {
      "entropy": 1.4780482947826385,
      "epoch": 2.6361242759347023,
      "grad_norm": 0.2663821280002594,
      "learning_rate": 0.0001490297786573542,
      "loss": 0.16736972332000732,
      "mean_token_accuracy": 0.9276704490184784,
      "num_tokens": 61751960.0,
      "step": 5006
    },
    {
      "entropy": 1.5240800082683563,
      "epoch": 2.636650868878357,
      "grad_norm": 0.29903721809387207,
      "learning_rate": 0.00014901001847715772,
      "loss": 0.21255864202976227,
      "mean_token_accuracy": 0.914804145693779,
      "num_tokens": 61764296.0,
      "step": 5007
    },
    {
      "entropy": 1.4933780431747437,
      "epoch": 2.6371774618220116,
      "grad_norm": 0.29435548186302185,
      "learning_rate": 0.00014899025598116378,
      "loss": 0.21587303280830383,
      "mean_token_accuracy": 0.9131506383419037,
      "num_tokens": 61776632.0,
      "step": 5008
    },
    {
      "entropy": 1.4735883176326752,
      "epoch": 2.637704054765666,
      "grad_norm": 0.26635274291038513,
      "learning_rate": 0.00014897049117054562,
      "loss": 0.18540072441101074,
      "mean_token_accuracy": 0.9208427369594574,
      "num_tokens": 61788968.0,
      "step": 5009
    },
    {
      "entropy": 1.482529640197754,
      "epoch": 2.638230647709321,
      "grad_norm": 0.2780616283416748,
      "learning_rate": 0.00014895072404647652,
      "loss": 0.1992691457271576,
      "mean_token_accuracy": 0.9197406619787216,
      "num_tokens": 61801304.0,
      "step": 5010
    },
    {
      "entropy": 1.4966471791267395,
      "epoch": 2.6387572406529753,
      "grad_norm": 0.2593666911125183,
      "learning_rate": 0.00014893095461012996,
      "loss": 0.1759515404701233,
      "mean_token_accuracy": 0.9271847456693649,
      "num_tokens": 61813640.0,
      "step": 5011
    },
    {
      "entropy": 1.5135290026664734,
      "epoch": 2.6392838335966298,
      "grad_norm": 0.29530930519104004,
      "learning_rate": 0.00014891118286267953,
      "loss": 0.1886937916278839,
      "mean_token_accuracy": 0.9284406006336212,
      "num_tokens": 61825976.0,
      "step": 5012
    },
    {
      "entropy": 1.5016413629055023,
      "epoch": 2.639810426540284,
      "grad_norm": 0.2679826021194458,
      "learning_rate": 0.00014889140880529895,
      "loss": 0.18548890948295593,
      "mean_token_accuracy": 0.9230465292930603,
      "num_tokens": 61838312.0,
      "step": 5013
    },
    {
      "entropy": 1.4605536758899689,
      "epoch": 2.640337019483939,
      "grad_norm": 0.26483213901519775,
      "learning_rate": 0.00014887163243916212,
      "loss": 0.1955711394548416,
      "mean_token_accuracy": 0.9219475090503693,
      "num_tokens": 61850648.0,
      "step": 5014
    },
    {
      "entropy": 1.4169133007526398,
      "epoch": 2.6408636124275935,
      "grad_norm": 0.25406575202941895,
      "learning_rate": 0.00014885185376544303,
      "loss": 0.18453362584114075,
      "mean_token_accuracy": 0.9296320825815201,
      "num_tokens": 61862984.0,
      "step": 5015
    },
    {
      "entropy": 1.4207545518875122,
      "epoch": 2.641390205371248,
      "grad_norm": 0.26548707485198975,
      "learning_rate": 0.00014883207278531583,
      "loss": 0.1941414475440979,
      "mean_token_accuracy": 0.9198261648416519,
      "num_tokens": 61875320.0,
      "step": 5016
    },
    {
      "entropy": 1.4815255403518677,
      "epoch": 2.6419167983149023,
      "grad_norm": 0.30611151456832886,
      "learning_rate": 0.0001488122894999548,
      "loss": 0.19840019941329956,
      "mean_token_accuracy": 0.9162526428699493,
      "num_tokens": 61887656.0,
      "step": 5017
    },
    {
      "entropy": 1.459589034318924,
      "epoch": 2.642443391258557,
      "grad_norm": 0.26431792974472046,
      "learning_rate": 0.00014879250391053436,
      "loss": 0.19742351770401,
      "mean_token_accuracy": 0.9196898192167282,
      "num_tokens": 61899992.0,
      "step": 5018
    },
    {
      "entropy": 1.4588222205638885,
      "epoch": 2.6429699842022116,
      "grad_norm": 0.2513519823551178,
      "learning_rate": 0.00014877271601822906,
      "loss": 0.17419758439064026,
      "mean_token_accuracy": 0.9283565282821655,
      "num_tokens": 61912328.0,
      "step": 5019
    },
    {
      "entropy": 1.4621175527572632,
      "epoch": 2.643496577145866,
      "grad_norm": 0.2847839295864105,
      "learning_rate": 0.00014875292582421361,
      "loss": 0.20356279611587524,
      "mean_token_accuracy": 0.9177330881357193,
      "num_tokens": 61924664.0,
      "step": 5020
    },
    {
      "entropy": 1.4893657863140106,
      "epoch": 2.644023170089521,
      "grad_norm": 0.2779315710067749,
      "learning_rate": 0.0001487331333296628,
      "loss": 0.18966048955917358,
      "mean_token_accuracy": 0.9230923056602478,
      "num_tokens": 61937000.0,
      "step": 5021
    },
    {
      "entropy": 1.4491442739963531,
      "epoch": 2.6445497630331753,
      "grad_norm": 0.26622483134269714,
      "learning_rate": 0.00014871333853575162,
      "loss": 0.17434798181056976,
      "mean_token_accuracy": 0.9300359487533569,
      "num_tokens": 61949336.0,
      "step": 5022
    },
    {
      "entropy": 1.4646160006523132,
      "epoch": 2.6450763559768298,
      "grad_norm": 0.2670898139476776,
      "learning_rate": 0.0001486935414436552,
      "loss": 0.18202300369739532,
      "mean_token_accuracy": 0.9229418784379959,
      "num_tokens": 61961672.0,
      "step": 5023
    },
    {
      "entropy": 1.4607856273651123,
      "epoch": 2.6456029489204846,
      "grad_norm": 0.2811572551727295,
      "learning_rate": 0.0001486737420545487,
      "loss": 0.191569522023201,
      "mean_token_accuracy": 0.9196878373622894,
      "num_tokens": 61974008.0,
      "step": 5024
    },
    {
      "entropy": 1.4494915902614594,
      "epoch": 2.646129541864139,
      "grad_norm": 0.280921071767807,
      "learning_rate": 0.00014865394036960757,
      "loss": 0.20201906561851501,
      "mean_token_accuracy": 0.9147041141986847,
      "num_tokens": 61986344.0,
      "step": 5025
    },
    {
      "entropy": 1.4409070909023285,
      "epoch": 2.6466561348077935,
      "grad_norm": 0.30701422691345215,
      "learning_rate": 0.00014863413639000728,
      "loss": 0.21084938943386078,
      "mean_token_accuracy": 0.9105718582868576,
      "num_tokens": 61998680.0,
      "step": 5026
    },
    {
      "entropy": 1.488195687532425,
      "epoch": 2.6471827277514484,
      "grad_norm": 0.2876113951206207,
      "learning_rate": 0.0001486143301169235,
      "loss": 0.19386553764343262,
      "mean_token_accuracy": 0.9238238483667374,
      "num_tokens": 62011016.0,
      "step": 5027
    },
    {
      "entropy": 1.4491178691387177,
      "epoch": 2.647709320695103,
      "grad_norm": 0.27359431982040405,
      "learning_rate": 0.000148594521551532,
      "loss": 0.1898740977048874,
      "mean_token_accuracy": 0.9226633757352829,
      "num_tokens": 62023352.0,
      "step": 5028
    },
    {
      "entropy": 1.458969622850418,
      "epoch": 2.648235913638757,
      "grad_norm": 0.2540356516838074,
      "learning_rate": 0.0001485747106950087,
      "loss": 0.16982772946357727,
      "mean_token_accuracy": 0.9287703484296799,
      "num_tokens": 62035688.0,
      "step": 5029
    },
    {
      "entropy": 1.513466477394104,
      "epoch": 2.6487625065824116,
      "grad_norm": 0.3041362166404724,
      "learning_rate": 0.00014855489754852968,
      "loss": 0.19860799610614777,
      "mean_token_accuracy": 0.9202546030282974,
      "num_tokens": 62048024.0,
      "step": 5030
    },
    {
      "entropy": 1.4717740714550018,
      "epoch": 2.6492890995260665,
      "grad_norm": 0.26170480251312256,
      "learning_rate": 0.00014853508211327106,
      "loss": 0.18246625363826752,
      "mean_token_accuracy": 0.9252501279115677,
      "num_tokens": 62060360.0,
      "step": 5031
    },
    {
      "entropy": 1.485759437084198,
      "epoch": 2.649815692469721,
      "grad_norm": 0.2745889127254486,
      "learning_rate": 0.00014851526439040922,
      "loss": 0.18612299859523773,
      "mean_token_accuracy": 0.9231321662664413,
      "num_tokens": 62072696.0,
      "step": 5032
    },
    {
      "entropy": 1.4918552339076996,
      "epoch": 2.6503422854133754,
      "grad_norm": 0.282909095287323,
      "learning_rate": 0.00014849544438112067,
      "loss": 0.1870800256729126,
      "mean_token_accuracy": 0.9241845160722733,
      "num_tokens": 62085032.0,
      "step": 5033
    },
    {
      "entropy": 1.468372255563736,
      "epoch": 2.65086887835703,
      "grad_norm": 0.25634321570396423,
      "learning_rate": 0.00014847562208658192,
      "loss": 0.19367656111717224,
      "mean_token_accuracy": 0.9194540530443192,
      "num_tokens": 62097368.0,
      "step": 5034
    },
    {
      "entropy": 1.4729166626930237,
      "epoch": 2.6513954713006846,
      "grad_norm": 0.2615182101726532,
      "learning_rate": 0.00014845579750796974,
      "loss": 0.18999339640140533,
      "mean_token_accuracy": 0.918884739279747,
      "num_tokens": 62109704.0,
      "step": 5035
    },
    {
      "entropy": 1.4983617663383484,
      "epoch": 2.651922064244339,
      "grad_norm": 0.2739679515361786,
      "learning_rate": 0.00014843597064646098,
      "loss": 0.19533707201480865,
      "mean_token_accuracy": 0.9183563441038132,
      "num_tokens": 62122040.0,
      "step": 5036
    },
    {
      "entropy": 1.5050725042819977,
      "epoch": 2.6524486571879935,
      "grad_norm": 0.2634129822254181,
      "learning_rate": 0.0001484161415032327,
      "loss": 0.17496910691261292,
      "mean_token_accuracy": 0.9283895641565323,
      "num_tokens": 62134376.0,
      "step": 5037
    },
    {
      "entropy": 1.4989458620548248,
      "epoch": 2.6529752501316484,
      "grad_norm": 0.2645623981952667,
      "learning_rate": 0.000148396310079462,
      "loss": 0.1856268048286438,
      "mean_token_accuracy": 0.9227189868688583,
      "num_tokens": 62146712.0,
      "step": 5038
    },
    {
      "entropy": 1.437988817691803,
      "epoch": 2.653501843075303,
      "grad_norm": 0.24866002798080444,
      "learning_rate": 0.0001483764763763261,
      "loss": 0.19240614771842957,
      "mean_token_accuracy": 0.9217969477176666,
      "num_tokens": 62159048.0,
      "step": 5039
    },
    {
      "entropy": 1.495307207107544,
      "epoch": 2.654028436018957,
      "grad_norm": 0.2985769510269165,
      "learning_rate": 0.00014835664039500256,
      "loss": 0.18848103284835815,
      "mean_token_accuracy": 0.9216516762971878,
      "num_tokens": 62171384.0,
      "step": 5040
    },
    {
      "entropy": 1.5302987098693848,
      "epoch": 2.654555028962612,
      "grad_norm": 0.29026493430137634,
      "learning_rate": 0.0001483368021366688,
      "loss": 0.19517871737480164,
      "mean_token_accuracy": 0.919246032834053,
      "num_tokens": 62183720.0,
      "step": 5041
    },
    {
      "entropy": 1.5133644342422485,
      "epoch": 2.6550816219062665,
      "grad_norm": 0.29291194677352905,
      "learning_rate": 0.00014831696160250255,
      "loss": 0.1887335330247879,
      "mean_token_accuracy": 0.9231208711862564,
      "num_tokens": 62196056.0,
      "step": 5042
    },
    {
      "entropy": 1.5374446511268616,
      "epoch": 2.655608214849921,
      "grad_norm": 0.30532917380332947,
      "learning_rate": 0.0001482971187936816,
      "loss": 0.1930762231349945,
      "mean_token_accuracy": 0.9206221103668213,
      "num_tokens": 62208392.0,
      "step": 5043
    },
    {
      "entropy": 1.5095820128917694,
      "epoch": 2.656134807793576,
      "grad_norm": 0.2619345188140869,
      "learning_rate": 0.00014827727371138392,
      "loss": 0.16936379671096802,
      "mean_token_accuracy": 0.9295063018798828,
      "num_tokens": 62220728.0,
      "step": 5044
    },
    {
      "entropy": 1.4872773885726929,
      "epoch": 2.6566614007372302,
      "grad_norm": 0.259647399187088,
      "learning_rate": 0.00014825742635678763,
      "loss": 0.18199995160102844,
      "mean_token_accuracy": 0.9230765402317047,
      "num_tokens": 62233064.0,
      "step": 5045
    },
    {
      "entropy": 1.481743037700653,
      "epoch": 2.6571879936808847,
      "grad_norm": 0.2543366551399231,
      "learning_rate": 0.00014823757673107087,
      "loss": 0.1779671460390091,
      "mean_token_accuracy": 0.9248911142349243,
      "num_tokens": 62245400.0,
      "step": 5046
    },
    {
      "entropy": 1.4478746354579926,
      "epoch": 2.657714586624539,
      "grad_norm": 0.41329747438430786,
      "learning_rate": 0.00014821772483541206,
      "loss": 0.2164127230644226,
      "mean_token_accuracy": 0.9103289246559143,
      "num_tokens": 62257736.0,
      "step": 5047
    },
    {
      "entropy": 1.3805460035800934,
      "epoch": 2.658241179568194,
      "grad_norm": 0.2816043198108673,
      "learning_rate": 0.00014819787067098974,
      "loss": 0.16988952457904816,
      "mean_token_accuracy": 0.9290281236171722,
      "num_tokens": 62270072.0,
      "step": 5048
    },
    {
      "entropy": 1.4640763998031616,
      "epoch": 2.6587677725118484,
      "grad_norm": 0.3088597357273102,
      "learning_rate": 0.00014817801423898242,
      "loss": 0.2151375412940979,
      "mean_token_accuracy": 0.9137706756591797,
      "num_tokens": 62282408.0,
      "step": 5049
    },
    {
      "entropy": 1.4645437598228455,
      "epoch": 2.659294365455503,
      "grad_norm": 0.2925296425819397,
      "learning_rate": 0.00014815815554056888,
      "loss": 0.20622608065605164,
      "mean_token_accuracy": 0.916460320353508,
      "num_tokens": 62294744.0,
      "step": 5050
    },
    {
      "entropy": 1.4329487681388855,
      "epoch": 2.659820958399157,
      "grad_norm": 0.2808581292629242,
      "learning_rate": 0.00014813829457692811,
      "loss": 0.19199764728546143,
      "mean_token_accuracy": 0.9220515191555023,
      "num_tokens": 62307080.0,
      "step": 5051
    },
    {
      "entropy": 1.4280554354190826,
      "epoch": 2.660347551342812,
      "grad_norm": 0.3025852143764496,
      "learning_rate": 0.00014811843134923905,
      "loss": 0.20314553380012512,
      "mean_token_accuracy": 0.9143315702676773,
      "num_tokens": 62319416.0,
      "step": 5052
    },
    {
      "entropy": 1.4104456305503845,
      "epoch": 2.6608741442864665,
      "grad_norm": 0.25764983892440796,
      "learning_rate": 0.0001480985658586809,
      "loss": 0.19061550498008728,
      "mean_token_accuracy": 0.9198824912309647,
      "num_tokens": 62331752.0,
      "step": 5053
    },
    {
      "entropy": 1.3653182983398438,
      "epoch": 2.661400737230121,
      "grad_norm": 0.25570061802864075,
      "learning_rate": 0.00014807869810643293,
      "loss": 0.185185968875885,
      "mean_token_accuracy": 0.9199551194906235,
      "num_tokens": 62344088.0,
      "step": 5054
    },
    {
      "entropy": 1.406097024679184,
      "epoch": 2.661927330173776,
      "grad_norm": 0.2705627679824829,
      "learning_rate": 0.0001480588280936746,
      "loss": 0.18984414637088776,
      "mean_token_accuracy": 0.922783151268959,
      "num_tokens": 62356424.0,
      "step": 5055
    },
    {
      "entropy": 1.3973156809806824,
      "epoch": 2.6624539231174302,
      "grad_norm": 0.2610485255718231,
      "learning_rate": 0.0001480389558215855,
      "loss": 0.18623566627502441,
      "mean_token_accuracy": 0.9179131835699081,
      "num_tokens": 62368760.0,
      "step": 5056
    },
    {
      "entropy": 1.4304659962654114,
      "epoch": 2.6629805160610847,
      "grad_norm": 0.2898637056350708,
      "learning_rate": 0.00014801908129134526,
      "loss": 0.19678384065628052,
      "mean_token_accuracy": 0.9213199466466904,
      "num_tokens": 62381096.0,
      "step": 5057
    },
    {
      "entropy": 1.4300953447818756,
      "epoch": 2.6635071090047395,
      "grad_norm": 0.2751653492450714,
      "learning_rate": 0.0001479992045041338,
      "loss": 0.20221449434757233,
      "mean_token_accuracy": 0.9203903675079346,
      "num_tokens": 62393432.0,
      "step": 5058
    },
    {
      "entropy": 1.427178055047989,
      "epoch": 2.664033701948394,
      "grad_norm": 0.290485680103302,
      "learning_rate": 0.000147979325461131,
      "loss": 0.21843308210372925,
      "mean_token_accuracy": 0.913594514131546,
      "num_tokens": 62405768.0,
      "step": 5059
    },
    {
      "entropy": 1.385628342628479,
      "epoch": 2.6645602948920484,
      "grad_norm": 0.25159385800361633,
      "learning_rate": 0.000147959444163517,
      "loss": 0.1820984035730362,
      "mean_token_accuracy": 0.9247033447027206,
      "num_tokens": 62418104.0,
      "step": 5060
    },
    {
      "entropy": 1.4184448421001434,
      "epoch": 2.6650868878357032,
      "grad_norm": 0.27652180194854736,
      "learning_rate": 0.00014793956061247206,
      "loss": 0.19931438565254211,
      "mean_token_accuracy": 0.9216466546058655,
      "num_tokens": 62430440.0,
      "step": 5061
    },
    {
      "entropy": 1.4194599390029907,
      "epoch": 2.6656134807793577,
      "grad_norm": 0.25531280040740967,
      "learning_rate": 0.00014791967480917657,
      "loss": 0.20077663660049438,
      "mean_token_accuracy": 0.9188417941331863,
      "num_tokens": 62442776.0,
      "step": 5062
    },
    {
      "entropy": 1.4733093976974487,
      "epoch": 2.666140073723012,
      "grad_norm": 0.2748359739780426,
      "learning_rate": 0.00014789978675481099,
      "loss": 0.20483776926994324,
      "mean_token_accuracy": 0.9151559323072433,
      "num_tokens": 62455112.0,
      "step": 5063
    },
    {
      "entropy": 1.45448699593544,
      "epoch": 2.6666666666666665,
      "grad_norm": 0.27252256870269775,
      "learning_rate": 0.00014787989645055592,
      "loss": 0.1859803944826126,
      "mean_token_accuracy": 0.9244741946458817,
      "num_tokens": 62467448.0,
      "step": 5064
    },
    {
      "entropy": 1.4732069373130798,
      "epoch": 2.6671932596103214,
      "grad_norm": 0.2930166721343994,
      "learning_rate": 0.0001478600038975922,
      "loss": 0.19164472818374634,
      "mean_token_accuracy": 0.9163028299808502,
      "num_tokens": 62479784.0,
      "step": 5065
    },
    {
      "entropy": 1.5055941045284271,
      "epoch": 2.667719852553976,
      "grad_norm": 0.2679279148578644,
      "learning_rate": 0.0001478401090971007,
      "loss": 0.19617860019207,
      "mean_token_accuracy": 0.9202954471111298,
      "num_tokens": 62492120.0,
      "step": 5066
    },
    {
      "entropy": 1.460567593574524,
      "epoch": 2.6682464454976302,
      "grad_norm": 0.27816247940063477,
      "learning_rate": 0.00014782021205026253,
      "loss": 0.19984135031700134,
      "mean_token_accuracy": 0.9161521047353745,
      "num_tokens": 62504456.0,
      "step": 5067
    },
    {
      "entropy": 1.5180878043174744,
      "epoch": 2.6687730384412847,
      "grad_norm": 0.2608720064163208,
      "learning_rate": 0.00014780031275825873,
      "loss": 0.17761757969856262,
      "mean_token_accuracy": 0.9275451302528381,
      "num_tokens": 62516792.0,
      "step": 5068
    },
    {
      "entropy": 1.472340166568756,
      "epoch": 2.6692996313849395,
      "grad_norm": 0.2798198461532593,
      "learning_rate": 0.00014778041122227075,
      "loss": 0.19814461469650269,
      "mean_token_accuracy": 0.9174378514289856,
      "num_tokens": 62529128.0,
      "step": 5069
    },
    {
      "entropy": 1.4553137123584747,
      "epoch": 2.669826224328594,
      "grad_norm": 0.26443323493003845,
      "learning_rate": 0.0001477605074434799,
      "loss": 0.1944284290075302,
      "mean_token_accuracy": 0.9212834388017654,
      "num_tokens": 62541464.0,
      "step": 5070
    },
    {
      "entropy": 1.4736585021018982,
      "epoch": 2.6703528172722484,
      "grad_norm": 0.26062700152397156,
      "learning_rate": 0.00014774060142306783,
      "loss": 0.18508225679397583,
      "mean_token_accuracy": 0.9232953488826752,
      "num_tokens": 62553800.0,
      "step": 5071
    },
    {
      "entropy": 1.4530866146087646,
      "epoch": 2.6708794102159032,
      "grad_norm": 0.2670004665851593,
      "learning_rate": 0.00014772069316221624,
      "loss": 0.18442218005657196,
      "mean_token_accuracy": 0.9211182296276093,
      "num_tokens": 62566136.0,
      "step": 5072
    },
    {
      "entropy": 1.5300174951553345,
      "epoch": 2.6714060031595577,
      "grad_norm": 0.29020318388938904,
      "learning_rate": 0.00014770078266210693,
      "loss": 0.19545911252498627,
      "mean_token_accuracy": 0.9184393137693405,
      "num_tokens": 62578472.0,
      "step": 5073
    },
    {
      "entropy": 1.481099247932434,
      "epoch": 2.671932596103212,
      "grad_norm": 0.2822254002094269,
      "learning_rate": 0.00014768086992392187,
      "loss": 0.21277567744255066,
      "mean_token_accuracy": 0.9113451242446899,
      "num_tokens": 62590808.0,
      "step": 5074
    },
    {
      "entropy": 1.4741357564926147,
      "epoch": 2.672459189046867,
      "grad_norm": 0.2677938640117645,
      "learning_rate": 0.0001476609549488432,
      "loss": 0.18364793062210083,
      "mean_token_accuracy": 0.9212507903575897,
      "num_tokens": 62603144.0,
      "step": 5075
    },
    {
      "entropy": 1.4302943646907806,
      "epoch": 2.6729857819905214,
      "grad_norm": 0.28869354724884033,
      "learning_rate": 0.00014764103773805317,
      "loss": 0.19501850008964539,
      "mean_token_accuracy": 0.9206612259149551,
      "num_tokens": 62615480.0,
      "step": 5076
    },
    {
      "entropy": 1.4747921526432037,
      "epoch": 2.673512374934176,
      "grad_norm": 0.28305885195732117,
      "learning_rate": 0.00014762111829273408,
      "loss": 0.19867157936096191,
      "mean_token_accuracy": 0.9136838912963867,
      "num_tokens": 62627816.0,
      "step": 5077
    },
    {
      "entropy": 1.477564960718155,
      "epoch": 2.6740389678778307,
      "grad_norm": 0.29496532678604126,
      "learning_rate": 0.00014760119661406848,
      "loss": 0.2132798433303833,
      "mean_token_accuracy": 0.9153429567813873,
      "num_tokens": 62640152.0,
      "step": 5078
    },
    {
      "entropy": 1.4560523927211761,
      "epoch": 2.674565560821485,
      "grad_norm": 0.2781224548816681,
      "learning_rate": 0.00014758127270323901,
      "loss": 0.19361227750778198,
      "mean_token_accuracy": 0.9156886488199234,
      "num_tokens": 62652488.0,
      "step": 5079
    },
    {
      "entropy": 1.4673229455947876,
      "epoch": 2.6750921537651395,
      "grad_norm": 0.2802741527557373,
      "learning_rate": 0.00014756134656142842,
      "loss": 0.19040311872959137,
      "mean_token_accuracy": 0.9192695170640945,
      "num_tokens": 62664824.0,
      "step": 5080
    },
    {
      "entropy": 1.4941490292549133,
      "epoch": 2.675618746708794,
      "grad_norm": 0.289615273475647,
      "learning_rate": 0.00014754141818981958,
      "loss": 0.2048090100288391,
      "mean_token_accuracy": 0.9167844653129578,
      "num_tokens": 62677160.0,
      "step": 5081
    },
    {
      "entropy": 1.5785216391086578,
      "epoch": 2.6761453396524484,
      "grad_norm": 0.29720696806907654,
      "learning_rate": 0.00014752148758959556,
      "loss": 0.2018081247806549,
      "mean_token_accuracy": 0.9176874905824661,
      "num_tokens": 62689496.0,
      "step": 5082
    },
    {
      "entropy": 1.4724806249141693,
      "epoch": 2.6766719325961033,
      "grad_norm": 0.26232361793518066,
      "learning_rate": 0.0001475015547619395,
      "loss": 0.18163368105888367,
      "mean_token_accuracy": 0.9263022541999817,
      "num_tokens": 62701832.0,
      "step": 5083
    },
    {
      "entropy": 1.458247810602188,
      "epoch": 2.6771985255397577,
      "grad_norm": 0.25973543524742126,
      "learning_rate": 0.00014748161970803476,
      "loss": 0.18482840061187744,
      "mean_token_accuracy": 0.9212622344493866,
      "num_tokens": 62714168.0,
      "step": 5084
    },
    {
      "entropy": 1.4636139869689941,
      "epoch": 2.677725118483412,
      "grad_norm": 0.2422919124364853,
      "learning_rate": 0.00014746168242906466,
      "loss": 0.18235735595226288,
      "mean_token_accuracy": 0.925435408949852,
      "num_tokens": 62726504.0,
      "step": 5085
    },
    {
      "entropy": 1.5218181014060974,
      "epoch": 2.678251711427067,
      "grad_norm": 0.2919144928455353,
      "learning_rate": 0.00014744174292621284,
      "loss": 0.20702354609966278,
      "mean_token_accuracy": 0.9145085662603378,
      "num_tokens": 62738840.0,
      "step": 5086
    },
    {
      "entropy": 1.525699645280838,
      "epoch": 2.6787783043707214,
      "grad_norm": 0.2882302701473236,
      "learning_rate": 0.000147421801200663,
      "loss": 0.1976340264081955,
      "mean_token_accuracy": 0.9134345352649689,
      "num_tokens": 62751176.0,
      "step": 5087
    },
    {
      "entropy": 1.4348933696746826,
      "epoch": 2.679304897314376,
      "grad_norm": 0.2594358026981354,
      "learning_rate": 0.00014740185725359888,
      "loss": 0.18077734112739563,
      "mean_token_accuracy": 0.9274876117706299,
      "num_tokens": 62763512.0,
      "step": 5088
    },
    {
      "entropy": 1.5126748085021973,
      "epoch": 2.6798314902580307,
      "grad_norm": 0.27899715304374695,
      "learning_rate": 0.00014738191108620453,
      "loss": 0.16763128340244293,
      "mean_token_accuracy": 0.9275881052017212,
      "num_tokens": 62775848.0,
      "step": 5089
    },
    {
      "entropy": 1.4852455258369446,
      "epoch": 2.680358083201685,
      "grad_norm": 0.2729766368865967,
      "learning_rate": 0.00014736196269966398,
      "loss": 0.19001656770706177,
      "mean_token_accuracy": 0.920672133564949,
      "num_tokens": 62788184.0,
      "step": 5090
    },
    {
      "entropy": 1.5065564215183258,
      "epoch": 2.6808846761453395,
      "grad_norm": 0.2911035120487213,
      "learning_rate": 0.00014734201209516147,
      "loss": 0.1995036005973816,
      "mean_token_accuracy": 0.9154681265354156,
      "num_tokens": 62800520.0,
      "step": 5091
    },
    {
      "entropy": 1.533303052186966,
      "epoch": 2.6814112690889944,
      "grad_norm": 0.2991733253002167,
      "learning_rate": 0.00014732205927388135,
      "loss": 0.2060600221157074,
      "mean_token_accuracy": 0.9143246114253998,
      "num_tokens": 62812856.0,
      "step": 5092
    },
    {
      "entropy": 1.460000604391098,
      "epoch": 2.681937862032649,
      "grad_norm": 0.29276540875434875,
      "learning_rate": 0.00014730210423700808,
      "loss": 0.21440377831459045,
      "mean_token_accuracy": 0.9087410867214203,
      "num_tokens": 62825192.0,
      "step": 5093
    },
    {
      "entropy": 1.5091078579425812,
      "epoch": 2.6824644549763033,
      "grad_norm": 0.2839752435684204,
      "learning_rate": 0.00014728214698572631,
      "loss": 0.18155799806118011,
      "mean_token_accuracy": 0.9258093386888504,
      "num_tokens": 62837528.0,
      "step": 5094
    },
    {
      "entropy": 1.4895299971103668,
      "epoch": 2.682991047919958,
      "grad_norm": 0.29584982991218567,
      "learning_rate": 0.00014726218752122077,
      "loss": 0.1990211009979248,
      "mean_token_accuracy": 0.9221978783607483,
      "num_tokens": 62849864.0,
      "step": 5095
    },
    {
      "entropy": 1.5201840102672577,
      "epoch": 2.6835176408636126,
      "grad_norm": 0.2847629487514496,
      "learning_rate": 0.00014724222584467636,
      "loss": 0.18885351717472076,
      "mean_token_accuracy": 0.9195914417505264,
      "num_tokens": 62862200.0,
      "step": 5096
    },
    {
      "entropy": 1.5187770426273346,
      "epoch": 2.684044233807267,
      "grad_norm": 0.2975653111934662,
      "learning_rate": 0.00014722226195727805,
      "loss": 0.18812790513038635,
      "mean_token_accuracy": 0.9264185279607773,
      "num_tokens": 62874536.0,
      "step": 5097
    },
    {
      "entropy": 1.5271228551864624,
      "epoch": 2.6845708267509214,
      "grad_norm": 0.30130523443222046,
      "learning_rate": 0.00014720229586021098,
      "loss": 0.20495375990867615,
      "mean_token_accuracy": 0.9173427224159241,
      "num_tokens": 62886872.0,
      "step": 5098
    },
    {
      "entropy": 1.4671573638916016,
      "epoch": 2.685097419694576,
      "grad_norm": 0.2789948880672455,
      "learning_rate": 0.00014718232755466044,
      "loss": 0.19155564904212952,
      "mean_token_accuracy": 0.9244812875986099,
      "num_tokens": 62899208.0,
      "step": 5099
    },
    {
      "entropy": 1.4602892696857452,
      "epoch": 2.6856240126382307,
      "grad_norm": 0.282321959733963,
      "learning_rate": 0.00014716235704181183,
      "loss": 0.19326922297477722,
      "mean_token_accuracy": 0.9190951883792877,
      "num_tokens": 62911544.0,
      "step": 5100
    },
    {
      "entropy": 1.4617981910705566,
      "epoch": 2.686150605581885,
      "grad_norm": 0.2758800685405731,
      "learning_rate": 0.00014714238432285068,
      "loss": 0.18824873864650726,
      "mean_token_accuracy": 0.9187182486057281,
      "num_tokens": 62923880.0,
      "step": 5101
    },
    {
      "entropy": 1.569584608078003,
      "epoch": 2.6866771985255395,
      "grad_norm": 0.3216269314289093,
      "learning_rate": 0.00014712240939896267,
      "loss": 0.20186135172843933,
      "mean_token_accuracy": 0.9188182950019836,
      "num_tokens": 62936216.0,
      "step": 5102
    },
    {
      "entropy": 1.557395488023758,
      "epoch": 2.6872037914691944,
      "grad_norm": 0.299017071723938,
      "learning_rate": 0.00014710243227133358,
      "loss": 0.2022053301334381,
      "mean_token_accuracy": 0.9197170734405518,
      "num_tokens": 62948552.0,
      "step": 5103
    },
    {
      "entropy": 1.4289929568767548,
      "epoch": 2.687730384412849,
      "grad_norm": 0.2511787712574005,
      "learning_rate": 0.00014708245294114933,
      "loss": 0.17053452134132385,
      "mean_token_accuracy": 0.9278395920991898,
      "num_tokens": 62960888.0,
      "step": 5104
    },
    {
      "entropy": 1.5246893167495728,
      "epoch": 2.6882569773565033,
      "grad_norm": 0.3040166199207306,
      "learning_rate": 0.00014706247140959598,
      "loss": 0.211963951587677,
      "mean_token_accuracy": 0.9108811169862747,
      "num_tokens": 62973224.0,
      "step": 5105
    },
    {
      "entropy": 1.5441584885120392,
      "epoch": 2.688783570300158,
      "grad_norm": 0.28725534677505493,
      "learning_rate": 0.00014704248767785972,
      "loss": 0.20567721128463745,
      "mean_token_accuracy": 0.9181210845708847,
      "num_tokens": 62985560.0,
      "step": 5106
    },
    {
      "entropy": 1.5119162201881409,
      "epoch": 2.6893101632438126,
      "grad_norm": 0.29949843883514404,
      "learning_rate": 0.00014702250174712683,
      "loss": 0.2123764157295227,
      "mean_token_accuracy": 0.9134354293346405,
      "num_tokens": 62997896.0,
      "step": 5107
    },
    {
      "entropy": 1.4511745274066925,
      "epoch": 2.689836756187467,
      "grad_norm": 0.2614595592021942,
      "learning_rate": 0.00014700251361858386,
      "loss": 0.18792091310024261,
      "mean_token_accuracy": 0.9241390973329544,
      "num_tokens": 63010232.0,
      "step": 5108
    },
    {
      "entropy": 1.4284895956516266,
      "epoch": 2.690363349131122,
      "grad_norm": 0.26647090911865234,
      "learning_rate": 0.0001469825232934173,
      "loss": 0.18051902949810028,
      "mean_token_accuracy": 0.9248806685209274,
      "num_tokens": 63022568.0,
      "step": 5109
    },
    {
      "entropy": 1.4971970617771149,
      "epoch": 2.6908899420747763,
      "grad_norm": 0.2562994062900543,
      "learning_rate": 0.00014696253077281385,
      "loss": 0.18611495196819305,
      "mean_token_accuracy": 0.9216926395893097,
      "num_tokens": 63034904.0,
      "step": 5110
    },
    {
      "entropy": 1.4280113577842712,
      "epoch": 2.6914165350184307,
      "grad_norm": 0.27627092599868774,
      "learning_rate": 0.00014694253605796042,
      "loss": 0.19952592253684998,
      "mean_token_accuracy": 0.9170150011777878,
      "num_tokens": 63047240.0,
      "step": 5111
    },
    {
      "entropy": 1.38975790143013,
      "epoch": 2.6919431279620856,
      "grad_norm": 0.25191327929496765,
      "learning_rate": 0.00014692253915004393,
      "loss": 0.16972798109054565,
      "mean_token_accuracy": 0.9325916618108749,
      "num_tokens": 63059576.0,
      "step": 5112
    },
    {
      "entropy": 1.4815746545791626,
      "epoch": 2.69246972090574,
      "grad_norm": 0.27529120445251465,
      "learning_rate": 0.00014690254005025148,
      "loss": 0.1895769238471985,
      "mean_token_accuracy": 0.9245963543653488,
      "num_tokens": 63071912.0,
      "step": 5113
    },
    {
      "entropy": 1.3822405934333801,
      "epoch": 2.6929963138493944,
      "grad_norm": 0.28958284854888916,
      "learning_rate": 0.0001468825387597703,
      "loss": 0.1995624601840973,
      "mean_token_accuracy": 0.9220484644174576,
      "num_tokens": 63084248.0,
      "step": 5114
    },
    {
      "entropy": 1.3855730891227722,
      "epoch": 2.693522906793049,
      "grad_norm": 0.250946044921875,
      "learning_rate": 0.0001468625352797878,
      "loss": 0.1722625494003296,
      "mean_token_accuracy": 0.9266123324632645,
      "num_tokens": 63096584.0,
      "step": 5115
    },
    {
      "entropy": 1.4091150760650635,
      "epoch": 2.6940494997367033,
      "grad_norm": 0.2849009037017822,
      "learning_rate": 0.00014684252961149144,
      "loss": 0.2039925754070282,
      "mean_token_accuracy": 0.9155628532171249,
      "num_tokens": 63108920.0,
      "step": 5116
    },
    {
      "entropy": 1.3787662386894226,
      "epoch": 2.694576092680358,
      "grad_norm": 0.28099095821380615,
      "learning_rate": 0.00014682252175606878,
      "loss": 0.19700410962104797,
      "mean_token_accuracy": 0.9183400124311447,
      "num_tokens": 63121256.0,
      "step": 5117
    },
    {
      "entropy": 1.3753366470336914,
      "epoch": 2.6951026856240126,
      "grad_norm": 0.2916547656059265,
      "learning_rate": 0.00014680251171470766,
      "loss": 0.2051456719636917,
      "mean_token_accuracy": 0.9148056507110596,
      "num_tokens": 63133592.0,
      "step": 5118
    },
    {
      "entropy": 1.3619878888130188,
      "epoch": 2.695629278567667,
      "grad_norm": 0.27181702852249146,
      "learning_rate": 0.0001467824994885959,
      "loss": 0.16953186690807343,
      "mean_token_accuracy": 0.9321550875902176,
      "num_tokens": 63145928.0,
      "step": 5119
    },
    {
      "entropy": 1.3182980716228485,
      "epoch": 2.696155871511322,
      "grad_norm": 0.2878861427307129,
      "learning_rate": 0.0001467624850789215,
      "loss": 0.189472496509552,
      "mean_token_accuracy": 0.9215676188468933,
      "num_tokens": 63158264.0,
      "step": 5120
    },
    {
      "entropy": 1.3578261733055115,
      "epoch": 2.6966824644549763,
      "grad_norm": 0.29031458497047424,
      "learning_rate": 0.0001467424684868727,
      "loss": 0.19785207509994507,
      "mean_token_accuracy": 0.9179916083812714,
      "num_tokens": 63170600.0,
      "step": 5121
    },
    {
      "entropy": 1.3003853261470795,
      "epoch": 2.6972090573986307,
      "grad_norm": 0.2773895859718323,
      "learning_rate": 0.00014672244971363768,
      "loss": 0.18851050734519958,
      "mean_token_accuracy": 0.9205857366323471,
      "num_tokens": 63182936.0,
      "step": 5122
    },
    {
      "entropy": 1.348473995923996,
      "epoch": 2.6977356503422856,
      "grad_norm": 0.28892043232917786,
      "learning_rate": 0.00014670242876040483,
      "loss": 0.19180741906166077,
      "mean_token_accuracy": 0.9160315543413162,
      "num_tokens": 63195272.0,
      "step": 5123
    },
    {
      "entropy": 1.2486159205436707,
      "epoch": 2.69826224328594,
      "grad_norm": 0.28478798270225525,
      "learning_rate": 0.0001466824056283627,
      "loss": 0.199224591255188,
      "mean_token_accuracy": 0.9219311624765396,
      "num_tokens": 63207608.0,
      "step": 5124
    },
    {
      "entropy": 1.2535159289836884,
      "epoch": 2.6987888362295944,
      "grad_norm": 0.24823498725891113,
      "learning_rate": 0.00014666238031869994,
      "loss": 0.1766614466905594,
      "mean_token_accuracy": 0.9269942939281464,
      "num_tokens": 63219944.0,
      "step": 5125
    },
    {
      "entropy": 1.2995186150074005,
      "epoch": 2.6993154291732493,
      "grad_norm": 0.2622680962085724,
      "learning_rate": 0.0001466423528326054,
      "loss": 0.181438148021698,
      "mean_token_accuracy": 0.9263859242200851,
      "num_tokens": 63232280.0,
      "step": 5126
    },
    {
      "entropy": 1.3106255531311035,
      "epoch": 2.6998420221169037,
      "grad_norm": 0.300825834274292,
      "learning_rate": 0.00014662232317126788,
      "loss": 0.1900949627161026,
      "mean_token_accuracy": 0.919785350561142,
      "num_tokens": 63244616.0,
      "step": 5127
    },
    {
      "entropy": 1.2831209599971771,
      "epoch": 2.700368615060558,
      "grad_norm": 0.2816452085971832,
      "learning_rate": 0.00014660229133587653,
      "loss": 0.197813481092453,
      "mean_token_accuracy": 0.9157050102949142,
      "num_tokens": 63256952.0,
      "step": 5128
    },
    {
      "entropy": 1.2818153500556946,
      "epoch": 2.700895208004213,
      "grad_norm": 0.25588682293891907,
      "learning_rate": 0.00014658225732762045,
      "loss": 0.17363756895065308,
      "mean_token_accuracy": 0.9299523234367371,
      "num_tokens": 63269288.0,
      "step": 5129
    },
    {
      "entropy": 1.2921530902385712,
      "epoch": 2.7014218009478674,
      "grad_norm": 0.2761573791503906,
      "learning_rate": 0.00014656222114768898,
      "loss": 0.18385839462280273,
      "mean_token_accuracy": 0.9249665588140488,
      "num_tokens": 63281624.0,
      "step": 5130
    },
    {
      "entropy": 1.2621832191944122,
      "epoch": 2.701948393891522,
      "grad_norm": 0.2666982412338257,
      "learning_rate": 0.00014654218279727154,
      "loss": 0.18635493516921997,
      "mean_token_accuracy": 0.9211985170841217,
      "num_tokens": 63293960.0,
      "step": 5131
    },
    {
      "entropy": 1.263505905866623,
      "epoch": 2.7024749868351763,
      "grad_norm": 0.27494779229164124,
      "learning_rate": 0.0001465221422775577,
      "loss": 0.18919119238853455,
      "mean_token_accuracy": 0.9211875349283218,
      "num_tokens": 63306296.0,
      "step": 5132
    },
    {
      "entropy": 1.2421028017997742,
      "epoch": 2.7030015797788307,
      "grad_norm": 0.2690388560295105,
      "learning_rate": 0.00014650209958973713,
      "loss": 0.19874891638755798,
      "mean_token_accuracy": 0.9149452149868011,
      "num_tokens": 63318632.0,
      "step": 5133
    },
    {
      "entropy": 1.3355876505374908,
      "epoch": 2.7035281727224856,
      "grad_norm": 0.27606603503227234,
      "learning_rate": 0.00014648205473499963,
      "loss": 0.18637162446975708,
      "mean_token_accuracy": 0.9224206060171127,
      "num_tokens": 63330968.0,
      "step": 5134
    },
    {
      "entropy": 1.307241529226303,
      "epoch": 2.70405476566614,
      "grad_norm": 0.25661829113960266,
      "learning_rate": 0.0001464620077145352,
      "loss": 0.18009352684020996,
      "mean_token_accuracy": 0.9281450062990189,
      "num_tokens": 63343304.0,
      "step": 5135
    },
    {
      "entropy": 1.2898215353488922,
      "epoch": 2.7045813586097944,
      "grad_norm": 0.2869699001312256,
      "learning_rate": 0.0001464419585295339,
      "loss": 0.20290739834308624,
      "mean_token_accuracy": 0.9138100743293762,
      "num_tokens": 63355640.0,
      "step": 5136
    },
    {
      "entropy": 1.292168915271759,
      "epoch": 2.7051079515534493,
      "grad_norm": 0.273143470287323,
      "learning_rate": 0.00014642190718118593,
      "loss": 0.19269871711730957,
      "mean_token_accuracy": 0.9203228056430817,
      "num_tokens": 63367976.0,
      "step": 5137
    },
    {
      "entropy": 1.3003365695476532,
      "epoch": 2.7056345444971037,
      "grad_norm": 0.29047563672065735,
      "learning_rate": 0.0001464018536706816,
      "loss": 0.19433802366256714,
      "mean_token_accuracy": 0.9222328662872314,
      "num_tokens": 63380312.0,
      "step": 5138
    },
    {
      "entropy": 1.3424124121665955,
      "epoch": 2.706161137440758,
      "grad_norm": 0.28726476430892944,
      "learning_rate": 0.00014638179799921137,
      "loss": 0.1908145397901535,
      "mean_token_accuracy": 0.9210158437490463,
      "num_tokens": 63392648.0,
      "step": 5139
    },
    {
      "entropy": 1.2957504391670227,
      "epoch": 2.706687730384413,
      "grad_norm": 0.27413058280944824,
      "learning_rate": 0.00014636174016796583,
      "loss": 0.19619929790496826,
      "mean_token_accuracy": 0.9191281795501709,
      "num_tokens": 63404984.0,
      "step": 5140
    },
    {
      "entropy": 1.2907661497592926,
      "epoch": 2.7072143233280674,
      "grad_norm": 0.2575305700302124,
      "learning_rate": 0.00014634168017813574,
      "loss": 0.18570370972156525,
      "mean_token_accuracy": 0.9248242080211639,
      "num_tokens": 63417320.0,
      "step": 5141
    },
    {
      "entropy": 1.2979350090026855,
      "epoch": 2.707740916271722,
      "grad_norm": 0.25969117879867554,
      "learning_rate": 0.0001463216180309119,
      "loss": 0.1702679693698883,
      "mean_token_accuracy": 0.9241595566272736,
      "num_tokens": 63429656.0,
      "step": 5142
    },
    {
      "entropy": 1.303638905286789,
      "epoch": 2.7082675092153767,
      "grad_norm": 0.304808646440506,
      "learning_rate": 0.0001463015537274853,
      "loss": 0.21302911639213562,
      "mean_token_accuracy": 0.9168027937412262,
      "num_tokens": 63441992.0,
      "step": 5143
    },
    {
      "entropy": 1.3027910888195038,
      "epoch": 2.708794102159031,
      "grad_norm": 0.27357497811317444,
      "learning_rate": 0.00014628148726904706,
      "loss": 0.1874362826347351,
      "mean_token_accuracy": 0.9246872961521149,
      "num_tokens": 63454328.0,
      "step": 5144
    },
    {
      "entropy": 1.3277390003204346,
      "epoch": 2.7093206951026856,
      "grad_norm": 0.3108993470668793,
      "learning_rate": 0.00014626141865678837,
      "loss": 0.22254815697669983,
      "mean_token_accuracy": 0.9114834666252136,
      "num_tokens": 63466664.0,
      "step": 5145
    },
    {
      "entropy": 1.2865792214870453,
      "epoch": 2.7098472880463405,
      "grad_norm": 0.26160353422164917,
      "learning_rate": 0.0001462413478919006,
      "loss": 0.18020738661289215,
      "mean_token_accuracy": 0.9248409420251846,
      "num_tokens": 63479000.0,
      "step": 5146
    },
    {
      "entropy": 1.2977813482284546,
      "epoch": 2.710373880989995,
      "grad_norm": 0.27224621176719666,
      "learning_rate": 0.0001462212749755752,
      "loss": 0.17955082654953003,
      "mean_token_accuracy": 0.921922042965889,
      "num_tokens": 63491336.0,
      "step": 5147
    },
    {
      "entropy": 1.2977706491947174,
      "epoch": 2.7109004739336493,
      "grad_norm": 0.2755237817764282,
      "learning_rate": 0.00014620119990900384,
      "loss": 0.1703076809644699,
      "mean_token_accuracy": 0.9293846786022186,
      "num_tokens": 63503672.0,
      "step": 5148
    },
    {
      "entropy": 1.288289338350296,
      "epoch": 2.7114270668773037,
      "grad_norm": 0.2902235984802246,
      "learning_rate": 0.00014618112269337827,
      "loss": 0.20000024139881134,
      "mean_token_accuracy": 0.9120529741048813,
      "num_tokens": 63516008.0,
      "step": 5149
    },
    {
      "entropy": 1.3351134061813354,
      "epoch": 2.711953659820958,
      "grad_norm": 0.28813472390174866,
      "learning_rate": 0.0001461610433298903,
      "loss": 0.19814369082450867,
      "mean_token_accuracy": 0.9182053804397583,
      "num_tokens": 63528344.0,
      "step": 5150
    },
    {
      "entropy": 1.3223653137683868,
      "epoch": 2.712480252764613,
      "grad_norm": 0.2728074789047241,
      "learning_rate": 0.000146140961819732,
      "loss": 0.18013453483581543,
      "mean_token_accuracy": 0.9253092706203461,
      "num_tokens": 63540680.0,
      "step": 5151
    },
    {
      "entropy": 1.2829354405403137,
      "epoch": 2.7130068457082674,
      "grad_norm": 0.26212215423583984,
      "learning_rate": 0.00014612087816409533,
      "loss": 0.17448779940605164,
      "mean_token_accuracy": 0.9293143302202225,
      "num_tokens": 63553016.0,
      "step": 5152
    },
    {
      "entropy": 1.282748967409134,
      "epoch": 2.713533438651922,
      "grad_norm": 0.27840161323547363,
      "learning_rate": 0.00014610079236417273,
      "loss": 0.17790716886520386,
      "mean_token_accuracy": 0.9291750490665436,
      "num_tokens": 63565352.0,
      "step": 5153
    },
    {
      "entropy": 1.2521959245204926,
      "epoch": 2.7140600315955767,
      "grad_norm": 0.26914894580841064,
      "learning_rate": 0.00014608070442115648,
      "loss": 0.19306844472885132,
      "mean_token_accuracy": 0.9231783151626587,
      "num_tokens": 63577688.0,
      "step": 5154
    },
    {
      "entropy": 1.3317798674106598,
      "epoch": 2.714586624539231,
      "grad_norm": 0.30790823698043823,
      "learning_rate": 0.00014606061433623913,
      "loss": 0.19364537298679352,
      "mean_token_accuracy": 0.9153637439012527,
      "num_tokens": 63590024.0,
      "step": 5155
    },
    {
      "entropy": 1.3261720836162567,
      "epoch": 2.7151132174828856,
      "grad_norm": 0.27726447582244873,
      "learning_rate": 0.00014604052211061327,
      "loss": 0.1857507824897766,
      "mean_token_accuracy": 0.9212318807840347,
      "num_tokens": 63602360.0,
      "step": 5156
    },
    {
      "entropy": 1.3425556123256683,
      "epoch": 2.7156398104265405,
      "grad_norm": 0.28193601965904236,
      "learning_rate": 0.00014602042774547165,
      "loss": 0.17949122190475464,
      "mean_token_accuracy": 0.9272536188364029,
      "num_tokens": 63614696.0,
      "step": 5157
    },
    {
      "entropy": 1.2356054782867432,
      "epoch": 2.716166403370195,
      "grad_norm": 0.29531651735305786,
      "learning_rate": 0.00014600033124200718,
      "loss": 0.1958901435136795,
      "mean_token_accuracy": 0.9217116087675095,
      "num_tokens": 63627032.0,
      "step": 5158
    },
    {
      "entropy": 1.255002111196518,
      "epoch": 2.7166929963138493,
      "grad_norm": 0.28950661420822144,
      "learning_rate": 0.00014598023260141286,
      "loss": 0.204793781042099,
      "mean_token_accuracy": 0.9128590226173401,
      "num_tokens": 63639368.0,
      "step": 5159
    },
    {
      "entropy": 1.2777542769908905,
      "epoch": 2.717219589257504,
      "grad_norm": 0.30766382813453674,
      "learning_rate": 0.0001459601318248819,
      "loss": 0.20999419689178467,
      "mean_token_accuracy": 0.9123559296131134,
      "num_tokens": 63651704.0,
      "step": 5160
    },
    {
      "entropy": 1.3307510614395142,
      "epoch": 2.7177461822011586,
      "grad_norm": 0.2608412206172943,
      "learning_rate": 0.00014594002891360749,
      "loss": 0.18216381967067719,
      "mean_token_accuracy": 0.923665001988411,
      "num_tokens": 63664040.0,
      "step": 5161
    },
    {
      "entropy": 1.3185522556304932,
      "epoch": 2.718272775144813,
      "grad_norm": 0.2726461589336395,
      "learning_rate": 0.00014591992386878299,
      "loss": 0.17323972284793854,
      "mean_token_accuracy": 0.9279865473508835,
      "num_tokens": 63676376.0,
      "step": 5162
    },
    {
      "entropy": 1.2769018113613129,
      "epoch": 2.7187993680884674,
      "grad_norm": 0.26045459508895874,
      "learning_rate": 0.00014589981669160197,
      "loss": 0.1939665824174881,
      "mean_token_accuracy": 0.9193387180566788,
      "num_tokens": 63688712.0,
      "step": 5163
    },
    {
      "entropy": 1.3116481602191925,
      "epoch": 2.7193259610321223,
      "grad_norm": 0.2655951976776123,
      "learning_rate": 0.00014587970738325808,
      "loss": 0.19033044576644897,
      "mean_token_accuracy": 0.920961931347847,
      "num_tokens": 63701048.0,
      "step": 5164
    },
    {
      "entropy": 1.3548246026039124,
      "epoch": 2.7198525539757767,
      "grad_norm": 0.26746660470962524,
      "learning_rate": 0.00014585959594494514,
      "loss": 0.19389790296554565,
      "mean_token_accuracy": 0.9159002602100372,
      "num_tokens": 63713384.0,
      "step": 5165
    },
    {
      "entropy": 1.270853042602539,
      "epoch": 2.720379146919431,
      "grad_norm": 0.2794607877731323,
      "learning_rate": 0.00014583948237785693,
      "loss": 0.20356296002864838,
      "mean_token_accuracy": 0.9151900112628937,
      "num_tokens": 63725720.0,
      "step": 5166
    },
    {
      "entropy": 1.3171870708465576,
      "epoch": 2.7209057398630856,
      "grad_norm": 0.2875237762928009,
      "learning_rate": 0.00014581936668318755,
      "loss": 0.20418274402618408,
      "mean_token_accuracy": 0.916813850402832,
      "num_tokens": 63738056.0,
      "step": 5167
    },
    {
      "entropy": 1.2941392660140991,
      "epoch": 2.7214323328067405,
      "grad_norm": 0.2574351131916046,
      "learning_rate": 0.00014579924886213118,
      "loss": 0.1572742909193039,
      "mean_token_accuracy": 0.9385139495134354,
      "num_tokens": 63750392.0,
      "step": 5168
    },
    {
      "entropy": 1.3139513731002808,
      "epoch": 2.721958925750395,
      "grad_norm": 0.271741658449173,
      "learning_rate": 0.00014577912891588203,
      "loss": 0.1982661485671997,
      "mean_token_accuracy": 0.9115052968263626,
      "num_tokens": 63762728.0,
      "step": 5169
    },
    {
      "entropy": 1.2726760506629944,
      "epoch": 2.7224855186940493,
      "grad_norm": 0.25223416090011597,
      "learning_rate": 0.00014575900684563452,
      "loss": 0.1943812072277069,
      "mean_token_accuracy": 0.9204421192407608,
      "num_tokens": 63775064.0,
      "step": 5170
    },
    {
      "entropy": 1.372187465429306,
      "epoch": 2.723012111637704,
      "grad_norm": 0.2644391655921936,
      "learning_rate": 0.00014573888265258323,
      "loss": 0.19780956208705902,
      "mean_token_accuracy": 0.9187611937522888,
      "num_tokens": 63787400.0,
      "step": 5171
    },
    {
      "entropy": 1.3559328317642212,
      "epoch": 2.7235387045813586,
      "grad_norm": 0.2788529098033905,
      "learning_rate": 0.00014571875633792277,
      "loss": 0.20060011744499207,
      "mean_token_accuracy": 0.9212859869003296,
      "num_tokens": 63799736.0,
      "step": 5172
    },
    {
      "entropy": 1.309464156627655,
      "epoch": 2.724065297525013,
      "grad_norm": 0.280444860458374,
      "learning_rate": 0.0001456986279028479,
      "loss": 0.1905817836523056,
      "mean_token_accuracy": 0.9269229620695114,
      "num_tokens": 63812072.0,
      "step": 5173
    },
    {
      "entropy": 1.2712817192077637,
      "epoch": 2.724591890468668,
      "grad_norm": 0.24894411861896515,
      "learning_rate": 0.00014567849734855356,
      "loss": 0.171059712767601,
      "mean_token_accuracy": 0.9284334927797318,
      "num_tokens": 63824408.0,
      "step": 5174
    },
    {
      "entropy": 1.2753881216049194,
      "epoch": 2.7251184834123223,
      "grad_norm": 0.2422960251569748,
      "learning_rate": 0.00014565836467623483,
      "loss": 0.17582392692565918,
      "mean_token_accuracy": 0.9304329454898834,
      "num_tokens": 63836744.0,
      "step": 5175
    },
    {
      "entropy": 1.2553509175777435,
      "epoch": 2.7256450763559767,
      "grad_norm": 0.29317882657051086,
      "learning_rate": 0.0001456382298870868,
      "loss": 0.19796429574489594,
      "mean_token_accuracy": 0.9160526990890503,
      "num_tokens": 63849080.0,
      "step": 5176
    },
    {
      "entropy": 1.2531674802303314,
      "epoch": 2.7261716692996316,
      "grad_norm": 0.2940789759159088,
      "learning_rate": 0.00014561809298230474,
      "loss": 0.1948608160018921,
      "mean_token_accuracy": 0.9217565655708313,
      "num_tokens": 63861416.0,
      "step": 5177
    },
    {
      "entropy": 1.2800334095954895,
      "epoch": 2.726698262243286,
      "grad_norm": 0.27374762296676636,
      "learning_rate": 0.00014559795396308414,
      "loss": 0.1936039924621582,
      "mean_token_accuracy": 0.9136587381362915,
      "num_tokens": 63873752.0,
      "step": 5178
    },
    {
      "entropy": 1.2157056629657745,
      "epoch": 2.7272248551869405,
      "grad_norm": 0.25958263874053955,
      "learning_rate": 0.0001455778128306205,
      "loss": 0.1764679104089737,
      "mean_token_accuracy": 0.9242573529481888,
      "num_tokens": 63886088.0,
      "step": 5179
    },
    {
      "entropy": 1.316890001296997,
      "epoch": 2.727751448130595,
      "grad_norm": 0.26737043261528015,
      "learning_rate": 0.00014555766958610945,
      "loss": 0.1727948784828186,
      "mean_token_accuracy": 0.9259275048971176,
      "num_tokens": 63898424.0,
      "step": 5180
    },
    {
      "entropy": 1.2939126789569855,
      "epoch": 2.7282780410742498,
      "grad_norm": 0.2589419484138489,
      "learning_rate": 0.0001455375242307468,
      "loss": 0.19002790749073029,
      "mean_token_accuracy": 0.9219155609607697,
      "num_tokens": 63910760.0,
      "step": 5181
    },
    {
      "entropy": 1.2866698801517487,
      "epoch": 2.728804634017904,
      "grad_norm": 0.2805895507335663,
      "learning_rate": 0.00014551737676572846,
      "loss": 0.1931273192167282,
      "mean_token_accuracy": 0.9201782494783401,
      "num_tokens": 63923096.0,
      "step": 5182
    },
    {
      "entropy": 1.3028650879859924,
      "epoch": 2.7293312269615586,
      "grad_norm": 0.31638723611831665,
      "learning_rate": 0.00014549722719225047,
      "loss": 0.18288934230804443,
      "mean_token_accuracy": 0.9220278561115265,
      "num_tokens": 63935432.0,
      "step": 5183
    },
    {
      "entropy": 1.324157953262329,
      "epoch": 2.729857819905213,
      "grad_norm": 0.2694256007671356,
      "learning_rate": 0.00014547707551150908,
      "loss": 0.1794993281364441,
      "mean_token_accuracy": 0.9276534020900726,
      "num_tokens": 63947768.0,
      "step": 5184
    },
    {
      "entropy": 1.204854130744934,
      "epoch": 2.730384412848868,
      "grad_norm": 0.2793176770210266,
      "learning_rate": 0.00014545692172470041,
      "loss": 0.18844300508499146,
      "mean_token_accuracy": 0.9218170195817947,
      "num_tokens": 63960104.0,
      "step": 5185
    },
    {
      "entropy": 1.2477740049362183,
      "epoch": 2.7309110057925223,
      "grad_norm": 0.2856691777706146,
      "learning_rate": 0.00014543676583302095,
      "loss": 0.2099856734275818,
      "mean_token_accuracy": 0.9131463170051575,
      "num_tokens": 63972440.0,
      "step": 5186
    },
    {
      "entropy": 1.255644053220749,
      "epoch": 2.7314375987361768,
      "grad_norm": 0.2527347803115845,
      "learning_rate": 0.00014541660783766723,
      "loss": 0.1695612072944641,
      "mean_token_accuracy": 0.9271181374788284,
      "num_tokens": 63984776.0,
      "step": 5187
    },
    {
      "entropy": 1.3577595055103302,
      "epoch": 2.7319641916798316,
      "grad_norm": 0.2855788469314575,
      "learning_rate": 0.00014539644773983599,
      "loss": 0.19383448362350464,
      "mean_token_accuracy": 0.9202312529087067,
      "num_tokens": 63997112.0,
      "step": 5188
    },
    {
      "entropy": 1.3106902539730072,
      "epoch": 2.732490784623486,
      "grad_norm": 0.2907700836658478,
      "learning_rate": 0.00014537628554072392,
      "loss": 0.1855756938457489,
      "mean_token_accuracy": 0.9252744168043137,
      "num_tokens": 64009448.0,
      "step": 5189
    },
    {
      "entropy": 1.2726895809173584,
      "epoch": 2.7330173775671405,
      "grad_norm": 0.26627859473228455,
      "learning_rate": 0.000145356121241528,
      "loss": 0.17433853447437286,
      "mean_token_accuracy": 0.9330893158912659,
      "num_tokens": 64021784.0,
      "step": 5190
    },
    {
      "entropy": 1.2664414644241333,
      "epoch": 2.7335439705107953,
      "grad_norm": 0.31092652678489685,
      "learning_rate": 0.0001453359548434452,
      "loss": 0.20331533253192902,
      "mean_token_accuracy": 0.9178122878074646,
      "num_tokens": 64034120.0,
      "step": 5191
    },
    {
      "entropy": 1.3027536571025848,
      "epoch": 2.7340705634544498,
      "grad_norm": 0.2834033966064453,
      "learning_rate": 0.0001453157863476727,
      "loss": 0.2043476402759552,
      "mean_token_accuracy": 0.9179918766021729,
      "num_tokens": 64046456.0,
      "step": 5192
    },
    {
      "entropy": 1.2744874060153961,
      "epoch": 2.734597156398104,
      "grad_norm": 0.25400489568710327,
      "learning_rate": 0.0001452956157554078,
      "loss": 0.18056827783584595,
      "mean_token_accuracy": 0.9235711693763733,
      "num_tokens": 64058792.0,
      "step": 5193
    },
    {
      "entropy": 1.3080530762672424,
      "epoch": 2.735123749341759,
      "grad_norm": 0.26861700415611267,
      "learning_rate": 0.00014527544306784792,
      "loss": 0.18566924333572388,
      "mean_token_accuracy": 0.9226769655942917,
      "num_tokens": 64071128.0,
      "step": 5194
    },
    {
      "entropy": 1.2273122072219849,
      "epoch": 2.7356503422854135,
      "grad_norm": 0.29266297817230225,
      "learning_rate": 0.00014525526828619063,
      "loss": 0.20136258006095886,
      "mean_token_accuracy": 0.9192598909139633,
      "num_tokens": 64083464.0,
      "step": 5195
    },
    {
      "entropy": 1.2667721211910248,
      "epoch": 2.736176935229068,
      "grad_norm": 0.2782294750213623,
      "learning_rate": 0.00014523509141163349,
      "loss": 0.19713616371154785,
      "mean_token_accuracy": 0.9188579767942429,
      "num_tokens": 64095800.0,
      "step": 5196
    },
    {
      "entropy": 1.2403879165649414,
      "epoch": 2.7367035281727223,
      "grad_norm": 0.25267916917800903,
      "learning_rate": 0.00014521491244537434,
      "loss": 0.1698072850704193,
      "mean_token_accuracy": 0.9281718283891678,
      "num_tokens": 64108136.0,
      "step": 5197
    },
    {
      "entropy": 1.288205772638321,
      "epoch": 2.737230121116377,
      "grad_norm": 0.2854776680469513,
      "learning_rate": 0.0001451947313886111,
      "loss": 0.19052140414714813,
      "mean_token_accuracy": 0.9202392548322678,
      "num_tokens": 64120472.0,
      "step": 5198
    },
    {
      "entropy": 1.2585653364658356,
      "epoch": 2.7377567140600316,
      "grad_norm": 0.2774556279182434,
      "learning_rate": 0.0001451745482425418,
      "loss": 0.18510374426841736,
      "mean_token_accuracy": 0.9226521104574203,
      "num_tokens": 64132808.0,
      "step": 5199
    },
    {
      "entropy": 1.2777175903320312,
      "epoch": 2.738283307003686,
      "grad_norm": 0.2825121283531189,
      "learning_rate": 0.0001451543630083646,
      "loss": 0.18815097212791443,
      "mean_token_accuracy": 0.92147396504879,
      "num_tokens": 64145144.0,
      "step": 5200
    },
    {
      "entropy": 1.2806333005428314,
      "epoch": 2.7388098999473405,
      "grad_norm": 0.265458881855011,
      "learning_rate": 0.0001451341756872777,
      "loss": 0.1735052764415741,
      "mean_token_accuracy": 0.9319272637367249,
      "num_tokens": 64157480.0,
      "step": 5201
    },
    {
      "entropy": 1.232534110546112,
      "epoch": 2.7393364928909953,
      "grad_norm": 0.26001816987991333,
      "learning_rate": 0.00014511398628047964,
      "loss": 0.18147748708724976,
      "mean_token_accuracy": 0.9249249398708344,
      "num_tokens": 64169816.0,
      "step": 5202
    },
    {
      "entropy": 1.2178304493427277,
      "epoch": 2.7398630858346498,
      "grad_norm": 0.25026735663414,
      "learning_rate": 0.00014509379478916883,
      "loss": 0.18916335701942444,
      "mean_token_accuracy": 0.9198585599660873,
      "num_tokens": 64182152.0,
      "step": 5203
    },
    {
      "entropy": 1.2706682085990906,
      "epoch": 2.740389678778304,
      "grad_norm": 0.3009374737739563,
      "learning_rate": 0.00014507360121454398,
      "loss": 0.21226197481155396,
      "mean_token_accuracy": 0.914648026227951,
      "num_tokens": 64194488.0,
      "step": 5204
    },
    {
      "entropy": 1.2476151287555695,
      "epoch": 2.740916271721959,
      "grad_norm": 0.2885046899318695,
      "learning_rate": 0.00014505340555780383,
      "loss": 0.20312687754631042,
      "mean_token_accuracy": 0.9144037365913391,
      "num_tokens": 64206824.0,
      "step": 5205
    },
    {
      "entropy": 1.2514026165008545,
      "epoch": 2.7414428646656135,
      "grad_norm": 0.2752453088760376,
      "learning_rate": 0.00014503320782014735,
      "loss": 0.20632526278495789,
      "mean_token_accuracy": 0.9120770543813705,
      "num_tokens": 64219160.0,
      "step": 5206
    },
    {
      "entropy": 1.1842446029186249,
      "epoch": 2.741969457609268,
      "grad_norm": 0.24866744875907898,
      "learning_rate": 0.00014501300800277345,
      "loss": 0.163893461227417,
      "mean_token_accuracy": 0.9325317144393921,
      "num_tokens": 64231496.0,
      "step": 5207
    },
    {
      "entropy": 1.2263388633728027,
      "epoch": 2.742496050552923,
      "grad_norm": 0.285268634557724,
      "learning_rate": 0.00014499280610688137,
      "loss": 0.2103516310453415,
      "mean_token_accuracy": 0.9175826460123062,
      "num_tokens": 64243832.0,
      "step": 5208
    },
    {
      "entropy": 1.2222225368022919,
      "epoch": 2.743022643496577,
      "grad_norm": 0.26084497570991516,
      "learning_rate": 0.00014497260213367036,
      "loss": 0.1803930550813675,
      "mean_token_accuracy": 0.9286944568157196,
      "num_tokens": 64256168.0,
      "step": 5209
    },
    {
      "entropy": 1.2412607073783875,
      "epoch": 2.7435492364402316,
      "grad_norm": 0.25516054034233093,
      "learning_rate": 0.00014495239608433978,
      "loss": 0.17983819544315338,
      "mean_token_accuracy": 0.9248096346855164,
      "num_tokens": 64268504.0,
      "step": 5210
    },
    {
      "entropy": 1.2084096372127533,
      "epoch": 2.7440758293838865,
      "grad_norm": 0.24899202585220337,
      "learning_rate": 0.00014493218796008913,
      "loss": 0.18748557567596436,
      "mean_token_accuracy": 0.9227083474397659,
      "num_tokens": 64280840.0,
      "step": 5211
    },
    {
      "entropy": 1.2613386809825897,
      "epoch": 2.744602422327541,
      "grad_norm": 0.26508280634880066,
      "learning_rate": 0.0001449119777621181,
      "loss": 0.18772578239440918,
      "mean_token_accuracy": 0.9219543635845184,
      "num_tokens": 64293176.0,
      "step": 5212
    },
    {
      "entropy": 1.255203664302826,
      "epoch": 2.7451290152711953,
      "grad_norm": 0.2667011320590973,
      "learning_rate": 0.00014489176549162646,
      "loss": 0.1858472377061844,
      "mean_token_accuracy": 0.924461156129837,
      "num_tokens": 64305512.0,
      "step": 5213
    },
    {
      "entropy": 1.2351004779338837,
      "epoch": 2.7456556082148498,
      "grad_norm": 0.280694842338562,
      "learning_rate": 0.00014487155114981405,
      "loss": 0.1888059377670288,
      "mean_token_accuracy": 0.9221524894237518,
      "num_tokens": 64317848.0,
      "step": 5214
    },
    {
      "entropy": 1.2104078531265259,
      "epoch": 2.746182201158504,
      "grad_norm": 0.26034653186798096,
      "learning_rate": 0.00014485133473788087,
      "loss": 0.19770097732543945,
      "mean_token_accuracy": 0.9191710352897644,
      "num_tokens": 64330184.0,
      "step": 5215
    },
    {
      "entropy": 1.1654371917247772,
      "epoch": 2.746708794102159,
      "grad_norm": 0.25307705998420715,
      "learning_rate": 0.0001448311162570271,
      "loss": 0.18185806274414062,
      "mean_token_accuracy": 0.9264684021472931,
      "num_tokens": 64342520.0,
      "step": 5216
    },
    {
      "entropy": 1.1659926176071167,
      "epoch": 2.7472353870458135,
      "grad_norm": 0.25822415947914124,
      "learning_rate": 0.00014481089570845298,
      "loss": 0.1906038373708725,
      "mean_token_accuracy": 0.9225552678108215,
      "num_tokens": 64354856.0,
      "step": 5217
    },
    {
      "entropy": 1.1801254749298096,
      "epoch": 2.747761979989468,
      "grad_norm": 0.27742892503738403,
      "learning_rate": 0.00014479067309335888,
      "loss": 0.18619799613952637,
      "mean_token_accuracy": 0.9215237945318222,
      "num_tokens": 64367192.0,
      "step": 5218
    },
    {
      "entropy": 1.1851786077022552,
      "epoch": 2.748288572933123,
      "grad_norm": 0.2780286371707916,
      "learning_rate": 0.00014477044841294527,
      "loss": 0.19493329524993896,
      "mean_token_accuracy": 0.9193734228610992,
      "num_tokens": 64379528.0,
      "step": 5219
    },
    {
      "entropy": 1.233342856168747,
      "epoch": 2.748815165876777,
      "grad_norm": 0.28098252415657043,
      "learning_rate": 0.00014475022166841281,
      "loss": 0.19491320848464966,
      "mean_token_accuracy": 0.9158923923969269,
      "num_tokens": 64391864.0,
      "step": 5220
    },
    {
      "entropy": 1.2095706462860107,
      "epoch": 2.7493417588204316,
      "grad_norm": 0.2857399582862854,
      "learning_rate": 0.00014472999286096222,
      "loss": 0.2000798135995865,
      "mean_token_accuracy": 0.9185933470726013,
      "num_tokens": 64404200.0,
      "step": 5221
    },
    {
      "entropy": 1.1988272964954376,
      "epoch": 2.7498683517640865,
      "grad_norm": 0.24530266225337982,
      "learning_rate": 0.00014470976199179437,
      "loss": 0.17582175135612488,
      "mean_token_accuracy": 0.9265767633914948,
      "num_tokens": 64416536.0,
      "step": 5222
    },
    {
      "entropy": 1.1718643307685852,
      "epoch": 2.750394944707741,
      "grad_norm": 0.2735878527164459,
      "learning_rate": 0.00014468952906211024,
      "loss": 0.19718340039253235,
      "mean_token_accuracy": 0.9156841188669205,
      "num_tokens": 64428872.0,
      "step": 5223
    },
    {
      "entropy": 1.1433568000793457,
      "epoch": 2.7509215376513954,
      "grad_norm": 0.2572272717952728,
      "learning_rate": 0.00014466929407311102,
      "loss": 0.17076869308948517,
      "mean_token_accuracy": 0.9290028512477875,
      "num_tokens": 64441208.0,
      "step": 5224
    },
    {
      "entropy": 1.1182581186294556,
      "epoch": 2.7514481305950502,
      "grad_norm": 0.2657313048839569,
      "learning_rate": 0.0001446490570259978,
      "loss": 0.18352721631526947,
      "mean_token_accuracy": 0.9193586707115173,
      "num_tokens": 64453544.0,
      "step": 5225
    },
    {
      "entropy": 1.2004330456256866,
      "epoch": 2.7519747235387046,
      "grad_norm": 0.29233238101005554,
      "learning_rate": 0.00014462881792197207,
      "loss": 0.19094334542751312,
      "mean_token_accuracy": 0.9189015328884125,
      "num_tokens": 64465880.0,
      "step": 5226
    },
    {
      "entropy": 1.1328704953193665,
      "epoch": 2.752501316482359,
      "grad_norm": 0.302139550447464,
      "learning_rate": 0.00014460857676223523,
      "loss": 0.19258467853069305,
      "mean_token_accuracy": 0.9215465188026428,
      "num_tokens": 64478216.0,
      "step": 5227
    },
    {
      "entropy": 1.1518427729606628,
      "epoch": 2.753027909426014,
      "grad_norm": 0.30873894691467285,
      "learning_rate": 0.00014458833354798892,
      "loss": 0.21177682280540466,
      "mean_token_accuracy": 0.9149043411016464,
      "num_tokens": 64490552.0,
      "step": 5228
    },
    {
      "entropy": 1.1240712106227875,
      "epoch": 2.7535545023696684,
      "grad_norm": 0.2717989385128021,
      "learning_rate": 0.00014456808828043483,
      "loss": 0.19752712547779083,
      "mean_token_accuracy": 0.9208426624536514,
      "num_tokens": 64502888.0,
      "step": 5229
    },
    {
      "entropy": 1.1326690316200256,
      "epoch": 2.754081095313323,
      "grad_norm": 0.3072580099105835,
      "learning_rate": 0.0001445478409607748,
      "loss": 0.21544644236564636,
      "mean_token_accuracy": 0.9124970734119415,
      "num_tokens": 64515224.0,
      "step": 5230
    },
    {
      "entropy": 1.1086510717868805,
      "epoch": 2.754607688256977,
      "grad_norm": 0.27127352356910706,
      "learning_rate": 0.0001445275915902108,
      "loss": 0.17774805426597595,
      "mean_token_accuracy": 0.9219206869602203,
      "num_tokens": 64527560.0,
      "step": 5231
    },
    {
      "entropy": 1.1655531525611877,
      "epoch": 2.7551342812006316,
      "grad_norm": 0.2691557705402374,
      "learning_rate": 0.00014450734016994496,
      "loss": 0.17809565365314484,
      "mean_token_accuracy": 0.9255595207214355,
      "num_tokens": 64539896.0,
      "step": 5232
    },
    {
      "entropy": 1.1543531119823456,
      "epoch": 2.7556608741442865,
      "grad_norm": 0.27001041173934937,
      "learning_rate": 0.00014448708670117947,
      "loss": 0.1864895522594452,
      "mean_token_accuracy": 0.9234678447246552,
      "num_tokens": 64552232.0,
      "step": 5233
    },
    {
      "entropy": 1.1790067255496979,
      "epoch": 2.756187467087941,
      "grad_norm": 0.26234033703804016,
      "learning_rate": 0.00014446683118511663,
      "loss": 0.18123625218868256,
      "mean_token_accuracy": 0.9251542240381241,
      "num_tokens": 64564568.0,
      "step": 5234
    },
    {
      "entropy": 1.2072664499282837,
      "epoch": 2.7567140600315954,
      "grad_norm": 0.2714584767818451,
      "learning_rate": 0.00014444657362295893,
      "loss": 0.1822567880153656,
      "mean_token_accuracy": 0.9235909283161163,
      "num_tokens": 64576904.0,
      "step": 5235
    },
    {
      "entropy": 1.2083474397659302,
      "epoch": 2.7572406529752502,
      "grad_norm": 0.28527194261550903,
      "learning_rate": 0.00014442631401590889,
      "loss": 0.18012532591819763,
      "mean_token_accuracy": 0.92564357817173,
      "num_tokens": 64589240.0,
      "step": 5236
    },
    {
      "entropy": 1.2130491435527802,
      "epoch": 2.7577672459189047,
      "grad_norm": 0.28817492723464966,
      "learning_rate": 0.00014440605236516927,
      "loss": 0.2071560025215149,
      "mean_token_accuracy": 0.9179072976112366,
      "num_tokens": 64601576.0,
      "step": 5237
    },
    {
      "entropy": 1.1850566864013672,
      "epoch": 2.758293838862559,
      "grad_norm": 0.27280986309051514,
      "learning_rate": 0.00014438578867194282,
      "loss": 0.19194592535495758,
      "mean_token_accuracy": 0.9210992604494095,
      "num_tokens": 64613912.0,
      "step": 5238
    },
    {
      "entropy": 1.1494078934192657,
      "epoch": 2.758820431806214,
      "grad_norm": 0.25719964504241943,
      "learning_rate": 0.00014436552293743254,
      "loss": 0.18231096863746643,
      "mean_token_accuracy": 0.9223944246768951,
      "num_tokens": 64626248.0,
      "step": 5239
    },
    {
      "entropy": 1.2233496606349945,
      "epoch": 2.7593470247498684,
      "grad_norm": 0.2691393494606018,
      "learning_rate": 0.00014434525516284145,
      "loss": 0.19390058517456055,
      "mean_token_accuracy": 0.9192745685577393,
      "num_tokens": 64638584.0,
      "step": 5240
    },
    {
      "entropy": 1.2024221122264862,
      "epoch": 2.759873617693523,
      "grad_norm": 0.27692461013793945,
      "learning_rate": 0.00014432498534937274,
      "loss": 0.20155805349349976,
      "mean_token_accuracy": 0.9162195473909378,
      "num_tokens": 64650920.0,
      "step": 5241
    },
    {
      "entropy": 1.2103087902069092,
      "epoch": 2.7604002106371777,
      "grad_norm": 0.26054757833480835,
      "learning_rate": 0.00014430471349822973,
      "loss": 0.1835065633058548,
      "mean_token_accuracy": 0.9229078888893127,
      "num_tokens": 64663256.0,
      "step": 5242
    },
    {
      "entropy": 1.2184035181999207,
      "epoch": 2.760926803580832,
      "grad_norm": 0.2641240358352661,
      "learning_rate": 0.00014428443961061582,
      "loss": 0.1787245273590088,
      "mean_token_accuracy": 0.927379846572876,
      "num_tokens": 64675592.0,
      "step": 5243
    },
    {
      "entropy": 1.2011678516864777,
      "epoch": 2.7614533965244865,
      "grad_norm": 0.27148598432540894,
      "learning_rate": 0.00014426416368773455,
      "loss": 0.18567071855068207,
      "mean_token_accuracy": 0.9205454736948013,
      "num_tokens": 64687928.0,
      "step": 5244
    },
    {
      "entropy": 1.216088354587555,
      "epoch": 2.7619799894681414,
      "grad_norm": 0.26671719551086426,
      "learning_rate": 0.00014424388573078958,
      "loss": 0.20205147564411163,
      "mean_token_accuracy": 0.9125513881444931,
      "num_tokens": 64700264.0,
      "step": 5245
    },
    {
      "entropy": 1.1836239397525787,
      "epoch": 2.762506582411796,
      "grad_norm": 0.27321091294288635,
      "learning_rate": 0.00014422360574098468,
      "loss": 0.1828138530254364,
      "mean_token_accuracy": 0.9238127917051315,
      "num_tokens": 64712600.0,
      "step": 5246
    },
    {
      "entropy": 1.245295375585556,
      "epoch": 2.7630331753554502,
      "grad_norm": 0.28472384810447693,
      "learning_rate": 0.00014420332371952382,
      "loss": 0.18590515851974487,
      "mean_token_accuracy": 0.920741856098175,
      "num_tokens": 64724936.0,
      "step": 5247
    },
    {
      "entropy": 1.2197468280792236,
      "epoch": 2.7635597682991047,
      "grad_norm": 0.28858470916748047,
      "learning_rate": 0.00014418303966761095,
      "loss": 0.18920327723026276,
      "mean_token_accuracy": 0.9211139678955078,
      "num_tokens": 64737272.0,
      "step": 5248
    },
    {
      "entropy": 1.238138347864151,
      "epoch": 2.764086361242759,
      "grad_norm": 0.28883299231529236,
      "learning_rate": 0.00014416275358645027,
      "loss": 0.19570133090019226,
      "mean_token_accuracy": 0.9165759533643723,
      "num_tokens": 64749608.0,
      "step": 5249
    },
    {
      "entropy": 1.1910213232040405,
      "epoch": 2.764612954186414,
      "grad_norm": 0.2884749472141266,
      "learning_rate": 0.000144142465477246,
      "loss": 0.19064441323280334,
      "mean_token_accuracy": 0.9157634079456329,
      "num_tokens": 64761944.0,
      "step": 5250
    },
    {
      "entropy": 1.2231583893299103,
      "epoch": 2.7651395471300684,
      "grad_norm": 0.2951197922229767,
      "learning_rate": 0.00014412217534120255,
      "loss": 0.20204216241836548,
      "mean_token_accuracy": 0.9111807197332382,
      "num_tokens": 64774280.0,
      "step": 5251
    },
    {
      "entropy": 1.2133639454841614,
      "epoch": 2.765666140073723,
      "grad_norm": 0.2701243460178375,
      "learning_rate": 0.0001441018831795245,
      "loss": 0.17370574176311493,
      "mean_token_accuracy": 0.9307438880205154,
      "num_tokens": 64786616.0,
      "step": 5252
    },
    {
      "entropy": 1.2652662694454193,
      "epoch": 2.7661927330173777,
      "grad_norm": 0.29625698924064636,
      "learning_rate": 0.00014408158899341633,
      "loss": 0.1891971081495285,
      "mean_token_accuracy": 0.9214306026697159,
      "num_tokens": 64798952.0,
      "step": 5253
    },
    {
      "entropy": 1.2378207743167877,
      "epoch": 2.766719325961032,
      "grad_norm": 0.28450554609298706,
      "learning_rate": 0.0001440612927840829,
      "loss": 0.19953784346580505,
      "mean_token_accuracy": 0.9168573915958405,
      "num_tokens": 64811288.0,
      "step": 5254
    },
    {
      "entropy": 1.22379869222641,
      "epoch": 2.7672459189046865,
      "grad_norm": 0.2879558205604553,
      "learning_rate": 0.000144040994552729,
      "loss": 0.18986776471138,
      "mean_token_accuracy": 0.9215910136699677,
      "num_tokens": 64823624.0,
      "step": 5255
    },
    {
      "entropy": 1.2522724568843842,
      "epoch": 2.7677725118483414,
      "grad_norm": 0.2794259190559387,
      "learning_rate": 0.00014402069430055972,
      "loss": 0.1900891661643982,
      "mean_token_accuracy": 0.9175702631473541,
      "num_tokens": 64835960.0,
      "step": 5256
    },
    {
      "entropy": 1.2027116417884827,
      "epoch": 2.768299104791996,
      "grad_norm": 0.25514569878578186,
      "learning_rate": 0.00014400039202878007,
      "loss": 0.1653316169977188,
      "mean_token_accuracy": 0.9280052781105042,
      "num_tokens": 64848296.0,
      "step": 5257
    },
    {
      "entropy": 1.2198582589626312,
      "epoch": 2.7688256977356502,
      "grad_norm": 0.2743576765060425,
      "learning_rate": 0.00014398008773859535,
      "loss": 0.18544761836528778,
      "mean_token_accuracy": 0.9242594391107559,
      "num_tokens": 64860632.0,
      "step": 5258
    },
    {
      "entropy": 1.2221347987651825,
      "epoch": 2.769352290679305,
      "grad_norm": 0.2561204135417938,
      "learning_rate": 0.00014395978143121082,
      "loss": 0.18239974975585938,
      "mean_token_accuracy": 0.9222259372472763,
      "num_tokens": 64872968.0,
      "step": 5259
    },
    {
      "entropy": 1.1785187125205994,
      "epoch": 2.7698788836229595,
      "grad_norm": 0.25451043248176575,
      "learning_rate": 0.00014393947310783204,
      "loss": 0.1851981282234192,
      "mean_token_accuracy": 0.9230374693870544,
      "num_tokens": 64885304.0,
      "step": 5260
    },
    {
      "entropy": 1.238612949848175,
      "epoch": 2.770405476566614,
      "grad_norm": 0.27332982420921326,
      "learning_rate": 0.00014391916276966453,
      "loss": 0.18217436969280243,
      "mean_token_accuracy": 0.9256610721349716,
      "num_tokens": 64897640.0,
      "step": 5261
    },
    {
      "entropy": 1.2903965413570404,
      "epoch": 2.770932069510269,
      "grad_norm": 0.28146326541900635,
      "learning_rate": 0.000143898850417914,
      "loss": 0.19812659919261932,
      "mean_token_accuracy": 0.9138201922178268,
      "num_tokens": 64909976.0,
      "step": 5262
    },
    {
      "entropy": 1.2601101100444794,
      "epoch": 2.7714586624539232,
      "grad_norm": 0.2966877818107605,
      "learning_rate": 0.00014387853605378635,
      "loss": 0.19086399674415588,
      "mean_token_accuracy": 0.9252540320158005,
      "num_tokens": 64922312.0,
      "step": 5263
    },
    {
      "entropy": 1.269772619009018,
      "epoch": 2.7719852553975777,
      "grad_norm": 0.2694266438484192,
      "learning_rate": 0.00014385821967848744,
      "loss": 0.1787468045949936,
      "mean_token_accuracy": 0.9245084077119827,
      "num_tokens": 64934648.0,
      "step": 5264
    },
    {
      "entropy": 1.2981602251529694,
      "epoch": 2.772511848341232,
      "grad_norm": 0.308047354221344,
      "learning_rate": 0.0001438379012932234,
      "loss": 0.2297125905752182,
      "mean_token_accuracy": 0.9067529439926147,
      "num_tokens": 64946984.0,
      "step": 5265
    },
    {
      "entropy": 1.2986539602279663,
      "epoch": 2.7730384412848865,
      "grad_norm": 0.27238884568214417,
      "learning_rate": 0.00014381758089920037,
      "loss": 0.18589141964912415,
      "mean_token_accuracy": 0.9196806848049164,
      "num_tokens": 64959320.0,
      "step": 5266
    },
    {
      "entropy": 1.2592036128044128,
      "epoch": 2.7735650342285414,
      "grad_norm": 0.2565900683403015,
      "learning_rate": 0.00014379725849762468,
      "loss": 0.17390495538711548,
      "mean_token_accuracy": 0.9275496900081635,
      "num_tokens": 64971656.0,
      "step": 5267
    },
    {
      "entropy": 1.3076839447021484,
      "epoch": 2.774091627172196,
      "grad_norm": 0.27231037616729736,
      "learning_rate": 0.00014377693408970267,
      "loss": 0.1904066503047943,
      "mean_token_accuracy": 0.9181473851203918,
      "num_tokens": 64983992.0,
      "step": 5268
    },
    {
      "entropy": 1.347479671239853,
      "epoch": 2.7746182201158502,
      "grad_norm": 0.2822572588920593,
      "learning_rate": 0.00014375660767664103,
      "loss": 0.19602887332439423,
      "mean_token_accuracy": 0.9184332489967346,
      "num_tokens": 64996328.0,
      "step": 5269
    },
    {
      "entropy": 1.2493623495101929,
      "epoch": 2.775144813059505,
      "grad_norm": 0.2655910551548004,
      "learning_rate": 0.0001437362792596463,
      "loss": 0.19221758842468262,
      "mean_token_accuracy": 0.9155505746603012,
      "num_tokens": 65008664.0,
      "step": 5270
    },
    {
      "entropy": 1.3480267524719238,
      "epoch": 2.7756714060031595,
      "grad_norm": 0.28599438071250916,
      "learning_rate": 0.0001437159488399253,
      "loss": 0.19499412178993225,
      "mean_token_accuracy": 0.9208921939134598,
      "num_tokens": 65021000.0,
      "step": 5271
    },
    {
      "entropy": 1.2931790351867676,
      "epoch": 2.776197998946814,
      "grad_norm": 0.25337666273117065,
      "learning_rate": 0.00014369561641868497,
      "loss": 0.19652974605560303,
      "mean_token_accuracy": 0.9209015667438507,
      "num_tokens": 65033336.0,
      "step": 5272
    },
    {
      "entropy": 1.280663251876831,
      "epoch": 2.776724591890469,
      "grad_norm": 0.26555371284484863,
      "learning_rate": 0.00014367528199713225,
      "loss": 0.1882692575454712,
      "mean_token_accuracy": 0.9232436418533325,
      "num_tokens": 65045672.0,
      "step": 5273
    },
    {
      "entropy": 1.3029052913188934,
      "epoch": 2.7772511848341233,
      "grad_norm": 0.28663209080696106,
      "learning_rate": 0.0001436549455764743,
      "loss": 0.2037314623594284,
      "mean_token_accuracy": 0.9130341559648514,
      "num_tokens": 65058008.0,
      "step": 5274
    },
    {
      "entropy": 1.3463798463344574,
      "epoch": 2.7777777777777777,
      "grad_norm": 0.29982322454452515,
      "learning_rate": 0.00014363460715791838,
      "loss": 0.20301522314548492,
      "mean_token_accuracy": 0.9173043817281723,
      "num_tokens": 65070344.0,
      "step": 5275
    },
    {
      "entropy": 1.3196207582950592,
      "epoch": 2.7783043707214325,
      "grad_norm": 0.2660505175590515,
      "learning_rate": 0.00014361426674267192,
      "loss": 0.19917218387126923,
      "mean_token_accuracy": 0.917966678738594,
      "num_tokens": 65082680.0,
      "step": 5276
    },
    {
      "entropy": 1.2840626537799835,
      "epoch": 2.778830963665087,
      "grad_norm": 0.26385557651519775,
      "learning_rate": 0.00014359392433194233,
      "loss": 0.17794926464557648,
      "mean_token_accuracy": 0.9250075668096542,
      "num_tokens": 65095016.0,
      "step": 5277
    },
    {
      "entropy": 1.314213752746582,
      "epoch": 2.7793575566087414,
      "grad_norm": 0.28997546434402466,
      "learning_rate": 0.00014357357992693726,
      "loss": 0.1899823695421219,
      "mean_token_accuracy": 0.9185740947723389,
      "num_tokens": 65107352.0,
      "step": 5278
    },
    {
      "entropy": 1.3273931741714478,
      "epoch": 2.7798841495523963,
      "grad_norm": 0.31028249859809875,
      "learning_rate": 0.00014355323352886444,
      "loss": 0.18756736814975739,
      "mean_token_accuracy": 0.9197654277086258,
      "num_tokens": 65119688.0,
      "step": 5279
    },
    {
      "entropy": 1.2950712740421295,
      "epoch": 2.7804107424960507,
      "grad_norm": 0.27508464455604553,
      "learning_rate": 0.0001435328851389317,
      "loss": 0.19255675375461578,
      "mean_token_accuracy": 0.9202096164226532,
      "num_tokens": 65132024.0,
      "step": 5280
    },
    {
      "entropy": 1.2744563817977905,
      "epoch": 2.780937335439705,
      "grad_norm": 0.26312628388404846,
      "learning_rate": 0.000143512534758347,
      "loss": 0.18344777822494507,
      "mean_token_accuracy": 0.9259104430675507,
      "num_tokens": 65144360.0,
      "step": 5281
    },
    {
      "entropy": 1.2877683341503143,
      "epoch": 2.7814639283833595,
      "grad_norm": 0.2613600790500641,
      "learning_rate": 0.00014349218238831845,
      "loss": 0.17790953814983368,
      "mean_token_accuracy": 0.9296992272138596,
      "num_tokens": 65156696.0,
      "step": 5282
    },
    {
      "entropy": 1.2867728471755981,
      "epoch": 2.781990521327014,
      "grad_norm": 0.2650804817676544,
      "learning_rate": 0.00014347182803005427,
      "loss": 0.1828990876674652,
      "mean_token_accuracy": 0.9258298873901367,
      "num_tokens": 65169032.0,
      "step": 5283
    },
    {
      "entropy": 1.2489764094352722,
      "epoch": 2.782517114270669,
      "grad_norm": 0.2658465504646301,
      "learning_rate": 0.0001434514716847627,
      "loss": 0.18873544037342072,
      "mean_token_accuracy": 0.9171571433544159,
      "num_tokens": 65181368.0,
      "step": 5284
    },
    {
      "entropy": 1.2812611758708954,
      "epoch": 2.7830437072143233,
      "grad_norm": 0.2830650508403778,
      "learning_rate": 0.00014343111335365222,
      "loss": 0.18863148987293243,
      "mean_token_accuracy": 0.9229893535375595,
      "num_tokens": 65193704.0,
      "step": 5285
    },
    {
      "entropy": 1.2569757997989655,
      "epoch": 2.7835703001579777,
      "grad_norm": 0.2600972652435303,
      "learning_rate": 0.00014341075303793142,
      "loss": 0.1723325252532959,
      "mean_token_accuracy": 0.9290564060211182,
      "num_tokens": 65206040.0,
      "step": 5286
    },
    {
      "entropy": 1.27193483710289,
      "epoch": 2.7840968931016326,
      "grad_norm": 0.2815724015235901,
      "learning_rate": 0.00014339039073880895,
      "loss": 0.18959340453147888,
      "mean_token_accuracy": 0.9206994324922562,
      "num_tokens": 65218376.0,
      "step": 5287
    },
    {
      "entropy": 1.2226817309856415,
      "epoch": 2.784623486045287,
      "grad_norm": 0.2693398892879486,
      "learning_rate": 0.0001433700264574936,
      "loss": 0.19547715783119202,
      "mean_token_accuracy": 0.9208534508943558,
      "num_tokens": 65230712.0,
      "step": 5288
    },
    {
      "entropy": 1.2494825720787048,
      "epoch": 2.7851500789889414,
      "grad_norm": 0.2742237150669098,
      "learning_rate": 0.00014334966019519425,
      "loss": 0.18227458000183105,
      "mean_token_accuracy": 0.9274836480617523,
      "num_tokens": 65243048.0,
      "step": 5289
    },
    {
      "entropy": 1.260303258895874,
      "epoch": 2.7856766719325963,
      "grad_norm": 0.28628888726234436,
      "learning_rate": 0.00014332929195311997,
      "loss": 0.20284047722816467,
      "mean_token_accuracy": 0.9181239157915115,
      "num_tokens": 65255384.0,
      "step": 5290
    },
    {
      "entropy": 1.2684848606586456,
      "epoch": 2.7862032648762507,
      "grad_norm": 0.27382534742355347,
      "learning_rate": 0.00014330892173247984,
      "loss": 0.18891406059265137,
      "mean_token_accuracy": 0.9240432977676392,
      "num_tokens": 65267720.0,
      "step": 5291
    },
    {
      "entropy": 1.222090721130371,
      "epoch": 2.786729857819905,
      "grad_norm": 0.2744442820549011,
      "learning_rate": 0.0001432885495344832,
      "loss": 0.18809741735458374,
      "mean_token_accuracy": 0.9243640154600143,
      "num_tokens": 65280056.0,
      "step": 5292
    },
    {
      "entropy": 1.2474882304668427,
      "epoch": 2.78725645076356,
      "grad_norm": 0.27189698815345764,
      "learning_rate": 0.00014326817536033942,
      "loss": 0.18525537848472595,
      "mean_token_accuracy": 0.9237808287143707,
      "num_tokens": 65292392.0,
      "step": 5293
    },
    {
      "entropy": 1.2759239077568054,
      "epoch": 2.7877830437072144,
      "grad_norm": 0.2819172143936157,
      "learning_rate": 0.00014324779921125796,
      "loss": 0.20205055177211761,
      "mean_token_accuracy": 0.9156257063150406,
      "num_tokens": 65304728.0,
      "step": 5294
    },
    {
      "entropy": 1.273671805858612,
      "epoch": 2.788309636650869,
      "grad_norm": 0.27508002519607544,
      "learning_rate": 0.0001432274210884484,
      "loss": 0.19093528389930725,
      "mean_token_accuracy": 0.9219478368759155,
      "num_tokens": 65317064.0,
      "step": 5295
    },
    {
      "entropy": 1.3292940258979797,
      "epoch": 2.7888362295945237,
      "grad_norm": 0.28093087673187256,
      "learning_rate": 0.00014320704099312053,
      "loss": 0.2062903195619583,
      "mean_token_accuracy": 0.9147017002105713,
      "num_tokens": 65329400.0,
      "step": 5296
    },
    {
      "entropy": 1.2952646017074585,
      "epoch": 2.789362822538178,
      "grad_norm": 0.28177374601364136,
      "learning_rate": 0.00014318665892648424,
      "loss": 0.20459504425525665,
      "mean_token_accuracy": 0.9161693304777145,
      "num_tokens": 65341736.0,
      "step": 5297
    },
    {
      "entropy": 1.304824024438858,
      "epoch": 2.7898894154818326,
      "grad_norm": 0.2694205045700073,
      "learning_rate": 0.00014316627488974938,
      "loss": 0.1909191906452179,
      "mean_token_accuracy": 0.9177864342927933,
      "num_tokens": 65354072.0,
      "step": 5298
    },
    {
      "entropy": 1.3001328110694885,
      "epoch": 2.790416008425487,
      "grad_norm": 0.2765825390815735,
      "learning_rate": 0.00014314588888412607,
      "loss": 0.1955200880765915,
      "mean_token_accuracy": 0.921666830778122,
      "num_tokens": 65366408.0,
      "step": 5299
    },
    {
      "entropy": 1.2705854177474976,
      "epoch": 2.7909426013691414,
      "grad_norm": 0.27244964241981506,
      "learning_rate": 0.00014312550091082456,
      "loss": 0.18523868918418884,
      "mean_token_accuracy": 0.924994632601738,
      "num_tokens": 65378744.0,
      "step": 5300
    },
    {
      "entropy": 1.2694044709205627,
      "epoch": 2.7914691943127963,
      "grad_norm": 0.2661687731742859,
      "learning_rate": 0.00014310511097105513,
      "loss": 0.1898202747106552,
      "mean_token_accuracy": 0.9222358614206314,
      "num_tokens": 65391080.0,
      "step": 5301
    },
    {
      "entropy": 1.29021418094635,
      "epoch": 2.7919957872564507,
      "grad_norm": 0.2877323627471924,
      "learning_rate": 0.00014308471906602824,
      "loss": 0.19828127324581146,
      "mean_token_accuracy": 0.9166731089353561,
      "num_tokens": 65403416.0,
      "step": 5302
    },
    {
      "entropy": 1.2674870193004608,
      "epoch": 2.792522380200105,
      "grad_norm": 0.275603324174881,
      "learning_rate": 0.00014306432519695436,
      "loss": 0.19587281346321106,
      "mean_token_accuracy": 0.9162575602531433,
      "num_tokens": 65415752.0,
      "step": 5303
    },
    {
      "entropy": 1.2863653898239136,
      "epoch": 2.79304897314376,
      "grad_norm": 0.27803856134414673,
      "learning_rate": 0.00014304392936504424,
      "loss": 0.19489046931266785,
      "mean_token_accuracy": 0.9222230017185211,
      "num_tokens": 65428088.0,
      "step": 5304
    },
    {
      "entropy": 1.3094123899936676,
      "epoch": 2.7935755660874144,
      "grad_norm": 0.31087836623191833,
      "learning_rate": 0.00014302353157150862,
      "loss": 0.19567900896072388,
      "mean_token_accuracy": 0.9196409285068512,
      "num_tokens": 65440424.0,
      "step": 5305
    },
    {
      "entropy": 1.2640268802642822,
      "epoch": 2.794102159031069,
      "grad_norm": 0.27602046728134155,
      "learning_rate": 0.0001430031318175584,
      "loss": 0.19994071125984192,
      "mean_token_accuracy": 0.9169894754886627,
      "num_tokens": 65452760.0,
      "step": 5306
    },
    {
      "entropy": 1.3626899421215057,
      "epoch": 2.7946287519747237,
      "grad_norm": 0.2630988359451294,
      "learning_rate": 0.00014298273010440458,
      "loss": 0.17848266661167145,
      "mean_token_accuracy": 0.9260770082473755,
      "num_tokens": 65465096.0,
      "step": 5307
    },
    {
      "entropy": 1.3198005855083466,
      "epoch": 2.795155344918378,
      "grad_norm": 0.254104346036911,
      "learning_rate": 0.00014296232643325836,
      "loss": 0.17157261073589325,
      "mean_token_accuracy": 0.929302304983139,
      "num_tokens": 65477432.0,
      "step": 5308
    },
    {
      "entropy": 1.3163739144802094,
      "epoch": 2.7956819378620326,
      "grad_norm": 0.2930879294872284,
      "learning_rate": 0.00014294192080533088,
      "loss": 0.18230167031288147,
      "mean_token_accuracy": 0.9215236604213715,
      "num_tokens": 65489768.0,
      "step": 5309
    },
    {
      "entropy": 1.331737995147705,
      "epoch": 2.7962085308056874,
      "grad_norm": 0.2841273546218872,
      "learning_rate": 0.0001429215132218336,
      "loss": 0.21370966732501984,
      "mean_token_accuracy": 0.9168280363082886,
      "num_tokens": 65502104.0,
      "step": 5310
    },
    {
      "entropy": 1.3565723598003387,
      "epoch": 2.796735123749342,
      "grad_norm": 0.2748474180698395,
      "learning_rate": 0.00014290110368397795,
      "loss": 0.1781301200389862,
      "mean_token_accuracy": 0.93081134557724,
      "num_tokens": 65514440.0,
      "step": 5311
    },
    {
      "entropy": 1.332760989665985,
      "epoch": 2.7972617166929963,
      "grad_norm": 0.2885291576385498,
      "learning_rate": 0.0001428806921929756,
      "loss": 0.19329184293746948,
      "mean_token_accuracy": 0.9193468242883682,
      "num_tokens": 65526776.0,
      "step": 5312
    },
    {
      "entropy": 1.3488106727600098,
      "epoch": 2.7977883096366507,
      "grad_norm": 0.3057740330696106,
      "learning_rate": 0.0001428602787500381,
      "loss": 0.2155866026878357,
      "mean_token_accuracy": 0.9115214347839355,
      "num_tokens": 65539112.0,
      "step": 5313
    },
    {
      "entropy": 1.3490218222141266,
      "epoch": 2.7983149025803056,
      "grad_norm": 0.2725186049938202,
      "learning_rate": 0.00014283986335637743,
      "loss": 0.1859339028596878,
      "mean_token_accuracy": 0.9261448532342911,
      "num_tokens": 65551448.0,
      "step": 5314
    },
    {
      "entropy": 1.310249239206314,
      "epoch": 2.79884149552396,
      "grad_norm": 0.2817498445510864,
      "learning_rate": 0.00014281944601320546,
      "loss": 0.1796184778213501,
      "mean_token_accuracy": 0.9227155297994614,
      "num_tokens": 65563784.0,
      "step": 5315
    },
    {
      "entropy": 1.3516960740089417,
      "epoch": 2.7993680884676144,
      "grad_norm": 0.31386610865592957,
      "learning_rate": 0.00014279902672173427,
      "loss": 0.2113589197397232,
      "mean_token_accuracy": 0.9109881967306137,
      "num_tokens": 65576120.0,
      "step": 5316
    },
    {
      "entropy": 1.3502106964588165,
      "epoch": 2.799894681411269,
      "grad_norm": 0.3107154369354248,
      "learning_rate": 0.00014277860548317603,
      "loss": 0.1957147717475891,
      "mean_token_accuracy": 0.9212307333946228,
      "num_tokens": 65588456.0,
      "step": 5317
    },
    {
      "entropy": 1.286857783794403,
      "epoch": 2.8004212743549237,
      "grad_norm": 0.2969721853733063,
      "learning_rate": 0.00014275818229874301,
      "loss": 0.18805904686450958,
      "mean_token_accuracy": 0.9243521392345428,
      "num_tokens": 65600792.0,
      "step": 5318
    },
    {
      "entropy": 1.3212200999259949,
      "epoch": 2.800947867298578,
      "grad_norm": 0.28291863203048706,
      "learning_rate": 0.00014273775716964766,
      "loss": 0.18257564306259155,
      "mean_token_accuracy": 0.9234202355146408,
      "num_tokens": 65613128.0,
      "step": 5319
    },
    {
      "entropy": 1.2968879044055939,
      "epoch": 2.8014744602422326,
      "grad_norm": 0.27695876359939575,
      "learning_rate": 0.00014271733009710245,
      "loss": 0.1660367250442505,
      "mean_token_accuracy": 0.9291543513536453,
      "num_tokens": 65625464.0,
      "step": 5320
    },
    {
      "entropy": 1.2907300889492035,
      "epoch": 2.8020010531858874,
      "grad_norm": 0.29541856050491333,
      "learning_rate": 0.0001426969010823201,
      "loss": 0.21649664640426636,
      "mean_token_accuracy": 0.9106380492448807,
      "num_tokens": 65637800.0,
      "step": 5321
    },
    {
      "entropy": 1.315805196762085,
      "epoch": 2.802527646129542,
      "grad_norm": 0.28097328543663025,
      "learning_rate": 0.0001426764701265133,
      "loss": 0.188256174325943,
      "mean_token_accuracy": 0.9229840487241745,
      "num_tokens": 65650136.0,
      "step": 5322
    },
    {
      "entropy": 1.3322736322879791,
      "epoch": 2.8030542390731963,
      "grad_norm": 0.45639967918395996,
      "learning_rate": 0.00014265603723089484,
      "loss": 0.18657879531383514,
      "mean_token_accuracy": 0.9199790060520172,
      "num_tokens": 65662472.0,
      "step": 5323
    },
    {
      "entropy": 1.2863487601280212,
      "epoch": 2.803580832016851,
      "grad_norm": 0.2746271789073944,
      "learning_rate": 0.00014263560239667785,
      "loss": 0.18844881653785706,
      "mean_token_accuracy": 0.925334244966507,
      "num_tokens": 65674808.0,
      "step": 5324
    },
    {
      "entropy": 1.2828960418701172,
      "epoch": 2.8041074249605056,
      "grad_norm": 0.2776462733745575,
      "learning_rate": 0.00014261516562507533,
      "loss": 0.1838618367910385,
      "mean_token_accuracy": 0.9251666814088821,
      "num_tokens": 65687144.0,
      "step": 5325
    },
    {
      "entropy": 1.2915571928024292,
      "epoch": 2.80463401790416,
      "grad_norm": 0.28010231256484985,
      "learning_rate": 0.0001425947269173006,
      "loss": 0.1980217695236206,
      "mean_token_accuracy": 0.9215652793645859,
      "num_tokens": 65699480.0,
      "step": 5326
    },
    {
      "entropy": 1.2577224969863892,
      "epoch": 2.805160610847815,
      "grad_norm": 0.2838359475135803,
      "learning_rate": 0.00014257428627456682,
      "loss": 0.20210154354572296,
      "mean_token_accuracy": 0.9163983017206192,
      "num_tokens": 65711816.0,
      "step": 5327
    },
    {
      "entropy": 1.2922307252883911,
      "epoch": 2.8056872037914693,
      "grad_norm": 0.2706328332424164,
      "learning_rate": 0.00014255384369808756,
      "loss": 0.19801615178585052,
      "mean_token_accuracy": 0.915440022945404,
      "num_tokens": 65724152.0,
      "step": 5328
    },
    {
      "entropy": 1.2760253250598907,
      "epoch": 2.8062137967351237,
      "grad_norm": 0.2580713629722595,
      "learning_rate": 0.0001425333991890763,
      "loss": 0.16899164021015167,
      "mean_token_accuracy": 0.9279545992612839,
      "num_tokens": 65736488.0,
      "step": 5329
    },
    {
      "entropy": 1.2592010498046875,
      "epoch": 2.806740389678778,
      "grad_norm": 0.2732999622821808,
      "learning_rate": 0.00014251295274874678,
      "loss": 0.1941339373588562,
      "mean_token_accuracy": 0.9203190058469772,
      "num_tokens": 65748824.0,
      "step": 5330
    },
    {
      "entropy": 1.3266755938529968,
      "epoch": 2.807266982622433,
      "grad_norm": 0.2721918821334839,
      "learning_rate": 0.00014249250437831277,
      "loss": 0.18719229102134705,
      "mean_token_accuracy": 0.9207041710615158,
      "num_tokens": 65761160.0,
      "step": 5331
    },
    {
      "entropy": 1.3405596911907196,
      "epoch": 2.8077935755660874,
      "grad_norm": 0.29513224959373474,
      "learning_rate": 0.00014247205407898813,
      "loss": 0.20140871405601501,
      "mean_token_accuracy": 0.919532835483551,
      "num_tokens": 65773496.0,
      "step": 5332
    },
    {
      "entropy": 1.3203657865524292,
      "epoch": 2.808320168509742,
      "grad_norm": 0.26965105533599854,
      "learning_rate": 0.0001424516018519869,
      "loss": 0.1990198791027069,
      "mean_token_accuracy": 0.917657345533371,
      "num_tokens": 65785832.0,
      "step": 5333
    },
    {
      "entropy": 1.2802288830280304,
      "epoch": 2.8088467614533963,
      "grad_norm": 0.29502639174461365,
      "learning_rate": 0.00014243114769852322,
      "loss": 0.21779046952724457,
      "mean_token_accuracy": 0.9080840200185776,
      "num_tokens": 65798168.0,
      "step": 5334
    },
    {
      "entropy": 1.2174814939498901,
      "epoch": 2.809373354397051,
      "grad_norm": 0.23770982027053833,
      "learning_rate": 0.00014241069161981132,
      "loss": 0.18031489849090576,
      "mean_token_accuracy": 0.9241122305393219,
      "num_tokens": 65810504.0,
      "step": 5335
    },
    {
      "entropy": 1.303638517856598,
      "epoch": 2.8098999473407056,
      "grad_norm": 0.250471830368042,
      "learning_rate": 0.00014239023361706558,
      "loss": 0.18099047243595123,
      "mean_token_accuracy": 0.9224986284971237,
      "num_tokens": 65822840.0,
      "step": 5336
    },
    {
      "entropy": 1.3555626571178436,
      "epoch": 2.81042654028436,
      "grad_norm": 0.2877865433692932,
      "learning_rate": 0.00014236977369150043,
      "loss": 0.19794785976409912,
      "mean_token_accuracy": 0.916514053940773,
      "num_tokens": 65835176.0,
      "step": 5337
    },
    {
      "entropy": 1.3452889621257782,
      "epoch": 2.810953133228015,
      "grad_norm": 0.29638126492500305,
      "learning_rate": 0.0001423493118443305,
      "loss": 0.20515380799770355,
      "mean_token_accuracy": 0.9165419936180115,
      "num_tokens": 65847512.0,
      "step": 5338
    },
    {
      "entropy": 1.3289613127708435,
      "epoch": 2.8114797261716693,
      "grad_norm": 0.31008410453796387,
      "learning_rate": 0.00014232884807677045,
      "loss": 0.20547129213809967,
      "mean_token_accuracy": 0.9179491698741913,
      "num_tokens": 65859848.0,
      "step": 5339
    },
    {
      "entropy": 1.3396630585193634,
      "epoch": 2.8120063191153237,
      "grad_norm": 0.24820318818092346,
      "learning_rate": 0.00014230838239003514,
      "loss": 0.18229185044765472,
      "mean_token_accuracy": 0.9222933501005173,
      "num_tokens": 65872184.0,
      "step": 5340
    },
    {
      "entropy": 1.3142532706260681,
      "epoch": 2.8125329120589786,
      "grad_norm": 0.27396664023399353,
      "learning_rate": 0.00014228791478533944,
      "loss": 0.2007269263267517,
      "mean_token_accuracy": 0.9183091521263123,
      "num_tokens": 65884520.0,
      "step": 5341
    },
    {
      "entropy": 1.3794560730457306,
      "epoch": 2.813059505002633,
      "grad_norm": 0.276889443397522,
      "learning_rate": 0.0001422674452638985,
      "loss": 0.19162897765636444,
      "mean_token_accuracy": 0.9210671484470367,
      "num_tokens": 65896856.0,
      "step": 5342
    },
    {
      "entropy": 1.3106403052806854,
      "epoch": 2.8135860979462874,
      "grad_norm": 0.24311819672584534,
      "learning_rate": 0.00014224697382692733,
      "loss": 0.17249861359596252,
      "mean_token_accuracy": 0.926313728094101,
      "num_tokens": 65909192.0,
      "step": 5343
    },
    {
      "entropy": 1.357205718755722,
      "epoch": 2.8141126908899423,
      "grad_norm": 0.2881726622581482,
      "learning_rate": 0.00014222650047564128,
      "loss": 0.19397282600402832,
      "mean_token_accuracy": 0.9207228273153305,
      "num_tokens": 65921528.0,
      "step": 5344
    },
    {
      "entropy": 1.2870065569877625,
      "epoch": 2.8146392838335967,
      "grad_norm": 0.2502511739730835,
      "learning_rate": 0.00014220602521125577,
      "loss": 0.19269129633903503,
      "mean_token_accuracy": 0.9196359366178513,
      "num_tokens": 65933864.0,
      "step": 5345
    },
    {
      "entropy": 1.2929141223430634,
      "epoch": 2.815165876777251,
      "grad_norm": 0.2680942714214325,
      "learning_rate": 0.00014218554803498623,
      "loss": 0.19123870134353638,
      "mean_token_accuracy": 0.9213044196367264,
      "num_tokens": 65946200.0,
      "step": 5346
    },
    {
      "entropy": 1.2766457498073578,
      "epoch": 2.8156924697209056,
      "grad_norm": 0.266617089509964,
      "learning_rate": 0.0001421650689480483,
      "loss": 0.19216907024383545,
      "mean_token_accuracy": 0.923948273062706,
      "num_tokens": 65958536.0,
      "step": 5347
    },
    {
      "entropy": 1.279809296131134,
      "epoch": 2.8162190626645605,
      "grad_norm": 0.25692835450172424,
      "learning_rate": 0.00014214458795165768,
      "loss": 0.16894961893558502,
      "mean_token_accuracy": 0.9307194501161575,
      "num_tokens": 65970872.0,
      "step": 5348
    },
    {
      "entropy": 1.2850238978862762,
      "epoch": 2.816745655608215,
      "grad_norm": 0.2544446587562561,
      "learning_rate": 0.00014212410504703025,
      "loss": 0.18745863437652588,
      "mean_token_accuracy": 0.9202258735895157,
      "num_tokens": 65983208.0,
      "step": 5349
    },
    {
      "entropy": 1.3343496918678284,
      "epoch": 2.8172722485518693,
      "grad_norm": 0.2888110280036926,
      "learning_rate": 0.00014210362023538194,
      "loss": 0.19372104108333588,
      "mean_token_accuracy": 0.917907640337944,
      "num_tokens": 65995544.0,
      "step": 5350
    },
    {
      "entropy": 1.248584896326065,
      "epoch": 2.8177988414955237,
      "grad_norm": 0.258998841047287,
      "learning_rate": 0.0001420831335179288,
      "loss": 0.18754670023918152,
      "mean_token_accuracy": 0.9216349273920059,
      "num_tokens": 66007880.0,
      "step": 5351
    },
    {
      "entropy": 1.2730433344841003,
      "epoch": 2.8183254344391786,
      "grad_norm": 0.2844271957874298,
      "learning_rate": 0.000142062644895887,
      "loss": 0.19392099976539612,
      "mean_token_accuracy": 0.9160989373922348,
      "num_tokens": 66020216.0,
      "step": 5352
    },
    {
      "entropy": 1.239387720823288,
      "epoch": 2.818852027382833,
      "grad_norm": 0.2637011408805847,
      "learning_rate": 0.00014204215437047289,
      "loss": 0.16577531397342682,
      "mean_token_accuracy": 0.930683970451355,
      "num_tokens": 66032552.0,
      "step": 5353
    },
    {
      "entropy": 1.2617818415164948,
      "epoch": 2.8193786203264875,
      "grad_norm": 0.27310746908187866,
      "learning_rate": 0.0001420216619429028,
      "loss": 0.18738313019275665,
      "mean_token_accuracy": 0.922308012843132,
      "num_tokens": 66044888.0,
      "step": 5354
    },
    {
      "entropy": 1.2251103222370148,
      "epoch": 2.8199052132701423,
      "grad_norm": 0.2810131013393402,
      "learning_rate": 0.00014200116761439326,
      "loss": 0.18970806896686554,
      "mean_token_accuracy": 0.9238426834344864,
      "num_tokens": 66057224.0,
      "step": 5355
    },
    {
      "entropy": 1.300120770931244,
      "epoch": 2.8204318062137967,
      "grad_norm": 0.3125358819961548,
      "learning_rate": 0.00014198067138616096,
      "loss": 0.19446682929992676,
      "mean_token_accuracy": 0.9172956049442291,
      "num_tokens": 66069560.0,
      "step": 5356
    },
    {
      "entropy": 1.232959359884262,
      "epoch": 2.820958399157451,
      "grad_norm": 0.26678264141082764,
      "learning_rate": 0.00014196017325942257,
      "loss": 0.19019223749637604,
      "mean_token_accuracy": 0.919284850358963,
      "num_tokens": 66081896.0,
      "step": 5357
    },
    {
      "entropy": 1.2427859604358673,
      "epoch": 2.821484992101106,
      "grad_norm": 0.2783074975013733,
      "learning_rate": 0.00014193967323539497,
      "loss": 0.18141202628612518,
      "mean_token_accuracy": 0.9261755645275116,
      "num_tokens": 66094232.0,
      "step": 5358
    },
    {
      "entropy": 1.2050098180770874,
      "epoch": 2.8220115850447605,
      "grad_norm": 0.23615165054798126,
      "learning_rate": 0.0001419191713152951,
      "loss": 0.1723269373178482,
      "mean_token_accuracy": 0.9296863228082657,
      "num_tokens": 66106568.0,
      "step": 5359
    },
    {
      "entropy": 1.259114533662796,
      "epoch": 2.822538177988415,
      "grad_norm": 0.2952539324760437,
      "learning_rate": 0.00014189866750034016,
      "loss": 0.1994144320487976,
      "mean_token_accuracy": 0.9215876311063766,
      "num_tokens": 66118904.0,
      "step": 5360
    },
    {
      "entropy": 1.3054438531398773,
      "epoch": 2.8230647709320698,
      "grad_norm": 0.2893647849559784,
      "learning_rate": 0.00014187816179174717,
      "loss": 0.18889150023460388,
      "mean_token_accuracy": 0.9236411154270172,
      "num_tokens": 66131240.0,
      "step": 5361
    },
    {
      "entropy": 1.213493674993515,
      "epoch": 2.823591363875724,
      "grad_norm": 0.28194475173950195,
      "learning_rate": 0.00014185765419073352,
      "loss": 0.19393613934516907,
      "mean_token_accuracy": 0.9203670918941498,
      "num_tokens": 66143576.0,
      "step": 5362
    },
    {
      "entropy": 1.31081622838974,
      "epoch": 2.8241179568193786,
      "grad_norm": 0.26179248094558716,
      "learning_rate": 0.00014183714469851664,
      "loss": 0.17356306314468384,
      "mean_token_accuracy": 0.9289624840021133,
      "num_tokens": 66155912.0,
      "step": 5363
    },
    {
      "entropy": 1.2900842726230621,
      "epoch": 2.824644549763033,
      "grad_norm": 0.2782558798789978,
      "learning_rate": 0.00014181663331631404,
      "loss": 0.19263340532779694,
      "mean_token_accuracy": 0.9261777848005295,
      "num_tokens": 66168248.0,
      "step": 5364
    },
    {
      "entropy": 1.2860835790634155,
      "epoch": 2.8251711427066875,
      "grad_norm": 0.3041127920150757,
      "learning_rate": 0.00014179612004534334,
      "loss": 0.19561003148555756,
      "mean_token_accuracy": 0.9226749986410141,
      "num_tokens": 66180584.0,
      "step": 5365
    },
    {
      "entropy": 1.306857168674469,
      "epoch": 2.8256977356503423,
      "grad_norm": 0.2649271786212921,
      "learning_rate": 0.00014177560488682237,
      "loss": 0.18361037969589233,
      "mean_token_accuracy": 0.923071101307869,
      "num_tokens": 66192920.0,
      "step": 5366
    },
    {
      "entropy": 1.3504366278648376,
      "epoch": 2.8262243285939967,
      "grad_norm": 0.30323871970176697,
      "learning_rate": 0.00014175508784196892,
      "loss": 0.20840725302696228,
      "mean_token_accuracy": 0.9156047105789185,
      "num_tokens": 66205256.0,
      "step": 5367
    },
    {
      "entropy": 1.3343227803707123,
      "epoch": 2.826750921537651,
      "grad_norm": 0.30277058482170105,
      "learning_rate": 0.00014173456891200097,
      "loss": 0.2028798907995224,
      "mean_token_accuracy": 0.9148493260145187,
      "num_tokens": 66217592.0,
      "step": 5368
    },
    {
      "entropy": 1.2690812051296234,
      "epoch": 2.827277514481306,
      "grad_norm": 0.2847183346748352,
      "learning_rate": 0.0001417140480981366,
      "loss": 0.18881455063819885,
      "mean_token_accuracy": 0.9242660850286484,
      "num_tokens": 66229928.0,
      "step": 5369
    },
    {
      "entropy": 1.3083016276359558,
      "epoch": 2.8278041074249605,
      "grad_norm": 0.30137863755226135,
      "learning_rate": 0.00014169352540159406,
      "loss": 0.19169145822525024,
      "mean_token_accuracy": 0.9188456684350967,
      "num_tokens": 66242264.0,
      "step": 5370
    },
    {
      "entropy": 1.3139874637126923,
      "epoch": 2.828330700368615,
      "grad_norm": 0.2890307307243347,
      "learning_rate": 0.00014167300082359167,
      "loss": 0.19907841086387634,
      "mean_token_accuracy": 0.9193234592676163,
      "num_tokens": 66254600.0,
      "step": 5371
    },
    {
      "entropy": 1.2936797440052032,
      "epoch": 2.8288572933122698,
      "grad_norm": 0.27607885003089905,
      "learning_rate": 0.00014165247436534782,
      "loss": 0.1906982958316803,
      "mean_token_accuracy": 0.9183892905712128,
      "num_tokens": 66266936.0,
      "step": 5372
    },
    {
      "entropy": 1.3275218605995178,
      "epoch": 2.829383886255924,
      "grad_norm": 0.29767462611198425,
      "learning_rate": 0.000141631946028081,
      "loss": 0.21200454235076904,
      "mean_token_accuracy": 0.9133586883544922,
      "num_tokens": 66279272.0,
      "step": 5373
    },
    {
      "entropy": 1.3114037215709686,
      "epoch": 2.8299104791995786,
      "grad_norm": 0.2859516441822052,
      "learning_rate": 0.00014161141581300993,
      "loss": 0.2003038227558136,
      "mean_token_accuracy": 0.9200000613927841,
      "num_tokens": 66291608.0,
      "step": 5374
    },
    {
      "entropy": 1.2808501720428467,
      "epoch": 2.8304370721432335,
      "grad_norm": 0.2558257579803467,
      "learning_rate": 0.0001415908837213534,
      "loss": 0.17073266208171844,
      "mean_token_accuracy": 0.9285646677017212,
      "num_tokens": 66303944.0,
      "step": 5375
    },
    {
      "entropy": 1.2928644716739655,
      "epoch": 2.830963665086888,
      "grad_norm": 0.2917443811893463,
      "learning_rate": 0.00014157034975433017,
      "loss": 0.20764575898647308,
      "mean_token_accuracy": 0.915138840675354,
      "num_tokens": 66316280.0,
      "step": 5376
    },
    {
      "entropy": 1.2757077515125275,
      "epoch": 2.8314902580305423,
      "grad_norm": 0.2791418731212616,
      "learning_rate": 0.00014154981391315932,
      "loss": 0.18959811329841614,
      "mean_token_accuracy": 0.9251828342676163,
      "num_tokens": 66328616.0,
      "step": 5377
    },
    {
      "entropy": 1.319762110710144,
      "epoch": 2.832016850974197,
      "grad_norm": 0.26925116777420044,
      "learning_rate": 0.00014152927619905987,
      "loss": 0.17835290729999542,
      "mean_token_accuracy": 0.9292363375425339,
      "num_tokens": 66340952.0,
      "step": 5378
    },
    {
      "entropy": 1.2947478890419006,
      "epoch": 2.8325434439178516,
      "grad_norm": 0.2954257130622864,
      "learning_rate": 0.00014150873661325107,
      "loss": 0.19611039757728577,
      "mean_token_accuracy": 0.9170423597097397,
      "num_tokens": 66353288.0,
      "step": 5379
    },
    {
      "entropy": 1.3365015387535095,
      "epoch": 2.833070036861506,
      "grad_norm": 0.28138986229896545,
      "learning_rate": 0.00014148819515695226,
      "loss": 0.19645000994205475,
      "mean_token_accuracy": 0.9223781526088715,
      "num_tokens": 66365624.0,
      "step": 5380
    },
    {
      "entropy": 1.2312410771846771,
      "epoch": 2.8335966298051605,
      "grad_norm": 0.25112465023994446,
      "learning_rate": 0.0001414676518313828,
      "loss": 0.17653129994869232,
      "mean_token_accuracy": 0.9302770644426346,
      "num_tokens": 66377960.0,
      "step": 5381
    },
    {
      "entropy": 1.29915651679039,
      "epoch": 2.834123222748815,
      "grad_norm": 0.28604868054389954,
      "learning_rate": 0.00014144710663776224,
      "loss": 0.18507474660873413,
      "mean_token_accuracy": 0.9225872606039047,
      "num_tokens": 66390296.0,
      "step": 5382
    },
    {
      "entropy": 1.2989193797111511,
      "epoch": 2.8346498156924698,
      "grad_norm": 0.29358354210853577,
      "learning_rate": 0.00014142655957731027,
      "loss": 0.20054125785827637,
      "mean_token_accuracy": 0.9196604937314987,
      "num_tokens": 66402632.0,
      "step": 5383
    },
    {
      "entropy": 1.3038256466388702,
      "epoch": 2.835176408636124,
      "grad_norm": 0.2762550115585327,
      "learning_rate": 0.0001414060106512466,
      "loss": 0.18558046221733093,
      "mean_token_accuracy": 0.9259804487228394,
      "num_tokens": 66414968.0,
      "step": 5384
    },
    {
      "entropy": 1.2950901687145233,
      "epoch": 2.8357030015797786,
      "grad_norm": 0.2872573435306549,
      "learning_rate": 0.00014138545986079119,
      "loss": 0.1845257580280304,
      "mean_token_accuracy": 0.9250736236572266,
      "num_tokens": 66427304.0,
      "step": 5385
    },
    {
      "entropy": 1.243366152048111,
      "epoch": 2.8362295945234335,
      "grad_norm": 0.272707998752594,
      "learning_rate": 0.0001413649072071639,
      "loss": 0.19566035270690918,
      "mean_token_accuracy": 0.9192866683006287,
      "num_tokens": 66439640.0,
      "step": 5386
    },
    {
      "entropy": 1.283011943101883,
      "epoch": 2.836756187467088,
      "grad_norm": 0.24862247705459595,
      "learning_rate": 0.00014134435269158487,
      "loss": 0.18011629581451416,
      "mean_token_accuracy": 0.9231017231941223,
      "num_tokens": 66451976.0,
      "step": 5387
    },
    {
      "entropy": 1.2681494057178497,
      "epoch": 2.8372827804107423,
      "grad_norm": 0.2905445396900177,
      "learning_rate": 0.00014132379631527433,
      "loss": 0.19375208020210266,
      "mean_token_accuracy": 0.919303685426712,
      "num_tokens": 66464312.0,
      "step": 5388
    },
    {
      "entropy": 1.2690399587154388,
      "epoch": 2.837809373354397,
      "grad_norm": 0.32442450523376465,
      "learning_rate": 0.0001413032380794526,
      "loss": 0.22549712657928467,
      "mean_token_accuracy": 0.9056236445903778,
      "num_tokens": 66476648.0,
      "step": 5389
    },
    {
      "entropy": 1.2229408919811249,
      "epoch": 2.8383359662980516,
      "grad_norm": 0.2401140034198761,
      "learning_rate": 0.00014128267798534002,
      "loss": 0.16507481038570404,
      "mean_token_accuracy": 0.9342491179704666,
      "num_tokens": 66488984.0,
      "step": 5390
    },
    {
      "entropy": 1.2351162135601044,
      "epoch": 2.838862559241706,
      "grad_norm": 0.2740425765514374,
      "learning_rate": 0.00014126211603415723,
      "loss": 0.1834060549736023,
      "mean_token_accuracy": 0.9211757332086563,
      "num_tokens": 66501320.0,
      "step": 5391
    },
    {
      "entropy": 1.2271034121513367,
      "epoch": 2.839389152185361,
      "grad_norm": 0.2942942976951599,
      "learning_rate": 0.00014124155222712477,
      "loss": 0.20962479710578918,
      "mean_token_accuracy": 0.9142883569002151,
      "num_tokens": 66513656.0,
      "step": 5392
    },
    {
      "entropy": 1.2456031441688538,
      "epoch": 2.8399157451290153,
      "grad_norm": 0.2792077958583832,
      "learning_rate": 0.00014122098656546347,
      "loss": 0.19956699013710022,
      "mean_token_accuracy": 0.9134206175804138,
      "num_tokens": 66525992.0,
      "step": 5393
    },
    {
      "entropy": 1.2200534343719482,
      "epoch": 2.8404423380726698,
      "grad_norm": 0.2667393386363983,
      "learning_rate": 0.00014120041905039417,
      "loss": 0.18422703444957733,
      "mean_token_accuracy": 0.9217170625925064,
      "num_tokens": 66538328.0,
      "step": 5394
    },
    {
      "entropy": 1.2396110892295837,
      "epoch": 2.8409689310163246,
      "grad_norm": 0.29442811012268066,
      "learning_rate": 0.00014117984968313785,
      "loss": 0.19714084267616272,
      "mean_token_accuracy": 0.920742392539978,
      "num_tokens": 66550664.0,
      "step": 5395
    },
    {
      "entropy": 1.1788528263568878,
      "epoch": 2.841495523959979,
      "grad_norm": 0.2758336067199707,
      "learning_rate": 0.00014115927846491556,
      "loss": 0.19068320095539093,
      "mean_token_accuracy": 0.9210142344236374,
      "num_tokens": 66563000.0,
      "step": 5396
    },
    {
      "entropy": 1.1362981796264648,
      "epoch": 2.8420221169036335,
      "grad_norm": 0.25416529178619385,
      "learning_rate": 0.00014113870539694852,
      "loss": 0.16557230055332184,
      "mean_token_accuracy": 0.9342633932828903,
      "num_tokens": 66575336.0,
      "step": 5397
    },
    {
      "entropy": 1.1327734589576721,
      "epoch": 2.842548709847288,
      "grad_norm": 0.2415960431098938,
      "learning_rate": 0.00014111813048045804,
      "loss": 0.16808874905109406,
      "mean_token_accuracy": 0.9288180619478226,
      "num_tokens": 66587672.0,
      "step": 5398
    },
    {
      "entropy": 1.1094682216644287,
      "epoch": 2.8430753027909423,
      "grad_norm": 0.2572054862976074,
      "learning_rate": 0.00014109755371666553,
      "loss": 0.18255862593650818,
      "mean_token_accuracy": 0.9252709001302719,
      "num_tokens": 66600008.0,
      "step": 5399
    },
    {
      "entropy": 1.167961448431015,
      "epoch": 2.843601895734597,
      "grad_norm": 0.26458224654197693,
      "learning_rate": 0.0001410769751067925,
      "loss": 0.18526852130889893,
      "mean_token_accuracy": 0.9216188937425613,
      "num_tokens": 66612344.0,
      "step": 5400
    },
    {
      "entropy": 1.1674366891384125,
      "epoch": 2.8441284886782516,
      "grad_norm": 0.2747248411178589,
      "learning_rate": 0.00014105639465206057,
      "loss": 0.17862823605537415,
      "mean_token_accuracy": 0.9248768389225006,
      "num_tokens": 66624680.0,
      "step": 5401
    },
    {
      "entropy": 1.1781964004039764,
      "epoch": 2.844655081621906,
      "grad_norm": 0.2930940091609955,
      "learning_rate": 0.00014103581235369152,
      "loss": 0.19545236229896545,
      "mean_token_accuracy": 0.921414703130722,
      "num_tokens": 66637016.0,
      "step": 5402
    },
    {
      "entropy": 1.137371987104416,
      "epoch": 2.845181674565561,
      "grad_norm": 0.3015097975730896,
      "learning_rate": 0.00014101522821290718,
      "loss": 0.1982129067182541,
      "mean_token_accuracy": 0.9206677973270416,
      "num_tokens": 66649352.0,
      "step": 5403
    },
    {
      "entropy": 1.1155942976474762,
      "epoch": 2.8457082675092154,
      "grad_norm": 0.2667999863624573,
      "learning_rate": 0.00014099464223092951,
      "loss": 0.18535880744457245,
      "mean_token_accuracy": 0.9235810935497284,
      "num_tokens": 66661688.0,
      "step": 5404
    },
    {
      "entropy": 1.1362461149692535,
      "epoch": 2.8462348604528698,
      "grad_norm": 0.28863558173179626,
      "learning_rate": 0.00014097405440898059,
      "loss": 0.20071467757225037,
      "mean_token_accuracy": 0.9162945002317429,
      "num_tokens": 66674024.0,
      "step": 5405
    },
    {
      "entropy": 1.1156060993671417,
      "epoch": 2.8467614533965246,
      "grad_norm": 0.28235065937042236,
      "learning_rate": 0.0001409534647482826,
      "loss": 0.18811024725437164,
      "mean_token_accuracy": 0.9257550984621048,
      "num_tokens": 66686360.0,
      "step": 5406
    },
    {
      "entropy": 1.1049256920814514,
      "epoch": 2.847288046340179,
      "grad_norm": 0.27387118339538574,
      "learning_rate": 0.0001409328732500578,
      "loss": 0.18561598658561707,
      "mean_token_accuracy": 0.9243172854185104,
      "num_tokens": 66698696.0,
      "step": 5407
    },
    {
      "entropy": 1.0878219902515411,
      "epoch": 2.8478146392838335,
      "grad_norm": 0.24213702976703644,
      "learning_rate": 0.0001409122799155286,
      "loss": 0.17320024967193604,
      "mean_token_accuracy": 0.9250019490718842,
      "num_tokens": 66711032.0,
      "step": 5408
    },
    {
      "entropy": 1.1544892191886902,
      "epoch": 2.8483412322274884,
      "grad_norm": 0.297046959400177,
      "learning_rate": 0.00014089168474591752,
      "loss": 0.2007860541343689,
      "mean_token_accuracy": 0.9169928729534149,
      "num_tokens": 66723368.0,
      "step": 5409
    },
    {
      "entropy": 1.09200057387352,
      "epoch": 2.848867825171143,
      "grad_norm": 0.26236802339553833,
      "learning_rate": 0.00014087108774244714,
      "loss": 0.171499103307724,
      "mean_token_accuracy": 0.9301941990852356,
      "num_tokens": 66735704.0,
      "step": 5410
    },
    {
      "entropy": 1.127855896949768,
      "epoch": 2.849394418114797,
      "grad_norm": 0.2707575261592865,
      "learning_rate": 0.00014085048890634023,
      "loss": 0.18420180678367615,
      "mean_token_accuracy": 0.9241728633642197,
      "num_tokens": 66748040.0,
      "step": 5411
    },
    {
      "entropy": 1.1189501583576202,
      "epoch": 2.849921011058452,
      "grad_norm": 0.2789716124534607,
      "learning_rate": 0.00014082988823881963,
      "loss": 0.18347470462322235,
      "mean_token_accuracy": 0.9260276108980179,
      "num_tokens": 66760376.0,
      "step": 5412
    },
    {
      "entropy": 1.1610067784786224,
      "epoch": 2.8504476040021065,
      "grad_norm": 0.28152376413345337,
      "learning_rate": 0.0001408092857411082,
      "loss": 0.18292734026908875,
      "mean_token_accuracy": 0.9208507537841797,
      "num_tokens": 66772712.0,
      "step": 5413
    },
    {
      "entropy": 1.180832177400589,
      "epoch": 2.850974196945761,
      "grad_norm": 0.27147287130355835,
      "learning_rate": 0.00014078868141442912,
      "loss": 0.17122916877269745,
      "mean_token_accuracy": 0.9286279231309891,
      "num_tokens": 66785048.0,
      "step": 5414
    },
    {
      "entropy": 1.1299280226230621,
      "epoch": 2.8515007898894154,
      "grad_norm": 0.3026059865951538,
      "learning_rate": 0.00014076807526000539,
      "loss": 0.19383591413497925,
      "mean_token_accuracy": 0.9125325083732605,
      "num_tokens": 66797384.0,
      "step": 5415
    },
    {
      "entropy": 1.1946425437927246,
      "epoch": 2.85202738283307,
      "grad_norm": 0.309876412153244,
      "learning_rate": 0.00014074746727906046,
      "loss": 0.21104586124420166,
      "mean_token_accuracy": 0.9126588106155396,
      "num_tokens": 66809720.0,
      "step": 5416
    },
    {
      "entropy": 1.123658001422882,
      "epoch": 2.8525539757767246,
      "grad_norm": 0.25807276368141174,
      "learning_rate": 0.00014072685747281753,
      "loss": 0.17802190780639648,
      "mean_token_accuracy": 0.9260611236095428,
      "num_tokens": 66822056.0,
      "step": 5417
    },
    {
      "entropy": 1.1184225380420685,
      "epoch": 2.853080568720379,
      "grad_norm": 0.2761235237121582,
      "learning_rate": 0.00014070624584250018,
      "loss": 0.20021545886993408,
      "mean_token_accuracy": 0.9165898710489273,
      "num_tokens": 66834392.0,
      "step": 5418
    },
    {
      "entropy": 1.1407923996448517,
      "epoch": 2.8536071616640335,
      "grad_norm": 0.2857515811920166,
      "learning_rate": 0.000140685632389332,
      "loss": 0.19885635375976562,
      "mean_token_accuracy": 0.9142837524414062,
      "num_tokens": 66846728.0,
      "step": 5419
    },
    {
      "entropy": 1.138054519891739,
      "epoch": 2.8541337546076884,
      "grad_norm": 0.27679088711738586,
      "learning_rate": 0.00014066501711453664,
      "loss": 0.19095392525196075,
      "mean_token_accuracy": 0.9210739135742188,
      "num_tokens": 66859064.0,
      "step": 5420
    },
    {
      "entropy": 1.1621717810630798,
      "epoch": 2.854660347551343,
      "grad_norm": 0.2896009385585785,
      "learning_rate": 0.00014064440001933798,
      "loss": 0.1932981014251709,
      "mean_token_accuracy": 0.9193224459886551,
      "num_tokens": 66871400.0,
      "step": 5421
    },
    {
      "entropy": 1.1080684959888458,
      "epoch": 2.855186940494997,
      "grad_norm": 0.2747195065021515,
      "learning_rate": 0.00014062378110495989,
      "loss": 0.19618317484855652,
      "mean_token_accuracy": 0.919920340180397,
      "num_tokens": 66883736.0,
      "step": 5422
    },
    {
      "entropy": 1.1182142198085785,
      "epoch": 2.855713533438652,
      "grad_norm": 0.2994195222854614,
      "learning_rate": 0.0001406031603726264,
      "loss": 0.19491897523403168,
      "mean_token_accuracy": 0.9171514958143234,
      "num_tokens": 66896072.0,
      "step": 5423
    },
    {
      "entropy": 1.0934617221355438,
      "epoch": 2.8562401263823065,
      "grad_norm": 0.2570323646068573,
      "learning_rate": 0.00014058253782356166,
      "loss": 0.17098519206047058,
      "mean_token_accuracy": 0.9273951053619385,
      "num_tokens": 66908408.0,
      "step": 5424
    },
    {
      "entropy": 1.0887425243854523,
      "epoch": 2.856766719325961,
      "grad_norm": 0.24847312271595,
      "learning_rate": 0.00014056191345898986,
      "loss": 0.1830718219280243,
      "mean_token_accuracy": 0.9246447682380676,
      "num_tokens": 66920744.0,
      "step": 5425
    },
    {
      "entropy": 1.0863420367240906,
      "epoch": 2.857293312269616,
      "grad_norm": 0.27926573157310486,
      "learning_rate": 0.0001405412872801354,
      "loss": 0.1928982138633728,
      "mean_token_accuracy": 0.9180453270673752,
      "num_tokens": 66933080.0,
      "step": 5426
    },
    {
      "entropy": 1.0742215812206268,
      "epoch": 2.8578199052132702,
      "grad_norm": 0.26996538043022156,
      "learning_rate": 0.00014052065928822275,
      "loss": 0.19069387018680573,
      "mean_token_accuracy": 0.9181806743144989,
      "num_tokens": 66945416.0,
      "step": 5427
    },
    {
      "entropy": 1.0830220580101013,
      "epoch": 2.8583464981569247,
      "grad_norm": 0.28085827827453613,
      "learning_rate": 0.00014050002948447644,
      "loss": 0.2127879559993744,
      "mean_token_accuracy": 0.9116558134555817,
      "num_tokens": 66957752.0,
      "step": 5428
    },
    {
      "entropy": 1.1039027869701385,
      "epoch": 2.8588730911005795,
      "grad_norm": 0.27143600583076477,
      "learning_rate": 0.00014047939787012115,
      "loss": 0.1901354193687439,
      "mean_token_accuracy": 0.9206545948982239,
      "num_tokens": 66970088.0,
      "step": 5429
    },
    {
      "entropy": 1.170148491859436,
      "epoch": 2.859399684044234,
      "grad_norm": 0.30527952313423157,
      "learning_rate": 0.00014045876444638165,
      "loss": 0.19135817885398865,
      "mean_token_accuracy": 0.9210909754037857,
      "num_tokens": 66982424.0,
      "step": 5430
    },
    {
      "entropy": 1.122731328010559,
      "epoch": 2.8599262769878884,
      "grad_norm": 0.2801341116428375,
      "learning_rate": 0.0001404381292144828,
      "loss": 0.1940993219614029,
      "mean_token_accuracy": 0.9169181287288666,
      "num_tokens": 66994760.0,
      "step": 5431
    },
    {
      "entropy": 1.1264297366142273,
      "epoch": 2.860452869931543,
      "grad_norm": 0.27544909715652466,
      "learning_rate": 0.00014041749217564965,
      "loss": 0.17596957087516785,
      "mean_token_accuracy": 0.9287864416837692,
      "num_tokens": 67007096.0,
      "step": 5432
    },
    {
      "entropy": 1.072750210762024,
      "epoch": 2.860979462875197,
      "grad_norm": 0.2531747817993164,
      "learning_rate": 0.0001403968533311073,
      "loss": 0.1748124659061432,
      "mean_token_accuracy": 0.9279569089412689,
      "num_tokens": 67019432.0,
      "step": 5433
    },
    {
      "entropy": 1.06135892868042,
      "epoch": 2.861506055818852,
      "grad_norm": 0.2722875773906708,
      "learning_rate": 0.00014037621268208093,
      "loss": 0.1929028183221817,
      "mean_token_accuracy": 0.9204573929309845,
      "num_tokens": 67031768.0,
      "step": 5434
    },
    {
      "entropy": 1.1115979850292206,
      "epoch": 2.8620326487625065,
      "grad_norm": 0.30371981859207153,
      "learning_rate": 0.00014035557022979583,
      "loss": 0.21404585242271423,
      "mean_token_accuracy": 0.9116851538419724,
      "num_tokens": 67044104.0,
      "step": 5435
    },
    {
      "entropy": 1.1167850196361542,
      "epoch": 2.862559241706161,
      "grad_norm": 0.3018032908439636,
      "learning_rate": 0.00014033492597547744,
      "loss": 0.21622182428836823,
      "mean_token_accuracy": 0.9091343581676483,
      "num_tokens": 67056440.0,
      "step": 5436
    },
    {
      "entropy": 1.1104548573493958,
      "epoch": 2.863085834649816,
      "grad_norm": 0.25832709670066833,
      "learning_rate": 0.0001403142799203513,
      "loss": 0.18880854547023773,
      "mean_token_accuracy": 0.9243634939193726,
      "num_tokens": 67068776.0,
      "step": 5437
    },
    {
      "entropy": 1.1190284192562103,
      "epoch": 2.8636124275934702,
      "grad_norm": 0.2891407310962677,
      "learning_rate": 0.00014029363206564306,
      "loss": 0.19353890419006348,
      "mean_token_accuracy": 0.9211379885673523,
      "num_tokens": 67081112.0,
      "step": 5438
    },
    {
      "entropy": 1.1418761312961578,
      "epoch": 2.8641390205371247,
      "grad_norm": 0.26448798179626465,
      "learning_rate": 0.0001402729824125785,
      "loss": 0.1813502162694931,
      "mean_token_accuracy": 0.9249635934829712,
      "num_tokens": 67093448.0,
      "step": 5439
    },
    {
      "entropy": 1.1411679685115814,
      "epoch": 2.8646656134807795,
      "grad_norm": 0.2673298716545105,
      "learning_rate": 0.00014025233096238337,
      "loss": 0.2000432014465332,
      "mean_token_accuracy": 0.92094986140728,
      "num_tokens": 67105784.0,
      "step": 5440
    },
    {
      "entropy": 1.126186192035675,
      "epoch": 2.865192206424434,
      "grad_norm": 0.26789215207099915,
      "learning_rate": 0.0001402316777162837,
      "loss": 0.1887417435646057,
      "mean_token_accuracy": 0.9270607680082321,
      "num_tokens": 67118120.0,
      "step": 5441
    },
    {
      "entropy": 1.1385208070278168,
      "epoch": 2.8657187993680884,
      "grad_norm": 0.25131142139434814,
      "learning_rate": 0.0001402110226755055,
      "loss": 0.1738722175359726,
      "mean_token_accuracy": 0.930255264043808,
      "num_tokens": 67130456.0,
      "step": 5442
    },
    {
      "entropy": 1.122873455286026,
      "epoch": 2.8662453923117432,
      "grad_norm": 0.26481401920318604,
      "learning_rate": 0.000140190365841275,
      "loss": 0.18120309710502625,
      "mean_token_accuracy": 0.9251967370510101,
      "num_tokens": 67142792.0,
      "step": 5443
    },
    {
      "entropy": 1.1093613505363464,
      "epoch": 2.8667719852553977,
      "grad_norm": 0.2555820941925049,
      "learning_rate": 0.00014016970721481838,
      "loss": 0.1706184446811676,
      "mean_token_accuracy": 0.9285883456468582,
      "num_tokens": 67155128.0,
      "step": 5444
    },
    {
      "entropy": 1.1308659613132477,
      "epoch": 2.867298578199052,
      "grad_norm": 0.2681415379047394,
      "learning_rate": 0.00014014904679736213,
      "loss": 0.19834303855895996,
      "mean_token_accuracy": 0.9212728142738342,
      "num_tokens": 67167464.0,
      "step": 5445
    },
    {
      "entropy": 1.117288589477539,
      "epoch": 2.867825171142707,
      "grad_norm": 0.28140929341316223,
      "learning_rate": 0.0001401283845901327,
      "loss": 0.2111014872789383,
      "mean_token_accuracy": 0.9120144248008728,
      "num_tokens": 67179800.0,
      "step": 5446
    },
    {
      "entropy": 1.141126126050949,
      "epoch": 2.8683517640863614,
      "grad_norm": 0.2594495117664337,
      "learning_rate": 0.00014010772059435664,
      "loss": 0.18402037024497986,
      "mean_token_accuracy": 0.921778216958046,
      "num_tokens": 67192136.0,
      "step": 5447
    },
    {
      "entropy": 1.1240050196647644,
      "epoch": 2.868878357030016,
      "grad_norm": 0.2530897855758667,
      "learning_rate": 0.00014008705481126073,
      "loss": 0.1877949982881546,
      "mean_token_accuracy": 0.9208306074142456,
      "num_tokens": 67204472.0,
      "step": 5448
    },
    {
      "entropy": 1.2034168243408203,
      "epoch": 2.8694049499736702,
      "grad_norm": 0.28608205914497375,
      "learning_rate": 0.00014006638724207174,
      "loss": 0.19868814945220947,
      "mean_token_accuracy": 0.9173824042081833,
      "num_tokens": 67216808.0,
      "step": 5449
    },
    {
      "entropy": 1.2032455205917358,
      "epoch": 2.8699315429173247,
      "grad_norm": 0.29074713587760925,
      "learning_rate": 0.00014004571788801652,
      "loss": 0.20050571858882904,
      "mean_token_accuracy": 0.9168554842472076,
      "num_tokens": 67229144.0,
      "step": 5450
    },
    {
      "entropy": 1.185323804616928,
      "epoch": 2.8704581358609795,
      "grad_norm": 0.2699410319328308,
      "learning_rate": 0.0001400250467503222,
      "loss": 0.19191224873065948,
      "mean_token_accuracy": 0.9225668460130692,
      "num_tokens": 67241480.0,
      "step": 5451
    },
    {
      "entropy": 1.2091895043849945,
      "epoch": 2.870984728804634,
      "grad_norm": 0.268320232629776,
      "learning_rate": 0.00014000437383021586,
      "loss": 0.18603497743606567,
      "mean_token_accuracy": 0.9205420464277267,
      "num_tokens": 67253816.0,
      "step": 5452
    },
    {
      "entropy": 1.2147535979747772,
      "epoch": 2.8715113217482884,
      "grad_norm": 0.2614723742008209,
      "learning_rate": 0.0001399836991289247,
      "loss": 0.18065370619297028,
      "mean_token_accuracy": 0.9245912879705429,
      "num_tokens": 67266152.0,
      "step": 5453
    },
    {
      "entropy": 1.1979366540908813,
      "epoch": 2.8720379146919433,
      "grad_norm": 0.2876172661781311,
      "learning_rate": 0.00013996302264767608,
      "loss": 0.19929926097393036,
      "mean_token_accuracy": 0.9158833473920822,
      "num_tokens": 67278488.0,
      "step": 5454
    },
    {
      "entropy": 1.2220218181610107,
      "epoch": 2.8725645076355977,
      "grad_norm": 0.2671680748462677,
      "learning_rate": 0.00013994234438769747,
      "loss": 0.18833203613758087,
      "mean_token_accuracy": 0.9217775911092758,
      "num_tokens": 67290824.0,
      "step": 5455
    },
    {
      "entropy": 1.1728639602661133,
      "epoch": 2.873091100579252,
      "grad_norm": 0.3314996361732483,
      "learning_rate": 0.00013992166435021635,
      "loss": 0.20119228959083557,
      "mean_token_accuracy": 0.9148829132318497,
      "num_tokens": 67303160.0,
      "step": 5456
    },
    {
      "entropy": 1.227847933769226,
      "epoch": 2.873617693522907,
      "grad_norm": 0.27688175439834595,
      "learning_rate": 0.00013990098253646043,
      "loss": 0.18525132536888123,
      "mean_token_accuracy": 0.9234299957752228,
      "num_tokens": 67315496.0,
      "step": 5457
    },
    {
      "entropy": 1.229965090751648,
      "epoch": 2.8741442864665614,
      "grad_norm": 0.28945228457450867,
      "learning_rate": 0.00013988029894765748,
      "loss": 0.19980910420417786,
      "mean_token_accuracy": 0.9183862209320068,
      "num_tokens": 67327832.0,
      "step": 5458
    },
    {
      "entropy": 1.1130054593086243,
      "epoch": 2.874670879410216,
      "grad_norm": 0.26660850644111633,
      "learning_rate": 0.0001398596135850353,
      "loss": 0.18399474024772644,
      "mean_token_accuracy": 0.9241610169410706,
      "num_tokens": 67340168.0,
      "step": 5459
    },
    {
      "entropy": 1.1546690165996552,
      "epoch": 2.8751974723538707,
      "grad_norm": 0.28937751054763794,
      "learning_rate": 0.0001398389264498219,
      "loss": 0.2176598459482193,
      "mean_token_accuracy": 0.9133116751909256,
      "num_tokens": 67352504.0,
      "step": 5460
    },
    {
      "entropy": 1.1990187168121338,
      "epoch": 2.875724065297525,
      "grad_norm": 0.28130099177360535,
      "learning_rate": 0.00013981823754324538,
      "loss": 0.19636386632919312,
      "mean_token_accuracy": 0.9205514341592789,
      "num_tokens": 67364840.0,
      "step": 5461
    },
    {
      "entropy": 1.1869704127311707,
      "epoch": 2.8762506582411795,
      "grad_norm": 0.28470396995544434,
      "learning_rate": 0.00013979754686653387,
      "loss": 0.18688493967056274,
      "mean_token_accuracy": 0.9220611900091171,
      "num_tokens": 67377176.0,
      "step": 5462
    },
    {
      "entropy": 1.1728688478469849,
      "epoch": 2.876777251184834,
      "grad_norm": 0.2566015422344208,
      "learning_rate": 0.0001397768544209157,
      "loss": 0.179299995303154,
      "mean_token_accuracy": 0.9220699667930603,
      "num_tokens": 67389512.0,
      "step": 5463
    },
    {
      "entropy": 1.1600215435028076,
      "epoch": 2.877303844128489,
      "grad_norm": 0.27697089314460754,
      "learning_rate": 0.00013975616020761922,
      "loss": 0.18880513310432434,
      "mean_token_accuracy": 0.9230993837118149,
      "num_tokens": 67401848.0,
      "step": 5464
    },
    {
      "entropy": 1.1880110800266266,
      "epoch": 2.8778304370721433,
      "grad_norm": 0.30111849308013916,
      "learning_rate": 0.0001397354642278729,
      "loss": 0.19892379641532898,
      "mean_token_accuracy": 0.9172447323799133,
      "num_tokens": 67414184.0,
      "step": 5465
    },
    {
      "entropy": 1.1921871602535248,
      "epoch": 2.8783570300157977,
      "grad_norm": 0.2692202925682068,
      "learning_rate": 0.00013971476648290545,
      "loss": 0.18941126763820648,
      "mean_token_accuracy": 0.9192425012588501,
      "num_tokens": 67426520.0,
      "step": 5466
    },
    {
      "entropy": 1.2540110051631927,
      "epoch": 2.878883622959452,
      "grad_norm": 0.29185977578163147,
      "learning_rate": 0.00013969406697394548,
      "loss": 0.19037245213985443,
      "mean_token_accuracy": 0.9188786596059799,
      "num_tokens": 67438856.0,
      "step": 5467
    },
    {
      "entropy": 1.2021810710430145,
      "epoch": 2.879410215903107,
      "grad_norm": 0.2722068727016449,
      "learning_rate": 0.00013967336570222183,
      "loss": 0.19480682909488678,
      "mean_token_accuracy": 0.9211893379688263,
      "num_tokens": 67451192.0,
      "step": 5468
    },
    {
      "entropy": 1.2452360689640045,
      "epoch": 2.8799368088467614,
      "grad_norm": 0.30236995220184326,
      "learning_rate": 0.0001396526626689634,
      "loss": 0.19539473950862885,
      "mean_token_accuracy": 0.9220133870840073,
      "num_tokens": 67463528.0,
      "step": 5469
    },
    {
      "entropy": 1.1860556304454803,
      "epoch": 2.880463401790416,
      "grad_norm": 0.2610684335231781,
      "learning_rate": 0.0001396319578753992,
      "loss": 0.2000972032546997,
      "mean_token_accuracy": 0.9178863316774368,
      "num_tokens": 67475864.0,
      "step": 5470
    },
    {
      "entropy": 1.1721631586551666,
      "epoch": 2.8809899947340707,
      "grad_norm": 0.2493157833814621,
      "learning_rate": 0.00013961125132275836,
      "loss": 0.17627960443496704,
      "mean_token_accuracy": 0.9253731369972229,
      "num_tokens": 67488200.0,
      "step": 5471
    },
    {
      "entropy": 1.1537190973758698,
      "epoch": 2.881516587677725,
      "grad_norm": 0.2524517774581909,
      "learning_rate": 0.00013959054301227011,
      "loss": 0.19084662199020386,
      "mean_token_accuracy": 0.922870934009552,
      "num_tokens": 67500536.0,
      "step": 5472
    },
    {
      "entropy": 1.2264028787612915,
      "epoch": 2.8820431806213795,
      "grad_norm": 0.25612008571624756,
      "learning_rate": 0.00013956983294516386,
      "loss": 0.1767480969429016,
      "mean_token_accuracy": 0.9253916144371033,
      "num_tokens": 67512872.0,
      "step": 5473
    },
    {
      "entropy": 1.2065593600273132,
      "epoch": 2.8825697735650344,
      "grad_norm": 0.2584444582462311,
      "learning_rate": 0.0001395491211226689,
      "loss": 0.18127556145191193,
      "mean_token_accuracy": 0.9242168962955475,
      "num_tokens": 67525208.0,
      "step": 5474
    },
    {
      "entropy": 1.1801214814186096,
      "epoch": 2.883096366508689,
      "grad_norm": 0.24572916328907013,
      "learning_rate": 0.00013952840754601488,
      "loss": 0.18749801814556122,
      "mean_token_accuracy": 0.9196372032165527,
      "num_tokens": 67537544.0,
      "step": 5475
    },
    {
      "entropy": 1.2200948596000671,
      "epoch": 2.8836229594523433,
      "grad_norm": 0.281856507062912,
      "learning_rate": 0.0001395076922164314,
      "loss": 0.194240540266037,
      "mean_token_accuracy": 0.9169766008853912,
      "num_tokens": 67549880.0,
      "step": 5476
    },
    {
      "entropy": 1.2109468281269073,
      "epoch": 2.884149552395998,
      "grad_norm": 0.28432926535606384,
      "learning_rate": 0.00013948697513514822,
      "loss": 0.18888379633426666,
      "mean_token_accuracy": 0.9218895584344864,
      "num_tokens": 67562216.0,
      "step": 5477
    },
    {
      "entropy": 1.1744347214698792,
      "epoch": 2.8846761453396526,
      "grad_norm": 0.24902574717998505,
      "learning_rate": 0.00013946625630339518,
      "loss": 0.1767992079257965,
      "mean_token_accuracy": 0.926706999540329,
      "num_tokens": 67574552.0,
      "step": 5478
    },
    {
      "entropy": 1.1746321320533752,
      "epoch": 2.885202738283307,
      "grad_norm": 0.2674560546875,
      "learning_rate": 0.00013944553572240226,
      "loss": 0.1794002652168274,
      "mean_token_accuracy": 0.9238214045763016,
      "num_tokens": 67586888.0,
      "step": 5479
    },
    {
      "entropy": 1.1886841356754303,
      "epoch": 2.8857293312269614,
      "grad_norm": 0.28249746561050415,
      "learning_rate": 0.0001394248133933995,
      "loss": 0.19088155031204224,
      "mean_token_accuracy": 0.919599786400795,
      "num_tokens": 67599224.0,
      "step": 5480
    },
    {
      "entropy": 1.1575458347797394,
      "epoch": 2.8862559241706163,
      "grad_norm": 0.2625466585159302,
      "learning_rate": 0.00013940408931761706,
      "loss": 0.18857283890247345,
      "mean_token_accuracy": 0.9230266064405441,
      "num_tokens": 67611560.0,
      "step": 5481
    },
    {
      "entropy": 1.164193332195282,
      "epoch": 2.8867825171142707,
      "grad_norm": 0.2611542046070099,
      "learning_rate": 0.00013938336349628524,
      "loss": 0.1958429515361786,
      "mean_token_accuracy": 0.9188902080059052,
      "num_tokens": 67623896.0,
      "step": 5482
    },
    {
      "entropy": 1.1729185581207275,
      "epoch": 2.887309110057925,
      "grad_norm": 0.2818647623062134,
      "learning_rate": 0.00013936263593063438,
      "loss": 0.20438428223133087,
      "mean_token_accuracy": 0.9178862571716309,
      "num_tokens": 67636232.0,
      "step": 5483
    },
    {
      "entropy": 1.2090890109539032,
      "epoch": 2.8878357030015795,
      "grad_norm": 0.27847352623939514,
      "learning_rate": 0.00013934190662189496,
      "loss": 0.19311118125915527,
      "mean_token_accuracy": 0.920319065451622,
      "num_tokens": 67648568.0,
      "step": 5484
    },
    {
      "entropy": 1.1742522418498993,
      "epoch": 2.8883622959452344,
      "grad_norm": 0.2799610197544098,
      "learning_rate": 0.00013932117557129755,
      "loss": 0.20451730489730835,
      "mean_token_accuracy": 0.9123682975769043,
      "num_tokens": 67660904.0,
      "step": 5485
    },
    {
      "entropy": 1.1871919333934784,
      "epoch": 2.888888888888889,
      "grad_norm": 0.27488139271736145,
      "learning_rate": 0.00013930044278007283,
      "loss": 0.1982969343662262,
      "mean_token_accuracy": 0.9207281917333603,
      "num_tokens": 67673240.0,
      "step": 5486
    },
    {
      "entropy": 1.2073151469230652,
      "epoch": 2.8894154818325433,
      "grad_norm": 0.28361013531684875,
      "learning_rate": 0.00013927970824945165,
      "loss": 0.19221793115139008,
      "mean_token_accuracy": 0.9205655306577682,
      "num_tokens": 67685576.0,
      "step": 5487
    },
    {
      "entropy": 1.1463909447193146,
      "epoch": 2.889942074776198,
      "grad_norm": 0.2542709708213806,
      "learning_rate": 0.0001392589719806648,
      "loss": 0.18216948211193085,
      "mean_token_accuracy": 0.9249231219291687,
      "num_tokens": 67697912.0,
      "step": 5488
    },
    {
      "entropy": 1.1823024451732635,
      "epoch": 2.8904686677198526,
      "grad_norm": 0.28278055787086487,
      "learning_rate": 0.00013923823397494334,
      "loss": 0.18571798503398895,
      "mean_token_accuracy": 0.9252088069915771,
      "num_tokens": 67710248.0,
      "step": 5489
    },
    {
      "entropy": 1.1500611007213593,
      "epoch": 2.890995260663507,
      "grad_norm": 0.2649606764316559,
      "learning_rate": 0.00013921749423351833,
      "loss": 0.18341246247291565,
      "mean_token_accuracy": 0.925861045718193,
      "num_tokens": 67722584.0,
      "step": 5490
    },
    {
      "entropy": 1.1709353029727936,
      "epoch": 2.891521853607162,
      "grad_norm": 0.2752512991428375,
      "learning_rate": 0.00013919675275762098,
      "loss": 0.18391147255897522,
      "mean_token_accuracy": 0.9232052862644196,
      "num_tokens": 67734920.0,
      "step": 5491
    },
    {
      "entropy": 1.1525313258171082,
      "epoch": 2.8920484465508163,
      "grad_norm": 0.3017459213733673,
      "learning_rate": 0.0001391760095484826,
      "loss": 0.20819026231765747,
      "mean_token_accuracy": 0.9137952923774719,
      "num_tokens": 67747256.0,
      "step": 5492
    },
    {
      "entropy": 1.1915410459041595,
      "epoch": 2.8925750394944707,
      "grad_norm": 0.26276955008506775,
      "learning_rate": 0.00013915526460733462,
      "loss": 0.17978140711784363,
      "mean_token_accuracy": 0.9292542785406113,
      "num_tokens": 67759592.0,
      "step": 5493
    },
    {
      "entropy": 1.2160204350948334,
      "epoch": 2.8931016324381256,
      "grad_norm": 0.28528469800949097,
      "learning_rate": 0.00013913451793540844,
      "loss": 0.19692108035087585,
      "mean_token_accuracy": 0.9169259369373322,
      "num_tokens": 67771928.0,
      "step": 5494
    },
    {
      "entropy": 1.1477001011371613,
      "epoch": 2.89362822538178,
      "grad_norm": 0.2837713956832886,
      "learning_rate": 0.00013911376953393577,
      "loss": 0.18914955854415894,
      "mean_token_accuracy": 0.9206328988075256,
      "num_tokens": 67784264.0,
      "step": 5495
    },
    {
      "entropy": 1.157814860343933,
      "epoch": 2.8941548183254344,
      "grad_norm": 0.2714032530784607,
      "learning_rate": 0.00013909301940414828,
      "loss": 0.1825396716594696,
      "mean_token_accuracy": 0.9275189191102982,
      "num_tokens": 67796600.0,
      "step": 5496
    },
    {
      "entropy": 1.1470504999160767,
      "epoch": 2.894681411269089,
      "grad_norm": 0.2933061420917511,
      "learning_rate": 0.00013907226754727783,
      "loss": 0.19350317120552063,
      "mean_token_accuracy": 0.9198871105909348,
      "num_tokens": 67808936.0,
      "step": 5497
    },
    {
      "entropy": 1.1557423174381256,
      "epoch": 2.8952080042127437,
      "grad_norm": 0.26497289538383484,
      "learning_rate": 0.00013905151396455628,
      "loss": 0.18695615231990814,
      "mean_token_accuracy": 0.9201800376176834,
      "num_tokens": 67821272.0,
      "step": 5498
    },
    {
      "entropy": 1.2245990633964539,
      "epoch": 2.895734597156398,
      "grad_norm": 0.31506285071372986,
      "learning_rate": 0.0001390307586572156,
      "loss": 0.19346579909324646,
      "mean_token_accuracy": 0.9197398275136948,
      "num_tokens": 67833608.0,
      "step": 5499
    },
    {
      "entropy": 1.194277048110962,
      "epoch": 2.8962611901000526,
      "grad_norm": 0.3318503499031067,
      "learning_rate": 0.0001390100016264881,
      "loss": 0.19116128981113434,
      "mean_token_accuracy": 0.923472136259079,
      "num_tokens": 67845944.0,
      "step": 5500
    },
    {
      "entropy": 1.147051066160202,
      "epoch": 2.896787783043707,
      "grad_norm": 0.2742278575897217,
      "learning_rate": 0.00013898924287360584,
      "loss": 0.2055630087852478,
      "mean_token_accuracy": 0.9172082245349884,
      "num_tokens": 67858280.0,
      "step": 5501
    },
    {
      "entropy": 1.1588106751441956,
      "epoch": 2.897314375987362,
      "grad_norm": 0.27538150548934937,
      "learning_rate": 0.00013896848239980118,
      "loss": 0.1679210364818573,
      "mean_token_accuracy": 0.9297711849212646,
      "num_tokens": 67870616.0,
      "step": 5502
    },
    {
      "entropy": 1.1855699121952057,
      "epoch": 2.8978409689310163,
      "grad_norm": 0.2715197801589966,
      "learning_rate": 0.00013894772020630656,
      "loss": 0.20283342897891998,
      "mean_token_accuracy": 0.9191963374614716,
      "num_tokens": 67882952.0,
      "step": 5503
    },
    {
      "entropy": 1.1894906759262085,
      "epoch": 2.8983675618746707,
      "grad_norm": 0.28704389929771423,
      "learning_rate": 0.00013892695629435454,
      "loss": 0.19675171375274658,
      "mean_token_accuracy": 0.9200365841388702,
      "num_tokens": 67895288.0,
      "step": 5504
    },
    {
      "entropy": 1.160377025604248,
      "epoch": 2.8988941548183256,
      "grad_norm": 0.2784435451030731,
      "learning_rate": 0.0001389061906651777,
      "loss": 0.18051111698150635,
      "mean_token_accuracy": 0.9258342087268829,
      "num_tokens": 67907624.0,
      "step": 5505
    },
    {
      "entropy": 1.1449159979820251,
      "epoch": 2.89942074776198,
      "grad_norm": 0.2636511027812958,
      "learning_rate": 0.00013888542332000882,
      "loss": 0.17140930891036987,
      "mean_token_accuracy": 0.9295295923948288,
      "num_tokens": 67919960.0,
      "step": 5506
    },
    {
      "entropy": 1.1670503914356232,
      "epoch": 2.8999473407056344,
      "grad_norm": 0.2751680612564087,
      "learning_rate": 0.00013886465426008067,
      "loss": 0.19141370058059692,
      "mean_token_accuracy": 0.9227615296840668,
      "num_tokens": 67932296.0,
      "step": 5507
    },
    {
      "entropy": 1.176351010799408,
      "epoch": 2.9004739336492893,
      "grad_norm": 0.2705549895763397,
      "learning_rate": 0.0001388438834866263,
      "loss": 0.19590432941913605,
      "mean_token_accuracy": 0.9217877984046936,
      "num_tokens": 67944632.0,
      "step": 5508
    },
    {
      "entropy": 1.1857755184173584,
      "epoch": 2.9010005265929437,
      "grad_norm": 0.2822164297103882,
      "learning_rate": 0.00013882311100087862,
      "loss": 0.20122680068016052,
      "mean_token_accuracy": 0.9154031872749329,
      "num_tokens": 67956968.0,
      "step": 5509
    },
    {
      "entropy": 1.1495457887649536,
      "epoch": 2.901527119536598,
      "grad_norm": 0.29185017943382263,
      "learning_rate": 0.00013880233680407087,
      "loss": 0.190867617726326,
      "mean_token_accuracy": 0.9220561236143112,
      "num_tokens": 67969304.0,
      "step": 5510
    },
    {
      "entropy": 1.106411337852478,
      "epoch": 2.902053712480253,
      "grad_norm": 0.24719662964344025,
      "learning_rate": 0.0001387815608974362,
      "loss": 0.1714990735054016,
      "mean_token_accuracy": 0.9271316230297089,
      "num_tokens": 67981640.0,
      "step": 5511
    },
    {
      "entropy": 1.1132754385471344,
      "epoch": 2.9025803054239074,
      "grad_norm": 0.2547193765640259,
      "learning_rate": 0.0001387607832822081,
      "loss": 0.1815553456544876,
      "mean_token_accuracy": 0.9266224503517151,
      "num_tokens": 67993976.0,
      "step": 5512
    },
    {
      "entropy": 1.1470238864421844,
      "epoch": 2.903106898367562,
      "grad_norm": 0.27963170409202576,
      "learning_rate": 0.00013874000395961987,
      "loss": 0.19604888558387756,
      "mean_token_accuracy": 0.9155842065811157,
      "num_tokens": 68006312.0,
      "step": 5513
    },
    {
      "entropy": 1.1404377818107605,
      "epoch": 2.9036334913112163,
      "grad_norm": 0.2531251013278961,
      "learning_rate": 0.0001387192229309051,
      "loss": 0.17346608638763428,
      "mean_token_accuracy": 0.9276323318481445,
      "num_tokens": 68018648.0,
      "step": 5514
    },
    {
      "entropy": 1.1547462940216064,
      "epoch": 2.9041600842548707,
      "grad_norm": 0.2719525396823883,
      "learning_rate": 0.0001386984401972975,
      "loss": 0.19307196140289307,
      "mean_token_accuracy": 0.9185868501663208,
      "num_tokens": 68030984.0,
      "step": 5515
    },
    {
      "entropy": 1.154180645942688,
      "epoch": 2.9046866771985256,
      "grad_norm": 0.28286588191986084,
      "learning_rate": 0.00013867765576003075,
      "loss": 0.19983167946338654,
      "mean_token_accuracy": 0.919625997543335,
      "num_tokens": 68043320.0,
      "step": 5516
    },
    {
      "entropy": 1.1318946182727814,
      "epoch": 2.90521327014218,
      "grad_norm": 0.2880934476852417,
      "learning_rate": 0.00013865686962033876,
      "loss": 0.2102263867855072,
      "mean_token_accuracy": 0.9138824343681335,
      "num_tokens": 68055656.0,
      "step": 5517
    },
    {
      "entropy": 1.17483389377594,
      "epoch": 2.9057398630858344,
      "grad_norm": 0.2941156029701233,
      "learning_rate": 0.0001386360817794554,
      "loss": 0.19800475239753723,
      "mean_token_accuracy": 0.9159929603338242,
      "num_tokens": 68067992.0,
      "step": 5518
    },
    {
      "entropy": 1.1367288827896118,
      "epoch": 2.9062664560294893,
      "grad_norm": 0.2663951516151428,
      "learning_rate": 0.00013861529223861476,
      "loss": 0.18763451278209686,
      "mean_token_accuracy": 0.9239258766174316,
      "num_tokens": 68080328.0,
      "step": 5519
    },
    {
      "entropy": 1.1588850617408752,
      "epoch": 2.9067930489731437,
      "grad_norm": 0.2822934091091156,
      "learning_rate": 0.00013859450099905102,
      "loss": 0.1937345266342163,
      "mean_token_accuracy": 0.9200102388858795,
      "num_tokens": 68092664.0,
      "step": 5520
    },
    {
      "entropy": 1.1312378942966461,
      "epoch": 2.907319641916798,
      "grad_norm": 0.29542550444602966,
      "learning_rate": 0.00013857370806199843,
      "loss": 0.22057875990867615,
      "mean_token_accuracy": 0.9083669930696487,
      "num_tokens": 68105000.0,
      "step": 5521
    },
    {
      "entropy": 1.1559078097343445,
      "epoch": 2.907846234860453,
      "grad_norm": 0.281016081571579,
      "learning_rate": 0.0001385529134286913,
      "loss": 0.20015105605125427,
      "mean_token_accuracy": 0.9185057580471039,
      "num_tokens": 68117336.0,
      "step": 5522
    },
    {
      "entropy": 1.1671436727046967,
      "epoch": 2.9083728278041074,
      "grad_norm": 0.27406924962997437,
      "learning_rate": 0.00013853211710036415,
      "loss": 0.19811135530471802,
      "mean_token_accuracy": 0.9185190796852112,
      "num_tokens": 68129672.0,
      "step": 5523
    },
    {
      "entropy": 1.1776782274246216,
      "epoch": 2.908899420747762,
      "grad_norm": 0.26774463057518005,
      "learning_rate": 0.00013851131907825152,
      "loss": 0.1784535050392151,
      "mean_token_accuracy": 0.9311188608407974,
      "num_tokens": 68142008.0,
      "step": 5524
    },
    {
      "entropy": 1.1597624719142914,
      "epoch": 2.9094260136914167,
      "grad_norm": 0.29032954573631287,
      "learning_rate": 0.000138490519363588,
      "loss": 0.1870250701904297,
      "mean_token_accuracy": 0.921985000371933,
      "num_tokens": 68154344.0,
      "step": 5525
    },
    {
      "entropy": 1.198120653629303,
      "epoch": 2.909952606635071,
      "grad_norm": 0.295183390378952,
      "learning_rate": 0.00013846971795760844,
      "loss": 0.1946275234222412,
      "mean_token_accuracy": 0.9213293045759201,
      "num_tokens": 68166680.0,
      "step": 5526
    },
    {
      "entropy": 1.1845296919345856,
      "epoch": 2.9104791995787256,
      "grad_norm": 0.2734968662261963,
      "learning_rate": 0.00013844891486154766,
      "loss": 0.19152450561523438,
      "mean_token_accuracy": 0.9245569109916687,
      "num_tokens": 68179016.0,
      "step": 5527
    },
    {
      "entropy": 1.1325328648090363,
      "epoch": 2.9110057925223805,
      "grad_norm": 0.2562635540962219,
      "learning_rate": 0.0001384281100766406,
      "loss": 0.15855859220027924,
      "mean_token_accuracy": 0.9361820667982101,
      "num_tokens": 68191352.0,
      "step": 5528
    },
    {
      "entropy": 1.1461539268493652,
      "epoch": 2.911532385466035,
      "grad_norm": 0.29077833890914917,
      "learning_rate": 0.00013840730360412234,
      "loss": 0.19791193306446075,
      "mean_token_accuracy": 0.9110380262136459,
      "num_tokens": 68203688.0,
      "step": 5529
    },
    {
      "entropy": 1.1295833885669708,
      "epoch": 2.9120589784096893,
      "grad_norm": 0.30538904666900635,
      "learning_rate": 0.00013838649544522803,
      "loss": 0.20151552557945251,
      "mean_token_accuracy": 0.9133839756250381,
      "num_tokens": 68216024.0,
      "step": 5530
    },
    {
      "entropy": 1.0852110385894775,
      "epoch": 2.9125855713533437,
      "grad_norm": 0.2697118818759918,
      "learning_rate": 0.00013836568560119297,
      "loss": 0.19141703844070435,
      "mean_token_accuracy": 0.920340895652771,
      "num_tokens": 68228360.0,
      "step": 5531
    },
    {
      "entropy": 1.0908479988574982,
      "epoch": 2.913112164296998,
      "grad_norm": 0.2654431462287903,
      "learning_rate": 0.00013834487407325246,
      "loss": 0.17703814804553986,
      "mean_token_accuracy": 0.928848385810852,
      "num_tokens": 68240696.0,
      "step": 5532
    },
    {
      "entropy": 1.102216750383377,
      "epoch": 2.913638757240653,
      "grad_norm": 0.2828739285469055,
      "learning_rate": 0.000138324060862642,
      "loss": 0.19559074938297272,
      "mean_token_accuracy": 0.9155698269605637,
      "num_tokens": 68253032.0,
      "step": 5533
    },
    {
      "entropy": 1.0754432678222656,
      "epoch": 2.9141653501843074,
      "grad_norm": 0.25650569796562195,
      "learning_rate": 0.00013830324597059707,
      "loss": 0.1764622926712036,
      "mean_token_accuracy": 0.9295746833086014,
      "num_tokens": 68265368.0,
      "step": 5534
    },
    {
      "entropy": 1.1392635107040405,
      "epoch": 2.914691943127962,
      "grad_norm": 0.28012871742248535,
      "learning_rate": 0.00013828242939835342,
      "loss": 0.18477457761764526,
      "mean_token_accuracy": 0.9233775585889816,
      "num_tokens": 68277704.0,
      "step": 5535
    },
    {
      "entropy": 1.0746980011463165,
      "epoch": 2.9152185360716167,
      "grad_norm": 0.2698800265789032,
      "learning_rate": 0.00013826161114714682,
      "loss": 0.17916232347488403,
      "mean_token_accuracy": 0.9196085929870605,
      "num_tokens": 68290040.0,
      "step": 5536
    },
    {
      "entropy": 1.0633073151111603,
      "epoch": 2.915745129015271,
      "grad_norm": 0.291299432516098,
      "learning_rate": 0.00013824079121821305,
      "loss": 0.20749138295650482,
      "mean_token_accuracy": 0.9145458191633224,
      "num_tokens": 68302376.0,
      "step": 5537
    },
    {
      "entropy": 1.0806976854801178,
      "epoch": 2.9162717219589256,
      "grad_norm": 0.2732320725917816,
      "learning_rate": 0.0001382199696127881,
      "loss": 0.16982217133045197,
      "mean_token_accuracy": 0.9300663471221924,
      "num_tokens": 68314712.0,
      "step": 5538
    },
    {
      "entropy": 1.1155247688293457,
      "epoch": 2.9167983149025805,
      "grad_norm": 0.2718318700790405,
      "learning_rate": 0.00013819914633210806,
      "loss": 0.1862022876739502,
      "mean_token_accuracy": 0.9197591245174408,
      "num_tokens": 68327048.0,
      "step": 5539
    },
    {
      "entropy": 1.0436288714408875,
      "epoch": 2.917324907846235,
      "grad_norm": 0.27272531390190125,
      "learning_rate": 0.00013817832137740904,
      "loss": 0.18183693289756775,
      "mean_token_accuracy": 0.9253615438938141,
      "num_tokens": 68339384.0,
      "step": 5540
    },
    {
      "entropy": 1.080330640077591,
      "epoch": 2.9178515007898893,
      "grad_norm": 0.26378878951072693,
      "learning_rate": 0.00013815749474992733,
      "loss": 0.18048113584518433,
      "mean_token_accuracy": 0.9251774549484253,
      "num_tokens": 68351720.0,
      "step": 5541
    },
    {
      "entropy": 1.0835255086421967,
      "epoch": 2.918378093733544,
      "grad_norm": 0.2663480043411255,
      "learning_rate": 0.00013813666645089926,
      "loss": 0.1744183599948883,
      "mean_token_accuracy": 0.9293684512376785,
      "num_tokens": 68364056.0,
      "step": 5542
    },
    {
      "entropy": 1.0906038284301758,
      "epoch": 2.9189046866771986,
      "grad_norm": 0.27525952458381653,
      "learning_rate": 0.0001381158364815613,
      "loss": 0.1827646642923355,
      "mean_token_accuracy": 0.9212264567613602,
      "num_tokens": 68376392.0,
      "step": 5543
    },
    {
      "entropy": 1.0514227747917175,
      "epoch": 2.919431279620853,
      "grad_norm": 0.26160329580307007,
      "learning_rate": 0.00013809500484314997,
      "loss": 0.17371343076229095,
      "mean_token_accuracy": 0.9279900044202805,
      "num_tokens": 68388728.0,
      "step": 5544
    },
    {
      "entropy": 1.0968440175056458,
      "epoch": 2.919957872564508,
      "grad_norm": 0.2624936103820801,
      "learning_rate": 0.000138074171536902,
      "loss": 0.17753034830093384,
      "mean_token_accuracy": 0.9319224506616592,
      "num_tokens": 68401064.0,
      "step": 5545
    },
    {
      "entropy": 1.1374029517173767,
      "epoch": 2.9204844655081623,
      "grad_norm": 0.32361873984336853,
      "learning_rate": 0.0001380533365640541,
      "loss": 0.22792309522628784,
      "mean_token_accuracy": 0.9060098677873611,
      "num_tokens": 68413400.0,
      "step": 5546
    },
    {
      "entropy": 1.12219899892807,
      "epoch": 2.9210110584518167,
      "grad_norm": 0.30003783106803894,
      "learning_rate": 0.00013803249992584308,
      "loss": 0.19746218621730804,
      "mean_token_accuracy": 0.9168110936880112,
      "num_tokens": 68425736.0,
      "step": 5547
    },
    {
      "entropy": 1.1274137794971466,
      "epoch": 2.921537651395471,
      "grad_norm": 0.29291093349456787,
      "learning_rate": 0.0001380116616235059,
      "loss": 0.19064177572727203,
      "mean_token_accuracy": 0.9205840528011322,
      "num_tokens": 68438072.0,
      "step": 5548
    },
    {
      "entropy": 1.0552725493907928,
      "epoch": 2.9220642443391256,
      "grad_norm": 0.275552898645401,
      "learning_rate": 0.0001379908216582797,
      "loss": 0.19442486763000488,
      "mean_token_accuracy": 0.9182971268892288,
      "num_tokens": 68450408.0,
      "step": 5549
    },
    {
      "entropy": 1.129012256860733,
      "epoch": 2.9225908372827805,
      "grad_norm": 0.28331005573272705,
      "learning_rate": 0.0001379699800314015,
      "loss": 0.1991719901561737,
      "mean_token_accuracy": 0.9190961122512817,
      "num_tokens": 68462744.0,
      "step": 5550
    },
    {
      "entropy": 1.1428042948246002,
      "epoch": 2.923117430226435,
      "grad_norm": 0.2907262146472931,
      "learning_rate": 0.00013794913674410866,
      "loss": 0.1955050528049469,
      "mean_token_accuracy": 0.9185226857662201,
      "num_tokens": 68475080.0,
      "step": 5551
    },
    {
      "entropy": 1.1355925500392914,
      "epoch": 2.9236440231700893,
      "grad_norm": 0.2629597783088684,
      "learning_rate": 0.00013792829179763843,
      "loss": 0.16810822486877441,
      "mean_token_accuracy": 0.9289321601390839,
      "num_tokens": 68487416.0,
      "step": 5552
    },
    {
      "entropy": 1.1617874205112457,
      "epoch": 2.924170616113744,
      "grad_norm": 0.27793678641319275,
      "learning_rate": 0.0001379074451932283,
      "loss": 0.19166631996631622,
      "mean_token_accuracy": 0.9207886010408401,
      "num_tokens": 68499752.0,
      "step": 5553
    },
    {
      "entropy": 1.1701002717018127,
      "epoch": 2.9246972090573986,
      "grad_norm": 0.2866170108318329,
      "learning_rate": 0.00013788659693211584,
      "loss": 0.19832077622413635,
      "mean_token_accuracy": 0.919544905424118,
      "num_tokens": 68512088.0,
      "step": 5554
    },
    {
      "entropy": 1.1990052163600922,
      "epoch": 2.925223802001053,
      "grad_norm": 0.30915117263793945,
      "learning_rate": 0.00013786574701553866,
      "loss": 0.2161540687084198,
      "mean_token_accuracy": 0.9073906093835831,
      "num_tokens": 68524424.0,
      "step": 5555
    },
    {
      "entropy": 1.2210666835308075,
      "epoch": 2.925750394944708,
      "grad_norm": 0.28099068999290466,
      "learning_rate": 0.00013784489544473445,
      "loss": 0.2011336386203766,
      "mean_token_accuracy": 0.9171841740608215,
      "num_tokens": 68536760.0,
      "step": 5556
    },
    {
      "entropy": 1.181823045015335,
      "epoch": 2.9262769878883623,
      "grad_norm": 0.2659522593021393,
      "learning_rate": 0.00013782404222094109,
      "loss": 0.17810407280921936,
      "mean_token_accuracy": 0.924164891242981,
      "num_tokens": 68549096.0,
      "step": 5557
    },
    {
      "entropy": 1.157679796218872,
      "epoch": 2.9268035808320167,
      "grad_norm": 0.27199587225914,
      "learning_rate": 0.00013780318734539653,
      "loss": 0.1843995898962021,
      "mean_token_accuracy": 0.9218865782022476,
      "num_tokens": 68561432.0,
      "step": 5558
    },
    {
      "entropy": 1.192594975233078,
      "epoch": 2.9273301737756716,
      "grad_norm": 0.2700129747390747,
      "learning_rate": 0.0001377823308193388,
      "loss": 0.18727098405361176,
      "mean_token_accuracy": 0.9208443760871887,
      "num_tokens": 68573768.0,
      "step": 5559
    },
    {
      "entropy": 1.1255095601081848,
      "epoch": 2.927856766719326,
      "grad_norm": 0.26654013991355896,
      "learning_rate": 0.000137761472644006,
      "loss": 0.1948063224554062,
      "mean_token_accuracy": 0.9209031462669373,
      "num_tokens": 68586104.0,
      "step": 5560
    },
    {
      "entropy": 1.2041710019111633,
      "epoch": 2.9283833596629805,
      "grad_norm": 0.27617987990379333,
      "learning_rate": 0.0001377406128206364,
      "loss": 0.2021069973707199,
      "mean_token_accuracy": 0.9189692735671997,
      "num_tokens": 68598440.0,
      "step": 5561
    },
    {
      "entropy": 1.1256828010082245,
      "epoch": 2.9289099526066353,
      "grad_norm": 0.2534157633781433,
      "learning_rate": 0.00013771975135046827,
      "loss": 0.17834888398647308,
      "mean_token_accuracy": 0.9273785054683685,
      "num_tokens": 68610776.0,
      "step": 5562
    },
    {
      "entropy": 1.1691264808177948,
      "epoch": 2.9294365455502898,
      "grad_norm": 0.30477964878082275,
      "learning_rate": 0.00013769888823474005,
      "loss": 0.20396670699119568,
      "mean_token_accuracy": 0.9175550490617752,
      "num_tokens": 68623112.0,
      "step": 5563
    },
    {
      "entropy": 1.1377472281455994,
      "epoch": 2.929963138493944,
      "grad_norm": 0.2771162986755371,
      "learning_rate": 0.0001376780234746903,
      "loss": 0.18264469504356384,
      "mean_token_accuracy": 0.922416552901268,
      "num_tokens": 68635448.0,
      "step": 5564
    },
    {
      "entropy": 1.1698735654354095,
      "epoch": 2.9304897314375986,
      "grad_norm": 0.30147019028663635,
      "learning_rate": 0.00013765715707155765,
      "loss": 0.1867901235818863,
      "mean_token_accuracy": 0.9241589903831482,
      "num_tokens": 68647784.0,
      "step": 5565
    },
    {
      "entropy": 1.166481763124466,
      "epoch": 2.931016324381253,
      "grad_norm": 0.2786411643028259,
      "learning_rate": 0.00013763628902658075,
      "loss": 0.17687778174877167,
      "mean_token_accuracy": 0.9243270307779312,
      "num_tokens": 68660120.0,
      "step": 5566
    },
    {
      "entropy": 1.1443816721439362,
      "epoch": 2.931542917324908,
      "grad_norm": 0.28672003746032715,
      "learning_rate": 0.00013761541934099845,
      "loss": 0.20579706132411957,
      "mean_token_accuracy": 0.9173473715782166,
      "num_tokens": 68672456.0,
      "step": 5567
    },
    {
      "entropy": 1.1425879895687103,
      "epoch": 2.9320695102685623,
      "grad_norm": 0.28145724534988403,
      "learning_rate": 0.00013759454801604966,
      "loss": 0.2017051726579666,
      "mean_token_accuracy": 0.915640190243721,
      "num_tokens": 68684792.0,
      "step": 5568
    },
    {
      "entropy": 1.2048570811748505,
      "epoch": 2.9325961032122168,
      "grad_norm": 0.3142516613006592,
      "learning_rate": 0.00013757367505297336,
      "loss": 0.20900258421897888,
      "mean_token_accuracy": 0.9165971428155899,
      "num_tokens": 68697128.0,
      "step": 5569
    },
    {
      "entropy": 1.1582045555114746,
      "epoch": 2.9331226961558716,
      "grad_norm": 0.2675777077674866,
      "learning_rate": 0.00013755280045300874,
      "loss": 0.18169574439525604,
      "mean_token_accuracy": 0.9225203096866608,
      "num_tokens": 68709464.0,
      "step": 5570
    },
    {
      "entropy": 1.1519646644592285,
      "epoch": 2.933649289099526,
      "grad_norm": 0.2735153138637543,
      "learning_rate": 0.00013753192421739493,
      "loss": 0.1958380937576294,
      "mean_token_accuracy": 0.9182684272527695,
      "num_tokens": 68721800.0,
      "step": 5571
    },
    {
      "entropy": 1.1638001501560211,
      "epoch": 2.9341758820431805,
      "grad_norm": 0.2899276316165924,
      "learning_rate": 0.0001375110463473712,
      "loss": 0.17327781021595,
      "mean_token_accuracy": 0.9242293983697891,
      "num_tokens": 68734136.0,
      "step": 5572
    },
    {
      "entropy": 1.2003310024738312,
      "epoch": 2.9347024749868353,
      "grad_norm": 0.2884262502193451,
      "learning_rate": 0.00013749016684417702,
      "loss": 0.18135476112365723,
      "mean_token_accuracy": 0.9279640018939972,
      "num_tokens": 68746472.0,
      "step": 5573
    },
    {
      "entropy": 1.1927189528942108,
      "epoch": 2.9352290679304898,
      "grad_norm": 0.28940659761428833,
      "learning_rate": 0.00013746928570905188,
      "loss": 0.18795032799243927,
      "mean_token_accuracy": 0.9253685474395752,
      "num_tokens": 68758808.0,
      "step": 5574
    },
    {
      "entropy": 1.1657047271728516,
      "epoch": 2.935755660874144,
      "grad_norm": 0.2832283079624176,
      "learning_rate": 0.00013744840294323537,
      "loss": 0.18321970105171204,
      "mean_token_accuracy": 0.9225218445062637,
      "num_tokens": 68771144.0,
      "step": 5575
    },
    {
      "entropy": 1.152781069278717,
      "epoch": 2.936282253817799,
      "grad_norm": 0.28583258390426636,
      "learning_rate": 0.00013742751854796712,
      "loss": 0.1874796450138092,
      "mean_token_accuracy": 0.922102153301239,
      "num_tokens": 68783480.0,
      "step": 5576
    },
    {
      "entropy": 1.2105313837528229,
      "epoch": 2.9368088467614535,
      "grad_norm": 0.27317848801612854,
      "learning_rate": 0.00013740663252448694,
      "loss": 0.18234558403491974,
      "mean_token_accuracy": 0.9219536781311035,
      "num_tokens": 68795816.0,
      "step": 5577
    },
    {
      "entropy": 1.2403792440891266,
      "epoch": 2.937335439705108,
      "grad_norm": 0.2711533010005951,
      "learning_rate": 0.00013738574487403475,
      "loss": 0.19365224242210388,
      "mean_token_accuracy": 0.9241596162319183,
      "num_tokens": 68808152.0,
      "step": 5578
    },
    {
      "entropy": 1.1860476732254028,
      "epoch": 2.9378620326487628,
      "grad_norm": 0.2992393970489502,
      "learning_rate": 0.00013736485559785047,
      "loss": 0.1919434666633606,
      "mean_token_accuracy": 0.9213222712278366,
      "num_tokens": 68820488.0,
      "step": 5579
    },
    {
      "entropy": 1.2470246851444244,
      "epoch": 2.938388625592417,
      "grad_norm": 0.287310391664505,
      "learning_rate": 0.00013734396469717425,
      "loss": 0.19495737552642822,
      "mean_token_accuracy": 0.9231348186731339,
      "num_tokens": 68832824.0,
      "step": 5580
    },
    {
      "entropy": 1.2436520159244537,
      "epoch": 2.9389152185360716,
      "grad_norm": 0.3332797586917877,
      "learning_rate": 0.00013732307217324623,
      "loss": 0.19875866174697876,
      "mean_token_accuracy": 0.9161509722471237,
      "num_tokens": 68845160.0,
      "step": 5581
    },
    {
      "entropy": 1.2003789842128754,
      "epoch": 2.939441811479726,
      "grad_norm": 0.2667289972305298,
      "learning_rate": 0.00013730217802730664,
      "loss": 0.16958105564117432,
      "mean_token_accuracy": 0.9293218553066254,
      "num_tokens": 68857496.0,
      "step": 5582
    },
    {
      "entropy": 1.1950378119945526,
      "epoch": 2.9399684044233805,
      "grad_norm": 0.2579761743545532,
      "learning_rate": 0.00013728128226059584,
      "loss": 0.18693850934505463,
      "mean_token_accuracy": 0.9228319078683853,
      "num_tokens": 68869832.0,
      "step": 5583
    },
    {
      "entropy": 1.2720504403114319,
      "epoch": 2.9404949973670353,
      "grad_norm": 0.2997112572193146,
      "learning_rate": 0.00013726038487435436,
      "loss": 0.2124652862548828,
      "mean_token_accuracy": 0.9126077145338058,
      "num_tokens": 68882168.0,
      "step": 5584
    },
    {
      "entropy": 1.228448748588562,
      "epoch": 2.9410215903106898,
      "grad_norm": 0.2893296182155609,
      "learning_rate": 0.00013723948586982275,
      "loss": 0.2037881761789322,
      "mean_token_accuracy": 0.9126189202070236,
      "num_tokens": 68894504.0,
      "step": 5585
    },
    {
      "entropy": 1.1871382594108582,
      "epoch": 2.941548183254344,
      "grad_norm": 0.26550090312957764,
      "learning_rate": 0.00013721858524824157,
      "loss": 0.19779062271118164,
      "mean_token_accuracy": 0.9172335267066956,
      "num_tokens": 68906840.0,
      "step": 5586
    },
    {
      "entropy": 1.2471431493759155,
      "epoch": 2.942074776197999,
      "grad_norm": 0.29272446036338806,
      "learning_rate": 0.0001371976830108516,
      "loss": 0.19359451532363892,
      "mean_token_accuracy": 0.9180149137973785,
      "num_tokens": 68919176.0,
      "step": 5587
    },
    {
      "entropy": 1.2173875868320465,
      "epoch": 2.9426013691416535,
      "grad_norm": 0.29134252667427063,
      "learning_rate": 0.00013717677915889377,
      "loss": 0.2119414508342743,
      "mean_token_accuracy": 0.9115025997161865,
      "num_tokens": 68931512.0,
      "step": 5588
    },
    {
      "entropy": 1.2124152481555939,
      "epoch": 2.943127962085308,
      "grad_norm": 0.2665821313858032,
      "learning_rate": 0.00013715587369360894,
      "loss": 0.17928844690322876,
      "mean_token_accuracy": 0.9267236292362213,
      "num_tokens": 68943848.0,
      "step": 5589
    },
    {
      "entropy": 1.2175507247447968,
      "epoch": 2.943654555028963,
      "grad_norm": 0.28769803047180176,
      "learning_rate": 0.00013713496661623816,
      "loss": 0.19170348346233368,
      "mean_token_accuracy": 0.9222258031368256,
      "num_tokens": 68956184.0,
      "step": 5590
    },
    {
      "entropy": 1.2272031605243683,
      "epoch": 2.944181147972617,
      "grad_norm": 0.27572914958000183,
      "learning_rate": 0.00013711405792802257,
      "loss": 0.18151387572288513,
      "mean_token_accuracy": 0.9221614301204681,
      "num_tokens": 68968520.0,
      "step": 5591
    },
    {
      "entropy": 1.2204799950122833,
      "epoch": 2.9447077409162716,
      "grad_norm": 0.27613502740859985,
      "learning_rate": 0.0001370931476302034,
      "loss": 0.2014046609401703,
      "mean_token_accuracy": 0.9178587943315506,
      "num_tokens": 68980856.0,
      "step": 5592
    },
    {
      "entropy": 1.231167882680893,
      "epoch": 2.9452343338599265,
      "grad_norm": 0.2966768443584442,
      "learning_rate": 0.00013707223572402197,
      "loss": 0.21449437737464905,
      "mean_token_accuracy": 0.908181220293045,
      "num_tokens": 68993192.0,
      "step": 5593
    },
    {
      "entropy": 1.2601788640022278,
      "epoch": 2.945760926803581,
      "grad_norm": 0.28138867020606995,
      "learning_rate": 0.00013705132221071968,
      "loss": 0.17587848007678986,
      "mean_token_accuracy": 0.9247183203697205,
      "num_tokens": 69005528.0,
      "step": 5594
    },
    {
      "entropy": 1.1992701292037964,
      "epoch": 2.9462875197472354,
      "grad_norm": 0.2590610384941101,
      "learning_rate": 0.00013703040709153806,
      "loss": 0.1704290211200714,
      "mean_token_accuracy": 0.9288462251424789,
      "num_tokens": 69017864.0,
      "step": 5595
    },
    {
      "entropy": 1.2185679376125336,
      "epoch": 2.94681411269089,
      "grad_norm": 0.2668219208717346,
      "learning_rate": 0.00013700949036771874,
      "loss": 0.16232874989509583,
      "mean_token_accuracy": 0.9315534979104996,
      "num_tokens": 69030200.0,
      "step": 5596
    },
    {
      "entropy": 1.1763940453529358,
      "epoch": 2.9473407056345446,
      "grad_norm": 0.2598697245121002,
      "learning_rate": 0.00013698857204050334,
      "loss": 0.1769179254770279,
      "mean_token_accuracy": 0.9267005920410156,
      "num_tokens": 69042536.0,
      "step": 5597
    },
    {
      "entropy": 1.202961027622223,
      "epoch": 2.947867298578199,
      "grad_norm": 0.27415749430656433,
      "learning_rate": 0.00013696765211113375,
      "loss": 0.19501955807209015,
      "mean_token_accuracy": 0.9173151254653931,
      "num_tokens": 69054872.0,
      "step": 5598
    },
    {
      "entropy": 1.209532380104065,
      "epoch": 2.9483938915218535,
      "grad_norm": 0.2773316502571106,
      "learning_rate": 0.00013694673058085185,
      "loss": 0.17899590730667114,
      "mean_token_accuracy": 0.9237614125013351,
      "num_tokens": 69067208.0,
      "step": 5599
    },
    {
      "entropy": 1.1219777464866638,
      "epoch": 2.948920484465508,
      "grad_norm": 0.2790414094924927,
      "learning_rate": 0.00013692580745089963,
      "loss": 0.19162613153457642,
      "mean_token_accuracy": 0.9177257567644119,
      "num_tokens": 69079544.0,
      "step": 5600
    },
    {
      "entropy": 1.118897259235382,
      "epoch": 2.949447077409163,
      "grad_norm": 0.2677527666091919,
      "learning_rate": 0.0001369048827225191,
      "loss": 0.18540120124816895,
      "mean_token_accuracy": 0.9220905154943466,
      "num_tokens": 69091880.0,
      "step": 5601
    },
    {
      "entropy": 1.1859447062015533,
      "epoch": 2.949973670352817,
      "grad_norm": 0.3110491931438446,
      "learning_rate": 0.00013688395639695252,
      "loss": 0.1812945008277893,
      "mean_token_accuracy": 0.92556431889534,
      "num_tokens": 69104216.0,
      "step": 5602
    },
    {
      "entropy": 1.178531676530838,
      "epoch": 2.9505002632964716,
      "grad_norm": 0.3070037066936493,
      "learning_rate": 0.00013686302847544217,
      "loss": 0.19497287273406982,
      "mean_token_accuracy": 0.9196391552686691,
      "num_tokens": 69116552.0,
      "step": 5603
    },
    {
      "entropy": 1.191015750169754,
      "epoch": 2.9510268562401265,
      "grad_norm": 0.2829379141330719,
      "learning_rate": 0.00013684209895923038,
      "loss": 0.1905146837234497,
      "mean_token_accuracy": 0.9247777760028839,
      "num_tokens": 69128888.0,
      "step": 5604
    },
    {
      "entropy": 1.1308483183383942,
      "epoch": 2.951553449183781,
      "grad_norm": 0.27800098061561584,
      "learning_rate": 0.0001368211678495596,
      "loss": 0.17672711610794067,
      "mean_token_accuracy": 0.9262357801198959,
      "num_tokens": 69141224.0,
      "step": 5605
    },
    {
      "entropy": 1.0918956398963928,
      "epoch": 2.9520800421274354,
      "grad_norm": 0.25561094284057617,
      "learning_rate": 0.00013680023514767243,
      "loss": 0.17155037820339203,
      "mean_token_accuracy": 0.9289103746414185,
      "num_tokens": 69153560.0,
      "step": 5606
    },
    {
      "entropy": 1.1246024668216705,
      "epoch": 2.9526066350710902,
      "grad_norm": 0.27591127157211304,
      "learning_rate": 0.00013677930085481148,
      "loss": 0.18423430621623993,
      "mean_token_accuracy": 0.9247561097145081,
      "num_tokens": 69165896.0,
      "step": 5607
    },
    {
      "entropy": 1.1606371998786926,
      "epoch": 2.9531332280147446,
      "grad_norm": 0.291604220867157,
      "learning_rate": 0.00013675836497221953,
      "loss": 0.19698481261730194,
      "mean_token_accuracy": 0.9247680753469467,
      "num_tokens": 69178232.0,
      "step": 5608
    },
    {
      "entropy": 1.1188312768936157,
      "epoch": 2.953659820958399,
      "grad_norm": 0.27514177560806274,
      "learning_rate": 0.00013673742750113942,
      "loss": 0.18783308565616608,
      "mean_token_accuracy": 0.9232326596975327,
      "num_tokens": 69190568.0,
      "step": 5609
    },
    {
      "entropy": 1.1653947532176971,
      "epoch": 2.954186413902054,
      "grad_norm": 0.31085118651390076,
      "learning_rate": 0.00013671648844281408,
      "loss": 0.2055719643831253,
      "mean_token_accuracy": 0.917727917432785,
      "num_tokens": 69202904.0,
      "step": 5610
    },
    {
      "entropy": 1.1252780854701996,
      "epoch": 2.9547130068457084,
      "grad_norm": 0.3070995807647705,
      "learning_rate": 0.00013669554779848652,
      "loss": 0.19462232291698456,
      "mean_token_accuracy": 0.9197984337806702,
      "num_tokens": 69215240.0,
      "step": 5611
    },
    {
      "entropy": 1.1574952900409698,
      "epoch": 2.955239599789363,
      "grad_norm": 0.3188292384147644,
      "learning_rate": 0.0001366746055693999,
      "loss": 0.19384144246578217,
      "mean_token_accuracy": 0.9179680198431015,
      "num_tokens": 69227576.0,
      "step": 5612
    },
    {
      "entropy": 1.1276047825813293,
      "epoch": 2.955766192733017,
      "grad_norm": 0.28531914949417114,
      "learning_rate": 0.0001366536617567974,
      "loss": 0.19046741724014282,
      "mean_token_accuracy": 0.9173817783594131,
      "num_tokens": 69239912.0,
      "step": 5613
    },
    {
      "entropy": 1.0744889974594116,
      "epoch": 2.956292785676672,
      "grad_norm": 0.3078976571559906,
      "learning_rate": 0.00013663271636192234,
      "loss": 0.1820499300956726,
      "mean_token_accuracy": 0.9228551387786865,
      "num_tokens": 69252248.0,
      "step": 5614
    },
    {
      "entropy": 1.070107638835907,
      "epoch": 2.9568193786203265,
      "grad_norm": 0.2992478311061859,
      "learning_rate": 0.00013661176938601816,
      "loss": 0.19892115890979767,
      "mean_token_accuracy": 0.9189724177122116,
      "num_tokens": 69264584.0,
      "step": 5615
    },
    {
      "entropy": 1.1251972019672394,
      "epoch": 2.957345971563981,
      "grad_norm": 0.2962946891784668,
      "learning_rate": 0.00013659082083032831,
      "loss": 0.19916094839572906,
      "mean_token_accuracy": 0.9157926440238953,
      "num_tokens": 69276920.0,
      "step": 5616
    },
    {
      "entropy": 1.098711758852005,
      "epoch": 2.9578725645076354,
      "grad_norm": 0.30655038356781006,
      "learning_rate": 0.00013656987069609645,
      "loss": 0.18999090790748596,
      "mean_token_accuracy": 0.9222403913736343,
      "num_tokens": 69289256.0,
      "step": 5617
    },
    {
      "entropy": 1.1321788430213928,
      "epoch": 2.9583991574512902,
      "grad_norm": 0.25325635075569153,
      "learning_rate": 0.0001365489189845662,
      "loss": 0.15657661855220795,
      "mean_token_accuracy": 0.937756285071373,
      "num_tokens": 69301592.0,
      "step": 5618
    },
    {
      "entropy": 1.1218098402023315,
      "epoch": 2.9589257503949447,
      "grad_norm": 0.2875848412513733,
      "learning_rate": 0.0001365279656969814,
      "loss": 0.19016146659851074,
      "mean_token_accuracy": 0.9236724972724915,
      "num_tokens": 69313928.0,
      "step": 5619
    },
    {
      "entropy": 1.1002118289470673,
      "epoch": 2.959452343338599,
      "grad_norm": 0.23385827243328094,
      "learning_rate": 0.00013650701083458585,
      "loss": 0.16082242131233215,
      "mean_token_accuracy": 0.9308483749628067,
      "num_tokens": 69326264.0,
      "step": 5620
    },
    {
      "entropy": 1.1413602828979492,
      "epoch": 2.959978936282254,
      "grad_norm": 0.2711458206176758,
      "learning_rate": 0.0001364860543986236,
      "loss": 0.1713000237941742,
      "mean_token_accuracy": 0.9290253072977066,
      "num_tokens": 69338600.0,
      "step": 5621
    },
    {
      "entropy": 1.1146959960460663,
      "epoch": 2.9605055292259084,
      "grad_norm": 0.269614577293396,
      "learning_rate": 0.00013646509639033864,
      "loss": 0.19589899480342865,
      "mean_token_accuracy": 0.9199978858232498,
      "num_tokens": 69350936.0,
      "step": 5622
    },
    {
      "entropy": 1.1638236343860626,
      "epoch": 2.961032122169563,
      "grad_norm": 0.2893790304660797,
      "learning_rate": 0.00013644413681097517,
      "loss": 0.18870526552200317,
      "mean_token_accuracy": 0.9185014367103577,
      "num_tokens": 69363272.0,
      "step": 5623
    },
    {
      "entropy": 1.1203119456768036,
      "epoch": 2.9615587151132177,
      "grad_norm": 0.27260667085647583,
      "learning_rate": 0.00013642317566177746,
      "loss": 0.18144097924232483,
      "mean_token_accuracy": 0.9256602972745895,
      "num_tokens": 69375608.0,
      "step": 5624
    },
    {
      "entropy": 1.1110114455223083,
      "epoch": 2.962085308056872,
      "grad_norm": 0.27438077330589294,
      "learning_rate": 0.00013640221294398977,
      "loss": 0.1718054860830307,
      "mean_token_accuracy": 0.9250148683786392,
      "num_tokens": 69387944.0,
      "step": 5625
    },
    {
      "entropy": 1.1150699257850647,
      "epoch": 2.9626119010005265,
      "grad_norm": 0.2791772186756134,
      "learning_rate": 0.0001363812486588566,
      "loss": 0.193474680185318,
      "mean_token_accuracy": 0.9222072809934616,
      "num_tokens": 69400280.0,
      "step": 5626
    },
    {
      "entropy": 1.1526327729225159,
      "epoch": 2.9631384939441814,
      "grad_norm": 0.29484543204307556,
      "learning_rate": 0.00013636028280762243,
      "loss": 0.18960119783878326,
      "mean_token_accuracy": 0.922269880771637,
      "num_tokens": 69412616.0,
      "step": 5627
    },
    {
      "entropy": 1.1298829913139343,
      "epoch": 2.963665086887836,
      "grad_norm": 0.3016878664493561,
      "learning_rate": 0.00013633931539153195,
      "loss": 0.20321956276893616,
      "mean_token_accuracy": 0.9152559489011765,
      "num_tokens": 69424952.0,
      "step": 5628
    },
    {
      "entropy": 1.1410238444805145,
      "epoch": 2.9641916798314902,
      "grad_norm": 0.28455066680908203,
      "learning_rate": 0.00013631834641182982,
      "loss": 0.18510159850120544,
      "mean_token_accuracy": 0.9224362522363663,
      "num_tokens": 69437288.0,
      "step": 5629
    },
    {
      "entropy": 1.1718485355377197,
      "epoch": 2.9647182727751447,
      "grad_norm": 0.2834126055240631,
      "learning_rate": 0.00013629737586976087,
      "loss": 0.1816817671060562,
      "mean_token_accuracy": 0.9258575141429901,
      "num_tokens": 69449624.0,
      "step": 5630
    },
    {
      "entropy": 1.172424703836441,
      "epoch": 2.9652448657187995,
      "grad_norm": 0.3318967819213867,
      "learning_rate": 0.00013627640376656996,
      "loss": 0.20343869924545288,
      "mean_token_accuracy": 0.918372318148613,
      "num_tokens": 69461960.0,
      "step": 5631
    },
    {
      "entropy": 1.1813274025917053,
      "epoch": 2.965771458662454,
      "grad_norm": 0.3032076060771942,
      "learning_rate": 0.0001362554301035021,
      "loss": 0.18673169612884521,
      "mean_token_accuracy": 0.9254914075136185,
      "num_tokens": 69474296.0,
      "step": 5632
    },
    {
      "entropy": 1.17621248960495,
      "epoch": 2.9662980516061084,
      "grad_norm": 0.2967461347579956,
      "learning_rate": 0.00013623445488180242,
      "loss": 0.20186461508274078,
      "mean_token_accuracy": 0.9179050922393799,
      "num_tokens": 69486632.0,
      "step": 5633
    },
    {
      "entropy": 1.1494372189044952,
      "epoch": 2.966824644549763,
      "grad_norm": 0.2860412299633026,
      "learning_rate": 0.00013621347810271605,
      "loss": 0.18691250681877136,
      "mean_token_accuracy": 0.9263933897018433,
      "num_tokens": 69498968.0,
      "step": 5634
    },
    {
      "entropy": 1.218750774860382,
      "epoch": 2.9673512374934177,
      "grad_norm": 0.2923490107059479,
      "learning_rate": 0.00013619249976748826,
      "loss": 0.18676425516605377,
      "mean_token_accuracy": 0.9209718555212021,
      "num_tokens": 69511304.0,
      "step": 5635
    },
    {
      "entropy": 1.1901514530181885,
      "epoch": 2.967877830437072,
      "grad_norm": 0.28049829602241516,
      "learning_rate": 0.00013617151987736442,
      "loss": 0.18535064160823822,
      "mean_token_accuracy": 0.9233685433864594,
      "num_tokens": 69523640.0,
      "step": 5636
    },
    {
      "entropy": 1.2013523578643799,
      "epoch": 2.9684044233807265,
      "grad_norm": 0.28062236309051514,
      "learning_rate": 0.00013615053843358998,
      "loss": 0.17983686923980713,
      "mean_token_accuracy": 0.9255612045526505,
      "num_tokens": 69535976.0,
      "step": 5637
    },
    {
      "entropy": 1.1624835133552551,
      "epoch": 2.9689310163243814,
      "grad_norm": 0.2737025320529938,
      "learning_rate": 0.0001361295554374105,
      "loss": 0.18844152987003326,
      "mean_token_accuracy": 0.9273969531059265,
      "num_tokens": 69548312.0,
      "step": 5638
    },
    {
      "entropy": 1.1991266310214996,
      "epoch": 2.969457609268036,
      "grad_norm": 0.2777165174484253,
      "learning_rate": 0.00013610857089007162,
      "loss": 0.19535896182060242,
      "mean_token_accuracy": 0.9200830161571503,
      "num_tokens": 69560648.0,
      "step": 5639
    },
    {
      "entropy": 1.144371896982193,
      "epoch": 2.9699842022116902,
      "grad_norm": 0.2615988850593567,
      "learning_rate": 0.00013608758479281903,
      "loss": 0.1780557632446289,
      "mean_token_accuracy": 0.9246671944856644,
      "num_tokens": 69572984.0,
      "step": 5640
    },
    {
      "entropy": 1.1618445813655853,
      "epoch": 2.970510795155345,
      "grad_norm": 0.2715195417404175,
      "learning_rate": 0.0001360665971468986,
      "loss": 0.19814664125442505,
      "mean_token_accuracy": 0.9190391451120377,
      "num_tokens": 69585320.0,
      "step": 5641
    },
    {
      "entropy": 1.1613898575305939,
      "epoch": 2.9710373880989995,
      "grad_norm": 0.2754245400428772,
      "learning_rate": 0.00013604560795355623,
      "loss": 0.1853109747171402,
      "mean_token_accuracy": 0.921645313501358,
      "num_tokens": 69597656.0,
      "step": 5642
    },
    {
      "entropy": 1.1355976164340973,
      "epoch": 2.971563981042654,
      "grad_norm": 0.257498562335968,
      "learning_rate": 0.0001360246172140379,
      "loss": 0.17994259297847748,
      "mean_token_accuracy": 0.9238065183162689,
      "num_tokens": 69609992.0,
      "step": 5643
    },
    {
      "entropy": 1.112012267112732,
      "epoch": 2.972090573986309,
      "grad_norm": 0.252828985452652,
      "learning_rate": 0.00013600362492958976,
      "loss": 0.18598134815692902,
      "mean_token_accuracy": 0.9217955470085144,
      "num_tokens": 69622328.0,
      "step": 5644
    },
    {
      "entropy": 1.126198172569275,
      "epoch": 2.9726171669299633,
      "grad_norm": 0.26551613211631775,
      "learning_rate": 0.00013598263110145792,
      "loss": 0.19965651631355286,
      "mean_token_accuracy": 0.9193202555179596,
      "num_tokens": 69634664.0,
      "step": 5645
    },
    {
      "entropy": 1.1668301224708557,
      "epoch": 2.9731437598736177,
      "grad_norm": 0.28958189487457275,
      "learning_rate": 0.00013596163573088876,
      "loss": 0.18356260657310486,
      "mean_token_accuracy": 0.9249053448438644,
      "num_tokens": 69647000.0,
      "step": 5646
    },
    {
      "entropy": 1.1143859326839447,
      "epoch": 2.973670352817272,
      "grad_norm": 0.3172629177570343,
      "learning_rate": 0.00013594063881912854,
      "loss": 0.1905253529548645,
      "mean_token_accuracy": 0.921536773443222,
      "num_tokens": 69659336.0,
      "step": 5647
    },
    {
      "entropy": 1.094642162322998,
      "epoch": 2.974196945760927,
      "grad_norm": 0.2848716974258423,
      "learning_rate": 0.00013591964036742385,
      "loss": 0.19140908122062683,
      "mean_token_accuracy": 0.9186394363641739,
      "num_tokens": 69671672.0,
      "step": 5648
    },
    {
      "entropy": 1.1297516524791718,
      "epoch": 2.9747235387045814,
      "grad_norm": 0.2905782461166382,
      "learning_rate": 0.00013589864037702118,
      "loss": 0.19544151425361633,
      "mean_token_accuracy": 0.9194969236850739,
      "num_tokens": 69684008.0,
      "step": 5649
    },
    {
      "entropy": 1.0739370584487915,
      "epoch": 2.975250131648236,
      "grad_norm": 0.2813047766685486,
      "learning_rate": 0.00013587763884916716,
      "loss": 0.18597646057605743,
      "mean_token_accuracy": 0.9232052862644196,
      "num_tokens": 69696344.0,
      "step": 5650
    },
    {
      "entropy": 1.1179830431938171,
      "epoch": 2.9757767245918902,
      "grad_norm": 0.2737787365913391,
      "learning_rate": 0.00013585663578510852,
      "loss": 0.17890819907188416,
      "mean_token_accuracy": 0.928794801235199,
      "num_tokens": 69708680.0,
      "step": 5651
    },
    {
      "entropy": 1.0576025247573853,
      "epoch": 2.976303317535545,
      "grad_norm": 0.29176604747772217,
      "learning_rate": 0.00013583563118609217,
      "loss": 0.20427364110946655,
      "mean_token_accuracy": 0.9133568853139877,
      "num_tokens": 69721016.0,
      "step": 5652
    },
    {
      "entropy": 1.0724064707756042,
      "epoch": 2.9768299104791995,
      "grad_norm": 0.29897189140319824,
      "learning_rate": 0.00013581462505336494,
      "loss": 0.19813014566898346,
      "mean_token_accuracy": 0.9181051701307297,
      "num_tokens": 69733352.0,
      "step": 5653
    },
    {
      "entropy": 1.0667171627283096,
      "epoch": 2.977356503422854,
      "grad_norm": 0.2759121060371399,
      "learning_rate": 0.00013579361738817392,
      "loss": 0.18483181297779083,
      "mean_token_accuracy": 0.9265222549438477,
      "num_tokens": 69745688.0,
      "step": 5654
    },
    {
      "entropy": 1.0673297494649887,
      "epoch": 2.977883096366509,
      "grad_norm": 0.2818041443824768,
      "learning_rate": 0.00013577260819176614,
      "loss": 0.1941726803779602,
      "mean_token_accuracy": 0.9198822677135468,
      "num_tokens": 69758024.0,
      "step": 5655
    },
    {
      "entropy": 1.1064940094947815,
      "epoch": 2.9784096893101633,
      "grad_norm": 0.274727463722229,
      "learning_rate": 0.0001357515974653888,
      "loss": 0.1813822090625763,
      "mean_token_accuracy": 0.9252461194992065,
      "num_tokens": 69770360.0,
      "step": 5656
    },
    {
      "entropy": 1.063794195652008,
      "epoch": 2.9789362822538177,
      "grad_norm": 0.26077696681022644,
      "learning_rate": 0.00013573058521028925,
      "loss": 0.18046130239963531,
      "mean_token_accuracy": 0.9256167858839035,
      "num_tokens": 69782696.0,
      "step": 5657
    },
    {
      "entropy": 1.0971498489379883,
      "epoch": 2.9794628751974725,
      "grad_norm": 0.2734004259109497,
      "learning_rate": 0.00013570957142771482,
      "loss": 0.17433691024780273,
      "mean_token_accuracy": 0.9247601628303528,
      "num_tokens": 69795032.0,
      "step": 5658
    },
    {
      "entropy": 1.0478947162628174,
      "epoch": 2.979989468141127,
      "grad_norm": 0.25824877619743347,
      "learning_rate": 0.00013568855611891298,
      "loss": 0.17879095673561096,
      "mean_token_accuracy": 0.9230777621269226,
      "num_tokens": 69807368.0,
      "step": 5659
    },
    {
      "entropy": 1.059202879667282,
      "epoch": 2.9805160610847814,
      "grad_norm": 0.2494882196187973,
      "learning_rate": 0.00013566753928513127,
      "loss": 0.17027448117733002,
      "mean_token_accuracy": 0.9276464134454727,
      "num_tokens": 69819704.0,
      "step": 5660
    },
    {
      "entropy": 1.0435924679040909,
      "epoch": 2.9810426540284363,
      "grad_norm": 0.2691742777824402,
      "learning_rate": 0.00013564652092761738,
      "loss": 0.18695388734340668,
      "mean_token_accuracy": 0.9181651771068573,
      "num_tokens": 69832040.0,
      "step": 5661
    },
    {
      "entropy": 1.0967499911785126,
      "epoch": 2.9815692469720907,
      "grad_norm": 0.28094589710235596,
      "learning_rate": 0.000135625501047619,
      "loss": 0.1988556832075119,
      "mean_token_accuracy": 0.9151445478200912,
      "num_tokens": 69844376.0,
      "step": 5662
    },
    {
      "entropy": 1.0836659967899323,
      "epoch": 2.982095839915745,
      "grad_norm": 0.27233564853668213,
      "learning_rate": 0.000135604479646384,
      "loss": 0.19193214178085327,
      "mean_token_accuracy": 0.9200305938720703,
      "num_tokens": 69856712.0,
      "step": 5663
    },
    {
      "entropy": 1.1166917979717255,
      "epoch": 2.9826224328593995,
      "grad_norm": 0.27832654118537903,
      "learning_rate": 0.00013558345672516026,
      "loss": 0.1840783655643463,
      "mean_token_accuracy": 0.9224764257669449,
      "num_tokens": 69869048.0,
      "step": 5664
    },
    {
      "entropy": 1.071811318397522,
      "epoch": 2.983149025803054,
      "grad_norm": 0.27846652269363403,
      "learning_rate": 0.00013556243228519583,
      "loss": 0.18135033547878265,
      "mean_token_accuracy": 0.9215449392795563,
      "num_tokens": 69881384.0,
      "step": 5665
    },
    {
      "entropy": 1.0782484710216522,
      "epoch": 2.983675618746709,
      "grad_norm": 0.28974413871765137,
      "learning_rate": 0.00013554140632773874,
      "loss": 0.2007175236940384,
      "mean_token_accuracy": 0.9168247580528259,
      "num_tokens": 69893720.0,
      "step": 5666
    },
    {
      "entropy": 1.1107634007930756,
      "epoch": 2.9842022116903633,
      "grad_norm": 0.283978670835495,
      "learning_rate": 0.00013552037885403728,
      "loss": 0.17787045240402222,
      "mean_token_accuracy": 0.9235873967409134,
      "num_tokens": 69906056.0,
      "step": 5667
    },
    {
      "entropy": 1.107334554195404,
      "epoch": 2.9847288046340177,
      "grad_norm": 0.2613908648490906,
      "learning_rate": 0.00013549934986533966,
      "loss": 0.17772354185581207,
      "mean_token_accuracy": 0.9233258366584778,
      "num_tokens": 69918392.0,
      "step": 5668
    },
    {
      "entropy": 1.0990718305110931,
      "epoch": 2.9852553975776726,
      "grad_norm": 0.2806550860404968,
      "learning_rate": 0.00013547831936289422,
      "loss": 0.20128649473190308,
      "mean_token_accuracy": 0.9159611463546753,
      "num_tokens": 69930728.0,
      "step": 5669
    },
    {
      "entropy": 1.0974575281143188,
      "epoch": 2.985781990521327,
      "grad_norm": 0.2574922740459442,
      "learning_rate": 0.00013545728734794949,
      "loss": 0.1948179453611374,
      "mean_token_accuracy": 0.9183401465415955,
      "num_tokens": 69943064.0,
      "step": 5670
    },
    {
      "entropy": 1.1051023602485657,
      "epoch": 2.9863085834649814,
      "grad_norm": 0.27157384157180786,
      "learning_rate": 0.00013543625382175396,
      "loss": 0.18090611696243286,
      "mean_token_accuracy": 0.9248793721199036,
      "num_tokens": 69955400.0,
      "step": 5671
    },
    {
      "entropy": 1.0544289648532867,
      "epoch": 2.9868351764086363,
      "grad_norm": 0.271531879901886,
      "learning_rate": 0.00013541521878555633,
      "loss": 0.18519523739814758,
      "mean_token_accuracy": 0.9253270477056503,
      "num_tokens": 69967736.0,
      "step": 5672
    },
    {
      "entropy": 1.038584515452385,
      "epoch": 2.9873617693522907,
      "grad_norm": 0.26272615790367126,
      "learning_rate": 0.00013539418224060528,
      "loss": 0.18200033903121948,
      "mean_token_accuracy": 0.9267624914646149,
      "num_tokens": 69980072.0,
      "step": 5673
    },
    {
      "entropy": 1.1364141404628754,
      "epoch": 2.987888362295945,
      "grad_norm": 0.29418516159057617,
      "learning_rate": 0.0001353731441881496,
      "loss": 0.1954139918088913,
      "mean_token_accuracy": 0.9196117371320724,
      "num_tokens": 69992408.0,
      "step": 5674
    },
    {
      "entropy": 1.0358662456274033,
      "epoch": 2.9884149552396,
      "grad_norm": 0.24295592308044434,
      "learning_rate": 0.0001353521046294383,
      "loss": 0.17815330624580383,
      "mean_token_accuracy": 0.9219689667224884,
      "num_tokens": 70004744.0,
      "step": 5675
    },
    {
      "entropy": 1.0438331812620163,
      "epoch": 2.9889415481832544,
      "grad_norm": 0.27891722321510315,
      "learning_rate": 0.00013533106356572024,
      "loss": 0.20683948695659637,
      "mean_token_accuracy": 0.9159213751554489,
      "num_tokens": 70017080.0,
      "step": 5676
    },
    {
      "entropy": 1.0625006258487701,
      "epoch": 2.989468141126909,
      "grad_norm": 0.2645440995693207,
      "learning_rate": 0.00013531002099824465,
      "loss": 0.19023014605045319,
      "mean_token_accuracy": 0.9216374158859253,
      "num_tokens": 70029416.0,
      "step": 5677
    },
    {
      "entropy": 1.0237343311309814,
      "epoch": 2.9899947340705637,
      "grad_norm": 0.29645782709121704,
      "learning_rate": 0.0001352889769282606,
      "loss": 0.19946202635765076,
      "mean_token_accuracy": 0.9190236330032349,
      "num_tokens": 70041752.0,
      "step": 5678
    },
    {
      "entropy": 0.9966332167387009,
      "epoch": 2.990521327014218,
      "grad_norm": 0.27027246356010437,
      "learning_rate": 0.0001352679313570174,
      "loss": 0.1885770857334137,
      "mean_token_accuracy": 0.9219847321510315,
      "num_tokens": 70054088.0,
      "step": 5679
    },
    {
      "entropy": 1.020631194114685,
      "epoch": 2.9910479199578726,
      "grad_norm": 0.2740084230899811,
      "learning_rate": 0.00013524688428576435,
      "loss": 0.17383310198783875,
      "mean_token_accuracy": 0.9266319870948792,
      "num_tokens": 70066424.0,
      "step": 5680
    },
    {
      "entropy": 0.9709053188562393,
      "epoch": 2.991574512901527,
      "grad_norm": 0.25379857420921326,
      "learning_rate": 0.00013522583571575094,
      "loss": 0.17841334640979767,
      "mean_token_accuracy": 0.9239553660154343,
      "num_tokens": 70078760.0,
      "step": 5681
    },
    {
      "entropy": 1.0074795484542847,
      "epoch": 2.9921011058451814,
      "grad_norm": 0.30197978019714355,
      "learning_rate": 0.0001352047856482267,
      "loss": 0.2074137181043625,
      "mean_token_accuracy": 0.9108300656080246,
      "num_tokens": 70091096.0,
      "step": 5682
    },
    {
      "entropy": 1.0503327548503876,
      "epoch": 2.9926276987888363,
      "grad_norm": 0.3334502875804901,
      "learning_rate": 0.00013518373408444127,
      "loss": 0.1890716552734375,
      "mean_token_accuracy": 0.9206590056419373,
      "num_tokens": 70103432.0,
      "step": 5683
    },
    {
      "entropy": 0.9906752109527588,
      "epoch": 2.9931542917324907,
      "grad_norm": 0.27631211280822754,
      "learning_rate": 0.00013516268102564428,
      "loss": 0.19861096143722534,
      "mean_token_accuracy": 0.9207852929830551,
      "num_tokens": 70115768.0,
      "step": 5684
    },
    {
      "entropy": 1.0107875168323517,
      "epoch": 2.993680884676145,
      "grad_norm": 0.2852430045604706,
      "learning_rate": 0.0001351416264730856,
      "loss": 0.19408808648586273,
      "mean_token_accuracy": 0.9197646528482437,
      "num_tokens": 70128104.0,
      "step": 5685
    },
    {
      "entropy": 1.0509508550167084,
      "epoch": 2.9942074776198,
      "grad_norm": 0.2755467891693115,
      "learning_rate": 0.0001351205704280151,
      "loss": 0.19173826277256012,
      "mean_token_accuracy": 0.9201934188604355,
      "num_tokens": 70140440.0,
      "step": 5686
    },
    {
      "entropy": 0.9853422343730927,
      "epoch": 2.9947340705634544,
      "grad_norm": 0.2897648811340332,
      "learning_rate": 0.00013509951289168273,
      "loss": 0.1806781142950058,
      "mean_token_accuracy": 0.9233992248773575,
      "num_tokens": 70152776.0,
      "step": 5687
    },
    {
      "entropy": 0.9925468116998672,
      "epoch": 2.995260663507109,
      "grad_norm": 0.2701224684715271,
      "learning_rate": 0.00013507845386533858,
      "loss": 0.19695024192333221,
      "mean_token_accuracy": 0.9201646447181702,
      "num_tokens": 70165112.0,
      "step": 5688
    },
    {
      "entropy": 1.03090438246727,
      "epoch": 2.9957872564507637,
      "grad_norm": 0.2682291865348816,
      "learning_rate": 0.0001350573933502328,
      "loss": 0.17822284996509552,
      "mean_token_accuracy": 0.9288727343082428,
      "num_tokens": 70177448.0,
      "step": 5689
    },
    {
      "entropy": 1.086290031671524,
      "epoch": 2.996313849394418,
      "grad_norm": 0.28797224164009094,
      "learning_rate": 0.00013503633134761563,
      "loss": 0.20880261063575745,
      "mean_token_accuracy": 0.9159017503261566,
      "num_tokens": 70189784.0,
      "step": 5690
    },
    {
      "entropy": 1.0291350483894348,
      "epoch": 2.9968404423380726,
      "grad_norm": 0.27310577034950256,
      "learning_rate": 0.00013501526785873738,
      "loss": 0.1941332072019577,
      "mean_token_accuracy": 0.9144625663757324,
      "num_tokens": 70202120.0,
      "step": 5691
    },
    {
      "entropy": 1.062148556113243,
      "epoch": 2.9973670352817274,
      "grad_norm": 0.2835206687450409,
      "learning_rate": 0.00013499420288484842,
      "loss": 0.187083899974823,
      "mean_token_accuracy": 0.9213053584098816,
      "num_tokens": 70214456.0,
      "step": 5692
    },
    {
      "entropy": 1.0626972317695618,
      "epoch": 2.997893628225382,
      "grad_norm": 0.29710713028907776,
      "learning_rate": 0.0001349731364271994,
      "loss": 0.19888778030872345,
      "mean_token_accuracy": 0.9213065057992935,
      "num_tokens": 70226792.0,
      "step": 5693
    },
    {
      "entropy": 1.0669539272785187,
      "epoch": 2.9984202211690363,
      "grad_norm": 0.29479360580444336,
      "learning_rate": 0.00013495206848704078,
      "loss": 0.1920504868030548,
      "mean_token_accuracy": 0.9204203486442566,
      "num_tokens": 70239128.0,
      "step": 5694
    },
    {
      "entropy": 1.116328477859497,
      "epoch": 2.998946814112691,
      "grad_norm": 0.2870364487171173,
      "learning_rate": 0.0001349309990656233,
      "loss": 0.18787749111652374,
      "mean_token_accuracy": 0.9229247868061066,
      "num_tokens": 70251464.0,
      "step": 5695
    },
    {
      "entropy": 1.0815511643886566,
      "epoch": 2.9994734070563456,
      "grad_norm": 0.28028604388237,
      "learning_rate": 0.00013490992816419769,
      "loss": 0.1949234902858734,
      "mean_token_accuracy": 0.9191743582487106,
      "num_tokens": 70263800.0,
      "step": 5696
    },
    {
      "entropy": 1.1147053837776184,
      "epoch": 3.0,
      "grad_norm": 0.28414386510849,
      "learning_rate": 0.00013488885578401486,
      "loss": 0.18812201917171478,
      "mean_token_accuracy": 0.9226029217243195,
      "num_tokens": 70275108.0,
      "step": 5697
    },
    {
      "epoch": 3.0,
      "eval_entropy": 1.1281561226578622,
      "eval_loss": 0.19836974143981934,
      "eval_mean_token_accuracy": 0.9189834011862816,
      "eval_num_tokens": 70275108.0,
      "eval_runtime": 665.1541,
      "eval_samples_per_second": 8.565,
      "eval_steps_per_second": 2.855,
      "step": 5697
    },
    {
      "entropy": 1.0955340713262558,
      "epoch": 3.0005265929436544,
      "grad_norm": 0.26776036620140076,
      "learning_rate": 0.00013486778192632574,
      "loss": 0.17236140370368958,
      "mean_token_accuracy": 0.9287768304347992,
      "num_tokens": 70287444.0,
      "step": 5698
    },
    {
      "entropy": 1.1214534044265747,
      "epoch": 3.0010531858873093,
      "grad_norm": 0.26134249567985535,
      "learning_rate": 0.0001348467065923813,
      "loss": 0.1755731999874115,
      "mean_token_accuracy": 0.9287609606981277,
      "num_tokens": 70299780.0,
      "step": 5699
    },
    {
      "entropy": 1.18242946267128,
      "epoch": 3.0015797788309637,
      "grad_norm": 0.2665163576602936,
      "learning_rate": 0.00013482562978343274,
      "loss": 0.17313434183597565,
      "mean_token_accuracy": 0.9301868975162506,
      "num_tokens": 70312116.0,
      "step": 5700
    },
    {
      "entropy": 1.1398328244686127,
      "epoch": 3.002106371774618,
      "grad_norm": 0.2637473940849304,
      "learning_rate": 0.0001348045515007312,
      "loss": 0.1913183033466339,
      "mean_token_accuracy": 0.9221321642398834,
      "num_tokens": 70324452.0,
      "step": 5701
    },
    {
      "entropy": 1.1458471715450287,
      "epoch": 3.0026329647182726,
      "grad_norm": 0.2840864062309265,
      "learning_rate": 0.00013478347174552806,
      "loss": 0.1964738816022873,
      "mean_token_accuracy": 0.9198814779520035,
      "num_tokens": 70336788.0,
      "step": 5702
    },
    {
      "entropy": 1.1397323310375214,
      "epoch": 3.0031595576619274,
      "grad_norm": 0.27136555314064026,
      "learning_rate": 0.00013476239051907462,
      "loss": 0.17828154563903809,
      "mean_token_accuracy": 0.9233542084693909,
      "num_tokens": 70349124.0,
      "step": 5703
    },
    {
      "entropy": 1.1608546674251556,
      "epoch": 3.003686150605582,
      "grad_norm": 0.26567453145980835,
      "learning_rate": 0.0001347413078226224,
      "loss": 0.17471587657928467,
      "mean_token_accuracy": 0.9279931634664536,
      "num_tokens": 70361460.0,
      "step": 5704
    },
    {
      "entropy": 1.1472634077072144,
      "epoch": 3.0042127435492363,
      "grad_norm": 0.30112066864967346,
      "learning_rate": 0.0001347202236574229,
      "loss": 0.1816388964653015,
      "mean_token_accuracy": 0.923245444893837,
      "num_tokens": 70373796.0,
      "step": 5705
    },
    {
      "entropy": 1.151903212070465,
      "epoch": 3.004739336492891,
      "grad_norm": 0.293971449136734,
      "learning_rate": 0.00013469913802472783,
      "loss": 0.18356375396251678,
      "mean_token_accuracy": 0.9272286742925644,
      "num_tokens": 70386132.0,
      "step": 5706
    },
    {
      "entropy": 1.1657882332801819,
      "epoch": 3.0052659294365456,
      "grad_norm": 0.28911593556404114,
      "learning_rate": 0.0001346780509257889,
      "loss": 0.18047240376472473,
      "mean_token_accuracy": 0.9277788549661636,
      "num_tokens": 70398468.0,
      "step": 5707
    },
    {
      "entropy": 1.1399290561676025,
      "epoch": 3.0057925223802,
      "grad_norm": 0.29552772641181946,
      "learning_rate": 0.0001346569623618579,
      "loss": 0.19448219239711761,
      "mean_token_accuracy": 0.9166269600391388,
      "num_tokens": 70410804.0,
      "step": 5708
    },
    {
      "entropy": 1.1525852382183075,
      "epoch": 3.006319115323855,
      "grad_norm": 0.3083997070789337,
      "learning_rate": 0.00013463587233418677,
      "loss": 0.19267818331718445,
      "mean_token_accuracy": 0.9213844537734985,
      "num_tokens": 70423140.0,
      "step": 5709
    },
    {
      "entropy": 1.1313498616218567,
      "epoch": 3.0068457082675093,
      "grad_norm": 0.2973012626171112,
      "learning_rate": 0.00013461478084402745,
      "loss": 0.1891457885503769,
      "mean_token_accuracy": 0.9198461174964905,
      "num_tokens": 70435476.0,
      "step": 5710
    },
    {
      "entropy": 1.115710586309433,
      "epoch": 3.0073723012111637,
      "grad_norm": 0.290634423494339,
      "learning_rate": 0.00013459368789263212,
      "loss": 0.18500478565692902,
      "mean_token_accuracy": 0.9200536608695984,
      "num_tokens": 70447812.0,
      "step": 5711
    },
    {
      "entropy": 1.100893348455429,
      "epoch": 3.007898894154818,
      "grad_norm": 0.2578546702861786,
      "learning_rate": 0.00013457259348125287,
      "loss": 0.16856224834918976,
      "mean_token_accuracy": 0.9306410551071167,
      "num_tokens": 70460148.0,
      "step": 5712
    },
    {
      "entropy": 1.139553964138031,
      "epoch": 3.008425487098473,
      "grad_norm": 0.2893870174884796,
      "learning_rate": 0.00013455149761114194,
      "loss": 0.18014776706695557,
      "mean_token_accuracy": 0.9216979593038559,
      "num_tokens": 70472484.0,
      "step": 5713
    },
    {
      "entropy": 1.1330049335956573,
      "epoch": 3.0089520800421274,
      "grad_norm": 0.27027586102485657,
      "learning_rate": 0.0001345304002835517,
      "loss": 0.18928660452365875,
      "mean_token_accuracy": 0.9179441779851913,
      "num_tokens": 70484820.0,
      "step": 5714
    },
    {
      "entropy": 1.0694325268268585,
      "epoch": 3.009478672985782,
      "grad_norm": 0.2793356776237488,
      "learning_rate": 0.00013450930149973457,
      "loss": 0.16622862219810486,
      "mean_token_accuracy": 0.9302334487438202,
      "num_tokens": 70497156.0,
      "step": 5715
    },
    {
      "entropy": 1.1282596588134766,
      "epoch": 3.0100052659294367,
      "grad_norm": 0.30310606956481934,
      "learning_rate": 0.00013448820126094307,
      "loss": 0.18254932761192322,
      "mean_token_accuracy": 0.926398292183876,
      "num_tokens": 70509492.0,
      "step": 5716
    },
    {
      "entropy": 1.0743071734905243,
      "epoch": 3.010531858873091,
      "grad_norm": 0.2671526372432709,
      "learning_rate": 0.00013446709956842977,
      "loss": 0.17425042390823364,
      "mean_token_accuracy": 0.9269017428159714,
      "num_tokens": 70521828.0,
      "step": 5717
    },
    {
      "entropy": 1.140877604484558,
      "epoch": 3.0110584518167456,
      "grad_norm": 0.28673312067985535,
      "learning_rate": 0.0001344459964234474,
      "loss": 0.17692753672599792,
      "mean_token_accuracy": 0.9303084462881088,
      "num_tokens": 70534164.0,
      "step": 5718
    },
    {
      "entropy": 1.0774936378002167,
      "epoch": 3.0115850447604,
      "grad_norm": 0.28812918066978455,
      "learning_rate": 0.0001344248918272487,
      "loss": 0.1853148192167282,
      "mean_token_accuracy": 0.9222434610128403,
      "num_tokens": 70546500.0,
      "step": 5719
    },
    {
      "entropy": 1.0731661915779114,
      "epoch": 3.012111637704055,
      "grad_norm": 0.2482844740152359,
      "learning_rate": 0.0001344037857810865,
      "loss": 0.16944734752178192,
      "mean_token_accuracy": 0.9281590282917023,
      "num_tokens": 70558836.0,
      "step": 5720
    },
    {
      "entropy": 1.0630016028881073,
      "epoch": 3.0126382306477093,
      "grad_norm": 0.27899959683418274,
      "learning_rate": 0.0001343826782862138,
      "loss": 0.17717432975769043,
      "mean_token_accuracy": 0.9256823360919952,
      "num_tokens": 70571172.0,
      "step": 5721
    },
    {
      "entropy": 1.0696978271007538,
      "epoch": 3.0131648235913637,
      "grad_norm": 0.28155818581581116,
      "learning_rate": 0.0001343615693438836,
      "loss": 0.17846429347991943,
      "mean_token_accuracy": 0.9255829900503159,
      "num_tokens": 70583508.0,
      "step": 5722
    },
    {
      "entropy": 1.0763144791126251,
      "epoch": 3.0136914165350186,
      "grad_norm": 0.28091639280319214,
      "learning_rate": 0.000134340458955349,
      "loss": 0.165150985121727,
      "mean_token_accuracy": 0.9292021095752716,
      "num_tokens": 70595844.0,
      "step": 5723
    },
    {
      "entropy": 1.0497006177902222,
      "epoch": 3.014218009478673,
      "grad_norm": 0.28318631649017334,
      "learning_rate": 0.00013431934712186328,
      "loss": 0.17467093467712402,
      "mean_token_accuracy": 0.9236724972724915,
      "num_tokens": 70608180.0,
      "step": 5724
    },
    {
      "entropy": 1.0530437529087067,
      "epoch": 3.0147446024223274,
      "grad_norm": 0.30431923270225525,
      "learning_rate": 0.00013429823384467958,
      "loss": 0.1854005753993988,
      "mean_token_accuracy": 0.9183477163314819,
      "num_tokens": 70620516.0,
      "step": 5725
    },
    {
      "entropy": 1.0483869463205338,
      "epoch": 3.0152711953659823,
      "grad_norm": 0.32633599638938904,
      "learning_rate": 0.0001342771191250514,
      "loss": 0.18871113657951355,
      "mean_token_accuracy": 0.9216799437999725,
      "num_tokens": 70632852.0,
      "step": 5726
    },
    {
      "entropy": 1.0900256037712097,
      "epoch": 3.0157977883096367,
      "grad_norm": 0.29401761293411255,
      "learning_rate": 0.00013425600296423218,
      "loss": 0.18117111921310425,
      "mean_token_accuracy": 0.9225471466779709,
      "num_tokens": 70645188.0,
      "step": 5727
    },
    {
      "entropy": 1.0106222331523895,
      "epoch": 3.016324381253291,
      "grad_norm": 0.31423333287239075,
      "learning_rate": 0.0001342348853634754,
      "loss": 0.19358471035957336,
      "mean_token_accuracy": 0.9177974313497543,
      "num_tokens": 70657524.0,
      "step": 5728
    },
    {
      "entropy": 1.0555269122123718,
      "epoch": 3.0168509741969456,
      "grad_norm": 0.3006555438041687,
      "learning_rate": 0.00013421376632403474,
      "loss": 0.18857192993164062,
      "mean_token_accuracy": 0.9178750962018967,
      "num_tokens": 70669860.0,
      "step": 5729
    },
    {
      "entropy": 1.0142070949077606,
      "epoch": 3.0173775671406005,
      "grad_norm": 0.28049519658088684,
      "learning_rate": 0.00013419264584716388,
      "loss": 0.17920871078968048,
      "mean_token_accuracy": 0.9233438372612,
      "num_tokens": 70682196.0,
      "step": 5730
    },
    {
      "entropy": 1.0379585921764374,
      "epoch": 3.017904160084255,
      "grad_norm": 0.3113473951816559,
      "learning_rate": 0.0001341715239341167,
      "loss": 0.19308455288410187,
      "mean_token_accuracy": 0.9195876121520996,
      "num_tokens": 70694532.0,
      "step": 5731
    },
    {
      "entropy": 1.0213616341352463,
      "epoch": 3.0184307530279093,
      "grad_norm": 0.27108868956565857,
      "learning_rate": 0.000134150400586147,
      "loss": 0.1776331663131714,
      "mean_token_accuracy": 0.9214796870946884,
      "num_tokens": 70706868.0,
      "step": 5732
    },
    {
      "entropy": 1.0628427565097809,
      "epoch": 3.018957345971564,
      "grad_norm": 0.29523083567619324,
      "learning_rate": 0.00013412927580450876,
      "loss": 0.16967853903770447,
      "mean_token_accuracy": 0.9279854595661163,
      "num_tokens": 70719204.0,
      "step": 5733
    },
    {
      "entropy": 1.0506865829229355,
      "epoch": 3.0194839389152186,
      "grad_norm": 0.2957015037536621,
      "learning_rate": 0.00013410814959045607,
      "loss": 0.19069498777389526,
      "mean_token_accuracy": 0.9195856302976608,
      "num_tokens": 70731540.0,
      "step": 5734
    },
    {
      "entropy": 1.0332060009241104,
      "epoch": 3.020010531858873,
      "grad_norm": 0.2643355429172516,
      "learning_rate": 0.000134087021945243,
      "loss": 0.16467365622520447,
      "mean_token_accuracy": 0.9302788972854614,
      "num_tokens": 70743876.0,
      "step": 5735
    },
    {
      "entropy": 1.0697474479675293,
      "epoch": 3.0205371248025275,
      "grad_norm": 0.2627510726451874,
      "learning_rate": 0.0001340658928701239,
      "loss": 0.17428165674209595,
      "mean_token_accuracy": 0.9276122748851776,
      "num_tokens": 70756212.0,
      "step": 5736
    },
    {
      "entropy": 1.0646617710590363,
      "epoch": 3.0210637177461823,
      "grad_norm": 0.2824498414993286,
      "learning_rate": 0.00013404476236635298,
      "loss": 0.1826760172843933,
      "mean_token_accuracy": 0.9254467487335205,
      "num_tokens": 70768548.0,
      "step": 5737
    },
    {
      "entropy": 1.0966342091560364,
      "epoch": 3.0215903106898367,
      "grad_norm": 0.3121627867221832,
      "learning_rate": 0.00013402363043518465,
      "loss": 0.2022552490234375,
      "mean_token_accuracy": 0.9123482406139374,
      "num_tokens": 70780884.0,
      "step": 5738
    },
    {
      "entropy": 1.0816529989242554,
      "epoch": 3.022116903633491,
      "grad_norm": 0.29687079787254333,
      "learning_rate": 0.0001340024970778734,
      "loss": 0.1894153505563736,
      "mean_token_accuracy": 0.9202007204294205,
      "num_tokens": 70793220.0,
      "step": 5739
    },
    {
      "entropy": 1.0896082818508148,
      "epoch": 3.022643496577146,
      "grad_norm": 0.2755042016506195,
      "learning_rate": 0.00013398136229567383,
      "loss": 0.16922765970230103,
      "mean_token_accuracy": 0.9306073486804962,
      "num_tokens": 70805556.0,
      "step": 5740
    },
    {
      "entropy": 1.0459871590137482,
      "epoch": 3.0231700895208005,
      "grad_norm": 0.2520120441913605,
      "learning_rate": 0.0001339602260898405,
      "loss": 0.17134934663772583,
      "mean_token_accuracy": 0.9289929568767548,
      "num_tokens": 70817892.0,
      "step": 5741
    },
    {
      "entropy": 1.0361694693565369,
      "epoch": 3.023696682464455,
      "grad_norm": 0.250061959028244,
      "learning_rate": 0.0001339390884616282,
      "loss": 0.16119395196437836,
      "mean_token_accuracy": 0.9309128522872925,
      "num_tokens": 70830228.0,
      "step": 5742
    },
    {
      "entropy": 1.071681022644043,
      "epoch": 3.0242232754081093,
      "grad_norm": 0.28559988737106323,
      "learning_rate": 0.0001339179494122918,
      "loss": 0.19253742694854736,
      "mean_token_accuracy": 0.9223012626171112,
      "num_tokens": 70842564.0,
      "step": 5743
    },
    {
      "entropy": 1.096194326877594,
      "epoch": 3.024749868351764,
      "grad_norm": 0.2857185900211334,
      "learning_rate": 0.00013389680894308612,
      "loss": 0.1934041976928711,
      "mean_token_accuracy": 0.918251097202301,
      "num_tokens": 70854900.0,
      "step": 5744
    },
    {
      "entropy": 1.0822049677371979,
      "epoch": 3.0252764612954186,
      "grad_norm": 0.2802906632423401,
      "learning_rate": 0.00013387566705526618,
      "loss": 0.16483765840530396,
      "mean_token_accuracy": 0.9284011572599411,
      "num_tokens": 70867236.0,
      "step": 5745
    },
    {
      "entropy": 1.108426570892334,
      "epoch": 3.025803054239073,
      "grad_norm": 0.2935839593410492,
      "learning_rate": 0.00013385452375008704,
      "loss": 0.18292377889156342,
      "mean_token_accuracy": 0.9248574525117874,
      "num_tokens": 70879572.0,
      "step": 5746
    },
    {
      "entropy": 1.056396335363388,
      "epoch": 3.026329647182728,
      "grad_norm": 0.2763398587703705,
      "learning_rate": 0.00013383337902880384,
      "loss": 0.16577623784542084,
      "mean_token_accuracy": 0.9300137013196945,
      "num_tokens": 70891908.0,
      "step": 5747
    },
    {
      "entropy": 1.1094647943973541,
      "epoch": 3.0268562401263823,
      "grad_norm": 0.29576507210731506,
      "learning_rate": 0.00013381223289267186,
      "loss": 0.18659496307373047,
      "mean_token_accuracy": 0.9255060255527496,
      "num_tokens": 70904244.0,
      "step": 5748
    },
    {
      "entropy": 1.0577907860279083,
      "epoch": 3.0273828330700367,
      "grad_norm": 0.2823028564453125,
      "learning_rate": 0.0001337910853429464,
      "loss": 0.1799258142709732,
      "mean_token_accuracy": 0.9247395992279053,
      "num_tokens": 70916580.0,
      "step": 5749
    },
    {
      "entropy": 1.0622846484184265,
      "epoch": 3.0279094260136916,
      "grad_norm": 0.2962205111980438,
      "learning_rate": 0.00013376993638088285,
      "loss": 0.19302281737327576,
      "mean_token_accuracy": 0.9213569909334183,
      "num_tokens": 70928916.0,
      "step": 5750
    },
    {
      "entropy": 1.1032023429870605,
      "epoch": 3.028436018957346,
      "grad_norm": 0.3092851936817169,
      "learning_rate": 0.00013374878600773676,
      "loss": 0.18477167189121246,
      "mean_token_accuracy": 0.9206407815217972,
      "num_tokens": 70941252.0,
      "step": 5751
    },
    {
      "entropy": 1.0572893917560577,
      "epoch": 3.0289626119010005,
      "grad_norm": 0.2842503786087036,
      "learning_rate": 0.00013372763422476365,
      "loss": 0.16644704341888428,
      "mean_token_accuracy": 0.9310343712568283,
      "num_tokens": 70953588.0,
      "step": 5752
    },
    {
      "entropy": 1.0782719254493713,
      "epoch": 3.029489204844655,
      "grad_norm": 0.2930472791194916,
      "learning_rate": 0.00013370648103321915,
      "loss": 0.17955656349658966,
      "mean_token_accuracy": 0.9243563711643219,
      "num_tokens": 70965924.0,
      "step": 5753
    },
    {
      "entropy": 1.0949573516845703,
      "epoch": 3.0300157977883098,
      "grad_norm": 0.2939299941062927,
      "learning_rate": 0.00013368532643435907,
      "loss": 0.18444274365901947,
      "mean_token_accuracy": 0.9251150488853455,
      "num_tokens": 70978260.0,
      "step": 5754
    },
    {
      "entropy": 1.1037635505199432,
      "epoch": 3.030542390731964,
      "grad_norm": 0.3208787739276886,
      "learning_rate": 0.0001336641704294392,
      "loss": 0.19437548518180847,
      "mean_token_accuracy": 0.9186409413814545,
      "num_tokens": 70990596.0,
      "step": 5755
    },
    {
      "entropy": 1.1095591187477112,
      "epoch": 3.0310689836756186,
      "grad_norm": 0.3024508059024811,
      "learning_rate": 0.00013364301301971548,
      "loss": 0.18942897021770477,
      "mean_token_accuracy": 0.9229587614536285,
      "num_tokens": 71002932.0,
      "step": 5756
    },
    {
      "entropy": 1.0845279693603516,
      "epoch": 3.0315955766192735,
      "grad_norm": 0.28569653630256653,
      "learning_rate": 0.00013362185420644386,
      "loss": 0.17577636241912842,
      "mean_token_accuracy": 0.9218069911003113,
      "num_tokens": 71015268.0,
      "step": 5757
    },
    {
      "entropy": 1.1118775308132172,
      "epoch": 3.032122169562928,
      "grad_norm": 0.258633553981781,
      "learning_rate": 0.00013360069399088044,
      "loss": 0.1673622876405716,
      "mean_token_accuracy": 0.9336172491312027,
      "num_tokens": 71027604.0,
      "step": 5758
    },
    {
      "entropy": 1.0533491671085358,
      "epoch": 3.0326487625065823,
      "grad_norm": 0.29966580867767334,
      "learning_rate": 0.00013357953237428137,
      "loss": 0.19163398444652557,
      "mean_token_accuracy": 0.9217822551727295,
      "num_tokens": 71039940.0,
      "step": 5759
    },
    {
      "entropy": 1.1153427362442017,
      "epoch": 3.0331753554502368,
      "grad_norm": 0.267869234085083,
      "learning_rate": 0.00013355836935790292,
      "loss": 0.16859161853790283,
      "mean_token_accuracy": 0.9335697442293167,
      "num_tokens": 71052276.0,
      "step": 5760
    },
    {
      "entropy": 1.1055462658405304,
      "epoch": 3.0337019483938916,
      "grad_norm": 0.28465765714645386,
      "learning_rate": 0.00013353720494300138,
      "loss": 0.17460112273693085,
      "mean_token_accuracy": 0.9235497415065765,
      "num_tokens": 71064612.0,
      "step": 5761
    },
    {
      "entropy": 1.1153737902641296,
      "epoch": 3.034228541337546,
      "grad_norm": 0.26903560757637024,
      "learning_rate": 0.00013351603913083315,
      "loss": 0.16478265821933746,
      "mean_token_accuracy": 0.9304361194372177,
      "num_tokens": 71076948.0,
      "step": 5762
    },
    {
      "entropy": 1.1499566733837128,
      "epoch": 3.0347551342812005,
      "grad_norm": 0.2969987392425537,
      "learning_rate": 0.00013349487192265472,
      "loss": 0.1865185797214508,
      "mean_token_accuracy": 0.9161245077848434,
      "num_tokens": 71089284.0,
      "step": 5763
    },
    {
      "entropy": 1.1352367103099823,
      "epoch": 3.0352817272248553,
      "grad_norm": 0.2893647253513336,
      "learning_rate": 0.00013347370331972272,
      "loss": 0.17901989817619324,
      "mean_token_accuracy": 0.9262292981147766,
      "num_tokens": 71101620.0,
      "step": 5764
    },
    {
      "entropy": 1.0742166340351105,
      "epoch": 3.0358083201685098,
      "grad_norm": 0.2570260763168335,
      "learning_rate": 0.00013345253332329375,
      "loss": 0.16443604230880737,
      "mean_token_accuracy": 0.9334456920623779,
      "num_tokens": 71113956.0,
      "step": 5765
    },
    {
      "entropy": 1.0759824514389038,
      "epoch": 3.036334913112164,
      "grad_norm": 0.2551080882549286,
      "learning_rate": 0.00013343136193462457,
      "loss": 0.1790422797203064,
      "mean_token_accuracy": 0.9245909750461578,
      "num_tokens": 71126292.0,
      "step": 5766
    },
    {
      "entropy": 1.1418220102787018,
      "epoch": 3.036861506055819,
      "grad_norm": 0.29069963097572327,
      "learning_rate": 0.000133410189154972,
      "loss": 0.17374365031719208,
      "mean_token_accuracy": 0.9302159249782562,
      "num_tokens": 71138628.0,
      "step": 5767
    },
    {
      "entropy": 1.0893676578998566,
      "epoch": 3.0373880989994735,
      "grad_norm": 0.25986167788505554,
      "learning_rate": 0.00013338901498559296,
      "loss": 0.15803691744804382,
      "mean_token_accuracy": 0.9312460720539093,
      "num_tokens": 71150964.0,
      "step": 5768
    },
    {
      "entropy": 1.1315541565418243,
      "epoch": 3.037914691943128,
      "grad_norm": 0.29130497574806213,
      "learning_rate": 0.0001333678394277444,
      "loss": 0.16455236077308655,
      "mean_token_accuracy": 0.9309317022562027,
      "num_tokens": 71163300.0,
      "step": 5769
    },
    {
      "entropy": 1.1340096592903137,
      "epoch": 3.0384412848867823,
      "grad_norm": 0.29143375158309937,
      "learning_rate": 0.0001333466624826834,
      "loss": 0.17422541975975037,
      "mean_token_accuracy": 0.9201169013977051,
      "num_tokens": 71175636.0,
      "step": 5770
    },
    {
      "entropy": 1.1219017505645752,
      "epoch": 3.038967877830437,
      "grad_norm": 0.29477885365486145,
      "learning_rate": 0.0001333254841516672,
      "loss": 0.1977756917476654,
      "mean_token_accuracy": 0.9190415889024734,
      "num_tokens": 71187972.0,
      "step": 5771
    },
    {
      "entropy": 1.15945702791214,
      "epoch": 3.0394944707740916,
      "grad_norm": 0.30047401785850525,
      "learning_rate": 0.00013330430443595287,
      "loss": 0.16675785183906555,
      "mean_token_accuracy": 0.9310600459575653,
      "num_tokens": 71200308.0,
      "step": 5772
    },
    {
      "entropy": 1.1178282797336578,
      "epoch": 3.040021063717746,
      "grad_norm": 0.2938012182712555,
      "learning_rate": 0.00013328312333679785,
      "loss": 0.20171600580215454,
      "mean_token_accuracy": 0.9149762839078903,
      "num_tokens": 71212644.0,
      "step": 5773
    },
    {
      "entropy": 1.1635751724243164,
      "epoch": 3.040547656661401,
      "grad_norm": 0.2849313020706177,
      "learning_rate": 0.00013326194085545948,
      "loss": 0.1755695790052414,
      "mean_token_accuracy": 0.9265087693929672,
      "num_tokens": 71224980.0,
      "step": 5774
    },
    {
      "entropy": 1.1646137833595276,
      "epoch": 3.0410742496050553,
      "grad_norm": 0.2710713744163513,
      "learning_rate": 0.0001332407569931953,
      "loss": 0.1735789030790329,
      "mean_token_accuracy": 0.9277262687683105,
      "num_tokens": 71237316.0,
      "step": 5775
    },
    {
      "entropy": 1.1816756129264832,
      "epoch": 3.0416008425487098,
      "grad_norm": 0.2794504463672638,
      "learning_rate": 0.0001332195717512628,
      "loss": 0.17608892917633057,
      "mean_token_accuracy": 0.9275052696466446,
      "num_tokens": 71249652.0,
      "step": 5776
    },
    {
      "entropy": 1.1921135187149048,
      "epoch": 3.042127435492364,
      "grad_norm": 0.2821495532989502,
      "learning_rate": 0.00013319838513091968,
      "loss": 0.18013815581798553,
      "mean_token_accuracy": 0.9232759028673172,
      "num_tokens": 71261988.0,
      "step": 5777
    },
    {
      "entropy": 1.1501494944095612,
      "epoch": 3.042654028436019,
      "grad_norm": 0.2837124168872833,
      "learning_rate": 0.00013317719713342368,
      "loss": 0.18457761406898499,
      "mean_token_accuracy": 0.9243838042020798,
      "num_tokens": 71274324.0,
      "step": 5778
    },
    {
      "entropy": 1.192774474620819,
      "epoch": 3.0431806213796735,
      "grad_norm": 0.30532482266426086,
      "learning_rate": 0.00013315600776003247,
      "loss": 0.18408426642417908,
      "mean_token_accuracy": 0.9200680702924728,
      "num_tokens": 71286660.0,
      "step": 5779
    },
    {
      "entropy": 1.136399120092392,
      "epoch": 3.043707214323328,
      "grad_norm": 0.26497748494148254,
      "learning_rate": 0.00013313481701200414,
      "loss": 0.16611620783805847,
      "mean_token_accuracy": 0.9339018017053604,
      "num_tokens": 71298996.0,
      "step": 5780
    },
    {
      "entropy": 1.1677029728889465,
      "epoch": 3.044233807266983,
      "grad_norm": 0.2965880334377289,
      "learning_rate": 0.00013311362489059654,
      "loss": 0.17575864493846893,
      "mean_token_accuracy": 0.927231565117836,
      "num_tokens": 71311332.0,
      "step": 5781
    },
    {
      "entropy": 1.1640919148921967,
      "epoch": 3.044760400210637,
      "grad_norm": 0.27395427227020264,
      "learning_rate": 0.00013309243139706772,
      "loss": 0.1687280684709549,
      "mean_token_accuracy": 0.9323958903551102,
      "num_tokens": 71323668.0,
      "step": 5782
    },
    {
      "entropy": 1.142760157585144,
      "epoch": 3.0452869931542916,
      "grad_norm": 0.3184906840324402,
      "learning_rate": 0.00013307123653267587,
      "loss": 0.18931150436401367,
      "mean_token_accuracy": 0.9203809797763824,
      "num_tokens": 71336004.0,
      "step": 5783
    },
    {
      "entropy": 1.1388859152793884,
      "epoch": 3.0458135860979465,
      "grad_norm": 0.2914552092552185,
      "learning_rate": 0.00013305004029867914,
      "loss": 0.1891023814678192,
      "mean_token_accuracy": 0.9190599769353867,
      "num_tokens": 71348340.0,
      "step": 5784
    },
    {
      "entropy": 1.1550920009613037,
      "epoch": 3.046340179041601,
      "grad_norm": 0.3001711964607239,
      "learning_rate": 0.0001330288426963359,
      "loss": 0.1784719079732895,
      "mean_token_accuracy": 0.9228701889514923,
      "num_tokens": 71360676.0,
      "step": 5785
    },
    {
      "entropy": 1.163255751132965,
      "epoch": 3.0468667719852554,
      "grad_norm": 0.3006163537502289,
      "learning_rate": 0.00013300764372690447,
      "loss": 0.20307712256908417,
      "mean_token_accuracy": 0.915765792131424,
      "num_tokens": 71373012.0,
      "step": 5786
    },
    {
      "entropy": 1.1174058318138123,
      "epoch": 3.0473933649289098,
      "grad_norm": 0.28055649995803833,
      "learning_rate": 0.00013298644339164334,
      "loss": 0.17206884920597076,
      "mean_token_accuracy": 0.9293131530284882,
      "num_tokens": 71385348.0,
      "step": 5787
    },
    {
      "entropy": 1.1481737196445465,
      "epoch": 3.0479199578725646,
      "grad_norm": 0.2757937014102936,
      "learning_rate": 0.00013296524169181107,
      "loss": 0.1666310727596283,
      "mean_token_accuracy": 0.9299214482307434,
      "num_tokens": 71397684.0,
      "step": 5788
    },
    {
      "entropy": 1.1161286234855652,
      "epoch": 3.048446550816219,
      "grad_norm": 0.28487345576286316,
      "learning_rate": 0.0001329440386286662,
      "loss": 0.17753304541110992,
      "mean_token_accuracy": 0.9214348793029785,
      "num_tokens": 71410020.0,
      "step": 5789
    },
    {
      "entropy": 1.1513926088809967,
      "epoch": 3.0489731437598735,
      "grad_norm": 0.2864345610141754,
      "learning_rate": 0.0001329228342034675,
      "loss": 0.17936980724334717,
      "mean_token_accuracy": 0.9247641712427139,
      "num_tokens": 71422356.0,
      "step": 5790
    },
    {
      "entropy": 1.1690279245376587,
      "epoch": 3.0494997367035284,
      "grad_norm": 0.29427191615104675,
      "learning_rate": 0.00013290162841747376,
      "loss": 0.18683072924613953,
      "mean_token_accuracy": 0.9196323603391647,
      "num_tokens": 71434692.0,
      "step": 5791
    },
    {
      "entropy": 1.1411975026130676,
      "epoch": 3.050026329647183,
      "grad_norm": 0.29006507992744446,
      "learning_rate": 0.0001328804212719438,
      "loss": 0.17810499668121338,
      "mean_token_accuracy": 0.9264704585075378,
      "num_tokens": 71447028.0,
      "step": 5792
    },
    {
      "entropy": 1.0939705073833466,
      "epoch": 3.050552922590837,
      "grad_norm": 0.28467386960983276,
      "learning_rate": 0.0001328592127681366,
      "loss": 0.16793327033519745,
      "mean_token_accuracy": 0.9279095828533173,
      "num_tokens": 71459364.0,
      "step": 5793
    },
    {
      "entropy": 1.0950967967510223,
      "epoch": 3.0510795155344916,
      "grad_norm": 0.2878323793411255,
      "learning_rate": 0.00013283800290731114,
      "loss": 0.17144641280174255,
      "mean_token_accuracy": 0.9291983097791672,
      "num_tokens": 71471700.0,
      "step": 5794
    },
    {
      "entropy": 1.1139646768569946,
      "epoch": 3.0516061084781465,
      "grad_norm": 0.2899021804332733,
      "learning_rate": 0.0001328167916907266,
      "loss": 0.1691819280385971,
      "mean_token_accuracy": 0.9288174957036972,
      "num_tokens": 71484036.0,
      "step": 5795
    },
    {
      "entropy": 1.0817930102348328,
      "epoch": 3.052132701421801,
      "grad_norm": 0.2783130705356598,
      "learning_rate": 0.0001327955791196421,
      "loss": 0.16818395256996155,
      "mean_token_accuracy": 0.9270586669445038,
      "num_tokens": 71496372.0,
      "step": 5796
    },
    {
      "entropy": 1.1195127665996552,
      "epoch": 3.0526592943654554,
      "grad_norm": 0.29349660873413086,
      "learning_rate": 0.0001327743651953169,
      "loss": 0.17434924840927124,
      "mean_token_accuracy": 0.9261120557785034,
      "num_tokens": 71508708.0,
      "step": 5797
    },
    {
      "entropy": 1.1135068237781525,
      "epoch": 3.0531858873091102,
      "grad_norm": 0.2993173897266388,
      "learning_rate": 0.00013275314991901044,
      "loss": 0.18269886076450348,
      "mean_token_accuracy": 0.9268842190504074,
      "num_tokens": 71521044.0,
      "step": 5798
    },
    {
      "entropy": 1.0761390030384064,
      "epoch": 3.0537124802527646,
      "grad_norm": 0.3032976984977722,
      "learning_rate": 0.00013273193329198205,
      "loss": 0.18568283319473267,
      "mean_token_accuracy": 0.9237289726734161,
      "num_tokens": 71533380.0,
      "step": 5799
    },
    {
      "entropy": 1.0722538530826569,
      "epoch": 3.054239073196419,
      "grad_norm": 0.3053904175758362,
      "learning_rate": 0.0001327107153154913,
      "loss": 0.18497273325920105,
      "mean_token_accuracy": 0.921196460723877,
      "num_tokens": 71545716.0,
      "step": 5800
    },
    {
      "entropy": 1.0652728378772736,
      "epoch": 3.0547656661400735,
      "grad_norm": 0.2781349718570709,
      "learning_rate": 0.0001326894959907977,
      "loss": 0.18898527324199677,
      "mean_token_accuracy": 0.9184873849153519,
      "num_tokens": 71558052.0,
      "step": 5801
    },
    {
      "entropy": 1.0627508759498596,
      "epoch": 3.0552922590837284,
      "grad_norm": 0.30204126238822937,
      "learning_rate": 0.000132668275319161,
      "loss": 0.19641637802124023,
      "mean_token_accuracy": 0.922800600528717,
      "num_tokens": 71570388.0,
      "step": 5802
    },
    {
      "entropy": 1.0257129222154617,
      "epoch": 3.055818852027383,
      "grad_norm": 0.2787397503852844,
      "learning_rate": 0.00013264705330184092,
      "loss": 0.17782054841518402,
      "mean_token_accuracy": 0.9246032387018204,
      "num_tokens": 71582724.0,
      "step": 5803
    },
    {
      "entropy": 1.0518714487552643,
      "epoch": 3.056345444971037,
      "grad_norm": 0.27710703015327454,
      "learning_rate": 0.0001326258299400973,
      "loss": 0.18275544047355652,
      "mean_token_accuracy": 0.9222777634859085,
      "num_tokens": 71595060.0,
      "step": 5804
    },
    {
      "entropy": 1.0704227536916733,
      "epoch": 3.056872037914692,
      "grad_norm": 0.3062211573123932,
      "learning_rate": 0.00013260460523519,
      "loss": 0.18094666302204132,
      "mean_token_accuracy": 0.9187027215957642,
      "num_tokens": 71607396.0,
      "step": 5805
    },
    {
      "entropy": 1.051187515258789,
      "epoch": 3.0573986308583465,
      "grad_norm": 0.28965574502944946,
      "learning_rate": 0.00013258337918837905,
      "loss": 0.18023884296417236,
      "mean_token_accuracy": 0.9190809279680252,
      "num_tokens": 71619732.0,
      "step": 5806
    },
    {
      "entropy": 1.1001330316066742,
      "epoch": 3.057925223802001,
      "grad_norm": 0.30051374435424805,
      "learning_rate": 0.00013256215180092447,
      "loss": 0.18936161696910858,
      "mean_token_accuracy": 0.9188069850206375,
      "num_tokens": 71632068.0,
      "step": 5807
    },
    {
      "entropy": 1.0515406131744385,
      "epoch": 3.058451816745656,
      "grad_norm": 0.2791909873485565,
      "learning_rate": 0.0001325409230740865,
      "loss": 0.17704379558563232,
      "mean_token_accuracy": 0.9265051782131195,
      "num_tokens": 71644404.0,
      "step": 5808
    },
    {
      "entropy": 1.0193083584308624,
      "epoch": 3.0589784096893102,
      "grad_norm": 0.285780668258667,
      "learning_rate": 0.00013251969300912523,
      "loss": 0.16971568763256073,
      "mean_token_accuracy": 0.9304918050765991,
      "num_tokens": 71656740.0,
      "step": 5809
    },
    {
      "entropy": 0.9988278597593307,
      "epoch": 3.0595050026329647,
      "grad_norm": 0.2850463390350342,
      "learning_rate": 0.0001324984616073011,
      "loss": 0.17589221894741058,
      "mean_token_accuracy": 0.9308154135942459,
      "num_tokens": 71669076.0,
      "step": 5810
    },
    {
      "entropy": 1.0253951698541641,
      "epoch": 3.060031595576619,
      "grad_norm": 0.28472432494163513,
      "learning_rate": 0.00013247722886987443,
      "loss": 0.1638464778661728,
      "mean_token_accuracy": 0.9300073832273483,
      "num_tokens": 71681412.0,
      "step": 5811
    },
    {
      "entropy": 1.023746833205223,
      "epoch": 3.060558188520274,
      "grad_norm": 0.28646063804626465,
      "learning_rate": 0.00013245599479810564,
      "loss": 0.1735069900751114,
      "mean_token_accuracy": 0.9263989329338074,
      "num_tokens": 71693748.0,
      "step": 5812
    },
    {
      "entropy": 0.9947127848863602,
      "epoch": 3.0610847814639284,
      "grad_norm": 0.27799171209335327,
      "learning_rate": 0.00013243475939325538,
      "loss": 0.17884136736392975,
      "mean_token_accuracy": 0.9239570498466492,
      "num_tokens": 71706084.0,
      "step": 5813
    },
    {
      "entropy": 1.0348500907421112,
      "epoch": 3.061611374407583,
      "grad_norm": 0.29147857427597046,
      "learning_rate": 0.00013241352265658417,
      "loss": 0.17199379205703735,
      "mean_token_accuracy": 0.9261778295040131,
      "num_tokens": 71718420.0,
      "step": 5814
    },
    {
      "entropy": 0.9770552963018417,
      "epoch": 3.0621379673512377,
      "grad_norm": 0.26427361369132996,
      "learning_rate": 0.0001323922845893528,
      "loss": 0.17064280807971954,
      "mean_token_accuracy": 0.9300706535577774,
      "num_tokens": 71730756.0,
      "step": 5815
    },
    {
      "entropy": 1.0042769014835358,
      "epoch": 3.062664560294892,
      "grad_norm": 0.274197518825531,
      "learning_rate": 0.00013237104519282198,
      "loss": 0.15780170261859894,
      "mean_token_accuracy": 0.9329145699739456,
      "num_tokens": 71743092.0,
      "step": 5816
    },
    {
      "entropy": 1.0151778310537338,
      "epoch": 3.0631911532385465,
      "grad_norm": 0.26683470606803894,
      "learning_rate": 0.00013234980446825257,
      "loss": 0.1659105122089386,
      "mean_token_accuracy": 0.9312007278203964,
      "num_tokens": 71755428.0,
      "step": 5817
    },
    {
      "entropy": 0.9765558987855911,
      "epoch": 3.063717746182201,
      "grad_norm": 0.2901502847671509,
      "learning_rate": 0.00013232856241690555,
      "loss": 0.18892541527748108,
      "mean_token_accuracy": 0.918955847620964,
      "num_tokens": 71767764.0,
      "step": 5818
    },
    {
      "entropy": 1.0077444463968277,
      "epoch": 3.064244339125856,
      "grad_norm": 0.3046390414237976,
      "learning_rate": 0.000132307319040042,
      "loss": 0.18454664945602417,
      "mean_token_accuracy": 0.9204002022743225,
      "num_tokens": 71780100.0,
      "step": 5819
    },
    {
      "entropy": 1.030941441655159,
      "epoch": 3.0647709320695102,
      "grad_norm": 0.31215187907218933,
      "learning_rate": 0.00013228607433892284,
      "loss": 0.20272856950759888,
      "mean_token_accuracy": 0.9166811406612396,
      "num_tokens": 71792436.0,
      "step": 5820
    },
    {
      "entropy": 0.9970352947711945,
      "epoch": 3.0652975250131647,
      "grad_norm": 0.26677122712135315,
      "learning_rate": 0.00013226482831480937,
      "loss": 0.17062878608703613,
      "mean_token_accuracy": 0.9292921721935272,
      "num_tokens": 71804772.0,
      "step": 5821
    },
    {
      "entropy": 0.9916958510875702,
      "epoch": 3.0658241179568195,
      "grad_norm": 0.3014390170574188,
      "learning_rate": 0.00013224358096896279,
      "loss": 0.1784030795097351,
      "mean_token_accuracy": 0.9230726957321167,
      "num_tokens": 71817108.0,
      "step": 5822
    },
    {
      "entropy": 1.003650039434433,
      "epoch": 3.066350710900474,
      "grad_norm": 0.30364856123924255,
      "learning_rate": 0.0001322223323026445,
      "loss": 0.18567705154418945,
      "mean_token_accuracy": 0.9203255027532578,
      "num_tokens": 71829444.0,
      "step": 5823
    },
    {
      "entropy": 0.9895204901695251,
      "epoch": 3.0668773038441284,
      "grad_norm": 0.28889593482017517,
      "learning_rate": 0.0001322010823171158,
      "loss": 0.1736719310283661,
      "mean_token_accuracy": 0.9261168986558914,
      "num_tokens": 71841780.0,
      "step": 5824
    },
    {
      "entropy": 0.9563289880752563,
      "epoch": 3.0674038967877832,
      "grad_norm": 0.2949967682361603,
      "learning_rate": 0.0001321798310136383,
      "loss": 0.18222150206565857,
      "mean_token_accuracy": 0.9264791905879974,
      "num_tokens": 71854116.0,
      "step": 5825
    },
    {
      "entropy": 0.9729903489351273,
      "epoch": 3.0679304897314377,
      "grad_norm": 0.2907126843929291,
      "learning_rate": 0.00013215857839347347,
      "loss": 0.18862363696098328,
      "mean_token_accuracy": 0.9187849313020706,
      "num_tokens": 71866452.0,
      "step": 5826
    },
    {
      "entropy": 0.9874851405620575,
      "epoch": 3.068457082675092,
      "grad_norm": 0.3004235029220581,
      "learning_rate": 0.00013213732445788303,
      "loss": 0.19510529935359955,
      "mean_token_accuracy": 0.9176079481840134,
      "num_tokens": 71878788.0,
      "step": 5827
    },
    {
      "entropy": 0.9737606048583984,
      "epoch": 3.0689836756187465,
      "grad_norm": 0.29955679178237915,
      "learning_rate": 0.00013211606920812864,
      "loss": 0.1905735284090042,
      "mean_token_accuracy": 0.9228790551424026,
      "num_tokens": 71891124.0,
      "step": 5828
    },
    {
      "entropy": 0.9671222120523453,
      "epoch": 3.0695102685624014,
      "grad_norm": 0.2844952642917633,
      "learning_rate": 0.0001320948126454721,
      "loss": 0.18964330852031708,
      "mean_token_accuracy": 0.918824627995491,
      "num_tokens": 71903460.0,
      "step": 5829
    },
    {
      "entropy": 0.9817742556333542,
      "epoch": 3.070036861506056,
      "grad_norm": 0.2865328788757324,
      "learning_rate": 0.00013207355477117534,
      "loss": 0.184239462018013,
      "mean_token_accuracy": 0.9218413978815079,
      "num_tokens": 71915796.0,
      "step": 5830
    },
    {
      "entropy": 1.0103529393672943,
      "epoch": 3.0705634544497102,
      "grad_norm": 0.2857781946659088,
      "learning_rate": 0.00013205229558650028,
      "loss": 0.1801823377609253,
      "mean_token_accuracy": 0.9265488833189011,
      "num_tokens": 71928132.0,
      "step": 5831
    },
    {
      "entropy": 1.0271504372358322,
      "epoch": 3.071090047393365,
      "grad_norm": 0.27296462655067444,
      "learning_rate": 0.00013203103509270894,
      "loss": 0.16963928937911987,
      "mean_token_accuracy": 0.9278423190116882,
      "num_tokens": 71940468.0,
      "step": 5832
    },
    {
      "entropy": 1.0373024642467499,
      "epoch": 3.0716166403370195,
      "grad_norm": 0.32619747519493103,
      "learning_rate": 0.0001320097732910635,
      "loss": 0.20516563951969147,
      "mean_token_accuracy": 0.9147870093584061,
      "num_tokens": 71952804.0,
      "step": 5833
    },
    {
      "entropy": 1.0268867909908295,
      "epoch": 3.072143233280674,
      "grad_norm": 0.27325931191444397,
      "learning_rate": 0.00013198851018282606,
      "loss": 0.18094776570796967,
      "mean_token_accuracy": 0.9250821322202682,
      "num_tokens": 71965140.0,
      "step": 5834
    },
    {
      "entropy": 0.9949807524681091,
      "epoch": 3.0726698262243284,
      "grad_norm": 0.2740168869495392,
      "learning_rate": 0.00013196724576925898,
      "loss": 0.18525102734565735,
      "mean_token_accuracy": 0.9229671359062195,
      "num_tokens": 71977476.0,
      "step": 5835
    },
    {
      "entropy": 1.00574392080307,
      "epoch": 3.0731964191679833,
      "grad_norm": 0.2765384614467621,
      "learning_rate": 0.00013194598005162447,
      "loss": 0.16539379954338074,
      "mean_token_accuracy": 0.9317118972539902,
      "num_tokens": 71989812.0,
      "step": 5836
    },
    {
      "entropy": 1.0072713494300842,
      "epoch": 3.0737230121116377,
      "grad_norm": 0.2896250784397125,
      "learning_rate": 0.0001319247130311851,
      "loss": 0.1674274504184723,
      "mean_token_accuracy": 0.9269036948680878,
      "num_tokens": 72002148.0,
      "step": 5837
    },
    {
      "entropy": 1.0159356147050858,
      "epoch": 3.074249605055292,
      "grad_norm": 0.2608317732810974,
      "learning_rate": 0.00013190344470920327,
      "loss": 0.17006650567054749,
      "mean_token_accuracy": 0.9248785823583603,
      "num_tokens": 72014484.0,
      "step": 5838
    },
    {
      "entropy": 1.0277487933635712,
      "epoch": 3.074776197998947,
      "grad_norm": 0.29685619473457336,
      "learning_rate": 0.00013188217508694163,
      "loss": 0.19034484028816223,
      "mean_token_accuracy": 0.9209220856428146,
      "num_tokens": 72026820.0,
      "step": 5839
    },
    {
      "entropy": 0.9729027599096298,
      "epoch": 3.0753027909426014,
      "grad_norm": 0.2588353157043457,
      "learning_rate": 0.00013186090416566278,
      "loss": 0.16739526391029358,
      "mean_token_accuracy": 0.9292702972888947,
      "num_tokens": 72039156.0,
      "step": 5840
    },
    {
      "entropy": 0.946832001209259,
      "epoch": 3.075829383886256,
      "grad_norm": 0.29065054655075073,
      "learning_rate": 0.00013183963194662946,
      "loss": 0.17941366136074066,
      "mean_token_accuracy": 0.9269576519727707,
      "num_tokens": 72051492.0,
      "step": 5841
    },
    {
      "entropy": 0.9500826895236969,
      "epoch": 3.0763559768299107,
      "grad_norm": 0.30167078971862793,
      "learning_rate": 0.00013181835843110448,
      "loss": 0.18562158942222595,
      "mean_token_accuracy": 0.9208818674087524,
      "num_tokens": 72063828.0,
      "step": 5842
    },
    {
      "entropy": 0.9446297138929367,
      "epoch": 3.076882569773565,
      "grad_norm": 0.2649385929107666,
      "learning_rate": 0.00013179708362035073,
      "loss": 0.18260982632637024,
      "mean_token_accuracy": 0.9223494529724121,
      "num_tokens": 72076164.0,
      "step": 5843
    },
    {
      "entropy": 0.9216752499341965,
      "epoch": 3.0774091627172195,
      "grad_norm": 0.29369136691093445,
      "learning_rate": 0.00013177580751563118,
      "loss": 0.17421871423721313,
      "mean_token_accuracy": 0.9291995912790298,
      "num_tokens": 72088500.0,
      "step": 5844
    },
    {
      "entropy": 1.0294009149074554,
      "epoch": 3.077935755660874,
      "grad_norm": 0.2864926755428314,
      "learning_rate": 0.0001317545301182088,
      "loss": 0.16719837486743927,
      "mean_token_accuracy": 0.9276296347379684,
      "num_tokens": 72100836.0,
      "step": 5845
    },
    {
      "entropy": 1.0061109811067581,
      "epoch": 3.078462348604529,
      "grad_norm": 0.31383106112480164,
      "learning_rate": 0.00013173325142934683,
      "loss": 0.1888384222984314,
      "mean_token_accuracy": 0.9204787760972977,
      "num_tokens": 72113172.0,
      "step": 5846
    },
    {
      "entropy": 0.946212887763977,
      "epoch": 3.0789889415481833,
      "grad_norm": 0.2886209189891815,
      "learning_rate": 0.00013171197145030837,
      "loss": 0.18259429931640625,
      "mean_token_accuracy": 0.9196333736181259,
      "num_tokens": 72125508.0,
      "step": 5847
    },
    {
      "entropy": 1.0319083482027054,
      "epoch": 3.0795155344918377,
      "grad_norm": 0.3208879232406616,
      "learning_rate": 0.0001316906901823567,
      "loss": 0.19582988321781158,
      "mean_token_accuracy": 0.9173017740249634,
      "num_tokens": 72137844.0,
      "step": 5848
    },
    {
      "entropy": 0.9948568791151047,
      "epoch": 3.0800421274354925,
      "grad_norm": 0.2987675666809082,
      "learning_rate": 0.00013166940762675522,
      "loss": 0.18619319796562195,
      "mean_token_accuracy": 0.9224425554275513,
      "num_tokens": 72150180.0,
      "step": 5849
    },
    {
      "entropy": 0.9998792558908463,
      "epoch": 3.080568720379147,
      "grad_norm": 0.29199832677841187,
      "learning_rate": 0.00013164812378476726,
      "loss": 0.1841306984424591,
      "mean_token_accuracy": 0.9235689789056778,
      "num_tokens": 72162516.0,
      "step": 5850
    },
    {
      "entropy": 0.9340603649616241,
      "epoch": 3.0810953133228014,
      "grad_norm": 0.29936087131500244,
      "learning_rate": 0.00013162683865765637,
      "loss": 0.18153662979602814,
      "mean_token_accuracy": 0.920834869146347,
      "num_tokens": 72174852.0,
      "step": 5851
    },
    {
      "entropy": 0.9640193283557892,
      "epoch": 3.081621906266456,
      "grad_norm": 0.32455864548683167,
      "learning_rate": 0.00013160555224668615,
      "loss": 0.18892093002796173,
      "mean_token_accuracy": 0.9217241108417511,
      "num_tokens": 72187188.0,
      "step": 5852
    },
    {
      "entropy": 0.9838850498199463,
      "epoch": 3.0821484992101107,
      "grad_norm": 0.29325219988822937,
      "learning_rate": 0.00013158426455312018,
      "loss": 0.1794615536928177,
      "mean_token_accuracy": 0.9225403517484665,
      "num_tokens": 72199524.0,
      "step": 5853
    },
    {
      "entropy": 0.9487807154655457,
      "epoch": 3.082675092153765,
      "grad_norm": 0.2752475440502167,
      "learning_rate": 0.00013156297557822224,
      "loss": 0.15716516971588135,
      "mean_token_accuracy": 0.9343967288732529,
      "num_tokens": 72211860.0,
      "step": 5854
    },
    {
      "entropy": 1.0301673710346222,
      "epoch": 3.0832016850974195,
      "grad_norm": 0.3180379569530487,
      "learning_rate": 0.00013154168532325615,
      "loss": 0.20467087626457214,
      "mean_token_accuracy": 0.9166474789381027,
      "num_tokens": 72224196.0,
      "step": 5855
    },
    {
      "entropy": 1.004869282245636,
      "epoch": 3.0837282780410744,
      "grad_norm": 0.28707006573677063,
      "learning_rate": 0.00013152039378948574,
      "loss": 0.18228140473365784,
      "mean_token_accuracy": 0.9262249171733856,
      "num_tokens": 72236532.0,
      "step": 5856
    },
    {
      "entropy": 1.0051259696483612,
      "epoch": 3.084254870984729,
      "grad_norm": 0.27821314334869385,
      "learning_rate": 0.00013149910097817494,
      "loss": 0.16908696293830872,
      "mean_token_accuracy": 0.9320413321256638,
      "num_tokens": 72248868.0,
      "step": 5857
    },
    {
      "entropy": 1.000169724225998,
      "epoch": 3.0847814639283833,
      "grad_norm": 0.2890099287033081,
      "learning_rate": 0.00013147780689058784,
      "loss": 0.16586747765541077,
      "mean_token_accuracy": 0.9278808683156967,
      "num_tokens": 72261204.0,
      "step": 5858
    },
    {
      "entropy": 1.0450342297554016,
      "epoch": 3.085308056872038,
      "grad_norm": 0.2950303256511688,
      "learning_rate": 0.00013145651152798857,
      "loss": 0.17741480469703674,
      "mean_token_accuracy": 0.9268434792757034,
      "num_tokens": 72273540.0,
      "step": 5859
    },
    {
      "entropy": 1.031152680516243,
      "epoch": 3.0858346498156926,
      "grad_norm": 0.31254854798316956,
      "learning_rate": 0.00013143521489164124,
      "loss": 0.1890886127948761,
      "mean_token_accuracy": 0.9225743561983109,
      "num_tokens": 72285876.0,
      "step": 5860
    },
    {
      "entropy": 1.0067125707864761,
      "epoch": 3.086361242759347,
      "grad_norm": 0.2938244938850403,
      "learning_rate": 0.00013141391698281012,
      "loss": 0.18479709327220917,
      "mean_token_accuracy": 0.9220405369997025,
      "num_tokens": 72298212.0,
      "step": 5861
    },
    {
      "entropy": 0.974154531955719,
      "epoch": 3.0868878357030014,
      "grad_norm": 0.26539933681488037,
      "learning_rate": 0.00013139261780275958,
      "loss": 0.1614721566438675,
      "mean_token_accuracy": 0.9337578117847443,
      "num_tokens": 72310548.0,
      "step": 5862
    },
    {
      "entropy": 0.9784013479948044,
      "epoch": 3.0874144286466563,
      "grad_norm": 0.2748902142047882,
      "learning_rate": 0.00013137131735275402,
      "loss": 0.16838830709457397,
      "mean_token_accuracy": 0.9306375980377197,
      "num_tokens": 72322884.0,
      "step": 5863
    },
    {
      "entropy": 1.03346548974514,
      "epoch": 3.0879410215903107,
      "grad_norm": 0.33242136240005493,
      "learning_rate": 0.00013135001563405787,
      "loss": 0.1944301426410675,
      "mean_token_accuracy": 0.913677766919136,
      "num_tokens": 72335220.0,
      "step": 5864
    },
    {
      "entropy": 0.9649850279092789,
      "epoch": 3.088467614533965,
      "grad_norm": 0.28117993474006653,
      "learning_rate": 0.00013132871264793573,
      "loss": 0.17021432518959045,
      "mean_token_accuracy": 0.928614541888237,
      "num_tokens": 72347556.0,
      "step": 5865
    },
    {
      "entropy": 0.9285301268100739,
      "epoch": 3.08899420747762,
      "grad_norm": 0.27827489376068115,
      "learning_rate": 0.00013130740839565228,
      "loss": 0.15523271262645721,
      "mean_token_accuracy": 0.9350172430276871,
      "num_tokens": 72359892.0,
      "step": 5866
    },
    {
      "entropy": 0.9720413982868195,
      "epoch": 3.0895208004212744,
      "grad_norm": 0.28834617137908936,
      "learning_rate": 0.00013128610287847214,
      "loss": 0.1742348074913025,
      "mean_token_accuracy": 0.9259906262159348,
      "num_tokens": 72372228.0,
      "step": 5867
    },
    {
      "entropy": 0.9477108120918274,
      "epoch": 3.090047393364929,
      "grad_norm": 0.2890585958957672,
      "learning_rate": 0.00013126479609766017,
      "loss": 0.17141348123550415,
      "mean_token_accuracy": 0.9350418746471405,
      "num_tokens": 72384564.0,
      "step": 5868
    },
    {
      "entropy": 0.9333702474832535,
      "epoch": 3.0905739863085833,
      "grad_norm": 0.3177509009838104,
      "learning_rate": 0.00013124348805448118,
      "loss": 0.17872607707977295,
      "mean_token_accuracy": 0.9237185716629028,
      "num_tokens": 72396900.0,
      "step": 5869
    },
    {
      "entropy": 0.9048350304365158,
      "epoch": 3.091100579252238,
      "grad_norm": 0.29665181040763855,
      "learning_rate": 0.00013122217875020012,
      "loss": 0.1873965859413147,
      "mean_token_accuracy": 0.9245747178792953,
      "num_tokens": 72409236.0,
      "step": 5870
    },
    {
      "entropy": 0.9140274971723557,
      "epoch": 3.0916271721958926,
      "grad_norm": 0.2923314869403839,
      "learning_rate": 0.000131200868186082,
      "loss": 0.1777024120092392,
      "mean_token_accuracy": 0.9223998188972473,
      "num_tokens": 72421572.0,
      "step": 5871
    },
    {
      "entropy": 0.93026502430439,
      "epoch": 3.092153765139547,
      "grad_norm": 0.2896692156791687,
      "learning_rate": 0.0001311795563633919,
      "loss": 0.17808280885219574,
      "mean_token_accuracy": 0.9236349910497665,
      "num_tokens": 72433908.0,
      "step": 5872
    },
    {
      "entropy": 0.9688552618026733,
      "epoch": 3.092680358083202,
      "grad_norm": 0.30245140194892883,
      "learning_rate": 0.00013115824328339498,
      "loss": 0.1754363775253296,
      "mean_token_accuracy": 0.9290272146463394,
      "num_tokens": 72446244.0,
      "step": 5873
    },
    {
      "entropy": 0.9436691403388977,
      "epoch": 3.0932069510268563,
      "grad_norm": 0.3126986622810364,
      "learning_rate": 0.00013113692894735647,
      "loss": 0.19607993960380554,
      "mean_token_accuracy": 0.919350653886795,
      "num_tokens": 72458580.0,
      "step": 5874
    },
    {
      "entropy": 0.9906564652919769,
      "epoch": 3.0937335439705107,
      "grad_norm": 0.288845956325531,
      "learning_rate": 0.00013111561335654173,
      "loss": 0.1837921440601349,
      "mean_token_accuracy": 0.923243373632431,
      "num_tokens": 72470916.0,
      "step": 5875
    },
    {
      "entropy": 0.9391609728336334,
      "epoch": 3.0942601369141656,
      "grad_norm": 0.28532668948173523,
      "learning_rate": 0.00013109429651221604,
      "loss": 0.15645398199558258,
      "mean_token_accuracy": 0.9355584979057312,
      "num_tokens": 72483252.0,
      "step": 5876
    },
    {
      "entropy": 0.9522290229797363,
      "epoch": 3.09478672985782,
      "grad_norm": 0.2875584065914154,
      "learning_rate": 0.00013107297841564495,
      "loss": 0.17390328645706177,
      "mean_token_accuracy": 0.926831528544426,
      "num_tokens": 72495588.0,
      "step": 5877
    },
    {
      "entropy": 1.0038781613111496,
      "epoch": 3.0953133228014744,
      "grad_norm": 0.31649264693260193,
      "learning_rate": 0.00013105165906809394,
      "loss": 0.17345508933067322,
      "mean_token_accuracy": 0.9273377507925034,
      "num_tokens": 72507924.0,
      "step": 5878
    },
    {
      "entropy": 0.953633263707161,
      "epoch": 3.095839915745129,
      "grad_norm": 0.27122682332992554,
      "learning_rate": 0.00013103033847082862,
      "loss": 0.17186909914016724,
      "mean_token_accuracy": 0.9262486696243286,
      "num_tokens": 72520260.0,
      "step": 5879
    },
    {
      "entropy": 0.9963507354259491,
      "epoch": 3.0963665086887837,
      "grad_norm": 0.29431045055389404,
      "learning_rate": 0.0001310090166251147,
      "loss": 0.1882622390985489,
      "mean_token_accuracy": 0.9177558422088623,
      "num_tokens": 72532596.0,
      "step": 5880
    },
    {
      "entropy": 1.0024547725915909,
      "epoch": 3.096893101632438,
      "grad_norm": 0.28813937306404114,
      "learning_rate": 0.00013098769353221787,
      "loss": 0.18425771594047546,
      "mean_token_accuracy": 0.9189359992742538,
      "num_tokens": 72544932.0,
      "step": 5881
    },
    {
      "entropy": 0.9621204137802124,
      "epoch": 3.0974196945760926,
      "grad_norm": 0.27608561515808105,
      "learning_rate": 0.00013096636919340403,
      "loss": 0.1865629106760025,
      "mean_token_accuracy": 0.9199081361293793,
      "num_tokens": 72557268.0,
      "step": 5882
    },
    {
      "entropy": 1.003304123878479,
      "epoch": 3.0979462875197474,
      "grad_norm": 0.29859909415245056,
      "learning_rate": 0.00013094504360993908,
      "loss": 0.1924053430557251,
      "mean_token_accuracy": 0.9214104861021042,
      "num_tokens": 72569604.0,
      "step": 5883
    },
    {
      "entropy": 1.025260254740715,
      "epoch": 3.098472880463402,
      "grad_norm": 0.3071950078010559,
      "learning_rate": 0.00013092371678308896,
      "loss": 0.20280207693576813,
      "mean_token_accuracy": 0.9109026789665222,
      "num_tokens": 72581940.0,
      "step": 5884
    },
    {
      "entropy": 1.00009123980999,
      "epoch": 3.0989994734070563,
      "grad_norm": 0.29816606640815735,
      "learning_rate": 0.0001309023887141197,
      "loss": 0.17498430609703064,
      "mean_token_accuracy": 0.9293450862169266,
      "num_tokens": 72594276.0,
      "step": 5885
    },
    {
      "entropy": 0.995394691824913,
      "epoch": 3.0995260663507107,
      "grad_norm": 0.28363287448883057,
      "learning_rate": 0.00013088105940429744,
      "loss": 0.1702607125043869,
      "mean_token_accuracy": 0.9287439733743668,
      "num_tokens": 72606612.0,
      "step": 5886
    },
    {
      "entropy": 1.014031559228897,
      "epoch": 3.1000526592943656,
      "grad_norm": 0.31453272700309753,
      "learning_rate": 0.00013085972885488843,
      "loss": 0.19794932007789612,
      "mean_token_accuracy": 0.9198706299066544,
      "num_tokens": 72618948.0,
      "step": 5887
    },
    {
      "entropy": 1.0270172506570816,
      "epoch": 3.10057925223802,
      "grad_norm": 0.31530511379241943,
      "learning_rate": 0.00013083839706715884,
      "loss": 0.19103944301605225,
      "mean_token_accuracy": 0.9153321981430054,
      "num_tokens": 72631284.0,
      "step": 5888
    },
    {
      "entropy": 1.0285425931215286,
      "epoch": 3.1011058451816744,
      "grad_norm": 0.2896331548690796,
      "learning_rate": 0.0001308170640423751,
      "loss": 0.1762722134590149,
      "mean_token_accuracy": 0.9229384660720825,
      "num_tokens": 72643620.0,
      "step": 5889
    },
    {
      "entropy": 1.0397944748401642,
      "epoch": 3.1016324381253293,
      "grad_norm": 0.2989019453525543,
      "learning_rate": 0.0001307957297818036,
      "loss": 0.18894504010677338,
      "mean_token_accuracy": 0.916764959692955,
      "num_tokens": 72655956.0,
      "step": 5890
    },
    {
      "entropy": 0.9798354804515839,
      "epoch": 3.1021590310689837,
      "grad_norm": 0.27662745118141174,
      "learning_rate": 0.00013077439428671084,
      "loss": 0.1908535659313202,
      "mean_token_accuracy": 0.9216964095830917,
      "num_tokens": 72668292.0,
      "step": 5891
    },
    {
      "entropy": 1.0175761878490448,
      "epoch": 3.102685624012638,
      "grad_norm": 0.3136708438396454,
      "learning_rate": 0.00013075305755836338,
      "loss": 0.195576474070549,
      "mean_token_accuracy": 0.9208648949861526,
      "num_tokens": 72680628.0,
      "step": 5892
    },
    {
      "entropy": 1.0088795125484467,
      "epoch": 3.103212216956293,
      "grad_norm": 0.2834429442882538,
      "learning_rate": 0.00013073171959802785,
      "loss": 0.17227596044540405,
      "mean_token_accuracy": 0.9266511499881744,
      "num_tokens": 72692964.0,
      "step": 5893
    },
    {
      "entropy": 1.0117809623479843,
      "epoch": 3.1037388098999474,
      "grad_norm": 0.3066660165786743,
      "learning_rate": 0.00013071038040697095,
      "loss": 0.18589381873607635,
      "mean_token_accuracy": 0.9194468706846237,
      "num_tokens": 72705300.0,
      "step": 5894
    },
    {
      "entropy": 1.0212982147932053,
      "epoch": 3.104265402843602,
      "grad_norm": 0.282331645488739,
      "learning_rate": 0.00013068903998645945,
      "loss": 0.18355408310890198,
      "mean_token_accuracy": 0.922243520617485,
      "num_tokens": 72717636.0,
      "step": 5895
    },
    {
      "entropy": 1.0034818798303604,
      "epoch": 3.1047919957872563,
      "grad_norm": 0.2822534739971161,
      "learning_rate": 0.00013066769833776026,
      "loss": 0.18468129634857178,
      "mean_token_accuracy": 0.9255609661340714,
      "num_tokens": 72729972.0,
      "step": 5896
    },
    {
      "entropy": 0.9707738161087036,
      "epoch": 3.105318588730911,
      "grad_norm": 0.28754761815071106,
      "learning_rate": 0.00013064635546214027,
      "loss": 0.18167972564697266,
      "mean_token_accuracy": 0.9241908490657806,
      "num_tokens": 72742308.0,
      "step": 5897
    },
    {
      "entropy": 0.9499937295913696,
      "epoch": 3.1058451816745656,
      "grad_norm": 0.28328055143356323,
      "learning_rate": 0.0001306250113608665,
      "loss": 0.18320026993751526,
      "mean_token_accuracy": 0.9243544191122055,
      "num_tokens": 72754644.0,
      "step": 5898
    },
    {
      "entropy": 1.0261553823947906,
      "epoch": 3.10637177461822,
      "grad_norm": 0.29419898986816406,
      "learning_rate": 0.00013060366603520601,
      "loss": 0.1817263513803482,
      "mean_token_accuracy": 0.9239636808633804,
      "num_tokens": 72766980.0,
      "step": 5899
    },
    {
      "entropy": 0.9795711040496826,
      "epoch": 3.106898367561875,
      "grad_norm": 0.2906143367290497,
      "learning_rate": 0.00013058231948642597,
      "loss": 0.17092804610729218,
      "mean_token_accuracy": 0.9313593059778214,
      "num_tokens": 72779316.0,
      "step": 5900
    },
    {
      "entropy": 0.9814235419034958,
      "epoch": 3.1074249605055293,
      "grad_norm": 0.28841039538383484,
      "learning_rate": 0.00013056097171579357,
      "loss": 0.17476145923137665,
      "mean_token_accuracy": 0.9297506809234619,
      "num_tokens": 72791652.0,
      "step": 5901
    },
    {
      "entropy": 0.9798099398612976,
      "epoch": 3.1079515534491837,
      "grad_norm": 0.2775534987449646,
      "learning_rate": 0.00013053962272457613,
      "loss": 0.1680983006954193,
      "mean_token_accuracy": 0.9283261299133301,
      "num_tokens": 72803988.0,
      "step": 5902
    },
    {
      "entropy": 1.0072557479143143,
      "epoch": 3.108478146392838,
      "grad_norm": 0.2890489399433136,
      "learning_rate": 0.000130518272514041,
      "loss": 0.17888548970222473,
      "mean_token_accuracy": 0.9257342368364334,
      "num_tokens": 72816324.0,
      "step": 5903
    },
    {
      "entropy": 0.9900611490011215,
      "epoch": 3.109004739336493,
      "grad_norm": 0.3024149239063263,
      "learning_rate": 0.00013049692108545562,
      "loss": 0.18220660090446472,
      "mean_token_accuracy": 0.9204623997211456,
      "num_tokens": 72828660.0,
      "step": 5904
    },
    {
      "entropy": 1.0078524500131607,
      "epoch": 3.1095313322801474,
      "grad_norm": 0.31779125332832336,
      "learning_rate": 0.00013047556844008747,
      "loss": 0.194386824965477,
      "mean_token_accuracy": 0.9209152162075043,
      "num_tokens": 72840996.0,
      "step": 5905
    },
    {
      "entropy": 0.9710298627614975,
      "epoch": 3.110057925223802,
      "grad_norm": 0.2938542664051056,
      "learning_rate": 0.00013045421457920416,
      "loss": 0.17387202382087708,
      "mean_token_accuracy": 0.9304901510477066,
      "num_tokens": 72853332.0,
      "step": 5906
    },
    {
      "entropy": 1.0137410461902618,
      "epoch": 3.1105845181674567,
      "grad_norm": 0.31136083602905273,
      "learning_rate": 0.00013043285950407342,
      "loss": 0.18210887908935547,
      "mean_token_accuracy": 0.9235569536685944,
      "num_tokens": 72865668.0,
      "step": 5907
    },
    {
      "entropy": 0.9845224022865295,
      "epoch": 3.111111111111111,
      "grad_norm": 0.2719968855381012,
      "learning_rate": 0.00013041150321596286,
      "loss": 0.16895705461502075,
      "mean_token_accuracy": 0.9294350296258926,
      "num_tokens": 72878004.0,
      "step": 5908
    },
    {
      "entropy": 0.9864397495985031,
      "epoch": 3.1116377040547656,
      "grad_norm": 0.301286518573761,
      "learning_rate": 0.00013039014571614028,
      "loss": 0.18717461824417114,
      "mean_token_accuracy": 0.9226465821266174,
      "num_tokens": 72890340.0,
      "step": 5909
    },
    {
      "entropy": 0.9790801554918289,
      "epoch": 3.11216429699842,
      "grad_norm": 0.2874787449836731,
      "learning_rate": 0.00013036878700587365,
      "loss": 0.18862608075141907,
      "mean_token_accuracy": 0.9194082766771317,
      "num_tokens": 72902676.0,
      "step": 5910
    },
    {
      "entropy": 1.0117427706718445,
      "epoch": 3.112690889942075,
      "grad_norm": 0.26695552468299866,
      "learning_rate": 0.00013034742708643084,
      "loss": 0.173864483833313,
      "mean_token_accuracy": 0.9279522448778152,
      "num_tokens": 72915012.0,
      "step": 5911
    },
    {
      "entropy": 0.942085400223732,
      "epoch": 3.1132174828857293,
      "grad_norm": 0.2916507124900818,
      "learning_rate": 0.00013032606595907989,
      "loss": 0.17523030936717987,
      "mean_token_accuracy": 0.9323842823505402,
      "num_tokens": 72927348.0,
      "step": 5912
    },
    {
      "entropy": 1.0327578485012054,
      "epoch": 3.1137440758293837,
      "grad_norm": 0.30272775888442993,
      "learning_rate": 0.00013030470362508884,
      "loss": 0.1859118491411209,
      "mean_token_accuracy": 0.9215680062770844,
      "num_tokens": 72939684.0,
      "step": 5913
    },
    {
      "entropy": 0.9773134887218475,
      "epoch": 3.1142706687730386,
      "grad_norm": 0.2893243730068207,
      "learning_rate": 0.00013028334008572588,
      "loss": 0.17654062807559967,
      "mean_token_accuracy": 0.9239582568407059,
      "num_tokens": 72952020.0,
      "step": 5914
    },
    {
      "entropy": 0.9801084250211716,
      "epoch": 3.114797261716693,
      "grad_norm": 0.2631586194038391,
      "learning_rate": 0.00013026197534225925,
      "loss": 0.15703842043876648,
      "mean_token_accuracy": 0.9351452887058258,
      "num_tokens": 72964356.0,
      "step": 5915
    },
    {
      "entropy": 1.0578633546829224,
      "epoch": 3.1153238546603474,
      "grad_norm": 0.2890935242176056,
      "learning_rate": 0.00013024060939595723,
      "loss": 0.1817035675048828,
      "mean_token_accuracy": 0.9252161383628845,
      "num_tokens": 72976692.0,
      "step": 5916
    },
    {
      "entropy": 1.0292495489120483,
      "epoch": 3.1158504476040023,
      "grad_norm": 0.288146436214447,
      "learning_rate": 0.00013021924224808823,
      "loss": 0.19315966963768005,
      "mean_token_accuracy": 0.9209520369768143,
      "num_tokens": 72989028.0,
      "step": 5917
    },
    {
      "entropy": 0.9683614373207092,
      "epoch": 3.1163770405476567,
      "grad_norm": 0.28280383348464966,
      "learning_rate": 0.00013019787389992066,
      "loss": 0.18978121876716614,
      "mean_token_accuracy": 0.9187489449977875,
      "num_tokens": 73001364.0,
      "step": 5918
    },
    {
      "entropy": 0.9939278662204742,
      "epoch": 3.116903633491311,
      "grad_norm": 0.2836332321166992,
      "learning_rate": 0.00013017650435272305,
      "loss": 0.17158405482769012,
      "mean_token_accuracy": 0.9307135939598083,
      "num_tokens": 73013700.0,
      "step": 5919
    },
    {
      "entropy": 1.040738195180893,
      "epoch": 3.1174302264349656,
      "grad_norm": 0.2956024408340454,
      "learning_rate": 0.00013015513360776392,
      "loss": 0.1901095062494278,
      "mean_token_accuracy": 0.9226278364658356,
      "num_tokens": 73026036.0,
      "step": 5920
    },
    {
      "entropy": 1.0458858758211136,
      "epoch": 3.1179568193786205,
      "grad_norm": 0.336071640253067,
      "learning_rate": 0.00013013376166631205,
      "loss": 0.20758454501628876,
      "mean_token_accuracy": 0.9135317951440811,
      "num_tokens": 73038372.0,
      "step": 5921
    },
    {
      "entropy": 1.0126291811466217,
      "epoch": 3.118483412322275,
      "grad_norm": 0.29755648970603943,
      "learning_rate": 0.00013011238852963605,
      "loss": 0.19317781925201416,
      "mean_token_accuracy": 0.9182720631361008,
      "num_tokens": 73050708.0,
      "step": 5922
    },
    {
      "entropy": 0.9963730573654175,
      "epoch": 3.1190100052659293,
      "grad_norm": 0.2682756781578064,
      "learning_rate": 0.00013009101419900475,
      "loss": 0.17621959745883942,
      "mean_token_accuracy": 0.925027385354042,
      "num_tokens": 73063044.0,
      "step": 5923
    },
    {
      "entropy": 1.0447964668273926,
      "epoch": 3.119536598209584,
      "grad_norm": 0.3019618093967438,
      "learning_rate": 0.0001300696386756871,
      "loss": 0.1902092695236206,
      "mean_token_accuracy": 0.9218323826789856,
      "num_tokens": 73075380.0,
      "step": 5924
    },
    {
      "entropy": 0.9710809290409088,
      "epoch": 3.1200631911532386,
      "grad_norm": 0.2597827613353729,
      "learning_rate": 0.0001300482619609519,
      "loss": 0.17227184772491455,
      "mean_token_accuracy": 0.92453433573246,
      "num_tokens": 73087716.0,
      "step": 5925
    },
    {
      "entropy": 1.0418781042099,
      "epoch": 3.120589784096893,
      "grad_norm": 0.29908981919288635,
      "learning_rate": 0.00013002688405606828,
      "loss": 0.17560279369354248,
      "mean_token_accuracy": 0.9268177151679993,
      "num_tokens": 73100052.0,
      "step": 5926
    },
    {
      "entropy": 0.9824186861515045,
      "epoch": 3.1211163770405475,
      "grad_norm": 0.303090363740921,
      "learning_rate": 0.0001300055049623053,
      "loss": 0.1912716180086136,
      "mean_token_accuracy": 0.9202925115823746,
      "num_tokens": 73112388.0,
      "step": 5927
    },
    {
      "entropy": 0.9852057695388794,
      "epoch": 3.1216429699842023,
      "grad_norm": 0.2938767671585083,
      "learning_rate": 0.0001299841246809321,
      "loss": 0.17864194512367249,
      "mean_token_accuracy": 0.9232177287340164,
      "num_tokens": 73124724.0,
      "step": 5928
    },
    {
      "entropy": 1.02889284491539,
      "epoch": 3.1221695629278567,
      "grad_norm": 0.305049866437912,
      "learning_rate": 0.00012996274321321787,
      "loss": 0.18025071918964386,
      "mean_token_accuracy": 0.9296489357948303,
      "num_tokens": 73137060.0,
      "step": 5929
    },
    {
      "entropy": 0.9720607846975327,
      "epoch": 3.122696155871511,
      "grad_norm": 0.28418251872062683,
      "learning_rate": 0.00012994136056043186,
      "loss": 0.18828493356704712,
      "mean_token_accuracy": 0.9190543591976166,
      "num_tokens": 73149396.0,
      "step": 5930
    },
    {
      "entropy": 0.9504470825195312,
      "epoch": 3.123222748815166,
      "grad_norm": 0.28166502714157104,
      "learning_rate": 0.00012991997672384358,
      "loss": 0.18314103782176971,
      "mean_token_accuracy": 0.9210266470909119,
      "num_tokens": 73161732.0,
      "step": 5931
    },
    {
      "entropy": 1.0184333622455597,
      "epoch": 3.1237493417588205,
      "grad_norm": 0.31155115365982056,
      "learning_rate": 0.0001298985917047224,
      "loss": 0.1838856041431427,
      "mean_token_accuracy": 0.9216693192720413,
      "num_tokens": 73174068.0,
      "step": 5932
    },
    {
      "entropy": 0.9916276931762695,
      "epoch": 3.124275934702475,
      "grad_norm": 0.3261113166809082,
      "learning_rate": 0.00012987720550433775,
      "loss": 0.17985975742340088,
      "mean_token_accuracy": 0.9207182675600052,
      "num_tokens": 73186404.0,
      "step": 5933
    },
    {
      "entropy": 0.9735004603862762,
      "epoch": 3.1248025276461293,
      "grad_norm": 0.2895289361476898,
      "learning_rate": 0.0001298558181239593,
      "loss": 0.17738863825798035,
      "mean_token_accuracy": 0.92310531437397,
      "num_tokens": 73198740.0,
      "step": 5934
    },
    {
      "entropy": 0.9904636740684509,
      "epoch": 3.125329120589784,
      "grad_norm": 0.26589369773864746,
      "learning_rate": 0.00012983442956485662,
      "loss": 0.1570383608341217,
      "mean_token_accuracy": 0.9323722422122955,
      "num_tokens": 73211076.0,
      "step": 5935
    },
    {
      "entropy": 1.0544918030500412,
      "epoch": 3.1258557135334386,
      "grad_norm": 0.3020542860031128,
      "learning_rate": 0.00012981303982829948,
      "loss": 0.18856723606586456,
      "mean_token_accuracy": 0.9201971888542175,
      "num_tokens": 73223412.0,
      "step": 5936
    },
    {
      "entropy": 0.9709584712982178,
      "epoch": 3.126382306477093,
      "grad_norm": 0.2670464515686035,
      "learning_rate": 0.00012979164891555766,
      "loss": 0.167905792593956,
      "mean_token_accuracy": 0.9310225397348404,
      "num_tokens": 73235748.0,
      "step": 5937
    },
    {
      "entropy": 1.026566207408905,
      "epoch": 3.126908899420748,
      "grad_norm": 0.29108577966690063,
      "learning_rate": 0.000129770256827901,
      "loss": 0.17460668087005615,
      "mean_token_accuracy": 0.9300505667924881,
      "num_tokens": 73248084.0,
      "step": 5938
    },
    {
      "entropy": 1.0133537948131561,
      "epoch": 3.1274354923644023,
      "grad_norm": 0.3254956305027008,
      "learning_rate": 0.0001297488635665994,
      "loss": 0.2051609456539154,
      "mean_token_accuracy": 0.916284367442131,
      "num_tokens": 73260420.0,
      "step": 5939
    },
    {
      "entropy": 1.0277920812368393,
      "epoch": 3.1279620853080567,
      "grad_norm": 0.2922893166542053,
      "learning_rate": 0.00012972746913292288,
      "loss": 0.17920175194740295,
      "mean_token_accuracy": 0.9276208430528641,
      "num_tokens": 73272756.0,
      "step": 5940
    },
    {
      "entropy": 0.9691586345434189,
      "epoch": 3.1284886782517116,
      "grad_norm": 0.2643762230873108,
      "learning_rate": 0.0001297060735281415,
      "loss": 0.17805035412311554,
      "mean_token_accuracy": 0.9242268353700638,
      "num_tokens": 73285092.0,
      "step": 5941
    },
    {
      "entropy": 1.0250355750322342,
      "epoch": 3.129015271195366,
      "grad_norm": 0.28367775678634644,
      "learning_rate": 0.0001296846767535254,
      "loss": 0.18581588566303253,
      "mean_token_accuracy": 0.9214816987514496,
      "num_tokens": 73297428.0,
      "step": 5942
    },
    {
      "entropy": 1.0472411066293716,
      "epoch": 3.1295418641390205,
      "grad_norm": 0.2918038070201874,
      "learning_rate": 0.00012966327881034476,
      "loss": 0.18668335676193237,
      "mean_token_accuracy": 0.9221546500921249,
      "num_tokens": 73309764.0,
      "step": 5943
    },
    {
      "entropy": 1.0437220931053162,
      "epoch": 3.130068457082675,
      "grad_norm": 0.286050021648407,
      "learning_rate": 0.00012964187969986986,
      "loss": 0.17461931705474854,
      "mean_token_accuracy": 0.9245903939008713,
      "num_tokens": 73322100.0,
      "step": 5944
    },
    {
      "entropy": 1.0097281336784363,
      "epoch": 3.1305950500263298,
      "grad_norm": 0.2658098340034485,
      "learning_rate": 0.00012962047942337106,
      "loss": 0.16860619187355042,
      "mean_token_accuracy": 0.9307616651058197,
      "num_tokens": 73334436.0,
      "step": 5945
    },
    {
      "entropy": 0.9922635406255722,
      "epoch": 3.131121642969984,
      "grad_norm": 0.27987632155418396,
      "learning_rate": 0.00012959907798211872,
      "loss": 0.1775607317686081,
      "mean_token_accuracy": 0.9260025918483734,
      "num_tokens": 73346772.0,
      "step": 5946
    },
    {
      "entropy": 0.9956186562776566,
      "epoch": 3.1316482359136386,
      "grad_norm": 0.27126094698905945,
      "learning_rate": 0.0001295776753773834,
      "loss": 0.16783833503723145,
      "mean_token_accuracy": 0.9281550049781799,
      "num_tokens": 73359108.0,
      "step": 5947
    },
    {
      "entropy": 1.0380958318710327,
      "epoch": 3.1321748288572935,
      "grad_norm": 0.2689456343650818,
      "learning_rate": 0.00012955627161043558,
      "loss": 0.1678178310394287,
      "mean_token_accuracy": 0.9268752187490463,
      "num_tokens": 73371444.0,
      "step": 5948
    },
    {
      "entropy": 1.0214119404554367,
      "epoch": 3.132701421800948,
      "grad_norm": 0.281562477350235,
      "learning_rate": 0.00012953486668254588,
      "loss": 0.17973113059997559,
      "mean_token_accuracy": 0.9210423231124878,
      "num_tokens": 73383780.0,
      "step": 5949
    },
    {
      "entropy": 1.011966422200203,
      "epoch": 3.1332280147446023,
      "grad_norm": 0.28674042224884033,
      "learning_rate": 0.00012951346059498505,
      "loss": 0.18101146817207336,
      "mean_token_accuracy": 0.9260558187961578,
      "num_tokens": 73396116.0,
      "step": 5950
    },
    {
      "entropy": 1.0085774511098862,
      "epoch": 3.1337546076882568,
      "grad_norm": 0.27952900528907776,
      "learning_rate": 0.00012949205334902375,
      "loss": 0.18016311526298523,
      "mean_token_accuracy": 0.9254178255796432,
      "num_tokens": 73408452.0,
      "step": 5951
    },
    {
      "entropy": 1.0077911466360092,
      "epoch": 3.1342812006319116,
      "grad_norm": 0.2722882926464081,
      "learning_rate": 0.00012947064494593286,
      "loss": 0.18341054022312164,
      "mean_token_accuracy": 0.9214029014110565,
      "num_tokens": 73420788.0,
      "step": 5952
    },
    {
      "entropy": 1.038177341222763,
      "epoch": 3.134807793575566,
      "grad_norm": 0.2963654100894928,
      "learning_rate": 0.00012944923538698325,
      "loss": 0.18527300655841827,
      "mean_token_accuracy": 0.9238464385271072,
      "num_tokens": 73433124.0,
      "step": 5953
    },
    {
      "entropy": 1.0265491753816605,
      "epoch": 3.1353343865192205,
      "grad_norm": 0.27687832713127136,
      "learning_rate": 0.00012942782467344593,
      "loss": 0.1755915731191635,
      "mean_token_accuracy": 0.9277375042438507,
      "num_tokens": 73445460.0,
      "step": 5954
    },
    {
      "entropy": 1.0480819642543793,
      "epoch": 3.1358609794628753,
      "grad_norm": 0.2859427332878113,
      "learning_rate": 0.00012940641280659188,
      "loss": 0.18823222815990448,
      "mean_token_accuracy": 0.9216801226139069,
      "num_tokens": 73457796.0,
      "step": 5955
    },
    {
      "entropy": 1.0693493485450745,
      "epoch": 3.1363875724065298,
      "grad_norm": 0.3324831426143646,
      "learning_rate": 0.00012938499978769222,
      "loss": 0.1855066418647766,
      "mean_token_accuracy": 0.9238448441028595,
      "num_tokens": 73470132.0,
      "step": 5956
    },
    {
      "entropy": 1.0338356047868729,
      "epoch": 3.136914165350184,
      "grad_norm": 0.26955410838127136,
      "learning_rate": 0.00012936358561801807,
      "loss": 0.1669587939977646,
      "mean_token_accuracy": 0.9290878027677536,
      "num_tokens": 73482468.0,
      "step": 5957
    },
    {
      "entropy": 1.0918577015399933,
      "epoch": 3.137440758293839,
      "grad_norm": 0.3077058792114258,
      "learning_rate": 0.0001293421702988407,
      "loss": 0.17801736295223236,
      "mean_token_accuracy": 0.9242205321788788,
      "num_tokens": 73494804.0,
      "step": 5958
    },
    {
      "entropy": 1.0301255732774734,
      "epoch": 3.1379673512374935,
      "grad_norm": 0.287098228931427,
      "learning_rate": 0.00012932075383143148,
      "loss": 0.17325901985168457,
      "mean_token_accuracy": 0.9261231869459152,
      "num_tokens": 73507140.0,
      "step": 5959
    },
    {
      "entropy": 1.0039707273244858,
      "epoch": 3.138493944181148,
      "grad_norm": 0.2861899733543396,
      "learning_rate": 0.00012929933621706166,
      "loss": 0.17903730273246765,
      "mean_token_accuracy": 0.9243130832910538,
      "num_tokens": 73519476.0,
      "step": 5960
    },
    {
      "entropy": 1.0366846919059753,
      "epoch": 3.1390205371248023,
      "grad_norm": 0.2873387336730957,
      "learning_rate": 0.00012927791745700277,
      "loss": 0.1751435399055481,
      "mean_token_accuracy": 0.9282572865486145,
      "num_tokens": 73531812.0,
      "step": 5961
    },
    {
      "entropy": 1.093095302581787,
      "epoch": 3.139547130068457,
      "grad_norm": 0.2945985198020935,
      "learning_rate": 0.00012925649755252624,
      "loss": 0.19299519062042236,
      "mean_token_accuracy": 0.9205166846513748,
      "num_tokens": 73544148.0,
      "step": 5962
    },
    {
      "entropy": 1.0331242233514786,
      "epoch": 3.1400737230121116,
      "grad_norm": 0.2650090456008911,
      "learning_rate": 0.0001292350765049037,
      "loss": 0.18281593918800354,
      "mean_token_accuracy": 0.9210380017757416,
      "num_tokens": 73556484.0,
      "step": 5963
    },
    {
      "entropy": 1.0869034230709076,
      "epoch": 3.140600315955766,
      "grad_norm": 0.31165093183517456,
      "learning_rate": 0.0001292136543154068,
      "loss": 0.18366098403930664,
      "mean_token_accuracy": 0.9227862507104874,
      "num_tokens": 73568820.0,
      "step": 5964
    },
    {
      "entropy": 1.008332222700119,
      "epoch": 3.141126908899421,
      "grad_norm": 0.27395057678222656,
      "learning_rate": 0.0001291922309853072,
      "loss": 0.16771671175956726,
      "mean_token_accuracy": 0.9288883656263351,
      "num_tokens": 73581156.0,
      "step": 5965
    },
    {
      "entropy": 1.049354761838913,
      "epoch": 3.1416535018430753,
      "grad_norm": 0.296152800321579,
      "learning_rate": 0.00012917080651587672,
      "loss": 0.17943327128887177,
      "mean_token_accuracy": 0.9249696284532547,
      "num_tokens": 73593492.0,
      "step": 5966
    },
    {
      "entropy": 1.0088461339473724,
      "epoch": 3.1421800947867298,
      "grad_norm": 0.2626063823699951,
      "learning_rate": 0.0001291493809083872,
      "loss": 0.1711309850215912,
      "mean_token_accuracy": 0.9292300641536713,
      "num_tokens": 73605828.0,
      "step": 5967
    },
    {
      "entropy": 1.051283359527588,
      "epoch": 3.142706687730384,
      "grad_norm": 0.28641244769096375,
      "learning_rate": 0.00012912795416411056,
      "loss": 0.18154369294643402,
      "mean_token_accuracy": 0.9274348318576813,
      "num_tokens": 73618164.0,
      "step": 5968
    },
    {
      "entropy": 1.0577448010444641,
      "epoch": 3.143233280674039,
      "grad_norm": 0.2720896005630493,
      "learning_rate": 0.0001291065262843187,
      "loss": 0.1672617644071579,
      "mean_token_accuracy": 0.9289580285549164,
      "num_tokens": 73630500.0,
      "step": 5969
    },
    {
      "entropy": 1.0459100604057312,
      "epoch": 3.1437598736176935,
      "grad_norm": 0.30032745003700256,
      "learning_rate": 0.00012908509727028376,
      "loss": 0.17962820827960968,
      "mean_token_accuracy": 0.9249450415372849,
      "num_tokens": 73642836.0,
      "step": 5970
    },
    {
      "entropy": 1.054712176322937,
      "epoch": 3.144286466561348,
      "grad_norm": 0.31555280089378357,
      "learning_rate": 0.00012906366712327788,
      "loss": 0.18882232904434204,
      "mean_token_accuracy": 0.9210335910320282,
      "num_tokens": 73655172.0,
      "step": 5971
    },
    {
      "entropy": 1.0289185047149658,
      "epoch": 3.144813059505003,
      "grad_norm": 0.3134496510028839,
      "learning_rate": 0.00012904223584457315,
      "loss": 0.20182853937149048,
      "mean_token_accuracy": 0.9166600257158279,
      "num_tokens": 73667508.0,
      "step": 5972
    },
    {
      "entropy": 1.0345142036676407,
      "epoch": 3.145339652448657,
      "grad_norm": 0.27865415811538696,
      "learning_rate": 0.00012902080343544188,
      "loss": 0.1802331507205963,
      "mean_token_accuracy": 0.9225296080112457,
      "num_tokens": 73679844.0,
      "step": 5973
    },
    {
      "entropy": 1.027937650680542,
      "epoch": 3.1458662453923116,
      "grad_norm": 0.2987896800041199,
      "learning_rate": 0.0001289993698971564,
      "loss": 0.17813797295093536,
      "mean_token_accuracy": 0.9263286739587784,
      "num_tokens": 73692180.0,
      "step": 5974
    },
    {
      "entropy": 1.0327493250370026,
      "epoch": 3.1463928383359665,
      "grad_norm": 0.2850527763366699,
      "learning_rate": 0.00012897793523098904,
      "loss": 0.176524817943573,
      "mean_token_accuracy": 0.923951119184494,
      "num_tokens": 73704516.0,
      "step": 5975
    },
    {
      "entropy": 0.9963532090187073,
      "epoch": 3.146919431279621,
      "grad_norm": 0.2879466414451599,
      "learning_rate": 0.00012895649943821228,
      "loss": 0.18072587251663208,
      "mean_token_accuracy": 0.9252651482820511,
      "num_tokens": 73716852.0,
      "step": 5976
    },
    {
      "entropy": 0.9650211036205292,
      "epoch": 3.1474460242232754,
      "grad_norm": 0.2874644696712494,
      "learning_rate": 0.00012893506252009863,
      "loss": 0.1855263113975525,
      "mean_token_accuracy": 0.9205344468355179,
      "num_tokens": 73729188.0,
      "step": 5977
    },
    {
      "entropy": 0.9902265071868896,
      "epoch": 3.1479726171669298,
      "grad_norm": 0.30048924684524536,
      "learning_rate": 0.00012891362447792069,
      "loss": 0.19820508360862732,
      "mean_token_accuracy": 0.9142971634864807,
      "num_tokens": 73741524.0,
      "step": 5978
    },
    {
      "entropy": 0.9674103707075119,
      "epoch": 3.1484992101105846,
      "grad_norm": 0.294927716255188,
      "learning_rate": 0.00012889218531295107,
      "loss": 0.1868659257888794,
      "mean_token_accuracy": 0.9226386100053787,
      "num_tokens": 73753860.0,
      "step": 5979
    },
    {
      "entropy": 0.9716824889183044,
      "epoch": 3.149025803054239,
      "grad_norm": 0.2963883578777313,
      "learning_rate": 0.00012887074502646257,
      "loss": 0.1829419881105423,
      "mean_token_accuracy": 0.9211580753326416,
      "num_tokens": 73766196.0,
      "step": 5980
    },
    {
      "entropy": 0.9973532259464264,
      "epoch": 3.1495523959978935,
      "grad_norm": 0.2655693292617798,
      "learning_rate": 0.0001288493036197279,
      "loss": 0.16832715272903442,
      "mean_token_accuracy": 0.9321129769086838,
      "num_tokens": 73778532.0,
      "step": 5981
    },
    {
      "entropy": 0.9477798938751221,
      "epoch": 3.1500789889415484,
      "grad_norm": 0.28809699416160583,
      "learning_rate": 0.00012882786109401993,
      "loss": 0.17731884121894836,
      "mean_token_accuracy": 0.9298195987939835,
      "num_tokens": 73790868.0,
      "step": 5982
    },
    {
      "entropy": 0.9419539272785187,
      "epoch": 3.150605581885203,
      "grad_norm": 0.25741738080978394,
      "learning_rate": 0.0001288064174506116,
      "loss": 0.17238347232341766,
      "mean_token_accuracy": 0.9305710196495056,
      "num_tokens": 73803204.0,
      "step": 5983
    },
    {
      "entropy": 0.9464537054300308,
      "epoch": 3.151132174828857,
      "grad_norm": 0.30378761887550354,
      "learning_rate": 0.00012878497269077586,
      "loss": 0.18827325105667114,
      "mean_token_accuracy": 0.9179285317659378,
      "num_tokens": 73815540.0,
      "step": 5984
    },
    {
      "entropy": 0.9460196644067764,
      "epoch": 3.1516587677725116,
      "grad_norm": 0.286990225315094,
      "learning_rate": 0.0001287635268157858,
      "loss": 0.1831638216972351,
      "mean_token_accuracy": 0.920984223484993,
      "num_tokens": 73827876.0,
      "step": 5985
    },
    {
      "entropy": 0.9726483821868896,
      "epoch": 3.1521853607161665,
      "grad_norm": 0.30301207304000854,
      "learning_rate": 0.00012874207982691447,
      "loss": 0.18871383368968964,
      "mean_token_accuracy": 0.922227144241333,
      "num_tokens": 73840212.0,
      "step": 5986
    },
    {
      "entropy": 0.9950109422206879,
      "epoch": 3.152711953659821,
      "grad_norm": 0.31202518939971924,
      "learning_rate": 0.00012872063172543512,
      "loss": 0.19398647546768188,
      "mean_token_accuracy": 0.9219925999641418,
      "num_tokens": 73852548.0,
      "step": 5987
    },
    {
      "entropy": 0.9679990857839584,
      "epoch": 3.1532385466034754,
      "grad_norm": 0.2762276828289032,
      "learning_rate": 0.000128699182512621,
      "loss": 0.16103002429008484,
      "mean_token_accuracy": 0.9330836683511734,
      "num_tokens": 73864884.0,
      "step": 5988
    },
    {
      "entropy": 0.9781599789857864,
      "epoch": 3.1537651395471302,
      "grad_norm": 0.3027253746986389,
      "learning_rate": 0.00012867773218974539,
      "loss": 0.20093132555484772,
      "mean_token_accuracy": 0.9174567013978958,
      "num_tokens": 73877220.0,
      "step": 5989
    },
    {
      "entropy": 0.926795020699501,
      "epoch": 3.1542917324907846,
      "grad_norm": 0.28476205468177795,
      "learning_rate": 0.00012865628075808168,
      "loss": 0.1770479679107666,
      "mean_token_accuracy": 0.9274154454469681,
      "num_tokens": 73889556.0,
      "step": 5990
    },
    {
      "entropy": 0.9234910011291504,
      "epoch": 3.154818325434439,
      "grad_norm": 0.27796196937561035,
      "learning_rate": 0.00012863482821890332,
      "loss": 0.17518147826194763,
      "mean_token_accuracy": 0.9281267076730728,
      "num_tokens": 73901892.0,
      "step": 5991
    },
    {
      "entropy": 0.9753734618425369,
      "epoch": 3.155344918378094,
      "grad_norm": 0.28891271352767944,
      "learning_rate": 0.00012861337457348383,
      "loss": 0.17538145184516907,
      "mean_token_accuracy": 0.9244814366102219,
      "num_tokens": 73914228.0,
      "step": 5992
    },
    {
      "entropy": 0.9478357136249542,
      "epoch": 3.1558715113217484,
      "grad_norm": 0.28454625606536865,
      "learning_rate": 0.00012859191982309673,
      "loss": 0.178102508187294,
      "mean_token_accuracy": 0.9241252690553665,
      "num_tokens": 73926564.0,
      "step": 5993
    },
    {
      "entropy": 0.9561555236577988,
      "epoch": 3.156398104265403,
      "grad_norm": 0.27385735511779785,
      "learning_rate": 0.00012857046396901577,
      "loss": 0.17082807421684265,
      "mean_token_accuracy": 0.9306003004312515,
      "num_tokens": 73938900.0,
      "step": 5994
    },
    {
      "entropy": 0.9790538102388382,
      "epoch": 3.156924697209057,
      "grad_norm": 0.3030516803264618,
      "learning_rate": 0.0001285490070125146,
      "loss": 0.19459722936153412,
      "mean_token_accuracy": 0.9199800938367844,
      "num_tokens": 73951236.0,
      "step": 5995
    },
    {
      "entropy": 0.9638478457927704,
      "epoch": 3.157451290152712,
      "grad_norm": 0.3094443082809448,
      "learning_rate": 0.00012852754895486697,
      "loss": 0.18378609418869019,
      "mean_token_accuracy": 0.9208549112081528,
      "num_tokens": 73963572.0,
      "step": 5996
    },
    {
      "entropy": 0.9615375697612762,
      "epoch": 3.1579778830963665,
      "grad_norm": 0.30250605940818787,
      "learning_rate": 0.00012850608979734672,
      "loss": 0.18148188292980194,
      "mean_token_accuracy": 0.9263082146644592,
      "num_tokens": 73975908.0,
      "step": 5997
    },
    {
      "entropy": 1.0005483031272888,
      "epoch": 3.158504476040021,
      "grad_norm": 0.3065294623374939,
      "learning_rate": 0.0001284846295412278,
      "loss": 0.1826051026582718,
      "mean_token_accuracy": 0.9210184514522552,
      "num_tokens": 73988244.0,
      "step": 5998
    },
    {
      "entropy": 0.9474383294582367,
      "epoch": 3.159031068983676,
      "grad_norm": 0.27827489376068115,
      "learning_rate": 0.00012846316818778417,
      "loss": 0.17282012104988098,
      "mean_token_accuracy": 0.923882320523262,
      "num_tokens": 74000580.0,
      "step": 5999
    },
    {
      "entropy": 0.9649649113416672,
      "epoch": 3.1595576619273302,
      "grad_norm": 0.2762611508369446,
      "learning_rate": 0.00012844170573828988,
      "loss": 0.17711496353149414,
      "mean_token_accuracy": 0.9302596747875214,
      "num_tokens": 74012916.0,
      "step": 6000
    },
    {
      "entropy": 0.9723414182662964,
      "epoch": 3.1600842548709847,
      "grad_norm": 0.2944243550300598,
      "learning_rate": 0.00012842024219401898,
      "loss": 0.16719335317611694,
      "mean_token_accuracy": 0.9334683865308762,
      "num_tokens": 74025252.0,
      "step": 6001
    },
    {
      "entropy": 0.9083682596683502,
      "epoch": 3.160610847814639,
      "grad_norm": 0.275662362575531,
      "learning_rate": 0.00012839877755624565,
      "loss": 0.1716337651014328,
      "mean_token_accuracy": 0.9241354614496231,
      "num_tokens": 74037588.0,
      "step": 6002
    },
    {
      "entropy": 0.9236375540494919,
      "epoch": 3.161137440758294,
      "grad_norm": 0.2888052463531494,
      "learning_rate": 0.00012837731182624415,
      "loss": 0.17050597071647644,
      "mean_token_accuracy": 0.9272229671478271,
      "num_tokens": 74049924.0,
      "step": 6003
    },
    {
      "entropy": 0.954410120844841,
      "epoch": 3.1616640337019484,
      "grad_norm": 0.2948882281780243,
      "learning_rate": 0.00012835584500528875,
      "loss": 0.18716907501220703,
      "mean_token_accuracy": 0.9205042868852615,
      "num_tokens": 74062260.0,
      "step": 6004
    },
    {
      "entropy": 0.9476812034845352,
      "epoch": 3.162190626645603,
      "grad_norm": 0.2547721862792969,
      "learning_rate": 0.00012833437709465386,
      "loss": 0.1682538241147995,
      "mean_token_accuracy": 0.9294954538345337,
      "num_tokens": 74074596.0,
      "step": 6005
    },
    {
      "entropy": 0.9671272784471512,
      "epoch": 3.1627172195892577,
      "grad_norm": 0.2941117286682129,
      "learning_rate": 0.00012831290809561382,
      "loss": 0.1855352371931076,
      "mean_token_accuracy": 0.9234724044799805,
      "num_tokens": 74086932.0,
      "step": 6006
    },
    {
      "entropy": 0.9462792277336121,
      "epoch": 3.163243812532912,
      "grad_norm": 0.2800365686416626,
      "learning_rate": 0.00012829143800944314,
      "loss": 0.1678551584482193,
      "mean_token_accuracy": 0.9280218780040741,
      "num_tokens": 74099268.0,
      "step": 6007
    },
    {
      "entropy": 0.9736743271350861,
      "epoch": 3.1637704054765665,
      "grad_norm": 0.29358065128326416,
      "learning_rate": 0.00012826996683741644,
      "loss": 0.17688816785812378,
      "mean_token_accuracy": 0.9257215261459351,
      "num_tokens": 74111604.0,
      "step": 6008
    },
    {
      "entropy": 0.9866583496332169,
      "epoch": 3.1642969984202214,
      "grad_norm": 0.3088134229183197,
      "learning_rate": 0.00012824849458080827,
      "loss": 0.19596421718597412,
      "mean_token_accuracy": 0.916352167725563,
      "num_tokens": 74123940.0,
      "step": 6009
    },
    {
      "entropy": 0.9428907185792923,
      "epoch": 3.164823591363876,
      "grad_norm": 0.304568350315094,
      "learning_rate": 0.00012822702124089337,
      "loss": 0.1897311508655548,
      "mean_token_accuracy": 0.9233558624982834,
      "num_tokens": 74136276.0,
      "step": 6010
    },
    {
      "entropy": 0.9672934710979462,
      "epoch": 3.1653501843075302,
      "grad_norm": 0.3099748194217682,
      "learning_rate": 0.0001282055468189464,
      "loss": 0.2050492912530899,
      "mean_token_accuracy": 0.9136051535606384,
      "num_tokens": 74148612.0,
      "step": 6011
    },
    {
      "entropy": 0.8931698203086853,
      "epoch": 3.1658767772511847,
      "grad_norm": 0.2969735264778137,
      "learning_rate": 0.0001281840713162423,
      "loss": 0.17930302023887634,
      "mean_token_accuracy": 0.9252002835273743,
      "num_tokens": 74160948.0,
      "step": 6012
    },
    {
      "entropy": 0.9153631329536438,
      "epoch": 3.1664033701948395,
      "grad_norm": 0.29347336292266846,
      "learning_rate": 0.0001281625947340558,
      "loss": 0.1735287308692932,
      "mean_token_accuracy": 0.9257992506027222,
      "num_tokens": 74173284.0,
      "step": 6013
    },
    {
      "entropy": 0.9040100425481796,
      "epoch": 3.166929963138494,
      "grad_norm": 0.2996053695678711,
      "learning_rate": 0.00012814111707366195,
      "loss": 0.1787641942501068,
      "mean_token_accuracy": 0.9250529259443283,
      "num_tokens": 74185620.0,
      "step": 6014
    },
    {
      "entropy": 0.9010534137487411,
      "epoch": 3.1674565560821484,
      "grad_norm": 0.25485759973526,
      "learning_rate": 0.0001281196383363357,
      "loss": 0.159829780459404,
      "mean_token_accuracy": 0.9268502444028854,
      "num_tokens": 74197956.0,
      "step": 6015
    },
    {
      "entropy": 0.9162964373826981,
      "epoch": 3.1679831490258032,
      "grad_norm": 0.3184720277786255,
      "learning_rate": 0.00012809815852335213,
      "loss": 0.18885093927383423,
      "mean_token_accuracy": 0.9227601438760757,
      "num_tokens": 74210292.0,
      "step": 6016
    },
    {
      "entropy": 0.886315330862999,
      "epoch": 3.1685097419694577,
      "grad_norm": 0.2911457121372223,
      "learning_rate": 0.00012807667763598638,
      "loss": 0.17227713763713837,
      "mean_token_accuracy": 0.9295205473899841,
      "num_tokens": 74222628.0,
      "step": 6017
    },
    {
      "entropy": 0.9286379963159561,
      "epoch": 3.169036334913112,
      "grad_norm": 0.27264803647994995,
      "learning_rate": 0.0001280551956755136,
      "loss": 0.1657165288925171,
      "mean_token_accuracy": 0.9366953074932098,
      "num_tokens": 74234964.0,
      "step": 6018
    },
    {
      "entropy": 0.930014505982399,
      "epoch": 3.1695629278567665,
      "grad_norm": 0.3046049475669861,
      "learning_rate": 0.00012803371264320912,
      "loss": 0.1632966250181198,
      "mean_token_accuracy": 0.9310959875583649,
      "num_tokens": 74247300.0,
      "step": 6019
    },
    {
      "entropy": 0.9086096882820129,
      "epoch": 3.1700895208004214,
      "grad_norm": 0.2823864221572876,
      "learning_rate": 0.00012801222854034826,
      "loss": 0.1753690093755722,
      "mean_token_accuracy": 0.9284543991088867,
      "num_tokens": 74259636.0,
      "step": 6020
    },
    {
      "entropy": 0.8453738987445831,
      "epoch": 3.170616113744076,
      "grad_norm": 0.2871163487434387,
      "learning_rate": 0.00012799074336820635,
      "loss": 0.18023160099983215,
      "mean_token_accuracy": 0.9228629618883133,
      "num_tokens": 74271972.0,
      "step": 6021
    },
    {
      "entropy": 0.904409795999527,
      "epoch": 3.1711427066877302,
      "grad_norm": 0.32247886061668396,
      "learning_rate": 0.00012796925712805883,
      "loss": 0.20106787979602814,
      "mean_token_accuracy": 0.9166100919246674,
      "num_tokens": 74284308.0,
      "step": 6022
    },
    {
      "entropy": 0.8857452869415283,
      "epoch": 3.171669299631385,
      "grad_norm": 0.31473007798194885,
      "learning_rate": 0.00012794776982118127,
      "loss": 0.206316739320755,
      "mean_token_accuracy": 0.9149357676506042,
      "num_tokens": 74296644.0,
      "step": 6023
    },
    {
      "entropy": 0.8967972993850708,
      "epoch": 3.1721958925750395,
      "grad_norm": 0.2719268798828125,
      "learning_rate": 0.00012792628144884925,
      "loss": 0.1664860099554062,
      "mean_token_accuracy": 0.9297046065330505,
      "num_tokens": 74308980.0,
      "step": 6024
    },
    {
      "entropy": 0.919925644993782,
      "epoch": 3.172722485518694,
      "grad_norm": 0.30338454246520996,
      "learning_rate": 0.00012790479201233834,
      "loss": 0.17423170804977417,
      "mean_token_accuracy": 0.9267141819000244,
      "num_tokens": 74321316.0,
      "step": 6025
    },
    {
      "entropy": 0.9101917892694473,
      "epoch": 3.173249078462349,
      "grad_norm": 0.2964281737804413,
      "learning_rate": 0.00012788330151292432,
      "loss": 0.17755359411239624,
      "mean_token_accuracy": 0.925290584564209,
      "num_tokens": 74333652.0,
      "step": 6026
    },
    {
      "entropy": 0.868523046374321,
      "epoch": 3.1737756714060033,
      "grad_norm": 0.32196730375289917,
      "learning_rate": 0.0001278618099518829,
      "loss": 0.2060573548078537,
      "mean_token_accuracy": 0.9190028607845306,
      "num_tokens": 74345988.0,
      "step": 6027
    },
    {
      "entropy": 0.8719951957464218,
      "epoch": 3.1743022643496577,
      "grad_norm": 0.2882232964038849,
      "learning_rate": 0.00012784031733048992,
      "loss": 0.1669994592666626,
      "mean_token_accuracy": 0.9271730333566666,
      "num_tokens": 74358324.0,
      "step": 6028
    },
    {
      "entropy": 0.8512067198753357,
      "epoch": 3.174828857293312,
      "grad_norm": 0.29102852940559387,
      "learning_rate": 0.00012781882365002132,
      "loss": 0.1903591752052307,
      "mean_token_accuracy": 0.9173757582902908,
      "num_tokens": 74370660.0,
      "step": 6029
    },
    {
      "entropy": 0.8923112750053406,
      "epoch": 3.175355450236967,
      "grad_norm": 0.28172409534454346,
      "learning_rate": 0.000127797328911753,
      "loss": 0.17183968424797058,
      "mean_token_accuracy": 0.9267172068357468,
      "num_tokens": 74382996.0,
      "step": 6030
    },
    {
      "entropy": 0.8994222730398178,
      "epoch": 3.1758820431806214,
      "grad_norm": 0.30900153517723083,
      "learning_rate": 0.00012777583311696097,
      "loss": 0.19755873084068298,
      "mean_token_accuracy": 0.9184713661670685,
      "num_tokens": 74395332.0,
      "step": 6031
    },
    {
      "entropy": 0.9129990190267563,
      "epoch": 3.176408636124276,
      "grad_norm": 0.2915842533111572,
      "learning_rate": 0.00012775433626692131,
      "loss": 0.17764748632907867,
      "mean_token_accuracy": 0.9233520030975342,
      "num_tokens": 74407668.0,
      "step": 6032
    },
    {
      "entropy": 0.8965678662061691,
      "epoch": 3.1769352290679307,
      "grad_norm": 0.3085368573665619,
      "learning_rate": 0.00012773283836291027,
      "loss": 0.19080962240695953,
      "mean_token_accuracy": 0.9208065420389175,
      "num_tokens": 74420004.0,
      "step": 6033
    },
    {
      "entropy": 0.9515413045883179,
      "epoch": 3.177461822011585,
      "grad_norm": 0.28024330735206604,
      "learning_rate": 0.0001277113394062039,
      "loss": 0.1752488613128662,
      "mean_token_accuracy": 0.923393651843071,
      "num_tokens": 74432340.0,
      "step": 6034
    },
    {
      "entropy": 0.8868319243192673,
      "epoch": 3.1779884149552395,
      "grad_norm": 0.2637077867984772,
      "learning_rate": 0.00012768983939807858,
      "loss": 0.1692328155040741,
      "mean_token_accuracy": 0.9270408600568771,
      "num_tokens": 74444676.0,
      "step": 6035
    },
    {
      "entropy": 0.9367111325263977,
      "epoch": 3.178515007898894,
      "grad_norm": 0.30866900086402893,
      "learning_rate": 0.0001276683383398106,
      "loss": 0.19459642469882965,
      "mean_token_accuracy": 0.9200681746006012,
      "num_tokens": 74457012.0,
      "step": 6036
    },
    {
      "entropy": 0.9755773544311523,
      "epoch": 3.179041600842549,
      "grad_norm": 0.29067596793174744,
      "learning_rate": 0.00012764683623267632,
      "loss": 0.17644400894641876,
      "mean_token_accuracy": 0.9248355776071548,
      "num_tokens": 74469348.0,
      "step": 6037
    },
    {
      "entropy": 0.9238932579755783,
      "epoch": 3.1795681937862033,
      "grad_norm": 0.2967524528503418,
      "learning_rate": 0.00012762533307795226,
      "loss": 0.18571233749389648,
      "mean_token_accuracy": 0.9185805767774582,
      "num_tokens": 74481684.0,
      "step": 6038
    },
    {
      "entropy": 0.9369767755270004,
      "epoch": 3.1800947867298577,
      "grad_norm": 0.27729305624961853,
      "learning_rate": 0.0001276038288769149,
      "loss": 0.17377406358718872,
      "mean_token_accuracy": 0.9294021874666214,
      "num_tokens": 74494020.0,
      "step": 6039
    },
    {
      "entropy": 0.9587116092443466,
      "epoch": 3.1806213796735125,
      "grad_norm": 0.2849549651145935,
      "learning_rate": 0.0001275823236308408,
      "loss": 0.17863985896110535,
      "mean_token_accuracy": 0.9239156097173691,
      "num_tokens": 74506356.0,
      "step": 6040
    },
    {
      "entropy": 0.906728059053421,
      "epoch": 3.181147972617167,
      "grad_norm": 0.2766047418117523,
      "learning_rate": 0.00012756081734100662,
      "loss": 0.17260222136974335,
      "mean_token_accuracy": 0.9303994923830032,
      "num_tokens": 74518692.0,
      "step": 6041
    },
    {
      "entropy": 0.9309165328741074,
      "epoch": 3.1816745655608214,
      "grad_norm": 0.25417715311050415,
      "learning_rate": 0.00012753931000868908,
      "loss": 0.17008274793624878,
      "mean_token_accuracy": 0.9315279871225357,
      "num_tokens": 74531028.0,
      "step": 6042
    },
    {
      "entropy": 0.9460414797067642,
      "epoch": 3.1822011585044763,
      "grad_norm": 0.2841247320175171,
      "learning_rate": 0.0001275178016351649,
      "loss": 0.20334357023239136,
      "mean_token_accuracy": 0.9160185158252716,
      "num_tokens": 74543364.0,
      "step": 6043
    },
    {
      "entropy": 0.9646149277687073,
      "epoch": 3.1827277514481307,
      "grad_norm": 0.27899524569511414,
      "learning_rate": 0.000127496292221711,
      "loss": 0.16916584968566895,
      "mean_token_accuracy": 0.9278223812580109,
      "num_tokens": 74555700.0,
      "step": 6044
    },
    {
      "entropy": 1.0223030000925064,
      "epoch": 3.183254344391785,
      "grad_norm": 0.3250932991504669,
      "learning_rate": 0.0001274747817696042,
      "loss": 0.20184363424777985,
      "mean_token_accuracy": 0.9218162596225739,
      "num_tokens": 74568036.0,
      "step": 6045
    },
    {
      "entropy": 0.9621344953775406,
      "epoch": 3.1837809373354395,
      "grad_norm": 0.26898354291915894,
      "learning_rate": 0.0001274532702801214,
      "loss": 0.16600225865840912,
      "mean_token_accuracy": 0.9326440840959549,
      "num_tokens": 74580372.0,
      "step": 6046
    },
    {
      "entropy": 0.9494886696338654,
      "epoch": 3.1843075302790944,
      "grad_norm": 0.27224642038345337,
      "learning_rate": 0.0001274317577545397,
      "loss": 0.17049263417720795,
      "mean_token_accuracy": 0.9249418377876282,
      "num_tokens": 74592708.0,
      "step": 6047
    },
    {
      "entropy": 0.9812628626823425,
      "epoch": 3.184834123222749,
      "grad_norm": 0.2864963710308075,
      "learning_rate": 0.00012741024419413613,
      "loss": 0.1766996532678604,
      "mean_token_accuracy": 0.9260700345039368,
      "num_tokens": 74605044.0,
      "step": 6048
    },
    {
      "entropy": 1.0052879303693771,
      "epoch": 3.1853607161664033,
      "grad_norm": 0.3256306052207947,
      "learning_rate": 0.0001273887296001879,
      "loss": 0.19246399402618408,
      "mean_token_accuracy": 0.9181402176618576,
      "num_tokens": 74617380.0,
      "step": 6049
    },
    {
      "entropy": 1.0402062833309174,
      "epoch": 3.185887309110058,
      "grad_norm": 0.31314828991889954,
      "learning_rate": 0.00012736721397397206,
      "loss": 0.18360930681228638,
      "mean_token_accuracy": 0.9209848344326019,
      "num_tokens": 74629716.0,
      "step": 6050
    },
    {
      "entropy": 1.0098388940095901,
      "epoch": 3.1864139020537126,
      "grad_norm": 0.2711458206176758,
      "learning_rate": 0.000127345697316766,
      "loss": 0.16650447249412537,
      "mean_token_accuracy": 0.927087277173996,
      "num_tokens": 74642052.0,
      "step": 6051
    },
    {
      "entropy": 1.044172078371048,
      "epoch": 3.186940494997367,
      "grad_norm": 0.3235626518726349,
      "learning_rate": 0.00012732417962984697,
      "loss": 0.19019578397274017,
      "mean_token_accuracy": 0.9197899550199509,
      "num_tokens": 74654388.0,
      "step": 6052
    },
    {
      "entropy": 1.0233334600925446,
      "epoch": 3.1874670879410214,
      "grad_norm": 0.28503936529159546,
      "learning_rate": 0.0001273026609144924,
      "loss": 0.18028733134269714,
      "mean_token_accuracy": 0.9271443486213684,
      "num_tokens": 74666724.0,
      "step": 6053
    },
    {
      "entropy": 1.0430432260036469,
      "epoch": 3.1879936808846763,
      "grad_norm": 0.28121379017829895,
      "learning_rate": 0.00012728114117197963,
      "loss": 0.17312172055244446,
      "mean_token_accuracy": 0.9267847239971161,
      "num_tokens": 74679060.0,
      "step": 6054
    },
    {
      "entropy": 0.9970735609531403,
      "epoch": 3.1885202738283307,
      "grad_norm": 0.31420785188674927,
      "learning_rate": 0.0001272596204035863,
      "loss": 0.1869233250617981,
      "mean_token_accuracy": 0.9196929782629013,
      "num_tokens": 74691396.0,
      "step": 6055
    },
    {
      "entropy": 1.0581863820552826,
      "epoch": 3.189046866771985,
      "grad_norm": 0.31081312894821167,
      "learning_rate": 0.00012723809861058986,
      "loss": 0.17754991352558136,
      "mean_token_accuracy": 0.9275716245174408,
      "num_tokens": 74703732.0,
      "step": 6056
    },
    {
      "entropy": 1.0594633221626282,
      "epoch": 3.18957345971564,
      "grad_norm": 0.2916032373905182,
      "learning_rate": 0.00012721657579426796,
      "loss": 0.16658541560173035,
      "mean_token_accuracy": 0.9299917966127396,
      "num_tokens": 74716068.0,
      "step": 6057
    },
    {
      "entropy": 1.0787553191184998,
      "epoch": 3.1901000526592944,
      "grad_norm": 0.2959164083003998,
      "learning_rate": 0.00012719505195589833,
      "loss": 0.18187478184700012,
      "mean_token_accuracy": 0.9289588928222656,
      "num_tokens": 74728404.0,
      "step": 6058
    },
    {
      "entropy": 1.0568149238824844,
      "epoch": 3.190626645602949,
      "grad_norm": 0.26274511218070984,
      "learning_rate": 0.00012717352709675872,
      "loss": 0.16812138259410858,
      "mean_token_accuracy": 0.9347347170114517,
      "num_tokens": 74740740.0,
      "step": 6059
    },
    {
      "entropy": 1.0166764706373215,
      "epoch": 3.1911532385466037,
      "grad_norm": 0.27479109168052673,
      "learning_rate": 0.00012715200121812687,
      "loss": 0.17319561541080475,
      "mean_token_accuracy": 0.9268957078456879,
      "num_tokens": 74753076.0,
      "step": 6060
    },
    {
      "entropy": 1.090210884809494,
      "epoch": 3.191679831490258,
      "grad_norm": 0.30568063259124756,
      "learning_rate": 0.00012713047432128067,
      "loss": 0.17797154188156128,
      "mean_token_accuracy": 0.9297853410243988,
      "num_tokens": 74765412.0,
      "step": 6061
    },
    {
      "entropy": 1.0944711565971375,
      "epoch": 3.1922064244339126,
      "grad_norm": 0.29322412610054016,
      "learning_rate": 0.00012710894640749807,
      "loss": 0.17470967769622803,
      "mean_token_accuracy": 0.9270069003105164,
      "num_tokens": 74777748.0,
      "step": 6062
    },
    {
      "entropy": 1.0665898323059082,
      "epoch": 3.192733017377567,
      "grad_norm": 0.2914141118526459,
      "learning_rate": 0.00012708741747805702,
      "loss": 0.19307973980903625,
      "mean_token_accuracy": 0.9191839843988419,
      "num_tokens": 74790084.0,
      "step": 6063
    },
    {
      "entropy": 1.0894026160240173,
      "epoch": 3.193259610321222,
      "grad_norm": 0.285011351108551,
      "learning_rate": 0.0001270658875342356,
      "loss": 0.17769688367843628,
      "mean_token_accuracy": 0.9278746247291565,
      "num_tokens": 74802420.0,
      "step": 6064
    },
    {
      "entropy": 1.113456517457962,
      "epoch": 3.1937862032648763,
      "grad_norm": 0.31497251987457275,
      "learning_rate": 0.00012704435657731188,
      "loss": 0.1936781406402588,
      "mean_token_accuracy": 0.9238170832395554,
      "num_tokens": 74814756.0,
      "step": 6065
    },
    {
      "entropy": 1.0806358754634857,
      "epoch": 3.1943127962085307,
      "grad_norm": 0.28313109278678894,
      "learning_rate": 0.00012702282460856407,
      "loss": 0.18170379102230072,
      "mean_token_accuracy": 0.9257088601589203,
      "num_tokens": 74827092.0,
      "step": 6066
    },
    {
      "entropy": 1.1051075458526611,
      "epoch": 3.194839389152185,
      "grad_norm": 0.2947232723236084,
      "learning_rate": 0.0001270012916292704,
      "loss": 0.1816052347421646,
      "mean_token_accuracy": 0.9246692359447479,
      "num_tokens": 74839428.0,
      "step": 6067
    },
    {
      "entropy": 1.17019122838974,
      "epoch": 3.19536598209584,
      "grad_norm": 0.31820914149284363,
      "learning_rate": 0.00012697975764070914,
      "loss": 0.18826723098754883,
      "mean_token_accuracy": 0.9142595082521439,
      "num_tokens": 74851764.0,
      "step": 6068
    },
    {
      "entropy": 1.102179765701294,
      "epoch": 3.1958925750394944,
      "grad_norm": 0.30241599678993225,
      "learning_rate": 0.0001269582226441586,
      "loss": 0.2013278603553772,
      "mean_token_accuracy": 0.9154189229011536,
      "num_tokens": 74864100.0,
      "step": 6069
    },
    {
      "entropy": 1.1176361441612244,
      "epoch": 3.196419167983149,
      "grad_norm": 0.2859967052936554,
      "learning_rate": 0.00012693668664089724,
      "loss": 0.18881945312023163,
      "mean_token_accuracy": 0.9228017330169678,
      "num_tokens": 74876436.0,
      "step": 6070
    },
    {
      "entropy": 1.147952139377594,
      "epoch": 3.1969457609268037,
      "grad_norm": 0.3129325807094574,
      "learning_rate": 0.0001269151496322035,
      "loss": 0.20676347613334656,
      "mean_token_accuracy": 0.9172220826148987,
      "num_tokens": 74888772.0,
      "step": 6071
    },
    {
      "entropy": 1.0977469086647034,
      "epoch": 3.197472353870458,
      "grad_norm": 0.2926788926124573,
      "learning_rate": 0.0001268936116193559,
      "loss": 0.17107011377811432,
      "mean_token_accuracy": 0.929108127951622,
      "num_tokens": 74901108.0,
      "step": 6072
    },
    {
      "entropy": 1.117818146944046,
      "epoch": 3.1979989468141126,
      "grad_norm": 0.28832677006721497,
      "learning_rate": 0.00012687207260363308,
      "loss": 0.18728822469711304,
      "mean_token_accuracy": 0.9181636422872543,
      "num_tokens": 74913444.0,
      "step": 6073
    },
    {
      "entropy": 1.1389150619506836,
      "epoch": 3.1985255397577674,
      "grad_norm": 0.28520458936691284,
      "learning_rate": 0.00012685053258631364,
      "loss": 0.18073593080043793,
      "mean_token_accuracy": 0.9208191931247711,
      "num_tokens": 74925780.0,
      "step": 6074
    },
    {
      "entropy": 1.139819324016571,
      "epoch": 3.199052132701422,
      "grad_norm": 0.3071376383304596,
      "learning_rate": 0.00012682899156867626,
      "loss": 0.19548124074935913,
      "mean_token_accuracy": 0.9135144054889679,
      "num_tokens": 74938116.0,
      "step": 6075
    },
    {
      "entropy": 1.1386458277702332,
      "epoch": 3.1995787256450763,
      "grad_norm": 0.2790388762950897,
      "learning_rate": 0.00012680744955199976,
      "loss": 0.17659339308738708,
      "mean_token_accuracy": 0.9237524420022964,
      "num_tokens": 74950452.0,
      "step": 6076
    },
    {
      "entropy": 1.1673567295074463,
      "epoch": 3.2001053185887307,
      "grad_norm": 0.2836650013923645,
      "learning_rate": 0.00012678590653756295,
      "loss": 0.18439726531505585,
      "mean_token_accuracy": 0.9258717596530914,
      "num_tokens": 74962788.0,
      "step": 6077
    },
    {
      "entropy": 1.1256002485752106,
      "epoch": 3.2006319115323856,
      "grad_norm": 0.2836996018886566,
      "learning_rate": 0.00012676436252664468,
      "loss": 0.1797911673784256,
      "mean_token_accuracy": 0.9222531169652939,
      "num_tokens": 74975124.0,
      "step": 6078
    },
    {
      "entropy": 1.139472633600235,
      "epoch": 3.20115850447604,
      "grad_norm": 0.29277554154396057,
      "learning_rate": 0.00012674281752052396,
      "loss": 0.1976083517074585,
      "mean_token_accuracy": 0.9170555472373962,
      "num_tokens": 74987460.0,
      "step": 6079
    },
    {
      "entropy": 1.1551057398319244,
      "epoch": 3.2016850974196944,
      "grad_norm": 0.2957473695278168,
      "learning_rate": 0.00012672127152047973,
      "loss": 0.1852739304304123,
      "mean_token_accuracy": 0.9232263416051865,
      "num_tokens": 74999796.0,
      "step": 6080
    },
    {
      "entropy": 1.1753365695476532,
      "epoch": 3.2022116903633493,
      "grad_norm": 0.29280924797058105,
      "learning_rate": 0.000126699724527791,
      "loss": 0.18121716380119324,
      "mean_token_accuracy": 0.9265182018280029,
      "num_tokens": 75012132.0,
      "step": 6081
    },
    {
      "entropy": 1.1696423590183258,
      "epoch": 3.2027382833070037,
      "grad_norm": 0.3267524838447571,
      "learning_rate": 0.00012667817654373704,
      "loss": 0.19591163098812103,
      "mean_token_accuracy": 0.9163945615291595,
      "num_tokens": 75024468.0,
      "step": 6082
    },
    {
      "entropy": 1.2067201733589172,
      "epoch": 3.203264876250658,
      "grad_norm": 0.2947297990322113,
      "learning_rate": 0.00012665662756959693,
      "loss": 0.1690807342529297,
      "mean_token_accuracy": 0.9303170293569565,
      "num_tokens": 75036804.0,
      "step": 6083
    },
    {
      "entropy": 1.1902517676353455,
      "epoch": 3.2037914691943126,
      "grad_norm": 0.29545536637306213,
      "learning_rate": 0.00012663507760664994,
      "loss": 0.18513049185276031,
      "mean_token_accuracy": 0.9230286926031113,
      "num_tokens": 75049140.0,
      "step": 6084
    },
    {
      "entropy": 1.1767702102661133,
      "epoch": 3.2043180621379674,
      "grad_norm": 0.29497984051704407,
      "learning_rate": 0.0001266135266561753,
      "loss": 0.17692333459854126,
      "mean_token_accuracy": 0.9241546839475632,
      "num_tokens": 75061476.0,
      "step": 6085
    },
    {
      "entropy": 1.1620597839355469,
      "epoch": 3.204844655081622,
      "grad_norm": 0.28324300050735474,
      "learning_rate": 0.00012659197471945246,
      "loss": 0.18718908727169037,
      "mean_token_accuracy": 0.9241278767585754,
      "num_tokens": 75073812.0,
      "step": 6086
    },
    {
      "entropy": 1.1229282915592194,
      "epoch": 3.2053712480252763,
      "grad_norm": 0.25627800822257996,
      "learning_rate": 0.00012657042179776078,
      "loss": 0.1552576869726181,
      "mean_token_accuracy": 0.9354147613048553,
      "num_tokens": 75086148.0,
      "step": 6087
    },
    {
      "entropy": 1.1936973929405212,
      "epoch": 3.205897840968931,
      "grad_norm": 0.28294000029563904,
      "learning_rate": 0.0001265488678923797,
      "loss": 0.1731725037097931,
      "mean_token_accuracy": 0.925356075167656,
      "num_tokens": 75098484.0,
      "step": 6088
    },
    {
      "entropy": 1.1629082262516022,
      "epoch": 3.2064244339125856,
      "grad_norm": 0.2877937853336334,
      "learning_rate": 0.00012652731300458883,
      "loss": 0.17543064057826996,
      "mean_token_accuracy": 0.9272084087133408,
      "num_tokens": 75110820.0,
      "step": 6089
    },
    {
      "entropy": 1.2082597315311432,
      "epoch": 3.20695102685624,
      "grad_norm": 0.2778930962085724,
      "learning_rate": 0.00012650575713566767,
      "loss": 0.16450908780097961,
      "mean_token_accuracy": 0.9307198077440262,
      "num_tokens": 75123156.0,
      "step": 6090
    },
    {
      "entropy": 1.226649910211563,
      "epoch": 3.207477619799895,
      "grad_norm": 0.2595750391483307,
      "learning_rate": 0.00012648420028689596,
      "loss": 0.16251420974731445,
      "mean_token_accuracy": 0.9295267313718796,
      "num_tokens": 75135492.0,
      "step": 6091
    },
    {
      "entropy": 1.158011019229889,
      "epoch": 3.2080042127435493,
      "grad_norm": 0.2987614870071411,
      "learning_rate": 0.0001264626424595533,
      "loss": 0.18314017355442047,
      "mean_token_accuracy": 0.9236122965812683,
      "num_tokens": 75147828.0,
      "step": 6092
    },
    {
      "entropy": 1.1719801127910614,
      "epoch": 3.2085308056872037,
      "grad_norm": 0.308034211397171,
      "learning_rate": 0.00012644108365491958,
      "loss": 0.19339613616466522,
      "mean_token_accuracy": 0.9197325706481934,
      "num_tokens": 75160164.0,
      "step": 6093
    },
    {
      "entropy": 1.1360925734043121,
      "epoch": 3.209057398630858,
      "grad_norm": 0.2838227152824402,
      "learning_rate": 0.00012641952387427448,
      "loss": 0.1735890656709671,
      "mean_token_accuracy": 0.9284939914941788,
      "num_tokens": 75172500.0,
      "step": 6094
    },
    {
      "entropy": 1.1555771827697754,
      "epoch": 3.209583991574513,
      "grad_norm": 0.3021514415740967,
      "learning_rate": 0.00012639796311889796,
      "loss": 0.18424902856349945,
      "mean_token_accuracy": 0.9217068552970886,
      "num_tokens": 75184836.0,
      "step": 6095
    },
    {
      "entropy": 1.1850262582302094,
      "epoch": 3.2101105845181674,
      "grad_norm": 0.29426127672195435,
      "learning_rate": 0.00012637640139006998,
      "loss": 0.1832590401172638,
      "mean_token_accuracy": 0.9267239272594452,
      "num_tokens": 75197172.0,
      "step": 6096
    },
    {
      "entropy": 1.174637347459793,
      "epoch": 3.210637177461822,
      "grad_norm": 0.2878929376602173,
      "learning_rate": 0.0001263548386890705,
      "loss": 0.18444159626960754,
      "mean_token_accuracy": 0.9251533895730972,
      "num_tokens": 75209508.0,
      "step": 6097
    },
    {
      "entropy": 1.2012336254119873,
      "epoch": 3.2111637704054767,
      "grad_norm": 0.29878565669059753,
      "learning_rate": 0.00012633327501717956,
      "loss": 0.18056027591228485,
      "mean_token_accuracy": 0.9282454997301102,
      "num_tokens": 75221844.0,
      "step": 6098
    },
    {
      "entropy": 1.1817958354949951,
      "epoch": 3.211690363349131,
      "grad_norm": 0.2876642346382141,
      "learning_rate": 0.00012631171037567727,
      "loss": 0.17095500230789185,
      "mean_token_accuracy": 0.924737811088562,
      "num_tokens": 75234180.0,
      "step": 6099
    },
    {
      "entropy": 1.1270457804203033,
      "epoch": 3.2122169562927856,
      "grad_norm": 0.24099025130271912,
      "learning_rate": 0.0001262901447658438,
      "loss": 0.15860724449157715,
      "mean_token_accuracy": 0.931660920381546,
      "num_tokens": 75246516.0,
      "step": 6100
    },
    {
      "entropy": 1.1745367348194122,
      "epoch": 3.21274354923644,
      "grad_norm": 0.2938980460166931,
      "learning_rate": 0.00012626857818895938,
      "loss": 0.18283730745315552,
      "mean_token_accuracy": 0.9220217019319534,
      "num_tokens": 75258852.0,
      "step": 6101
    },
    {
      "entropy": 1.1532530188560486,
      "epoch": 3.213270142180095,
      "grad_norm": 0.27625831961631775,
      "learning_rate": 0.00012624701064630433,
      "loss": 0.17850688099861145,
      "mean_token_accuracy": 0.9294148087501526,
      "num_tokens": 75271188.0,
      "step": 6102
    },
    {
      "entropy": 1.160976082086563,
      "epoch": 3.2137967351237493,
      "grad_norm": 0.29574382305145264,
      "learning_rate": 0.00012622544213915894,
      "loss": 0.17223690450191498,
      "mean_token_accuracy": 0.9264660626649857,
      "num_tokens": 75283524.0,
      "step": 6103
    },
    {
      "entropy": 1.1664827466011047,
      "epoch": 3.2143233280674037,
      "grad_norm": 0.29536303877830505,
      "learning_rate": 0.0001262038726688036,
      "loss": 0.18994994461536407,
      "mean_token_accuracy": 0.9243973344564438,
      "num_tokens": 75295860.0,
      "step": 6104
    },
    {
      "entropy": 1.1745307743549347,
      "epoch": 3.2148499210110586,
      "grad_norm": 0.30445703864097595,
      "learning_rate": 0.00012618230223651876,
      "loss": 0.1881660670042038,
      "mean_token_accuracy": 0.923809602856636,
      "num_tokens": 75308196.0,
      "step": 6105
    },
    {
      "entropy": 1.1292920410633087,
      "epoch": 3.215376513954713,
      "grad_norm": 0.2815026044845581,
      "learning_rate": 0.000126160730843585,
      "loss": 0.18442246317863464,
      "mean_token_accuracy": 0.9258362799882889,
      "num_tokens": 75320532.0,
      "step": 6106
    },
    {
      "entropy": 1.1680370271205902,
      "epoch": 3.2159031068983674,
      "grad_norm": 0.3535638451576233,
      "learning_rate": 0.0001261391584912828,
      "loss": 0.2101319283246994,
      "mean_token_accuracy": 0.9123208075761795,
      "num_tokens": 75332868.0,
      "step": 6107
    },
    {
      "entropy": 1.1719010472297668,
      "epoch": 3.2164296998420223,
      "grad_norm": 0.31875982880592346,
      "learning_rate": 0.00012611758518089286,
      "loss": 0.1957074999809265,
      "mean_token_accuracy": 0.9135269224643707,
      "num_tokens": 75345204.0,
      "step": 6108
    },
    {
      "entropy": 1.2349201440811157,
      "epoch": 3.2169562927856767,
      "grad_norm": 0.3066956102848053,
      "learning_rate": 0.00012609601091369583,
      "loss": 0.1689293086528778,
      "mean_token_accuracy": 0.9295897334814072,
      "num_tokens": 75357540.0,
      "step": 6109
    },
    {
      "entropy": 1.1840606331825256,
      "epoch": 3.217482885729331,
      "grad_norm": 0.29718321561813354,
      "learning_rate": 0.00012607443569097245,
      "loss": 0.20115578174591064,
      "mean_token_accuracy": 0.9171886295080185,
      "num_tokens": 75369876.0,
      "step": 6110
    },
    {
      "entropy": 1.1642614603042603,
      "epoch": 3.2180094786729856,
      "grad_norm": 0.2805807590484619,
      "learning_rate": 0.0001260528595140035,
      "loss": 0.17763221263885498,
      "mean_token_accuracy": 0.9212594777345657,
      "num_tokens": 75382212.0,
      "step": 6111
    },
    {
      "entropy": 1.2080804705619812,
      "epoch": 3.2185360716166405,
      "grad_norm": 0.284940242767334,
      "learning_rate": 0.00012603128238406985,
      "loss": 0.1861756145954132,
      "mean_token_accuracy": 0.9266208410263062,
      "num_tokens": 75394548.0,
      "step": 6112
    },
    {
      "entropy": 1.2124915719032288,
      "epoch": 3.219062664560295,
      "grad_norm": 0.3166573643684387,
      "learning_rate": 0.00012600970430245236,
      "loss": 0.20203432440757751,
      "mean_token_accuracy": 0.9154464155435562,
      "num_tokens": 75406884.0,
      "step": 6113
    },
    {
      "entropy": 1.1943967640399933,
      "epoch": 3.2195892575039493,
      "grad_norm": 0.2512170672416687,
      "learning_rate": 0.00012598812527043208,
      "loss": 0.16128602623939514,
      "mean_token_accuracy": 0.9311299026012421,
      "num_tokens": 75419220.0,
      "step": 6114
    },
    {
      "entropy": 1.2292022109031677,
      "epoch": 3.220115850447604,
      "grad_norm": 0.28310030698776245,
      "learning_rate": 0.00012596654528929,
      "loss": 0.1860269159078598,
      "mean_token_accuracy": 0.9250501245260239,
      "num_tokens": 75431556.0,
      "step": 6115
    },
    {
      "entropy": 1.2319591641426086,
      "epoch": 3.2206424433912586,
      "grad_norm": 0.3099038898944855,
      "learning_rate": 0.00012594496436030714,
      "loss": 0.19201034307479858,
      "mean_token_accuracy": 0.9213238954544067,
      "num_tokens": 75443892.0,
      "step": 6116
    },
    {
      "entropy": 1.2286298871040344,
      "epoch": 3.221169036334913,
      "grad_norm": 0.29794779419898987,
      "learning_rate": 0.0001259233824847647,
      "loss": 0.1706649363040924,
      "mean_token_accuracy": 0.9281015396118164,
      "num_tokens": 75456228.0,
      "step": 6117
    },
    {
      "entropy": 1.2060716450214386,
      "epoch": 3.2216956292785675,
      "grad_norm": 0.30427494645118713,
      "learning_rate": 0.00012590179966394388,
      "loss": 0.19397449493408203,
      "mean_token_accuracy": 0.9212569743394852,
      "num_tokens": 75468564.0,
      "step": 6118
    },
    {
      "entropy": 1.239819973707199,
      "epoch": 3.2222222222222223,
      "grad_norm": 0.2990228831768036,
      "learning_rate": 0.0001258802158991259,
      "loss": 0.18792952597141266,
      "mean_token_accuracy": 0.9206580072641373,
      "num_tokens": 75480900.0,
      "step": 6119
    },
    {
      "entropy": 1.23019939661026,
      "epoch": 3.2227488151658767,
      "grad_norm": 0.27720263600349426,
      "learning_rate": 0.00012585863119159198,
      "loss": 0.18679332733154297,
      "mean_token_accuracy": 0.9180274903774261,
      "num_tokens": 75493236.0,
      "step": 6120
    },
    {
      "entropy": 1.2414479851722717,
      "epoch": 3.223275408109531,
      "grad_norm": 0.28125715255737305,
      "learning_rate": 0.00012583704554262364,
      "loss": 0.16257213056087494,
      "mean_token_accuracy": 0.9326681643724442,
      "num_tokens": 75505572.0,
      "step": 6121
    },
    {
      "entropy": 1.2053897678852081,
      "epoch": 3.223802001053186,
      "grad_norm": 0.26854726672172546,
      "learning_rate": 0.00012581545895350217,
      "loss": 0.17040804028511047,
      "mean_token_accuracy": 0.9294180870056152,
      "num_tokens": 75517908.0,
      "step": 6122
    },
    {
      "entropy": 1.2712475657463074,
      "epoch": 3.2243285939968405,
      "grad_norm": 0.30547404289245605,
      "learning_rate": 0.0001257938714255091,
      "loss": 0.18066394329071045,
      "mean_token_accuracy": 0.9216051995754242,
      "num_tokens": 75530244.0,
      "step": 6123
    },
    {
      "entropy": 1.2184970080852509,
      "epoch": 3.224855186940495,
      "grad_norm": 0.2835502624511719,
      "learning_rate": 0.0001257722829599259,
      "loss": 0.1837792992591858,
      "mean_token_accuracy": 0.9190782308578491,
      "num_tokens": 75542580.0,
      "step": 6124
    },
    {
      "entropy": 1.191634327173233,
      "epoch": 3.2253817798841498,
      "grad_norm": 0.2832392156124115,
      "learning_rate": 0.00012575069355803427,
      "loss": 0.17726731300354004,
      "mean_token_accuracy": 0.9231899827718735,
      "num_tokens": 75554916.0,
      "step": 6125
    },
    {
      "entropy": 1.2391059696674347,
      "epoch": 3.225908372827804,
      "grad_norm": 0.30272725224494934,
      "learning_rate": 0.0001257291032211157,
      "loss": 0.18769414722919464,
      "mean_token_accuracy": 0.9255936145782471,
      "num_tokens": 75567252.0,
      "step": 6126
    },
    {
      "entropy": 1.177028775215149,
      "epoch": 3.2264349657714586,
      "grad_norm": 0.260392963886261,
      "learning_rate": 0.00012570751195045197,
      "loss": 0.1694093942642212,
      "mean_token_accuracy": 0.9297531694173813,
      "num_tokens": 75579588.0,
      "step": 6127
    },
    {
      "entropy": 1.2347355782985687,
      "epoch": 3.226961558715113,
      "grad_norm": 0.2754638195037842,
      "learning_rate": 0.00012568591974732477,
      "loss": 0.17848004400730133,
      "mean_token_accuracy": 0.925995871424675,
      "num_tokens": 75591924.0,
      "step": 6128
    },
    {
      "entropy": 1.202033132314682,
      "epoch": 3.227488151658768,
      "grad_norm": 0.2989853322505951,
      "learning_rate": 0.00012566432661301598,
      "loss": 0.17350336909294128,
      "mean_token_accuracy": 0.9255778640508652,
      "num_tokens": 75604260.0,
      "step": 6129
    },
    {
      "entropy": 1.1826526522636414,
      "epoch": 3.2280147446024223,
      "grad_norm": 0.29631954431533813,
      "learning_rate": 0.0001256427325488074,
      "loss": 0.18491564691066742,
      "mean_token_accuracy": 0.92462258040905,
      "num_tokens": 75616596.0,
      "step": 6130
    },
    {
      "entropy": 1.212283730506897,
      "epoch": 3.2285413375460768,
      "grad_norm": 0.2688755691051483,
      "learning_rate": 0.00012562113755598095,
      "loss": 0.1655760407447815,
      "mean_token_accuracy": 0.9282697588205338,
      "num_tokens": 75628932.0,
      "step": 6131
    },
    {
      "entropy": 1.2409653961658478,
      "epoch": 3.2290679304897316,
      "grad_norm": 0.2876735329627991,
      "learning_rate": 0.00012559954163581866,
      "loss": 0.17220713198184967,
      "mean_token_accuracy": 0.9249308109283447,
      "num_tokens": 75641268.0,
      "step": 6132
    },
    {
      "entropy": 1.2126244008541107,
      "epoch": 3.229594523433386,
      "grad_norm": 0.3038305640220642,
      "learning_rate": 0.00012557794478960246,
      "loss": 0.1925899088382721,
      "mean_token_accuracy": 0.9231642782688141,
      "num_tokens": 75653604.0,
      "step": 6133
    },
    {
      "entropy": 1.2351355254650116,
      "epoch": 3.2301211163770405,
      "grad_norm": 0.32141029834747314,
      "learning_rate": 0.00012555634701861448,
      "loss": 0.19204822182655334,
      "mean_token_accuracy": 0.920773446559906,
      "num_tokens": 75665940.0,
      "step": 6134
    },
    {
      "entropy": 1.183781385421753,
      "epoch": 3.230647709320695,
      "grad_norm": 0.28224945068359375,
      "learning_rate": 0.00012553474832413685,
      "loss": 0.188804030418396,
      "mean_token_accuracy": 0.9181794226169586,
      "num_tokens": 75678276.0,
      "step": 6135
    },
    {
      "entropy": 1.2339649200439453,
      "epoch": 3.2311743022643498,
      "grad_norm": 0.26924869418144226,
      "learning_rate": 0.00012551314870745174,
      "loss": 0.16667309403419495,
      "mean_token_accuracy": 0.9302802532911301,
      "num_tokens": 75690612.0,
      "step": 6136
    },
    {
      "entropy": 1.2304637134075165,
      "epoch": 3.231700895208004,
      "grad_norm": 0.3316200375556946,
      "learning_rate": 0.00012549154816984142,
      "loss": 0.1978515088558197,
      "mean_token_accuracy": 0.9189146757125854,
      "num_tokens": 75702948.0,
      "step": 6137
    },
    {
      "entropy": 1.2245360910892487,
      "epoch": 3.2322274881516586,
      "grad_norm": 0.29681289196014404,
      "learning_rate": 0.0001254699467125882,
      "loss": 0.16820311546325684,
      "mean_token_accuracy": 0.9299140870571136,
      "num_tokens": 75715284.0,
      "step": 6138
    },
    {
      "entropy": 1.1647963225841522,
      "epoch": 3.2327540810953135,
      "grad_norm": 0.2810593545436859,
      "learning_rate": 0.0001254483443369744,
      "loss": 0.17601312696933746,
      "mean_token_accuracy": 0.9255238920450211,
      "num_tokens": 75727620.0,
      "step": 6139
    },
    {
      "entropy": 1.2284044921398163,
      "epoch": 3.233280674038968,
      "grad_norm": 0.3164808750152588,
      "learning_rate": 0.00012542674104428243,
      "loss": 0.19982348382472992,
      "mean_token_accuracy": 0.9159713089466095,
      "num_tokens": 75739956.0,
      "step": 6140
    },
    {
      "entropy": 1.2567321062088013,
      "epoch": 3.2338072669826223,
      "grad_norm": 0.3081972897052765,
      "learning_rate": 0.00012540513683579474,
      "loss": 0.19968608021736145,
      "mean_token_accuracy": 0.9218475669622421,
      "num_tokens": 75752292.0,
      "step": 6141
    },
    {
      "entropy": 1.2771920561790466,
      "epoch": 3.234333859926277,
      "grad_norm": 0.30709677934646606,
      "learning_rate": 0.00012538353171279387,
      "loss": 0.19182801246643066,
      "mean_token_accuracy": 0.9227764755487442,
      "num_tokens": 75764628.0,
      "step": 6142
    },
    {
      "entropy": 1.17875474691391,
      "epoch": 3.2348604528699316,
      "grad_norm": 0.28673815727233887,
      "learning_rate": 0.0001253619256765624,
      "loss": 0.18364201486110687,
      "mean_token_accuracy": 0.9248145222663879,
      "num_tokens": 75776964.0,
      "step": 6143
    },
    {
      "entropy": 1.1919936537742615,
      "epoch": 3.235387045813586,
      "grad_norm": 0.2708572447299957,
      "learning_rate": 0.00012534031872838292,
      "loss": 0.164185032248497,
      "mean_token_accuracy": 0.9241882562637329,
      "num_tokens": 75789300.0,
      "step": 6144
    },
    {
      "entropy": 1.2565065324306488,
      "epoch": 3.2359136387572405,
      "grad_norm": 0.32703325152397156,
      "learning_rate": 0.00012531871086953814,
      "loss": 0.18734511733055115,
      "mean_token_accuracy": 0.9243505895137787,
      "num_tokens": 75801636.0,
      "step": 6145
    },
    {
      "entropy": 1.1935489773750305,
      "epoch": 3.2364402317008953,
      "grad_norm": 0.29197418689727783,
      "learning_rate": 0.0001252971021013108,
      "loss": 0.18213504552841187,
      "mean_token_accuracy": 0.9209382683038712,
      "num_tokens": 75813972.0,
      "step": 6146
    },
    {
      "entropy": 1.2141109108924866,
      "epoch": 3.2369668246445498,
      "grad_norm": 0.28175437450408936,
      "learning_rate": 0.00012527549242498365,
      "loss": 0.16677020490169525,
      "mean_token_accuracy": 0.9321560859680176,
      "num_tokens": 75826308.0,
      "step": 6147
    },
    {
      "entropy": 1.2104218900203705,
      "epoch": 3.237493417588204,
      "grad_norm": 0.30901896953582764,
      "learning_rate": 0.00012525388184183952,
      "loss": 0.18270671367645264,
      "mean_token_accuracy": 0.9228143841028214,
      "num_tokens": 75838644.0,
      "step": 6148
    },
    {
      "entropy": 1.2070663571357727,
      "epoch": 3.238020010531859,
      "grad_norm": 0.31356850266456604,
      "learning_rate": 0.00012523227035316135,
      "loss": 0.16722548007965088,
      "mean_token_accuracy": 0.9297654330730438,
      "num_tokens": 75850980.0,
      "step": 6149
    },
    {
      "entropy": 1.2063406705856323,
      "epoch": 3.2385466034755135,
      "grad_norm": 0.29572710394859314,
      "learning_rate": 0.0001252106579602321,
      "loss": 0.1823430061340332,
      "mean_token_accuracy": 0.9221715778112411,
      "num_tokens": 75863316.0,
      "step": 6150
    },
    {
      "entropy": 1.202089548110962,
      "epoch": 3.239073196419168,
      "grad_norm": 0.41573962569236755,
      "learning_rate": 0.0001251890446643347,
      "loss": 0.18167167901992798,
      "mean_token_accuracy": 0.9231297075748444,
      "num_tokens": 75875652.0,
      "step": 6151
    },
    {
      "entropy": 1.2254632711410522,
      "epoch": 3.2395997893628223,
      "grad_norm": 0.30677860975265503,
      "learning_rate": 0.00012516743046675228,
      "loss": 0.17912791669368744,
      "mean_token_accuracy": 0.9259785264730453,
      "num_tokens": 75887988.0,
      "step": 6152
    },
    {
      "entropy": 1.2403289377689362,
      "epoch": 3.240126382306477,
      "grad_norm": 0.28188419342041016,
      "learning_rate": 0.00012514581536876787,
      "loss": 0.1662982851266861,
      "mean_token_accuracy": 0.9265069961547852,
      "num_tokens": 75900324.0,
      "step": 6153
    },
    {
      "entropy": 1.214851200580597,
      "epoch": 3.2406529752501316,
      "grad_norm": 0.29193687438964844,
      "learning_rate": 0.00012512419937166474,
      "loss": 0.18590956926345825,
      "mean_token_accuracy": 0.9184418022632599,
      "num_tokens": 75912660.0,
      "step": 6154
    },
    {
      "entropy": 1.206022471189499,
      "epoch": 3.241179568193786,
      "grad_norm": 0.2955065369606018,
      "learning_rate": 0.00012510258247672596,
      "loss": 0.1851632297039032,
      "mean_token_accuracy": 0.9227798581123352,
      "num_tokens": 75924996.0,
      "step": 6155
    },
    {
      "entropy": 1.2169673144817352,
      "epoch": 3.241706161137441,
      "grad_norm": 0.3072492778301239,
      "learning_rate": 0.00012508096468523498,
      "loss": 0.19872868061065674,
      "mean_token_accuracy": 0.9143420159816742,
      "num_tokens": 75937332.0,
      "step": 6156
    },
    {
      "entropy": 1.2228606343269348,
      "epoch": 3.2422327540810953,
      "grad_norm": 0.2958139479160309,
      "learning_rate": 0.00012505934599847497,
      "loss": 0.1759375035762787,
      "mean_token_accuracy": 0.9271186739206314,
      "num_tokens": 75949668.0,
      "step": 6157
    },
    {
      "entropy": 1.2175047397613525,
      "epoch": 3.2427593470247498,
      "grad_norm": 0.2919408679008484,
      "learning_rate": 0.00012503772641772934,
      "loss": 0.1769973337650299,
      "mean_token_accuracy": 0.9262620210647583,
      "num_tokens": 75962004.0,
      "step": 6158
    },
    {
      "entropy": 1.2171292901039124,
      "epoch": 3.2432859399684046,
      "grad_norm": 0.275551438331604,
      "learning_rate": 0.00012501610594428157,
      "loss": 0.1847660392522812,
      "mean_token_accuracy": 0.9262053966522217,
      "num_tokens": 75974340.0,
      "step": 6159
    },
    {
      "entropy": 1.234005868434906,
      "epoch": 3.243812532912059,
      "grad_norm": 0.3129563331604004,
      "learning_rate": 0.0001249944845794151,
      "loss": 0.19297876954078674,
      "mean_token_accuracy": 0.9163733869791031,
      "num_tokens": 75986676.0,
      "step": 6160
    },
    {
      "entropy": 1.2253689765930176,
      "epoch": 3.2443391258557135,
      "grad_norm": 0.28347060084342957,
      "learning_rate": 0.00012497286232441348,
      "loss": 0.17327147722244263,
      "mean_token_accuracy": 0.928035169839859,
      "num_tokens": 75999012.0,
      "step": 6161
    },
    {
      "entropy": 1.2434102296829224,
      "epoch": 3.244865718799368,
      "grad_norm": 0.2920335829257965,
      "learning_rate": 0.0001249512391805603,
      "loss": 0.2005065679550171,
      "mean_token_accuracy": 0.9157353788614273,
      "num_tokens": 76011348.0,
      "step": 6162
    },
    {
      "entropy": 1.302938461303711,
      "epoch": 3.245392311743023,
      "grad_norm": 0.2882930040359497,
      "learning_rate": 0.0001249296151491392,
      "loss": 0.1890123039484024,
      "mean_token_accuracy": 0.9203406572341919,
      "num_tokens": 76023684.0,
      "step": 6163
    },
    {
      "entropy": 1.2709252536296844,
      "epoch": 3.245918904686677,
      "grad_norm": 0.3011837899684906,
      "learning_rate": 0.00012490799023143385,
      "loss": 0.19571897387504578,
      "mean_token_accuracy": 0.9177806377410889,
      "num_tokens": 76036020.0,
      "step": 6164
    },
    {
      "entropy": 1.2109342217445374,
      "epoch": 3.2464454976303316,
      "grad_norm": 0.2555895149707794,
      "learning_rate": 0.000124886364428728,
      "loss": 0.16856005787849426,
      "mean_token_accuracy": 0.9285546839237213,
      "num_tokens": 76048356.0,
      "step": 6165
    },
    {
      "entropy": 1.2866427302360535,
      "epoch": 3.2469720905739865,
      "grad_norm": 0.31408652663230896,
      "learning_rate": 0.00012486473774230548,
      "loss": 0.20854584872722626,
      "mean_token_accuracy": 0.9128608703613281,
      "num_tokens": 76060692.0,
      "step": 6166
    },
    {
      "entropy": 1.2667243480682373,
      "epoch": 3.247498683517641,
      "grad_norm": 0.27591609954833984,
      "learning_rate": 0.0001248431101734501,
      "loss": 0.18142713606357574,
      "mean_token_accuracy": 0.922798290848732,
      "num_tokens": 76073028.0,
      "step": 6167
    },
    {
      "entropy": 1.2338934242725372,
      "epoch": 3.2480252764612954,
      "grad_norm": 0.2758982181549072,
      "learning_rate": 0.0001248214817234458,
      "loss": 0.18854719400405884,
      "mean_token_accuracy": 0.9161380380392075,
      "num_tokens": 76085364.0,
      "step": 6168
    },
    {
      "entropy": 1.2804130017757416,
      "epoch": 3.2485518694049498,
      "grad_norm": 0.29587045311927795,
      "learning_rate": 0.0001247998523935765,
      "loss": 0.18399131298065186,
      "mean_token_accuracy": 0.9224258214235306,
      "num_tokens": 76097700.0,
      "step": 6169
    },
    {
      "entropy": 1.3210213482379913,
      "epoch": 3.2490784623486046,
      "grad_norm": 0.3032381236553192,
      "learning_rate": 0.00012477822218512623,
      "loss": 0.186031311750412,
      "mean_token_accuracy": 0.921225294470787,
      "num_tokens": 76110036.0,
      "step": 6170
    },
    {
      "entropy": 1.2946960628032684,
      "epoch": 3.249605055292259,
      "grad_norm": 0.2837267220020294,
      "learning_rate": 0.00012475659109937906,
      "loss": 0.17635822296142578,
      "mean_token_accuracy": 0.9275114834308624,
      "num_tokens": 76122372.0,
      "step": 6171
    },
    {
      "entropy": 1.3051108717918396,
      "epoch": 3.2501316482359135,
      "grad_norm": 0.27409878373146057,
      "learning_rate": 0.00012473495913761906,
      "loss": 0.18140482902526855,
      "mean_token_accuracy": 0.922147199511528,
      "num_tokens": 76134708.0,
      "step": 6172
    },
    {
      "entropy": 1.2981198728084564,
      "epoch": 3.2506582411795684,
      "grad_norm": 0.27573296427726746,
      "learning_rate": 0.0001247133263011304,
      "loss": 0.173725426197052,
      "mean_token_accuracy": 0.9266543388366699,
      "num_tokens": 76147044.0,
      "step": 6173
    },
    {
      "entropy": 1.3327777683734894,
      "epoch": 3.251184834123223,
      "grad_norm": 0.29280543327331543,
      "learning_rate": 0.00012469169259119735,
      "loss": 0.17917679250240326,
      "mean_token_accuracy": 0.9230498522520065,
      "num_tokens": 76159380.0,
      "step": 6174
    },
    {
      "entropy": 1.273023247718811,
      "epoch": 3.251711427066877,
      "grad_norm": 0.2697977125644684,
      "learning_rate": 0.00012467005800910412,
      "loss": 0.17984896898269653,
      "mean_token_accuracy": 0.9266160875558853,
      "num_tokens": 76171716.0,
      "step": 6175
    },
    {
      "entropy": 1.3279211521148682,
      "epoch": 3.252238020010532,
      "grad_norm": 0.30282941460609436,
      "learning_rate": 0.0001246484225561351,
      "loss": 0.18830984830856323,
      "mean_token_accuracy": 0.9210638850927353,
      "num_tokens": 76184052.0,
      "step": 6176
    },
    {
      "entropy": 1.3487639725208282,
      "epoch": 3.2527646129541865,
      "grad_norm": 0.31188657879829407,
      "learning_rate": 0.00012462678623357455,
      "loss": 0.20160964131355286,
      "mean_token_accuracy": 0.9153773486614227,
      "num_tokens": 76196388.0,
      "step": 6177
    },
    {
      "entropy": 1.3300007581710815,
      "epoch": 3.253291205897841,
      "grad_norm": 0.2827070951461792,
      "learning_rate": 0.00012460514904270696,
      "loss": 0.19206058979034424,
      "mean_token_accuracy": 0.920170783996582,
      "num_tokens": 76208724.0,
      "step": 6178
    },
    {
      "entropy": 1.3335220217704773,
      "epoch": 3.2538177988414954,
      "grad_norm": 0.3128904104232788,
      "learning_rate": 0.0001245835109848168,
      "loss": 0.19153635203838348,
      "mean_token_accuracy": 0.919683575630188,
      "num_tokens": 76221060.0,
      "step": 6179
    },
    {
      "entropy": 1.329508662223816,
      "epoch": 3.2543443917851502,
      "grad_norm": 0.294729620218277,
      "learning_rate": 0.0001245618720611886,
      "loss": 0.17604173719882965,
      "mean_token_accuracy": 0.9236066043376923,
      "num_tokens": 76233396.0,
      "step": 6180
    },
    {
      "entropy": 1.3312866687774658,
      "epoch": 3.2548709847288047,
      "grad_norm": 0.3070805072784424,
      "learning_rate": 0.00012454023227310694,
      "loss": 0.19934087991714478,
      "mean_token_accuracy": 0.9166095405817032,
      "num_tokens": 76245732.0,
      "step": 6181
    },
    {
      "entropy": 1.2477159202098846,
      "epoch": 3.255397577672459,
      "grad_norm": 0.2840286195278168,
      "learning_rate": 0.00012451859162185643,
      "loss": 0.18146276473999023,
      "mean_token_accuracy": 0.9268984943628311,
      "num_tokens": 76258068.0,
      "step": 6182
    },
    {
      "entropy": 1.2960457801818848,
      "epoch": 3.2559241706161135,
      "grad_norm": 0.29388269782066345,
      "learning_rate": 0.00012449695010872177,
      "loss": 0.18993444740772247,
      "mean_token_accuracy": 0.9193546026945114,
      "num_tokens": 76270404.0,
      "step": 6183
    },
    {
      "entropy": 1.259146362543106,
      "epoch": 3.2564507635597684,
      "grad_norm": 0.26432740688323975,
      "learning_rate": 0.00012447530773498764,
      "loss": 0.17386454343795776,
      "mean_token_accuracy": 0.9284330308437347,
      "num_tokens": 76282740.0,
      "step": 6184
    },
    {
      "entropy": 1.2615996301174164,
      "epoch": 3.256977356503423,
      "grad_norm": 0.29614487290382385,
      "learning_rate": 0.00012445366450193886,
      "loss": 0.16952276229858398,
      "mean_token_accuracy": 0.9308066666126251,
      "num_tokens": 76295076.0,
      "step": 6185
    },
    {
      "entropy": 1.3251206576824188,
      "epoch": 3.257503949447077,
      "grad_norm": 0.3130854666233063,
      "learning_rate": 0.0001244320204108603,
      "loss": 0.1777714341878891,
      "mean_token_accuracy": 0.9288688153028488,
      "num_tokens": 76307412.0,
      "step": 6186
    },
    {
      "entropy": 1.2637268602848053,
      "epoch": 3.258030542390732,
      "grad_norm": 0.298968106508255,
      "learning_rate": 0.00012441037546303676,
      "loss": 0.1983793079853058,
      "mean_token_accuracy": 0.9180431962013245,
      "num_tokens": 76319748.0,
      "step": 6187
    },
    {
      "entropy": 1.2078056633472443,
      "epoch": 3.2585571353343865,
      "grad_norm": 0.2688981294631958,
      "learning_rate": 0.00012438872965975326,
      "loss": 0.17481450736522675,
      "mean_token_accuracy": 0.9246529191732407,
      "num_tokens": 76332084.0,
      "step": 6188
    },
    {
      "entropy": 1.2519322335720062,
      "epoch": 3.259083728278041,
      "grad_norm": 0.29199621081352234,
      "learning_rate": 0.00012436708300229472,
      "loss": 0.1926528662443161,
      "mean_token_accuracy": 0.9195021241903305,
      "num_tokens": 76344420.0,
      "step": 6189
    },
    {
      "entropy": 1.2125513851642609,
      "epoch": 3.259610321221696,
      "grad_norm": 0.2778526544570923,
      "learning_rate": 0.0001243454354919462,
      "loss": 0.16762031614780426,
      "mean_token_accuracy": 0.9271342903375626,
      "num_tokens": 76356756.0,
      "step": 6190
    },
    {
      "entropy": 1.2814996242523193,
      "epoch": 3.2601369141653502,
      "grad_norm": 0.30158013105392456,
      "learning_rate": 0.0001243237871299928,
      "loss": 0.16893842816352844,
      "mean_token_accuracy": 0.9282845556735992,
      "num_tokens": 76369092.0,
      "step": 6191
    },
    {
      "entropy": 1.2342957556247711,
      "epoch": 3.2606635071090047,
      "grad_norm": 0.28979063034057617,
      "learning_rate": 0.00012430213791771963,
      "loss": 0.17779289186000824,
      "mean_token_accuracy": 0.9218163043260574,
      "num_tokens": 76381428.0,
      "step": 6192
    },
    {
      "entropy": 1.2822068333625793,
      "epoch": 3.2611901000526595,
      "grad_norm": 0.327267050743103,
      "learning_rate": 0.0001242804878564119,
      "loss": 0.19853328168392181,
      "mean_token_accuracy": 0.9153106510639191,
      "num_tokens": 76393764.0,
      "step": 6193
    },
    {
      "entropy": 1.2986519038677216,
      "epoch": 3.261716692996314,
      "grad_norm": 0.3420284688472748,
      "learning_rate": 0.00012425883694735482,
      "loss": 0.1888749599456787,
      "mean_token_accuracy": 0.9159906208515167,
      "num_tokens": 76406100.0,
      "step": 6194
    },
    {
      "entropy": 1.2821783423423767,
      "epoch": 3.2622432859399684,
      "grad_norm": 0.30797383189201355,
      "learning_rate": 0.0001242371851918337,
      "loss": 0.18895478546619415,
      "mean_token_accuracy": 0.9191595017910004,
      "num_tokens": 76418436.0,
      "step": 6195
    },
    {
      "entropy": 1.2532289028167725,
      "epoch": 3.262769878883623,
      "grad_norm": 0.3051166832447052,
      "learning_rate": 0.00012421553259113393,
      "loss": 0.17775626480579376,
      "mean_token_accuracy": 0.9237864166498184,
      "num_tokens": 76430772.0,
      "step": 6196
    },
    {
      "entropy": 1.3062248826026917,
      "epoch": 3.2632964718272777,
      "grad_norm": 0.32124513387680054,
      "learning_rate": 0.00012419387914654076,
      "loss": 0.18037666380405426,
      "mean_token_accuracy": 0.9269574880599976,
      "num_tokens": 76443108.0,
      "step": 6197
    },
    {
      "entropy": 1.2437196373939514,
      "epoch": 3.263823064770932,
      "grad_norm": 0.2947427034378052,
      "learning_rate": 0.00012417222485933978,
      "loss": 0.18406802415847778,
      "mean_token_accuracy": 0.9227783232927322,
      "num_tokens": 76455444.0,
      "step": 6198
    },
    {
      "entropy": 1.2761265933513641,
      "epoch": 3.2643496577145865,
      "grad_norm": 0.2777099609375,
      "learning_rate": 0.00012415056973081638,
      "loss": 0.1712026447057724,
      "mean_token_accuracy": 0.9295600354671478,
      "num_tokens": 76467780.0,
      "step": 6199
    },
    {
      "entropy": 1.3158571124076843,
      "epoch": 3.264876250658241,
      "grad_norm": 0.2971872389316559,
      "learning_rate": 0.00012412891376225612,
      "loss": 0.17403270304203033,
      "mean_token_accuracy": 0.9262079894542694,
      "num_tokens": 76480116.0,
      "step": 6200
    },
    {
      "entropy": 1.2905315160751343,
      "epoch": 3.265402843601896,
      "grad_norm": 0.28583183884620667,
      "learning_rate": 0.00012410725695494464,
      "loss": 0.16530053317546844,
      "mean_token_accuracy": 0.929039940237999,
      "num_tokens": 76492452.0,
      "step": 6201
    },
    {
      "entropy": 1.252262145280838,
      "epoch": 3.2659294365455502,
      "grad_norm": 0.3061239719390869,
      "learning_rate": 0.00012408559931016753,
      "loss": 0.18879777193069458,
      "mean_token_accuracy": 0.9231210649013519,
      "num_tokens": 76504788.0,
      "step": 6202
    },
    {
      "entropy": 1.2744266092777252,
      "epoch": 3.2664560294892047,
      "grad_norm": 0.2914201319217682,
      "learning_rate": 0.00012406394082921046,
      "loss": 0.17487694323062897,
      "mean_token_accuracy": 0.9287125766277313,
      "num_tokens": 76517124.0,
      "step": 6203
    },
    {
      "entropy": 1.2931773960590363,
      "epoch": 3.2669826224328595,
      "grad_norm": 0.2819075882434845,
      "learning_rate": 0.0001240422815133592,
      "loss": 0.16330452263355255,
      "mean_token_accuracy": 0.9317172169685364,
      "num_tokens": 76529460.0,
      "step": 6204
    },
    {
      "entropy": 1.2473205626010895,
      "epoch": 3.267509215376514,
      "grad_norm": 0.3001861274242401,
      "learning_rate": 0.00012402062136389954,
      "loss": 0.18457162380218506,
      "mean_token_accuracy": 0.9188896268606186,
      "num_tokens": 76541796.0,
      "step": 6205
    },
    {
      "entropy": 1.2851862609386444,
      "epoch": 3.2680358083201684,
      "grad_norm": 0.29719528555870056,
      "learning_rate": 0.0001239989603821173,
      "loss": 0.17140623927116394,
      "mean_token_accuracy": 0.9282460957765579,
      "num_tokens": 76554132.0,
      "step": 6206
    },
    {
      "entropy": 1.2846690118312836,
      "epoch": 3.2685624012638232,
      "grad_norm": 0.2931407392024994,
      "learning_rate": 0.00012397729856929835,
      "loss": 0.17985954880714417,
      "mean_token_accuracy": 0.9300493150949478,
      "num_tokens": 76566468.0,
      "step": 6207
    },
    {
      "entropy": 1.3359651863574982,
      "epoch": 3.2690889942074777,
      "grad_norm": 0.2905985713005066,
      "learning_rate": 0.0001239556359267287,
      "loss": 0.1789180040359497,
      "mean_token_accuracy": 0.9208443909883499,
      "num_tokens": 76578804.0,
      "step": 6208
    },
    {
      "entropy": 1.2525035738945007,
      "epoch": 3.269615587151132,
      "grad_norm": 0.281355082988739,
      "learning_rate": 0.00012393397245569427,
      "loss": 0.18138766288757324,
      "mean_token_accuracy": 0.9221275150775909,
      "num_tokens": 76591140.0,
      "step": 6209
    },
    {
      "entropy": 1.3100119233131409,
      "epoch": 3.270142180094787,
      "grad_norm": 0.270400732755661,
      "learning_rate": 0.00012391230815748112,
      "loss": 0.17373928427696228,
      "mean_token_accuracy": 0.9232037961483002,
      "num_tokens": 76603476.0,
      "step": 6210
    },
    {
      "entropy": 1.3040125966072083,
      "epoch": 3.2706687730384414,
      "grad_norm": 0.29999953508377075,
      "learning_rate": 0.00012389064303337528,
      "loss": 0.1941448450088501,
      "mean_token_accuracy": 0.9185135662555695,
      "num_tokens": 76615812.0,
      "step": 6211
    },
    {
      "entropy": 1.317997932434082,
      "epoch": 3.271195365982096,
      "grad_norm": 0.28788891434669495,
      "learning_rate": 0.00012386897708466295,
      "loss": 0.18096309900283813,
      "mean_token_accuracy": 0.9225906133651733,
      "num_tokens": 76628148.0,
      "step": 6212
    },
    {
      "entropy": 1.3324589431285858,
      "epoch": 3.2717219589257502,
      "grad_norm": 0.308905690908432,
      "learning_rate": 0.00012384731031263028,
      "loss": 0.18385492265224457,
      "mean_token_accuracy": 0.9245429188013077,
      "num_tokens": 76640484.0,
      "step": 6213
    },
    {
      "entropy": 1.3233293890953064,
      "epoch": 3.272248551869405,
      "grad_norm": 0.2808932662010193,
      "learning_rate": 0.0001238256427185635,
      "loss": 0.16888180375099182,
      "mean_token_accuracy": 0.9283661246299744,
      "num_tokens": 76652820.0,
      "step": 6214
    },
    {
      "entropy": 1.2769432961940765,
      "epoch": 3.2727751448130595,
      "grad_norm": 0.2868754267692566,
      "learning_rate": 0.00012380397430374892,
      "loss": 0.18716515600681305,
      "mean_token_accuracy": 0.9205536544322968,
      "num_tokens": 76665156.0,
      "step": 6215
    },
    {
      "entropy": 1.3713759183883667,
      "epoch": 3.273301737756714,
      "grad_norm": 0.29639267921447754,
      "learning_rate": 0.00012378230506947285,
      "loss": 0.1748577207326889,
      "mean_token_accuracy": 0.9252920746803284,
      "num_tokens": 76677492.0,
      "step": 6216
    },
    {
      "entropy": 1.2873197197914124,
      "epoch": 3.2738283307003684,
      "grad_norm": 0.3188132345676422,
      "learning_rate": 0.0001237606350170216,
      "loss": 0.19027827680110931,
      "mean_token_accuracy": 0.9210074543952942,
      "num_tokens": 76689828.0,
      "step": 6217
    },
    {
      "entropy": 1.242147296667099,
      "epoch": 3.2743549236440233,
      "grad_norm": 0.2808919847011566,
      "learning_rate": 0.00012373896414768167,
      "loss": 0.15831927955150604,
      "mean_token_accuracy": 0.9348323941230774,
      "num_tokens": 76702164.0,
      "step": 6218
    },
    {
      "entropy": 1.3022652864456177,
      "epoch": 3.2748815165876777,
      "grad_norm": 0.2820264399051666,
      "learning_rate": 0.00012371729246273954,
      "loss": 0.18252812325954437,
      "mean_token_accuracy": 0.9279136508703232,
      "num_tokens": 76714500.0,
      "step": 6219
    },
    {
      "entropy": 1.258202999830246,
      "epoch": 3.275408109531332,
      "grad_norm": 0.29160061478614807,
      "learning_rate": 0.0001236956199634817,
      "loss": 0.17641672492027283,
      "mean_token_accuracy": 0.9247133880853653,
      "num_tokens": 76726836.0,
      "step": 6220
    },
    {
      "entropy": 1.2638160586357117,
      "epoch": 3.275934702474987,
      "grad_norm": 0.30546367168426514,
      "learning_rate": 0.0001236739466511947,
      "loss": 0.17519892752170563,
      "mean_token_accuracy": 0.9232414662837982,
      "num_tokens": 76739172.0,
      "step": 6221
    },
    {
      "entropy": 1.2811826467514038,
      "epoch": 3.2764612954186414,
      "grad_norm": 0.28423309326171875,
      "learning_rate": 0.00012365227252716522,
      "loss": 0.168607696890831,
      "mean_token_accuracy": 0.9301139712333679,
      "num_tokens": 76751508.0,
      "step": 6222
    },
    {
      "entropy": 1.3453866243362427,
      "epoch": 3.276987888362296,
      "grad_norm": 0.3236750066280365,
      "learning_rate": 0.00012363059759267988,
      "loss": 0.18853197991847992,
      "mean_token_accuracy": 0.9232541024684906,
      "num_tokens": 76763844.0,
      "step": 6223
    },
    {
      "entropy": 1.268451988697052,
      "epoch": 3.2775144813059507,
      "grad_norm": 0.2743200361728668,
      "learning_rate": 0.0001236089218490254,
      "loss": 0.15886418521404266,
      "mean_token_accuracy": 0.9334812015295029,
      "num_tokens": 76776180.0,
      "step": 6224
    },
    {
      "entropy": 1.3066802024841309,
      "epoch": 3.278041074249605,
      "grad_norm": 0.3363740146160126,
      "learning_rate": 0.00012358724529748857,
      "loss": 0.18869782984256744,
      "mean_token_accuracy": 0.9239563941955566,
      "num_tokens": 76788516.0,
      "step": 6225
    },
    {
      "entropy": 1.3019660413265228,
      "epoch": 3.2785676671932595,
      "grad_norm": 0.30076122283935547,
      "learning_rate": 0.00012356556793935615,
      "loss": 0.1893063187599182,
      "mean_token_accuracy": 0.9223721027374268,
      "num_tokens": 76800852.0,
      "step": 6226
    },
    {
      "entropy": 1.3066136240959167,
      "epoch": 3.2790942601369144,
      "grad_norm": 0.3011791706085205,
      "learning_rate": 0.00012354388977591505,
      "loss": 0.15764912962913513,
      "mean_token_accuracy": 0.9360844492912292,
      "num_tokens": 76813188.0,
      "step": 6227
    },
    {
      "entropy": 1.3428710401058197,
      "epoch": 3.279620853080569,
      "grad_norm": 0.30883345007896423,
      "learning_rate": 0.00012352221080845212,
      "loss": 0.18702161312103271,
      "mean_token_accuracy": 0.9179170429706573,
      "num_tokens": 76825524.0,
      "step": 6228
    },
    {
      "entropy": 1.2546886503696442,
      "epoch": 3.2801474460242233,
      "grad_norm": 0.31887173652648926,
      "learning_rate": 0.00012350053103825438,
      "loss": 0.18897868692874908,
      "mean_token_accuracy": 0.9202984124422073,
      "num_tokens": 76837860.0,
      "step": 6229
    },
    {
      "entropy": 1.2445046305656433,
      "epoch": 3.2806740389678777,
      "grad_norm": 0.27127400040626526,
      "learning_rate": 0.0001234788504666088,
      "loss": 0.17110732197761536,
      "mean_token_accuracy": 0.9267043769359589,
      "num_tokens": 76850196.0,
      "step": 6230
    },
    {
      "entropy": 1.2792969346046448,
      "epoch": 3.2812006319115326,
      "grad_norm": 0.29130396246910095,
      "learning_rate": 0.00012345716909480242,
      "loss": 0.1774936467409134,
      "mean_token_accuracy": 0.9278624057769775,
      "num_tokens": 76862532.0,
      "step": 6231
    },
    {
      "entropy": 1.284456580877304,
      "epoch": 3.281727224855187,
      "grad_norm": 0.30248746275901794,
      "learning_rate": 0.00012343548692412233,
      "loss": 0.1800181269645691,
      "mean_token_accuracy": 0.9215735793113708,
      "num_tokens": 76874868.0,
      "step": 6232
    },
    {
      "entropy": 1.2692930698394775,
      "epoch": 3.2822538177988414,
      "grad_norm": 0.2928861379623413,
      "learning_rate": 0.0001234138039558557,
      "loss": 0.19163967669010162,
      "mean_token_accuracy": 0.9179900139570236,
      "num_tokens": 76887204.0,
      "step": 6233
    },
    {
      "entropy": 1.247650921344757,
      "epoch": 3.282780410742496,
      "grad_norm": 0.31334158778190613,
      "learning_rate": 0.00012339212019128973,
      "loss": 0.19242893159389496,
      "mean_token_accuracy": 0.9193117916584015,
      "num_tokens": 76899540.0,
      "step": 6234
    },
    {
      "entropy": 1.2478911876678467,
      "epoch": 3.2833070036861507,
      "grad_norm": 0.3001970052719116,
      "learning_rate": 0.0001233704356317116,
      "loss": 0.17564845085144043,
      "mean_token_accuracy": 0.9316419661045074,
      "num_tokens": 76911876.0,
      "step": 6235
    },
    {
      "entropy": 1.3031432330608368,
      "epoch": 3.283833596629805,
      "grad_norm": 0.28892484307289124,
      "learning_rate": 0.0001233487502784087,
      "loss": 0.18837390840053558,
      "mean_token_accuracy": 0.9224282503128052,
      "num_tokens": 76924212.0,
      "step": 6236
    },
    {
      "entropy": 1.3214940428733826,
      "epoch": 3.2843601895734595,
      "grad_norm": 0.2903456687927246,
      "learning_rate": 0.00012332706413266823,
      "loss": 0.17962339520454407,
      "mean_token_accuracy": 0.9192012697458267,
      "num_tokens": 76936548.0,
      "step": 6237
    },
    {
      "entropy": 1.3401213586330414,
      "epoch": 3.2848867825171144,
      "grad_norm": 0.29672813415527344,
      "learning_rate": 0.00012330537719577766,
      "loss": 0.18679866194725037,
      "mean_token_accuracy": 0.9169516563415527,
      "num_tokens": 76948884.0,
      "step": 6238
    },
    {
      "entropy": 1.3582242131233215,
      "epoch": 3.285413375460769,
      "grad_norm": 0.2805413007736206,
      "learning_rate": 0.0001232836894690244,
      "loss": 0.18263594806194305,
      "mean_token_accuracy": 0.9259663075208664,
      "num_tokens": 76961220.0,
      "step": 6239
    },
    {
      "entropy": 1.2710779011249542,
      "epoch": 3.2859399684044233,
      "grad_norm": 0.2695918083190918,
      "learning_rate": 0.00012326200095369593,
      "loss": 0.178550124168396,
      "mean_token_accuracy": 0.9272159487009048,
      "num_tokens": 76973556.0,
      "step": 6240
    },
    {
      "entropy": 1.348080426454544,
      "epoch": 3.286466561348078,
      "grad_norm": 0.308569073677063,
      "learning_rate": 0.00012324031165107977,
      "loss": 0.1801525056362152,
      "mean_token_accuracy": 0.924568384885788,
      "num_tokens": 76985892.0,
      "step": 6241
    },
    {
      "entropy": 1.2940427362918854,
      "epoch": 3.2869931542917326,
      "grad_norm": 0.28971853852272034,
      "learning_rate": 0.00012321862156246342,
      "loss": 0.1751011162996292,
      "mean_token_accuracy": 0.9240054190158844,
      "num_tokens": 76998228.0,
      "step": 6242
    },
    {
      "entropy": 1.3105450570583344,
      "epoch": 3.287519747235387,
      "grad_norm": 0.2823137640953064,
      "learning_rate": 0.0001231969306891346,
      "loss": 0.18045909702777863,
      "mean_token_accuracy": 0.926315575838089,
      "num_tokens": 77010564.0,
      "step": 6243
    },
    {
      "entropy": 1.3067985475063324,
      "epoch": 3.288046340179042,
      "grad_norm": 0.30029135942459106,
      "learning_rate": 0.00012317523903238094,
      "loss": 0.18511660397052765,
      "mean_token_accuracy": 0.9207724779844284,
      "num_tokens": 77022900.0,
      "step": 6244
    },
    {
      "entropy": 1.333036631345749,
      "epoch": 3.2885729331226963,
      "grad_norm": 0.30139487981796265,
      "learning_rate": 0.0001231535465934901,
      "loss": 0.1940365582704544,
      "mean_token_accuracy": 0.9214530885219574,
      "num_tokens": 77035236.0,
      "step": 6245
    },
    {
      "entropy": 1.3007804155349731,
      "epoch": 3.2890995260663507,
      "grad_norm": 0.26908811926841736,
      "learning_rate": 0.00012313185337374985,
      "loss": 0.17049038410186768,
      "mean_token_accuracy": 0.9276867210865021,
      "num_tokens": 77047572.0,
      "step": 6246
    },
    {
      "entropy": 1.3777782022953033,
      "epoch": 3.289626119010005,
      "grad_norm": 0.31890666484832764,
      "learning_rate": 0.000123110159374448,
      "loss": 0.19547241926193237,
      "mean_token_accuracy": 0.9192814230918884,
      "num_tokens": 77059908.0,
      "step": 6247
    },
    {
      "entropy": 1.3698776066303253,
      "epoch": 3.29015271195366,
      "grad_norm": 0.30019745230674744,
      "learning_rate": 0.00012308846459687243,
      "loss": 0.17105017602443695,
      "mean_token_accuracy": 0.9322203248739243,
      "num_tokens": 77072244.0,
      "step": 6248
    },
    {
      "entropy": 1.2837482690811157,
      "epoch": 3.2906793048973144,
      "grad_norm": 0.28274422883987427,
      "learning_rate": 0.000123066769042311,
      "loss": 0.17917296290397644,
      "mean_token_accuracy": 0.9279322028160095,
      "num_tokens": 77084580.0,
      "step": 6249
    },
    {
      "entropy": 1.348880112171173,
      "epoch": 3.291205897840969,
      "grad_norm": 0.3124294877052307,
      "learning_rate": 0.00012304507271205167,
      "loss": 0.18216298520565033,
      "mean_token_accuracy": 0.9279361516237259,
      "num_tokens": 77096916.0,
      "step": 6250
    },
    {
      "entropy": 1.328457921743393,
      "epoch": 3.2917324907846233,
      "grad_norm": 0.29864731431007385,
      "learning_rate": 0.00012302337560738235,
      "loss": 0.19336733222007751,
      "mean_token_accuracy": 0.9219559580087662,
      "num_tokens": 77109252.0,
      "step": 6251
    },
    {
      "entropy": 1.308976262807846,
      "epoch": 3.292259083728278,
      "grad_norm": 0.3369043469429016,
      "learning_rate": 0.00012300167772959114,
      "loss": 0.20705236494541168,
      "mean_token_accuracy": 0.9159487783908844,
      "num_tokens": 77121588.0,
      "step": 6252
    },
    {
      "entropy": 1.3961984515190125,
      "epoch": 3.2927856766719326,
      "grad_norm": 0.2894037067890167,
      "learning_rate": 0.0001229799790799661,
      "loss": 0.1619293987751007,
      "mean_token_accuracy": 0.9350053817033768,
      "num_tokens": 77133924.0,
      "step": 6253
    },
    {
      "entropy": 1.3159714043140411,
      "epoch": 3.293312269615587,
      "grad_norm": 0.26713791489601135,
      "learning_rate": 0.0001229582796597954,
      "loss": 0.1694222390651703,
      "mean_token_accuracy": 0.9325111657381058,
      "num_tokens": 77146260.0,
      "step": 6254
    },
    {
      "entropy": 1.3922246992588043,
      "epoch": 3.293838862559242,
      "grad_norm": 0.2928251326084137,
      "learning_rate": 0.00012293657947036715,
      "loss": 0.17767813801765442,
      "mean_token_accuracy": 0.9284683912992477,
      "num_tokens": 77158596.0,
      "step": 6255
    },
    {
      "entropy": 1.3610136806964874,
      "epoch": 3.2943654555028963,
      "grad_norm": 0.277475506067276,
      "learning_rate": 0.00012291487851296955,
      "loss": 0.17403161525726318,
      "mean_token_accuracy": 0.9285924583673477,
      "num_tokens": 77170932.0,
      "step": 6256
    },
    {
      "entropy": 1.3568114638328552,
      "epoch": 3.2948920484465507,
      "grad_norm": 0.31263595819473267,
      "learning_rate": 0.0001228931767888909,
      "loss": 0.1871493011713028,
      "mean_token_accuracy": 0.925555557012558,
      "num_tokens": 77183268.0,
      "step": 6257
    },
    {
      "entropy": 1.3260321021080017,
      "epoch": 3.2954186413902056,
      "grad_norm": 0.2852775752544403,
      "learning_rate": 0.0001228714742994195,
      "loss": 0.16670246422290802,
      "mean_token_accuracy": 0.9286938607692719,
      "num_tokens": 77195604.0,
      "step": 6258
    },
    {
      "entropy": 1.3797948062419891,
      "epoch": 3.29594523433386,
      "grad_norm": 0.30797091126441956,
      "learning_rate": 0.00012284977104584367,
      "loss": 0.1951187551021576,
      "mean_token_accuracy": 0.9217554777860641,
      "num_tokens": 77207940.0,
      "step": 6259
    },
    {
      "entropy": 1.3536570966243744,
      "epoch": 3.2964718272775144,
      "grad_norm": 0.2918627858161926,
      "learning_rate": 0.00012282806702945185,
      "loss": 0.18449142575263977,
      "mean_token_accuracy": 0.9261642545461655,
      "num_tokens": 77220276.0,
      "step": 6260
    },
    {
      "entropy": 1.3510786592960358,
      "epoch": 3.296998420221169,
      "grad_norm": 0.2922392785549164,
      "learning_rate": 0.00012280636225153244,
      "loss": 0.1884160339832306,
      "mean_token_accuracy": 0.9224065691232681,
      "num_tokens": 77232612.0,
      "step": 6261
    },
    {
      "entropy": 1.3692474663257599,
      "epoch": 3.2975250131648237,
      "grad_norm": 0.3316969871520996,
      "learning_rate": 0.00012278465671337394,
      "loss": 0.20385709404945374,
      "mean_token_accuracy": 0.9145285934209824,
      "num_tokens": 77244948.0,
      "step": 6262
    },
    {
      "entropy": 1.3025420308113098,
      "epoch": 3.298051606108478,
      "grad_norm": 0.2928790748119354,
      "learning_rate": 0.00012276295041626492,
      "loss": 0.17966334521770477,
      "mean_token_accuracy": 0.9247696399688721,
      "num_tokens": 77257284.0,
      "step": 6263
    },
    {
      "entropy": 1.331528127193451,
      "epoch": 3.2985781990521326,
      "grad_norm": 0.2715156674385071,
      "learning_rate": 0.00012274124336149388,
      "loss": 0.15209448337554932,
      "mean_token_accuracy": 0.9380554258823395,
      "num_tokens": 77269620.0,
      "step": 6264
    },
    {
      "entropy": 1.2930799722671509,
      "epoch": 3.2991047919957874,
      "grad_norm": 0.3041073679924011,
      "learning_rate": 0.00012271953555034948,
      "loss": 0.16017360985279083,
      "mean_token_accuracy": 0.9354993849992752,
      "num_tokens": 77281956.0,
      "step": 6265
    },
    {
      "entropy": 1.288301259279251,
      "epoch": 3.299631384939442,
      "grad_norm": 0.34979477524757385,
      "learning_rate": 0.00012269782698412043,
      "loss": 0.18351054191589355,
      "mean_token_accuracy": 0.9256859570741653,
      "num_tokens": 77294292.0,
      "step": 6266
    },
    {
      "entropy": 1.2581088542938232,
      "epoch": 3.3001579778830963,
      "grad_norm": 0.29193171858787537,
      "learning_rate": 0.0001226761176640954,
      "loss": 0.18515905737876892,
      "mean_token_accuracy": 0.920244887471199,
      "num_tokens": 77306628.0,
      "step": 6267
    },
    {
      "entropy": 1.334260106086731,
      "epoch": 3.3006845708267507,
      "grad_norm": 0.31209471821784973,
      "learning_rate": 0.0001226544075915631,
      "loss": 0.1918957531452179,
      "mean_token_accuracy": 0.9205730259418488,
      "num_tokens": 77318964.0,
      "step": 6268
    },
    {
      "entropy": 1.2262840270996094,
      "epoch": 3.3012111637704056,
      "grad_norm": 0.2917366921901703,
      "learning_rate": 0.00012263269676781243,
      "loss": 0.1814029961824417,
      "mean_token_accuracy": 0.9222237467765808,
      "num_tokens": 77331300.0,
      "step": 6269
    },
    {
      "entropy": 1.291917860507965,
      "epoch": 3.30173775671406,
      "grad_norm": 0.3144878149032593,
      "learning_rate": 0.00012261098519413213,
      "loss": 0.182567298412323,
      "mean_token_accuracy": 0.9216288179159164,
      "num_tokens": 77343636.0,
      "step": 6270
    },
    {
      "entropy": 1.2892558574676514,
      "epoch": 3.3022643496577144,
      "grad_norm": 0.3424370586872101,
      "learning_rate": 0.00012258927287181118,
      "loss": 0.17693836987018585,
      "mean_token_accuracy": 0.9250752925872803,
      "num_tokens": 77355972.0,
      "step": 6271
    },
    {
      "entropy": 1.2943249344825745,
      "epoch": 3.3027909426013693,
      "grad_norm": 0.30987316370010376,
      "learning_rate": 0.00012256755980213845,
      "loss": 0.17040230333805084,
      "mean_token_accuracy": 0.9257796704769135,
      "num_tokens": 77368308.0,
      "step": 6272
    },
    {
      "entropy": 1.248966097831726,
      "epoch": 3.3033175355450237,
      "grad_norm": 0.3050011396408081,
      "learning_rate": 0.00012254584598640297,
      "loss": 0.1835927516222,
      "mean_token_accuracy": 0.9233296066522598,
      "num_tokens": 77380644.0,
      "step": 6273
    },
    {
      "entropy": 1.2827671766281128,
      "epoch": 3.303844128488678,
      "grad_norm": 0.3306349217891693,
      "learning_rate": 0.0001225241314258937,
      "loss": 0.17201808094978333,
      "mean_token_accuracy": 0.9289962649345398,
      "num_tokens": 77392980.0,
      "step": 6274
    },
    {
      "entropy": 1.2073831260204315,
      "epoch": 3.304370721432333,
      "grad_norm": 0.2555938959121704,
      "learning_rate": 0.0001225024161218998,
      "loss": 0.1644388884305954,
      "mean_token_accuracy": 0.9306099265813828,
      "num_tokens": 77405316.0,
      "step": 6275
    },
    {
      "entropy": 1.1848298013210297,
      "epoch": 3.3048973143759874,
      "grad_norm": 0.2741433084011078,
      "learning_rate": 0.00012248070007571027,
      "loss": 0.1641799509525299,
      "mean_token_accuracy": 0.9321964681148529,
      "num_tokens": 77417652.0,
      "step": 6276
    },
    {
      "entropy": 1.2541768550872803,
      "epoch": 3.305423907319642,
      "grad_norm": 0.33179324865341187,
      "learning_rate": 0.00012245898328861436,
      "loss": 0.20032846927642822,
      "mean_token_accuracy": 0.9160593599081039,
      "num_tokens": 77429988.0,
      "step": 6277
    },
    {
      "entropy": 1.2670972049236298,
      "epoch": 3.3059505002632963,
      "grad_norm": 0.32215121388435364,
      "learning_rate": 0.0001224372657619012,
      "loss": 0.18955227732658386,
      "mean_token_accuracy": 0.9195306450128555,
      "num_tokens": 77442324.0,
      "step": 6278
    },
    {
      "entropy": 1.243691623210907,
      "epoch": 3.306477093206951,
      "grad_norm": 0.26894742250442505,
      "learning_rate": 0.00012241554749686006,
      "loss": 0.17201484739780426,
      "mean_token_accuracy": 0.9331451952457428,
      "num_tokens": 77454660.0,
      "step": 6279
    },
    {
      "entropy": 1.2710082828998566,
      "epoch": 3.3070036861506056,
      "grad_norm": 0.277451753616333,
      "learning_rate": 0.00012239382849478026,
      "loss": 0.16356228291988373,
      "mean_token_accuracy": 0.9296573549509048,
      "num_tokens": 77466996.0,
      "step": 6280
    },
    {
      "entropy": 1.2021631002426147,
      "epoch": 3.30753027909426,
      "grad_norm": 0.28685319423675537,
      "learning_rate": 0.00012237210875695108,
      "loss": 0.175466850399971,
      "mean_token_accuracy": 0.9210334420204163,
      "num_tokens": 77479332.0,
      "step": 6281
    },
    {
      "entropy": 1.2092857658863068,
      "epoch": 3.308056872037915,
      "grad_norm": 0.31844884157180786,
      "learning_rate": 0.0001223503882846619,
      "loss": 0.18344981968402863,
      "mean_token_accuracy": 0.9227282404899597,
      "num_tokens": 77491668.0,
      "step": 6282
    },
    {
      "entropy": 1.189133197069168,
      "epoch": 3.3085834649815693,
      "grad_norm": 0.2735708951950073,
      "learning_rate": 0.00012232866707920215,
      "loss": 0.16277778148651123,
      "mean_token_accuracy": 0.9327268749475479,
      "num_tokens": 77504004.0,
      "step": 6283
    },
    {
      "entropy": 1.2456674873828888,
      "epoch": 3.3091100579252237,
      "grad_norm": 0.31346774101257324,
      "learning_rate": 0.00012230694514186134,
      "loss": 0.18626126646995544,
      "mean_token_accuracy": 0.9211814552545547,
      "num_tokens": 77516340.0,
      "step": 6284
    },
    {
      "entropy": 1.2482060492038727,
      "epoch": 3.309636650868878,
      "grad_norm": 0.30682602524757385,
      "learning_rate": 0.00012228522247392887,
      "loss": 0.17980405688285828,
      "mean_token_accuracy": 0.9244150966405869,
      "num_tokens": 77528676.0,
      "step": 6285
    },
    {
      "entropy": 1.1912206411361694,
      "epoch": 3.310163243812533,
      "grad_norm": 0.25870731472969055,
      "learning_rate": 0.0001222634990766944,
      "loss": 0.16307370364665985,
      "mean_token_accuracy": 0.9303781092166901,
      "num_tokens": 77541012.0,
      "step": 6286
    },
    {
      "entropy": 1.1591845452785492,
      "epoch": 3.3106898367561874,
      "grad_norm": 0.27272090315818787,
      "learning_rate": 0.0001222417749514474,
      "loss": 0.17256568372249603,
      "mean_token_accuracy": 0.9279441237449646,
      "num_tokens": 77553348.0,
      "step": 6287
    },
    {
      "entropy": 1.2483986914157867,
      "epoch": 3.311216429699842,
      "grad_norm": 0.2890327572822571,
      "learning_rate": 0.0001222200500994776,
      "loss": 0.16983464360237122,
      "mean_token_accuracy": 0.9310442209243774,
      "num_tokens": 77565684.0,
      "step": 6288
    },
    {
      "entropy": 1.206378698348999,
      "epoch": 3.3117430226434967,
      "grad_norm": 0.2675359547138214,
      "learning_rate": 0.00012219832452207467,
      "loss": 0.1648457944393158,
      "mean_token_accuracy": 0.926652655005455,
      "num_tokens": 77578020.0,
      "step": 6289
    },
    {
      "entropy": 1.2289733290672302,
      "epoch": 3.312269615587151,
      "grad_norm": 0.30007657408714294,
      "learning_rate": 0.00012217659822052828,
      "loss": 0.16314581036567688,
      "mean_token_accuracy": 0.9349983930587769,
      "num_tokens": 77590356.0,
      "step": 6290
    },
    {
      "entropy": 1.2137439548969269,
      "epoch": 3.3127962085308056,
      "grad_norm": 0.2984062731266022,
      "learning_rate": 0.00012215487119612826,
      "loss": 0.18604514002799988,
      "mean_token_accuracy": 0.9209646135568619,
      "num_tokens": 77602692.0,
      "step": 6291
    },
    {
      "entropy": 1.1965177357196808,
      "epoch": 3.3133228014744605,
      "grad_norm": 0.2772281765937805,
      "learning_rate": 0.00012213314345016434,
      "loss": 0.16918744146823883,
      "mean_token_accuracy": 0.9283002465963364,
      "num_tokens": 77615028.0,
      "step": 6292
    },
    {
      "entropy": 1.2328120470046997,
      "epoch": 3.313849394418115,
      "grad_norm": 0.29671624302864075,
      "learning_rate": 0.00012211141498392645,
      "loss": 0.17436084151268005,
      "mean_token_accuracy": 0.926695927977562,
      "num_tokens": 77627364.0,
      "step": 6293
    },
    {
      "entropy": 1.2302624881267548,
      "epoch": 3.3143759873617693,
      "grad_norm": 0.2875054180622101,
      "learning_rate": 0.00012208968579870442,
      "loss": 0.18385103344917297,
      "mean_token_accuracy": 0.9278070479631424,
      "num_tokens": 77639700.0,
      "step": 6294
    },
    {
      "entropy": 1.2216106355190277,
      "epoch": 3.3149025803054237,
      "grad_norm": 0.27963787317276,
      "learning_rate": 0.0001220679558957882,
      "loss": 0.16584040224552155,
      "mean_token_accuracy": 0.9339368790388107,
      "num_tokens": 77652036.0,
      "step": 6295
    },
    {
      "entropy": 1.214008629322052,
      "epoch": 3.3154291732490786,
      "grad_norm": 0.3183528780937195,
      "learning_rate": 0.00012204622527646781,
      "loss": 0.20121121406555176,
      "mean_token_accuracy": 0.9120351374149323,
      "num_tokens": 77664372.0,
      "step": 6296
    },
    {
      "entropy": 1.1923558712005615,
      "epoch": 3.315955766192733,
      "grad_norm": 0.30246466398239136,
      "learning_rate": 0.00012202449394203319,
      "loss": 0.17298796772956848,
      "mean_token_accuracy": 0.9281925112009048,
      "num_tokens": 77676708.0,
      "step": 6297
    },
    {
      "entropy": 1.1890799403190613,
      "epoch": 3.3164823591363874,
      "grad_norm": 0.29945051670074463,
      "learning_rate": 0.00012200276189377449,
      "loss": 0.16734355688095093,
      "mean_token_accuracy": 0.9296771138906479,
      "num_tokens": 77689044.0,
      "step": 6298
    },
    {
      "entropy": 1.155686229467392,
      "epoch": 3.3170089520800423,
      "grad_norm": 0.28938549757003784,
      "learning_rate": 0.00012198102913298178,
      "loss": 0.17813976109027863,
      "mean_token_accuracy": 0.9260826110839844,
      "num_tokens": 77701380.0,
      "step": 6299
    },
    {
      "entropy": 1.2165965139865875,
      "epoch": 3.3175355450236967,
      "grad_norm": 0.31674203276634216,
      "learning_rate": 0.00012195929566094515,
      "loss": 0.18068945407867432,
      "mean_token_accuracy": 0.9256344437599182,
      "num_tokens": 77713716.0,
      "step": 6300
    },
    {
      "entropy": 1.2314816415309906,
      "epoch": 3.318062137967351,
      "grad_norm": 0.30882787704467773,
      "learning_rate": 0.00012193756147895488,
      "loss": 0.18806438148021698,
      "mean_token_accuracy": 0.9189537465572357,
      "num_tokens": 77726052.0,
      "step": 6301
    },
    {
      "entropy": 1.1795615553855896,
      "epoch": 3.3185887309110056,
      "grad_norm": 0.28717485070228577,
      "learning_rate": 0.00012191582658830114,
      "loss": 0.1791655421257019,
      "mean_token_accuracy": 0.9259586930274963,
      "num_tokens": 77738388.0,
      "step": 6302
    },
    {
      "entropy": 1.2114834785461426,
      "epoch": 3.3191153238546605,
      "grad_norm": 0.3069087266921997,
      "learning_rate": 0.00012189409099027427,
      "loss": 0.1844533383846283,
      "mean_token_accuracy": 0.9256695508956909,
      "num_tokens": 77750724.0,
      "step": 6303
    },
    {
      "entropy": 1.189144790172577,
      "epoch": 3.319641916798315,
      "grad_norm": 0.2855905294418335,
      "learning_rate": 0.00012187235468616449,
      "loss": 0.17765678465366364,
      "mean_token_accuracy": 0.9295614361763,
      "num_tokens": 77763060.0,
      "step": 6304
    },
    {
      "entropy": 1.2269161939620972,
      "epoch": 3.3201685097419693,
      "grad_norm": 0.31023627519607544,
      "learning_rate": 0.00012185061767726225,
      "loss": 0.19039218127727509,
      "mean_token_accuracy": 0.9216252714395523,
      "num_tokens": 77775396.0,
      "step": 6305
    },
    {
      "entropy": 1.2304878532886505,
      "epoch": 3.320695102685624,
      "grad_norm": 0.2721764147281647,
      "learning_rate": 0.00012182887996485786,
      "loss": 0.17890873551368713,
      "mean_token_accuracy": 0.926185354590416,
      "num_tokens": 77787732.0,
      "step": 6306
    },
    {
      "entropy": 1.198279082775116,
      "epoch": 3.3212216956292786,
      "grad_norm": 0.3037056028842926,
      "learning_rate": 0.00012180714155024185,
      "loss": 0.19335156679153442,
      "mean_token_accuracy": 0.9204846918582916,
      "num_tokens": 77800068.0,
      "step": 6307
    },
    {
      "entropy": 1.198100596666336,
      "epoch": 3.321748288572933,
      "grad_norm": 0.28984254598617554,
      "learning_rate": 0.00012178540243470466,
      "loss": 0.18056681752204895,
      "mean_token_accuracy": 0.9277454763650894,
      "num_tokens": 77812404.0,
      "step": 6308
    },
    {
      "entropy": 1.2626543045043945,
      "epoch": 3.322274881516588,
      "grad_norm": 0.28939563035964966,
      "learning_rate": 0.00012176366261953677,
      "loss": 0.18535594642162323,
      "mean_token_accuracy": 0.917705699801445,
      "num_tokens": 77824740.0,
      "step": 6309
    },
    {
      "entropy": 1.1280077993869781,
      "epoch": 3.3228014744602423,
      "grad_norm": 0.2761329710483551,
      "learning_rate": 0.00012174192210602886,
      "loss": 0.16896678507328033,
      "mean_token_accuracy": 0.9300144165754318,
      "num_tokens": 77837076.0,
      "step": 6310
    },
    {
      "entropy": 1.1924511790275574,
      "epoch": 3.3233280674038967,
      "grad_norm": 0.2868991196155548,
      "learning_rate": 0.00012172018089547146,
      "loss": 0.1733475625514984,
      "mean_token_accuracy": 0.9288337081670761,
      "num_tokens": 77849412.0,
      "step": 6311
    },
    {
      "entropy": 1.2250944375991821,
      "epoch": 3.323854660347551,
      "grad_norm": 0.28686419129371643,
      "learning_rate": 0.00012169843898915521,
      "loss": 0.17834466695785522,
      "mean_token_accuracy": 0.9234661906957626,
      "num_tokens": 77861748.0,
      "step": 6312
    },
    {
      "entropy": 1.18630051612854,
      "epoch": 3.324381253291206,
      "grad_norm": 0.27554911375045776,
      "learning_rate": 0.00012167669638837083,
      "loss": 0.17016924917697906,
      "mean_token_accuracy": 0.9278373867273331,
      "num_tokens": 77874084.0,
      "step": 6313
    },
    {
      "entropy": 1.194406509399414,
      "epoch": 3.3249078462348605,
      "grad_norm": 0.2967835068702698,
      "learning_rate": 0.00012165495309440904,
      "loss": 0.1776619851589203,
      "mean_token_accuracy": 0.9240773171186447,
      "num_tokens": 77886420.0,
      "step": 6314
    },
    {
      "entropy": 1.1789636015892029,
      "epoch": 3.325434439178515,
      "grad_norm": 0.29467225074768066,
      "learning_rate": 0.00012163320910856062,
      "loss": 0.18165810406208038,
      "mean_token_accuracy": 0.9236823320388794,
      "num_tokens": 77898756.0,
      "step": 6315
    },
    {
      "entropy": 1.235141545534134,
      "epoch": 3.3259610321221693,
      "grad_norm": 0.288943350315094,
      "learning_rate": 0.00012161146443211635,
      "loss": 0.18570385873317719,
      "mean_token_accuracy": 0.9260200262069702,
      "num_tokens": 77911092.0,
      "step": 6316
    },
    {
      "entropy": 1.207500845193863,
      "epoch": 3.326487625065824,
      "grad_norm": 0.3039279282093048,
      "learning_rate": 0.00012158971906636716,
      "loss": 0.18949507176876068,
      "mean_token_accuracy": 0.9216862767934799,
      "num_tokens": 77923428.0,
      "step": 6317
    },
    {
      "entropy": 1.157785713672638,
      "epoch": 3.3270142180094786,
      "grad_norm": 0.28206175565719604,
      "learning_rate": 0.0001215679730126039,
      "loss": 0.16813766956329346,
      "mean_token_accuracy": 0.9271836131811142,
      "num_tokens": 77935764.0,
      "step": 6318
    },
    {
      "entropy": 1.1457824409008026,
      "epoch": 3.327540810953133,
      "grad_norm": 0.2728706896305084,
      "learning_rate": 0.00012154622627211748,
      "loss": 0.18301129341125488,
      "mean_token_accuracy": 0.9233203828334808,
      "num_tokens": 77948100.0,
      "step": 6319
    },
    {
      "entropy": 1.1551872491836548,
      "epoch": 3.328067403896788,
      "grad_norm": 0.30633148550987244,
      "learning_rate": 0.00012152447884619887,
      "loss": 0.1762765645980835,
      "mean_token_accuracy": 0.9293456971645355,
      "num_tokens": 77960436.0,
      "step": 6320
    },
    {
      "entropy": 1.1439066231250763,
      "epoch": 3.3285939968404423,
      "grad_norm": 0.30425259470939636,
      "learning_rate": 0.00012150273073613916,
      "loss": 0.17624837160110474,
      "mean_token_accuracy": 0.9292152673006058,
      "num_tokens": 77972772.0,
      "step": 6321
    },
    {
      "entropy": 1.1565793454647064,
      "epoch": 3.3291205897840968,
      "grad_norm": 0.28278112411499023,
      "learning_rate": 0.00012148098194322936,
      "loss": 0.17056959867477417,
      "mean_token_accuracy": 0.9300214350223541,
      "num_tokens": 77985108.0,
      "step": 6322
    },
    {
      "entropy": 1.1920295357704163,
      "epoch": 3.3296471827277516,
      "grad_norm": 0.32281213998794556,
      "learning_rate": 0.00012145923246876058,
      "loss": 0.18377986550331116,
      "mean_token_accuracy": 0.9265360087156296,
      "num_tokens": 77997444.0,
      "step": 6323
    },
    {
      "entropy": 1.1695996820926666,
      "epoch": 3.330173775671406,
      "grad_norm": 0.32021892070770264,
      "learning_rate": 0.00012143748231402396,
      "loss": 0.19081415235996246,
      "mean_token_accuracy": 0.9188333600759506,
      "num_tokens": 78009780.0,
      "step": 6324
    },
    {
      "entropy": 1.2148053050041199,
      "epoch": 3.3307003686150605,
      "grad_norm": 0.3022609353065491,
      "learning_rate": 0.00012141573148031069,
      "loss": 0.1708085983991623,
      "mean_token_accuracy": 0.9268668293952942,
      "num_tokens": 78022116.0,
      "step": 6325
    },
    {
      "entropy": 1.2103819847106934,
      "epoch": 3.3312269615587153,
      "grad_norm": 0.3088929057121277,
      "learning_rate": 0.00012139397996891195,
      "loss": 0.17563574016094208,
      "mean_token_accuracy": 0.9249402433633804,
      "num_tokens": 78034452.0,
      "step": 6326
    },
    {
      "entropy": 1.2167181074619293,
      "epoch": 3.3317535545023698,
      "grad_norm": 0.3257026672363281,
      "learning_rate": 0.00012137222778111907,
      "loss": 0.18812602758407593,
      "mean_token_accuracy": 0.9249632805585861,
      "num_tokens": 78046788.0,
      "step": 6327
    },
    {
      "entropy": 1.19572913646698,
      "epoch": 3.332280147446024,
      "grad_norm": 0.3196490705013275,
      "learning_rate": 0.00012135047491822329,
      "loss": 0.1719440370798111,
      "mean_token_accuracy": 0.9299701750278473,
      "num_tokens": 78059124.0,
      "step": 6328
    },
    {
      "entropy": 1.186407446861267,
      "epoch": 3.3328067403896786,
      "grad_norm": 0.27008503675460815,
      "learning_rate": 0.00012132872138151597,
      "loss": 0.1582118272781372,
      "mean_token_accuracy": 0.9298784285783768,
      "num_tokens": 78071460.0,
      "step": 6329
    },
    {
      "entropy": 1.2195619642734528,
      "epoch": 3.3333333333333335,
      "grad_norm": 0.2909710109233856,
      "learning_rate": 0.00012130696717228845,
      "loss": 0.1674053966999054,
      "mean_token_accuracy": 0.9280695915222168,
      "num_tokens": 78083796.0,
      "step": 6330
    },
    {
      "entropy": 1.2100169360637665,
      "epoch": 3.333859926276988,
      "grad_norm": 0.2931506335735321,
      "learning_rate": 0.00012128521229183223,
      "loss": 0.18500615656375885,
      "mean_token_accuracy": 0.9271363615989685,
      "num_tokens": 78096132.0,
      "step": 6331
    },
    {
      "entropy": 1.2285574674606323,
      "epoch": 3.3343865192206423,
      "grad_norm": 0.2904454469680786,
      "learning_rate": 0.00012126345674143873,
      "loss": 0.17087505757808685,
      "mean_token_accuracy": 0.9284899085760117,
      "num_tokens": 78108468.0,
      "step": 6332
    },
    {
      "entropy": 1.2296108901500702,
      "epoch": 3.3349131121642968,
      "grad_norm": 0.27629929780960083,
      "learning_rate": 0.00012124170052239947,
      "loss": 0.16646593809127808,
      "mean_token_accuracy": 0.928304061293602,
      "num_tokens": 78120804.0,
      "step": 6333
    },
    {
      "entropy": 1.2326724231243134,
      "epoch": 3.3354397051079516,
      "grad_norm": 0.2919878363609314,
      "learning_rate": 0.00012121994363600593,
      "loss": 0.16050314903259277,
      "mean_token_accuracy": 0.9345861971378326,
      "num_tokens": 78133140.0,
      "step": 6334
    },
    {
      "entropy": 1.2286337912082672,
      "epoch": 3.335966298051606,
      "grad_norm": 0.3131559491157532,
      "learning_rate": 0.00012119818608354976,
      "loss": 0.1890125274658203,
      "mean_token_accuracy": 0.9257236868143082,
      "num_tokens": 78145476.0,
      "step": 6335
    },
    {
      "entropy": 1.224203109741211,
      "epoch": 3.3364928909952605,
      "grad_norm": 0.30545997619628906,
      "learning_rate": 0.00012117642786632252,
      "loss": 0.1883762627840042,
      "mean_token_accuracy": 0.9211859107017517,
      "num_tokens": 78157812.0,
      "step": 6336
    },
    {
      "entropy": 1.2018961906433105,
      "epoch": 3.3370194839389153,
      "grad_norm": 0.28659677505493164,
      "learning_rate": 0.00012115466898561594,
      "loss": 0.16844090819358826,
      "mean_token_accuracy": 0.9275706559419632,
      "num_tokens": 78170148.0,
      "step": 6337
    },
    {
      "entropy": 1.2401127219200134,
      "epoch": 3.3375460768825698,
      "grad_norm": 0.2736494243144989,
      "learning_rate": 0.00012113290944272164,
      "loss": 0.17726998031139374,
      "mean_token_accuracy": 0.9241432696580887,
      "num_tokens": 78182484.0,
      "step": 6338
    },
    {
      "entropy": 1.3018797636032104,
      "epoch": 3.338072669826224,
      "grad_norm": 0.3217552900314331,
      "learning_rate": 0.00012111114923893143,
      "loss": 0.17722119390964508,
      "mean_token_accuracy": 0.9320580661296844,
      "num_tokens": 78194820.0,
      "step": 6339
    },
    {
      "entropy": 1.2583179473876953,
      "epoch": 3.338599262769879,
      "grad_norm": 0.303801953792572,
      "learning_rate": 0.00012108938837553703,
      "loss": 0.19931627810001373,
      "mean_token_accuracy": 0.9146895110607147,
      "num_tokens": 78207156.0,
      "step": 6340
    },
    {
      "entropy": 1.2409113943576813,
      "epoch": 3.3391258557135335,
      "grad_norm": 0.28290995955467224,
      "learning_rate": 0.00012106762685383023,
      "loss": 0.1745530366897583,
      "mean_token_accuracy": 0.9276322573423386,
      "num_tokens": 78219492.0,
      "step": 6341
    },
    {
      "entropy": 1.285282850265503,
      "epoch": 3.339652448657188,
      "grad_norm": 0.3015868663787842,
      "learning_rate": 0.00012104586467510302,
      "loss": 0.18549884855747223,
      "mean_token_accuracy": 0.9204204976558685,
      "num_tokens": 78231828.0,
      "step": 6342
    },
    {
      "entropy": 1.2624708712100983,
      "epoch": 3.340179041600843,
      "grad_norm": 0.2822583317756653,
      "learning_rate": 0.00012102410184064716,
      "loss": 0.1973864734172821,
      "mean_token_accuracy": 0.9140585213899612,
      "num_tokens": 78244164.0,
      "step": 6343
    },
    {
      "entropy": 1.262296050786972,
      "epoch": 3.340705634544497,
      "grad_norm": 0.26198214292526245,
      "learning_rate": 0.0001210023383517546,
      "loss": 0.16668781638145447,
      "mean_token_accuracy": 0.9317191243171692,
      "num_tokens": 78256500.0,
      "step": 6344
    },
    {
      "entropy": 1.2780193090438843,
      "epoch": 3.3412322274881516,
      "grad_norm": 0.27448320388793945,
      "learning_rate": 0.00012098057420971738,
      "loss": 0.1586671769618988,
      "mean_token_accuracy": 0.9334205389022827,
      "num_tokens": 78268836.0,
      "step": 6345
    },
    {
      "entropy": 1.234157383441925,
      "epoch": 3.341758820431806,
      "grad_norm": 0.3024885654449463,
      "learning_rate": 0.00012095880941582744,
      "loss": 0.20172396302223206,
      "mean_token_accuracy": 0.916020467877388,
      "num_tokens": 78281172.0,
      "step": 6346
    },
    {
      "entropy": 1.1784034371376038,
      "epoch": 3.342285413375461,
      "grad_norm": 0.265259712934494,
      "learning_rate": 0.00012093704397137691,
      "loss": 0.17870020866394043,
      "mean_token_accuracy": 0.927327424287796,
      "num_tokens": 78293508.0,
      "step": 6347
    },
    {
      "entropy": 1.2499751150608063,
      "epoch": 3.3428120063191153,
      "grad_norm": 0.2962338328361511,
      "learning_rate": 0.00012091527787765778,
      "loss": 0.1875610649585724,
      "mean_token_accuracy": 0.925552487373352,
      "num_tokens": 78305844.0,
      "step": 6348
    },
    {
      "entropy": 1.24460768699646,
      "epoch": 3.3433385992627698,
      "grad_norm": 0.28698840737342834,
      "learning_rate": 0.0001208935111359622,
      "loss": 0.18320082128047943,
      "mean_token_accuracy": 0.9247717559337616,
      "num_tokens": 78318180.0,
      "step": 6349
    },
    {
      "entropy": 1.1530158817768097,
      "epoch": 3.343865192206424,
      "grad_norm": 0.2892373502254486,
      "learning_rate": 0.00012087174374758242,
      "loss": 0.18444812297821045,
      "mean_token_accuracy": 0.9245470464229584,
      "num_tokens": 78330516.0,
      "step": 6350
    },
    {
      "entropy": 1.1886387765407562,
      "epoch": 3.344391785150079,
      "grad_norm": 0.2701685130596161,
      "learning_rate": 0.00012084997571381054,
      "loss": 0.1672668755054474,
      "mean_token_accuracy": 0.9284243434667587,
      "num_tokens": 78342852.0,
      "step": 6351
    },
    {
      "entropy": 1.156844139099121,
      "epoch": 3.3449183780937335,
      "grad_norm": 0.2838175892829895,
      "learning_rate": 0.00012082820703593885,
      "loss": 0.18693877756595612,
      "mean_token_accuracy": 0.9201638102531433,
      "num_tokens": 78355188.0,
      "step": 6352
    },
    {
      "entropy": 1.188556969165802,
      "epoch": 3.345444971037388,
      "grad_norm": 0.2990287244319916,
      "learning_rate": 0.00012080643771525961,
      "loss": 0.18148550391197205,
      "mean_token_accuracy": 0.9223086684942245,
      "num_tokens": 78367524.0,
      "step": 6353
    },
    {
      "entropy": 1.1083954274654388,
      "epoch": 3.345971563981043,
      "grad_norm": 0.257366806268692,
      "learning_rate": 0.00012078466775306516,
      "loss": 0.16781064867973328,
      "mean_token_accuracy": 0.9296265095472336,
      "num_tokens": 78379860.0,
      "step": 6354
    },
    {
      "entropy": 1.1946285963058472,
      "epoch": 3.346498156924697,
      "grad_norm": 0.27597206830978394,
      "learning_rate": 0.00012076289715064782,
      "loss": 0.17030231654644012,
      "mean_token_accuracy": 0.9265072345733643,
      "num_tokens": 78392196.0,
      "step": 6355
    },
    {
      "entropy": 1.183891624212265,
      "epoch": 3.3470247498683516,
      "grad_norm": 0.30806371569633484,
      "learning_rate": 0.00012074112590930002,
      "loss": 0.20386773347854614,
      "mean_token_accuracy": 0.9101837873458862,
      "num_tokens": 78404532.0,
      "step": 6356
    },
    {
      "entropy": 1.1817936897277832,
      "epoch": 3.3475513428120065,
      "grad_norm": 0.31423214077949524,
      "learning_rate": 0.0001207193540303142,
      "loss": 0.19481787085533142,
      "mean_token_accuracy": 0.9217385649681091,
      "num_tokens": 78416868.0,
      "step": 6357
    },
    {
      "entropy": 1.1405643820762634,
      "epoch": 3.348077935755661,
      "grad_norm": 0.2859838306903839,
      "learning_rate": 0.00012069758151498279,
      "loss": 0.16153916716575623,
      "mean_token_accuracy": 0.9346592873334885,
      "num_tokens": 78429204.0,
      "step": 6358
    },
    {
      "entropy": 1.1543972492218018,
      "epoch": 3.3486045286993154,
      "grad_norm": 0.24678775668144226,
      "learning_rate": 0.00012067580836459831,
      "loss": 0.15756991505622864,
      "mean_token_accuracy": 0.9298305809497833,
      "num_tokens": 78441540.0,
      "step": 6359
    },
    {
      "entropy": 1.1946463882923126,
      "epoch": 3.34913112164297,
      "grad_norm": 0.3142666816711426,
      "learning_rate": 0.00012065403458045333,
      "loss": 0.20290496945381165,
      "mean_token_accuracy": 0.9145755767822266,
      "num_tokens": 78453876.0,
      "step": 6360
    },
    {
      "entropy": 1.1657448410987854,
      "epoch": 3.3496577145866246,
      "grad_norm": 0.2870401442050934,
      "learning_rate": 0.00012063226016384039,
      "loss": 0.172034353017807,
      "mean_token_accuracy": 0.922562375664711,
      "num_tokens": 78466212.0,
      "step": 6361
    },
    {
      "entropy": 1.2565245032310486,
      "epoch": 3.350184307530279,
      "grad_norm": 0.3314197361469269,
      "learning_rate": 0.00012061048511605215,
      "loss": 0.20180392265319824,
      "mean_token_accuracy": 0.9127491116523743,
      "num_tokens": 78478548.0,
      "step": 6362
    },
    {
      "entropy": 1.177832692861557,
      "epoch": 3.3507109004739335,
      "grad_norm": 0.2937653362751007,
      "learning_rate": 0.00012058870943838126,
      "loss": 0.18437829613685608,
      "mean_token_accuracy": 0.9219544529914856,
      "num_tokens": 78490884.0,
      "step": 6363
    },
    {
      "entropy": 1.2124599814414978,
      "epoch": 3.3512374934175884,
      "grad_norm": 0.5006588697433472,
      "learning_rate": 0.0001205669331321204,
      "loss": 0.1952374279499054,
      "mean_token_accuracy": 0.9153865575790405,
      "num_tokens": 78503220.0,
      "step": 6364
    },
    {
      "entropy": 1.2190957367420197,
      "epoch": 3.351764086361243,
      "grad_norm": 0.2922239899635315,
      "learning_rate": 0.00012054515619856226,
      "loss": 0.1920488327741623,
      "mean_token_accuracy": 0.9183184653520584,
      "num_tokens": 78515556.0,
      "step": 6365
    },
    {
      "entropy": 1.2415741980075836,
      "epoch": 3.352290679304897,
      "grad_norm": 0.30346837639808655,
      "learning_rate": 0.00012052337863899974,
      "loss": 0.1981666088104248,
      "mean_token_accuracy": 0.9160395860671997,
      "num_tokens": 78527892.0,
      "step": 6366
    },
    {
      "entropy": 1.195659875869751,
      "epoch": 3.3528172722485516,
      "grad_norm": 0.2848213315010071,
      "learning_rate": 0.00012050160045472552,
      "loss": 0.1909000128507614,
      "mean_token_accuracy": 0.9224957972764969,
      "num_tokens": 78540228.0,
      "step": 6367
    },
    {
      "entropy": 1.2450491189956665,
      "epoch": 3.3533438651922065,
      "grad_norm": 0.2762988805770874,
      "learning_rate": 0.00012047982164703249,
      "loss": 0.18776358664035797,
      "mean_token_accuracy": 0.9186998158693314,
      "num_tokens": 78552564.0,
      "step": 6368
    },
    {
      "entropy": 1.207970678806305,
      "epoch": 3.353870458135861,
      "grad_norm": 0.2834496796131134,
      "learning_rate": 0.00012045804221721351,
      "loss": 0.18195226788520813,
      "mean_token_accuracy": 0.9273946136236191,
      "num_tokens": 78564900.0,
      "step": 6369
    },
    {
      "entropy": 1.2410457730293274,
      "epoch": 3.3543970510795154,
      "grad_norm": 0.27324333786964417,
      "learning_rate": 0.00012043626216656154,
      "loss": 0.16907982528209686,
      "mean_token_accuracy": 0.9275219589471817,
      "num_tokens": 78577236.0,
      "step": 6370
    },
    {
      "entropy": 1.2029935419559479,
      "epoch": 3.3549236440231702,
      "grad_norm": 0.2765830457210541,
      "learning_rate": 0.00012041448149636951,
      "loss": 0.17685846984386444,
      "mean_token_accuracy": 0.9254040271043777,
      "num_tokens": 78589572.0,
      "step": 6371
    },
    {
      "entropy": 1.2648300230503082,
      "epoch": 3.3554502369668247,
      "grad_norm": 0.27480435371398926,
      "learning_rate": 0.00012039270020793044,
      "loss": 0.17362096905708313,
      "mean_token_accuracy": 0.9295898973941803,
      "num_tokens": 78601908.0,
      "step": 6372
    },
    {
      "entropy": 1.2284594178199768,
      "epoch": 3.355976829910479,
      "grad_norm": 0.25838661193847656,
      "learning_rate": 0.00012037091830253731,
      "loss": 0.16208310425281525,
      "mean_token_accuracy": 0.9317517578601837,
      "num_tokens": 78614244.0,
      "step": 6373
    },
    {
      "entropy": 1.2273503541946411,
      "epoch": 3.356503422854134,
      "grad_norm": 0.2887210547924042,
      "learning_rate": 0.0001203491357814832,
      "loss": 0.19540974497795105,
      "mean_token_accuracy": 0.9187801033258438,
      "num_tokens": 78626580.0,
      "step": 6374
    },
    {
      "entropy": 1.2493335902690887,
      "epoch": 3.3570300157977884,
      "grad_norm": 0.2819010019302368,
      "learning_rate": 0.00012032735264606123,
      "loss": 0.17759953439235687,
      "mean_token_accuracy": 0.9274661093950272,
      "num_tokens": 78638916.0,
      "step": 6375
    },
    {
      "entropy": 1.2599061727523804,
      "epoch": 3.357556608741443,
      "grad_norm": 0.29018980264663696,
      "learning_rate": 0.00012030556889756451,
      "loss": 0.1770056188106537,
      "mean_token_accuracy": 0.9274345934391022,
      "num_tokens": 78651252.0,
      "step": 6376
    },
    {
      "entropy": 1.264013558626175,
      "epoch": 3.3580832016850977,
      "grad_norm": 0.3153349757194519,
      "learning_rate": 0.00012028378453728624,
      "loss": 0.17471736669540405,
      "mean_token_accuracy": 0.9202364683151245,
      "num_tokens": 78663588.0,
      "step": 6377
    },
    {
      "entropy": 1.1637910306453705,
      "epoch": 3.358609794628752,
      "grad_norm": 0.2624271512031555,
      "learning_rate": 0.00012026199956651962,
      "loss": 0.17016033828258514,
      "mean_token_accuracy": 0.9273744225502014,
      "num_tokens": 78675924.0,
      "step": 6378
    },
    {
      "entropy": 1.1808833479881287,
      "epoch": 3.3591363875724065,
      "grad_norm": 0.2628500163555145,
      "learning_rate": 0.00012024021398655787,
      "loss": 0.16058243811130524,
      "mean_token_accuracy": 0.927247166633606,
      "num_tokens": 78688260.0,
      "step": 6379
    },
    {
      "entropy": 1.2481871843338013,
      "epoch": 3.359662980516061,
      "grad_norm": 0.28747764229774475,
      "learning_rate": 0.00012021842779869434,
      "loss": 0.16280566155910492,
      "mean_token_accuracy": 0.9292982518672943,
      "num_tokens": 78700596.0,
      "step": 6380
    },
    {
      "entropy": 1.2191427946090698,
      "epoch": 3.360189573459716,
      "grad_norm": 0.27619126439094543,
      "learning_rate": 0.00012019664100422232,
      "loss": 0.1760042905807495,
      "mean_token_accuracy": 0.9255061149597168,
      "num_tokens": 78712932.0,
      "step": 6381
    },
    {
      "entropy": 1.1552186608314514,
      "epoch": 3.3607161664033702,
      "grad_norm": 0.30576804280281067,
      "learning_rate": 0.00012017485360443512,
      "loss": 0.18111807107925415,
      "mean_token_accuracy": 0.9282488375902176,
      "num_tokens": 78725268.0,
      "step": 6382
    },
    {
      "entropy": 1.1739835739135742,
      "epoch": 3.3612427593470247,
      "grad_norm": 0.30674469470977783,
      "learning_rate": 0.00012015306560062615,
      "loss": 0.1644260287284851,
      "mean_token_accuracy": 0.9309899657964706,
      "num_tokens": 78737604.0,
      "step": 6383
    },
    {
      "entropy": 1.2402102649211884,
      "epoch": 3.361769352290679,
      "grad_norm": 0.3177843987941742,
      "learning_rate": 0.0001201312769940889,
      "loss": 0.1768120974302292,
      "mean_token_accuracy": 0.9286710470914841,
      "num_tokens": 78749940.0,
      "step": 6384
    },
    {
      "entropy": 1.2038277089595795,
      "epoch": 3.362295945234334,
      "grad_norm": 0.3215712010860443,
      "learning_rate": 0.00012010948778611677,
      "loss": 0.1781700700521469,
      "mean_token_accuracy": 0.9258454293012619,
      "num_tokens": 78762276.0,
      "step": 6385
    },
    {
      "entropy": 1.1869046688079834,
      "epoch": 3.3628225381779884,
      "grad_norm": 0.2943532466888428,
      "learning_rate": 0.00012008769797800325,
      "loss": 0.1682606190443039,
      "mean_token_accuracy": 0.9279265105724335,
      "num_tokens": 78774612.0,
      "step": 6386
    },
    {
      "entropy": 1.2341134548187256,
      "epoch": 3.363349131121643,
      "grad_norm": 0.2947143316268921,
      "learning_rate": 0.00012006590757104193,
      "loss": 0.17378848791122437,
      "mean_token_accuracy": 0.9286226630210876,
      "num_tokens": 78786948.0,
      "step": 6387
    },
    {
      "entropy": 1.2471059262752533,
      "epoch": 3.3638757240652977,
      "grad_norm": 0.3317924439907074,
      "learning_rate": 0.00012004411656652629,
      "loss": 0.19607557356357574,
      "mean_token_accuracy": 0.9176600426435471,
      "num_tokens": 78799284.0,
      "step": 6388
    },
    {
      "entropy": 1.2376134991645813,
      "epoch": 3.364402317008952,
      "grad_norm": 0.31895434856414795,
      "learning_rate": 0.00012002232496575005,
      "loss": 0.19523006677627563,
      "mean_token_accuracy": 0.9213353097438812,
      "num_tokens": 78811620.0,
      "step": 6389
    },
    {
      "entropy": 1.2288467288017273,
      "epoch": 3.3649289099526065,
      "grad_norm": 0.29991111159324646,
      "learning_rate": 0.00012000053277000676,
      "loss": 0.18137109279632568,
      "mean_token_accuracy": 0.9244268536567688,
      "num_tokens": 78823956.0,
      "step": 6390
    },
    {
      "entropy": 1.1793527603149414,
      "epoch": 3.3654555028962614,
      "grad_norm": 0.29665106534957886,
      "learning_rate": 0.00011997873998059014,
      "loss": 0.1704409420490265,
      "mean_token_accuracy": 0.9303467869758606,
      "num_tokens": 78836292.0,
      "step": 6391
    },
    {
      "entropy": 1.1957817375659943,
      "epoch": 3.365982095839916,
      "grad_norm": 0.29373741149902344,
      "learning_rate": 0.00011995694659879388,
      "loss": 0.18752476572990417,
      "mean_token_accuracy": 0.9178687334060669,
      "num_tokens": 78848628.0,
      "step": 6392
    },
    {
      "entropy": 1.2042733132839203,
      "epoch": 3.3665086887835702,
      "grad_norm": 0.28693804144859314,
      "learning_rate": 0.00011993515262591173,
      "loss": 0.17888300120830536,
      "mean_token_accuracy": 0.9232824593782425,
      "num_tokens": 78860964.0,
      "step": 6393
    },
    {
      "entropy": 1.2232481241226196,
      "epoch": 3.367035281727225,
      "grad_norm": 0.2558070123195648,
      "learning_rate": 0.00011991335806323751,
      "loss": 0.16054439544677734,
      "mean_token_accuracy": 0.9309938848018646,
      "num_tokens": 78873300.0,
      "step": 6394
    },
    {
      "entropy": 1.2379162907600403,
      "epoch": 3.3675618746708795,
      "grad_norm": 0.2749565839767456,
      "learning_rate": 0.00011989156291206496,
      "loss": 0.18252678215503693,
      "mean_token_accuracy": 0.9256701171398163,
      "num_tokens": 78885636.0,
      "step": 6395
    },
    {
      "entropy": 1.2794814705848694,
      "epoch": 3.368088467614534,
      "grad_norm": 0.30195149779319763,
      "learning_rate": 0.00011986976717368801,
      "loss": 0.18169333040714264,
      "mean_token_accuracy": 0.9204668700695038,
      "num_tokens": 78897972.0,
      "step": 6396
    },
    {
      "entropy": 1.231975257396698,
      "epoch": 3.3686150605581884,
      "grad_norm": 0.27333301305770874,
      "learning_rate": 0.00011984797084940047,
      "loss": 0.16971755027770996,
      "mean_token_accuracy": 0.9289774149656296,
      "num_tokens": 78910308.0,
      "step": 6397
    },
    {
      "entropy": 1.2228410243988037,
      "epoch": 3.3691416535018432,
      "grad_norm": 0.28853610157966614,
      "learning_rate": 0.00011982617394049631,
      "loss": 0.18892638385295868,
      "mean_token_accuracy": 0.9222717136144638,
      "num_tokens": 78922644.0,
      "step": 6398
    },
    {
      "entropy": 1.249678373336792,
      "epoch": 3.3696682464454977,
      "grad_norm": 0.2797059416770935,
      "learning_rate": 0.0001198043764482695,
      "loss": 0.18563976883888245,
      "mean_token_accuracy": 0.9269446730613708,
      "num_tokens": 78934980.0,
      "step": 6399
    },
    {
      "entropy": 1.284705102443695,
      "epoch": 3.370194839389152,
      "grad_norm": 0.30524787306785583,
      "learning_rate": 0.00011978257837401396,
      "loss": 0.1918303370475769,
      "mean_token_accuracy": 0.9167343527078629,
      "num_tokens": 78947316.0,
      "step": 6400
    },
    {
      "entropy": 1.207202434539795,
      "epoch": 3.3707214323328065,
      "grad_norm": 0.29419729113578796,
      "learning_rate": 0.0001197607797190238,
      "loss": 0.184175044298172,
      "mean_token_accuracy": 0.9205076545476913,
      "num_tokens": 78959652.0,
      "step": 6401
    },
    {
      "entropy": 1.227299064397812,
      "epoch": 3.3712480252764614,
      "grad_norm": 0.2963317930698395,
      "learning_rate": 0.00011973898048459302,
      "loss": 0.1846417635679245,
      "mean_token_accuracy": 0.9231779873371124,
      "num_tokens": 78971988.0,
      "step": 6402
    },
    {
      "entropy": 1.2725982964038849,
      "epoch": 3.371774618220116,
      "grad_norm": 0.2637333273887634,
      "learning_rate": 0.00011971718067201571,
      "loss": 0.16954471170902252,
      "mean_token_accuracy": 0.9284968376159668,
      "num_tokens": 78984324.0,
      "step": 6403
    },
    {
      "entropy": 1.2824425995349884,
      "epoch": 3.3723012111637702,
      "grad_norm": 0.2682304382324219,
      "learning_rate": 0.00011969538028258605,
      "loss": 0.1740560233592987,
      "mean_token_accuracy": 0.9296318143606186,
      "num_tokens": 78996660.0,
      "step": 6404
    },
    {
      "entropy": 1.329407274723053,
      "epoch": 3.372827804107425,
      "grad_norm": 0.29967549443244934,
      "learning_rate": 0.00011967357931759817,
      "loss": 0.19346961379051208,
      "mean_token_accuracy": 0.9207515865564346,
      "num_tokens": 79008996.0,
      "step": 6405
    },
    {
      "entropy": 1.3016041815280914,
      "epoch": 3.3733543970510795,
      "grad_norm": 0.29610949754714966,
      "learning_rate": 0.00011965177777834627,
      "loss": 0.18378973007202148,
      "mean_token_accuracy": 0.9205546379089355,
      "num_tokens": 79021332.0,
      "step": 6406
    },
    {
      "entropy": 1.2999705076217651,
      "epoch": 3.373880989994734,
      "grad_norm": 0.2858199179172516,
      "learning_rate": 0.00011962997566612455,
      "loss": 0.18614935874938965,
      "mean_token_accuracy": 0.926463171839714,
      "num_tokens": 79033668.0,
      "step": 6407
    },
    {
      "entropy": 1.2828991115093231,
      "epoch": 3.374407582938389,
      "grad_norm": 0.2873018980026245,
      "learning_rate": 0.00011960817298222734,
      "loss": 0.1899888515472412,
      "mean_token_accuracy": 0.9183394014835358,
      "num_tokens": 79046004.0,
      "step": 6408
    },
    {
      "entropy": 1.3172793090343475,
      "epoch": 3.3749341758820433,
      "grad_norm": 0.275519460439682,
      "learning_rate": 0.00011958636972794885,
      "loss": 0.1644439399242401,
      "mean_token_accuracy": 0.9322114735841751,
      "num_tokens": 79058340.0,
      "step": 6409
    },
    {
      "entropy": 1.298895001411438,
      "epoch": 3.3754607688256977,
      "grad_norm": 0.2899971604347229,
      "learning_rate": 0.00011956456590458351,
      "loss": 0.19333909451961517,
      "mean_token_accuracy": 0.9222669005393982,
      "num_tokens": 79070676.0,
      "step": 6410
    },
    {
      "entropy": 1.2996078133583069,
      "epoch": 3.375987361769352,
      "grad_norm": 0.27553460001945496,
      "learning_rate": 0.00011954276151342563,
      "loss": 0.1796150505542755,
      "mean_token_accuracy": 0.9242543876171112,
      "num_tokens": 79083012.0,
      "step": 6411
    },
    {
      "entropy": 1.308868795633316,
      "epoch": 3.376513954713007,
      "grad_norm": 0.29169005155563354,
      "learning_rate": 0.0001195209565557696,
      "loss": 0.19432169198989868,
      "mean_token_accuracy": 0.9213165938854218,
      "num_tokens": 79095348.0,
      "step": 6412
    },
    {
      "entropy": 1.2722634673118591,
      "epoch": 3.3770405476566614,
      "grad_norm": 0.2994008958339691,
      "learning_rate": 0.0001194991510329099,
      "loss": 0.18884015083312988,
      "mean_token_accuracy": 0.9226324707269669,
      "num_tokens": 79107684.0,
      "step": 6413
    },
    {
      "entropy": 1.2974510788917542,
      "epoch": 3.377567140600316,
      "grad_norm": 0.2953341603279114,
      "learning_rate": 0.00011947734494614094,
      "loss": 0.17390935122966766,
      "mean_token_accuracy": 0.9245711266994476,
      "num_tokens": 79120020.0,
      "step": 6414
    },
    {
      "entropy": 1.3775281608104706,
      "epoch": 3.3780937335439707,
      "grad_norm": 0.3047100007534027,
      "learning_rate": 0.00011945553829675727,
      "loss": 0.19041329622268677,
      "mean_token_accuracy": 0.9213204681873322,
      "num_tokens": 79132356.0,
      "step": 6415
    },
    {
      "entropy": 1.275354027748108,
      "epoch": 3.378620326487625,
      "grad_norm": 0.284958153963089,
      "learning_rate": 0.0001194337310860534,
      "loss": 0.19259652495384216,
      "mean_token_accuracy": 0.9220735430717468,
      "num_tokens": 79144692.0,
      "step": 6416
    },
    {
      "entropy": 1.2742116153240204,
      "epoch": 3.3791469194312795,
      "grad_norm": 0.2963763475418091,
      "learning_rate": 0.0001194119233153239,
      "loss": 0.19243276119232178,
      "mean_token_accuracy": 0.9173785448074341,
      "num_tokens": 79157028.0,
      "step": 6417
    },
    {
      "entropy": 1.2317672371864319,
      "epoch": 3.379673512374934,
      "grad_norm": 0.2609945237636566,
      "learning_rate": 0.00011939011498586333,
      "loss": 0.1713690608739853,
      "mean_token_accuracy": 0.9274909198284149,
      "num_tokens": 79169364.0,
      "step": 6418
    },
    {
      "entropy": 1.2643351554870605,
      "epoch": 3.380200105318589,
      "grad_norm": 0.29731491208076477,
      "learning_rate": 0.0001193683060989664,
      "loss": 0.1900637447834015,
      "mean_token_accuracy": 0.9176663905382156,
      "num_tokens": 79181700.0,
      "step": 6419
    },
    {
      "entropy": 1.22477388381958,
      "epoch": 3.3807266982622433,
      "grad_norm": 0.26480039954185486,
      "learning_rate": 0.00011934649665592775,
      "loss": 0.1722249835729599,
      "mean_token_accuracy": 0.928803339600563,
      "num_tokens": 79194036.0,
      "step": 6420
    },
    {
      "entropy": 1.3332690596580505,
      "epoch": 3.3812532912058977,
      "grad_norm": 0.301868200302124,
      "learning_rate": 0.00011932468665804204,
      "loss": 0.19132918119430542,
      "mean_token_accuracy": 0.9140020310878754,
      "num_tokens": 79206372.0,
      "step": 6421
    },
    {
      "entropy": 1.2536985278129578,
      "epoch": 3.3817798841495526,
      "grad_norm": 0.2905433475971222,
      "learning_rate": 0.00011930287610660404,
      "loss": 0.18709422647953033,
      "mean_token_accuracy": 0.9244557172060013,
      "num_tokens": 79218708.0,
      "step": 6422
    },
    {
      "entropy": 1.2495616376399994,
      "epoch": 3.382306477093207,
      "grad_norm": 0.28424185514450073,
      "learning_rate": 0.00011928106500290852,
      "loss": 0.1886255443096161,
      "mean_token_accuracy": 0.9215163737535477,
      "num_tokens": 79231044.0,
      "step": 6423
    },
    {
      "entropy": 1.2593499422073364,
      "epoch": 3.3828330700368614,
      "grad_norm": 0.31604936718940735,
      "learning_rate": 0.00011925925334825026,
      "loss": 0.18721966445446014,
      "mean_token_accuracy": 0.9196096658706665,
      "num_tokens": 79243380.0,
      "step": 6424
    },
    {
      "entropy": 1.2623234689235687,
      "epoch": 3.3833596629805163,
      "grad_norm": 0.2880619764328003,
      "learning_rate": 0.0001192374411439241,
      "loss": 0.1715773046016693,
      "mean_token_accuracy": 0.923977866768837,
      "num_tokens": 79255716.0,
      "step": 6425
    },
    {
      "entropy": 1.3184354901313782,
      "epoch": 3.3838862559241707,
      "grad_norm": 0.2996920049190521,
      "learning_rate": 0.00011921562839122488,
      "loss": 0.17126569151878357,
      "mean_token_accuracy": 0.9306629002094269,
      "num_tokens": 79268052.0,
      "step": 6426
    },
    {
      "entropy": 1.3011780977249146,
      "epoch": 3.384412848867825,
      "grad_norm": 0.3026772141456604,
      "learning_rate": 0.00011919381509144754,
      "loss": 0.19099418818950653,
      "mean_token_accuracy": 0.9181606322526932,
      "num_tokens": 79280388.0,
      "step": 6427
    },
    {
      "entropy": 1.2803550064563751,
      "epoch": 3.3849394418114795,
      "grad_norm": 0.29658186435699463,
      "learning_rate": 0.00011917200124588697,
      "loss": 0.18606960773468018,
      "mean_token_accuracy": 0.9221401065587997,
      "num_tokens": 79292724.0,
      "step": 6428
    },
    {
      "entropy": 1.2547394335269928,
      "epoch": 3.3854660347551344,
      "grad_norm": 0.2973290681838989,
      "learning_rate": 0.00011915018685583812,
      "loss": 0.1786566823720932,
      "mean_token_accuracy": 0.9251726716756821,
      "num_tokens": 79305060.0,
      "step": 6429
    },
    {
      "entropy": 1.2376038432121277,
      "epoch": 3.385992627698789,
      "grad_norm": 0.2691799998283386,
      "learning_rate": 0.00011912837192259605,
      "loss": 0.16852682828903198,
      "mean_token_accuracy": 0.9302997440099716,
      "num_tokens": 79317396.0,
      "step": 6430
    },
    {
      "entropy": 1.2441319227218628,
      "epoch": 3.3865192206424433,
      "grad_norm": 0.33154886960983276,
      "learning_rate": 0.00011910655644745575,
      "loss": 0.18511882424354553,
      "mean_token_accuracy": 0.9216586500406265,
      "num_tokens": 79329732.0,
      "step": 6431
    },
    {
      "entropy": 1.2591944336891174,
      "epoch": 3.387045813586098,
      "grad_norm": 0.28172236680984497,
      "learning_rate": 0.0001190847404317122,
      "loss": 0.1694486439228058,
      "mean_token_accuracy": 0.9262288212776184,
      "num_tokens": 79342068.0,
      "step": 6432
    },
    {
      "entropy": 1.2733435928821564,
      "epoch": 3.3875724065297526,
      "grad_norm": 0.2878052890300751,
      "learning_rate": 0.00011906292387666062,
      "loss": 0.17619693279266357,
      "mean_token_accuracy": 0.9262575060129166,
      "num_tokens": 79354404.0,
      "step": 6433
    },
    {
      "entropy": 1.2411152124404907,
      "epoch": 3.388098999473407,
      "grad_norm": 0.2835223078727722,
      "learning_rate": 0.00011904110678359609,
      "loss": 0.16783303022384644,
      "mean_token_accuracy": 0.9300921261310577,
      "num_tokens": 79366740.0,
      "step": 6434
    },
    {
      "entropy": 1.231415331363678,
      "epoch": 3.3886255924170614,
      "grad_norm": 0.29557785391807556,
      "learning_rate": 0.00011901928915381368,
      "loss": 0.17754420638084412,
      "mean_token_accuracy": 0.9224963933229446,
      "num_tokens": 79379076.0,
      "step": 6435
    },
    {
      "entropy": 1.2087534964084625,
      "epoch": 3.3891521853607163,
      "grad_norm": 0.2918643057346344,
      "learning_rate": 0.0001189974709886087,
      "loss": 0.16821406781673431,
      "mean_token_accuracy": 0.9269145727157593,
      "num_tokens": 79391412.0,
      "step": 6436
    },
    {
      "entropy": 1.227657288312912,
      "epoch": 3.3896787783043707,
      "grad_norm": 0.2932935357093811,
      "learning_rate": 0.0001189756522892763,
      "loss": 0.18248559534549713,
      "mean_token_accuracy": 0.9262166172266006,
      "num_tokens": 79403748.0,
      "step": 6437
    },
    {
      "entropy": 1.1676511764526367,
      "epoch": 3.390205371248025,
      "grad_norm": 0.27335324883461,
      "learning_rate": 0.0001189538330571117,
      "loss": 0.17541730403900146,
      "mean_token_accuracy": 0.9259244501590729,
      "num_tokens": 79416084.0,
      "step": 6438
    },
    {
      "entropy": 1.1459696888923645,
      "epoch": 3.39073196419168,
      "grad_norm": 0.28938010334968567,
      "learning_rate": 0.00011893201329341024,
      "loss": 0.19099408388137817,
      "mean_token_accuracy": 0.9196609854698181,
      "num_tokens": 79428420.0,
      "step": 6439
    },
    {
      "entropy": 1.1173287332057953,
      "epoch": 3.3912585571353344,
      "grad_norm": 0.2747189700603485,
      "learning_rate": 0.00011891019299946722,
      "loss": 0.17232942581176758,
      "mean_token_accuracy": 0.9275729954242706,
      "num_tokens": 79440756.0,
      "step": 6440
    },
    {
      "entropy": 1.1371882557868958,
      "epoch": 3.391785150078989,
      "grad_norm": 0.2940896153450012,
      "learning_rate": 0.00011888837217657796,
      "loss": 0.18742087483406067,
      "mean_token_accuracy": 0.9245633780956268,
      "num_tokens": 79453092.0,
      "step": 6441
    },
    {
      "entropy": 1.155574470758438,
      "epoch": 3.3923117430226437,
      "grad_norm": 0.29194408655166626,
      "learning_rate": 0.00011886655082603784,
      "loss": 0.18641987442970276,
      "mean_token_accuracy": 0.917420044541359,
      "num_tokens": 79465428.0,
      "step": 6442
    },
    {
      "entropy": 1.1233960390090942,
      "epoch": 3.392838335966298,
      "grad_norm": 0.26886695623397827,
      "learning_rate": 0.00011884472894914229,
      "loss": 0.16601043939590454,
      "mean_token_accuracy": 0.9291917532682419,
      "num_tokens": 79477764.0,
      "step": 6443
    },
    {
      "entropy": 1.1976157426834106,
      "epoch": 3.3933649289099526,
      "grad_norm": 0.29385286569595337,
      "learning_rate": 0.00011882290654718675,
      "loss": 0.18610018491744995,
      "mean_token_accuracy": 0.9244470000267029,
      "num_tokens": 79490100.0,
      "step": 6444
    },
    {
      "entropy": 1.203144907951355,
      "epoch": 3.393891521853607,
      "grad_norm": 0.29804056882858276,
      "learning_rate": 0.00011880108362146667,
      "loss": 0.1752803772687912,
      "mean_token_accuracy": 0.9256718158721924,
      "num_tokens": 79502436.0,
      "step": 6445
    },
    {
      "entropy": 1.1739285290241241,
      "epoch": 3.394418114797262,
      "grad_norm": 0.2718615233898163,
      "learning_rate": 0.00011877926017327754,
      "loss": 0.1714663803577423,
      "mean_token_accuracy": 0.9250137358903885,
      "num_tokens": 79514772.0,
      "step": 6446
    },
    {
      "entropy": 1.1488629579544067,
      "epoch": 3.3949447077409163,
      "grad_norm": 0.3127129077911377,
      "learning_rate": 0.00011875743620391488,
      "loss": 0.20285893976688385,
      "mean_token_accuracy": 0.917593389749527,
      "num_tokens": 79527108.0,
      "step": 6447
    },
    {
      "entropy": 1.1775774657726288,
      "epoch": 3.3954713006845707,
      "grad_norm": 0.29271870851516724,
      "learning_rate": 0.00011873561171467428,
      "loss": 0.17101457715034485,
      "mean_token_accuracy": 0.9278821647167206,
      "num_tokens": 79539444.0,
      "step": 6448
    },
    {
      "entropy": 1.1941743195056915,
      "epoch": 3.3959978936282256,
      "grad_norm": 0.27094972133636475,
      "learning_rate": 0.00011871378670685136,
      "loss": 0.17561134696006775,
      "mean_token_accuracy": 0.9257544875144958,
      "num_tokens": 79551780.0,
      "step": 6449
    },
    {
      "entropy": 1.178059995174408,
      "epoch": 3.39652448657188,
      "grad_norm": 0.3000102639198303,
      "learning_rate": 0.00011869196118174171,
      "loss": 0.18154367804527283,
      "mean_token_accuracy": 0.9236386120319366,
      "num_tokens": 79564116.0,
      "step": 6450
    },
    {
      "entropy": 1.1910910308361053,
      "epoch": 3.3970510795155344,
      "grad_norm": 0.2832523286342621,
      "learning_rate": 0.00011867013514064099,
      "loss": 0.17263087630271912,
      "mean_token_accuracy": 0.9248661249876022,
      "num_tokens": 79576452.0,
      "step": 6451
    },
    {
      "entropy": 1.2076069116592407,
      "epoch": 3.397577672459189,
      "grad_norm": 0.2817825973033905,
      "learning_rate": 0.00011864830858484483,
      "loss": 0.17139405012130737,
      "mean_token_accuracy": 0.9288913309574127,
      "num_tokens": 79588788.0,
      "step": 6452
    },
    {
      "entropy": 1.2060706615447998,
      "epoch": 3.3981042654028437,
      "grad_norm": 0.2979954481124878,
      "learning_rate": 0.00011862648151564901,
      "loss": 0.18236488103866577,
      "mean_token_accuracy": 0.9236573427915573,
      "num_tokens": 79601124.0,
      "step": 6453
    },
    {
      "entropy": 1.1929865777492523,
      "epoch": 3.398630858346498,
      "grad_norm": 0.2678283751010895,
      "learning_rate": 0.0001186046539343493,
      "loss": 0.16911433637142181,
      "mean_token_accuracy": 0.9306193888187408,
      "num_tokens": 79613460.0,
      "step": 6454
    },
    {
      "entropy": 1.233336627483368,
      "epoch": 3.3991574512901526,
      "grad_norm": 0.29617202281951904,
      "learning_rate": 0.0001185828258422414,
      "loss": 0.18631018698215485,
      "mean_token_accuracy": 0.9211310297250748,
      "num_tokens": 79625796.0,
      "step": 6455
    },
    {
      "entropy": 1.1534165441989899,
      "epoch": 3.3996840442338074,
      "grad_norm": 0.27414995431900024,
      "learning_rate": 0.00011856099724062118,
      "loss": 0.16907277703285217,
      "mean_token_accuracy": 0.9311786592006683,
      "num_tokens": 79638132.0,
      "step": 6456
    },
    {
      "entropy": 1.1924218833446503,
      "epoch": 3.400210637177462,
      "grad_norm": 0.2850103974342346,
      "learning_rate": 0.00011853916813078443,
      "loss": 0.17662088572978973,
      "mean_token_accuracy": 0.921423465013504,
      "num_tokens": 79650468.0,
      "step": 6457
    },
    {
      "entropy": 1.1488845944404602,
      "epoch": 3.4007372301211163,
      "grad_norm": 0.28349876403808594,
      "learning_rate": 0.00011851733851402705,
      "loss": 0.1736859381198883,
      "mean_token_accuracy": 0.9290521740913391,
      "num_tokens": 79662804.0,
      "step": 6458
    },
    {
      "entropy": 1.1178289353847504,
      "epoch": 3.401263823064771,
      "grad_norm": 0.2724773585796356,
      "learning_rate": 0.00011849550839164494,
      "loss": 0.16534751653671265,
      "mean_token_accuracy": 0.9299388378858566,
      "num_tokens": 79675140.0,
      "step": 6459
    },
    {
      "entropy": 1.1308171153068542,
      "epoch": 3.4017904160084256,
      "grad_norm": 0.32460305094718933,
      "learning_rate": 0.00011847367776493398,
      "loss": 0.1942930817604065,
      "mean_token_accuracy": 0.9228388965129852,
      "num_tokens": 79687476.0,
      "step": 6460
    },
    {
      "entropy": 1.1373795866966248,
      "epoch": 3.40231700895208,
      "grad_norm": 0.29590973258018494,
      "learning_rate": 0.00011845184663519017,
      "loss": 0.19952240586280823,
      "mean_token_accuracy": 0.9187348335981369,
      "num_tokens": 79699812.0,
      "step": 6461
    },
    {
      "entropy": 1.150606632232666,
      "epoch": 3.4028436018957344,
      "grad_norm": 0.30514517426490784,
      "learning_rate": 0.0001184300150037095,
      "loss": 0.19055360555648804,
      "mean_token_accuracy": 0.9176119863986969,
      "num_tokens": 79712148.0,
      "step": 6462
    },
    {
      "entropy": 1.1520099937915802,
      "epoch": 3.4033701948393893,
      "grad_norm": 0.3059881031513214,
      "learning_rate": 0.00011840818287178794,
      "loss": 0.18636028468608856,
      "mean_token_accuracy": 0.9222996830940247,
      "num_tokens": 79724484.0,
      "step": 6463
    },
    {
      "entropy": 1.1693687438964844,
      "epoch": 3.4038967877830437,
      "grad_norm": 0.2850539982318878,
      "learning_rate": 0.00011838635024072163,
      "loss": 0.18698805570602417,
      "mean_token_accuracy": 0.9174241870641708,
      "num_tokens": 79736820.0,
      "step": 6464
    },
    {
      "entropy": 1.114208161830902,
      "epoch": 3.404423380726698,
      "grad_norm": 0.3076694905757904,
      "learning_rate": 0.00011836451711180653,
      "loss": 0.19643941521644592,
      "mean_token_accuracy": 0.9194021672010422,
      "num_tokens": 79749156.0,
      "step": 6465
    },
    {
      "entropy": 1.1309235095977783,
      "epoch": 3.4049499736703526,
      "grad_norm": 0.28560197353363037,
      "learning_rate": 0.00011834268348633883,
      "loss": 0.1844782531261444,
      "mean_token_accuracy": 0.9242021590471268,
      "num_tokens": 79761492.0,
      "step": 6466
    },
    {
      "entropy": 1.148236870765686,
      "epoch": 3.4054765666140074,
      "grad_norm": 0.28917980194091797,
      "learning_rate": 0.00011832084936561463,
      "loss": 0.18366330862045288,
      "mean_token_accuracy": 0.9233373254537582,
      "num_tokens": 79773828.0,
      "step": 6467
    },
    {
      "entropy": 1.070885330438614,
      "epoch": 3.406003159557662,
      "grad_norm": 0.2740862965583801,
      "learning_rate": 0.00011829901475093008,
      "loss": 0.1761816143989563,
      "mean_token_accuracy": 0.9252870231866837,
      "num_tokens": 79786164.0,
      "step": 6468
    },
    {
      "entropy": 1.177080899477005,
      "epoch": 3.4065297525013163,
      "grad_norm": 0.3115704357624054,
      "learning_rate": 0.00011827717964358143,
      "loss": 0.18946781754493713,
      "mean_token_accuracy": 0.917195126414299,
      "num_tokens": 79798500.0,
      "step": 6469
    },
    {
      "entropy": 1.0926099717617035,
      "epoch": 3.407056345444971,
      "grad_norm": 0.25463196635246277,
      "learning_rate": 0.00011825534404486482,
      "loss": 0.17440186440944672,
      "mean_token_accuracy": 0.9275397509336472,
      "num_tokens": 79810836.0,
      "step": 6470
    },
    {
      "entropy": 1.0430421829223633,
      "epoch": 3.4075829383886256,
      "grad_norm": 0.26303786039352417,
      "learning_rate": 0.00011823350795607659,
      "loss": 0.1669006198644638,
      "mean_token_accuracy": 0.9294939339160919,
      "num_tokens": 79823172.0,
      "step": 6471
    },
    {
      "entropy": 1.1282849311828613,
      "epoch": 3.40810953133228,
      "grad_norm": 0.2979987561702728,
      "learning_rate": 0.00011821167137851299,
      "loss": 0.18477582931518555,
      "mean_token_accuracy": 0.922090008854866,
      "num_tokens": 79835508.0,
      "step": 6472
    },
    {
      "entropy": 1.1561923325061798,
      "epoch": 3.408636124275935,
      "grad_norm": 0.29177531599998474,
      "learning_rate": 0.00011818983431347028,
      "loss": 0.17757226526737213,
      "mean_token_accuracy": 0.9250189960002899,
      "num_tokens": 79847844.0,
      "step": 6473
    },
    {
      "entropy": 1.0861241519451141,
      "epoch": 3.4091627172195893,
      "grad_norm": 0.2876853942871094,
      "learning_rate": 0.00011816799676224488,
      "loss": 0.1904408186674118,
      "mean_token_accuracy": 0.9173862636089325,
      "num_tokens": 79860180.0,
      "step": 6474
    },
    {
      "entropy": 1.17974391579628,
      "epoch": 3.4096893101632437,
      "grad_norm": 0.3049277067184448,
      "learning_rate": 0.00011814615872613311,
      "loss": 0.18711437284946442,
      "mean_token_accuracy": 0.9238569587469101,
      "num_tokens": 79872516.0,
      "step": 6475
    },
    {
      "entropy": 1.0536549985408783,
      "epoch": 3.4102159031068986,
      "grad_norm": 0.28174707293510437,
      "learning_rate": 0.00011812432020643137,
      "loss": 0.1753990352153778,
      "mean_token_accuracy": 0.9222654402256012,
      "num_tokens": 79884852.0,
      "step": 6476
    },
    {
      "entropy": 1.1042083203792572,
      "epoch": 3.410742496050553,
      "grad_norm": 0.32421329617500305,
      "learning_rate": 0.00011810248120443608,
      "loss": 0.19320152699947357,
      "mean_token_accuracy": 0.920337051153183,
      "num_tokens": 79897188.0,
      "step": 6477
    },
    {
      "entropy": 1.140292465686798,
      "epoch": 3.4112690889942074,
      "grad_norm": 0.2862115502357483,
      "learning_rate": 0.0001180806417214437,
      "loss": 0.1740087866783142,
      "mean_token_accuracy": 0.9301891177892685,
      "num_tokens": 79909524.0,
      "step": 6478
    },
    {
      "entropy": 1.1167017817497253,
      "epoch": 3.411795681937862,
      "grad_norm": 0.2692073881626129,
      "learning_rate": 0.00011805880175875075,
      "loss": 0.1728738248348236,
      "mean_token_accuracy": 0.927298367023468,
      "num_tokens": 79921860.0,
      "step": 6479
    },
    {
      "entropy": 1.1420233249664307,
      "epoch": 3.4123222748815167,
      "grad_norm": 0.29182690382003784,
      "learning_rate": 0.00011803696131765369,
      "loss": 0.19215109944343567,
      "mean_token_accuracy": 0.9238179624080658,
      "num_tokens": 79934196.0,
      "step": 6480
    },
    {
      "entropy": 1.1108609437942505,
      "epoch": 3.412848867825171,
      "grad_norm": 0.2644272446632385,
      "learning_rate": 0.00011801512039944906,
      "loss": 0.1557019054889679,
      "mean_token_accuracy": 0.9326039105653763,
      "num_tokens": 79946532.0,
      "step": 6481
    },
    {
      "entropy": 1.1349286437034607,
      "epoch": 3.4133754607688256,
      "grad_norm": 0.26609179377555847,
      "learning_rate": 0.00011799327900543348,
      "loss": 0.17688050866127014,
      "mean_token_accuracy": 0.9275576919317245,
      "num_tokens": 79958868.0,
      "step": 6482
    },
    {
      "entropy": 1.1302656531333923,
      "epoch": 3.41390205371248,
      "grad_norm": 0.28293290734291077,
      "learning_rate": 0.00011797143713690351,
      "loss": 0.1745883971452713,
      "mean_token_accuracy": 0.9272965341806412,
      "num_tokens": 79971204.0,
      "step": 6483
    },
    {
      "entropy": 1.1133773624897003,
      "epoch": 3.414428646656135,
      "grad_norm": 0.30129706859588623,
      "learning_rate": 0.00011794959479515577,
      "loss": 0.17393873631954193,
      "mean_token_accuracy": 0.9289537519216537,
      "num_tokens": 79983540.0,
      "step": 6484
    },
    {
      "entropy": 1.079831063747406,
      "epoch": 3.4149552395997893,
      "grad_norm": 0.2830981910228729,
      "learning_rate": 0.00011792775198148694,
      "loss": 0.17949578166007996,
      "mean_token_accuracy": 0.9253024756908417,
      "num_tokens": 79995876.0,
      "step": 6485
    },
    {
      "entropy": 1.1521460115909576,
      "epoch": 3.4154818325434437,
      "grad_norm": 0.2825452387332916,
      "learning_rate": 0.00011790590869719364,
      "loss": 0.17278841137886047,
      "mean_token_accuracy": 0.9261876493692398,
      "num_tokens": 80008212.0,
      "step": 6486
    },
    {
      "entropy": 1.139787495136261,
      "epoch": 3.4160084254870986,
      "grad_norm": 0.2988513708114624,
      "learning_rate": 0.00011788406494357267,
      "loss": 0.16685795783996582,
      "mean_token_accuracy": 0.928671270608902,
      "num_tokens": 80020548.0,
      "step": 6487
    },
    {
      "entropy": 1.1628269851207733,
      "epoch": 3.416535018430753,
      "grad_norm": 0.31854918599128723,
      "learning_rate": 0.0001178622207219207,
      "loss": 0.181523859500885,
      "mean_token_accuracy": 0.9237556904554367,
      "num_tokens": 80032884.0,
      "step": 6488
    },
    {
      "entropy": 1.1285809874534607,
      "epoch": 3.4170616113744074,
      "grad_norm": 0.30883485078811646,
      "learning_rate": 0.00011784037603353448,
      "loss": 0.19393117725849152,
      "mean_token_accuracy": 0.9213438779115677,
      "num_tokens": 80045220.0,
      "step": 6489
    },
    {
      "entropy": 1.1644846498966217,
      "epoch": 3.4175882043180623,
      "grad_norm": 0.3214302361011505,
      "learning_rate": 0.00011781853087971087,
      "loss": 0.19055591523647308,
      "mean_token_accuracy": 0.917168065905571,
      "num_tokens": 80057556.0,
      "step": 6490
    },
    {
      "entropy": 1.121892660856247,
      "epoch": 3.4181147972617167,
      "grad_norm": 0.2962856590747833,
      "learning_rate": 0.0001177966852617466,
      "loss": 0.18106937408447266,
      "mean_token_accuracy": 0.9241335690021515,
      "num_tokens": 80069892.0,
      "step": 6491
    },
    {
      "entropy": 1.1662413477897644,
      "epoch": 3.418641390205371,
      "grad_norm": 0.2937811017036438,
      "learning_rate": 0.00011777483918093861,
      "loss": 0.18079042434692383,
      "mean_token_accuracy": 0.9276749789714813,
      "num_tokens": 80082228.0,
      "step": 6492
    },
    {
      "entropy": 1.1140519380569458,
      "epoch": 3.419167983149026,
      "grad_norm": 0.31999075412750244,
      "learning_rate": 0.00011775299263858375,
      "loss": 0.19316482543945312,
      "mean_token_accuracy": 0.913131058216095,
      "num_tokens": 80094564.0,
      "step": 6493
    },
    {
      "entropy": 1.1816271245479584,
      "epoch": 3.4196945760926805,
      "grad_norm": 0.3379019498825073,
      "learning_rate": 0.00011773114563597889,
      "loss": 0.1968633532524109,
      "mean_token_accuracy": 0.9174982309341431,
      "num_tokens": 80106900.0,
      "step": 6494
    },
    {
      "entropy": 1.1929983496665955,
      "epoch": 3.420221169036335,
      "grad_norm": 0.3005848526954651,
      "learning_rate": 0.00011770929817442095,
      "loss": 0.19482584297657013,
      "mean_token_accuracy": 0.9178702384233475,
      "num_tokens": 80119236.0,
      "step": 6495
    },
    {
      "entropy": 1.1554315090179443,
      "epoch": 3.4207477619799893,
      "grad_norm": 0.307674378156662,
      "learning_rate": 0.00011768745025520694,
      "loss": 0.19060266017913818,
      "mean_token_accuracy": 0.9200639873743057,
      "num_tokens": 80131572.0,
      "step": 6496
    },
    {
      "entropy": 1.1913148760795593,
      "epoch": 3.421274354923644,
      "grad_norm": 0.2751021683216095,
      "learning_rate": 0.00011766560187963378,
      "loss": 0.16455766558647156,
      "mean_token_accuracy": 0.9302422404289246,
      "num_tokens": 80143908.0,
      "step": 6497
    },
    {
      "entropy": 1.1741644442081451,
      "epoch": 3.4218009478672986,
      "grad_norm": 0.2983550429344177,
      "learning_rate": 0.00011764375304899852,
      "loss": 0.1866965889930725,
      "mean_token_accuracy": 0.9235400706529617,
      "num_tokens": 80156244.0,
      "step": 6498
    },
    {
      "entropy": 1.1782455444335938,
      "epoch": 3.422327540810953,
      "grad_norm": 0.27744725346565247,
      "learning_rate": 0.00011762190376459821,
      "loss": 0.17858903110027313,
      "mean_token_accuracy": 0.9297527521848679,
      "num_tokens": 80168580.0,
      "step": 6499
    },
    {
      "entropy": 1.1986348330974579,
      "epoch": 3.4228541337546075,
      "grad_norm": 0.3018032908439636,
      "learning_rate": 0.00011760005402772991,
      "loss": 0.1937883049249649,
      "mean_token_accuracy": 0.920555979013443,
      "num_tokens": 80180916.0,
      "step": 6500
    },
    {
      "entropy": 1.15380859375,
      "epoch": 3.4233807266982623,
      "grad_norm": 0.2850533723831177,
      "learning_rate": 0.00011757820383969067,
      "loss": 0.18893076479434967,
      "mean_token_accuracy": 0.9242502003908157,
      "num_tokens": 80193252.0,
      "step": 6501
    },
    {
      "entropy": 1.1450749039649963,
      "epoch": 3.4239073196419167,
      "grad_norm": 0.2731717526912689,
      "learning_rate": 0.00011755635320177765,
      "loss": 0.18395653367042542,
      "mean_token_accuracy": 0.9267681837081909,
      "num_tokens": 80205588.0,
      "step": 6502
    },
    {
      "entropy": 1.1836364269256592,
      "epoch": 3.424433912585571,
      "grad_norm": 0.27843019366264343,
      "learning_rate": 0.000117534502115288,
      "loss": 0.1765354722738266,
      "mean_token_accuracy": 0.9239010214805603,
      "num_tokens": 80217924.0,
      "step": 6503
    },
    {
      "entropy": 1.18682000041008,
      "epoch": 3.424960505529226,
      "grad_norm": 0.2813323140144348,
      "learning_rate": 0.00011751265058151886,
      "loss": 0.18243086338043213,
      "mean_token_accuracy": 0.92229263484478,
      "num_tokens": 80230260.0,
      "step": 6504
    },
    {
      "entropy": 1.1708821058273315,
      "epoch": 3.4254870984728805,
      "grad_norm": 0.2978111207485199,
      "learning_rate": 0.00011749079860176744,
      "loss": 0.1917700469493866,
      "mean_token_accuracy": 0.9173836261034012,
      "num_tokens": 80242596.0,
      "step": 6505
    },
    {
      "entropy": 1.1800695657730103,
      "epoch": 3.426013691416535,
      "grad_norm": 0.26697853207588196,
      "learning_rate": 0.00011746894617733095,
      "loss": 0.16811959445476532,
      "mean_token_accuracy": 0.9246619343757629,
      "num_tokens": 80254932.0,
      "step": 6506
    },
    {
      "entropy": 1.2367313504219055,
      "epoch": 3.4265402843601898,
      "grad_norm": 0.27477413415908813,
      "learning_rate": 0.00011744709330950668,
      "loss": 0.16949662566184998,
      "mean_token_accuracy": 0.9291162043809891,
      "num_tokens": 80267268.0,
      "step": 6507
    },
    {
      "entropy": 1.2026059925556183,
      "epoch": 3.427066877303844,
      "grad_norm": 0.3085479140281677,
      "learning_rate": 0.00011742523999959189,
      "loss": 0.1769024133682251,
      "mean_token_accuracy": 0.9260054975748062,
      "num_tokens": 80279604.0,
      "step": 6508
    },
    {
      "entropy": 1.1773946285247803,
      "epoch": 3.4275934702474986,
      "grad_norm": 0.2875780761241913,
      "learning_rate": 0.00011740338624888387,
      "loss": 0.19272294640541077,
      "mean_token_accuracy": 0.9172291159629822,
      "num_tokens": 80291940.0,
      "step": 6509
    },
    {
      "entropy": 1.1749962270259857,
      "epoch": 3.4281200631911535,
      "grad_norm": 0.30613988637924194,
      "learning_rate": 0.00011738153205867994,
      "loss": 0.20412415266036987,
      "mean_token_accuracy": 0.9147639125585556,
      "num_tokens": 80304276.0,
      "step": 6510
    },
    {
      "entropy": 1.160490334033966,
      "epoch": 3.428646656134808,
      "grad_norm": 0.2724848985671997,
      "learning_rate": 0.0001173596774302775,
      "loss": 0.1785711944103241,
      "mean_token_accuracy": 0.922538086771965,
      "num_tokens": 80316612.0,
      "step": 6511
    },
    {
      "entropy": 1.214755892753601,
      "epoch": 3.4291732490784623,
      "grad_norm": 0.28239157795906067,
      "learning_rate": 0.00011733782236497388,
      "loss": 0.18591734766960144,
      "mean_token_accuracy": 0.9207028150558472,
      "num_tokens": 80328948.0,
      "step": 6512
    },
    {
      "entropy": 1.2043667435646057,
      "epoch": 3.4296998420221168,
      "grad_norm": 0.3124966025352478,
      "learning_rate": 0.00011731596686406654,
      "loss": 0.19779212772846222,
      "mean_token_accuracy": 0.9196977764368057,
      "num_tokens": 80341284.0,
      "step": 6513
    },
    {
      "entropy": 1.2145142257213593,
      "epoch": 3.4302264349657716,
      "grad_norm": 0.2977919578552246,
      "learning_rate": 0.0001172941109288529,
      "loss": 0.18922749161720276,
      "mean_token_accuracy": 0.9222802817821503,
      "num_tokens": 80353620.0,
      "step": 6514
    },
    {
      "entropy": 1.15992671251297,
      "epoch": 3.430753027909426,
      "grad_norm": 0.2696559727191925,
      "learning_rate": 0.00011727225456063039,
      "loss": 0.18153204023838043,
      "mean_token_accuracy": 0.9248543679714203,
      "num_tokens": 80365956.0,
      "step": 6515
    },
    {
      "entropy": 1.2090700566768646,
      "epoch": 3.4312796208530805,
      "grad_norm": 0.3119373023509979,
      "learning_rate": 0.00011725039776069648,
      "loss": 0.20308628678321838,
      "mean_token_accuracy": 0.9092807322740555,
      "num_tokens": 80378292.0,
      "step": 6516
    },
    {
      "entropy": 1.2414629459381104,
      "epoch": 3.431806213796735,
      "grad_norm": 0.3116752505302429,
      "learning_rate": 0.00011722854053034873,
      "loss": 0.17774319648742676,
      "mean_token_accuracy": 0.9240649789571762,
      "num_tokens": 80390628.0,
      "step": 6517
    },
    {
      "entropy": 1.1835478842258453,
      "epoch": 3.4323328067403898,
      "grad_norm": 0.27387240529060364,
      "learning_rate": 0.0001172066828708847,
      "loss": 0.1747710108757019,
      "mean_token_accuracy": 0.923068955540657,
      "num_tokens": 80402964.0,
      "step": 6518
    },
    {
      "entropy": 1.1313741505146027,
      "epoch": 3.432859399684044,
      "grad_norm": 0.304924875497818,
      "learning_rate": 0.00011718482478360188,
      "loss": 0.17499877512454987,
      "mean_token_accuracy": 0.9271975159645081,
      "num_tokens": 80415300.0,
      "step": 6519
    },
    {
      "entropy": 1.1431568264961243,
      "epoch": 3.4333859926276986,
      "grad_norm": 0.28177160024642944,
      "learning_rate": 0.00011716296626979789,
      "loss": 0.19097380340099335,
      "mean_token_accuracy": 0.9180660396814346,
      "num_tokens": 80427636.0,
      "step": 6520
    },
    {
      "entropy": 1.1546144783496857,
      "epoch": 3.4339125855713535,
      "grad_norm": 0.3083032965660095,
      "learning_rate": 0.00011714110733077035,
      "loss": 0.1891811490058899,
      "mean_token_accuracy": 0.9216991811990738,
      "num_tokens": 80439972.0,
      "step": 6521
    },
    {
      "entropy": 1.1065263599157333,
      "epoch": 3.434439178515008,
      "grad_norm": 0.3084210455417633,
      "learning_rate": 0.00011711924796781688,
      "loss": 0.17542743682861328,
      "mean_token_accuracy": 0.9292380809783936,
      "num_tokens": 80452308.0,
      "step": 6522
    },
    {
      "entropy": 1.1676940321922302,
      "epoch": 3.4349657714586623,
      "grad_norm": 0.2608500123023987,
      "learning_rate": 0.00011709738818223514,
      "loss": 0.16403529047966003,
      "mean_token_accuracy": 0.9323164075613022,
      "num_tokens": 80464644.0,
      "step": 6523
    },
    {
      "entropy": 1.1683576107025146,
      "epoch": 3.435492364402317,
      "grad_norm": 0.28820911049842834,
      "learning_rate": 0.00011707552797532286,
      "loss": 0.18304772675037384,
      "mean_token_accuracy": 0.9200898706912994,
      "num_tokens": 80476980.0,
      "step": 6524
    },
    {
      "entropy": 1.185946136713028,
      "epoch": 3.4360189573459716,
      "grad_norm": 0.2908236086368561,
      "learning_rate": 0.00011705366734837771,
      "loss": 0.19122043251991272,
      "mean_token_accuracy": 0.9251454919576645,
      "num_tokens": 80489316.0,
      "step": 6525
    },
    {
      "entropy": 1.1847181618213654,
      "epoch": 3.436545550289626,
      "grad_norm": 0.27221596240997314,
      "learning_rate": 0.00011703180630269743,
      "loss": 0.18010379374027252,
      "mean_token_accuracy": 0.9298422038555145,
      "num_tokens": 80501652.0,
      "step": 6526
    },
    {
      "entropy": 1.2135211527347565,
      "epoch": 3.437072143233281,
      "grad_norm": 0.2955719530582428,
      "learning_rate": 0.00011700994483957982,
      "loss": 0.18285644054412842,
      "mean_token_accuracy": 0.9246233403682709,
      "num_tokens": 80513988.0,
      "step": 6527
    },
    {
      "entropy": 1.144883543252945,
      "epoch": 3.4375987361769353,
      "grad_norm": 0.26957473158836365,
      "learning_rate": 0.00011698808296032264,
      "loss": 0.14868520200252533,
      "mean_token_accuracy": 0.9365567862987518,
      "num_tokens": 80526324.0,
      "step": 6528
    },
    {
      "entropy": 1.1634040772914886,
      "epoch": 3.4381253291205898,
      "grad_norm": 0.2562483847141266,
      "learning_rate": 0.00011696622066622367,
      "loss": 0.16868998110294342,
      "mean_token_accuracy": 0.927982896566391,
      "num_tokens": 80538660.0,
      "step": 6529
    },
    {
      "entropy": 1.1752456724643707,
      "epoch": 3.438651922064244,
      "grad_norm": 0.2774886190891266,
      "learning_rate": 0.00011694435795858078,
      "loss": 0.16871336102485657,
      "mean_token_accuracy": 0.9282544255256653,
      "num_tokens": 80550996.0,
      "step": 6530
    },
    {
      "entropy": 1.1989236772060394,
      "epoch": 3.439178515007899,
      "grad_norm": 0.28946998715400696,
      "learning_rate": 0.00011692249483869186,
      "loss": 0.17763032019138336,
      "mean_token_accuracy": 0.9257817566394806,
      "num_tokens": 80563332.0,
      "step": 6531
    },
    {
      "entropy": 1.1545622050762177,
      "epoch": 3.4397051079515535,
      "grad_norm": 0.27830907702445984,
      "learning_rate": 0.00011690063130785478,
      "loss": 0.16989469528198242,
      "mean_token_accuracy": 0.926095500588417,
      "num_tokens": 80575668.0,
      "step": 6532
    },
    {
      "entropy": 1.1149566769599915,
      "epoch": 3.440231700895208,
      "grad_norm": 0.2736249566078186,
      "learning_rate": 0.00011687876736736745,
      "loss": 0.1887683868408203,
      "mean_token_accuracy": 0.9233509004116058,
      "num_tokens": 80588004.0,
      "step": 6533
    },
    {
      "entropy": 1.2303832173347473,
      "epoch": 3.4407582938388623,
      "grad_norm": 0.29226261377334595,
      "learning_rate": 0.00011685690301852776,
      "loss": 0.17193004488945007,
      "mean_token_accuracy": 0.9291989207267761,
      "num_tokens": 80600340.0,
      "step": 6534
    },
    {
      "entropy": 1.1652312874794006,
      "epoch": 3.441284886782517,
      "grad_norm": 0.2794043719768524,
      "learning_rate": 0.0001168350382626337,
      "loss": 0.177666574716568,
      "mean_token_accuracy": 0.9256787747144699,
      "num_tokens": 80612676.0,
      "step": 6535
    },
    {
      "entropy": 1.1783691346645355,
      "epoch": 3.4418114797261716,
      "grad_norm": 0.29861900210380554,
      "learning_rate": 0.00011681317310098328,
      "loss": 0.16599920392036438,
      "mean_token_accuracy": 0.9310362935066223,
      "num_tokens": 80625012.0,
      "step": 6536
    },
    {
      "entropy": 1.0840280950069427,
      "epoch": 3.442338072669826,
      "grad_norm": 0.26808103919029236,
      "learning_rate": 0.00011679130753487447,
      "loss": 0.16576027870178223,
      "mean_token_accuracy": 0.9295181483030319,
      "num_tokens": 80637348.0,
      "step": 6537
    },
    {
      "entropy": 1.1727299690246582,
      "epoch": 3.442864665613481,
      "grad_norm": 0.31399986147880554,
      "learning_rate": 0.00011676944156560532,
      "loss": 0.18242204189300537,
      "mean_token_accuracy": 0.9273713827133179,
      "num_tokens": 80649684.0,
      "step": 6538
    },
    {
      "entropy": 1.0957313776016235,
      "epoch": 3.4433912585571353,
      "grad_norm": 0.28226765990257263,
      "learning_rate": 0.00011674757519447386,
      "loss": 0.18178041279315948,
      "mean_token_accuracy": 0.9263020008802414,
      "num_tokens": 80662020.0,
      "step": 6539
    },
    {
      "entropy": 1.149830937385559,
      "epoch": 3.4439178515007898,
      "grad_norm": 0.31284838914871216,
      "learning_rate": 0.0001167257084227782,
      "loss": 0.2012108564376831,
      "mean_token_accuracy": 0.9146735966205597,
      "num_tokens": 80674356.0,
      "step": 6540
    },
    {
      "entropy": 1.1322484910488129,
      "epoch": 3.4444444444444446,
      "grad_norm": 0.31378963589668274,
      "learning_rate": 0.00011670384125181641,
      "loss": 0.18821854889392853,
      "mean_token_accuracy": 0.9224686771631241,
      "num_tokens": 80686692.0,
      "step": 6541
    },
    {
      "entropy": 1.1315763294696808,
      "epoch": 3.444971037388099,
      "grad_norm": 0.28233757615089417,
      "learning_rate": 0.0001166819736828867,
      "loss": 0.1841033399105072,
      "mean_token_accuracy": 0.9217792600393295,
      "num_tokens": 80699028.0,
      "step": 6542
    },
    {
      "entropy": 1.1389168202877045,
      "epoch": 3.4454976303317535,
      "grad_norm": 0.2919313609600067,
      "learning_rate": 0.00011666010571728712,
      "loss": 0.1807428002357483,
      "mean_token_accuracy": 0.9238739907741547,
      "num_tokens": 80711364.0,
      "step": 6543
    },
    {
      "entropy": 1.1506391763687134,
      "epoch": 3.446024223275408,
      "grad_norm": 0.30495715141296387,
      "learning_rate": 0.00011663823735631585,
      "loss": 0.17703501880168915,
      "mean_token_accuracy": 0.9246587008237839,
      "num_tokens": 80723700.0,
      "step": 6544
    },
    {
      "entropy": 1.149292677640915,
      "epoch": 3.446550816219063,
      "grad_norm": 0.32440587878227234,
      "learning_rate": 0.00011661636860127114,
      "loss": 0.18952536582946777,
      "mean_token_accuracy": 0.9217132180929184,
      "num_tokens": 80736036.0,
      "step": 6545
    },
    {
      "entropy": 1.143787831068039,
      "epoch": 3.447077409162717,
      "grad_norm": 0.3194747567176819,
      "learning_rate": 0.00011659449945345123,
      "loss": 0.19040217995643616,
      "mean_token_accuracy": 0.9150918424129486,
      "num_tokens": 80748372.0,
      "step": 6546
    },
    {
      "entropy": 1.133771300315857,
      "epoch": 3.4476040021063716,
      "grad_norm": 0.29951024055480957,
      "learning_rate": 0.00011657262991415428,
      "loss": 0.1916394680738449,
      "mean_token_accuracy": 0.9152339398860931,
      "num_tokens": 80760708.0,
      "step": 6547
    },
    {
      "entropy": 1.0690845847129822,
      "epoch": 3.4481305950500265,
      "grad_norm": 0.2761151194572449,
      "learning_rate": 0.00011655075998467861,
      "loss": 0.17733247578144073,
      "mean_token_accuracy": 0.924552246928215,
      "num_tokens": 80773044.0,
      "step": 6548
    },
    {
      "entropy": 1.125348061323166,
      "epoch": 3.448657187993681,
      "grad_norm": 0.3156280219554901,
      "learning_rate": 0.0001165288896663225,
      "loss": 0.19168615341186523,
      "mean_token_accuracy": 0.9187020510435104,
      "num_tokens": 80785380.0,
      "step": 6549
    },
    {
      "entropy": 1.1275436580181122,
      "epoch": 3.4491837809373354,
      "grad_norm": 0.316463828086853,
      "learning_rate": 0.00011650701896038428,
      "loss": 0.18441078066825867,
      "mean_token_accuracy": 0.9246549159288406,
      "num_tokens": 80797716.0,
      "step": 6550
    },
    {
      "entropy": 1.1101392209529877,
      "epoch": 3.4497103738809898,
      "grad_norm": 0.2858564853668213,
      "learning_rate": 0.00011648514786816225,
      "loss": 0.17851370573043823,
      "mean_token_accuracy": 0.9278943240642548,
      "num_tokens": 80810052.0,
      "step": 6551
    },
    {
      "entropy": 1.1429508030414581,
      "epoch": 3.4502369668246446,
      "grad_norm": 0.29304617643356323,
      "learning_rate": 0.0001164632763909548,
      "loss": 0.18404991924762726,
      "mean_token_accuracy": 0.9244458377361298,
      "num_tokens": 80822388.0,
      "step": 6552
    },
    {
      "entropy": 1.1314391195774078,
      "epoch": 3.450763559768299,
      "grad_norm": 0.3081319332122803,
      "learning_rate": 0.00011644140453006032,
      "loss": 0.17857228219509125,
      "mean_token_accuracy": 0.928916722536087,
      "num_tokens": 80834724.0,
      "step": 6553
    },
    {
      "entropy": 1.1212832033634186,
      "epoch": 3.4512901527119535,
      "grad_norm": 0.27429813146591187,
      "learning_rate": 0.00011641953228677715,
      "loss": 0.16691306233406067,
      "mean_token_accuracy": 0.9267957210540771,
      "num_tokens": 80847060.0,
      "step": 6554
    },
    {
      "entropy": 1.1509199142456055,
      "epoch": 3.4518167456556084,
      "grad_norm": 0.27302080392837524,
      "learning_rate": 0.00011639765966240377,
      "loss": 0.17131754755973816,
      "mean_token_accuracy": 0.93344546854496,
      "num_tokens": 80859396.0,
      "step": 6555
    },
    {
      "entropy": 1.1925573647022247,
      "epoch": 3.452343338599263,
      "grad_norm": 0.29770827293395996,
      "learning_rate": 0.00011637578665823865,
      "loss": 0.18186959624290466,
      "mean_token_accuracy": 0.9227061569690704,
      "num_tokens": 80871732.0,
      "step": 6556
    },
    {
      "entropy": 1.1526966989040375,
      "epoch": 3.452869931542917,
      "grad_norm": 0.30691203474998474,
      "learning_rate": 0.00011635391327558024,
      "loss": 0.18933729827404022,
      "mean_token_accuracy": 0.9198598265647888,
      "num_tokens": 80884068.0,
      "step": 6557
    },
    {
      "entropy": 1.1838021278381348,
      "epoch": 3.453396524486572,
      "grad_norm": 0.2861593961715698,
      "learning_rate": 0.00011633203951572702,
      "loss": 0.16275659203529358,
      "mean_token_accuracy": 0.928879976272583,
      "num_tokens": 80896404.0,
      "step": 6558
    },
    {
      "entropy": 1.1318974792957306,
      "epoch": 3.4539231174302265,
      "grad_norm": 0.2791045606136322,
      "learning_rate": 0.00011631016537997752,
      "loss": 0.17893241345882416,
      "mean_token_accuracy": 0.9246382415294647,
      "num_tokens": 80908740.0,
      "step": 6559
    },
    {
      "entropy": 1.1594657599925995,
      "epoch": 3.454449710373881,
      "grad_norm": 0.27512961626052856,
      "learning_rate": 0.00011628829086963027,
      "loss": 0.17512042820453644,
      "mean_token_accuracy": 0.9278998225927353,
      "num_tokens": 80921076.0,
      "step": 6560
    },
    {
      "entropy": 1.1768932342529297,
      "epoch": 3.4549763033175354,
      "grad_norm": 0.3068045973777771,
      "learning_rate": 0.00011626641598598387,
      "loss": 0.1870686411857605,
      "mean_token_accuracy": 0.9198578298091888,
      "num_tokens": 80933412.0,
      "step": 6561
    },
    {
      "entropy": 1.1646634340286255,
      "epoch": 3.4555028962611902,
      "grad_norm": 0.2588438093662262,
      "learning_rate": 0.00011624454073033686,
      "loss": 0.16256776452064514,
      "mean_token_accuracy": 0.933254674077034,
      "num_tokens": 80945748.0,
      "step": 6562
    },
    {
      "entropy": 1.151212453842163,
      "epoch": 3.4560294892048447,
      "grad_norm": 0.27041497826576233,
      "learning_rate": 0.00011622266510398788,
      "loss": 0.16854098439216614,
      "mean_token_accuracy": 0.9279792308807373,
      "num_tokens": 80958084.0,
      "step": 6563
    },
    {
      "entropy": 1.1997774839401245,
      "epoch": 3.456556082148499,
      "grad_norm": 0.29609692096710205,
      "learning_rate": 0.00011620078910823551,
      "loss": 0.17611055076122284,
      "mean_token_accuracy": 0.9268074929714203,
      "num_tokens": 80970420.0,
      "step": 6564
    },
    {
      "entropy": 1.170654296875,
      "epoch": 3.457082675092154,
      "grad_norm": 0.2995675802230835,
      "learning_rate": 0.00011617891274437842,
      "loss": 0.17296573519706726,
      "mean_token_accuracy": 0.9252562969923019,
      "num_tokens": 80982756.0,
      "step": 6565
    },
    {
      "entropy": 1.156025767326355,
      "epoch": 3.4576092680358084,
      "grad_norm": 0.2921859920024872,
      "learning_rate": 0.00011615703601371533,
      "loss": 0.17505501210689545,
      "mean_token_accuracy": 0.9234371185302734,
      "num_tokens": 80995092.0,
      "step": 6566
    },
    {
      "entropy": 1.161046952009201,
      "epoch": 3.458135860979463,
      "grad_norm": 0.31642282009124756,
      "learning_rate": 0.0001161351589175449,
      "loss": 0.18353474140167236,
      "mean_token_accuracy": 0.9227949529886246,
      "num_tokens": 81007428.0,
      "step": 6567
    },
    {
      "entropy": 1.1706399023532867,
      "epoch": 3.458662453923117,
      "grad_norm": 0.2912273108959198,
      "learning_rate": 0.00011611328145716582,
      "loss": 0.18829534947872162,
      "mean_token_accuracy": 0.9239943772554398,
      "num_tokens": 81019764.0,
      "step": 6568
    },
    {
      "entropy": 1.2326059937477112,
      "epoch": 3.459189046866772,
      "grad_norm": 0.3132517635822296,
      "learning_rate": 0.00011609140363387685,
      "loss": 0.1908145248889923,
      "mean_token_accuracy": 0.9225149899721146,
      "num_tokens": 81032100.0,
      "step": 6569
    },
    {
      "entropy": 1.2250230610370636,
      "epoch": 3.4597156398104265,
      "grad_norm": 0.30630579590797424,
      "learning_rate": 0.00011606952544897679,
      "loss": 0.17763450741767883,
      "mean_token_accuracy": 0.9313493818044662,
      "num_tokens": 81044436.0,
      "step": 6570
    },
    {
      "entropy": 1.1709964573383331,
      "epoch": 3.460242232754081,
      "grad_norm": 0.29354631900787354,
      "learning_rate": 0.00011604764690376433,
      "loss": 0.17751732468605042,
      "mean_token_accuracy": 0.9243370294570923,
      "num_tokens": 81056772.0,
      "step": 6571
    },
    {
      "entropy": 1.2080395221710205,
      "epoch": 3.460768825697736,
      "grad_norm": 0.36343979835510254,
      "learning_rate": 0.00011602576799953833,
      "loss": 0.23610971868038177,
      "mean_token_accuracy": 0.9005018621683121,
      "num_tokens": 81069108.0,
      "step": 6572
    },
    {
      "entropy": 1.1714594960212708,
      "epoch": 3.4612954186413902,
      "grad_norm": 0.29311808943748474,
      "learning_rate": 0.00011600388873759761,
      "loss": 0.1803843230009079,
      "mean_token_accuracy": 0.9217955619096756,
      "num_tokens": 81081444.0,
      "step": 6573
    },
    {
      "entropy": 1.220309853553772,
      "epoch": 3.4618220115850447,
      "grad_norm": 0.2932807207107544,
      "learning_rate": 0.00011598200911924104,
      "loss": 0.176841601729393,
      "mean_token_accuracy": 0.9230359196662903,
      "num_tokens": 81093780.0,
      "step": 6574
    },
    {
      "entropy": 1.2133452594280243,
      "epoch": 3.4623486045286995,
      "grad_norm": 0.29667747020721436,
      "learning_rate": 0.00011596012914576741,
      "loss": 0.18306420743465424,
      "mean_token_accuracy": 0.9236768782138824,
      "num_tokens": 81106116.0,
      "step": 6575
    },
    {
      "entropy": 1.1925814747810364,
      "epoch": 3.462875197472354,
      "grad_norm": 0.29117000102996826,
      "learning_rate": 0.00011593824881847568,
      "loss": 0.1902187466621399,
      "mean_token_accuracy": 0.9213954210281372,
      "num_tokens": 81118452.0,
      "step": 6576
    },
    {
      "entropy": 1.1880240738391876,
      "epoch": 3.4634017904160084,
      "grad_norm": 0.25567275285720825,
      "learning_rate": 0.00011591636813866473,
      "loss": 0.1658111959695816,
      "mean_token_accuracy": 0.9269106239080429,
      "num_tokens": 81130788.0,
      "step": 6577
    },
    {
      "entropy": 1.2391690611839294,
      "epoch": 3.463928383359663,
      "grad_norm": 0.29640063643455505,
      "learning_rate": 0.00011589448710763348,
      "loss": 0.19188007712364197,
      "mean_token_accuracy": 0.9219590872526169,
      "num_tokens": 81143124.0,
      "step": 6578
    },
    {
      "entropy": 1.2411593794822693,
      "epoch": 3.4644549763033177,
      "grad_norm": 0.3092680275440216,
      "learning_rate": 0.00011587260572668085,
      "loss": 0.1927184760570526,
      "mean_token_accuracy": 0.9193492829799652,
      "num_tokens": 81155460.0,
      "step": 6579
    },
    {
      "entropy": 1.1688562631607056,
      "epoch": 3.464981569246972,
      "grad_norm": 0.28550633788108826,
      "learning_rate": 0.00011585072399710588,
      "loss": 0.1812695562839508,
      "mean_token_accuracy": 0.9236355274915695,
      "num_tokens": 81167796.0,
      "step": 6580
    },
    {
      "entropy": 1.208341121673584,
      "epoch": 3.4655081621906265,
      "grad_norm": 0.33204883337020874,
      "learning_rate": 0.00011582884192020757,
      "loss": 0.21132433414459229,
      "mean_token_accuracy": 0.9120856076478958,
      "num_tokens": 81180132.0,
      "step": 6581
    },
    {
      "entropy": 1.1913246512413025,
      "epoch": 3.4660347551342814,
      "grad_norm": 0.2986827790737152,
      "learning_rate": 0.00011580695949728484,
      "loss": 0.1846022605895996,
      "mean_token_accuracy": 0.920821487903595,
      "num_tokens": 81192468.0,
      "step": 6582
    },
    {
      "entropy": 1.163210928440094,
      "epoch": 3.466561348077936,
      "grad_norm": 0.247142493724823,
      "learning_rate": 0.00011578507672963677,
      "loss": 0.17058244347572327,
      "mean_token_accuracy": 0.9306974112987518,
      "num_tokens": 81204804.0,
      "step": 6583
    },
    {
      "entropy": 1.184818983078003,
      "epoch": 3.4670879410215902,
      "grad_norm": 0.29265332221984863,
      "learning_rate": 0.00011576319361856241,
      "loss": 0.1763373613357544,
      "mean_token_accuracy": 0.9287092983722687,
      "num_tokens": 81217140.0,
      "step": 6584
    },
    {
      "entropy": 1.2022126913070679,
      "epoch": 3.4676145339652447,
      "grad_norm": 0.2635056674480438,
      "learning_rate": 0.00011574131016536084,
      "loss": 0.16333281993865967,
      "mean_token_accuracy": 0.9308081120252609,
      "num_tokens": 81229476.0,
      "step": 6585
    },
    {
      "entropy": 1.2003533244132996,
      "epoch": 3.4681411269088995,
      "grad_norm": 0.28320667147636414,
      "learning_rate": 0.00011571942637133115,
      "loss": 0.1698949933052063,
      "mean_token_accuracy": 0.9323157668113708,
      "num_tokens": 81241812.0,
      "step": 6586
    },
    {
      "entropy": 1.167895257472992,
      "epoch": 3.468667719852554,
      "grad_norm": 0.29053619503974915,
      "learning_rate": 0.00011569754223777245,
      "loss": 0.1694645881652832,
      "mean_token_accuracy": 0.9299274981021881,
      "num_tokens": 81254148.0,
      "step": 6587
    },
    {
      "entropy": 1.1586227118968964,
      "epoch": 3.4691943127962084,
      "grad_norm": 0.2703516185283661,
      "learning_rate": 0.00011567565776598388,
      "loss": 0.18138079345226288,
      "mean_token_accuracy": 0.9253202378749847,
      "num_tokens": 81266484.0,
      "step": 6588
    },
    {
      "entropy": 1.1255720853805542,
      "epoch": 3.4697209057398632,
      "grad_norm": 0.29182925820350647,
      "learning_rate": 0.00011565377295726459,
      "loss": 0.1912750005722046,
      "mean_token_accuracy": 0.9197563380002975,
      "num_tokens": 81278820.0,
      "step": 6589
    },
    {
      "entropy": 1.187912940979004,
      "epoch": 3.4702474986835177,
      "grad_norm": 0.31208208203315735,
      "learning_rate": 0.00011563188781291374,
      "loss": 0.19024209678173065,
      "mean_token_accuracy": 0.9244617819786072,
      "num_tokens": 81291156.0,
      "step": 6590
    },
    {
      "entropy": 1.1452976763248444,
      "epoch": 3.470774091627172,
      "grad_norm": 0.2791716158390045,
      "learning_rate": 0.00011561000233423056,
      "loss": 0.17118626832962036,
      "mean_token_accuracy": 0.9291893988847733,
      "num_tokens": 81303492.0,
      "step": 6591
    },
    {
      "entropy": 1.1394523978233337,
      "epoch": 3.471300684570827,
      "grad_norm": 0.2760827839374542,
      "learning_rate": 0.00011558811652251422,
      "loss": 0.1816837340593338,
      "mean_token_accuracy": 0.9236189126968384,
      "num_tokens": 81315828.0,
      "step": 6592
    },
    {
      "entropy": 1.1685126721858978,
      "epoch": 3.4718272775144814,
      "grad_norm": 0.296041876077652,
      "learning_rate": 0.00011556623037906395,
      "loss": 0.2019861936569214,
      "mean_token_accuracy": 0.9170893728733063,
      "num_tokens": 81328164.0,
      "step": 6593
    },
    {
      "entropy": 1.1197969913482666,
      "epoch": 3.472353870458136,
      "grad_norm": 0.2810327708721161,
      "learning_rate": 0.00011554434390517905,
      "loss": 0.17977027595043182,
      "mean_token_accuracy": 0.9234771877527237,
      "num_tokens": 81340500.0,
      "step": 6594
    },
    {
      "entropy": 1.098362535238266,
      "epoch": 3.4728804634017902,
      "grad_norm": 0.25688573718070984,
      "learning_rate": 0.00011552245710215879,
      "loss": 0.16313600540161133,
      "mean_token_accuracy": 0.9315668940544128,
      "num_tokens": 81352836.0,
      "step": 6595
    },
    {
      "entropy": 1.1131737232208252,
      "epoch": 3.473407056345445,
      "grad_norm": 0.27133485674858093,
      "learning_rate": 0.00011550056997130243,
      "loss": 0.16423733532428741,
      "mean_token_accuracy": 0.9326688498258591,
      "num_tokens": 81365172.0,
      "step": 6596
    },
    {
      "entropy": 1.1590063273906708,
      "epoch": 3.4739336492890995,
      "grad_norm": 0.297391414642334,
      "learning_rate": 0.00011547868251390925,
      "loss": 0.1711481511592865,
      "mean_token_accuracy": 0.9264299124479294,
      "num_tokens": 81377508.0,
      "step": 6597
    },
    {
      "entropy": 1.1494792103767395,
      "epoch": 3.474460242232754,
      "grad_norm": 0.3100283741950989,
      "learning_rate": 0.00011545679473127864,
      "loss": 0.19451028108596802,
      "mean_token_accuracy": 0.9161731451749802,
      "num_tokens": 81389844.0,
      "step": 6598
    },
    {
      "entropy": 1.133365660905838,
      "epoch": 3.474986835176409,
      "grad_norm": 0.2772313356399536,
      "learning_rate": 0.00011543490662470995,
      "loss": 0.1697755753993988,
      "mean_token_accuracy": 0.9305469989776611,
      "num_tokens": 81402180.0,
      "step": 6599
    },
    {
      "entropy": 1.1910265684127808,
      "epoch": 3.4755134281200633,
      "grad_norm": 0.3043367266654968,
      "learning_rate": 0.00011541301819550252,
      "loss": 0.20497861504554749,
      "mean_token_accuracy": 0.9168409556150436,
      "num_tokens": 81414516.0,
      "step": 6600
    },
    {
      "entropy": 1.1660354733467102,
      "epoch": 3.4760400210637177,
      "grad_norm": 0.2840496599674225,
      "learning_rate": 0.00011539112944495576,
      "loss": 0.18001000583171844,
      "mean_token_accuracy": 0.926645889878273,
      "num_tokens": 81426852.0,
      "step": 6601
    },
    {
      "entropy": 1.1751192808151245,
      "epoch": 3.476566614007372,
      "grad_norm": 0.28846150636672974,
      "learning_rate": 0.00011536924037436905,
      "loss": 0.1850300431251526,
      "mean_token_accuracy": 0.9225201606750488,
      "num_tokens": 81439188.0,
      "step": 6602
    },
    {
      "entropy": 1.1629333794116974,
      "epoch": 3.477093206951027,
      "grad_norm": 0.2861765921115875,
      "learning_rate": 0.00011534735098504183,
      "loss": 0.18943603336811066,
      "mean_token_accuracy": 0.9230009764432907,
      "num_tokens": 81451524.0,
      "step": 6603
    },
    {
      "entropy": 1.1608426570892334,
      "epoch": 3.4776197998946814,
      "grad_norm": 0.2963612675666809,
      "learning_rate": 0.00011532546127827355,
      "loss": 0.1895626187324524,
      "mean_token_accuracy": 0.9220454096794128,
      "num_tokens": 81463860.0,
      "step": 6604
    },
    {
      "entropy": 1.1968906819820404,
      "epoch": 3.478146392838336,
      "grad_norm": 0.30905681848526,
      "learning_rate": 0.00011530357125536368,
      "loss": 0.19823943078517914,
      "mean_token_accuracy": 0.920026570558548,
      "num_tokens": 81476196.0,
      "step": 6605
    },
    {
      "entropy": 1.1985065340995789,
      "epoch": 3.4786729857819907,
      "grad_norm": 0.34819188714027405,
      "learning_rate": 0.00011528168091761173,
      "loss": 0.19964611530303955,
      "mean_token_accuracy": 0.915794849395752,
      "num_tokens": 81488532.0,
      "step": 6606
    },
    {
      "entropy": 1.1689610183238983,
      "epoch": 3.479199578725645,
      "grad_norm": 0.27715957164764404,
      "learning_rate": 0.00011525979026631712,
      "loss": 0.16660673916339874,
      "mean_token_accuracy": 0.9315045028924942,
      "num_tokens": 81500868.0,
      "step": 6607
    },
    {
      "entropy": 1.2170887291431427,
      "epoch": 3.4797261716692995,
      "grad_norm": 0.2823488712310791,
      "learning_rate": 0.00011523789930277944,
      "loss": 0.1779467612504959,
      "mean_token_accuracy": 0.9199190735816956,
      "num_tokens": 81513204.0,
      "step": 6608
    },
    {
      "entropy": 1.162218153476715,
      "epoch": 3.4802527646129544,
      "grad_norm": 0.26582789421081543,
      "learning_rate": 0.00011521600802829819,
      "loss": 0.18130990862846375,
      "mean_token_accuracy": 0.9264215528964996,
      "num_tokens": 81525540.0,
      "step": 6609
    },
    {
      "entropy": 1.1923732459545135,
      "epoch": 3.480779357556609,
      "grad_norm": 0.2713901698589325,
      "learning_rate": 0.00011519411644417296,
      "loss": 0.16954302787780762,
      "mean_token_accuracy": 0.9345614463090897,
      "num_tokens": 81537876.0,
      "step": 6610
    },
    {
      "entropy": 1.1585627794265747,
      "epoch": 3.4813059505002633,
      "grad_norm": 0.27282825112342834,
      "learning_rate": 0.00011517222455170328,
      "loss": 0.1718309372663498,
      "mean_token_accuracy": 0.9259562790393829,
      "num_tokens": 81550212.0,
      "step": 6611
    },
    {
      "entropy": 1.1762122213840485,
      "epoch": 3.4818325434439177,
      "grad_norm": 0.3016502857208252,
      "learning_rate": 0.00011515033235218881,
      "loss": 0.20218555629253387,
      "mean_token_accuracy": 0.9207109361886978,
      "num_tokens": 81562548.0,
      "step": 6612
    },
    {
      "entropy": 1.215482771396637,
      "epoch": 3.4823591363875726,
      "grad_norm": 0.2939586937427521,
      "learning_rate": 0.00011512843984692914,
      "loss": 0.17666953802108765,
      "mean_token_accuracy": 0.9242789447307587,
      "num_tokens": 81574884.0,
      "step": 6613
    },
    {
      "entropy": 1.1481686234474182,
      "epoch": 3.482885729331227,
      "grad_norm": 0.2719016671180725,
      "learning_rate": 0.00011510654703722383,
      "loss": 0.15779361128807068,
      "mean_token_accuracy": 0.9320281744003296,
      "num_tokens": 81587220.0,
      "step": 6614
    },
    {
      "entropy": 1.1334491670131683,
      "epoch": 3.4834123222748814,
      "grad_norm": 0.3050325810909271,
      "learning_rate": 0.00011508465392437264,
      "loss": 0.18969212472438812,
      "mean_token_accuracy": 0.9199853092432022,
      "num_tokens": 81599556.0,
      "step": 6615
    },
    {
      "entropy": 1.1732980012893677,
      "epoch": 3.483938915218536,
      "grad_norm": 0.29076868295669556,
      "learning_rate": 0.00011506276050967518,
      "loss": 0.1871393620967865,
      "mean_token_accuracy": 0.921945109963417,
      "num_tokens": 81611892.0,
      "step": 6616
    },
    {
      "entropy": 1.0930970907211304,
      "epoch": 3.4844655081621907,
      "grad_norm": 0.2882307469844818,
      "learning_rate": 0.00011504086679443111,
      "loss": 0.17107221484184265,
      "mean_token_accuracy": 0.9295700341463089,
      "num_tokens": 81624228.0,
      "step": 6617
    },
    {
      "entropy": 1.1252177059650421,
      "epoch": 3.484992101105845,
      "grad_norm": 0.30228209495544434,
      "learning_rate": 0.00011501897277994015,
      "loss": 0.1892736852169037,
      "mean_token_accuracy": 0.9221300929784775,
      "num_tokens": 81636564.0,
      "step": 6618
    },
    {
      "entropy": 1.0760619044303894,
      "epoch": 3.4855186940494995,
      "grad_norm": 0.29396650195121765,
      "learning_rate": 0.00011499707846750205,
      "loss": 0.18223868310451508,
      "mean_token_accuracy": 0.923398420214653,
      "num_tokens": 81648900.0,
      "step": 6619
    },
    {
      "entropy": 1.1159788966178894,
      "epoch": 3.4860452869931544,
      "grad_norm": 0.2782004177570343,
      "learning_rate": 0.00011497518385841652,
      "loss": 0.17237144708633423,
      "mean_token_accuracy": 0.9255961179733276,
      "num_tokens": 81661236.0,
      "step": 6620
    },
    {
      "entropy": 1.1408976912498474,
      "epoch": 3.486571879936809,
      "grad_norm": 0.27848049998283386,
      "learning_rate": 0.00011495328895398332,
      "loss": 0.17134708166122437,
      "mean_token_accuracy": 0.9314141422510147,
      "num_tokens": 81673572.0,
      "step": 6621
    },
    {
      "entropy": 1.1072252094745636,
      "epoch": 3.4870984728804633,
      "grad_norm": 0.2836681306362152,
      "learning_rate": 0.00011493139375550222,
      "loss": 0.17910803854465485,
      "mean_token_accuracy": 0.9250835627317429,
      "num_tokens": 81685908.0,
      "step": 6622
    },
    {
      "entropy": 1.1477026045322418,
      "epoch": 3.487625065824118,
      "grad_norm": 0.2834495007991791,
      "learning_rate": 0.00011490949826427304,
      "loss": 0.1532410979270935,
      "mean_token_accuracy": 0.9367298632860184,
      "num_tokens": 81698244.0,
      "step": 6623
    },
    {
      "entropy": 1.1082239151000977,
      "epoch": 3.4881516587677726,
      "grad_norm": 0.28554338216781616,
      "learning_rate": 0.00011488760248159554,
      "loss": 0.18579819798469543,
      "mean_token_accuracy": 0.9214308559894562,
      "num_tokens": 81710580.0,
      "step": 6624
    },
    {
      "entropy": 1.1037914752960205,
      "epoch": 3.488678251711427,
      "grad_norm": 0.29008516669273376,
      "learning_rate": 0.00011486570640876959,
      "loss": 0.18042230606079102,
      "mean_token_accuracy": 0.925680547952652,
      "num_tokens": 81722916.0,
      "step": 6625
    },
    {
      "entropy": 1.0983848571777344,
      "epoch": 3.489204844655082,
      "grad_norm": 0.2795179486274719,
      "learning_rate": 0.00011484381004709499,
      "loss": 0.18224458396434784,
      "mean_token_accuracy": 0.9255627989768982,
      "num_tokens": 81735252.0,
      "step": 6626
    },
    {
      "entropy": 1.1155040860176086,
      "epoch": 3.4897314375987363,
      "grad_norm": 0.3062456250190735,
      "learning_rate": 0.00011482191339787161,
      "loss": 0.17353153228759766,
      "mean_token_accuracy": 0.9272499978542328,
      "num_tokens": 81747588.0,
      "step": 6627
    },
    {
      "entropy": 1.114521861076355,
      "epoch": 3.4902580305423907,
      "grad_norm": 0.2883119285106659,
      "learning_rate": 0.00011480001646239935,
      "loss": 0.18347394466400146,
      "mean_token_accuracy": 0.9220988899469376,
      "num_tokens": 81759924.0,
      "step": 6628
    },
    {
      "entropy": 1.1525468230247498,
      "epoch": 3.490784623486045,
      "grad_norm": 0.28988999128341675,
      "learning_rate": 0.00011477811924197807,
      "loss": 0.17436982691287994,
      "mean_token_accuracy": 0.9284380972385406,
      "num_tokens": 81772260.0,
      "step": 6629
    },
    {
      "entropy": 1.115129292011261,
      "epoch": 3.4913112164297,
      "grad_norm": 0.3117002844810486,
      "learning_rate": 0.00011475622173790775,
      "loss": 0.20904293656349182,
      "mean_token_accuracy": 0.9139687567949295,
      "num_tokens": 81784596.0,
      "step": 6630
    },
    {
      "entropy": 1.1141754984855652,
      "epoch": 3.4918378093733544,
      "grad_norm": 0.2861360013484955,
      "learning_rate": 0.00011473432395148824,
      "loss": 0.17029538750648499,
      "mean_token_accuracy": 0.926383838057518,
      "num_tokens": 81796932.0,
      "step": 6631
    },
    {
      "entropy": 1.13048455119133,
      "epoch": 3.492364402317009,
      "grad_norm": 0.28451624512672424,
      "learning_rate": 0.00011471242588401949,
      "loss": 0.17212338745594025,
      "mean_token_accuracy": 0.9275546222925186,
      "num_tokens": 81809268.0,
      "step": 6632
    },
    {
      "entropy": 1.1467244625091553,
      "epoch": 3.4928909952606633,
      "grad_norm": 0.3095795810222626,
      "learning_rate": 0.00011469052753680149,
      "loss": 0.17537778615951538,
      "mean_token_accuracy": 0.9277688413858414,
      "num_tokens": 81821604.0,
      "step": 6633
    },
    {
      "entropy": 1.1912976801395416,
      "epoch": 3.493417588204318,
      "grad_norm": 0.2997737228870392,
      "learning_rate": 0.00011466862891113424,
      "loss": 0.1905490607023239,
      "mean_token_accuracy": 0.9257136583328247,
      "num_tokens": 81833940.0,
      "step": 6634
    },
    {
      "entropy": 1.1551499962806702,
      "epoch": 3.4939441811479726,
      "grad_norm": 0.2778836786746979,
      "learning_rate": 0.00011464673000831767,
      "loss": 0.17382913827896118,
      "mean_token_accuracy": 0.931753933429718,
      "num_tokens": 81846276.0,
      "step": 6635
    },
    {
      "entropy": 1.135138601064682,
      "epoch": 3.494470774091627,
      "grad_norm": 0.2902297377586365,
      "learning_rate": 0.00011462483082965183,
      "loss": 0.17853501439094543,
      "mean_token_accuracy": 0.9257641136646271,
      "num_tokens": 81858612.0,
      "step": 6636
    },
    {
      "entropy": 1.1280643343925476,
      "epoch": 3.494997367035282,
      "grad_norm": 0.2728481888771057,
      "learning_rate": 0.00011460293137643673,
      "loss": 0.17710497975349426,
      "mean_token_accuracy": 0.9243131428956985,
      "num_tokens": 81870948.0,
      "step": 6637
    },
    {
      "entropy": 1.2509664297103882,
      "epoch": 3.4955239599789363,
      "grad_norm": 0.3230011761188507,
      "learning_rate": 0.00011458103164997244,
      "loss": 0.18864156305789948,
      "mean_token_accuracy": 0.9205769598484039,
      "num_tokens": 81883284.0,
      "step": 6638
    },
    {
      "entropy": 1.2034538090229034,
      "epoch": 3.4960505529225907,
      "grad_norm": 0.2898879647254944,
      "learning_rate": 0.00011455913165155898,
      "loss": 0.17311909794807434,
      "mean_token_accuracy": 0.925488069653511,
      "num_tokens": 81895620.0,
      "step": 6639
    },
    {
      "entropy": 1.198070228099823,
      "epoch": 3.4965771458662456,
      "grad_norm": 0.34274739027023315,
      "learning_rate": 0.00011453723138249647,
      "loss": 0.2226196974515915,
      "mean_token_accuracy": 0.9075485169887543,
      "num_tokens": 81907956.0,
      "step": 6640
    },
    {
      "entropy": 1.148144394159317,
      "epoch": 3.4971037388099,
      "grad_norm": 0.29571598768234253,
      "learning_rate": 0.000114515330844085,
      "loss": 0.1886741667985916,
      "mean_token_accuracy": 0.9233464449644089,
      "num_tokens": 81920292.0,
      "step": 6641
    },
    {
      "entropy": 1.221672236919403,
      "epoch": 3.4976303317535544,
      "grad_norm": 0.2766885757446289,
      "learning_rate": 0.00011449343003762459,
      "loss": 0.17083361744880676,
      "mean_token_accuracy": 0.9291570335626602,
      "num_tokens": 81932628.0,
      "step": 6642
    },
    {
      "entropy": 1.1299061477184296,
      "epoch": 3.4981569246972093,
      "grad_norm": 0.2906188368797302,
      "learning_rate": 0.00011447152896441544,
      "loss": 0.1777857095003128,
      "mean_token_accuracy": 0.9236705750226974,
      "num_tokens": 81944964.0,
      "step": 6643
    },
    {
      "entropy": 1.159753531217575,
      "epoch": 3.4986835176408637,
      "grad_norm": 0.2817007005214691,
      "learning_rate": 0.00011444962762575771,
      "loss": 0.16421736776828766,
      "mean_token_accuracy": 0.9325893074274063,
      "num_tokens": 81957300.0,
      "step": 6644
    },
    {
      "entropy": 1.1971134841442108,
      "epoch": 3.499210110584518,
      "grad_norm": 0.2881527841091156,
      "learning_rate": 0.00011442772602295154,
      "loss": 0.1794702112674713,
      "mean_token_accuracy": 0.9238597750663757,
      "num_tokens": 81969636.0,
      "step": 6645
    },
    {
      "entropy": 1.1957855224609375,
      "epoch": 3.4997367035281726,
      "grad_norm": 0.27637138962745667,
      "learning_rate": 0.00011440582415729704,
      "loss": 0.18101292848587036,
      "mean_token_accuracy": 0.9228304326534271,
      "num_tokens": 81981972.0,
      "step": 6646
    },
    {
      "entropy": 1.1890069842338562,
      "epoch": 3.5002632964718274,
      "grad_norm": 0.287105530500412,
      "learning_rate": 0.00011438392203009445,
      "loss": 0.16792245209217072,
      "mean_token_accuracy": 0.9272554069757462,
      "num_tokens": 81994308.0,
      "step": 6647
    },
    {
      "entropy": 1.1754546761512756,
      "epoch": 3.500789889415482,
      "grad_norm": 0.2850852310657501,
      "learning_rate": 0.00011436201964264397,
      "loss": 0.18495099246501923,
      "mean_token_accuracy": 0.9250370264053345,
      "num_tokens": 82006644.0,
      "step": 6648
    },
    {
      "entropy": 1.170114904642105,
      "epoch": 3.5013164823591363,
      "grad_norm": 0.3009004592895508,
      "learning_rate": 0.0001143401169962458,
      "loss": 0.17369845509529114,
      "mean_token_accuracy": 0.9245720952749252,
      "num_tokens": 82018980.0,
      "step": 6649
    },
    {
      "entropy": 1.1544669270515442,
      "epoch": 3.5018430753027907,
      "grad_norm": 0.2812612056732178,
      "learning_rate": 0.00011431821409220016,
      "loss": 0.15663373470306396,
      "mean_token_accuracy": 0.9367344081401825,
      "num_tokens": 82031316.0,
      "step": 6650
    },
    {
      "entropy": 1.166733056306839,
      "epoch": 3.5023696682464456,
      "grad_norm": 0.2908081114292145,
      "learning_rate": 0.00011429631093180735,
      "loss": 0.188845694065094,
      "mean_token_accuracy": 0.9181435257196426,
      "num_tokens": 82043652.0,
      "step": 6651
    },
    {
      "entropy": 1.1835882365703583,
      "epoch": 3.5028962611901,
      "grad_norm": 0.2998650074005127,
      "learning_rate": 0.0001142744075163676,
      "loss": 0.18816843628883362,
      "mean_token_accuracy": 0.9196074903011322,
      "num_tokens": 82055988.0,
      "step": 6652
    },
    {
      "entropy": 1.1490100622177124,
      "epoch": 3.5034228541337544,
      "grad_norm": 0.29029542207717896,
      "learning_rate": 0.00011425250384718113,
      "loss": 0.1794453114271164,
      "mean_token_accuracy": 0.9246736168861389,
      "num_tokens": 82068324.0,
      "step": 6653
    },
    {
      "entropy": 1.1529352962970734,
      "epoch": 3.5039494470774093,
      "grad_norm": 0.28752025961875916,
      "learning_rate": 0.00011423059992554836,
      "loss": 0.1690784990787506,
      "mean_token_accuracy": 0.9275295436382294,
      "num_tokens": 82080660.0,
      "step": 6654
    },
    {
      "entropy": 1.1477917730808258,
      "epoch": 3.5044760400210637,
      "grad_norm": 0.2715306282043457,
      "learning_rate": 0.00011420869575276954,
      "loss": 0.18134590983390808,
      "mean_token_accuracy": 0.923336461186409,
      "num_tokens": 82092996.0,
      "step": 6655
    },
    {
      "entropy": 1.1334083378314972,
      "epoch": 3.505002632964718,
      "grad_norm": 0.2830401360988617,
      "learning_rate": 0.00011418679133014492,
      "loss": 0.1798047125339508,
      "mean_token_accuracy": 0.9220581203699112,
      "num_tokens": 82105332.0,
      "step": 6656
    },
    {
      "entropy": 1.1629810333251953,
      "epoch": 3.505529225908373,
      "grad_norm": 0.29926276206970215,
      "learning_rate": 0.00011416488665897496,
      "loss": 0.17554765939712524,
      "mean_token_accuracy": 0.9291818886995316,
      "num_tokens": 82117668.0,
      "step": 6657
    },
    {
      "entropy": 1.189599096775055,
      "epoch": 3.5060558188520274,
      "grad_norm": 0.29402297735214233,
      "learning_rate": 0.0001141429817405599,
      "loss": 0.1820201873779297,
      "mean_token_accuracy": 0.9224792569875717,
      "num_tokens": 82130004.0,
      "step": 6658
    },
    {
      "entropy": 1.202353149652481,
      "epoch": 3.506582411795682,
      "grad_norm": 0.30211108922958374,
      "learning_rate": 0.00011412107657620022,
      "loss": 0.18185541033744812,
      "mean_token_accuracy": 0.9220666885375977,
      "num_tokens": 82142340.0,
      "step": 6659
    },
    {
      "entropy": 1.1482025682926178,
      "epoch": 3.5071090047393367,
      "grad_norm": 0.2861393094062805,
      "learning_rate": 0.0001140991711671962,
      "loss": 0.1571296602487564,
      "mean_token_accuracy": 0.9309457689523697,
      "num_tokens": 82154676.0,
      "step": 6660
    },
    {
      "entropy": 1.175143539905548,
      "epoch": 3.507635597682991,
      "grad_norm": 0.30909398198127747,
      "learning_rate": 0.00011407726551484829,
      "loss": 0.1912471503019333,
      "mean_token_accuracy": 0.9234516620635986,
      "num_tokens": 82167012.0,
      "step": 6661
    },
    {
      "entropy": 1.1644525229930878,
      "epoch": 3.5081621906266456,
      "grad_norm": 0.27131006121635437,
      "learning_rate": 0.0001140553596204569,
      "loss": 0.16325272619724274,
      "mean_token_accuracy": 0.9287524670362473,
      "num_tokens": 82179348.0,
      "step": 6662
    },
    {
      "entropy": 1.1756154894828796,
      "epoch": 3.5086887835703,
      "grad_norm": 0.2639835774898529,
      "learning_rate": 0.00011403345348532241,
      "loss": 0.16972073912620544,
      "mean_token_accuracy": 0.930676743388176,
      "num_tokens": 82191684.0,
      "step": 6663
    },
    {
      "entropy": 1.168285846710205,
      "epoch": 3.509215376513955,
      "grad_norm": 0.280975341796875,
      "learning_rate": 0.00011401154711074536,
      "loss": 0.17953525483608246,
      "mean_token_accuracy": 0.9248925000429153,
      "num_tokens": 82204020.0,
      "step": 6664
    },
    {
      "entropy": 1.1686240434646606,
      "epoch": 3.5097419694576093,
      "grad_norm": 0.3023900091648102,
      "learning_rate": 0.0001139896404980261,
      "loss": 0.19406768679618835,
      "mean_token_accuracy": 0.9197174608707428,
      "num_tokens": 82216356.0,
      "step": 6665
    },
    {
      "entropy": 1.1314858198165894,
      "epoch": 3.5102685624012637,
      "grad_norm": 0.2750619053840637,
      "learning_rate": 0.00011396773364846515,
      "loss": 0.18743553757667542,
      "mean_token_accuracy": 0.9228628724813461,
      "num_tokens": 82228692.0,
      "step": 6666
    },
    {
      "entropy": 1.1731296479701996,
      "epoch": 3.510795155344918,
      "grad_norm": 0.26219767332077026,
      "learning_rate": 0.00011394582656336294,
      "loss": 0.16224750876426697,
      "mean_token_accuracy": 0.9315479099750519,
      "num_tokens": 82241028.0,
      "step": 6667
    },
    {
      "entropy": 1.1279303133487701,
      "epoch": 3.511321748288573,
      "grad_norm": 0.2696554362773895,
      "learning_rate": 0.00011392391924402006,
      "loss": 0.1876666396856308,
      "mean_token_accuracy": 0.9174473583698273,
      "num_tokens": 82253364.0,
      "step": 6668
    },
    {
      "entropy": 1.1923480331897736,
      "epoch": 3.5118483412322274,
      "grad_norm": 0.3161199688911438,
      "learning_rate": 0.00011390201169173696,
      "loss": 0.18782727420330048,
      "mean_token_accuracy": 0.9201867580413818,
      "num_tokens": 82265700.0,
      "step": 6669
    },
    {
      "entropy": 1.13890740275383,
      "epoch": 3.512374934175882,
      "grad_norm": 0.28775709867477417,
      "learning_rate": 0.00011388010390781412,
      "loss": 0.18304120004177094,
      "mean_token_accuracy": 0.9230560064315796,
      "num_tokens": 82278036.0,
      "step": 6670
    },
    {
      "entropy": 1.1755778789520264,
      "epoch": 3.5129015271195367,
      "grad_norm": 0.29729554057121277,
      "learning_rate": 0.00011385819589355213,
      "loss": 0.19876186549663544,
      "mean_token_accuracy": 0.9160097390413284,
      "num_tokens": 82290372.0,
      "step": 6671
    },
    {
      "entropy": 1.1228938102722168,
      "epoch": 3.513428120063191,
      "grad_norm": 0.269479900598526,
      "learning_rate": 0.00011383628765025159,
      "loss": 0.16613386571407318,
      "mean_token_accuracy": 0.9278391003608704,
      "num_tokens": 82302708.0,
      "step": 6672
    },
    {
      "entropy": 1.1971102356910706,
      "epoch": 3.5139547130068456,
      "grad_norm": 0.29220128059387207,
      "learning_rate": 0.00011381437917921298,
      "loss": 0.19029895961284637,
      "mean_token_accuracy": 0.9208804666996002,
      "num_tokens": 82315044.0,
      "step": 6673
    },
    {
      "entropy": 1.1914150714874268,
      "epoch": 3.5144813059505005,
      "grad_norm": 0.26985040307044983,
      "learning_rate": 0.00011379247048173691,
      "loss": 0.1660212129354477,
      "mean_token_accuracy": 0.9299520552158356,
      "num_tokens": 82327380.0,
      "step": 6674
    },
    {
      "entropy": 1.1616050004959106,
      "epoch": 3.515007898894155,
      "grad_norm": 0.27132195234298706,
      "learning_rate": 0.00011377056155912398,
      "loss": 0.16966667771339417,
      "mean_token_accuracy": 0.9275068193674088,
      "num_tokens": 82339716.0,
      "step": 6675
    },
    {
      "entropy": 1.1854330003261566,
      "epoch": 3.5155344918378093,
      "grad_norm": 0.29107174277305603,
      "learning_rate": 0.00011374865241267478,
      "loss": 0.19449283182621002,
      "mean_token_accuracy": 0.9214458614587784,
      "num_tokens": 82352052.0,
      "step": 6676
    },
    {
      "entropy": 1.159307211637497,
      "epoch": 3.516061084781464,
      "grad_norm": 0.30077505111694336,
      "learning_rate": 0.0001137267430436899,
      "loss": 0.18535788357257843,
      "mean_token_accuracy": 0.9221621602773666,
      "num_tokens": 82364388.0,
      "step": 6677
    },
    {
      "entropy": 1.1502597630023956,
      "epoch": 3.5165876777251186,
      "grad_norm": 0.26556146144866943,
      "learning_rate": 0.00011370483345347005,
      "loss": 0.14751596748828888,
      "mean_token_accuracy": 0.9397944211959839,
      "num_tokens": 82376724.0,
      "step": 6678
    },
    {
      "entropy": 1.1277763843536377,
      "epoch": 3.517114270668773,
      "grad_norm": 0.26896631717681885,
      "learning_rate": 0.00011368292364331586,
      "loss": 0.16893108189105988,
      "mean_token_accuracy": 0.930274710059166,
      "num_tokens": 82389060.0,
      "step": 6679
    },
    {
      "entropy": 1.165677160024643,
      "epoch": 3.5176408636124274,
      "grad_norm": 0.2698739767074585,
      "learning_rate": 0.00011366101361452792,
      "loss": 0.16360832750797272,
      "mean_token_accuracy": 0.9305921196937561,
      "num_tokens": 82401396.0,
      "step": 6680
    },
    {
      "entropy": 1.1598637700080872,
      "epoch": 3.518167456556082,
      "grad_norm": 0.29341399669647217,
      "learning_rate": 0.00011363910336840693,
      "loss": 0.18213388323783875,
      "mean_token_accuracy": 0.9251310080289841,
      "num_tokens": 82413732.0,
      "step": 6681
    },
    {
      "entropy": 1.2141633033752441,
      "epoch": 3.5186940494997367,
      "grad_norm": 0.3217828869819641,
      "learning_rate": 0.00011361719290625359,
      "loss": 0.1961696296930313,
      "mean_token_accuracy": 0.9190056324005127,
      "num_tokens": 82426068.0,
      "step": 6682
    },
    {
      "entropy": 1.2312059104442596,
      "epoch": 3.519220642443391,
      "grad_norm": 0.29935598373413086,
      "learning_rate": 0.00011359528222936859,
      "loss": 0.17847727239131927,
      "mean_token_accuracy": 0.9217860400676727,
      "num_tokens": 82438404.0,
      "step": 6683
    },
    {
      "entropy": 1.1697284579277039,
      "epoch": 3.5197472353870456,
      "grad_norm": 0.2881301939487457,
      "learning_rate": 0.00011357337133905267,
      "loss": 0.1669689267873764,
      "mean_token_accuracy": 0.9282235205173492,
      "num_tokens": 82450740.0,
      "step": 6684
    },
    {
      "entropy": 1.1748462319374084,
      "epoch": 3.5202738283307005,
      "grad_norm": 0.2914867699146271,
      "learning_rate": 0.00011355146023660647,
      "loss": 0.16577622294425964,
      "mean_token_accuracy": 0.928860068321228,
      "num_tokens": 82463076.0,
      "step": 6685
    },
    {
      "entropy": 1.193132072687149,
      "epoch": 3.520800421274355,
      "grad_norm": 0.288672536611557,
      "learning_rate": 0.00011352954892333081,
      "loss": 0.16755886375904083,
      "mean_token_accuracy": 0.9266272485256195,
      "num_tokens": 82475412.0,
      "step": 6686
    },
    {
      "entropy": 1.1540884375572205,
      "epoch": 3.5213270142180093,
      "grad_norm": 0.28521907329559326,
      "learning_rate": 0.00011350763740052641,
      "loss": 0.16336628794670105,
      "mean_token_accuracy": 0.9258767068386078,
      "num_tokens": 82487748.0,
      "step": 6687
    },
    {
      "entropy": 1.1704376339912415,
      "epoch": 3.521853607161664,
      "grad_norm": 0.285859078168869,
      "learning_rate": 0.000113485725669494,
      "loss": 0.174256831407547,
      "mean_token_accuracy": 0.9224302023649216,
      "num_tokens": 82500084.0,
      "step": 6688
    },
    {
      "entropy": 1.1904620826244354,
      "epoch": 3.5223802001053186,
      "grad_norm": 0.3063889145851135,
      "learning_rate": 0.00011346381373153437,
      "loss": 0.17402036488056183,
      "mean_token_accuracy": 0.9279422610998154,
      "num_tokens": 82512420.0,
      "step": 6689
    },
    {
      "entropy": 1.1764242053031921,
      "epoch": 3.522906793048973,
      "grad_norm": 0.3261604905128479,
      "learning_rate": 0.00011344190158794831,
      "loss": 0.19507795572280884,
      "mean_token_accuracy": 0.9163391441106796,
      "num_tokens": 82524756.0,
      "step": 6690
    },
    {
      "entropy": 1.1777270138263702,
      "epoch": 3.523433385992628,
      "grad_norm": 0.30891919136047363,
      "learning_rate": 0.0001134199892400366,
      "loss": 0.17530623078346252,
      "mean_token_accuracy": 0.9266311526298523,
      "num_tokens": 82537092.0,
      "step": 6691
    },
    {
      "entropy": 1.1871914863586426,
      "epoch": 3.5239599789362823,
      "grad_norm": 0.2826218008995056,
      "learning_rate": 0.0001133980766891001,
      "loss": 0.16185614466667175,
      "mean_token_accuracy": 0.9352222830057144,
      "num_tokens": 82549428.0,
      "step": 6692
    },
    {
      "entropy": 1.198514848947525,
      "epoch": 3.5244865718799367,
      "grad_norm": 0.31576254963874817,
      "learning_rate": 0.00011337616393643957,
      "loss": 0.19219081103801727,
      "mean_token_accuracy": 0.9173667877912521,
      "num_tokens": 82561764.0,
      "step": 6693
    },
    {
      "entropy": 1.110146313905716,
      "epoch": 3.5250131648235916,
      "grad_norm": 0.30720821022987366,
      "learning_rate": 0.0001133542509833559,
      "loss": 0.17404642701148987,
      "mean_token_accuracy": 0.9256002455949783,
      "num_tokens": 82574100.0,
      "step": 6694
    },
    {
      "entropy": 1.1758156716823578,
      "epoch": 3.525539757767246,
      "grad_norm": 0.28131312131881714,
      "learning_rate": 0.00011333233783114983,
      "loss": 0.16816528141498566,
      "mean_token_accuracy": 0.9299330562353134,
      "num_tokens": 82586436.0,
      "step": 6695
    },
    {
      "entropy": 1.2057862281799316,
      "epoch": 3.5260663507109005,
      "grad_norm": 0.27498504519462585,
      "learning_rate": 0.00011331042448112233,
      "loss": 0.16259035468101501,
      "mean_token_accuracy": 0.9331092834472656,
      "num_tokens": 82598772.0,
      "step": 6696
    },
    {
      "entropy": 1.254322737455368,
      "epoch": 3.526592943654555,
      "grad_norm": 0.3335970938205719,
      "learning_rate": 0.00011328851093457422,
      "loss": 0.19345268607139587,
      "mean_token_accuracy": 0.9160381555557251,
      "num_tokens": 82611108.0,
      "step": 6697
    },
    {
      "entropy": 1.189445048570633,
      "epoch": 3.5271195365982093,
      "grad_norm": 0.3048352599143982,
      "learning_rate": 0.00011326659719280639,
      "loss": 0.18552914261817932,
      "mean_token_accuracy": 0.9201581329107285,
      "num_tokens": 82623444.0,
      "step": 6698
    },
    {
      "entropy": 1.184005230665207,
      "epoch": 3.527646129541864,
      "grad_norm": 0.2636079788208008,
      "learning_rate": 0.0001132446832571197,
      "loss": 0.1744004786014557,
      "mean_token_accuracy": 0.9310558140277863,
      "num_tokens": 82635780.0,
      "step": 6699
    },
    {
      "entropy": 1.2675062119960785,
      "epoch": 3.5281727224855186,
      "grad_norm": 0.3155592978000641,
      "learning_rate": 0.00011322276912881509,
      "loss": 0.18889999389648438,
      "mean_token_accuracy": 0.9213542491197586,
      "num_tokens": 82648116.0,
      "step": 6700
    },
    {
      "entropy": 1.2281614542007446,
      "epoch": 3.528699315429173,
      "grad_norm": 0.3071075975894928,
      "learning_rate": 0.00011320085480919347,
      "loss": 0.18316353857517242,
      "mean_token_accuracy": 0.9199003130197525,
      "num_tokens": 82660452.0,
      "step": 6701
    },
    {
      "entropy": 1.2589176893234253,
      "epoch": 3.529225908372828,
      "grad_norm": 0.2643257975578308,
      "learning_rate": 0.00011317894029955571,
      "loss": 0.17070399224758148,
      "mean_token_accuracy": 0.9279015958309174,
      "num_tokens": 82672788.0,
      "step": 6702
    },
    {
      "entropy": 1.2546433806419373,
      "epoch": 3.5297525013164823,
      "grad_norm": 0.28589674830436707,
      "learning_rate": 0.00011315702560120287,
      "loss": 0.17748479545116425,
      "mean_token_accuracy": 0.9249358177185059,
      "num_tokens": 82685124.0,
      "step": 6703
    },
    {
      "entropy": 1.2395256757736206,
      "epoch": 3.5302790942601368,
      "grad_norm": 0.27757275104522705,
      "learning_rate": 0.00011313511071543577,
      "loss": 0.18514981865882874,
      "mean_token_accuracy": 0.9227488189935684,
      "num_tokens": 82697460.0,
      "step": 6704
    },
    {
      "entropy": 1.3025157749652863,
      "epoch": 3.5308056872037916,
      "grad_norm": 0.31811168789863586,
      "learning_rate": 0.00011311319564355542,
      "loss": 0.2027345597743988,
      "mean_token_accuracy": 0.919356182217598,
      "num_tokens": 82709796.0,
      "step": 6705
    },
    {
      "entropy": 1.2027359008789062,
      "epoch": 3.531332280147446,
      "grad_norm": 0.25090813636779785,
      "learning_rate": 0.00011309128038686278,
      "loss": 0.17038676142692566,
      "mean_token_accuracy": 0.9237317740917206,
      "num_tokens": 82722132.0,
      "step": 6706
    },
    {
      "entropy": 1.2685667276382446,
      "epoch": 3.5318588730911005,
      "grad_norm": 0.2764739394187927,
      "learning_rate": 0.00011306936494665887,
      "loss": 0.18713237345218658,
      "mean_token_accuracy": 0.9216730147600174,
      "num_tokens": 82734468.0,
      "step": 6707
    },
    {
      "entropy": 1.2293285727500916,
      "epoch": 3.5323854660347553,
      "grad_norm": 0.2929922342300415,
      "learning_rate": 0.00011304744932424464,
      "loss": 0.1961107850074768,
      "mean_token_accuracy": 0.9158258140087128,
      "num_tokens": 82746804.0,
      "step": 6708
    },
    {
      "entropy": 1.2187176942825317,
      "epoch": 3.5329120589784098,
      "grad_norm": 0.23983658850193024,
      "learning_rate": 0.00011302553352092111,
      "loss": 0.16560348868370056,
      "mean_token_accuracy": 0.932516485452652,
      "num_tokens": 82759140.0,
      "step": 6709
    },
    {
      "entropy": 1.202950358390808,
      "epoch": 3.533438651922064,
      "grad_norm": 0.2635363042354584,
      "learning_rate": 0.0001130036175379893,
      "loss": 0.15890291333198547,
      "mean_token_accuracy": 0.9299011379480362,
      "num_tokens": 82771476.0,
      "step": 6710
    },
    {
      "entropy": 1.2060719430446625,
      "epoch": 3.533965244865719,
      "grad_norm": 0.27711576223373413,
      "learning_rate": 0.0001129817013767502,
      "loss": 0.17853610217571259,
      "mean_token_accuracy": 0.9225573241710663,
      "num_tokens": 82783812.0,
      "step": 6711
    },
    {
      "entropy": 1.2049013674259186,
      "epoch": 3.5344918378093735,
      "grad_norm": 0.2870274484157562,
      "learning_rate": 0.00011295978503850487,
      "loss": 0.18564696609973907,
      "mean_token_accuracy": 0.9162482470273972,
      "num_tokens": 82796148.0,
      "step": 6712
    },
    {
      "entropy": 1.2103980481624603,
      "epoch": 3.535018430753028,
      "grad_norm": 0.26491767168045044,
      "learning_rate": 0.00011293786852455439,
      "loss": 0.16474959254264832,
      "mean_token_accuracy": 0.932452842593193,
      "num_tokens": 82808484.0,
      "step": 6713
    },
    {
      "entropy": 1.196048378944397,
      "epoch": 3.5355450236966823,
      "grad_norm": 0.2709740400314331,
      "learning_rate": 0.00011291595183619974,
      "loss": 0.17166750133037567,
      "mean_token_accuracy": 0.9255477041006088,
      "num_tokens": 82820820.0,
      "step": 6714
    },
    {
      "entropy": 1.1764866709709167,
      "epoch": 3.5360716166403368,
      "grad_norm": 0.29145118594169617,
      "learning_rate": 0.00011289403497474202,
      "loss": 0.17616400122642517,
      "mean_token_accuracy": 0.9311197251081467,
      "num_tokens": 82833156.0,
      "step": 6715
    },
    {
      "entropy": 1.1895210444927216,
      "epoch": 3.5365982095839916,
      "grad_norm": 0.3000554144382477,
      "learning_rate": 0.00011287211794148232,
      "loss": 0.18552231788635254,
      "mean_token_accuracy": 0.9238729327917099,
      "num_tokens": 82845492.0,
      "step": 6716
    },
    {
      "entropy": 1.1906383037567139,
      "epoch": 3.537124802527646,
      "grad_norm": 0.30997785925865173,
      "learning_rate": 0.00011285020073772172,
      "loss": 0.1873130053281784,
      "mean_token_accuracy": 0.9219983518123627,
      "num_tokens": 82857828.0,
      "step": 6717
    },
    {
      "entropy": 1.2437363266944885,
      "epoch": 3.5376513954713005,
      "grad_norm": 0.2874256670475006,
      "learning_rate": 0.00011282828336476134,
      "loss": 0.15653786063194275,
      "mean_token_accuracy": 0.9357765763998032,
      "num_tokens": 82870164.0,
      "step": 6718
    },
    {
      "entropy": 1.162617802619934,
      "epoch": 3.5381779884149553,
      "grad_norm": 0.274565190076828,
      "learning_rate": 0.00011280636582390222,
      "loss": 0.16253608465194702,
      "mean_token_accuracy": 0.9314239025115967,
      "num_tokens": 82882500.0,
      "step": 6719
    },
    {
      "entropy": 1.1853586435317993,
      "epoch": 3.5387045813586098,
      "grad_norm": 0.30693045258522034,
      "learning_rate": 0.00011278444811644553,
      "loss": 0.1883617341518402,
      "mean_token_accuracy": 0.9210736006498337,
      "num_tokens": 82894836.0,
      "step": 6720
    },
    {
      "entropy": 1.1961411833763123,
      "epoch": 3.539231174302264,
      "grad_norm": 0.7807809710502625,
      "learning_rate": 0.0001127625302436924,
      "loss": 0.18562889099121094,
      "mean_token_accuracy": 0.9230353385210037,
      "num_tokens": 82907172.0,
      "step": 6721
    },
    {
      "entropy": 1.1774947941303253,
      "epoch": 3.539757767245919,
      "grad_norm": 0.30910661816596985,
      "learning_rate": 0.00011274061220694396,
      "loss": 0.17419037222862244,
      "mean_token_accuracy": 0.924850583076477,
      "num_tokens": 82919508.0,
      "step": 6722
    },
    {
      "entropy": 1.1937752664089203,
      "epoch": 3.5402843601895735,
      "grad_norm": 0.32564640045166016,
      "learning_rate": 0.00011271869400750133,
      "loss": 0.19138182699680328,
      "mean_token_accuracy": 0.9183522164821625,
      "num_tokens": 82931844.0,
      "step": 6723
    },
    {
      "entropy": 1.1646762192249298,
      "epoch": 3.540810953133228,
      "grad_norm": 0.29397276043891907,
      "learning_rate": 0.00011269677564666565,
      "loss": 0.1737130582332611,
      "mean_token_accuracy": 0.9258291274309158,
      "num_tokens": 82944180.0,
      "step": 6724
    },
    {
      "entropy": 1.2043721377849579,
      "epoch": 3.541337546076883,
      "grad_norm": 0.30750924348831177,
      "learning_rate": 0.00011267485712573818,
      "loss": 0.18758301436901093,
      "mean_token_accuracy": 0.9227220118045807,
      "num_tokens": 82956516.0,
      "step": 6725
    },
    {
      "entropy": 1.205798625946045,
      "epoch": 3.541864139020537,
      "grad_norm": 0.30392226576805115,
      "learning_rate": 0.00011265293844601997,
      "loss": 0.17270836234092712,
      "mean_token_accuracy": 0.9301195740699768,
      "num_tokens": 82968852.0,
      "step": 6726
    },
    {
      "entropy": 1.1954290270805359,
      "epoch": 3.5423907319641916,
      "grad_norm": 0.28532564640045166,
      "learning_rate": 0.00011263101960881228,
      "loss": 0.17308568954467773,
      "mean_token_accuracy": 0.9245552122592926,
      "num_tokens": 82981188.0,
      "step": 6727
    },
    {
      "entropy": 1.232110172510147,
      "epoch": 3.5429173249078465,
      "grad_norm": 0.33423352241516113,
      "learning_rate": 0.00011260910061541637,
      "loss": 0.18927906453609467,
      "mean_token_accuracy": 0.9178860485553741,
      "num_tokens": 82993524.0,
      "step": 6728
    },
    {
      "entropy": 1.2852757275104523,
      "epoch": 3.543443917851501,
      "grad_norm": 0.2983771562576294,
      "learning_rate": 0.0001125871814671333,
      "loss": 0.17770478129386902,
      "mean_token_accuracy": 0.9284773916006088,
      "num_tokens": 83005860.0,
      "step": 6729
    },
    {
      "entropy": 1.2685899138450623,
      "epoch": 3.5439705107951553,
      "grad_norm": 0.29142412543296814,
      "learning_rate": 0.00011256526216526433,
      "loss": 0.1880548745393753,
      "mean_token_accuracy": 0.923332005739212,
      "num_tokens": 83018196.0,
      "step": 6730
    },
    {
      "entropy": 1.2598896324634552,
      "epoch": 3.5444971037388098,
      "grad_norm": 0.2949274480342865,
      "learning_rate": 0.0001125433427111107,
      "loss": 0.17068499326705933,
      "mean_token_accuracy": 0.9293069690465927,
      "num_tokens": 83030532.0,
      "step": 6731
    },
    {
      "entropy": 1.2281588017940521,
      "epoch": 3.545023696682464,
      "grad_norm": 0.2889408469200134,
      "learning_rate": 0.00011252142310597364,
      "loss": 0.17698845267295837,
      "mean_token_accuracy": 0.9286070317029953,
      "num_tokens": 83042868.0,
      "step": 6732
    },
    {
      "entropy": 1.269063025712967,
      "epoch": 3.545550289626119,
      "grad_norm": 0.29950693249702454,
      "learning_rate": 0.00011249950335115443,
      "loss": 0.16892579197883606,
      "mean_token_accuracy": 0.9297320991754532,
      "num_tokens": 83055204.0,
      "step": 6733
    },
    {
      "entropy": 1.2776862978935242,
      "epoch": 3.5460768825697735,
      "grad_norm": 0.31639254093170166,
      "learning_rate": 0.00011247758344795423,
      "loss": 0.19716720283031464,
      "mean_token_accuracy": 0.9158044904470444,
      "num_tokens": 83067540.0,
      "step": 6734
    },
    {
      "entropy": 1.317240983247757,
      "epoch": 3.546603475513428,
      "grad_norm": 0.28840455412864685,
      "learning_rate": 0.00011245566339767435,
      "loss": 0.18107880651950836,
      "mean_token_accuracy": 0.9243495315313339,
      "num_tokens": 83079876.0,
      "step": 6735
    },
    {
      "entropy": 1.2626528143882751,
      "epoch": 3.547130068457083,
      "grad_norm": 0.2795928716659546,
      "learning_rate": 0.00011243374320161607,
      "loss": 0.16566278040409088,
      "mean_token_accuracy": 0.9298494160175323,
      "num_tokens": 83092212.0,
      "step": 6736
    },
    {
      "entropy": 1.2726233303546906,
      "epoch": 3.547656661400737,
      "grad_norm": 0.315793514251709,
      "learning_rate": 0.0001124118228610806,
      "loss": 0.1717158704996109,
      "mean_token_accuracy": 0.9273900836706161,
      "num_tokens": 83104548.0,
      "step": 6737
    },
    {
      "entropy": 1.3019500076770782,
      "epoch": 3.5481832543443916,
      "grad_norm": 0.3635459244251251,
      "learning_rate": 0.0001123899023773693,
      "loss": 0.18998388946056366,
      "mean_token_accuracy": 0.919920027256012,
      "num_tokens": 83116884.0,
      "step": 6738
    },
    {
      "entropy": 1.2970272600650787,
      "epoch": 3.5487098472880465,
      "grad_norm": 0.2943572998046875,
      "learning_rate": 0.0001123679817517834,
      "loss": 0.18436715006828308,
      "mean_token_accuracy": 0.9233886748552322,
      "num_tokens": 83129220.0,
      "step": 6739
    },
    {
      "entropy": 1.2755154073238373,
      "epoch": 3.549236440231701,
      "grad_norm": 0.3052658140659332,
      "learning_rate": 0.00011234606098562424,
      "loss": 0.18061406910419464,
      "mean_token_accuracy": 0.9223170429468155,
      "num_tokens": 83141556.0,
      "step": 6740
    },
    {
      "entropy": 1.2355510890483856,
      "epoch": 3.5497630331753554,
      "grad_norm": 0.30959102511405945,
      "learning_rate": 0.0001123241400801931,
      "loss": 0.18960706889629364,
      "mean_token_accuracy": 0.9217345267534256,
      "num_tokens": 83153892.0,
      "step": 6741
    },
    {
      "entropy": 1.2396555840969086,
      "epoch": 3.5502896261190102,
      "grad_norm": 0.29119154810905457,
      "learning_rate": 0.0001123022190367913,
      "loss": 0.17762398719787598,
      "mean_token_accuracy": 0.9257684648036957,
      "num_tokens": 83166228.0,
      "step": 6742
    },
    {
      "entropy": 1.1585346460342407,
      "epoch": 3.5508162190626646,
      "grad_norm": 0.2666080892086029,
      "learning_rate": 0.00011228029785672023,
      "loss": 0.1687043160200119,
      "mean_token_accuracy": 0.9259619563817978,
      "num_tokens": 83178564.0,
      "step": 6743
    },
    {
      "entropy": 1.1754328608512878,
      "epoch": 3.551342812006319,
      "grad_norm": 0.35181063413619995,
      "learning_rate": 0.0001122583765412811,
      "loss": 0.18153096735477448,
      "mean_token_accuracy": 0.9212551862001419,
      "num_tokens": 83190900.0,
      "step": 6744
    },
    {
      "entropy": 1.1387519538402557,
      "epoch": 3.551869404949974,
      "grad_norm": 0.24666999280452728,
      "learning_rate": 0.00011223645509177535,
      "loss": 0.14854127168655396,
      "mean_token_accuracy": 0.9388223141431808,
      "num_tokens": 83203236.0,
      "step": 6745
    },
    {
      "entropy": 1.122187227010727,
      "epoch": 3.5523959978936284,
      "grad_norm": 0.2791399657726288,
      "learning_rate": 0.00011221453350950431,
      "loss": 0.1779361367225647,
      "mean_token_accuracy": 0.9246224761009216,
      "num_tokens": 83215572.0,
      "step": 6746
    },
    {
      "entropy": 1.1300030052661896,
      "epoch": 3.552922590837283,
      "grad_norm": 0.29074087738990784,
      "learning_rate": 0.00011219261179576928,
      "loss": 0.1794925183057785,
      "mean_token_accuracy": 0.9241160750389099,
      "num_tokens": 83227908.0,
      "step": 6747
    },
    {
      "entropy": 1.1947501301765442,
      "epoch": 3.553449183780937,
      "grad_norm": 0.3051902949810028,
      "learning_rate": 0.00011217068995187172,
      "loss": 0.18434175848960876,
      "mean_token_accuracy": 0.9224974811077118,
      "num_tokens": 83240244.0,
      "step": 6748
    },
    {
      "entropy": 1.1320153772830963,
      "epoch": 3.5539757767245916,
      "grad_norm": 0.2957667112350464,
      "learning_rate": 0.00011214876797911292,
      "loss": 0.18279367685317993,
      "mean_token_accuracy": 0.9267748147249222,
      "num_tokens": 83252580.0,
      "step": 6749
    },
    {
      "entropy": 1.1726331114768982,
      "epoch": 3.5545023696682465,
      "grad_norm": 0.27514028549194336,
      "learning_rate": 0.00011212684587879427,
      "loss": 0.1648423969745636,
      "mean_token_accuracy": 0.9331820756196976,
      "num_tokens": 83264916.0,
      "step": 6750
    },
    {
      "entropy": 1.1044942140579224,
      "epoch": 3.555028962611901,
      "grad_norm": 0.2936144769191742,
      "learning_rate": 0.00011210492365221719,
      "loss": 0.18164953589439392,
      "mean_token_accuracy": 0.9242898970842361,
      "num_tokens": 83277252.0,
      "step": 6751
    },
    {
      "entropy": 1.1623157262802124,
      "epoch": 3.5555555555555554,
      "grad_norm": 0.2973748743534088,
      "learning_rate": 0.00011208300130068308,
      "loss": 0.1812804937362671,
      "mean_token_accuracy": 0.9253558069467545,
      "num_tokens": 83289588.0,
      "step": 6752
    },
    {
      "entropy": 1.2013108432292938,
      "epoch": 3.5560821484992102,
      "grad_norm": 0.3230435848236084,
      "learning_rate": 0.00011206107882549333,
      "loss": 0.19521446526050568,
      "mean_token_accuracy": 0.913422480225563,
      "num_tokens": 83301924.0,
      "step": 6753
    },
    {
      "entropy": 1.1237129867076874,
      "epoch": 3.5566087414428647,
      "grad_norm": 0.25985023379325867,
      "learning_rate": 0.00011203915622794934,
      "loss": 0.16712124645709991,
      "mean_token_accuracy": 0.9284080117940903,
      "num_tokens": 83314260.0,
      "step": 6754
    },
    {
      "entropy": 1.1246764659881592,
      "epoch": 3.557135334386519,
      "grad_norm": 0.2639875113964081,
      "learning_rate": 0.00011201723350935249,
      "loss": 0.1713811457157135,
      "mean_token_accuracy": 0.92882439494133,
      "num_tokens": 83326596.0,
      "step": 6755
    },
    {
      "entropy": 1.1077988743782043,
      "epoch": 3.557661927330174,
      "grad_norm": 0.27503833174705505,
      "learning_rate": 0.00011199531067100426,
      "loss": 0.17916113138198853,
      "mean_token_accuracy": 0.9328770935535431,
      "num_tokens": 83338932.0,
      "step": 6756
    },
    {
      "entropy": 1.1336831152439117,
      "epoch": 3.5581885202738284,
      "grad_norm": 0.3099113702774048,
      "learning_rate": 0.00011197338771420608,
      "loss": 0.18660598993301392,
      "mean_token_accuracy": 0.9220423400402069,
      "num_tokens": 83351268.0,
      "step": 6757
    },
    {
      "entropy": 1.1971750855445862,
      "epoch": 3.558715113217483,
      "grad_norm": 0.27214622497558594,
      "learning_rate": 0.00011195146464025936,
      "loss": 0.17578336596488953,
      "mean_token_accuracy": 0.9273701459169388,
      "num_tokens": 83363604.0,
      "step": 6758
    },
    {
      "entropy": 1.1493057310581207,
      "epoch": 3.5592417061611377,
      "grad_norm": 0.29546862840652466,
      "learning_rate": 0.00011192954145046556,
      "loss": 0.1825748085975647,
      "mean_token_accuracy": 0.9244335293769836,
      "num_tokens": 83375940.0,
      "step": 6759
    },
    {
      "entropy": 1.1059208512306213,
      "epoch": 3.559768299104792,
      "grad_norm": 0.26903000473976135,
      "learning_rate": 0.00011190761814612616,
      "loss": 0.18101567029953003,
      "mean_token_accuracy": 0.9220552444458008,
      "num_tokens": 83388276.0,
      "step": 6760
    },
    {
      "entropy": 1.150664508342743,
      "epoch": 3.5602948920484465,
      "grad_norm": 0.30451393127441406,
      "learning_rate": 0.00011188569472854256,
      "loss": 0.20869070291519165,
      "mean_token_accuracy": 0.9107763916254044,
      "num_tokens": 83400612.0,
      "step": 6761
    },
    {
      "entropy": 1.1247061789035797,
      "epoch": 3.5608214849921014,
      "grad_norm": 0.28411510586738586,
      "learning_rate": 0.00011186377119901625,
      "loss": 0.1772649884223938,
      "mean_token_accuracy": 0.922274649143219,
      "num_tokens": 83412948.0,
      "step": 6762
    },
    {
      "entropy": 1.091322273015976,
      "epoch": 3.561348077935756,
      "grad_norm": 0.30867794156074524,
      "learning_rate": 0.00011184184755884872,
      "loss": 0.20318832993507385,
      "mean_token_accuracy": 0.9137501269578934,
      "num_tokens": 83425284.0,
      "step": 6763
    },
    {
      "entropy": 1.1086713373661041,
      "epoch": 3.5618746708794102,
      "grad_norm": 0.28916677832603455,
      "learning_rate": 0.00011181992380934144,
      "loss": 0.18094968795776367,
      "mean_token_accuracy": 0.9253650158643723,
      "num_tokens": 83437620.0,
      "step": 6764
    },
    {
      "entropy": 1.0924776494503021,
      "epoch": 3.5624012638230647,
      "grad_norm": 0.2812732756137848,
      "learning_rate": 0.00011179799995179585,
      "loss": 0.1750286966562271,
      "mean_token_accuracy": 0.9251724034547806,
      "num_tokens": 83449956.0,
      "step": 6765
    },
    {
      "entropy": 1.0866563320159912,
      "epoch": 3.562927856766719,
      "grad_norm": 0.30919456481933594,
      "learning_rate": 0.00011177607598751354,
      "loss": 0.1673109233379364,
      "mean_token_accuracy": 0.9293659329414368,
      "num_tokens": 83462292.0,
      "step": 6766
    },
    {
      "entropy": 1.0872103571891785,
      "epoch": 3.563454449710374,
      "grad_norm": 0.2803879976272583,
      "learning_rate": 0.00011175415191779593,
      "loss": 0.17549145221710205,
      "mean_token_accuracy": 0.9249834716320038,
      "num_tokens": 83474628.0,
      "step": 6767
    },
    {
      "entropy": 1.1090264022350311,
      "epoch": 3.5639810426540284,
      "grad_norm": 0.295563668012619,
      "learning_rate": 0.00011173222774394455,
      "loss": 0.19598843157291412,
      "mean_token_accuracy": 0.9154873639345169,
      "num_tokens": 83486964.0,
      "step": 6768
    },
    {
      "entropy": 1.0583168864250183,
      "epoch": 3.564507635597683,
      "grad_norm": 0.2749026417732239,
      "learning_rate": 0.00011171030346726088,
      "loss": 0.17145390808582306,
      "mean_token_accuracy": 0.9302745312452316,
      "num_tokens": 83499300.0,
      "step": 6769
    },
    {
      "entropy": 1.0702010095119476,
      "epoch": 3.5650342285413377,
      "grad_norm": 0.2993081510066986,
      "learning_rate": 0.00011168837908904648,
      "loss": 0.18920329213142395,
      "mean_token_accuracy": 0.9198135733604431,
      "num_tokens": 83511636.0,
      "step": 6770
    },
    {
      "entropy": 1.0419896095991135,
      "epoch": 3.565560821484992,
      "grad_norm": 0.31577932834625244,
      "learning_rate": 0.00011166645461060285,
      "loss": 0.18755492568016052,
      "mean_token_accuracy": 0.9221476912498474,
      "num_tokens": 83523972.0,
      "step": 6771
    },
    {
      "entropy": 1.0922668874263763,
      "epoch": 3.5660874144286465,
      "grad_norm": 0.31884366273880005,
      "learning_rate": 0.00011164453003323152,
      "loss": 0.191270649433136,
      "mean_token_accuracy": 0.9214844703674316,
      "num_tokens": 83536308.0,
      "step": 6772
    },
    {
      "entropy": 1.0515078604221344,
      "epoch": 3.5666140073723014,
      "grad_norm": 0.27819815278053284,
      "learning_rate": 0.00011162260535823404,
      "loss": 0.17676641047000885,
      "mean_token_accuracy": 0.9244179129600525,
      "num_tokens": 83548644.0,
      "step": 6773
    },
    {
      "entropy": 1.076279640197754,
      "epoch": 3.567140600315956,
      "grad_norm": 0.3012269139289856,
      "learning_rate": 0.00011160068058691193,
      "loss": 0.17410975694656372,
      "mean_token_accuracy": 0.9300138354301453,
      "num_tokens": 83560980.0,
      "step": 6774
    },
    {
      "entropy": 1.0549605637788773,
      "epoch": 3.5676671932596102,
      "grad_norm": 0.2836728096008301,
      "learning_rate": 0.00011157875572056673,
      "loss": 0.177901491522789,
      "mean_token_accuracy": 0.9266462922096252,
      "num_tokens": 83573316.0,
      "step": 6775
    },
    {
      "entropy": 1.1353142261505127,
      "epoch": 3.568193786203265,
      "grad_norm": 0.28448861837387085,
      "learning_rate": 0.00011155683076050003,
      "loss": 0.17505420744419098,
      "mean_token_accuracy": 0.9255066215991974,
      "num_tokens": 83585652.0,
      "step": 6776
    },
    {
      "entropy": 1.0923458933830261,
      "epoch": 3.5687203791469195,
      "grad_norm": 0.29703381657600403,
      "learning_rate": 0.00011153490570801333,
      "loss": 0.1849551647901535,
      "mean_token_accuracy": 0.9206576198339462,
      "num_tokens": 83597988.0,
      "step": 6777
    },
    {
      "entropy": 1.1172213554382324,
      "epoch": 3.569246972090574,
      "grad_norm": 0.30837827920913696,
      "learning_rate": 0.00011151298056440825,
      "loss": 0.19131802022457123,
      "mean_token_accuracy": 0.9169535785913467,
      "num_tokens": 83610324.0,
      "step": 6778
    },
    {
      "entropy": 1.1013193428516388,
      "epoch": 3.5697735650342284,
      "grad_norm": 0.26760244369506836,
      "learning_rate": 0.0001114910553309863,
      "loss": 0.16245457530021667,
      "mean_token_accuracy": 0.9318066239356995,
      "num_tokens": 83622660.0,
      "step": 6779
    },
    {
      "entropy": 1.069222867488861,
      "epoch": 3.5703001579778832,
      "grad_norm": 0.2659068703651428,
      "learning_rate": 0.0001114691300090491,
      "loss": 0.16208085417747498,
      "mean_token_accuracy": 0.9289850294589996,
      "num_tokens": 83634996.0,
      "step": 6780
    },
    {
      "entropy": 1.0956320464611053,
      "epoch": 3.5708267509215377,
      "grad_norm": 0.319746196269989,
      "learning_rate": 0.00011144720459989824,
      "loss": 0.19753004610538483,
      "mean_token_accuracy": 0.919308990240097,
      "num_tokens": 83647332.0,
      "step": 6781
    },
    {
      "entropy": 1.081997960805893,
      "epoch": 3.571353343865192,
      "grad_norm": 0.2700563371181488,
      "learning_rate": 0.00011142527910483527,
      "loss": 0.18001338839530945,
      "mean_token_accuracy": 0.9277038872241974,
      "num_tokens": 83659668.0,
      "step": 6782
    },
    {
      "entropy": 1.0502025187015533,
      "epoch": 3.5718799368088465,
      "grad_norm": 0.27929866313934326,
      "learning_rate": 0.00011140335352516179,
      "loss": 0.1719617396593094,
      "mean_token_accuracy": 0.9303514063358307,
      "num_tokens": 83672004.0,
      "step": 6783
    },
    {
      "entropy": 1.0725300312042236,
      "epoch": 3.5724065297525014,
      "grad_norm": 0.3040179908275604,
      "learning_rate": 0.00011138142786217936,
      "loss": 0.19116222858428955,
      "mean_token_accuracy": 0.9210270047187805,
      "num_tokens": 83684340.0,
      "step": 6784
    },
    {
      "entropy": 1.0981976985931396,
      "epoch": 3.572933122696156,
      "grad_norm": 0.2722747325897217,
      "learning_rate": 0.00011135950211718961,
      "loss": 0.1681644171476364,
      "mean_token_accuracy": 0.9292253851890564,
      "num_tokens": 83696676.0,
      "step": 6785
    },
    {
      "entropy": 1.1399394571781158,
      "epoch": 3.5734597156398102,
      "grad_norm": 0.30738040804862976,
      "learning_rate": 0.00011133757629149417,
      "loss": 0.1938926875591278,
      "mean_token_accuracy": 0.9183665215969086,
      "num_tokens": 83709012.0,
      "step": 6786
    },
    {
      "entropy": 1.0733324587345123,
      "epoch": 3.573986308583465,
      "grad_norm": 0.2664124369621277,
      "learning_rate": 0.00011131565038639462,
      "loss": 0.16366492211818695,
      "mean_token_accuracy": 0.9280208200216293,
      "num_tokens": 83721348.0,
      "step": 6787
    },
    {
      "entropy": 1.070346713066101,
      "epoch": 3.5745129015271195,
      "grad_norm": 0.2684425413608551,
      "learning_rate": 0.00011129372440319256,
      "loss": 0.17055614292621613,
      "mean_token_accuracy": 0.9309419095516205,
      "num_tokens": 83733684.0,
      "step": 6788
    },
    {
      "entropy": 1.1445283591747284,
      "epoch": 3.575039494470774,
      "grad_norm": 0.30912500619888306,
      "learning_rate": 0.00011127179834318964,
      "loss": 0.18295526504516602,
      "mean_token_accuracy": 0.9247478097677231,
      "num_tokens": 83746020.0,
      "step": 6789
    },
    {
      "entropy": 1.1068629920482635,
      "epoch": 3.575566087414429,
      "grad_norm": 0.3023226857185364,
      "learning_rate": 0.00011124987220768743,
      "loss": 0.18361878395080566,
      "mean_token_accuracy": 0.9251338094472885,
      "num_tokens": 83758356.0,
      "step": 6790
    },
    {
      "entropy": 1.0960930585861206,
      "epoch": 3.5760926803580833,
      "grad_norm": 0.32193610072135925,
      "learning_rate": 0.00011122794599798765,
      "loss": 0.1902855485677719,
      "mean_token_accuracy": 0.9249085038900375,
      "num_tokens": 83770692.0,
      "step": 6791
    },
    {
      "entropy": 1.102184534072876,
      "epoch": 3.5766192733017377,
      "grad_norm": 0.30219942331314087,
      "learning_rate": 0.00011120601971539184,
      "loss": 0.18872946500778198,
      "mean_token_accuracy": 0.9205964207649231,
      "num_tokens": 83783028.0,
      "step": 6792
    },
    {
      "entropy": 1.0089440494775772,
      "epoch": 3.5771458662453925,
      "grad_norm": 0.2962239980697632,
      "learning_rate": 0.00011118409336120163,
      "loss": 0.17877568304538727,
      "mean_token_accuracy": 0.9258351624011993,
      "num_tokens": 83795364.0,
      "step": 6793
    },
    {
      "entropy": 1.0831408500671387,
      "epoch": 3.577672459189047,
      "grad_norm": 0.3217921853065491,
      "learning_rate": 0.00011116216693671875,
      "loss": 0.19278764724731445,
      "mean_token_accuracy": 0.9192754030227661,
      "num_tokens": 83807700.0,
      "step": 6794
    },
    {
      "entropy": 1.0765918493270874,
      "epoch": 3.5781990521327014,
      "grad_norm": 0.3078082203865051,
      "learning_rate": 0.00011114024044324478,
      "loss": 0.18732166290283203,
      "mean_token_accuracy": 0.9194670766592026,
      "num_tokens": 83820036.0,
      "step": 6795
    },
    {
      "entropy": 1.0819332301616669,
      "epoch": 3.578725645076356,
      "grad_norm": 0.3063753545284271,
      "learning_rate": 0.00011111831388208138,
      "loss": 0.18549399077892303,
      "mean_token_accuracy": 0.9213854074478149,
      "num_tokens": 83832372.0,
      "step": 6796
    },
    {
      "entropy": 1.051520049571991,
      "epoch": 3.5792522380200107,
      "grad_norm": 0.30566105246543884,
      "learning_rate": 0.0001110963872545302,
      "loss": 0.18486547470092773,
      "mean_token_accuracy": 0.9276604950428009,
      "num_tokens": 83844708.0,
      "step": 6797
    },
    {
      "entropy": 1.0421344339847565,
      "epoch": 3.579778830963665,
      "grad_norm": 0.27134665846824646,
      "learning_rate": 0.00011107446056189292,
      "loss": 0.16037064790725708,
      "mean_token_accuracy": 0.9303885400295258,
      "num_tokens": 83857044.0,
      "step": 6798
    },
    {
      "entropy": 1.0744150280952454,
      "epoch": 3.5803054239073195,
      "grad_norm": 0.29975277185440063,
      "learning_rate": 0.00011105253380547116,
      "loss": 0.17214947938919067,
      "mean_token_accuracy": 0.9243510961532593,
      "num_tokens": 83869380.0,
      "step": 6799
    },
    {
      "entropy": 1.0933693647384644,
      "epoch": 3.580832016850974,
      "grad_norm": 0.3158538341522217,
      "learning_rate": 0.00011103060698656657,
      "loss": 0.17609092593193054,
      "mean_token_accuracy": 0.9244079291820526,
      "num_tokens": 83881716.0,
      "step": 6800
    },
    {
      "entropy": 1.0673887133598328,
      "epoch": 3.581358609794629,
      "grad_norm": 0.31816163659095764,
      "learning_rate": 0.00011100868010648092,
      "loss": 0.18575707077980042,
      "mean_token_accuracy": 0.922477513551712,
      "num_tokens": 83894052.0,
      "step": 6801
    },
    {
      "entropy": 1.1075579822063446,
      "epoch": 3.5818852027382833,
      "grad_norm": 0.3141384720802307,
      "learning_rate": 0.00011098675316651576,
      "loss": 0.19477181136608124,
      "mean_token_accuracy": 0.9194188117980957,
      "num_tokens": 83906388.0,
      "step": 6802
    },
    {
      "entropy": 1.083218827843666,
      "epoch": 3.5824117956819377,
      "grad_norm": 0.27618029713630676,
      "learning_rate": 0.00011096482616797284,
      "loss": 0.16593825817108154,
      "mean_token_accuracy": 0.9309423565864563,
      "num_tokens": 83918724.0,
      "step": 6803
    },
    {
      "entropy": 1.099007934331894,
      "epoch": 3.5829383886255926,
      "grad_norm": 0.2944222688674927,
      "learning_rate": 0.00011094289911215377,
      "loss": 0.18428628146648407,
      "mean_token_accuracy": 0.9252902120351791,
      "num_tokens": 83931060.0,
      "step": 6804
    },
    {
      "entropy": 1.068886861205101,
      "epoch": 3.583464981569247,
      "grad_norm": 0.28556880354881287,
      "learning_rate": 0.00011092097200036032,
      "loss": 0.17528092861175537,
      "mean_token_accuracy": 0.9258764088153839,
      "num_tokens": 83943396.0,
      "step": 6805
    },
    {
      "entropy": 1.0801942050457,
      "epoch": 3.5839915745129014,
      "grad_norm": 0.2890860140323639,
      "learning_rate": 0.00011089904483389415,
      "loss": 0.17374573647975922,
      "mean_token_accuracy": 0.9233235269784927,
      "num_tokens": 83955732.0,
      "step": 6806
    },
    {
      "entropy": 1.1448765099048615,
      "epoch": 3.5845181674565563,
      "grad_norm": 0.30026209354400635,
      "learning_rate": 0.00011087711761405688,
      "loss": 0.18570362031459808,
      "mean_token_accuracy": 0.9220779836177826,
      "num_tokens": 83968068.0,
      "step": 6807
    },
    {
      "entropy": 1.0914206057786942,
      "epoch": 3.5850447604002107,
      "grad_norm": 0.2962283492088318,
      "learning_rate": 0.00011085519034215027,
      "loss": 0.18485112488269806,
      "mean_token_accuracy": 0.9181759804487228,
      "num_tokens": 83980404.0,
      "step": 6808
    },
    {
      "entropy": 1.117035686969757,
      "epoch": 3.585571353343865,
      "grad_norm": 0.29223090410232544,
      "learning_rate": 0.00011083326301947599,
      "loss": 0.18125474452972412,
      "mean_token_accuracy": 0.9249403476715088,
      "num_tokens": 83992740.0,
      "step": 6809
    },
    {
      "entropy": 1.115389347076416,
      "epoch": 3.58609794628752,
      "grad_norm": 0.3016378581523895,
      "learning_rate": 0.00011081133564733572,
      "loss": 0.1807543933391571,
      "mean_token_accuracy": 0.9210816323757172,
      "num_tokens": 84005076.0,
      "step": 6810
    },
    {
      "entropy": 1.0712342858314514,
      "epoch": 3.5866245392311744,
      "grad_norm": 0.2989068925380707,
      "learning_rate": 0.00011078940822703121,
      "loss": 0.18486231565475464,
      "mean_token_accuracy": 0.9215465188026428,
      "num_tokens": 84017412.0,
      "step": 6811
    },
    {
      "entropy": 1.0903160274028778,
      "epoch": 3.587151132174829,
      "grad_norm": 0.27834755182266235,
      "learning_rate": 0.00011076748075986413,
      "loss": 0.18570701777935028,
      "mean_token_accuracy": 0.9223471432924271,
      "num_tokens": 84029748.0,
      "step": 6812
    },
    {
      "entropy": 1.0887775719165802,
      "epoch": 3.5876777251184833,
      "grad_norm": 0.3086070716381073,
      "learning_rate": 0.0001107455532471362,
      "loss": 0.2004110962152481,
      "mean_token_accuracy": 0.9159822016954422,
      "num_tokens": 84042084.0,
      "step": 6813
    },
    {
      "entropy": 1.0728232264518738,
      "epoch": 3.588204318062138,
      "grad_norm": 0.2820192873477936,
      "learning_rate": 0.0001107236256901491,
      "loss": 0.17461654543876648,
      "mean_token_accuracy": 0.925453782081604,
      "num_tokens": 84054420.0,
      "step": 6814
    },
    {
      "entropy": 1.0823879539966583,
      "epoch": 3.5887309110057926,
      "grad_norm": 0.2939518392086029,
      "learning_rate": 0.00011070169809020456,
      "loss": 0.1872849017381668,
      "mean_token_accuracy": 0.9165738672018051,
      "num_tokens": 84066756.0,
      "step": 6815
    },
    {
      "entropy": 1.1180018484592438,
      "epoch": 3.589257503949447,
      "grad_norm": 0.2900797128677368,
      "learning_rate": 0.00011067977044860432,
      "loss": 0.18645426630973816,
      "mean_token_accuracy": 0.9243784248828888,
      "num_tokens": 84079092.0,
      "step": 6816
    },
    {
      "entropy": 1.0934679508209229,
      "epoch": 3.5897840968931014,
      "grad_norm": 0.2787036895751953,
      "learning_rate": 0.00011065784276665005,
      "loss": 0.16626539826393127,
      "mean_token_accuracy": 0.9302684217691422,
      "num_tokens": 84091428.0,
      "step": 6817
    },
    {
      "entropy": 1.0777550637722015,
      "epoch": 3.5903106898367563,
      "grad_norm": 0.29463809728622437,
      "learning_rate": 0.00011063591504564348,
      "loss": 0.18691596388816833,
      "mean_token_accuracy": 0.922309547662735,
      "num_tokens": 84103764.0,
      "step": 6818
    },
    {
      "entropy": 1.0706429183483124,
      "epoch": 3.5908372827804107,
      "grad_norm": 0.29603153467178345,
      "learning_rate": 0.00011061398728688634,
      "loss": 0.18845272064208984,
      "mean_token_accuracy": 0.9194611012935638,
      "num_tokens": 84116100.0,
      "step": 6819
    },
    {
      "entropy": 1.0385090708732605,
      "epoch": 3.591363875724065,
      "grad_norm": 0.2972482442855835,
      "learning_rate": 0.00011059205949168037,
      "loss": 0.19239693880081177,
      "mean_token_accuracy": 0.9183678478002548,
      "num_tokens": 84128436.0,
      "step": 6820
    },
    {
      "entropy": 1.1171515882015228,
      "epoch": 3.59189046866772,
      "grad_norm": 0.2884787321090698,
      "learning_rate": 0.00011057013166132729,
      "loss": 0.17812785506248474,
      "mean_token_accuracy": 0.9246303290128708,
      "num_tokens": 84140772.0,
      "step": 6821
    },
    {
      "entropy": 1.0698346197605133,
      "epoch": 3.5924170616113744,
      "grad_norm": 0.27562543749809265,
      "learning_rate": 0.0001105482037971288,
      "loss": 0.16333125531673431,
      "mean_token_accuracy": 0.9276170432567596,
      "num_tokens": 84153108.0,
      "step": 6822
    },
    {
      "entropy": 1.105311542749405,
      "epoch": 3.592943654555029,
      "grad_norm": 0.3589690327644348,
      "learning_rate": 0.00011052627590038665,
      "loss": 0.19951340556144714,
      "mean_token_accuracy": 0.9194656014442444,
      "num_tokens": 84165444.0,
      "step": 6823
    },
    {
      "entropy": 1.056771695613861,
      "epoch": 3.5934702474986837,
      "grad_norm": 0.31134459376335144,
      "learning_rate": 0.00011050434797240257,
      "loss": 0.18938994407653809,
      "mean_token_accuracy": 0.9203962981700897,
      "num_tokens": 84177780.0,
      "step": 6824
    },
    {
      "entropy": 1.0244185477495193,
      "epoch": 3.593996840442338,
      "grad_norm": 0.31450948119163513,
      "learning_rate": 0.00011048242001447827,
      "loss": 0.20221704244613647,
      "mean_token_accuracy": 0.9144396930932999,
      "num_tokens": 84190116.0,
      "step": 6825
    },
    {
      "entropy": 1.0762320160865784,
      "epoch": 3.5945234333859926,
      "grad_norm": 0.27855023741722107,
      "learning_rate": 0.00011046049202791553,
      "loss": 0.1769963800907135,
      "mean_token_accuracy": 0.9241359382867813,
      "num_tokens": 84202452.0,
      "step": 6826
    },
    {
      "entropy": 1.0686099827289581,
      "epoch": 3.5950500263296474,
      "grad_norm": 0.3128301203250885,
      "learning_rate": 0.00011043856401401605,
      "loss": 0.1926872730255127,
      "mean_token_accuracy": 0.917589545249939,
      "num_tokens": 84214788.0,
      "step": 6827
    },
    {
      "entropy": 1.1129842102527618,
      "epoch": 3.595576619273302,
      "grad_norm": 0.2821250259876251,
      "learning_rate": 0.00011041663597408157,
      "loss": 0.18698513507843018,
      "mean_token_accuracy": 0.925433337688446,
      "num_tokens": 84227124.0,
      "step": 6828
    },
    {
      "entropy": 1.0851635932922363,
      "epoch": 3.5961032122169563,
      "grad_norm": 0.29759037494659424,
      "learning_rate": 0.00011039470790941384,
      "loss": 0.17480885982513428,
      "mean_token_accuracy": 0.9230799973011017,
      "num_tokens": 84239460.0,
      "step": 6829
    },
    {
      "entropy": 1.0806046724319458,
      "epoch": 3.5966298051606107,
      "grad_norm": 0.28347960114479065,
      "learning_rate": 0.00011037277982131463,
      "loss": 0.187611386179924,
      "mean_token_accuracy": 0.918888509273529,
      "num_tokens": 84251796.0,
      "step": 6830
    },
    {
      "entropy": 1.04977385699749,
      "epoch": 3.597156398104265,
      "grad_norm": 0.2821602523326874,
      "learning_rate": 0.00011035085171108565,
      "loss": 0.18557719886302948,
      "mean_token_accuracy": 0.9250794351100922,
      "num_tokens": 84264132.0,
      "step": 6831
    },
    {
      "entropy": 1.0975022614002228,
      "epoch": 3.59768299104792,
      "grad_norm": 0.29911336302757263,
      "learning_rate": 0.00011032892358002862,
      "loss": 0.17656078934669495,
      "mean_token_accuracy": 0.9215593487024307,
      "num_tokens": 84276468.0,
      "step": 6832
    },
    {
      "entropy": 1.1453993022441864,
      "epoch": 3.5982095839915744,
      "grad_norm": 0.2856663167476654,
      "learning_rate": 0.00011030699542944533,
      "loss": 0.17330417037010193,
      "mean_token_accuracy": 0.9261975884437561,
      "num_tokens": 84288804.0,
      "step": 6833
    },
    {
      "entropy": 1.0889165699481964,
      "epoch": 3.598736176935229,
      "grad_norm": 0.2788821756839752,
      "learning_rate": 0.00011028506726063752,
      "loss": 0.17634747922420502,
      "mean_token_accuracy": 0.9255717843770981,
      "num_tokens": 84301140.0,
      "step": 6834
    },
    {
      "entropy": 1.0964802503585815,
      "epoch": 3.5992627698788837,
      "grad_norm": 0.28933143615722656,
      "learning_rate": 0.00011026313907490693,
      "loss": 0.1901606321334839,
      "mean_token_accuracy": 0.9201119393110275,
      "num_tokens": 84313476.0,
      "step": 6835
    },
    {
      "entropy": 1.1324889659881592,
      "epoch": 3.599789362822538,
      "grad_norm": 0.2915611267089844,
      "learning_rate": 0.00011024121087355527,
      "loss": 0.1741735190153122,
      "mean_token_accuracy": 0.9331314861774445,
      "num_tokens": 84325812.0,
      "step": 6836
    },
    {
      "entropy": 1.0679437518119812,
      "epoch": 3.6003159557661926,
      "grad_norm": 0.3005986511707306,
      "learning_rate": 0.0001102192826578844,
      "loss": 0.18817023932933807,
      "mean_token_accuracy": 0.9224089980125427,
      "num_tokens": 84338148.0,
      "step": 6837
    },
    {
      "entropy": 1.0960251837968826,
      "epoch": 3.6008425487098474,
      "grad_norm": 0.2999762296676636,
      "learning_rate": 0.00011019735442919594,
      "loss": 0.18257468938827515,
      "mean_token_accuracy": 0.9245083183050156,
      "num_tokens": 84350484.0,
      "step": 6838
    },
    {
      "entropy": 1.0532985776662827,
      "epoch": 3.601369141653502,
      "grad_norm": 0.31091398000717163,
      "learning_rate": 0.00011017542618879166,
      "loss": 0.18717285990715027,
      "mean_token_accuracy": 0.9183699786663055,
      "num_tokens": 84362820.0,
      "step": 6839
    },
    {
      "entropy": 1.0669248700141907,
      "epoch": 3.6018957345971563,
      "grad_norm": 0.31256037950515747,
      "learning_rate": 0.00011015349793797345,
      "loss": 0.17956934869289398,
      "mean_token_accuracy": 0.9235763698816299,
      "num_tokens": 84375156.0,
      "step": 6840
    },
    {
      "entropy": 1.1000414490699768,
      "epoch": 3.602422327540811,
      "grad_norm": 0.3200954794883728,
      "learning_rate": 0.0001101315696780429,
      "loss": 0.18431906402111053,
      "mean_token_accuracy": 0.9205802828073502,
      "num_tokens": 84387492.0,
      "step": 6841
    },
    {
      "entropy": 1.0772568136453629,
      "epoch": 3.6029489204844656,
      "grad_norm": 0.2726421356201172,
      "learning_rate": 0.00011010964141030182,
      "loss": 0.16852417588233948,
      "mean_token_accuracy": 0.92500339448452,
      "num_tokens": 84399828.0,
      "step": 6842
    },
    {
      "entropy": 1.0504442304372787,
      "epoch": 3.60347551342812,
      "grad_norm": 0.2797470688819885,
      "learning_rate": 0.00011008771313605199,
      "loss": 0.17749673128128052,
      "mean_token_accuracy": 0.9265327602624893,
      "num_tokens": 84412164.0,
      "step": 6843
    },
    {
      "entropy": 1.098019003868103,
      "epoch": 3.604002106371775,
      "grad_norm": 0.3110448718070984,
      "learning_rate": 0.00011006578485659513,
      "loss": 0.18846482038497925,
      "mean_token_accuracy": 0.92154660820961,
      "num_tokens": 84424500.0,
      "step": 6844
    },
    {
      "entropy": 1.0312111526727676,
      "epoch": 3.6045286993154293,
      "grad_norm": 0.30218741297721863,
      "learning_rate": 0.00011004385657323301,
      "loss": 0.19102369248867035,
      "mean_token_accuracy": 0.9235749393701553,
      "num_tokens": 84436836.0,
      "step": 6845
    },
    {
      "entropy": 1.0944945514202118,
      "epoch": 3.6050552922590837,
      "grad_norm": 0.3308318853378296,
      "learning_rate": 0.0001100219282872674,
      "loss": 0.17298060655593872,
      "mean_token_accuracy": 0.928741455078125,
      "num_tokens": 84449172.0,
      "step": 6846
    },
    {
      "entropy": 1.0642427206039429,
      "epoch": 3.605581885202738,
      "grad_norm": 0.28731769323349,
      "learning_rate": 0.00011000000000000002,
      "loss": 0.1798696517944336,
      "mean_token_accuracy": 0.9239630848169327,
      "num_tokens": 84461508.0,
      "step": 6847
    },
    {
      "entropy": 1.123925805091858,
      "epoch": 3.6061084781463926,
      "grad_norm": 0.3246334195137024,
      "learning_rate": 0.00010997807171273265,
      "loss": 0.1852877140045166,
      "mean_token_accuracy": 0.9206551760435104,
      "num_tokens": 84473844.0,
      "step": 6848
    },
    {
      "entropy": 1.086002677679062,
      "epoch": 3.6066350710900474,
      "grad_norm": 0.281442254781723,
      "learning_rate": 0.000109956143426767,
      "loss": 0.1753401905298233,
      "mean_token_accuracy": 0.9272176027297974,
      "num_tokens": 84486180.0,
      "step": 6849
    },
    {
      "entropy": 1.0604727566242218,
      "epoch": 3.607161664033702,
      "grad_norm": 0.2794533967971802,
      "learning_rate": 0.00010993421514340489,
      "loss": 0.16479840874671936,
      "mean_token_accuracy": 0.9340286701917648,
      "num_tokens": 84498516.0,
      "step": 6850
    },
    {
      "entropy": 1.067774474620819,
      "epoch": 3.6076882569773563,
      "grad_norm": 0.2668810784816742,
      "learning_rate": 0.00010991228686394802,
      "loss": 0.1580381691455841,
      "mean_token_accuracy": 0.9306191056966782,
      "num_tokens": 84510852.0,
      "step": 6851
    },
    {
      "entropy": 1.0851312279701233,
      "epoch": 3.608214849921011,
      "grad_norm": 0.2869192659854889,
      "learning_rate": 0.00010989035858969818,
      "loss": 0.1692989021539688,
      "mean_token_accuracy": 0.9321866929531097,
      "num_tokens": 84523188.0,
      "step": 6852
    },
    {
      "entropy": 1.0702165961265564,
      "epoch": 3.6087414428646656,
      "grad_norm": 0.32713115215301514,
      "learning_rate": 0.00010986843032195713,
      "loss": 0.1853421926498413,
      "mean_token_accuracy": 0.9256032258272171,
      "num_tokens": 84535524.0,
      "step": 6853
    },
    {
      "entropy": 1.063175618648529,
      "epoch": 3.60926803580832,
      "grad_norm": 0.31001904606819153,
      "learning_rate": 0.00010984650206202657,
      "loss": 0.16657377779483795,
      "mean_token_accuracy": 0.9314382374286652,
      "num_tokens": 84547860.0,
      "step": 6854
    },
    {
      "entropy": 1.0820069909095764,
      "epoch": 3.609794628751975,
      "grad_norm": 0.2937212586402893,
      "learning_rate": 0.00010982457381120834,
      "loss": 0.17677544057369232,
      "mean_token_accuracy": 0.9208458214998245,
      "num_tokens": 84560196.0,
      "step": 6855
    },
    {
      "entropy": 1.0830555260181427,
      "epoch": 3.6103212216956293,
      "grad_norm": 0.32006511092185974,
      "learning_rate": 0.0001098026455708041,
      "loss": 0.19498637318611145,
      "mean_token_accuracy": 0.9215183705091476,
      "num_tokens": 84572532.0,
      "step": 6856
    },
    {
      "entropy": 1.0698910057544708,
      "epoch": 3.6108478146392837,
      "grad_norm": 0.28908658027648926,
      "learning_rate": 0.00010978071734211566,
      "loss": 0.19524383544921875,
      "mean_token_accuracy": 0.9204393923282623,
      "num_tokens": 84584868.0,
      "step": 6857
    },
    {
      "entropy": 1.0425672829151154,
      "epoch": 3.6113744075829386,
      "grad_norm": 0.28950250148773193,
      "learning_rate": 0.0001097587891264447,
      "loss": 0.1864977329969406,
      "mean_token_accuracy": 0.9201027452945709,
      "num_tokens": 84597204.0,
      "step": 6858
    },
    {
      "entropy": 1.1237571835517883,
      "epoch": 3.611901000526593,
      "grad_norm": 0.30378156900405884,
      "learning_rate": 0.0001097368609250931,
      "loss": 0.17740464210510254,
      "mean_token_accuracy": 0.9226698577404022,
      "num_tokens": 84609540.0,
      "step": 6859
    },
    {
      "entropy": 1.0304887294769287,
      "epoch": 3.6124275934702474,
      "grad_norm": 0.28008705377578735,
      "learning_rate": 0.00010971493273936251,
      "loss": 0.19065578281879425,
      "mean_token_accuracy": 0.9184155315160751,
      "num_tokens": 84621876.0,
      "step": 6860
    },
    {
      "entropy": 1.0502332746982574,
      "epoch": 3.6129541864139023,
      "grad_norm": 0.3209384083747864,
      "learning_rate": 0.00010969300457055469,
      "loss": 0.20223379135131836,
      "mean_token_accuracy": 0.9163955748081207,
      "num_tokens": 84634212.0,
      "step": 6861
    },
    {
      "entropy": 1.0601438879966736,
      "epoch": 3.6134807793575567,
      "grad_norm": 0.32106125354766846,
      "learning_rate": 0.00010967107641997141,
      "loss": 0.184079110622406,
      "mean_token_accuracy": 0.9217220991849899,
      "num_tokens": 84646548.0,
      "step": 6862
    },
    {
      "entropy": 1.071303516626358,
      "epoch": 3.614007372301211,
      "grad_norm": 0.28086385130882263,
      "learning_rate": 0.0001096491482889144,
      "loss": 0.1744384467601776,
      "mean_token_accuracy": 0.9287065863609314,
      "num_tokens": 84658884.0,
      "step": 6863
    },
    {
      "entropy": 1.091096580028534,
      "epoch": 3.6145339652448656,
      "grad_norm": 0.28602415323257446,
      "learning_rate": 0.00010962722017868541,
      "loss": 0.19257156550884247,
      "mean_token_accuracy": 0.9204146265983582,
      "num_tokens": 84671220.0,
      "step": 6864
    },
    {
      "entropy": 1.0848321914672852,
      "epoch": 3.61506055818852,
      "grad_norm": 0.2787522077560425,
      "learning_rate": 0.00010960529209058617,
      "loss": 0.18023037910461426,
      "mean_token_accuracy": 0.9271013885736465,
      "num_tokens": 84683556.0,
      "step": 6865
    },
    {
      "entropy": 1.1429481208324432,
      "epoch": 3.615587151132175,
      "grad_norm": 0.30959388613700867,
      "learning_rate": 0.00010958336402591844,
      "loss": 0.1963968276977539,
      "mean_token_accuracy": 0.9185330718755722,
      "num_tokens": 84695892.0,
      "step": 6866
    },
    {
      "entropy": 1.1271454393863678,
      "epoch": 3.6161137440758293,
      "grad_norm": 0.29112374782562256,
      "learning_rate": 0.00010956143598598399,
      "loss": 0.18528416752815247,
      "mean_token_accuracy": 0.9206521213054657,
      "num_tokens": 84708228.0,
      "step": 6867
    },
    {
      "entropy": 1.081824392080307,
      "epoch": 3.6166403370194837,
      "grad_norm": 0.2598550617694855,
      "learning_rate": 0.0001095395079720845,
      "loss": 0.16786476969718933,
      "mean_token_accuracy": 0.9285723119974136,
      "num_tokens": 84720564.0,
      "step": 6868
    },
    {
      "entropy": 1.068942368030548,
      "epoch": 3.6171669299631386,
      "grad_norm": 0.28704607486724854,
      "learning_rate": 0.00010951757998552174,
      "loss": 0.19206884503364563,
      "mean_token_accuracy": 0.9195718169212341,
      "num_tokens": 84732900.0,
      "step": 6869
    },
    {
      "entropy": 1.0923133790493011,
      "epoch": 3.617693522906793,
      "grad_norm": 0.27027350664138794,
      "learning_rate": 0.00010949565202759749,
      "loss": 0.17678622901439667,
      "mean_token_accuracy": 0.9282222241163254,
      "num_tokens": 84745236.0,
      "step": 6870
    },
    {
      "entropy": 1.155369907617569,
      "epoch": 3.6182201158504474,
      "grad_norm": 0.2801276743412018,
      "learning_rate": 0.0001094737240996134,
      "loss": 0.1720304787158966,
      "mean_token_accuracy": 0.9314807504415512,
      "num_tokens": 84757572.0,
      "step": 6871
    },
    {
      "entropy": 1.0734200477600098,
      "epoch": 3.6187467087941023,
      "grad_norm": 0.2531544864177704,
      "learning_rate": 0.0001094517962028712,
      "loss": 0.17671561241149902,
      "mean_token_accuracy": 0.9305606037378311,
      "num_tokens": 84769908.0,
      "step": 6872
    },
    {
      "entropy": 1.0713300704956055,
      "epoch": 3.6192733017377567,
      "grad_norm": 0.2790954113006592,
      "learning_rate": 0.00010942986833867272,
      "loss": 0.18087071180343628,
      "mean_token_accuracy": 0.9217016994953156,
      "num_tokens": 84782244.0,
      "step": 6873
    },
    {
      "entropy": 1.0994652807712555,
      "epoch": 3.619799894681411,
      "grad_norm": 0.3021692633628845,
      "learning_rate": 0.00010940794050831964,
      "loss": 0.18766359984874725,
      "mean_token_accuracy": 0.9191304445266724,
      "num_tokens": 84794580.0,
      "step": 6874
    },
    {
      "entropy": 1.0470281392335892,
      "epoch": 3.620326487625066,
      "grad_norm": 0.2776238024234772,
      "learning_rate": 0.00010938601271311368,
      "loss": 0.17030346393585205,
      "mean_token_accuracy": 0.9237184375524521,
      "num_tokens": 84806916.0,
      "step": 6875
    },
    {
      "entropy": 1.0921571552753448,
      "epoch": 3.6208530805687205,
      "grad_norm": 0.284041166305542,
      "learning_rate": 0.00010936408495435654,
      "loss": 0.16998879611492157,
      "mean_token_accuracy": 0.9288204312324524,
      "num_tokens": 84819252.0,
      "step": 6876
    },
    {
      "entropy": 1.0739878714084625,
      "epoch": 3.621379673512375,
      "grad_norm": 0.2863426208496094,
      "learning_rate": 0.00010934215723335,
      "loss": 0.1693313866853714,
      "mean_token_accuracy": 0.9308694303035736,
      "num_tokens": 84831588.0,
      "step": 6877
    },
    {
      "entropy": 1.072409301996231,
      "epoch": 3.6219062664560298,
      "grad_norm": 0.2869056463241577,
      "learning_rate": 0.00010932022955139572,
      "loss": 0.18741311132907867,
      "mean_token_accuracy": 0.9215927571058273,
      "num_tokens": 84843924.0,
      "step": 6878
    },
    {
      "entropy": 1.1229929625988007,
      "epoch": 3.622432859399684,
      "grad_norm": 0.3232174813747406,
      "learning_rate": 0.00010929830190979545,
      "loss": 0.19894477725028992,
      "mean_token_accuracy": 0.9182947725057602,
      "num_tokens": 84856260.0,
      "step": 6879
    },
    {
      "entropy": 1.0645553469657898,
      "epoch": 3.6229594523433386,
      "grad_norm": 0.26264068484306335,
      "learning_rate": 0.00010927637430985091,
      "loss": 0.16409020125865936,
      "mean_token_accuracy": 0.9330554455518723,
      "num_tokens": 84868596.0,
      "step": 6880
    },
    {
      "entropy": 1.0455782115459442,
      "epoch": 3.623486045286993,
      "grad_norm": 0.3416646718978882,
      "learning_rate": 0.00010925444675286382,
      "loss": 0.21813735365867615,
      "mean_token_accuracy": 0.9076962023973465,
      "num_tokens": 84880932.0,
      "step": 6881
    },
    {
      "entropy": 1.076392948627472,
      "epoch": 3.6240126382306475,
      "grad_norm": 0.30169451236724854,
      "learning_rate": 0.00010923251924013589,
      "loss": 0.17133498191833496,
      "mean_token_accuracy": 0.9298492968082428,
      "num_tokens": 84893268.0,
      "step": 6882
    },
    {
      "entropy": 1.0720700770616531,
      "epoch": 3.6245392311743023,
      "grad_norm": 0.29317283630371094,
      "learning_rate": 0.0001092105917729688,
      "loss": 0.16995534300804138,
      "mean_token_accuracy": 0.9279211014509201,
      "num_tokens": 84905604.0,
      "step": 6883
    },
    {
      "entropy": 1.0382179468870163,
      "epoch": 3.6250658241179567,
      "grad_norm": 0.2807539701461792,
      "learning_rate": 0.00010918866435266428,
      "loss": 0.158788800239563,
      "mean_token_accuracy": 0.9308933913707733,
      "num_tokens": 84917940.0,
      "step": 6884
    },
    {
      "entropy": 1.0360213816165924,
      "epoch": 3.625592417061611,
      "grad_norm": 0.2820185124874115,
      "learning_rate": 0.00010916673698052407,
      "loss": 0.17735148966312408,
      "mean_token_accuracy": 0.9254759401082993,
      "num_tokens": 84930276.0,
      "step": 6885
    },
    {
      "entropy": 1.0321247577667236,
      "epoch": 3.626119010005266,
      "grad_norm": 0.277366042137146,
      "learning_rate": 0.00010914480965784974,
      "loss": 0.16998949646949768,
      "mean_token_accuracy": 0.9264341443777084,
      "num_tokens": 84942612.0,
      "step": 6886
    },
    {
      "entropy": 1.11885204911232,
      "epoch": 3.6266456029489205,
      "grad_norm": 0.3434223532676697,
      "learning_rate": 0.00010912288238594312,
      "loss": 0.19510222971439362,
      "mean_token_accuracy": 0.9249685257673264,
      "num_tokens": 84954948.0,
      "step": 6887
    },
    {
      "entropy": 1.0723938047885895,
      "epoch": 3.627172195892575,
      "grad_norm": 0.33464300632476807,
      "learning_rate": 0.00010910095516610587,
      "loss": 0.18581683933734894,
      "mean_token_accuracy": 0.9200177937746048,
      "num_tokens": 84967284.0,
      "step": 6888
    },
    {
      "entropy": 1.0627216696739197,
      "epoch": 3.6276987888362298,
      "grad_norm": 0.29970863461494446,
      "learning_rate": 0.00010907902799963972,
      "loss": 0.1742255538702011,
      "mean_token_accuracy": 0.9224542528390884,
      "num_tokens": 84979620.0,
      "step": 6889
    },
    {
      "entropy": 1.0519684851169586,
      "epoch": 3.628225381779884,
      "grad_norm": 0.3227628767490387,
      "learning_rate": 0.00010905710088784624,
      "loss": 0.18886661529541016,
      "mean_token_accuracy": 0.9240933060646057,
      "num_tokens": 84991956.0,
      "step": 6890
    },
    {
      "entropy": 1.0896053314208984,
      "epoch": 3.6287519747235386,
      "grad_norm": 0.2841867208480835,
      "learning_rate": 0.00010903517383202721,
      "loss": 0.17191782593727112,
      "mean_token_accuracy": 0.9256446659564972,
      "num_tokens": 85004292.0,
      "step": 6891
    },
    {
      "entropy": 1.0260752737522125,
      "epoch": 3.6292785676671935,
      "grad_norm": 0.2733602821826935,
      "learning_rate": 0.00010901324683348428,
      "loss": 0.1687346249818802,
      "mean_token_accuracy": 0.9257003366947174,
      "num_tokens": 85016628.0,
      "step": 6892
    },
    {
      "entropy": 1.071472853422165,
      "epoch": 3.629805160610848,
      "grad_norm": 0.31747159361839294,
      "learning_rate": 0.00010899131989351912,
      "loss": 0.18207454681396484,
      "mean_token_accuracy": 0.92514668405056,
      "num_tokens": 85028964.0,
      "step": 6893
    },
    {
      "entropy": 1.1075446009635925,
      "epoch": 3.6303317535545023,
      "grad_norm": 0.28832435607910156,
      "learning_rate": 0.00010896939301343344,
      "loss": 0.16811193525791168,
      "mean_token_accuracy": 0.9267793744802475,
      "num_tokens": 85041300.0,
      "step": 6894
    },
    {
      "entropy": 1.0397011488676071,
      "epoch": 3.630858346498157,
      "grad_norm": 0.2771309018135071,
      "learning_rate": 0.00010894746619452888,
      "loss": 0.16865265369415283,
      "mean_token_accuracy": 0.9260818362236023,
      "num_tokens": 85053636.0,
      "step": 6895
    },
    {
      "entropy": 1.058392345905304,
      "epoch": 3.6313849394418116,
      "grad_norm": 0.3174492120742798,
      "learning_rate": 0.0001089255394381071,
      "loss": 0.17661213874816895,
      "mean_token_accuracy": 0.9262597411870956,
      "num_tokens": 85065972.0,
      "step": 6896
    },
    {
      "entropy": 1.0779773145914078,
      "epoch": 3.631911532385466,
      "grad_norm": 0.2963669002056122,
      "learning_rate": 0.00010890361274546983,
      "loss": 0.18939608335494995,
      "mean_token_accuracy": 0.9218048602342606,
      "num_tokens": 85078308.0,
      "step": 6897
    },
    {
      "entropy": 1.0529873967170715,
      "epoch": 3.6324381253291205,
      "grad_norm": 0.2809640169143677,
      "learning_rate": 0.00010888168611791864,
      "loss": 0.187717467546463,
      "mean_token_accuracy": 0.9181215912103653,
      "num_tokens": 85090644.0,
      "step": 6898
    },
    {
      "entropy": 1.08050075173378,
      "epoch": 3.632964718272775,
      "grad_norm": 0.29269665479660034,
      "learning_rate": 0.00010885975955675524,
      "loss": 0.18015234172344208,
      "mean_token_accuracy": 0.9219858050346375,
      "num_tokens": 85102980.0,
      "step": 6899
    },
    {
      "entropy": 1.1340832561254501,
      "epoch": 3.6334913112164298,
      "grad_norm": 0.29490646719932556,
      "learning_rate": 0.00010883783306328126,
      "loss": 0.18035733699798584,
      "mean_token_accuracy": 0.9175348877906799,
      "num_tokens": 85115316.0,
      "step": 6900
    },
    {
      "entropy": 1.078901320695877,
      "epoch": 3.634017904160084,
      "grad_norm": 0.2814980149269104,
      "learning_rate": 0.00010881590663879837,
      "loss": 0.17936724424362183,
      "mean_token_accuracy": 0.9242068231105804,
      "num_tokens": 85127652.0,
      "step": 6901
    },
    {
      "entropy": 1.0839639902114868,
      "epoch": 3.6345444971037386,
      "grad_norm": 0.2766796946525574,
      "learning_rate": 0.0001087939802846082,
      "loss": 0.18048717081546783,
      "mean_token_accuracy": 0.9242569804191589,
      "num_tokens": 85139988.0,
      "step": 6902
    },
    {
      "entropy": 1.0977406799793243,
      "epoch": 3.6350710900473935,
      "grad_norm": 0.28339245915412903,
      "learning_rate": 0.00010877205400201238,
      "loss": 0.17970004677772522,
      "mean_token_accuracy": 0.9218273609876633,
      "num_tokens": 85152324.0,
      "step": 6903
    },
    {
      "entropy": 1.0731355249881744,
      "epoch": 3.635597682991048,
      "grad_norm": 0.29195013642311096,
      "learning_rate": 0.0001087501277923126,
      "loss": 0.19581443071365356,
      "mean_token_accuracy": 0.9157865047454834,
      "num_tokens": 85164660.0,
      "step": 6904
    },
    {
      "entropy": 1.0776561498641968,
      "epoch": 3.6361242759347023,
      "grad_norm": 0.27677494287490845,
      "learning_rate": 0.00010872820165681041,
      "loss": 0.17254780232906342,
      "mean_token_accuracy": 0.9247097820043564,
      "num_tokens": 85176996.0,
      "step": 6905
    },
    {
      "entropy": 1.0955494940280914,
      "epoch": 3.636650868878357,
      "grad_norm": 0.2864554226398468,
      "learning_rate": 0.00010870627559680749,
      "loss": 0.1892591416835785,
      "mean_token_accuracy": 0.9218159466981888,
      "num_tokens": 85189332.0,
      "step": 6906
    },
    {
      "entropy": 1.1011942327022552,
      "epoch": 3.6371774618220116,
      "grad_norm": 0.279434472322464,
      "learning_rate": 0.00010868434961360539,
      "loss": 0.1690611094236374,
      "mean_token_accuracy": 0.9287009537220001,
      "num_tokens": 85201668.0,
      "step": 6907
    },
    {
      "entropy": 1.1294350922107697,
      "epoch": 3.637704054765666,
      "grad_norm": 0.3198857307434082,
      "learning_rate": 0.00010866242370850586,
      "loss": 0.19883982837200165,
      "mean_token_accuracy": 0.9200601279735565,
      "num_tokens": 85214004.0,
      "step": 6908
    },
    {
      "entropy": 1.1131604313850403,
      "epoch": 3.638230647709321,
      "grad_norm": 0.28839418292045593,
      "learning_rate": 0.00010864049788281041,
      "loss": 0.18458659946918488,
      "mean_token_accuracy": 0.9204800724983215,
      "num_tokens": 85226340.0,
      "step": 6909
    },
    {
      "entropy": 1.1100212335586548,
      "epoch": 3.6387572406529753,
      "grad_norm": 0.2910570502281189,
      "learning_rate": 0.00010861857213782068,
      "loss": 0.16764257848262787,
      "mean_token_accuracy": 0.9310833513736725,
      "num_tokens": 85238676.0,
      "step": 6910
    },
    {
      "entropy": 1.0589947998523712,
      "epoch": 3.6392838335966298,
      "grad_norm": 0.27973130345344543,
      "learning_rate": 0.00010859664647483825,
      "loss": 0.1870989054441452,
      "mean_token_accuracy": 0.9192935824394226,
      "num_tokens": 85251012.0,
      "step": 6911
    },
    {
      "entropy": 1.0890919864177704,
      "epoch": 3.639810426540284,
      "grad_norm": 0.31354379653930664,
      "learning_rate": 0.00010857472089516476,
      "loss": 0.19124506413936615,
      "mean_token_accuracy": 0.9161752760410309,
      "num_tokens": 85263348.0,
      "step": 6912
    },
    {
      "entropy": 1.1142856180667877,
      "epoch": 3.640337019483939,
      "grad_norm": 0.2831694483757019,
      "learning_rate": 0.0001085527954001018,
      "loss": 0.1791841685771942,
      "mean_token_accuracy": 0.9219368547201157,
      "num_tokens": 85275684.0,
      "step": 6913
    },
    {
      "entropy": 1.087710440158844,
      "epoch": 3.6408636124275935,
      "grad_norm": 0.27839532494544983,
      "learning_rate": 0.0001085308699909509,
      "loss": 0.15793544054031372,
      "mean_token_accuracy": 0.9308710992336273,
      "num_tokens": 85288020.0,
      "step": 6914
    },
    {
      "entropy": 1.104949176311493,
      "epoch": 3.641390205371248,
      "grad_norm": 0.3108198046684265,
      "learning_rate": 0.0001085089446690137,
      "loss": 0.20564743876457214,
      "mean_token_accuracy": 0.9142681956291199,
      "num_tokens": 85300356.0,
      "step": 6915
    },
    {
      "entropy": 1.089268833398819,
      "epoch": 3.6419167983149023,
      "grad_norm": 0.2661731541156769,
      "learning_rate": 0.00010848701943559176,
      "loss": 0.16525886952877045,
      "mean_token_accuracy": 0.928042009472847,
      "num_tokens": 85312692.0,
      "step": 6916
    },
    {
      "entropy": 1.1005620956420898,
      "epoch": 3.642443391258557,
      "grad_norm": 0.3060215711593628,
      "learning_rate": 0.00010846509429198671,
      "loss": 0.1820535808801651,
      "mean_token_accuracy": 0.9212777465581894,
      "num_tokens": 85325028.0,
      "step": 6917
    },
    {
      "entropy": 1.0968313217163086,
      "epoch": 3.6429699842022116,
      "grad_norm": 0.29798147082328796,
      "learning_rate": 0.00010844316923950002,
      "loss": 0.19336146116256714,
      "mean_token_accuracy": 0.9188212752342224,
      "num_tokens": 85337364.0,
      "step": 6918
    },
    {
      "entropy": 1.1311476826667786,
      "epoch": 3.643496577145866,
      "grad_norm": 0.33231881260871887,
      "learning_rate": 0.00010842124427943332,
      "loss": 0.19728441536426544,
      "mean_token_accuracy": 0.9228433519601822,
      "num_tokens": 85349700.0,
      "step": 6919
    },
    {
      "entropy": 1.0722969472408295,
      "epoch": 3.644023170089521,
      "grad_norm": 0.29335349798202515,
      "learning_rate": 0.00010839931941308815,
      "loss": 0.17789262533187866,
      "mean_token_accuracy": 0.9251670092344284,
      "num_tokens": 85362036.0,
      "step": 6920
    },
    {
      "entropy": 1.135120540857315,
      "epoch": 3.6445497630331753,
      "grad_norm": 0.31360843777656555,
      "learning_rate": 0.00010837739464176597,
      "loss": 0.19223442673683167,
      "mean_token_accuracy": 0.916714608669281,
      "num_tokens": 85374372.0,
      "step": 6921
    },
    {
      "entropy": 1.1161190867424011,
      "epoch": 3.6450763559768298,
      "grad_norm": 0.30781087279319763,
      "learning_rate": 0.00010835546996676848,
      "loss": 0.1870063841342926,
      "mean_token_accuracy": 0.922933965921402,
      "num_tokens": 85386708.0,
      "step": 6922
    },
    {
      "entropy": 1.1311403810977936,
      "epoch": 3.6456029489204846,
      "grad_norm": 0.30369794368743896,
      "learning_rate": 0.00010833354538939717,
      "loss": 0.18748193979263306,
      "mean_token_accuracy": 0.9215940535068512,
      "num_tokens": 85399044.0,
      "step": 6923
    },
    {
      "entropy": 1.100357174873352,
      "epoch": 3.646129541864139,
      "grad_norm": 0.3031750023365021,
      "learning_rate": 0.00010831162091095355,
      "loss": 0.1798853576183319,
      "mean_token_accuracy": 0.9241892546415329,
      "num_tokens": 85411380.0,
      "step": 6924
    },
    {
      "entropy": 1.125574380159378,
      "epoch": 3.6466561348077935,
      "grad_norm": 0.27512413263320923,
      "learning_rate": 0.00010828969653273915,
      "loss": 0.16666743159294128,
      "mean_token_accuracy": 0.9244058281183243,
      "num_tokens": 85423716.0,
      "step": 6925
    },
    {
      "entropy": 1.1576015055179596,
      "epoch": 3.6471827277514484,
      "grad_norm": 0.3020809590816498,
      "learning_rate": 0.00010826777225605549,
      "loss": 0.19194430112838745,
      "mean_token_accuracy": 0.9175030440092087,
      "num_tokens": 85436052.0,
      "step": 6926
    },
    {
      "entropy": 1.1278834640979767,
      "epoch": 3.647709320695103,
      "grad_norm": 0.26242175698280334,
      "learning_rate": 0.0001082458480822041,
      "loss": 0.17175456881523132,
      "mean_token_accuracy": 0.9282833635807037,
      "num_tokens": 85448388.0,
      "step": 6927
    },
    {
      "entropy": 1.1385831236839294,
      "epoch": 3.648235913638757,
      "grad_norm": 0.28382521867752075,
      "learning_rate": 0.00010822392401248649,
      "loss": 0.17114661633968353,
      "mean_token_accuracy": 0.9304853975772858,
      "num_tokens": 85460724.0,
      "step": 6928
    },
    {
      "entropy": 1.1503818333148956,
      "epoch": 3.6487625065824116,
      "grad_norm": 0.2844337522983551,
      "learning_rate": 0.00010820200004820414,
      "loss": 0.18441346287727356,
      "mean_token_accuracy": 0.9221296608448029,
      "num_tokens": 85473060.0,
      "step": 6929
    },
    {
      "entropy": 1.1435934901237488,
      "epoch": 3.6492890995260665,
      "grad_norm": 0.2975497543811798,
      "learning_rate": 0.0001081800761906586,
      "loss": 0.17758281528949738,
      "mean_token_accuracy": 0.9228782057762146,
      "num_tokens": 85485396.0,
      "step": 6930
    },
    {
      "entropy": 1.1359702944755554,
      "epoch": 3.649815692469721,
      "grad_norm": 0.2905472218990326,
      "learning_rate": 0.0001081581524411513,
      "loss": 0.1734602451324463,
      "mean_token_accuracy": 0.9265166819095612,
      "num_tokens": 85497732.0,
      "step": 6931
    },
    {
      "entropy": 1.0825994312763214,
      "epoch": 3.6503422854133754,
      "grad_norm": 0.2530800700187683,
      "learning_rate": 0.00010813622880098377,
      "loss": 0.16221484541893005,
      "mean_token_accuracy": 0.9289655685424805,
      "num_tokens": 85510068.0,
      "step": 6932
    },
    {
      "entropy": 1.1005148589611053,
      "epoch": 3.65086887835703,
      "grad_norm": 0.29540663957595825,
      "learning_rate": 0.00010811430527145747,
      "loss": 0.1880090832710266,
      "mean_token_accuracy": 0.9217631965875626,
      "num_tokens": 85522404.0,
      "step": 6933
    },
    {
      "entropy": 1.127230018377304,
      "epoch": 3.6513954713006846,
      "grad_norm": 0.30108895897865295,
      "learning_rate": 0.00010809238185387389,
      "loss": 0.18256250023841858,
      "mean_token_accuracy": 0.9247212707996368,
      "num_tokens": 85534740.0,
      "step": 6934
    },
    {
      "entropy": 1.1048816442489624,
      "epoch": 3.651922064244339,
      "grad_norm": 0.2992095351219177,
      "learning_rate": 0.00010807045854953443,
      "loss": 0.1793767511844635,
      "mean_token_accuracy": 0.919749990105629,
      "num_tokens": 85547076.0,
      "step": 6935
    },
    {
      "entropy": 1.0833216607570648,
      "epoch": 3.6524486571879935,
      "grad_norm": 0.27640217542648315,
      "learning_rate": 0.00010804853535974065,
      "loss": 0.17367729544639587,
      "mean_token_accuracy": 0.9288899600505829,
      "num_tokens": 85559412.0,
      "step": 6936
    },
    {
      "entropy": 1.0898649990558624,
      "epoch": 3.6529752501316484,
      "grad_norm": 0.3019793927669525,
      "learning_rate": 0.00010802661228579396,
      "loss": 0.17442551255226135,
      "mean_token_accuracy": 0.927208662033081,
      "num_tokens": 85571748.0,
      "step": 6937
    },
    {
      "entropy": 1.0895110815763474,
      "epoch": 3.653501843075303,
      "grad_norm": 0.27178964018821716,
      "learning_rate": 0.00010800468932899577,
      "loss": 0.1709968000650406,
      "mean_token_accuracy": 0.925664871931076,
      "num_tokens": 85584084.0,
      "step": 6938
    },
    {
      "entropy": 1.0956975221633911,
      "epoch": 3.654028436018957,
      "grad_norm": 0.29212814569473267,
      "learning_rate": 0.00010798276649064756,
      "loss": 0.17584949731826782,
      "mean_token_accuracy": 0.9293852895498276,
      "num_tokens": 85596420.0,
      "step": 6939
    },
    {
      "entropy": 1.0751636922359467,
      "epoch": 3.654555028962612,
      "grad_norm": 0.31412044167518616,
      "learning_rate": 0.00010796084377205071,
      "loss": 0.19069185853004456,
      "mean_token_accuracy": 0.9202124923467636,
      "num_tokens": 85608756.0,
      "step": 6940
    },
    {
      "entropy": 1.0722732841968536,
      "epoch": 3.6550816219062665,
      "grad_norm": 0.2728721499443054,
      "learning_rate": 0.00010793892117450672,
      "loss": 0.16478730738162994,
      "mean_token_accuracy": 0.9310110807418823,
      "num_tokens": 85621092.0,
      "step": 6941
    },
    {
      "entropy": 1.1176518499851227,
      "epoch": 3.655608214849921,
      "grad_norm": 0.3089950680732727,
      "learning_rate": 0.00010791699869931693,
      "loss": 0.18918341398239136,
      "mean_token_accuracy": 0.9210700541734695,
      "num_tokens": 85633428.0,
      "step": 6942
    },
    {
      "entropy": 1.0875534117221832,
      "epoch": 3.656134807793576,
      "grad_norm": 0.2986571788787842,
      "learning_rate": 0.00010789507634778281,
      "loss": 0.1821071207523346,
      "mean_token_accuracy": 0.924040675163269,
      "num_tokens": 85645764.0,
      "step": 6943
    },
    {
      "entropy": 1.104602336883545,
      "epoch": 3.6566614007372302,
      "grad_norm": 0.2819954752922058,
      "learning_rate": 0.00010787315412120574,
      "loss": 0.17842718958854675,
      "mean_token_accuracy": 0.9197171330451965,
      "num_tokens": 85658100.0,
      "step": 6944
    },
    {
      "entropy": 1.0745837092399597,
      "epoch": 3.6571879936808847,
      "grad_norm": 0.2845374643802643,
      "learning_rate": 0.00010785123202088712,
      "loss": 0.17215374112129211,
      "mean_token_accuracy": 0.9247159063816071,
      "num_tokens": 85670436.0,
      "step": 6945
    },
    {
      "entropy": 1.0754876732826233,
      "epoch": 3.657714586624539,
      "grad_norm": 0.30206388235092163,
      "learning_rate": 0.00010782931004812831,
      "loss": 0.18929605185985565,
      "mean_token_accuracy": 0.9201420098543167,
      "num_tokens": 85682772.0,
      "step": 6946
    },
    {
      "entropy": 1.0547612607479095,
      "epoch": 3.658241179568194,
      "grad_norm": 0.2812706232070923,
      "learning_rate": 0.00010780738820423073,
      "loss": 0.1781274378299713,
      "mean_token_accuracy": 0.9253074079751968,
      "num_tokens": 85695108.0,
      "step": 6947
    },
    {
      "entropy": 1.0610165894031525,
      "epoch": 3.6587677725118484,
      "grad_norm": 0.32550254464149475,
      "learning_rate": 0.00010778546649049574,
      "loss": 0.1952066421508789,
      "mean_token_accuracy": 0.9139260500669479,
      "num_tokens": 85707444.0,
      "step": 6948
    },
    {
      "entropy": 1.0588375329971313,
      "epoch": 3.659294365455503,
      "grad_norm": 0.287535160779953,
      "learning_rate": 0.00010776354490822464,
      "loss": 0.17247122526168823,
      "mean_token_accuracy": 0.9268250167369843,
      "num_tokens": 85719780.0,
      "step": 6949
    },
    {
      "entropy": 1.0423914641141891,
      "epoch": 3.659820958399157,
      "grad_norm": 0.2818460166454315,
      "learning_rate": 0.0001077416234587189,
      "loss": 0.17156368494033813,
      "mean_token_accuracy": 0.9256498515605927,
      "num_tokens": 85732116.0,
      "step": 6950
    },
    {
      "entropy": 1.0398815125226974,
      "epoch": 3.660347551342812,
      "grad_norm": 0.2628284692764282,
      "learning_rate": 0.0001077197021432798,
      "loss": 0.16416022181510925,
      "mean_token_accuracy": 0.9331476390361786,
      "num_tokens": 85744452.0,
      "step": 6951
    },
    {
      "entropy": 1.0208431482315063,
      "epoch": 3.6608741442864665,
      "grad_norm": 0.27734678983688354,
      "learning_rate": 0.00010769778096320873,
      "loss": 0.17575320601463318,
      "mean_token_accuracy": 0.92631596326828,
      "num_tokens": 85756788.0,
      "step": 6952
    },
    {
      "entropy": 1.062884896993637,
      "epoch": 3.661400737230121,
      "grad_norm": 0.30154260993003845,
      "learning_rate": 0.00010767585991980693,
      "loss": 0.18772351741790771,
      "mean_token_accuracy": 0.9216388911008835,
      "num_tokens": 85769124.0,
      "step": 6953
    },
    {
      "entropy": 1.0944669097661972,
      "epoch": 3.661927330173776,
      "grad_norm": 0.3269876539707184,
      "learning_rate": 0.00010765393901437581,
      "loss": 0.1878005564212799,
      "mean_token_accuracy": 0.9181205332279205,
      "num_tokens": 85781460.0,
      "step": 6954
    },
    {
      "entropy": 1.030048444867134,
      "epoch": 3.6624539231174302,
      "grad_norm": 0.2715197503566742,
      "learning_rate": 0.00010763201824821665,
      "loss": 0.1689884215593338,
      "mean_token_accuracy": 0.9284142404794693,
      "num_tokens": 85793796.0,
      "step": 6955
    },
    {
      "entropy": 1.053311824798584,
      "epoch": 3.6629805160610847,
      "grad_norm": 0.31180477142333984,
      "learning_rate": 0.00010761009762263074,
      "loss": 0.19359172880649567,
      "mean_token_accuracy": 0.9211233556270599,
      "num_tokens": 85806132.0,
      "step": 6956
    },
    {
      "entropy": 1.0647279024124146,
      "epoch": 3.6635071090047395,
      "grad_norm": 0.2950548529624939,
      "learning_rate": 0.00010758817713891942,
      "loss": 0.1748398393392563,
      "mean_token_accuracy": 0.9264338165521622,
      "num_tokens": 85818468.0,
      "step": 6957
    },
    {
      "entropy": 1.0664381682872772,
      "epoch": 3.664033701948394,
      "grad_norm": 0.2937673330307007,
      "learning_rate": 0.00010756625679838397,
      "loss": 0.18024608492851257,
      "mean_token_accuracy": 0.9259634613990784,
      "num_tokens": 85830804.0,
      "step": 6958
    },
    {
      "entropy": 1.0663295984268188,
      "epoch": 3.6645602948920484,
      "grad_norm": 0.28091055154800415,
      "learning_rate": 0.00010754433660232568,
      "loss": 0.1665952354669571,
      "mean_token_accuracy": 0.929583802819252,
      "num_tokens": 85843140.0,
      "step": 6959
    },
    {
      "entropy": 1.0400154292583466,
      "epoch": 3.6650868878357032,
      "grad_norm": 0.2815306484699249,
      "learning_rate": 0.0001075224165520458,
      "loss": 0.16376999020576477,
      "mean_token_accuracy": 0.930091381072998,
      "num_tokens": 85855476.0,
      "step": 6960
    },
    {
      "entropy": 1.0289010107517242,
      "epoch": 3.6656134807793577,
      "grad_norm": 0.28532126545906067,
      "learning_rate": 0.00010750049664884561,
      "loss": 0.17586354911327362,
      "mean_token_accuracy": 0.9279272109270096,
      "num_tokens": 85867812.0,
      "step": 6961
    },
    {
      "entropy": 1.0397853255271912,
      "epoch": 3.666140073723012,
      "grad_norm": 0.2653680145740509,
      "learning_rate": 0.00010747857689402637,
      "loss": 0.15809708833694458,
      "mean_token_accuracy": 0.9312074035406113,
      "num_tokens": 85880148.0,
      "step": 6962
    },
    {
      "entropy": 1.03483048081398,
      "epoch": 3.6666666666666665,
      "grad_norm": 0.3177836537361145,
      "learning_rate": 0.0001074566572888893,
      "loss": 0.1801329106092453,
      "mean_token_accuracy": 0.9240194410085678,
      "num_tokens": 85892484.0,
      "step": 6963
    },
    {
      "entropy": 1.0460475087165833,
      "epoch": 3.6671932596103214,
      "grad_norm": 0.2922004163265228,
      "learning_rate": 0.0001074347378347357,
      "loss": 0.20034848153591156,
      "mean_token_accuracy": 0.9176178127527237,
      "num_tokens": 85904820.0,
      "step": 6964
    },
    {
      "entropy": 1.0509907603263855,
      "epoch": 3.667719852553976,
      "grad_norm": 0.2957066297531128,
      "learning_rate": 0.00010741281853286675,
      "loss": 0.17528370022773743,
      "mean_token_accuracy": 0.9293950200080872,
      "num_tokens": 85917156.0,
      "step": 6965
    },
    {
      "entropy": 0.976374089717865,
      "epoch": 3.6682464454976302,
      "grad_norm": 0.2763223350048065,
      "learning_rate": 0.00010739089938458367,
      "loss": 0.17812931537628174,
      "mean_token_accuracy": 0.9236332774162292,
      "num_tokens": 85929492.0,
      "step": 6966
    },
    {
      "entropy": 1.076490581035614,
      "epoch": 3.6687730384412847,
      "grad_norm": 0.31013357639312744,
      "learning_rate": 0.00010736898039118773,
      "loss": 0.197622150182724,
      "mean_token_accuracy": 0.9179268032312393,
      "num_tokens": 85941828.0,
      "step": 6967
    },
    {
      "entropy": 1.0025828033685684,
      "epoch": 3.6692996313849395,
      "grad_norm": 0.28579673171043396,
      "learning_rate": 0.00010734706155398007,
      "loss": 0.16753435134887695,
      "mean_token_accuracy": 0.9261990189552307,
      "num_tokens": 85954164.0,
      "step": 6968
    },
    {
      "entropy": 1.02617509663105,
      "epoch": 3.669826224328594,
      "grad_norm": 0.3085153102874756,
      "learning_rate": 0.00010732514287426188,
      "loss": 0.19467276334762573,
      "mean_token_accuracy": 0.9151240140199661,
      "num_tokens": 85966500.0,
      "step": 6969
    },
    {
      "entropy": 1.0354234278202057,
      "epoch": 3.6703528172722484,
      "grad_norm": 0.2686126232147217,
      "learning_rate": 0.00010730322435333433,
      "loss": 0.1822713315486908,
      "mean_token_accuracy": 0.92469722032547,
      "num_tokens": 85978836.0,
      "step": 6970
    },
    {
      "entropy": 1.0647673606872559,
      "epoch": 3.6708794102159032,
      "grad_norm": 0.32261499762535095,
      "learning_rate": 0.0001072813059924987,
      "loss": 0.19555431604385376,
      "mean_token_accuracy": 0.9207833260297775,
      "num_tokens": 85991172.0,
      "step": 6971
    },
    {
      "entropy": 1.0193777978420258,
      "epoch": 3.6714060031595577,
      "grad_norm": 0.2969292402267456,
      "learning_rate": 0.00010725938779305609,
      "loss": 0.17920270562171936,
      "mean_token_accuracy": 0.9244997650384903,
      "num_tokens": 86003508.0,
      "step": 6972
    },
    {
      "entropy": 1.000598281621933,
      "epoch": 3.671932596103212,
      "grad_norm": 0.2639365792274475,
      "learning_rate": 0.00010723746975630761,
      "loss": 0.15792196989059448,
      "mean_token_accuracy": 0.9345228672027588,
      "num_tokens": 86015844.0,
      "step": 6973
    },
    {
      "entropy": 1.0382906347513199,
      "epoch": 3.672459189046867,
      "grad_norm": 0.2822777330875397,
      "learning_rate": 0.00010721555188355448,
      "loss": 0.18117579817771912,
      "mean_token_accuracy": 0.924898087978363,
      "num_tokens": 86028180.0,
      "step": 6974
    },
    {
      "entropy": 1.0620505809783936,
      "epoch": 3.6729857819905214,
      "grad_norm": 0.2701767086982727,
      "learning_rate": 0.0001071936341760978,
      "loss": 0.17222149670124054,
      "mean_token_accuracy": 0.9235534220933914,
      "num_tokens": 86040516.0,
      "step": 6975
    },
    {
      "entropy": 1.0112586617469788,
      "epoch": 3.673512374934176,
      "grad_norm": 0.26706138253211975,
      "learning_rate": 0.00010717171663523871,
      "loss": 0.1747536063194275,
      "mean_token_accuracy": 0.9254122376441956,
      "num_tokens": 86052852.0,
      "step": 6976
    },
    {
      "entropy": 1.039280742406845,
      "epoch": 3.6740389678778307,
      "grad_norm": 0.2836163640022278,
      "learning_rate": 0.00010714979926227828,
      "loss": 0.17702583968639374,
      "mean_token_accuracy": 0.9256714284420013,
      "num_tokens": 86065188.0,
      "step": 6977
    },
    {
      "entropy": 1.0016122907400131,
      "epoch": 3.674565560821485,
      "grad_norm": 0.2784358859062195,
      "learning_rate": 0.00010712788205851769,
      "loss": 0.17481949925422668,
      "mean_token_accuracy": 0.9267537146806717,
      "num_tokens": 86077524.0,
      "step": 6978
    },
    {
      "entropy": 1.0275622457265854,
      "epoch": 3.6750921537651395,
      "grad_norm": 0.31567052006721497,
      "learning_rate": 0.000107105965025258,
      "loss": 0.19603204727172852,
      "mean_token_accuracy": 0.9181112349033356,
      "num_tokens": 86089860.0,
      "step": 6979
    },
    {
      "entropy": 1.0318071246147156,
      "epoch": 3.675618746708794,
      "grad_norm": 0.2872801125049591,
      "learning_rate": 0.00010708404816380029,
      "loss": 0.1774379462003708,
      "mean_token_accuracy": 0.9286105185747147,
      "num_tokens": 86102196.0,
      "step": 6980
    },
    {
      "entropy": 1.0042276829481125,
      "epoch": 3.6761453396524484,
      "grad_norm": 0.30108317732810974,
      "learning_rate": 0.00010706213147544566,
      "loss": 0.18309810757637024,
      "mean_token_accuracy": 0.9228730797767639,
      "num_tokens": 86114532.0,
      "step": 6981
    },
    {
      "entropy": 0.9636296778917313,
      "epoch": 3.6766719325961033,
      "grad_norm": 0.2768324017524719,
      "learning_rate": 0.00010704021496149517,
      "loss": 0.1785029023885727,
      "mean_token_accuracy": 0.9208957850933075,
      "num_tokens": 86126868.0,
      "step": 6982
    },
    {
      "entropy": 1.031101867556572,
      "epoch": 3.6771985255397577,
      "grad_norm": 0.31000271439552307,
      "learning_rate": 0.00010701829862324984,
      "loss": 0.18313610553741455,
      "mean_token_accuracy": 0.9193488359451294,
      "num_tokens": 86139204.0,
      "step": 6983
    },
    {
      "entropy": 1.032566413283348,
      "epoch": 3.677725118483412,
      "grad_norm": 0.28918370604515076,
      "learning_rate": 0.00010699638246201071,
      "loss": 0.17839691042900085,
      "mean_token_accuracy": 0.9232564121484756,
      "num_tokens": 86151540.0,
      "step": 6984
    },
    {
      "entropy": 1.0230403393507004,
      "epoch": 3.678251711427067,
      "grad_norm": 0.29073366522789,
      "learning_rate": 0.0001069744664790789,
      "loss": 0.1784803867340088,
      "mean_token_accuracy": 0.9265795201063156,
      "num_tokens": 86163876.0,
      "step": 6985
    },
    {
      "entropy": 1.0558317303657532,
      "epoch": 3.6787783043707214,
      "grad_norm": 0.27190113067626953,
      "learning_rate": 0.00010695255067575538,
      "loss": 0.1662517786026001,
      "mean_token_accuracy": 0.9308346807956696,
      "num_tokens": 86176212.0,
      "step": 6986
    },
    {
      "entropy": 1.033191591501236,
      "epoch": 3.679304897314376,
      "grad_norm": 0.2988441586494446,
      "learning_rate": 0.00010693063505334115,
      "loss": 0.18694527447223663,
      "mean_token_accuracy": 0.9207980185747147,
      "num_tokens": 86188548.0,
      "step": 6987
    },
    {
      "entropy": 1.0609621107578278,
      "epoch": 3.6798314902580307,
      "grad_norm": 0.2892952263355255,
      "learning_rate": 0.00010690871961313724,
      "loss": 0.17166557908058167,
      "mean_token_accuracy": 0.931251272559166,
      "num_tokens": 86200884.0,
      "step": 6988
    },
    {
      "entropy": 1.0341887772083282,
      "epoch": 3.680358083201685,
      "grad_norm": 0.30244240164756775,
      "learning_rate": 0.0001068868043564446,
      "loss": 0.1853313148021698,
      "mean_token_accuracy": 0.923940971493721,
      "num_tokens": 86213220.0,
      "step": 6989
    },
    {
      "entropy": 1.049064576625824,
      "epoch": 3.6808846761453395,
      "grad_norm": 0.3076011538505554,
      "learning_rate": 0.00010686488928456428,
      "loss": 0.20103253424167633,
      "mean_token_accuracy": 0.9156415909528732,
      "num_tokens": 86225556.0,
      "step": 6990
    },
    {
      "entropy": 1.0623431205749512,
      "epoch": 3.6814112690889944,
      "grad_norm": 0.3282015919685364,
      "learning_rate": 0.00010684297439879715,
      "loss": 0.20622855424880981,
      "mean_token_accuracy": 0.9159451127052307,
      "num_tokens": 86237892.0,
      "step": 6991
    },
    {
      "entropy": 1.0407926738262177,
      "epoch": 3.681937862032649,
      "grad_norm": 0.28654420375823975,
      "learning_rate": 0.0001068210597004443,
      "loss": 0.17378610372543335,
      "mean_token_accuracy": 0.9291479289531708,
      "num_tokens": 86250228.0,
      "step": 6992
    },
    {
      "entropy": 1.128141850233078,
      "epoch": 3.6824644549763033,
      "grad_norm": 0.286441445350647,
      "learning_rate": 0.00010679914519080655,
      "loss": 0.18758511543273926,
      "mean_token_accuracy": 0.9224020689725876,
      "num_tokens": 86262564.0,
      "step": 6993
    },
    {
      "entropy": 1.1071661114692688,
      "epoch": 3.682991047919958,
      "grad_norm": 0.27097955346107483,
      "learning_rate": 0.00010677723087118495,
      "loss": 0.17104873061180115,
      "mean_token_accuracy": 0.9285868257284164,
      "num_tokens": 86274900.0,
      "step": 6994
    },
    {
      "entropy": 1.1507742702960968,
      "epoch": 3.6835176408636126,
      "grad_norm": 0.3454969525337219,
      "learning_rate": 0.00010675531674288033,
      "loss": 0.1876622587442398,
      "mean_token_accuracy": 0.9162978529930115,
      "num_tokens": 86287236.0,
      "step": 6995
    },
    {
      "entropy": 1.1414210498332977,
      "epoch": 3.684044233807267,
      "grad_norm": 0.28427883982658386,
      "learning_rate": 0.00010673340280719364,
      "loss": 0.1635189950466156,
      "mean_token_accuracy": 0.935024231672287,
      "num_tokens": 86299572.0,
      "step": 6996
    },
    {
      "entropy": 1.0934878289699554,
      "epoch": 3.6845708267509214,
      "grad_norm": 0.30094486474990845,
      "learning_rate": 0.00010671148906542583,
      "loss": 0.20287704467773438,
      "mean_token_accuracy": 0.9153666496276855,
      "num_tokens": 86311908.0,
      "step": 6997
    },
    {
      "entropy": 1.1483174860477448,
      "epoch": 3.685097419694576,
      "grad_norm": 0.30192604660987854,
      "learning_rate": 0.00010668957551887768,
      "loss": 0.20245520770549774,
      "mean_token_accuracy": 0.9145628660917282,
      "num_tokens": 86324244.0,
      "step": 6998
    },
    {
      "entropy": 1.1466923654079437,
      "epoch": 3.6856240126382307,
      "grad_norm": 0.2737888991832733,
      "learning_rate": 0.00010666766216885016,
      "loss": 0.16557517647743225,
      "mean_token_accuracy": 0.9325868636369705,
      "num_tokens": 86336580.0,
      "step": 6999
    },
    {
      "entropy": 1.1706739664077759,
      "epoch": 3.686150605581885,
      "grad_norm": 0.309152752161026,
      "learning_rate": 0.00010664574901664415,
      "loss": 0.19145433604717255,
      "mean_token_accuracy": 0.9230051189661026,
      "num_tokens": 86348916.0,
      "step": 7000
    },
    {
      "entropy": 1.188638299703598,
      "epoch": 3.6866771985255395,
      "grad_norm": 0.3105132579803467,
      "learning_rate": 0.00010662383606356045,
      "loss": 0.1874886304140091,
      "mean_token_accuracy": 0.9241949319839478,
      "num_tokens": 86361252.0,
      "step": 7001
    },
    {
      "entropy": 1.2060999870300293,
      "epoch": 3.6872037914691944,
      "grad_norm": 0.28197094798088074,
      "learning_rate": 0.00010660192331089994,
      "loss": 0.18086087703704834,
      "mean_token_accuracy": 0.9176561236381531,
      "num_tokens": 86373588.0,
      "step": 7002
    },
    {
      "entropy": 1.1158166229724884,
      "epoch": 3.687730384412849,
      "grad_norm": 0.2853429615497589,
      "learning_rate": 0.00010658001075996342,
      "loss": 0.18830300867557526,
      "mean_token_accuracy": 0.9259162545204163,
      "num_tokens": 86385924.0,
      "step": 7003
    },
    {
      "entropy": 1.1557681560516357,
      "epoch": 3.6882569773565033,
      "grad_norm": 0.31260934472084045,
      "learning_rate": 0.00010655809841205173,
      "loss": 0.18399247527122498,
      "mean_token_accuracy": 0.9262349903583527,
      "num_tokens": 86398260.0,
      "step": 7004
    },
    {
      "entropy": 1.162243366241455,
      "epoch": 3.688783570300158,
      "grad_norm": 0.3255998194217682,
      "learning_rate": 0.00010653618626846564,
      "loss": 0.1905967891216278,
      "mean_token_accuracy": 0.9187018573284149,
      "num_tokens": 86410596.0,
      "step": 7005
    },
    {
      "entropy": 1.16911181807518,
      "epoch": 3.6893101632438126,
      "grad_norm": 0.28588801622390747,
      "learning_rate": 0.00010651427433050603,
      "loss": 0.17679578065872192,
      "mean_token_accuracy": 0.9221796095371246,
      "num_tokens": 86422932.0,
      "step": 7006
    },
    {
      "entropy": 1.1541486084461212,
      "epoch": 3.689836756187467,
      "grad_norm": 0.3123913109302521,
      "learning_rate": 0.00010649236259947362,
      "loss": 0.1948072761297226,
      "mean_token_accuracy": 0.9239502102136612,
      "num_tokens": 86435268.0,
      "step": 7007
    },
    {
      "entropy": 1.0990737974643707,
      "epoch": 3.690363349131122,
      "grad_norm": 0.28149572014808655,
      "learning_rate": 0.00010647045107666922,
      "loss": 0.1947208046913147,
      "mean_token_accuracy": 0.9193027168512344,
      "num_tokens": 86447604.0,
      "step": 7008
    },
    {
      "entropy": 1.113670527935028,
      "epoch": 3.6908899420747763,
      "grad_norm": 0.29361167550086975,
      "learning_rate": 0.00010644853976339355,
      "loss": 0.18384166061878204,
      "mean_token_accuracy": 0.9195683002471924,
      "num_tokens": 86459940.0,
      "step": 7009
    },
    {
      "entropy": 1.1258732974529266,
      "epoch": 3.6914165350184307,
      "grad_norm": 0.2840478718280792,
      "learning_rate": 0.00010642662866094737,
      "loss": 0.17376568913459778,
      "mean_token_accuracy": 0.9311702698469162,
      "num_tokens": 86472276.0,
      "step": 7010
    },
    {
      "entropy": 1.0605426728725433,
      "epoch": 3.6919431279620856,
      "grad_norm": 0.27593398094177246,
      "learning_rate": 0.00010640471777063143,
      "loss": 0.18008655309677124,
      "mean_token_accuracy": 0.9253093004226685,
      "num_tokens": 86484612.0,
      "step": 7011
    },
    {
      "entropy": 1.1514114439487457,
      "epoch": 3.69246972090574,
      "grad_norm": 0.29216331243515015,
      "learning_rate": 0.00010638280709374642,
      "loss": 0.17637720704078674,
      "mean_token_accuracy": 0.9234650433063507,
      "num_tokens": 86496948.0,
      "step": 7012
    },
    {
      "entropy": 1.0970701277256012,
      "epoch": 3.6929963138493944,
      "grad_norm": 0.28925836086273193,
      "learning_rate": 0.00010636089663159307,
      "loss": 0.1796543151140213,
      "mean_token_accuracy": 0.9250158369541168,
      "num_tokens": 86509284.0,
      "step": 7013
    },
    {
      "entropy": 1.1331265270709991,
      "epoch": 3.693522906793049,
      "grad_norm": 0.27358677983283997,
      "learning_rate": 0.0001063389863854721,
      "loss": 0.17291176319122314,
      "mean_token_accuracy": 0.9264134615659714,
      "num_tokens": 86521620.0,
      "step": 7014
    },
    {
      "entropy": 1.1059588491916656,
      "epoch": 3.6940494997367033,
      "grad_norm": 0.2972795069217682,
      "learning_rate": 0.00010631707635668418,
      "loss": 0.17801031470298767,
      "mean_token_accuracy": 0.9305655509233475,
      "num_tokens": 86533956.0,
      "step": 7015
    },
    {
      "entropy": 1.1910763084888458,
      "epoch": 3.694576092680358,
      "grad_norm": 0.32364970445632935,
      "learning_rate": 0.00010629516654652996,
      "loss": 0.2020956575870514,
      "mean_token_accuracy": 0.919253408908844,
      "num_tokens": 86546292.0,
      "step": 7016
    },
    {
      "entropy": 1.140288382768631,
      "epoch": 3.6951026856240126,
      "grad_norm": 0.2970587909221649,
      "learning_rate": 0.00010627325695631012,
      "loss": 0.1709657460451126,
      "mean_token_accuracy": 0.9261876344680786,
      "num_tokens": 86558628.0,
      "step": 7017
    },
    {
      "entropy": 1.1525245010852814,
      "epoch": 3.695629278567667,
      "grad_norm": 0.3080959916114807,
      "learning_rate": 0.00010625134758732527,
      "loss": 0.19118379056453705,
      "mean_token_accuracy": 0.9182205647230148,
      "num_tokens": 86570964.0,
      "step": 7018
    },
    {
      "entropy": 1.091363787651062,
      "epoch": 3.696155871511322,
      "grad_norm": 0.27670302987098694,
      "learning_rate": 0.00010622943844087603,
      "loss": 0.16978196799755096,
      "mean_token_accuracy": 0.9306501001119614,
      "num_tokens": 86583300.0,
      "step": 7019
    },
    {
      "entropy": 1.1769406497478485,
      "epoch": 3.6966824644549763,
      "grad_norm": 0.3435772657394409,
      "learning_rate": 0.00010620752951826312,
      "loss": 0.193936288356781,
      "mean_token_accuracy": 0.9195510298013687,
      "num_tokens": 86595636.0,
      "step": 7020
    },
    {
      "entropy": 1.0992920696735382,
      "epoch": 3.6972090573986307,
      "grad_norm": 0.2789323627948761,
      "learning_rate": 0.00010618562082078705,
      "loss": 0.17331257462501526,
      "mean_token_accuracy": 0.9258032143115997,
      "num_tokens": 86607972.0,
      "step": 7021
    },
    {
      "entropy": 1.1387426257133484,
      "epoch": 3.6977356503422856,
      "grad_norm": 0.2974391281604767,
      "learning_rate": 0.00010616371234974845,
      "loss": 0.18600372970104218,
      "mean_token_accuracy": 0.9238637983798981,
      "num_tokens": 86620308.0,
      "step": 7022
    },
    {
      "entropy": 1.145241916179657,
      "epoch": 3.69826224328594,
      "grad_norm": 0.26601076126098633,
      "learning_rate": 0.00010614180410644788,
      "loss": 0.1715870052576065,
      "mean_token_accuracy": 0.9243782013654709,
      "num_tokens": 86632644.0,
      "step": 7023
    },
    {
      "entropy": 1.1596001982688904,
      "epoch": 3.6987888362295944,
      "grad_norm": 0.2737334072589874,
      "learning_rate": 0.0001061198960921859,
      "loss": 0.17031270265579224,
      "mean_token_accuracy": 0.9278070032596588,
      "num_tokens": 86644980.0,
      "step": 7024
    },
    {
      "entropy": 1.1383101642131805,
      "epoch": 3.6993154291732493,
      "grad_norm": 0.29354214668273926,
      "learning_rate": 0.00010609798830826309,
      "loss": 0.17928820848464966,
      "mean_token_accuracy": 0.921865239739418,
      "num_tokens": 86657316.0,
      "step": 7025
    },
    {
      "entropy": 1.1404908895492554,
      "epoch": 3.6998420221169037,
      "grad_norm": 0.2634413242340088,
      "learning_rate": 0.00010607608075597997,
      "loss": 0.17588280141353607,
      "mean_token_accuracy": 0.9216579794883728,
      "num_tokens": 86669652.0,
      "step": 7026
    },
    {
      "entropy": 1.1622402966022491,
      "epoch": 3.700368615060558,
      "grad_norm": 0.2730373442173004,
      "learning_rate": 0.00010605417343663707,
      "loss": 0.1770833134651184,
      "mean_token_accuracy": 0.9242895394563675,
      "num_tokens": 86681988.0,
      "step": 7027
    },
    {
      "entropy": 1.137715607881546,
      "epoch": 3.700895208004213,
      "grad_norm": 0.2774525284767151,
      "learning_rate": 0.00010603226635153487,
      "loss": 0.18399879336357117,
      "mean_token_accuracy": 0.9218821078538895,
      "num_tokens": 86694324.0,
      "step": 7028
    },
    {
      "entropy": 1.0905758738517761,
      "epoch": 3.7014218009478674,
      "grad_norm": 0.2718823254108429,
      "learning_rate": 0.00010601035950197391,
      "loss": 0.1788283884525299,
      "mean_token_accuracy": 0.9250014126300812,
      "num_tokens": 86706660.0,
      "step": 7029
    },
    {
      "entropy": 1.1376757323741913,
      "epoch": 3.701948393891522,
      "grad_norm": 0.2840229570865631,
      "learning_rate": 0.00010598845288925465,
      "loss": 0.19945470988750458,
      "mean_token_accuracy": 0.9160834699869156,
      "num_tokens": 86718996.0,
      "step": 7030
    },
    {
      "entropy": 1.1366423964500427,
      "epoch": 3.7024749868351763,
      "grad_norm": 0.2850939929485321,
      "learning_rate": 0.00010596654651467761,
      "loss": 0.18090331554412842,
      "mean_token_accuracy": 0.9235516041517258,
      "num_tokens": 86731332.0,
      "step": 7031
    },
    {
      "entropy": 1.1042983531951904,
      "epoch": 3.7030015797788307,
      "grad_norm": 0.2823299169540405,
      "learning_rate": 0.00010594464037954315,
      "loss": 0.17295169830322266,
      "mean_token_accuracy": 0.9247632920742035,
      "num_tokens": 86743668.0,
      "step": 7032
    },
    {
      "entropy": 1.1178973019123077,
      "epoch": 3.7035281727224856,
      "grad_norm": 0.2881527543067932,
      "learning_rate": 0.00010592273448515171,
      "loss": 0.19073688983917236,
      "mean_token_accuracy": 0.9142032116651535,
      "num_tokens": 86756004.0,
      "step": 7033
    },
    {
      "entropy": 1.1280372738838196,
      "epoch": 3.70405476566614,
      "grad_norm": 0.2831754982471466,
      "learning_rate": 0.00010590082883280381,
      "loss": 0.18510054051876068,
      "mean_token_accuracy": 0.9240960776805878,
      "num_tokens": 86768340.0,
      "step": 7034
    },
    {
      "entropy": 1.1257420778274536,
      "epoch": 3.7045813586097944,
      "grad_norm": 0.2848588228225708,
      "learning_rate": 0.00010587892342379982,
      "loss": 0.17345714569091797,
      "mean_token_accuracy": 0.9254269748926163,
      "num_tokens": 86780676.0,
      "step": 7035
    },
    {
      "entropy": 1.1075385212898254,
      "epoch": 3.7051079515534493,
      "grad_norm": 0.26853349804878235,
      "learning_rate": 0.0001058570182594401,
      "loss": 0.163071870803833,
      "mean_token_accuracy": 0.9308745414018631,
      "num_tokens": 86793012.0,
      "step": 7036
    },
    {
      "entropy": 1.0885754227638245,
      "epoch": 3.7056345444971037,
      "grad_norm": 0.3040599226951599,
      "learning_rate": 0.00010583511334102507,
      "loss": 0.20325782895088196,
      "mean_token_accuracy": 0.9152310937643051,
      "num_tokens": 86805348.0,
      "step": 7037
    },
    {
      "entropy": 1.143157958984375,
      "epoch": 3.706161137440758,
      "grad_norm": 0.3034161627292633,
      "learning_rate": 0.00010581320866985509,
      "loss": 0.17412713170051575,
      "mean_token_accuracy": 0.9315810650587082,
      "num_tokens": 86817684.0,
      "step": 7038
    },
    {
      "entropy": 1.12423837184906,
      "epoch": 3.706687730384413,
      "grad_norm": 0.32303717732429504,
      "learning_rate": 0.00010579130424723052,
      "loss": 0.19759589433670044,
      "mean_token_accuracy": 0.9174207895994186,
      "num_tokens": 86830020.0,
      "step": 7039
    },
    {
      "entropy": 1.1076728105545044,
      "epoch": 3.7072143233280674,
      "grad_norm": 0.2884438633918762,
      "learning_rate": 0.00010576940007445165,
      "loss": 0.17870476841926575,
      "mean_token_accuracy": 0.9225609600543976,
      "num_tokens": 86842356.0,
      "step": 7040
    },
    {
      "entropy": 1.1127937734127045,
      "epoch": 3.707740916271722,
      "grad_norm": 0.3068588376045227,
      "learning_rate": 0.00010574749615281885,
      "loss": 0.1765388399362564,
      "mean_token_accuracy": 0.9259982705116272,
      "num_tokens": 86854692.0,
      "step": 7041
    },
    {
      "entropy": 1.092573642730713,
      "epoch": 3.7082675092153767,
      "grad_norm": 0.28363847732543945,
      "learning_rate": 0.00010572559248363244,
      "loss": 0.18522685766220093,
      "mean_token_accuracy": 0.9234550446271896,
      "num_tokens": 86867028.0,
      "step": 7042
    },
    {
      "entropy": 1.1002579033374786,
      "epoch": 3.708794102159031,
      "grad_norm": 0.4295247197151184,
      "learning_rate": 0.00010570368906819268,
      "loss": 0.18969866633415222,
      "mean_token_accuracy": 0.9216194003820419,
      "num_tokens": 86879364.0,
      "step": 7043
    },
    {
      "entropy": 1.0092736929655075,
      "epoch": 3.7093206951026856,
      "grad_norm": 0.29199665784835815,
      "learning_rate": 0.00010568178590779986,
      "loss": 0.18738186359405518,
      "mean_token_accuracy": 0.92066690325737,
      "num_tokens": 86891700.0,
      "step": 7044
    },
    {
      "entropy": 1.0848056077957153,
      "epoch": 3.7098472880463405,
      "grad_norm": 0.284700870513916,
      "learning_rate": 0.00010565988300375425,
      "loss": 0.17852607369422913,
      "mean_token_accuracy": 0.9268171042203903,
      "num_tokens": 86904036.0,
      "step": 7045
    },
    {
      "entropy": 1.0516371130943298,
      "epoch": 3.710373880989995,
      "grad_norm": 0.2858789265155792,
      "learning_rate": 0.00010563798035735608,
      "loss": 0.17847073078155518,
      "mean_token_accuracy": 0.9252327531576157,
      "num_tokens": 86916372.0,
      "step": 7046
    },
    {
      "entropy": 1.1216425895690918,
      "epoch": 3.7109004739336493,
      "grad_norm": 0.29138556122779846,
      "learning_rate": 0.00010561607796990556,
      "loss": 0.17626778781414032,
      "mean_token_accuracy": 0.9225746989250183,
      "num_tokens": 86928708.0,
      "step": 7047
    },
    {
      "entropy": 1.0883024036884308,
      "epoch": 3.7114270668773037,
      "grad_norm": 0.32015201449394226,
      "learning_rate": 0.00010559417584270297,
      "loss": 0.18942302465438843,
      "mean_token_accuracy": 0.9230293780565262,
      "num_tokens": 86941044.0,
      "step": 7048
    },
    {
      "entropy": 1.1369044482707977,
      "epoch": 3.711953659820958,
      "grad_norm": 0.3087009787559509,
      "learning_rate": 0.0001055722739770485,
      "loss": 0.19450201094150543,
      "mean_token_accuracy": 0.918872594833374,
      "num_tokens": 86953380.0,
      "step": 7049
    },
    {
      "entropy": 1.0756378769874573,
      "epoch": 3.712480252764613,
      "grad_norm": 0.300651490688324,
      "learning_rate": 0.00010555037237424233,
      "loss": 0.18945585191249847,
      "mean_token_accuracy": 0.9195737093687057,
      "num_tokens": 86965716.0,
      "step": 7050
    },
    {
      "entropy": 1.0762072801589966,
      "epoch": 3.7130068457082674,
      "grad_norm": 0.3127416968345642,
      "learning_rate": 0.00010552847103558457,
      "loss": 0.19951653480529785,
      "mean_token_accuracy": 0.9176425486803055,
      "num_tokens": 86978052.0,
      "step": 7051
    },
    {
      "entropy": 1.0640795826911926,
      "epoch": 3.713533438651922,
      "grad_norm": 0.2833158075809479,
      "learning_rate": 0.00010550656996237545,
      "loss": 0.17779900133609772,
      "mean_token_accuracy": 0.9267347902059555,
      "num_tokens": 86990388.0,
      "step": 7052
    },
    {
      "entropy": 1.0813401639461517,
      "epoch": 3.7140600315955767,
      "grad_norm": 0.2971404790878296,
      "learning_rate": 0.00010548466915591508,
      "loss": 0.18745601177215576,
      "mean_token_accuracy": 0.9218907654285431,
      "num_tokens": 87002724.0,
      "step": 7053
    },
    {
      "entropy": 1.054432064294815,
      "epoch": 3.714586624539231,
      "grad_norm": 0.2983354330062866,
      "learning_rate": 0.00010546276861750355,
      "loss": 0.1724255532026291,
      "mean_token_accuracy": 0.926984503865242,
      "num_tokens": 87015060.0,
      "step": 7054
    },
    {
      "entropy": 1.10925954580307,
      "epoch": 3.7151132174828856,
      "grad_norm": 0.3013254404067993,
      "learning_rate": 0.00010544086834844102,
      "loss": 0.18378786742687225,
      "mean_token_accuracy": 0.9229534566402435,
      "num_tokens": 87027396.0,
      "step": 7055
    },
    {
      "entropy": 1.0945369601249695,
      "epoch": 3.7156398104265405,
      "grad_norm": 0.2869090139865875,
      "learning_rate": 0.00010541896835002758,
      "loss": 0.17971734702587128,
      "mean_token_accuracy": 0.9238785058259964,
      "num_tokens": 87039732.0,
      "step": 7056
    },
    {
      "entropy": 1.0483132302761078,
      "epoch": 3.716166403370195,
      "grad_norm": 0.28307098150253296,
      "learning_rate": 0.00010539706862356329,
      "loss": 0.16245801746845245,
      "mean_token_accuracy": 0.9303528666496277,
      "num_tokens": 87052068.0,
      "step": 7057
    },
    {
      "entropy": 1.0750107169151306,
      "epoch": 3.7166929963138493,
      "grad_norm": 0.280137300491333,
      "learning_rate": 0.0001053751691703482,
      "loss": 0.1640922725200653,
      "mean_token_accuracy": 0.9329541176557541,
      "num_tokens": 87064404.0,
      "step": 7058
    },
    {
      "entropy": 1.0682209730148315,
      "epoch": 3.717219589257504,
      "grad_norm": 0.2897961139678955,
      "learning_rate": 0.00010535326999168235,
      "loss": 0.1660948097705841,
      "mean_token_accuracy": 0.9257662445306778,
      "num_tokens": 87076740.0,
      "step": 7059
    },
    {
      "entropy": 1.0374273657798767,
      "epoch": 3.7177461822011586,
      "grad_norm": 0.27872800827026367,
      "learning_rate": 0.0001053313710888658,
      "loss": 0.17981553077697754,
      "mean_token_accuracy": 0.9291188567876816,
      "num_tokens": 87089076.0,
      "step": 7060
    },
    {
      "entropy": 1.0463912636041641,
      "epoch": 3.718272775144813,
      "grad_norm": 0.29372626543045044,
      "learning_rate": 0.0001053094724631985,
      "loss": 0.1634107530117035,
      "mean_token_accuracy": 0.9343952238559723,
      "num_tokens": 87101412.0,
      "step": 7061
    },
    {
      "entropy": 1.0434607565402985,
      "epoch": 3.7187993680884674,
      "grad_norm": 0.3085126280784607,
      "learning_rate": 0.00010528757411598052,
      "loss": 0.18878869712352753,
      "mean_token_accuracy": 0.9215325862169266,
      "num_tokens": 87113748.0,
      "step": 7062
    },
    {
      "entropy": 1.0086200386285782,
      "epoch": 3.7193259610321223,
      "grad_norm": 0.31608739495277405,
      "learning_rate": 0.00010526567604851178,
      "loss": 0.19069062173366547,
      "mean_token_accuracy": 0.9255263060331345,
      "num_tokens": 87126084.0,
      "step": 7063
    },
    {
      "entropy": 1.043920412659645,
      "epoch": 3.7198525539757767,
      "grad_norm": 0.3183654248714447,
      "learning_rate": 0.00010524377826209229,
      "loss": 0.1812736988067627,
      "mean_token_accuracy": 0.9252707064151764,
      "num_tokens": 87138420.0,
      "step": 7064
    },
    {
      "entropy": 0.9960617870092392,
      "epoch": 3.720379146919431,
      "grad_norm": 0.26266780495643616,
      "learning_rate": 0.00010522188075802195,
      "loss": 0.1656363606452942,
      "mean_token_accuracy": 0.9341549277305603,
      "num_tokens": 87150756.0,
      "step": 7065
    },
    {
      "entropy": 1.0182566493749619,
      "epoch": 3.7209057398630856,
      "grad_norm": 0.27417296171188354,
      "learning_rate": 0.00010519998353760068,
      "loss": 0.16538512706756592,
      "mean_token_accuracy": 0.9307667762041092,
      "num_tokens": 87163092.0,
      "step": 7066
    },
    {
      "entropy": 1.0149471759796143,
      "epoch": 3.7214323328067405,
      "grad_norm": 0.27810150384902954,
      "learning_rate": 0.00010517808660212842,
      "loss": 0.18155770003795624,
      "mean_token_accuracy": 0.9208849668502808,
      "num_tokens": 87175428.0,
      "step": 7067
    },
    {
      "entropy": 1.0324757099151611,
      "epoch": 3.721958925750395,
      "grad_norm": 0.32252272963523865,
      "learning_rate": 0.00010515618995290502,
      "loss": 0.19960591197013855,
      "mean_token_accuracy": 0.9167255610227585,
      "num_tokens": 87187764.0,
      "step": 7068
    },
    {
      "entropy": 1.0477562993764877,
      "epoch": 3.7224855186940493,
      "grad_norm": 0.30965742468833923,
      "learning_rate": 0.00010513429359123046,
      "loss": 0.20774033665657043,
      "mean_token_accuracy": 0.9094702899456024,
      "num_tokens": 87200100.0,
      "step": 7069
    },
    {
      "entropy": 1.0455181300640106,
      "epoch": 3.723012111637704,
      "grad_norm": 0.28178003430366516,
      "learning_rate": 0.0001051123975184045,
      "loss": 0.16794756054878235,
      "mean_token_accuracy": 0.9258260428905487,
      "num_tokens": 87212436.0,
      "step": 7070
    },
    {
      "entropy": 1.027303621172905,
      "epoch": 3.7235387045813586,
      "grad_norm": 0.288570374250412,
      "learning_rate": 0.00010509050173572699,
      "loss": 0.1763126403093338,
      "mean_token_accuracy": 0.9280012547969818,
      "num_tokens": 87224772.0,
      "step": 7071
    },
    {
      "entropy": 1.0355613380670547,
      "epoch": 3.724065297525013,
      "grad_norm": 0.3108993172645569,
      "learning_rate": 0.00010506860624449779,
      "loss": 0.18876023590564728,
      "mean_token_accuracy": 0.9230781495571136,
      "num_tokens": 87237108.0,
      "step": 7072
    },
    {
      "entropy": 1.0131930708885193,
      "epoch": 3.724591890468668,
      "grad_norm": 0.24582242965698242,
      "learning_rate": 0.00010504671104601671,
      "loss": 0.1521173119544983,
      "mean_token_accuracy": 0.9366391003131866,
      "num_tokens": 87249444.0,
      "step": 7073
    },
    {
      "entropy": 1.016406238079071,
      "epoch": 3.7251184834123223,
      "grad_norm": 0.29857704043388367,
      "learning_rate": 0.00010502481614158353,
      "loss": 0.17020674049854279,
      "mean_token_accuracy": 0.9255031496286392,
      "num_tokens": 87261780.0,
      "step": 7074
    },
    {
      "entropy": 0.9883182942867279,
      "epoch": 3.7256450763559767,
      "grad_norm": 0.2852786183357239,
      "learning_rate": 0.00010500292153249798,
      "loss": 0.189723938703537,
      "mean_token_accuracy": 0.9228391647338867,
      "num_tokens": 87274116.0,
      "step": 7075
    },
    {
      "entropy": 1.013285368680954,
      "epoch": 3.7261716692996316,
      "grad_norm": 0.30359360575675964,
      "learning_rate": 0.00010498102722005987,
      "loss": 0.1763514131307602,
      "mean_token_accuracy": 0.9228106886148453,
      "num_tokens": 87286452.0,
      "step": 7076
    },
    {
      "entropy": 0.9790749102830887,
      "epoch": 3.726698262243286,
      "grad_norm": 0.2750509977340698,
      "learning_rate": 0.00010495913320556891,
      "loss": 0.18288496136665344,
      "mean_token_accuracy": 0.9239853918552399,
      "num_tokens": 87298788.0,
      "step": 7077
    },
    {
      "entropy": 1.0321734845638275,
      "epoch": 3.7272248551869405,
      "grad_norm": 0.29339393973350525,
      "learning_rate": 0.00010493723949032486,
      "loss": 0.17784325778484344,
      "mean_token_accuracy": 0.9215517640113831,
      "num_tokens": 87311124.0,
      "step": 7078
    },
    {
      "entropy": 1.030923455953598,
      "epoch": 3.727751448130595,
      "grad_norm": 0.27889496088027954,
      "learning_rate": 0.00010491534607562739,
      "loss": 0.1686307191848755,
      "mean_token_accuracy": 0.9292322397232056,
      "num_tokens": 87323460.0,
      "step": 7079
    },
    {
      "entropy": 1.0209147930145264,
      "epoch": 3.7282780410742498,
      "grad_norm": 0.29664507508277893,
      "learning_rate": 0.0001048934529627762,
      "loss": 0.18868425488471985,
      "mean_token_accuracy": 0.9178652465343475,
      "num_tokens": 87335796.0,
      "step": 7080
    },
    {
      "entropy": 1.016210913658142,
      "epoch": 3.728804634017904,
      "grad_norm": 0.3059195876121521,
      "learning_rate": 0.00010487156015307092,
      "loss": 0.1921241283416748,
      "mean_token_accuracy": 0.9246607571840286,
      "num_tokens": 87348132.0,
      "step": 7081
    },
    {
      "entropy": 1.028878003358841,
      "epoch": 3.7293312269615586,
      "grad_norm": 0.29554203152656555,
      "learning_rate": 0.00010484966764781123,
      "loss": 0.18789461255073547,
      "mean_token_accuracy": 0.9193727970123291,
      "num_tokens": 87360468.0,
      "step": 7082
    },
    {
      "entropy": 0.964047983288765,
      "epoch": 3.729857819905213,
      "grad_norm": 0.2756931781768799,
      "learning_rate": 0.00010482777544829671,
      "loss": 0.17855557799339294,
      "mean_token_accuracy": 0.921904906630516,
      "num_tokens": 87372804.0,
      "step": 7083
    },
    {
      "entropy": 1.0087389796972275,
      "epoch": 3.730384412848868,
      "grad_norm": 0.27971115708351135,
      "learning_rate": 0.00010480588355582708,
      "loss": 0.17549286782741547,
      "mean_token_accuracy": 0.9275772571563721,
      "num_tokens": 87385140.0,
      "step": 7084
    },
    {
      "entropy": 1.0169537961483002,
      "epoch": 3.7309110057925223,
      "grad_norm": 0.2893778383731842,
      "learning_rate": 0.00010478399197170183,
      "loss": 0.17516522109508514,
      "mean_token_accuracy": 0.9268490374088287,
      "num_tokens": 87397476.0,
      "step": 7085
    },
    {
      "entropy": 1.0312487930059433,
      "epoch": 3.7314375987361768,
      "grad_norm": 0.27305132150650024,
      "learning_rate": 0.0001047621006972206,
      "loss": 0.1587783694267273,
      "mean_token_accuracy": 0.9328574538230896,
      "num_tokens": 87409812.0,
      "step": 7086
    },
    {
      "entropy": 1.0589696168899536,
      "epoch": 3.7319641916798316,
      "grad_norm": 0.31207895278930664,
      "learning_rate": 0.00010474020973368293,
      "loss": 0.18918633460998535,
      "mean_token_accuracy": 0.9184137284755707,
      "num_tokens": 87422148.0,
      "step": 7087
    },
    {
      "entropy": 0.9988068491220474,
      "epoch": 3.732490784623486,
      "grad_norm": 0.2731066048145294,
      "learning_rate": 0.00010471831908238834,
      "loss": 0.16324186325073242,
      "mean_token_accuracy": 0.9312383532524109,
      "num_tokens": 87434484.0,
      "step": 7088
    },
    {
      "entropy": 0.9938575476408005,
      "epoch": 3.7330173775671405,
      "grad_norm": 0.27274930477142334,
      "learning_rate": 0.00010469642874463634,
      "loss": 0.1695559173822403,
      "mean_token_accuracy": 0.9315649718046188,
      "num_tokens": 87446820.0,
      "step": 7089
    },
    {
      "entropy": 1.025950849056244,
      "epoch": 3.7335439705107953,
      "grad_norm": 0.3154991567134857,
      "learning_rate": 0.00010467453872172646,
      "loss": 0.20251576602458954,
      "mean_token_accuracy": 0.9197258055210114,
      "num_tokens": 87459156.0,
      "step": 7090
    },
    {
      "entropy": 0.9701757431030273,
      "epoch": 3.7340705634544498,
      "grad_norm": 0.27344241738319397,
      "learning_rate": 0.00010465264901495819,
      "loss": 0.17037051916122437,
      "mean_token_accuracy": 0.9294610321521759,
      "num_tokens": 87471492.0,
      "step": 7091
    },
    {
      "entropy": 1.0319063663482666,
      "epoch": 3.734597156398104,
      "grad_norm": 0.34979090094566345,
      "learning_rate": 0.00010463075962563099,
      "loss": 0.19255414605140686,
      "mean_token_accuracy": 0.9198327660560608,
      "num_tokens": 87483828.0,
      "step": 7092
    },
    {
      "entropy": 1.0084194839000702,
      "epoch": 3.735123749341759,
      "grad_norm": 0.3136915862560272,
      "learning_rate": 0.00010460887055504427,
      "loss": 0.18214285373687744,
      "mean_token_accuracy": 0.9235022515058517,
      "num_tokens": 87496164.0,
      "step": 7093
    },
    {
      "entropy": 0.9943555891513824,
      "epoch": 3.7356503422854135,
      "grad_norm": 0.2898204028606415,
      "learning_rate": 0.0001045869818044975,
      "loss": 0.1824132204055786,
      "mean_token_accuracy": 0.9250787794589996,
      "num_tokens": 87508500.0,
      "step": 7094
    },
    {
      "entropy": 0.9767244309186935,
      "epoch": 3.736176935229068,
      "grad_norm": 0.31859782338142395,
      "learning_rate": 0.0001045650933752901,
      "loss": 0.18570992350578308,
      "mean_token_accuracy": 0.9204951971769333,
      "num_tokens": 87520836.0,
      "step": 7095
    },
    {
      "entropy": 0.9487767368555069,
      "epoch": 3.7367035281727223,
      "grad_norm": 0.30867719650268555,
      "learning_rate": 0.00010454320526872139,
      "loss": 0.19354388117790222,
      "mean_token_accuracy": 0.9184281229972839,
      "num_tokens": 87533172.0,
      "step": 7096
    },
    {
      "entropy": 0.9652345925569534,
      "epoch": 3.737230121116377,
      "grad_norm": 0.28436383605003357,
      "learning_rate": 0.00010452131748609076,
      "loss": 0.15788644552230835,
      "mean_token_accuracy": 0.9355803728103638,
      "num_tokens": 87545508.0,
      "step": 7097
    },
    {
      "entropy": 0.9447768330574036,
      "epoch": 3.7377567140600316,
      "grad_norm": 0.3082371950149536,
      "learning_rate": 0.00010449943002869762,
      "loss": 0.19267767667770386,
      "mean_token_accuracy": 0.9179702401161194,
      "num_tokens": 87557844.0,
      "step": 7098
    },
    {
      "entropy": 0.9711893647909164,
      "epoch": 3.738283307003686,
      "grad_norm": 0.3047843873500824,
      "learning_rate": 0.00010447754289784126,
      "loss": 0.17203961312770844,
      "mean_token_accuracy": 0.9273226410150528,
      "num_tokens": 87570180.0,
      "step": 7099
    },
    {
      "entropy": 0.9295950829982758,
      "epoch": 3.7388098999473405,
      "grad_norm": 0.287088543176651,
      "learning_rate": 0.00010445565609482097,
      "loss": 0.1747625172138214,
      "mean_token_accuracy": 0.9251388758420944,
      "num_tokens": 87582516.0,
      "step": 7100
    },
    {
      "entropy": 1.0014701783657074,
      "epoch": 3.7393364928909953,
      "grad_norm": 0.30902376770973206,
      "learning_rate": 0.00010443376962093607,
      "loss": 0.1791127324104309,
      "mean_token_accuracy": 0.925250843167305,
      "num_tokens": 87594852.0,
      "step": 7101
    },
    {
      "entropy": 0.9752064198255539,
      "epoch": 3.7398630858346498,
      "grad_norm": 0.294033944606781,
      "learning_rate": 0.00010441188347748583,
      "loss": 0.1893167346715927,
      "mean_token_accuracy": 0.9265682995319366,
      "num_tokens": 87607188.0,
      "step": 7102
    },
    {
      "entropy": 0.9655837267637253,
      "epoch": 3.740389678778304,
      "grad_norm": 0.30552423000335693,
      "learning_rate": 0.00010438999766576949,
      "loss": 0.1884727030992508,
      "mean_token_accuracy": 0.92208431661129,
      "num_tokens": 87619524.0,
      "step": 7103
    },
    {
      "entropy": 1.0147232711315155,
      "epoch": 3.740916271721959,
      "grad_norm": 0.3045092523097992,
      "learning_rate": 0.00010436811218708628,
      "loss": 0.19394133985042572,
      "mean_token_accuracy": 0.9220876395702362,
      "num_tokens": 87631860.0,
      "step": 7104
    },
    {
      "entropy": 0.9478363990783691,
      "epoch": 3.7414428646656135,
      "grad_norm": 0.29392796754837036,
      "learning_rate": 0.00010434622704273545,
      "loss": 0.17552950978279114,
      "mean_token_accuracy": 0.9269378036260605,
      "num_tokens": 87644196.0,
      "step": 7105
    },
    {
      "entropy": 0.9850336462259293,
      "epoch": 3.741969457609268,
      "grad_norm": 0.29128676652908325,
      "learning_rate": 0.00010432434223401615,
      "loss": 0.17779776453971863,
      "mean_token_accuracy": 0.9280469864606857,
      "num_tokens": 87656532.0,
      "step": 7106
    },
    {
      "entropy": 0.9837324321269989,
      "epoch": 3.742496050552923,
      "grad_norm": 0.3252773880958557,
      "learning_rate": 0.00010430245776222758,
      "loss": 0.19084057211875916,
      "mean_token_accuracy": 0.9230901747941971,
      "num_tokens": 87668868.0,
      "step": 7107
    },
    {
      "entropy": 1.038240522146225,
      "epoch": 3.743022643496577,
      "grad_norm": 0.28228163719177246,
      "learning_rate": 0.00010428057362866888,
      "loss": 0.17360344529151917,
      "mean_token_accuracy": 0.9263254404067993,
      "num_tokens": 87681204.0,
      "step": 7108
    },
    {
      "entropy": 1.0297512263059616,
      "epoch": 3.7435492364402316,
      "grad_norm": 0.2939990758895874,
      "learning_rate": 0.0001042586898346392,
      "loss": 0.19297507405281067,
      "mean_token_accuracy": 0.9198752194643021,
      "num_tokens": 87693540.0,
      "step": 7109
    },
    {
      "entropy": 0.9940708875656128,
      "epoch": 3.7440758293838865,
      "grad_norm": 0.273837685585022,
      "learning_rate": 0.00010423680638143764,
      "loss": 0.17688986659049988,
      "mean_token_accuracy": 0.9205555766820908,
      "num_tokens": 87705876.0,
      "step": 7110
    },
    {
      "entropy": 1.0277384221553802,
      "epoch": 3.744602422327541,
      "grad_norm": 0.28091904520988464,
      "learning_rate": 0.00010421492327036327,
      "loss": 0.1765642613172531,
      "mean_token_accuracy": 0.9265970140695572,
      "num_tokens": 87718212.0,
      "step": 7111
    },
    {
      "entropy": 0.9914429187774658,
      "epoch": 3.7451290152711953,
      "grad_norm": 0.2665352523326874,
      "learning_rate": 0.00010419304050271519,
      "loss": 0.17894355952739716,
      "mean_token_accuracy": 0.922010600566864,
      "num_tokens": 87730548.0,
      "step": 7112
    },
    {
      "entropy": 1.055834949016571,
      "epoch": 3.7456556082148498,
      "grad_norm": 0.28357288241386414,
      "learning_rate": 0.00010417115807979248,
      "loss": 0.17035986483097076,
      "mean_token_accuracy": 0.932903915643692,
      "num_tokens": 87742884.0,
      "step": 7113
    },
    {
      "entropy": 1.0203563868999481,
      "epoch": 3.746182201158504,
      "grad_norm": 0.3063550293445587,
      "learning_rate": 0.00010414927600289412,
      "loss": 0.18917737901210785,
      "mean_token_accuracy": 0.9191833138465881,
      "num_tokens": 87755220.0,
      "step": 7114
    },
    {
      "entropy": 1.0060118436813354,
      "epoch": 3.746708794102159,
      "grad_norm": 0.28606951236724854,
      "learning_rate": 0.00010412739427331917,
      "loss": 0.18628007173538208,
      "mean_token_accuracy": 0.9235791712999344,
      "num_tokens": 87767556.0,
      "step": 7115
    },
    {
      "entropy": 1.0224921852350235,
      "epoch": 3.7472353870458135,
      "grad_norm": 0.27402663230895996,
      "learning_rate": 0.00010410551289236656,
      "loss": 0.1804238259792328,
      "mean_token_accuracy": 0.9233539551496506,
      "num_tokens": 87779892.0,
      "step": 7116
    },
    {
      "entropy": 1.0892615020275116,
      "epoch": 3.747761979989468,
      "grad_norm": 0.28371381759643555,
      "learning_rate": 0.00010408363186133532,
      "loss": 0.17050835490226746,
      "mean_token_accuracy": 0.928670272231102,
      "num_tokens": 87792228.0,
      "step": 7117
    },
    {
      "entropy": 1.0294643342494965,
      "epoch": 3.748288572933123,
      "grad_norm": 0.2943047881126404,
      "learning_rate": 0.00010406175118152432,
      "loss": 0.18783560395240784,
      "mean_token_accuracy": 0.9241560995578766,
      "num_tokens": 87804564.0,
      "step": 7118
    },
    {
      "entropy": 1.0753727555274963,
      "epoch": 3.748815165876777,
      "grad_norm": 0.32707637548446655,
      "learning_rate": 0.0001040398708542326,
      "loss": 0.18916916847229004,
      "mean_token_accuracy": 0.9227591007947922,
      "num_tokens": 87816900.0,
      "step": 7119
    },
    {
      "entropy": 1.122930258512497,
      "epoch": 3.7493417588204316,
      "grad_norm": 0.30631089210510254,
      "learning_rate": 0.00010401799088075899,
      "loss": 0.16908381879329681,
      "mean_token_accuracy": 0.9244441390037537,
      "num_tokens": 87829236.0,
      "step": 7120
    },
    {
      "entropy": 1.0308923721313477,
      "epoch": 3.7498683517640865,
      "grad_norm": 0.2576659321784973,
      "learning_rate": 0.00010399611126240238,
      "loss": 0.17816726863384247,
      "mean_token_accuracy": 0.9263759553432465,
      "num_tokens": 87841572.0,
      "step": 7121
    },
    {
      "entropy": 1.0776856541633606,
      "epoch": 3.750394944707741,
      "grad_norm": 0.2968147397041321,
      "learning_rate": 0.00010397423200046168,
      "loss": 0.1893785148859024,
      "mean_token_accuracy": 0.9209145456552505,
      "num_tokens": 87853908.0,
      "step": 7122
    },
    {
      "entropy": 1.1022182703018188,
      "epoch": 3.7509215376513954,
      "grad_norm": 0.2902420163154602,
      "learning_rate": 0.0001039523530962357,
      "loss": 0.1806715726852417,
      "mean_token_accuracy": 0.9261004626750946,
      "num_tokens": 87866244.0,
      "step": 7123
    },
    {
      "entropy": 1.0705251097679138,
      "epoch": 3.7514481305950502,
      "grad_norm": 0.29407504200935364,
      "learning_rate": 0.00010393047455102328,
      "loss": 0.18582364916801453,
      "mean_token_accuracy": 0.9204881638288498,
      "num_tokens": 87878580.0,
      "step": 7124
    },
    {
      "entropy": 1.0737611651420593,
      "epoch": 3.7519747235387046,
      "grad_norm": 0.2850419580936432,
      "learning_rate": 0.00010390859636612316,
      "loss": 0.18563367426395416,
      "mean_token_accuracy": 0.9202277064323425,
      "num_tokens": 87890916.0,
      "step": 7125
    },
    {
      "entropy": 1.084518700838089,
      "epoch": 3.752501316482359,
      "grad_norm": 0.27986982464790344,
      "learning_rate": 0.0001038867185428342,
      "loss": 0.1888093650341034,
      "mean_token_accuracy": 0.9146895706653595,
      "num_tokens": 87903252.0,
      "step": 7126
    },
    {
      "entropy": 1.0868968069553375,
      "epoch": 3.753027909426014,
      "grad_norm": 0.2747521996498108,
      "learning_rate": 0.00010386484108245513,
      "loss": 0.1743554174900055,
      "mean_token_accuracy": 0.9248699843883514,
      "num_tokens": 87915588.0,
      "step": 7127
    },
    {
      "entropy": 1.12200728058815,
      "epoch": 3.7535545023696684,
      "grad_norm": 0.27726665139198303,
      "learning_rate": 0.0001038429639862847,
      "loss": 0.17595049738883972,
      "mean_token_accuracy": 0.9274223297834396,
      "num_tokens": 87927924.0,
      "step": 7128
    },
    {
      "entropy": 1.117217406630516,
      "epoch": 3.754081095313323,
      "grad_norm": 0.29257217049598694,
      "learning_rate": 0.00010382108725562161,
      "loss": 0.17220927774906158,
      "mean_token_accuracy": 0.9267859756946564,
      "num_tokens": 87940260.0,
      "step": 7129
    },
    {
      "entropy": 1.0715479254722595,
      "epoch": 3.754607688256977,
      "grad_norm": 0.2643260955810547,
      "learning_rate": 0.00010379921089176454,
      "loss": 0.14874723553657532,
      "mean_token_accuracy": 0.9353399723768234,
      "num_tokens": 87952596.0,
      "step": 7130
    },
    {
      "entropy": 1.1052345633506775,
      "epoch": 3.7551342812006316,
      "grad_norm": 0.2888478934764862,
      "learning_rate": 0.00010377733489601217,
      "loss": 0.16529148817062378,
      "mean_token_accuracy": 0.9293965697288513,
      "num_tokens": 87964932.0,
      "step": 7131
    },
    {
      "entropy": 1.0375074446201324,
      "epoch": 3.7556608741442865,
      "grad_norm": 0.26645591855049133,
      "learning_rate": 0.00010375545926966316,
      "loss": 0.17097528278827667,
      "mean_token_accuracy": 0.9303841888904572,
      "num_tokens": 87977268.0,
      "step": 7132
    },
    {
      "entropy": 1.0503558665513992,
      "epoch": 3.756187467087941,
      "grad_norm": 0.31690800189971924,
      "learning_rate": 0.00010373358401401616,
      "loss": 0.19033551216125488,
      "mean_token_accuracy": 0.9215103983879089,
      "num_tokens": 87989604.0,
      "step": 7133
    },
    {
      "entropy": 1.1111974716186523,
      "epoch": 3.7567140600315954,
      "grad_norm": 0.2914089262485504,
      "learning_rate": 0.00010371170913036974,
      "loss": 0.16939187049865723,
      "mean_token_accuracy": 0.9296749085187912,
      "num_tokens": 88001940.0,
      "step": 7134
    },
    {
      "entropy": 1.149122953414917,
      "epoch": 3.7572406529752502,
      "grad_norm": 0.31661757826805115,
      "learning_rate": 0.00010368983462002251,
      "loss": 0.17850740253925323,
      "mean_token_accuracy": 0.9292630702257156,
      "num_tokens": 88014276.0,
      "step": 7135
    },
    {
      "entropy": 1.104194015264511,
      "epoch": 3.7577672459189047,
      "grad_norm": 0.3128017783164978,
      "learning_rate": 0.000103667960484273,
      "loss": 0.18695411086082458,
      "mean_token_accuracy": 0.9218857288360596,
      "num_tokens": 88026612.0,
      "step": 7136
    },
    {
      "entropy": 1.098086178302765,
      "epoch": 3.758293838862559,
      "grad_norm": 0.289302259683609,
      "learning_rate": 0.0001036460867244198,
      "loss": 0.16781547665596008,
      "mean_token_accuracy": 0.9273310452699661,
      "num_tokens": 88038948.0,
      "step": 7137
    },
    {
      "entropy": 1.0770449936389923,
      "epoch": 3.758820431806214,
      "grad_norm": 0.31516364216804504,
      "learning_rate": 0.00010362421334176138,
      "loss": 0.18100041151046753,
      "mean_token_accuracy": 0.9250895977020264,
      "num_tokens": 88051284.0,
      "step": 7138
    },
    {
      "entropy": 1.082071304321289,
      "epoch": 3.7593470247498684,
      "grad_norm": 0.3046118915081024,
      "learning_rate": 0.00010360234033759621,
      "loss": 0.17777326703071594,
      "mean_token_accuracy": 0.9293287992477417,
      "num_tokens": 88063620.0,
      "step": 7139
    },
    {
      "entropy": 1.1441842019557953,
      "epoch": 3.759873617693523,
      "grad_norm": 0.3186715245246887,
      "learning_rate": 0.00010358046771322287,
      "loss": 0.1779317408800125,
      "mean_token_accuracy": 0.9298745840787888,
      "num_tokens": 88075956.0,
      "step": 7140
    },
    {
      "entropy": 1.1175787448883057,
      "epoch": 3.7604002106371777,
      "grad_norm": 0.3357693552970886,
      "learning_rate": 0.00010355859546993973,
      "loss": 0.20494714379310608,
      "mean_token_accuracy": 0.9127781838178635,
      "num_tokens": 88088292.0,
      "step": 7141
    },
    {
      "entropy": 1.1210220456123352,
      "epoch": 3.760926803580832,
      "grad_norm": 0.30628088116645813,
      "learning_rate": 0.00010353672360904522,
      "loss": 0.1719299852848053,
      "mean_token_accuracy": 0.9311191439628601,
      "num_tokens": 88100628.0,
      "step": 7142
    },
    {
      "entropy": 1.0917105376720428,
      "epoch": 3.7614533965244865,
      "grad_norm": 0.2801833152770996,
      "learning_rate": 0.0001035148521318378,
      "loss": 0.17148740589618683,
      "mean_token_accuracy": 0.9297465682029724,
      "num_tokens": 88112964.0,
      "step": 7143
    },
    {
      "entropy": 1.0566374361515045,
      "epoch": 3.7619799894681414,
      "grad_norm": 0.2890241742134094,
      "learning_rate": 0.00010349298103961577,
      "loss": 0.17020730674266815,
      "mean_token_accuracy": 0.9320740401744843,
      "num_tokens": 88125300.0,
      "step": 7144
    },
    {
      "entropy": 1.1605753600597382,
      "epoch": 3.762506582411796,
      "grad_norm": 0.3184696137905121,
      "learning_rate": 0.00010347111033367754,
      "loss": 0.18983620405197144,
      "mean_token_accuracy": 0.9197225421667099,
      "num_tokens": 88137636.0,
      "step": 7145
    },
    {
      "entropy": 1.1180182695388794,
      "epoch": 3.7630331753554502,
      "grad_norm": 0.30830660462379456,
      "learning_rate": 0.0001034492400153214,
      "loss": 0.17144691944122314,
      "mean_token_accuracy": 0.9264721274375916,
      "num_tokens": 88149972.0,
      "step": 7146
    },
    {
      "entropy": 1.1877989172935486,
      "epoch": 3.7635597682991047,
      "grad_norm": 0.29914480447769165,
      "learning_rate": 0.00010342737008584573,
      "loss": 0.1823132485151291,
      "mean_token_accuracy": 0.9239718466997147,
      "num_tokens": 88162308.0,
      "step": 7147
    },
    {
      "entropy": 1.0980007946491241,
      "epoch": 3.764086361242759,
      "grad_norm": 0.2649677097797394,
      "learning_rate": 0.0001034055005465488,
      "loss": 0.16298440098762512,
      "mean_token_accuracy": 0.9309890568256378,
      "num_tokens": 88174644.0,
      "step": 7148
    },
    {
      "entropy": 1.1103598475456238,
      "epoch": 3.764612954186414,
      "grad_norm": 0.27731263637542725,
      "learning_rate": 0.00010338363139872887,
      "loss": 0.16658708453178406,
      "mean_token_accuracy": 0.9308271110057831,
      "num_tokens": 88186980.0,
      "step": 7149
    },
    {
      "entropy": 1.0874804258346558,
      "epoch": 3.7651395471300684,
      "grad_norm": 0.3050479590892792,
      "learning_rate": 0.00010336176264368418,
      "loss": 0.1813635528087616,
      "mean_token_accuracy": 0.9246079325675964,
      "num_tokens": 88199316.0,
      "step": 7150
    },
    {
      "entropy": 1.0669877976179123,
      "epoch": 3.765666140073723,
      "grad_norm": 0.2986302673816681,
      "learning_rate": 0.00010333989428271294,
      "loss": 0.1748489886522293,
      "mean_token_accuracy": 0.9280883967876434,
      "num_tokens": 88211652.0,
      "step": 7151
    },
    {
      "entropy": 1.130952626466751,
      "epoch": 3.7661927330173777,
      "grad_norm": 0.3100493848323822,
      "learning_rate": 0.00010331802631711335,
      "loss": 0.19673530757427216,
      "mean_token_accuracy": 0.9181911945343018,
      "num_tokens": 88223988.0,
      "step": 7152
    },
    {
      "entropy": 1.1111083626747131,
      "epoch": 3.766719325961032,
      "grad_norm": 0.28959664702415466,
      "learning_rate": 0.00010329615874818359,
      "loss": 0.1911488175392151,
      "mean_token_accuracy": 0.9209056049585342,
      "num_tokens": 88236324.0,
      "step": 7153
    },
    {
      "entropy": 1.1012301445007324,
      "epoch": 3.7672459189046865,
      "grad_norm": 0.2829752564430237,
      "learning_rate": 0.00010327429157722181,
      "loss": 0.16403992474079132,
      "mean_token_accuracy": 0.9249609857797623,
      "num_tokens": 88248660.0,
      "step": 7154
    },
    {
      "entropy": 1.0873798429965973,
      "epoch": 3.7677725118483414,
      "grad_norm": 0.30785346031188965,
      "learning_rate": 0.00010325242480552617,
      "loss": 0.1962447166442871,
      "mean_token_accuracy": 0.9166468977928162,
      "num_tokens": 88260996.0,
      "step": 7155
    },
    {
      "entropy": 1.1623855829238892,
      "epoch": 3.768299104791996,
      "grad_norm": 0.28202804923057556,
      "learning_rate": 0.00010323055843439473,
      "loss": 0.17079147696495056,
      "mean_token_accuracy": 0.9259966462850571,
      "num_tokens": 88273332.0,
      "step": 7156
    },
    {
      "entropy": 1.1565687954425812,
      "epoch": 3.7688256977356502,
      "grad_norm": 0.2951604425907135,
      "learning_rate": 0.00010320869246512557,
      "loss": 0.18850970268249512,
      "mean_token_accuracy": 0.9177795052528381,
      "num_tokens": 88285668.0,
      "step": 7157
    },
    {
      "entropy": 1.0610617399215698,
      "epoch": 3.769352290679305,
      "grad_norm": 0.25743353366851807,
      "learning_rate": 0.00010318682689901677,
      "loss": 0.16640762984752655,
      "mean_token_accuracy": 0.9265040755271912,
      "num_tokens": 88298004.0,
      "step": 7158
    },
    {
      "entropy": 1.0700703561306,
      "epoch": 3.7698788836229595,
      "grad_norm": 0.2738151252269745,
      "learning_rate": 0.00010316496173736636,
      "loss": 0.1697658896446228,
      "mean_token_accuracy": 0.9291802495718002,
      "num_tokens": 88310340.0,
      "step": 7159
    },
    {
      "entropy": 1.1381312012672424,
      "epoch": 3.770405476566614,
      "grad_norm": 0.29495909810066223,
      "learning_rate": 0.00010314309698147226,
      "loss": 0.1771652102470398,
      "mean_token_accuracy": 0.9262112379074097,
      "num_tokens": 88322676.0,
      "step": 7160
    },
    {
      "entropy": 1.1073337495326996,
      "epoch": 3.770932069510269,
      "grad_norm": 0.2812015116214752,
      "learning_rate": 0.00010312123263263258,
      "loss": 0.1764044165611267,
      "mean_token_accuracy": 0.9280906170606613,
      "num_tokens": 88335012.0,
      "step": 7161
    },
    {
      "entropy": 1.0950949639081955,
      "epoch": 3.7714586624539232,
      "grad_norm": 0.2788774371147156,
      "learning_rate": 0.00010309936869214525,
      "loss": 0.1627884954214096,
      "mean_token_accuracy": 0.932257205247879,
      "num_tokens": 88347348.0,
      "step": 7162
    },
    {
      "entropy": 1.105437308549881,
      "epoch": 3.7719852553975777,
      "grad_norm": 0.28457725048065186,
      "learning_rate": 0.00010307750516130815,
      "loss": 0.17091962695121765,
      "mean_token_accuracy": 0.9235321432352066,
      "num_tokens": 88359684.0,
      "step": 7163
    },
    {
      "entropy": 1.091333419084549,
      "epoch": 3.772511848341232,
      "grad_norm": 0.29195544123649597,
      "learning_rate": 0.00010305564204141924,
      "loss": 0.18163371086120605,
      "mean_token_accuracy": 0.9292350262403488,
      "num_tokens": 88372020.0,
      "step": 7164
    },
    {
      "entropy": 1.0755595266819,
      "epoch": 3.7730384412848865,
      "grad_norm": 0.28921985626220703,
      "learning_rate": 0.00010303377933377636,
      "loss": 0.19467858970165253,
      "mean_token_accuracy": 0.9174851328134537,
      "num_tokens": 88384356.0,
      "step": 7165
    },
    {
      "entropy": 1.1279121041297913,
      "epoch": 3.7735650342285414,
      "grad_norm": 0.3318541646003723,
      "learning_rate": 0.00010301191703967742,
      "loss": 0.20074036717414856,
      "mean_token_accuracy": 0.9134417325258255,
      "num_tokens": 88396692.0,
      "step": 7166
    },
    {
      "entropy": 1.0769948661327362,
      "epoch": 3.774091627172196,
      "grad_norm": 0.2806553840637207,
      "learning_rate": 0.00010299005516042022,
      "loss": 0.17609083652496338,
      "mean_token_accuracy": 0.9280765205621719,
      "num_tokens": 88409028.0,
      "step": 7167
    },
    {
      "entropy": 1.100732684135437,
      "epoch": 3.7746182201158502,
      "grad_norm": 0.30120766162872314,
      "learning_rate": 0.00010296819369730258,
      "loss": 0.19822050631046295,
      "mean_token_accuracy": 0.9118075370788574,
      "num_tokens": 88421364.0,
      "step": 7168
    },
    {
      "entropy": 1.095085084438324,
      "epoch": 3.775144813059505,
      "grad_norm": 0.27734094858169556,
      "learning_rate": 0.00010294633265162233,
      "loss": 0.16742560267448425,
      "mean_token_accuracy": 0.9289995431900024,
      "num_tokens": 88433700.0,
      "step": 7169
    },
    {
      "entropy": 1.0630706548690796,
      "epoch": 3.7756714060031595,
      "grad_norm": 0.2720811665058136,
      "learning_rate": 0.00010292447202467716,
      "loss": 0.1635817140340805,
      "mean_token_accuracy": 0.9314915537834167,
      "num_tokens": 88446036.0,
      "step": 7170
    },
    {
      "entropy": 1.0302708148956299,
      "epoch": 3.776197998946814,
      "grad_norm": 0.28553077578544617,
      "learning_rate": 0.00010290261181776488,
      "loss": 0.18625357747077942,
      "mean_token_accuracy": 0.9199012070894241,
      "num_tokens": 88458372.0,
      "step": 7171
    },
    {
      "entropy": 1.0996154844760895,
      "epoch": 3.776724591890469,
      "grad_norm": 0.2932043969631195,
      "learning_rate": 0.00010288075203218315,
      "loss": 0.18521277606487274,
      "mean_token_accuracy": 0.9185535311698914,
      "num_tokens": 88470708.0,
      "step": 7172
    },
    {
      "entropy": 1.123921811580658,
      "epoch": 3.7772511848341233,
      "grad_norm": 0.2973147928714752,
      "learning_rate": 0.0001028588926692297,
      "loss": 0.18819671869277954,
      "mean_token_accuracy": 0.9193670153617859,
      "num_tokens": 88483044.0,
      "step": 7173
    },
    {
      "entropy": 1.0878064036369324,
      "epoch": 3.7777777777777777,
      "grad_norm": 0.27763670682907104,
      "learning_rate": 0.00010283703373020212,
      "loss": 0.17634183168411255,
      "mean_token_accuracy": 0.92812779545784,
      "num_tokens": 88495380.0,
      "step": 7174
    },
    {
      "entropy": 1.1034175157546997,
      "epoch": 3.7783043707214325,
      "grad_norm": 0.30256980657577515,
      "learning_rate": 0.00010281517521639814,
      "loss": 0.18281732499599457,
      "mean_token_accuracy": 0.9237858802080154,
      "num_tokens": 88507716.0,
      "step": 7175
    },
    {
      "entropy": 1.0836420059204102,
      "epoch": 3.778830963665087,
      "grad_norm": 0.28862160444259644,
      "learning_rate": 0.00010279331712911533,
      "loss": 0.16188812255859375,
      "mean_token_accuracy": 0.9328870177268982,
      "num_tokens": 88520052.0,
      "step": 7176
    },
    {
      "entropy": 1.130695253610611,
      "epoch": 3.7793575566087414,
      "grad_norm": 0.29959797859191895,
      "learning_rate": 0.00010277145946965127,
      "loss": 0.18327149748802185,
      "mean_token_accuracy": 0.9227287471294403,
      "num_tokens": 88532388.0,
      "step": 7177
    },
    {
      "entropy": 1.0789493918418884,
      "epoch": 3.7798841495523963,
      "grad_norm": 0.30192306637763977,
      "learning_rate": 0.00010274960223930354,
      "loss": 0.19548943638801575,
      "mean_token_accuracy": 0.9164448976516724,
      "num_tokens": 88544724.0,
      "step": 7178
    },
    {
      "entropy": 1.0900969803333282,
      "epoch": 3.7804107424960507,
      "grad_norm": 0.3188364803791046,
      "learning_rate": 0.00010272774543936968,
      "loss": 0.1938486248254776,
      "mean_token_accuracy": 0.9169595688581467,
      "num_tokens": 88557060.0,
      "step": 7179
    },
    {
      "entropy": 1.0823054313659668,
      "epoch": 3.780937335439705,
      "grad_norm": 0.2802466154098511,
      "learning_rate": 0.00010270588907114716,
      "loss": 0.1587461233139038,
      "mean_token_accuracy": 0.9326314032077789,
      "num_tokens": 88569396.0,
      "step": 7180
    },
    {
      "entropy": 1.068724662065506,
      "epoch": 3.7814639283833595,
      "grad_norm": 0.28032195568084717,
      "learning_rate": 0.00010268403313593349,
      "loss": 0.1607462763786316,
      "mean_token_accuracy": 0.9316603094339371,
      "num_tokens": 88581732.0,
      "step": 7181
    },
    {
      "entropy": 1.0886749029159546,
      "epoch": 3.781990521327014,
      "grad_norm": 0.30621814727783203,
      "learning_rate": 0.00010266217763502613,
      "loss": 0.1921982765197754,
      "mean_token_accuracy": 0.9193924516439438,
      "num_tokens": 88594068.0,
      "step": 7182
    },
    {
      "entropy": 1.0444151908159256,
      "epoch": 3.782517114270669,
      "grad_norm": 0.316541463136673,
      "learning_rate": 0.00010264032256972254,
      "loss": 0.19500555098056793,
      "mean_token_accuracy": 0.9189309179782867,
      "num_tokens": 88606404.0,
      "step": 7183
    },
    {
      "entropy": 1.05389966070652,
      "epoch": 3.7830437072143233,
      "grad_norm": 0.30011123418807983,
      "learning_rate": 0.00010261846794132009,
      "loss": 0.18319684267044067,
      "mean_token_accuracy": 0.92215596139431,
      "num_tokens": 88618740.0,
      "step": 7184
    },
    {
      "entropy": 1.0236373394727707,
      "epoch": 3.7835703001579777,
      "grad_norm": 0.27942049503326416,
      "learning_rate": 0.00010259661375111618,
      "loss": 0.1714269518852234,
      "mean_token_accuracy": 0.9292668700218201,
      "num_tokens": 88631076.0,
      "step": 7185
    },
    {
      "entropy": 1.0853042006492615,
      "epoch": 3.7840968931016326,
      "grad_norm": 0.30122044682502747,
      "learning_rate": 0.00010257476000040816,
      "loss": 0.1864994615316391,
      "mean_token_accuracy": 0.9235980659723282,
      "num_tokens": 88643412.0,
      "step": 7186
    },
    {
      "entropy": 1.0853917300701141,
      "epoch": 3.784623486045287,
      "grad_norm": 0.27920642495155334,
      "learning_rate": 0.00010255290669049336,
      "loss": 0.16710279881954193,
      "mean_token_accuracy": 0.9283783286809921,
      "num_tokens": 88655748.0,
      "step": 7187
    },
    {
      "entropy": 1.0690882205963135,
      "epoch": 3.7851500789889414,
      "grad_norm": 0.357407808303833,
      "learning_rate": 0.00010253105382266906,
      "loss": 0.21778488159179688,
      "mean_token_accuracy": 0.9111634343862534,
      "num_tokens": 88668084.0,
      "step": 7188
    },
    {
      "entropy": 1.0412110090255737,
      "epoch": 3.7856766719325963,
      "grad_norm": 0.30147868394851685,
      "learning_rate": 0.0001025092013982326,
      "loss": 0.17533808946609497,
      "mean_token_accuracy": 0.9283206611871719,
      "num_tokens": 88680420.0,
      "step": 7189
    },
    {
      "entropy": 1.0542258024215698,
      "epoch": 3.7862032648762507,
      "grad_norm": 0.29640671610832214,
      "learning_rate": 0.00010248734941848117,
      "loss": 0.17819085717201233,
      "mean_token_accuracy": 0.9252132475376129,
      "num_tokens": 88692756.0,
      "step": 7190
    },
    {
      "entropy": 1.0278376340866089,
      "epoch": 3.786729857819905,
      "grad_norm": 0.310051292181015,
      "learning_rate": 0.00010246549788471203,
      "loss": 0.19025927782058716,
      "mean_token_accuracy": 0.9163366258144379,
      "num_tokens": 88705092.0,
      "step": 7191
    },
    {
      "entropy": 1.0549766272306442,
      "epoch": 3.78725645076356,
      "grad_norm": 0.32528188824653625,
      "learning_rate": 0.00010244364679822238,
      "loss": 0.20853351056575775,
      "mean_token_accuracy": 0.9162613600492477,
      "num_tokens": 88717428.0,
      "step": 7192
    },
    {
      "entropy": 1.0296104848384857,
      "epoch": 3.7877830437072144,
      "grad_norm": 0.2889884412288666,
      "learning_rate": 0.00010242179616030935,
      "loss": 0.17803248763084412,
      "mean_token_accuracy": 0.93204265832901,
      "num_tokens": 88729764.0,
      "step": 7193
    },
    {
      "entropy": 1.06743985414505,
      "epoch": 3.788309636650869,
      "grad_norm": 0.3100864887237549,
      "learning_rate": 0.00010239994597227014,
      "loss": 0.17786242067813873,
      "mean_token_accuracy": 0.9230290651321411,
      "num_tokens": 88742100.0,
      "step": 7194
    },
    {
      "entropy": 1.0340757369995117,
      "epoch": 3.7888362295945237,
      "grad_norm": 0.2721974551677704,
      "learning_rate": 0.00010237809623540178,
      "loss": 0.18255293369293213,
      "mean_token_accuracy": 0.9202168434858322,
      "num_tokens": 88754436.0,
      "step": 7195
    },
    {
      "entropy": 0.9922773838043213,
      "epoch": 3.789362822538178,
      "grad_norm": 0.2785194218158722,
      "learning_rate": 0.00010235624695100147,
      "loss": 0.1667042225599289,
      "mean_token_accuracy": 0.9307727068662643,
      "num_tokens": 88766772.0,
      "step": 7196
    },
    {
      "entropy": 1.0679470002651215,
      "epoch": 3.7898894154818326,
      "grad_norm": 0.279930055141449,
      "learning_rate": 0.00010233439812036625,
      "loss": 0.1680060774087906,
      "mean_token_accuracy": 0.9259512722492218,
      "num_tokens": 88779108.0,
      "step": 7197
    },
    {
      "entropy": 1.0533702373504639,
      "epoch": 3.790416008425487,
      "grad_norm": 0.27056095004081726,
      "learning_rate": 0.00010231254974479312,
      "loss": 0.16815969347953796,
      "mean_token_accuracy": 0.9305378496646881,
      "num_tokens": 88791444.0,
      "step": 7198
    },
    {
      "entropy": 1.0571512877941132,
      "epoch": 3.7909426013691414,
      "grad_norm": 0.33323702216148376,
      "learning_rate": 0.0001022907018255791,
      "loss": 0.20830880105495453,
      "mean_token_accuracy": 0.913742259144783,
      "num_tokens": 88803780.0,
      "step": 7199
    },
    {
      "entropy": 1.0696991980075836,
      "epoch": 3.7914691943127963,
      "grad_norm": 0.2812376320362091,
      "learning_rate": 0.00010226885436402116,
      "loss": 0.17021866142749786,
      "mean_token_accuracy": 0.9285214692354202,
      "num_tokens": 88816116.0,
      "step": 7200
    },
    {
      "entropy": 1.0224677622318268,
      "epoch": 3.7919957872564507,
      "grad_norm": 0.2666688859462738,
      "learning_rate": 0.0001022470073614163,
      "loss": 0.15672501921653748,
      "mean_token_accuracy": 0.9360812902450562,
      "num_tokens": 88828452.0,
      "step": 7201
    },
    {
      "entropy": 1.0100512653589249,
      "epoch": 3.792522380200105,
      "grad_norm": 0.27672815322875977,
      "learning_rate": 0.00010222516081906139,
      "loss": 0.17214569449424744,
      "mean_token_accuracy": 0.9276096075773239,
      "num_tokens": 88840788.0,
      "step": 7202
    },
    {
      "entropy": 1.0761884450912476,
      "epoch": 3.79304897314376,
      "grad_norm": 0.31685155630111694,
      "learning_rate": 0.0001022033147382534,
      "loss": 0.18890151381492615,
      "mean_token_accuracy": 0.9192356020212173,
      "num_tokens": 88853124.0,
      "step": 7203
    },
    {
      "entropy": 1.0674984157085419,
      "epoch": 3.7935755660874144,
      "grad_norm": 0.28930044174194336,
      "learning_rate": 0.00010218146912028917,
      "loss": 0.18217478692531586,
      "mean_token_accuracy": 0.9230011254549026,
      "num_tokens": 88865460.0,
      "step": 7204
    },
    {
      "entropy": 1.0485822707414627,
      "epoch": 3.794102159031069,
      "grad_norm": 0.3020317256450653,
      "learning_rate": 0.00010215962396646554,
      "loss": 0.19102491438388824,
      "mean_token_accuracy": 0.9197740405797958,
      "num_tokens": 88877796.0,
      "step": 7205
    },
    {
      "entropy": 1.0452226996421814,
      "epoch": 3.7946287519747237,
      "grad_norm": 0.279055655002594,
      "learning_rate": 0.00010213777927807935,
      "loss": 0.17906352877616882,
      "mean_token_accuracy": 0.9296561032533646,
      "num_tokens": 88890132.0,
      "step": 7206
    },
    {
      "entropy": 1.043957456946373,
      "epoch": 3.795155344918378,
      "grad_norm": 0.27935850620269775,
      "learning_rate": 0.00010211593505642738,
      "loss": 0.17773735523223877,
      "mean_token_accuracy": 0.9236277043819427,
      "num_tokens": 88902468.0,
      "step": 7207
    },
    {
      "entropy": 1.0406917482614517,
      "epoch": 3.7956819378620326,
      "grad_norm": 0.2945674657821655,
      "learning_rate": 0.00010209409130280638,
      "loss": 0.17464163899421692,
      "mean_token_accuracy": 0.928332194685936,
      "num_tokens": 88914804.0,
      "step": 7208
    },
    {
      "entropy": 1.0502985268831253,
      "epoch": 3.7962085308056874,
      "grad_norm": 0.30945059657096863,
      "learning_rate": 0.00010207224801851307,
      "loss": 0.17953622341156006,
      "mean_token_accuracy": 0.9266233444213867,
      "num_tokens": 88927140.0,
      "step": 7209
    },
    {
      "entropy": 1.0640967786312103,
      "epoch": 3.796735123749342,
      "grad_norm": 0.3158370554447174,
      "learning_rate": 0.00010205040520484423,
      "loss": 0.17787714302539825,
      "mean_token_accuracy": 0.9302430152893066,
      "num_tokens": 88939476.0,
      "step": 7210
    },
    {
      "entropy": 1.0614418238401413,
      "epoch": 3.7972617166929963,
      "grad_norm": 0.2914262115955353,
      "learning_rate": 0.00010202856286309652,
      "loss": 0.18282458186149597,
      "mean_token_accuracy": 0.9210742712020874,
      "num_tokens": 88951812.0,
      "step": 7211
    },
    {
      "entropy": 1.052042379975319,
      "epoch": 3.7977883096366507,
      "grad_norm": 0.3105827867984772,
      "learning_rate": 0.00010200672099456654,
      "loss": 0.18357163667678833,
      "mean_token_accuracy": 0.92650206387043,
      "num_tokens": 88964148.0,
      "step": 7212
    },
    {
      "entropy": 1.0473193526268005,
      "epoch": 3.7983149025803056,
      "grad_norm": 0.2969362437725067,
      "learning_rate": 0.00010198487960055094,
      "loss": 0.1796019822359085,
      "mean_token_accuracy": 0.9263205826282501,
      "num_tokens": 88976484.0,
      "step": 7213
    },
    {
      "entropy": 0.9956441968679428,
      "epoch": 3.79884149552396,
      "grad_norm": 0.29836389422416687,
      "learning_rate": 0.00010196303868234636,
      "loss": 0.18609370291233063,
      "mean_token_accuracy": 0.9226072877645493,
      "num_tokens": 88988820.0,
      "step": 7214
    },
    {
      "entropy": 1.063538134098053,
      "epoch": 3.7993680884676144,
      "grad_norm": 0.30064263939857483,
      "learning_rate": 0.00010194119824124931,
      "loss": 0.19918721914291382,
      "mean_token_accuracy": 0.918574869632721,
      "num_tokens": 89001156.0,
      "step": 7215
    },
    {
      "entropy": 1.035595253109932,
      "epoch": 3.799894681411269,
      "grad_norm": 0.2827257215976715,
      "learning_rate": 0.0001019193582785563,
      "loss": 0.17484167218208313,
      "mean_token_accuracy": 0.9244060963392258,
      "num_tokens": 89013492.0,
      "step": 7216
    },
    {
      "entropy": 1.0159606635570526,
      "epoch": 3.8004212743549237,
      "grad_norm": 0.2863979935646057,
      "learning_rate": 0.00010189751879556395,
      "loss": 0.18061409890651703,
      "mean_token_accuracy": 0.9228454977273941,
      "num_tokens": 89025828.0,
      "step": 7217
    },
    {
      "entropy": 1.051970288157463,
      "epoch": 3.800947867298578,
      "grad_norm": 0.2826246917247772,
      "learning_rate": 0.00010187567979356867,
      "loss": 0.17374084889888763,
      "mean_token_accuracy": 0.930910050868988,
      "num_tokens": 89038164.0,
      "step": 7218
    },
    {
      "entropy": 1.0662977546453476,
      "epoch": 3.8014744602422326,
      "grad_norm": 0.29204562306404114,
      "learning_rate": 0.00010185384127386692,
      "loss": 0.18853583931922913,
      "mean_token_accuracy": 0.9232958853244781,
      "num_tokens": 89050500.0,
      "step": 7219
    },
    {
      "entropy": 1.131626844406128,
      "epoch": 3.8020010531858874,
      "grad_norm": 0.3503033220767975,
      "learning_rate": 0.00010183200323775515,
      "loss": 0.19388490915298462,
      "mean_token_accuracy": 0.9230544865131378,
      "num_tokens": 89062836.0,
      "step": 7220
    },
    {
      "entropy": 1.082968145608902,
      "epoch": 3.802527646129542,
      "grad_norm": 0.28977474570274353,
      "learning_rate": 0.00010181016568652975,
      "loss": 0.17399579286575317,
      "mean_token_accuracy": 0.9243248850107193,
      "num_tokens": 89075172.0,
      "step": 7221
    },
    {
      "entropy": 1.1116178333759308,
      "epoch": 3.8030542390731963,
      "grad_norm": 0.29531964659690857,
      "learning_rate": 0.00010178832862148706,
      "loss": 0.17993034422397614,
      "mean_token_accuracy": 0.9204618185758591,
      "num_tokens": 89087508.0,
      "step": 7222
    },
    {
      "entropy": 1.042968511581421,
      "epoch": 3.803580832016851,
      "grad_norm": 0.26373812556266785,
      "learning_rate": 0.00010176649204392342,
      "loss": 0.17210234701633453,
      "mean_token_accuracy": 0.9245043843984604,
      "num_tokens": 89099844.0,
      "step": 7223
    },
    {
      "entropy": 1.1081231236457825,
      "epoch": 3.8041074249605056,
      "grad_norm": 0.29073357582092285,
      "learning_rate": 0.00010174465595513516,
      "loss": 0.17837102711200714,
      "mean_token_accuracy": 0.9271133542060852,
      "num_tokens": 89112180.0,
      "step": 7224
    },
    {
      "entropy": 1.0379737168550491,
      "epoch": 3.80463401790416,
      "grad_norm": 0.2728631794452667,
      "learning_rate": 0.00010172282035641858,
      "loss": 0.1719096601009369,
      "mean_token_accuracy": 0.9286081343889236,
      "num_tokens": 89124516.0,
      "step": 7225
    },
    {
      "entropy": 1.0780547708272934,
      "epoch": 3.805160610847815,
      "grad_norm": 0.27084219455718994,
      "learning_rate": 0.00010170098524906994,
      "loss": 0.179472878575325,
      "mean_token_accuracy": 0.9250780194997787,
      "num_tokens": 89136852.0,
      "step": 7226
    },
    {
      "entropy": 1.078784555196762,
      "epoch": 3.8056872037914693,
      "grad_norm": 0.30687907338142395,
      "learning_rate": 0.00010167915063438542,
      "loss": 0.18293949961662292,
      "mean_token_accuracy": 0.9218228608369827,
      "num_tokens": 89149188.0,
      "step": 7227
    },
    {
      "entropy": 1.0638082027435303,
      "epoch": 3.8062137967351237,
      "grad_norm": 0.3069583773612976,
      "learning_rate": 0.00010165731651366122,
      "loss": 0.1985979974269867,
      "mean_token_accuracy": 0.9143917560577393,
      "num_tokens": 89161524.0,
      "step": 7228
    },
    {
      "entropy": 1.0987060964107513,
      "epoch": 3.806740389678778,
      "grad_norm": 0.2627451419830322,
      "learning_rate": 0.00010163548288819351,
      "loss": 0.17108958959579468,
      "mean_token_accuracy": 0.92933489382267,
      "num_tokens": 89173860.0,
      "step": 7229
    },
    {
      "entropy": 1.1162225008010864,
      "epoch": 3.807266982622433,
      "grad_norm": 0.293716698884964,
      "learning_rate": 0.00010161364975927841,
      "loss": 0.1702927052974701,
      "mean_token_accuracy": 0.9316524863243103,
      "num_tokens": 89186196.0,
      "step": 7230
    },
    {
      "entropy": 1.0891296565532684,
      "epoch": 3.8077935755660874,
      "grad_norm": 0.29380878806114197,
      "learning_rate": 0.00010159181712821207,
      "loss": 0.18582729995250702,
      "mean_token_accuracy": 0.9171992391347885,
      "num_tokens": 89198532.0,
      "step": 7231
    },
    {
      "entropy": 1.0300887823104858,
      "epoch": 3.808320168509742,
      "grad_norm": 0.28559356927871704,
      "learning_rate": 0.00010156998499629053,
      "loss": 0.16330835223197937,
      "mean_token_accuracy": 0.9303398281335831,
      "num_tokens": 89210868.0,
      "step": 7232
    },
    {
      "entropy": 1.135662466287613,
      "epoch": 3.8088467614533963,
      "grad_norm": 0.2866886854171753,
      "learning_rate": 0.00010154815336480985,
      "loss": 0.17558231949806213,
      "mean_token_accuracy": 0.9255096316337585,
      "num_tokens": 89223204.0,
      "step": 7233
    },
    {
      "entropy": 1.0409517288208008,
      "epoch": 3.809373354397051,
      "grad_norm": 0.27483031153678894,
      "learning_rate": 0.00010152632223506604,
      "loss": 0.17298176884651184,
      "mean_token_accuracy": 0.9237565845251083,
      "num_tokens": 89235540.0,
      "step": 7234
    },
    {
      "entropy": 1.0508598238229752,
      "epoch": 3.8098999473407056,
      "grad_norm": 0.2898114025592804,
      "learning_rate": 0.00010150449160835511,
      "loss": 0.16894441843032837,
      "mean_token_accuracy": 0.9310885369777679,
      "num_tokens": 89247876.0,
      "step": 7235
    },
    {
      "entropy": 1.07838636636734,
      "epoch": 3.81042654028436,
      "grad_norm": 0.30749765038490295,
      "learning_rate": 0.00010148266148597299,
      "loss": 0.18928919732570648,
      "mean_token_accuracy": 0.9212216436862946,
      "num_tokens": 89260212.0,
      "step": 7236
    },
    {
      "entropy": 1.0341950207948685,
      "epoch": 3.810953133228015,
      "grad_norm": 0.31431862711906433,
      "learning_rate": 0.00010146083186921558,
      "loss": 0.18621139228343964,
      "mean_token_accuracy": 0.9205119013786316,
      "num_tokens": 89272548.0,
      "step": 7237
    },
    {
      "entropy": 1.072775810956955,
      "epoch": 3.8114797261716693,
      "grad_norm": 0.3027096688747406,
      "learning_rate": 0.00010143900275937886,
      "loss": 0.1710643172264099,
      "mean_token_accuracy": 0.9289993047714233,
      "num_tokens": 89284884.0,
      "step": 7238
    },
    {
      "entropy": 1.0736041069030762,
      "epoch": 3.8120063191153237,
      "grad_norm": 0.2878359854221344,
      "learning_rate": 0.0001014171741577586,
      "loss": 0.17023108899593353,
      "mean_token_accuracy": 0.928799033164978,
      "num_tokens": 89297220.0,
      "step": 7239
    },
    {
      "entropy": 1.0706163048744202,
      "epoch": 3.8125329120589786,
      "grad_norm": 0.4564668834209442,
      "learning_rate": 0.00010139534606565073,
      "loss": 0.16438832879066467,
      "mean_token_accuracy": 0.9315458983182907,
      "num_tokens": 89309556.0,
      "step": 7240
    },
    {
      "entropy": 1.119351178407669,
      "epoch": 3.813059505002633,
      "grad_norm": 0.3326112926006317,
      "learning_rate": 0.00010137351848435101,
      "loss": 0.17732514441013336,
      "mean_token_accuracy": 0.9268805235624313,
      "num_tokens": 89321892.0,
      "step": 7241
    },
    {
      "entropy": 1.0823031663894653,
      "epoch": 3.8135860979462874,
      "grad_norm": 0.3071443438529968,
      "learning_rate": 0.00010135169141515522,
      "loss": 0.18601597845554352,
      "mean_token_accuracy": 0.9182300120592117,
      "num_tokens": 89334228.0,
      "step": 7242
    },
    {
      "entropy": 1.0551444292068481,
      "epoch": 3.8141126908899423,
      "grad_norm": 0.294696569442749,
      "learning_rate": 0.00010132986485935909,
      "loss": 0.18259482085704803,
      "mean_token_accuracy": 0.9259582459926605,
      "num_tokens": 89346564.0,
      "step": 7243
    },
    {
      "entropy": 1.066992163658142,
      "epoch": 3.8146392838335967,
      "grad_norm": 0.29047083854675293,
      "learning_rate": 0.0001013080388182583,
      "loss": 0.16876797378063202,
      "mean_token_accuracy": 0.9309166371822357,
      "num_tokens": 89358900.0,
      "step": 7244
    },
    {
      "entropy": 1.1132435202598572,
      "epoch": 3.815165876777251,
      "grad_norm": 0.3349572420120239,
      "learning_rate": 0.00010128621329314867,
      "loss": 0.19489669799804688,
      "mean_token_accuracy": 0.9177320748567581,
      "num_tokens": 89371236.0,
      "step": 7245
    },
    {
      "entropy": 1.1137051582336426,
      "epoch": 3.8156924697209056,
      "grad_norm": 0.34975436329841614,
      "learning_rate": 0.00010126438828532571,
      "loss": 0.17569835484027863,
      "mean_token_accuracy": 0.9270328730344772,
      "num_tokens": 89383572.0,
      "step": 7246
    },
    {
      "entropy": 1.112154096364975,
      "epoch": 3.8162190626645605,
      "grad_norm": 0.29454243183135986,
      "learning_rate": 0.00010124256379608514,
      "loss": 0.162591814994812,
      "mean_token_accuracy": 0.9338550865650177,
      "num_tokens": 89395908.0,
      "step": 7247
    },
    {
      "entropy": 1.0950568318367004,
      "epoch": 3.816745655608215,
      "grad_norm": 0.2949798107147217,
      "learning_rate": 0.0001012207398267225,
      "loss": 0.17738787829875946,
      "mean_token_accuracy": 0.9251436740159988,
      "num_tokens": 89408244.0,
      "step": 7248
    },
    {
      "entropy": 1.1034832000732422,
      "epoch": 3.8172722485518693,
      "grad_norm": 0.28425031900405884,
      "learning_rate": 0.00010119891637853339,
      "loss": 0.1770511418581009,
      "mean_token_accuracy": 0.92559814453125,
      "num_tokens": 89420580.0,
      "step": 7249
    },
    {
      "entropy": 1.098187506198883,
      "epoch": 3.8177988414955237,
      "grad_norm": 0.2896685302257538,
      "learning_rate": 0.0001011770934528133,
      "loss": 0.1746765673160553,
      "mean_token_accuracy": 0.9279729723930359,
      "num_tokens": 89432916.0,
      "step": 7250
    },
    {
      "entropy": 1.0950307846069336,
      "epoch": 3.8183254344391786,
      "grad_norm": 0.28907012939453125,
      "learning_rate": 0.00010115527105085772,
      "loss": 0.1761474758386612,
      "mean_token_accuracy": 0.9249496161937714,
      "num_tokens": 89445252.0,
      "step": 7251
    },
    {
      "entropy": 1.1701862215995789,
      "epoch": 3.818852027382833,
      "grad_norm": 0.34673872590065,
      "learning_rate": 0.00010113344917396215,
      "loss": 0.20624682307243347,
      "mean_token_accuracy": 0.9154643565416336,
      "num_tokens": 89457588.0,
      "step": 7252
    },
    {
      "entropy": 1.1130734384059906,
      "epoch": 3.8193786203264875,
      "grad_norm": 0.2743799090385437,
      "learning_rate": 0.00010111162782342205,
      "loss": 0.17597100138664246,
      "mean_token_accuracy": 0.9302635490894318,
      "num_tokens": 89469924.0,
      "step": 7253
    },
    {
      "entropy": 1.1158782541751862,
      "epoch": 3.8199052132701423,
      "grad_norm": 0.28568530082702637,
      "learning_rate": 0.00010108980700053281,
      "loss": 0.18606042861938477,
      "mean_token_accuracy": 0.9275815039873123,
      "num_tokens": 89482260.0,
      "step": 7254
    },
    {
      "entropy": 1.1550249755382538,
      "epoch": 3.8204318062137967,
      "grad_norm": 0.2865970730781555,
      "learning_rate": 0.00010106798670658979,
      "loss": 0.1869693100452423,
      "mean_token_accuracy": 0.9218897372484207,
      "num_tokens": 89494596.0,
      "step": 7255
    },
    {
      "entropy": 1.131758064031601,
      "epoch": 3.820958399157451,
      "grad_norm": 0.28293687105178833,
      "learning_rate": 0.00010104616694288835,
      "loss": 0.1857099086046219,
      "mean_token_accuracy": 0.9230599403381348,
      "num_tokens": 89506932.0,
      "step": 7256
    },
    {
      "entropy": 1.1339568197727203,
      "epoch": 3.821484992101106,
      "grad_norm": 0.28943586349487305,
      "learning_rate": 0.00010102434771072378,
      "loss": 0.1801375448703766,
      "mean_token_accuracy": 0.9199080318212509,
      "num_tokens": 89519268.0,
      "step": 7257
    },
    {
      "entropy": 1.1382434666156769,
      "epoch": 3.8220115850447605,
      "grad_norm": 0.2906562089920044,
      "learning_rate": 0.00010100252901139131,
      "loss": 0.1690700650215149,
      "mean_token_accuracy": 0.9321166574954987,
      "num_tokens": 89531604.0,
      "step": 7258
    },
    {
      "entropy": 1.1378513276576996,
      "epoch": 3.822538177988415,
      "grad_norm": 0.28166836500167847,
      "learning_rate": 0.00010098071084618632,
      "loss": 0.17545185983181,
      "mean_token_accuracy": 0.9237697720527649,
      "num_tokens": 89543940.0,
      "step": 7259
    },
    {
      "entropy": 1.1416104137897491,
      "epoch": 3.8230647709320698,
      "grad_norm": 0.29689937829971313,
      "learning_rate": 0.00010095889321640395,
      "loss": 0.1880737990140915,
      "mean_token_accuracy": 0.9221426397562027,
      "num_tokens": 89556276.0,
      "step": 7260
    },
    {
      "entropy": 1.1484348773956299,
      "epoch": 3.823591363875724,
      "grad_norm": 0.2994082272052765,
      "learning_rate": 0.0001009370761233394,
      "loss": 0.18645620346069336,
      "mean_token_accuracy": 0.9207324832677841,
      "num_tokens": 89568612.0,
      "step": 7261
    },
    {
      "entropy": 1.1428918242454529,
      "epoch": 3.8241179568193786,
      "grad_norm": 0.30418846011161804,
      "learning_rate": 0.0001009152595682878,
      "loss": 0.19167828559875488,
      "mean_token_accuracy": 0.9225038588047028,
      "num_tokens": 89580948.0,
      "step": 7262
    },
    {
      "entropy": 1.1049898266792297,
      "epoch": 3.824644549763033,
      "grad_norm": 0.25247815251350403,
      "learning_rate": 0.00010089344355254431,
      "loss": 0.1506982445716858,
      "mean_token_accuracy": 0.9416893571615219,
      "num_tokens": 89593284.0,
      "step": 7263
    },
    {
      "entropy": 1.1729646921157837,
      "epoch": 3.8251711427066875,
      "grad_norm": 0.2825002372264862,
      "learning_rate": 0.00010087162807740397,
      "loss": 0.1731032133102417,
      "mean_token_accuracy": 0.9266055077314377,
      "num_tokens": 89605620.0,
      "step": 7264
    },
    {
      "entropy": 1.1801711916923523,
      "epoch": 3.8256977356503423,
      "grad_norm": 0.2951127886772156,
      "learning_rate": 0.00010084981314416189,
      "loss": 0.17518743872642517,
      "mean_token_accuracy": 0.9253078103065491,
      "num_tokens": 89617956.0,
      "step": 7265
    },
    {
      "entropy": 1.1535230875015259,
      "epoch": 3.8262243285939967,
      "grad_norm": 0.2761785686016083,
      "learning_rate": 0.00010082799875411307,
      "loss": 0.17142808437347412,
      "mean_token_accuracy": 0.9257194846868515,
      "num_tokens": 89630292.0,
      "step": 7266
    },
    {
      "entropy": 1.207795798778534,
      "epoch": 3.826750921537651,
      "grad_norm": 0.30591529607772827,
      "learning_rate": 0.0001008061849085525,
      "loss": 0.19267059862613678,
      "mean_token_accuracy": 0.9200548529624939,
      "num_tokens": 89642628.0,
      "step": 7267
    },
    {
      "entropy": 1.1359332203865051,
      "epoch": 3.827277514481306,
      "grad_norm": 0.26560959219932556,
      "learning_rate": 0.00010078437160877514,
      "loss": 0.15736259520053864,
      "mean_token_accuracy": 0.9354649782180786,
      "num_tokens": 89654964.0,
      "step": 7268
    },
    {
      "entropy": 1.1676850616931915,
      "epoch": 3.8278041074249605,
      "grad_norm": 0.283997118473053,
      "learning_rate": 0.00010076255885607593,
      "loss": 0.18263943493366241,
      "mean_token_accuracy": 0.921161338686943,
      "num_tokens": 89667300.0,
      "step": 7269
    },
    {
      "entropy": 1.1453185975551605,
      "epoch": 3.828330700368615,
      "grad_norm": 0.25955888628959656,
      "learning_rate": 0.00010074074665174977,
      "loss": 0.1715111881494522,
      "mean_token_accuracy": 0.9292258322238922,
      "num_tokens": 89679636.0,
      "step": 7270
    },
    {
      "entropy": 1.1053584516048431,
      "epoch": 3.8288572933122698,
      "grad_norm": 0.32442501187324524,
      "learning_rate": 0.00010071893499709153,
      "loss": 0.18358784914016724,
      "mean_token_accuracy": 0.9204039126634598,
      "num_tokens": 89691972.0,
      "step": 7271
    },
    {
      "entropy": 1.1756221950054169,
      "epoch": 3.829383886255924,
      "grad_norm": 0.28485652804374695,
      "learning_rate": 0.00010069712389339596,
      "loss": 0.17328260838985443,
      "mean_token_accuracy": 0.9245188385248184,
      "num_tokens": 89704308.0,
      "step": 7272
    },
    {
      "entropy": 1.2232281565666199,
      "epoch": 3.8299104791995786,
      "grad_norm": 0.30336758494377136,
      "learning_rate": 0.00010067531334195797,
      "loss": 0.17735067009925842,
      "mean_token_accuracy": 0.9247820824384689,
      "num_tokens": 89716644.0,
      "step": 7273
    },
    {
      "entropy": 1.1930607855319977,
      "epoch": 3.8304370721432335,
      "grad_norm": 0.31025582551956177,
      "learning_rate": 0.00010065350334407226,
      "loss": 0.21691036224365234,
      "mean_token_accuracy": 0.9097111821174622,
      "num_tokens": 89728980.0,
      "step": 7274
    },
    {
      "entropy": 1.1593596935272217,
      "epoch": 3.830963665086888,
      "grad_norm": 0.29387667775154114,
      "learning_rate": 0.00010063169390103364,
      "loss": 0.18222147226333618,
      "mean_token_accuracy": 0.9236123710870743,
      "num_tokens": 89741316.0,
      "step": 7275
    },
    {
      "entropy": 1.165407657623291,
      "epoch": 3.8314902580305423,
      "grad_norm": 0.3327574133872986,
      "learning_rate": 0.00010060988501413668,
      "loss": 0.187834233045578,
      "mean_token_accuracy": 0.920792430639267,
      "num_tokens": 89753652.0,
      "step": 7276
    },
    {
      "entropy": 1.1497161388397217,
      "epoch": 3.832016850974197,
      "grad_norm": 0.2776005268096924,
      "learning_rate": 0.00010058807668467615,
      "loss": 0.16925781965255737,
      "mean_token_accuracy": 0.9254584312438965,
      "num_tokens": 89765988.0,
      "step": 7277
    },
    {
      "entropy": 1.149237334728241,
      "epoch": 3.8325434439178516,
      "grad_norm": 0.2985914647579193,
      "learning_rate": 0.00010056626891394665,
      "loss": 0.16763046383857727,
      "mean_token_accuracy": 0.925909012556076,
      "num_tokens": 89778324.0,
      "step": 7278
    },
    {
      "entropy": 1.1808510720729828,
      "epoch": 3.833070036861506,
      "grad_norm": 0.29439499974250793,
      "learning_rate": 0.00010054446170324277,
      "loss": 0.18559673428535461,
      "mean_token_accuracy": 0.9183323979377747,
      "num_tokens": 89790660.0,
      "step": 7279
    },
    {
      "entropy": 1.1405372321605682,
      "epoch": 3.8335966298051605,
      "grad_norm": 0.3132508099079132,
      "learning_rate": 0.00010052265505385907,
      "loss": 0.1839461326599121,
      "mean_token_accuracy": 0.9206763058900833,
      "num_tokens": 89802996.0,
      "step": 7280
    },
    {
      "entropy": 1.1559495031833649,
      "epoch": 3.834123222748815,
      "grad_norm": 0.3002241551876068,
      "learning_rate": 0.00010050084896709014,
      "loss": 0.19972693920135498,
      "mean_token_accuracy": 0.9159600734710693,
      "num_tokens": 89815332.0,
      "step": 7281
    },
    {
      "entropy": 1.1627310514450073,
      "epoch": 3.8346498156924698,
      "grad_norm": 0.2774801552295685,
      "learning_rate": 0.00010047904344423043,
      "loss": 0.16976918280124664,
      "mean_token_accuracy": 0.9299463927745819,
      "num_tokens": 89827668.0,
      "step": 7282
    },
    {
      "entropy": 1.1482534408569336,
      "epoch": 3.835176408636124,
      "grad_norm": 0.2664020359516144,
      "learning_rate": 0.0001004572384865744,
      "loss": 0.16775798797607422,
      "mean_token_accuracy": 0.9295926690101624,
      "num_tokens": 89840004.0,
      "step": 7283
    },
    {
      "entropy": 1.1820363402366638,
      "epoch": 3.8357030015797786,
      "grad_norm": 0.27489718794822693,
      "learning_rate": 0.00010043543409541651,
      "loss": 0.16735351085662842,
      "mean_token_accuracy": 0.9299400001764297,
      "num_tokens": 89852340.0,
      "step": 7284
    },
    {
      "entropy": 1.2132955193519592,
      "epoch": 3.8362295945234335,
      "grad_norm": 0.29108142852783203,
      "learning_rate": 0.00010041363027205115,
      "loss": 0.1773652583360672,
      "mean_token_accuracy": 0.9252279847860336,
      "num_tokens": 89864676.0,
      "step": 7285
    },
    {
      "entropy": 1.1338000297546387,
      "epoch": 3.836756187467088,
      "grad_norm": 0.30709701776504517,
      "learning_rate": 0.0001003918270177727,
      "loss": 0.18275758624076843,
      "mean_token_accuracy": 0.9216789901256561,
      "num_tokens": 89877012.0,
      "step": 7286
    },
    {
      "entropy": 1.1404915750026703,
      "epoch": 3.8372827804107423,
      "grad_norm": 0.29424789547920227,
      "learning_rate": 0.00010037002433387545,
      "loss": 0.18272551894187927,
      "mean_token_accuracy": 0.926551342010498,
      "num_tokens": 89889348.0,
      "step": 7287
    },
    {
      "entropy": 1.1835018694400787,
      "epoch": 3.837809373354397,
      "grad_norm": 0.3396196663379669,
      "learning_rate": 0.00010034822222165377,
      "loss": 0.21048396825790405,
      "mean_token_accuracy": 0.9132413268089294,
      "num_tokens": 89901684.0,
      "step": 7288
    },
    {
      "entropy": 1.1551093757152557,
      "epoch": 3.8383359662980516,
      "grad_norm": 0.28068169951438904,
      "learning_rate": 0.00010032642068240185,
      "loss": 0.17294597625732422,
      "mean_token_accuracy": 0.9275632053613663,
      "num_tokens": 89914020.0,
      "step": 7289
    },
    {
      "entropy": 1.1892271041870117,
      "epoch": 3.838862559241706,
      "grad_norm": 0.2994571626186371,
      "learning_rate": 0.00010030461971741399,
      "loss": 0.18415969610214233,
      "mean_token_accuracy": 0.9226613193750381,
      "num_tokens": 89926356.0,
      "step": 7290
    },
    {
      "entropy": 1.165372908115387,
      "epoch": 3.839389152185361,
      "grad_norm": 0.26454928517341614,
      "learning_rate": 0.00010028281932798431,
      "loss": 0.17179468274116516,
      "mean_token_accuracy": 0.928283154964447,
      "num_tokens": 89938692.0,
      "step": 7291
    },
    {
      "entropy": 1.1823021471500397,
      "epoch": 3.8399157451290153,
      "grad_norm": 0.28512468934059143,
      "learning_rate": 0.00010026101951540703,
      "loss": 0.1671009212732315,
      "mean_token_accuracy": 0.9315878450870514,
      "num_tokens": 89951028.0,
      "step": 7292
    },
    {
      "entropy": 1.1704443991184235,
      "epoch": 3.8404423380726698,
      "grad_norm": 0.2963160276412964,
      "learning_rate": 0.00010023922028097621,
      "loss": 0.1925249844789505,
      "mean_token_accuracy": 0.9225925803184509,
      "num_tokens": 89963364.0,
      "step": 7293
    },
    {
      "entropy": 1.178544044494629,
      "epoch": 3.8409689310163246,
      "grad_norm": 0.289794385433197,
      "learning_rate": 0.00010021742162598606,
      "loss": 0.17334839701652527,
      "mean_token_accuracy": 0.9263816773891449,
      "num_tokens": 89975700.0,
      "step": 7294
    },
    {
      "entropy": 1.192434698343277,
      "epoch": 3.841495523959979,
      "grad_norm": 0.3179748058319092,
      "learning_rate": 0.00010019562355173054,
      "loss": 0.19346535205841064,
      "mean_token_accuracy": 0.9140064567327499,
      "num_tokens": 89988036.0,
      "step": 7295
    },
    {
      "entropy": 1.1856091022491455,
      "epoch": 3.8420221169036335,
      "grad_norm": 0.3246690034866333,
      "learning_rate": 0.00010017382605950372,
      "loss": 0.19298748672008514,
      "mean_token_accuracy": 0.9217430502176285,
      "num_tokens": 90000372.0,
      "step": 7296
    },
    {
      "entropy": 1.215908706188202,
      "epoch": 3.842548709847288,
      "grad_norm": 0.29726147651672363,
      "learning_rate": 0.00010015202915059956,
      "loss": 0.1933974325656891,
      "mean_token_accuracy": 0.9189482480287552,
      "num_tokens": 90012708.0,
      "step": 7297
    },
    {
      "entropy": 1.1829409897327423,
      "epoch": 3.8430753027909423,
      "grad_norm": 0.2635277211666107,
      "learning_rate": 0.00010013023282631204,
      "loss": 0.16917887330055237,
      "mean_token_accuracy": 0.9315038621425629,
      "num_tokens": 90025044.0,
      "step": 7298
    },
    {
      "entropy": 1.1632716357707977,
      "epoch": 3.843601895734597,
      "grad_norm": 0.27100130915641785,
      "learning_rate": 0.00010010843708793507,
      "loss": 0.17784683406352997,
      "mean_token_accuracy": 0.9262017607688904,
      "num_tokens": 90037380.0,
      "step": 7299
    },
    {
      "entropy": 1.24711474776268,
      "epoch": 3.8441284886782516,
      "grad_norm": 0.3065671920776367,
      "learning_rate": 0.00010008664193676251,
      "loss": 0.17322944104671478,
      "mean_token_accuracy": 0.9291545897722244,
      "num_tokens": 90049716.0,
      "step": 7300
    },
    {
      "entropy": 1.1855449378490448,
      "epoch": 3.844655081621906,
      "grad_norm": 0.27872559428215027,
      "learning_rate": 0.00010006484737408825,
      "loss": 0.17799080908298492,
      "mean_token_accuracy": 0.9267636239528656,
      "num_tokens": 90062052.0,
      "step": 7301
    },
    {
      "entropy": 1.1732699871063232,
      "epoch": 3.845181674565561,
      "grad_norm": 0.2870476245880127,
      "learning_rate": 0.00010004305340120612,
      "loss": 0.1874730885028839,
      "mean_token_accuracy": 0.9199872314929962,
      "num_tokens": 90074388.0,
      "step": 7302
    },
    {
      "entropy": 1.1861219704151154,
      "epoch": 3.8457082675092154,
      "grad_norm": 0.3155590295791626,
      "learning_rate": 0.00010002126001940988,
      "loss": 0.196644589304924,
      "mean_token_accuracy": 0.9160917550325394,
      "num_tokens": 90086724.0,
      "step": 7303
    },
    {
      "entropy": 1.2212349772453308,
      "epoch": 3.8462348604528698,
      "grad_norm": 0.2800969183444977,
      "learning_rate": 9.999946722999328e-05,
      "loss": 0.1710227131843567,
      "mean_token_accuracy": 0.9273928701877594,
      "num_tokens": 90099060.0,
      "step": 7304
    },
    {
      "entropy": 1.205515056848526,
      "epoch": 3.8467614533965246,
      "grad_norm": 0.26504290103912354,
      "learning_rate": 9.997767503425e-05,
      "loss": 0.17037883400917053,
      "mean_token_accuracy": 0.9252675473690033,
      "num_tokens": 90111396.0,
      "step": 7305
    },
    {
      "entropy": 1.201006919145584,
      "epoch": 3.847288046340179,
      "grad_norm": 0.30139029026031494,
      "learning_rate": 9.995588343347373e-05,
      "loss": 0.1959807276725769,
      "mean_token_accuracy": 0.913934201002121,
      "num_tokens": 90123732.0,
      "step": 7306
    },
    {
      "entropy": 1.2046195566654205,
      "epoch": 3.8478146392838335,
      "grad_norm": 0.2724730372428894,
      "learning_rate": 9.993409242895808e-05,
      "loss": 0.18013715744018555,
      "mean_token_accuracy": 0.9217778146266937,
      "num_tokens": 90136068.0,
      "step": 7307
    },
    {
      "entropy": 1.2070066034793854,
      "epoch": 3.8483412322274884,
      "grad_norm": 0.2678380012512207,
      "learning_rate": 9.991230202199677e-05,
      "loss": 0.17368824779987335,
      "mean_token_accuracy": 0.9266010373830795,
      "num_tokens": 90148404.0,
      "step": 7308
    },
    {
      "entropy": 1.2179716527462006,
      "epoch": 3.848867825171143,
      "grad_norm": 0.31269174814224243,
      "learning_rate": 9.989051221388328e-05,
      "loss": 0.1937474012374878,
      "mean_token_accuracy": 0.9200000315904617,
      "num_tokens": 90160740.0,
      "step": 7309
    },
    {
      "entropy": 1.2247945666313171,
      "epoch": 3.849394418114797,
      "grad_norm": 0.32107073068618774,
      "learning_rate": 9.986872300591112e-05,
      "loss": 0.20229271054267883,
      "mean_token_accuracy": 0.9178968369960785,
      "num_tokens": 90173076.0,
      "step": 7310
    },
    {
      "entropy": 1.2144478559494019,
      "epoch": 3.849921011058452,
      "grad_norm": 0.3409155607223511,
      "learning_rate": 9.984693439937385e-05,
      "loss": 0.19109617173671722,
      "mean_token_accuracy": 0.9212620854377747,
      "num_tokens": 90185412.0,
      "step": 7311
    },
    {
      "entropy": 1.1986571252346039,
      "epoch": 3.8504476040021065,
      "grad_norm": 0.32058560848236084,
      "learning_rate": 9.98251463955649e-05,
      "loss": 0.18746724724769592,
      "mean_token_accuracy": 0.9225563108921051,
      "num_tokens": 90197748.0,
      "step": 7312
    },
    {
      "entropy": 1.163426399230957,
      "epoch": 3.850974196945761,
      "grad_norm": 0.25540074706077576,
      "learning_rate": 9.980335899577773e-05,
      "loss": 0.16777797043323517,
      "mean_token_accuracy": 0.9263751208782196,
      "num_tokens": 90210084.0,
      "step": 7313
    },
    {
      "entropy": 1.1606984734535217,
      "epoch": 3.8515007898894154,
      "grad_norm": 0.27272337675094604,
      "learning_rate": 9.978157220130566e-05,
      "loss": 0.17516924440860748,
      "mean_token_accuracy": 0.928168535232544,
      "num_tokens": 90222420.0,
      "step": 7314
    },
    {
      "entropy": 1.209740698337555,
      "epoch": 3.85202738283307,
      "grad_norm": 0.2787659168243408,
      "learning_rate": 9.975978601344211e-05,
      "loss": 0.17293882369995117,
      "mean_token_accuracy": 0.9250793009996414,
      "num_tokens": 90234756.0,
      "step": 7315
    },
    {
      "entropy": 1.2494952976703644,
      "epoch": 3.8525539757767246,
      "grad_norm": 0.3164391815662384,
      "learning_rate": 9.97380004334804e-05,
      "loss": 0.18004949390888214,
      "mean_token_accuracy": 0.9228081554174423,
      "num_tokens": 90247092.0,
      "step": 7316
    },
    {
      "entropy": 1.2021544873714447,
      "epoch": 3.853080568720379,
      "grad_norm": 0.31024160981178284,
      "learning_rate": 9.97162154627138e-05,
      "loss": 0.19099017977714539,
      "mean_token_accuracy": 0.9227265566587448,
      "num_tokens": 90259428.0,
      "step": 7317
    },
    {
      "entropy": 1.188399612903595,
      "epoch": 3.8536071616640335,
      "grad_norm": 0.28021976351737976,
      "learning_rate": 9.96944311024355e-05,
      "loss": 0.18051081895828247,
      "mean_token_accuracy": 0.922139585018158,
      "num_tokens": 90271764.0,
      "step": 7318
    },
    {
      "entropy": 1.2218237519264221,
      "epoch": 3.8541337546076884,
      "grad_norm": 0.30597159266471863,
      "learning_rate": 9.967264735393882e-05,
      "loss": 0.18588949739933014,
      "mean_token_accuracy": 0.9208100736141205,
      "num_tokens": 90284100.0,
      "step": 7319
    },
    {
      "entropy": 1.1899022459983826,
      "epoch": 3.854660347551343,
      "grad_norm": 0.2929156422615051,
      "learning_rate": 9.965086421851684e-05,
      "loss": 0.17718812823295593,
      "mean_token_accuracy": 0.9276111572980881,
      "num_tokens": 90296436.0,
      "step": 7320
    },
    {
      "entropy": 1.1932482421398163,
      "epoch": 3.855186940494997,
      "grad_norm": 0.293131947517395,
      "learning_rate": 9.96290816974627e-05,
      "loss": 0.17138488590717316,
      "mean_token_accuracy": 0.926831841468811,
      "num_tokens": 90308772.0,
      "step": 7321
    },
    {
      "entropy": 1.238861232995987,
      "epoch": 3.855713533438652,
      "grad_norm": 0.28433671593666077,
      "learning_rate": 9.960729979206957e-05,
      "loss": 0.17833887040615082,
      "mean_token_accuracy": 0.9238818287849426,
      "num_tokens": 90321108.0,
      "step": 7322
    },
    {
      "entropy": 1.213705599308014,
      "epoch": 3.8562401263823065,
      "grad_norm": 0.29272541403770447,
      "learning_rate": 9.95855185036305e-05,
      "loss": 0.18631236255168915,
      "mean_token_accuracy": 0.9275387078523636,
      "num_tokens": 90333444.0,
      "step": 7323
    },
    {
      "entropy": 1.2195116877555847,
      "epoch": 3.856766719325961,
      "grad_norm": 0.29437002539634705,
      "learning_rate": 9.956373783343847e-05,
      "loss": 0.18267959356307983,
      "mean_token_accuracy": 0.9190680831670761,
      "num_tokens": 90345780.0,
      "step": 7324
    },
    {
      "entropy": 1.1932564079761505,
      "epoch": 3.857293312269616,
      "grad_norm": 0.2885369062423706,
      "learning_rate": 9.954195778278653e-05,
      "loss": 0.17232061922550201,
      "mean_token_accuracy": 0.9254325926303864,
      "num_tokens": 90358116.0,
      "step": 7325
    },
    {
      "entropy": 1.2298594117164612,
      "epoch": 3.8578199052132702,
      "grad_norm": 0.30461183190345764,
      "learning_rate": 9.952017835296754e-05,
      "loss": 0.1930835098028183,
      "mean_token_accuracy": 0.9228019863367081,
      "num_tokens": 90370452.0,
      "step": 7326
    },
    {
      "entropy": 1.2099511325359344,
      "epoch": 3.8583464981569247,
      "grad_norm": 0.2939245402812958,
      "learning_rate": 9.949839954527454e-05,
      "loss": 0.19376038014888763,
      "mean_token_accuracy": 0.9188228696584702,
      "num_tokens": 90382788.0,
      "step": 7327
    },
    {
      "entropy": 1.1923102736473083,
      "epoch": 3.8588730911005795,
      "grad_norm": 0.2934664785861969,
      "learning_rate": 9.947662136100031e-05,
      "loss": 0.1802346110343933,
      "mean_token_accuracy": 0.9260061830282211,
      "num_tokens": 90395124.0,
      "step": 7328
    },
    {
      "entropy": 1.202105551958084,
      "epoch": 3.859399684044234,
      "grad_norm": 0.2979043126106262,
      "learning_rate": 9.945484380143773e-05,
      "loss": 0.19685573875904083,
      "mean_token_accuracy": 0.9163574129343033,
      "num_tokens": 90407460.0,
      "step": 7329
    },
    {
      "entropy": 1.2044208943843842,
      "epoch": 3.8599262769878884,
      "grad_norm": 0.29820647835731506,
      "learning_rate": 9.943306686787964e-05,
      "loss": 0.19858041405677795,
      "mean_token_accuracy": 0.9161760061979294,
      "num_tokens": 90419796.0,
      "step": 7330
    },
    {
      "entropy": 1.2462234199047089,
      "epoch": 3.860452869931543,
      "grad_norm": 0.2908308804035187,
      "learning_rate": 9.941129056161878e-05,
      "loss": 0.18359780311584473,
      "mean_token_accuracy": 0.9262418448925018,
      "num_tokens": 90432132.0,
      "step": 7331
    },
    {
      "entropy": 1.258163183927536,
      "epoch": 3.860979462875197,
      "grad_norm": 0.30769476294517517,
      "learning_rate": 9.938951488394788e-05,
      "loss": 0.19418224692344666,
      "mean_token_accuracy": 0.9224050343036652,
      "num_tokens": 90444468.0,
      "step": 7332
    },
    {
      "entropy": 1.190240353345871,
      "epoch": 3.861506055818852,
      "grad_norm": 0.2786393463611603,
      "learning_rate": 9.936773983615963e-05,
      "loss": 0.17566590011119843,
      "mean_token_accuracy": 0.9270531982183456,
      "num_tokens": 90456804.0,
      "step": 7333
    },
    {
      "entropy": 1.2426856756210327,
      "epoch": 3.8620326487625065,
      "grad_norm": 0.2850218713283539,
      "learning_rate": 9.93459654195467e-05,
      "loss": 0.18113307654857635,
      "mean_token_accuracy": 0.9268236309289932,
      "num_tokens": 90469140.0,
      "step": 7334
    },
    {
      "entropy": 1.20493283867836,
      "epoch": 3.862559241706161,
      "grad_norm": 0.27189159393310547,
      "learning_rate": 9.932419163540169e-05,
      "loss": 0.17980167269706726,
      "mean_token_accuracy": 0.9276924729347229,
      "num_tokens": 90481476.0,
      "step": 7335
    },
    {
      "entropy": 1.2181030809879303,
      "epoch": 3.863085834649816,
      "grad_norm": 0.24642521142959595,
      "learning_rate": 9.930241848501722e-05,
      "loss": 0.1630403995513916,
      "mean_token_accuracy": 0.9312721788883209,
      "num_tokens": 90493812.0,
      "step": 7336
    },
    {
      "entropy": 1.2517006993293762,
      "epoch": 3.8636124275934702,
      "grad_norm": 0.2872298061847687,
      "learning_rate": 9.928064596968581e-05,
      "loss": 0.1801758110523224,
      "mean_token_accuracy": 0.9221181422472,
      "num_tokens": 90506148.0,
      "step": 7337
    },
    {
      "entropy": 1.2343012988567352,
      "epoch": 3.8641390205371247,
      "grad_norm": 0.2791367769241333,
      "learning_rate": 9.925887409069998e-05,
      "loss": 0.17382994294166565,
      "mean_token_accuracy": 0.9254388809204102,
      "num_tokens": 90518484.0,
      "step": 7338
    },
    {
      "entropy": 1.250428467988968,
      "epoch": 3.8646656134807795,
      "grad_norm": 0.29819661378860474,
      "learning_rate": 9.923710284935219e-05,
      "loss": 0.20673494040966034,
      "mean_token_accuracy": 0.9112135022878647,
      "num_tokens": 90530820.0,
      "step": 7339
    },
    {
      "entropy": 1.2244381606578827,
      "epoch": 3.865192206424434,
      "grad_norm": 0.25664618611335754,
      "learning_rate": 9.921533224693488e-05,
      "loss": 0.1699652373790741,
      "mean_token_accuracy": 0.9253463894128799,
      "num_tokens": 90543156.0,
      "step": 7340
    },
    {
      "entropy": 1.1868961453437805,
      "epoch": 3.8657187993680884,
      "grad_norm": 0.27383333444595337,
      "learning_rate": 9.919356228474041e-05,
      "loss": 0.18589892983436584,
      "mean_token_accuracy": 0.920842707157135,
      "num_tokens": 90555492.0,
      "step": 7341
    },
    {
      "entropy": 1.2844567000865936,
      "epoch": 3.8662453923117432,
      "grad_norm": 0.30267804861068726,
      "learning_rate": 9.917179296406116e-05,
      "loss": 0.19701601564884186,
      "mean_token_accuracy": 0.9183336347341537,
      "num_tokens": 90567828.0,
      "step": 7342
    },
    {
      "entropy": 1.3010142147541046,
      "epoch": 3.8667719852553977,
      "grad_norm": 0.3052208423614502,
      "learning_rate": 9.915002428618947e-05,
      "loss": 0.19094228744506836,
      "mean_token_accuracy": 0.9229763150215149,
      "num_tokens": 90580164.0,
      "step": 7343
    },
    {
      "entropy": 1.1697178184986115,
      "epoch": 3.867298578199052,
      "grad_norm": 0.2604224383831024,
      "learning_rate": 9.912825625241759e-05,
      "loss": 0.16901522874832153,
      "mean_token_accuracy": 0.9302792698144913,
      "num_tokens": 90592500.0,
      "step": 7344
    },
    {
      "entropy": 1.2191344797611237,
      "epoch": 3.867825171142707,
      "grad_norm": 0.27005285024642944,
      "learning_rate": 9.910648886403779e-05,
      "loss": 0.192753404378891,
      "mean_token_accuracy": 0.9172493368387222,
      "num_tokens": 90604836.0,
      "step": 7345
    },
    {
      "entropy": 1.262347787618637,
      "epoch": 3.8683517640863614,
      "grad_norm": 0.29758551716804504,
      "learning_rate": 9.908472212234223e-05,
      "loss": 0.2073061764240265,
      "mean_token_accuracy": 0.9140929877758026,
      "num_tokens": 90617172.0,
      "step": 7346
    },
    {
      "entropy": 1.2290584444999695,
      "epoch": 3.868878357030016,
      "grad_norm": 0.287990003824234,
      "learning_rate": 9.906295602862312e-05,
      "loss": 0.19012236595153809,
      "mean_token_accuracy": 0.9192764163017273,
      "num_tokens": 90629508.0,
      "step": 7347
    },
    {
      "entropy": 1.205379992723465,
      "epoch": 3.8694049499736702,
      "grad_norm": 0.2572006583213806,
      "learning_rate": 9.904119058417256e-05,
      "loss": 0.16958723962306976,
      "mean_token_accuracy": 0.9291314631700516,
      "num_tokens": 90641844.0,
      "step": 7348
    },
    {
      "entropy": 1.2571605145931244,
      "epoch": 3.8699315429173247,
      "grad_norm": 0.2561599314212799,
      "learning_rate": 9.901942579028263e-05,
      "loss": 0.15259096026420593,
      "mean_token_accuracy": 0.9404455125331879,
      "num_tokens": 90654180.0,
      "step": 7349
    },
    {
      "entropy": 1.190529614686966,
      "epoch": 3.8704581358609795,
      "grad_norm": 0.41509267687797546,
      "learning_rate": 9.89976616482454e-05,
      "loss": 0.16672082245349884,
      "mean_token_accuracy": 0.9305219650268555,
      "num_tokens": 90666516.0,
      "step": 7350
    },
    {
      "entropy": 1.2418386042118073,
      "epoch": 3.870984728804634,
      "grad_norm": 0.3323068916797638,
      "learning_rate": 9.897589815935286e-05,
      "loss": 0.20139282941818237,
      "mean_token_accuracy": 0.9165514260530472,
      "num_tokens": 90678852.0,
      "step": 7351
    },
    {
      "entropy": 1.2107313871383667,
      "epoch": 3.8715113217482884,
      "grad_norm": 0.26844340562820435,
      "learning_rate": 9.895413532489701e-05,
      "loss": 0.15874214470386505,
      "mean_token_accuracy": 0.9347341060638428,
      "num_tokens": 90691188.0,
      "step": 7352
    },
    {
      "entropy": 1.1857129633426666,
      "epoch": 3.8720379146919433,
      "grad_norm": 0.2747436463832855,
      "learning_rate": 9.893237314616976e-05,
      "loss": 0.17956000566482544,
      "mean_token_accuracy": 0.9220409989356995,
      "num_tokens": 90703524.0,
      "step": 7353
    },
    {
      "entropy": 1.198570877313614,
      "epoch": 3.8725645076355977,
      "grad_norm": 0.2919812798500061,
      "learning_rate": 9.891061162446302e-05,
      "loss": 0.1921466886997223,
      "mean_token_accuracy": 0.916918158531189,
      "num_tokens": 90715860.0,
      "step": 7354
    },
    {
      "entropy": 1.2124747335910797,
      "epoch": 3.873091100579252,
      "grad_norm": 0.2959701120853424,
      "learning_rate": 9.888885076106861e-05,
      "loss": 0.17341776192188263,
      "mean_token_accuracy": 0.9281175881624222,
      "num_tokens": 90728196.0,
      "step": 7355
    },
    {
      "entropy": 1.1563901603221893,
      "epoch": 3.873617693522907,
      "grad_norm": 0.2615200877189636,
      "learning_rate": 9.886709055727834e-05,
      "loss": 0.16795092821121216,
      "mean_token_accuracy": 0.9334702044725418,
      "num_tokens": 90740532.0,
      "step": 7356
    },
    {
      "entropy": 1.2142418324947357,
      "epoch": 3.8741442864665614,
      "grad_norm": 0.3013656437397003,
      "learning_rate": 9.88453310143841e-05,
      "loss": 0.1777857393026352,
      "mean_token_accuracy": 0.9271297752857208,
      "num_tokens": 90752868.0,
      "step": 7357
    },
    {
      "entropy": 1.1635612845420837,
      "epoch": 3.874670879410216,
      "grad_norm": 0.28417861461639404,
      "learning_rate": 9.88235721336775e-05,
      "loss": 0.17131461203098297,
      "mean_token_accuracy": 0.9289181679487228,
      "num_tokens": 90765204.0,
      "step": 7358
    },
    {
      "entropy": 1.2065666019916534,
      "epoch": 3.8751974723538707,
      "grad_norm": 0.2814251780509949,
      "learning_rate": 9.880181391645027e-05,
      "loss": 0.17419341206550598,
      "mean_token_accuracy": 0.9259565472602844,
      "num_tokens": 90777540.0,
      "step": 7359
    },
    {
      "entropy": 1.1907231509685516,
      "epoch": 3.875724065297525,
      "grad_norm": 0.3082534968852997,
      "learning_rate": 9.87800563639941e-05,
      "loss": 0.18956147134304047,
      "mean_token_accuracy": 0.9228996187448502,
      "num_tokens": 90789876.0,
      "step": 7360
    },
    {
      "entropy": 1.2383562326431274,
      "epoch": 3.8762506582411795,
      "grad_norm": 0.2990850508213043,
      "learning_rate": 9.875829947760057e-05,
      "loss": 0.1762770563364029,
      "mean_token_accuracy": 0.9269504696130753,
      "num_tokens": 90802212.0,
      "step": 7361
    },
    {
      "entropy": 1.2204852104187012,
      "epoch": 3.876777251184834,
      "grad_norm": 0.29158973693847656,
      "learning_rate": 9.87365432585613e-05,
      "loss": 0.1674453616142273,
      "mean_token_accuracy": 0.9312193691730499,
      "num_tokens": 90814548.0,
      "step": 7362
    },
    {
      "entropy": 1.1590887606143951,
      "epoch": 3.877303844128489,
      "grad_norm": 0.26443982124328613,
      "learning_rate": 9.871478770816778e-05,
      "loss": 0.1666359156370163,
      "mean_token_accuracy": 0.9319517016410828,
      "num_tokens": 90826884.0,
      "step": 7363
    },
    {
      "entropy": 1.212668776512146,
      "epoch": 3.8778304370721433,
      "grad_norm": 0.30860036611557007,
      "learning_rate": 9.869303282771156e-05,
      "loss": 0.17729061841964722,
      "mean_token_accuracy": 0.924604594707489,
      "num_tokens": 90839220.0,
      "step": 7364
    },
    {
      "entropy": 1.2208824157714844,
      "epoch": 3.8783570300157977,
      "grad_norm": 0.31199607253074646,
      "learning_rate": 9.867127861848407e-05,
      "loss": 0.19631695747375488,
      "mean_token_accuracy": 0.9166485667228699,
      "num_tokens": 90851556.0,
      "step": 7365
    },
    {
      "entropy": 1.1799960136413574,
      "epoch": 3.878883622959452,
      "grad_norm": 0.301144540309906,
      "learning_rate": 9.864952508177673e-05,
      "loss": 0.1875683218240738,
      "mean_token_accuracy": 0.921687513589859,
      "num_tokens": 90863892.0,
      "step": 7366
    },
    {
      "entropy": 1.1755280196666718,
      "epoch": 3.879410215903107,
      "grad_norm": 0.29750579595565796,
      "learning_rate": 9.862777221888095e-05,
      "loss": 0.19362355768680573,
      "mean_token_accuracy": 0.9189414232969284,
      "num_tokens": 90876228.0,
      "step": 7367
    },
    {
      "entropy": 1.2063259780406952,
      "epoch": 3.8799368088467614,
      "grad_norm": 0.3054865896701813,
      "learning_rate": 9.860602003108807e-05,
      "loss": 0.18168985843658447,
      "mean_token_accuracy": 0.92440365254879,
      "num_tokens": 90888564.0,
      "step": 7368
    },
    {
      "entropy": 1.2309964001178741,
      "epoch": 3.880463401790416,
      "grad_norm": 0.30633923411369324,
      "learning_rate": 9.858426851968936e-05,
      "loss": 0.18840034306049347,
      "mean_token_accuracy": 0.9172258079051971,
      "num_tokens": 90900900.0,
      "step": 7369
    },
    {
      "entropy": 1.254058837890625,
      "epoch": 3.8809899947340707,
      "grad_norm": 0.2868705689907074,
      "learning_rate": 9.856251768597605e-05,
      "loss": 0.17972050607204437,
      "mean_token_accuracy": 0.9242666363716125,
      "num_tokens": 90913236.0,
      "step": 7370
    },
    {
      "entropy": 1.2371594905853271,
      "epoch": 3.881516587677725,
      "grad_norm": 0.33569571375846863,
      "learning_rate": 9.854076753123941e-05,
      "loss": 0.19579902291297913,
      "mean_token_accuracy": 0.9211303889751434,
      "num_tokens": 90925572.0,
      "step": 7371
    },
    {
      "entropy": 1.260173887014389,
      "epoch": 3.8820431806213795,
      "grad_norm": 0.2709573209285736,
      "learning_rate": 9.851901805677066e-05,
      "loss": 0.17761409282684326,
      "mean_token_accuracy": 0.9270109385251999,
      "num_tokens": 90937908.0,
      "step": 7372
    },
    {
      "entropy": 1.2311949133872986,
      "epoch": 3.8825697735650344,
      "grad_norm": 0.2965104579925537,
      "learning_rate": 9.849726926386087e-05,
      "loss": 0.17649513483047485,
      "mean_token_accuracy": 0.9215144068002701,
      "num_tokens": 90950244.0,
      "step": 7373
    },
    {
      "entropy": 1.2264364063739777,
      "epoch": 3.883096366508689,
      "grad_norm": 0.2903778553009033,
      "learning_rate": 9.847552115380117e-05,
      "loss": 0.19147197902202606,
      "mean_token_accuracy": 0.918977215886116,
      "num_tokens": 90962580.0,
      "step": 7374
    },
    {
      "entropy": 1.2604758143424988,
      "epoch": 3.8836229594523433,
      "grad_norm": 0.2860509753227234,
      "learning_rate": 9.845377372788259e-05,
      "loss": 0.16984225809574127,
      "mean_token_accuracy": 0.9307971298694611,
      "num_tokens": 90974916.0,
      "step": 7375
    },
    {
      "entropy": 1.1899789869785309,
      "epoch": 3.884149552395998,
      "grad_norm": 0.2879489064216614,
      "learning_rate": 9.843202698739616e-05,
      "loss": 0.17775698006153107,
      "mean_token_accuracy": 0.9248269200325012,
      "num_tokens": 90987252.0,
      "step": 7376
    },
    {
      "entropy": 1.281430572271347,
      "epoch": 3.8846761453396526,
      "grad_norm": 0.29807573556900024,
      "learning_rate": 9.841028093363284e-05,
      "loss": 0.18883194029331207,
      "mean_token_accuracy": 0.9234910309314728,
      "num_tokens": 90999588.0,
      "step": 7377
    },
    {
      "entropy": 1.2089340388774872,
      "epoch": 3.885202738283307,
      "grad_norm": 0.2950511872768402,
      "learning_rate": 9.838853556788366e-05,
      "loss": 0.20106419920921326,
      "mean_token_accuracy": 0.9162346571683884,
      "num_tokens": 91011924.0,
      "step": 7378
    },
    {
      "entropy": 1.2650335431098938,
      "epoch": 3.8857293312269614,
      "grad_norm": 0.3102474808692932,
      "learning_rate": 9.83667908914394e-05,
      "loss": 0.1975744515657425,
      "mean_token_accuracy": 0.9211560487747192,
      "num_tokens": 91024260.0,
      "step": 7379
    },
    {
      "entropy": 1.2446538507938385,
      "epoch": 3.8862559241706163,
      "grad_norm": 0.28649139404296875,
      "learning_rate": 9.834504690559099e-05,
      "loss": 0.1754469871520996,
      "mean_token_accuracy": 0.926554948091507,
      "num_tokens": 91036596.0,
      "step": 7380
    },
    {
      "entropy": 1.2238378524780273,
      "epoch": 3.8867825171142707,
      "grad_norm": 0.2706204056739807,
      "learning_rate": 9.83233036116292e-05,
      "loss": 0.16986840963363647,
      "mean_token_accuracy": 0.929158017039299,
      "num_tokens": 91048932.0,
      "step": 7381
    },
    {
      "entropy": 1.2629105746746063,
      "epoch": 3.887309110057925,
      "grad_norm": 0.30980411171913147,
      "learning_rate": 9.830156101084483e-05,
      "loss": 0.18923744559288025,
      "mean_token_accuracy": 0.9220970869064331,
      "num_tokens": 91061268.0,
      "step": 7382
    },
    {
      "entropy": 1.2490929961204529,
      "epoch": 3.8878357030015795,
      "grad_norm": 0.2837015390396118,
      "learning_rate": 9.827981910452859e-05,
      "loss": 0.20252351462841034,
      "mean_token_accuracy": 0.9161817729473114,
      "num_tokens": 91073604.0,
      "step": 7383
    },
    {
      "entropy": 1.3015010058879852,
      "epoch": 3.8883622959452344,
      "grad_norm": 0.27492520213127136,
      "learning_rate": 9.825807789397115e-05,
      "loss": 0.1569894552230835,
      "mean_token_accuracy": 0.9341099411249161,
      "num_tokens": 91085940.0,
      "step": 7384
    },
    {
      "entropy": 1.3030857145786285,
      "epoch": 3.888888888888889,
      "grad_norm": 0.28954070806503296,
      "learning_rate": 9.82363373804632e-05,
      "loss": 0.17589737474918365,
      "mean_token_accuracy": 0.9284959137439728,
      "num_tokens": 91098276.0,
      "step": 7385
    },
    {
      "entropy": 1.2944543659687042,
      "epoch": 3.8894154818325433,
      "grad_norm": 0.27544423937797546,
      "learning_rate": 9.821459756529536e-05,
      "loss": 0.16199910640716553,
      "mean_token_accuracy": 0.9316005408763885,
      "num_tokens": 91110612.0,
      "step": 7386
    },
    {
      "entropy": 1.2723598182201385,
      "epoch": 3.889942074776198,
      "grad_norm": 0.27726244926452637,
      "learning_rate": 9.819285844975819e-05,
      "loss": 0.17507168650627136,
      "mean_token_accuracy": 0.9228856265544891,
      "num_tokens": 91122948.0,
      "step": 7387
    },
    {
      "entropy": 1.3930054008960724,
      "epoch": 3.8904686677198526,
      "grad_norm": 0.3249645233154297,
      "learning_rate": 9.817112003514218e-05,
      "loss": 0.174981027841568,
      "mean_token_accuracy": 0.9266958832740784,
      "num_tokens": 91135284.0,
      "step": 7388
    },
    {
      "entropy": 1.3319167494773865,
      "epoch": 3.890995260663507,
      "grad_norm": 0.2998090982437134,
      "learning_rate": 9.81493823227378e-05,
      "loss": 0.18659012019634247,
      "mean_token_accuracy": 0.9238382875919342,
      "num_tokens": 91147620.0,
      "step": 7389
    },
    {
      "entropy": 1.280248075723648,
      "epoch": 3.891521853607162,
      "grad_norm": 0.30350276827812195,
      "learning_rate": 9.812764531383556e-05,
      "loss": 0.1922396868467331,
      "mean_token_accuracy": 0.9191283881664276,
      "num_tokens": 91159956.0,
      "step": 7390
    },
    {
      "entropy": 1.3184870779514313,
      "epoch": 3.8920484465508163,
      "grad_norm": 0.2840585708618164,
      "learning_rate": 9.810590900972576e-05,
      "loss": 0.1652616560459137,
      "mean_token_accuracy": 0.9287358373403549,
      "num_tokens": 91172292.0,
      "step": 7391
    },
    {
      "entropy": 1.3189331591129303,
      "epoch": 3.8925750394944707,
      "grad_norm": 0.30320435762405396,
      "learning_rate": 9.808417341169886e-05,
      "loss": 0.1749056875705719,
      "mean_token_accuracy": 0.9255077093839645,
      "num_tokens": 91184628.0,
      "step": 7392
    },
    {
      "entropy": 1.2997994422912598,
      "epoch": 3.8931016324381256,
      "grad_norm": 0.29542285203933716,
      "learning_rate": 9.806243852104516e-05,
      "loss": 0.18183457851409912,
      "mean_token_accuracy": 0.9283244013786316,
      "num_tokens": 91196964.0,
      "step": 7393
    },
    {
      "entropy": 1.273093044757843,
      "epoch": 3.89362822538178,
      "grad_norm": 0.2934871017932892,
      "learning_rate": 9.804070433905486e-05,
      "loss": 0.17550939321517944,
      "mean_token_accuracy": 0.9281349778175354,
      "num_tokens": 91209300.0,
      "step": 7394
    },
    {
      "entropy": 1.247098594903946,
      "epoch": 3.8941548183254344,
      "grad_norm": 0.26843369007110596,
      "learning_rate": 9.801897086701827e-05,
      "loss": 0.16914314031600952,
      "mean_token_accuracy": 0.9286849945783615,
      "num_tokens": 91221636.0,
      "step": 7395
    },
    {
      "entropy": 1.3457880020141602,
      "epoch": 3.894681411269089,
      "grad_norm": 0.31451231241226196,
      "learning_rate": 9.799723810622552e-05,
      "loss": 0.1843279004096985,
      "mean_token_accuracy": 0.9233946204185486,
      "num_tokens": 91233972.0,
      "step": 7396
    },
    {
      "entropy": 1.3243741989135742,
      "epoch": 3.8952080042127437,
      "grad_norm": 0.30700138211250305,
      "learning_rate": 9.797550605796681e-05,
      "loss": 0.18327546119689941,
      "mean_token_accuracy": 0.9218372255563736,
      "num_tokens": 91246308.0,
      "step": 7397
    },
    {
      "entropy": 1.2985079884529114,
      "epoch": 3.895734597156398,
      "grad_norm": 0.32867226004600525,
      "learning_rate": 9.79537747235322e-05,
      "loss": 0.1934511512517929,
      "mean_token_accuracy": 0.9179651439189911,
      "num_tokens": 91258644.0,
      "step": 7398
    },
    {
      "entropy": 1.3588362634181976,
      "epoch": 3.8962611901000526,
      "grad_norm": 0.3065076470375061,
      "learning_rate": 9.793204410421179e-05,
      "loss": 0.18493646383285522,
      "mean_token_accuracy": 0.9219991266727448,
      "num_tokens": 91270980.0,
      "step": 7399
    },
    {
      "entropy": 1.272426277399063,
      "epoch": 3.896787783043707,
      "grad_norm": 0.27240192890167236,
      "learning_rate": 9.79103142012956e-05,
      "loss": 0.16781708598136902,
      "mean_token_accuracy": 0.9306810200214386,
      "num_tokens": 91283316.0,
      "step": 7400
    },
    {
      "entropy": 1.3233053982257843,
      "epoch": 3.897314375987362,
      "grad_norm": 0.33670109510421753,
      "learning_rate": 9.788858501607357e-05,
      "loss": 0.19102466106414795,
      "mean_token_accuracy": 0.9206659644842148,
      "num_tokens": 91295652.0,
      "step": 7401
    },
    {
      "entropy": 1.305526077747345,
      "epoch": 3.8978409689310163,
      "grad_norm": 0.3031172752380371,
      "learning_rate": 9.786685654983567e-05,
      "loss": 0.18682445585727692,
      "mean_token_accuracy": 0.9239467233419418,
      "num_tokens": 91307988.0,
      "step": 7402
    },
    {
      "entropy": 1.3286968767642975,
      "epoch": 3.8983675618746707,
      "grad_norm": 0.30504098534584045,
      "learning_rate": 9.784512880387178e-05,
      "loss": 0.1931457668542862,
      "mean_token_accuracy": 0.9222109168767929,
      "num_tokens": 91320324.0,
      "step": 7403
    },
    {
      "entropy": 1.2467556595802307,
      "epoch": 3.8988941548183256,
      "grad_norm": 0.2830975651741028,
      "learning_rate": 9.782340177947174e-05,
      "loss": 0.1753942221403122,
      "mean_token_accuracy": 0.9269160330295563,
      "num_tokens": 91332660.0,
      "step": 7404
    },
    {
      "entropy": 1.2635121941566467,
      "epoch": 3.89942074776198,
      "grad_norm": 0.268989622592926,
      "learning_rate": 9.780167547792534e-05,
      "loss": 0.16541332006454468,
      "mean_token_accuracy": 0.9283698350191116,
      "num_tokens": 91344996.0,
      "step": 7405
    },
    {
      "entropy": 1.2748238146305084,
      "epoch": 3.8999473407056344,
      "grad_norm": 0.2895947992801666,
      "learning_rate": 9.777994990052242e-05,
      "loss": 0.18233062326908112,
      "mean_token_accuracy": 0.9232377111911774,
      "num_tokens": 91357332.0,
      "step": 7406
    },
    {
      "entropy": 1.371468424797058,
      "epoch": 3.9004739336492893,
      "grad_norm": 0.33178168535232544,
      "learning_rate": 9.775822504855261e-05,
      "loss": 0.17663978040218353,
      "mean_token_accuracy": 0.9220523536205292,
      "num_tokens": 91369668.0,
      "step": 7407
    },
    {
      "entropy": 1.3484316170215607,
      "epoch": 3.9010005265929437,
      "grad_norm": 0.3129909634590149,
      "learning_rate": 9.773650092330566e-05,
      "loss": 0.17559215426445007,
      "mean_token_accuracy": 0.923889771103859,
      "num_tokens": 91382004.0,
      "step": 7408
    },
    {
      "entropy": 1.334187537431717,
      "epoch": 3.901527119536598,
      "grad_norm": 0.3251679837703705,
      "learning_rate": 9.771477752607115e-05,
      "loss": 0.19099709391593933,
      "mean_token_accuracy": 0.9230151027441025,
      "num_tokens": 91394340.0,
      "step": 7409
    },
    {
      "entropy": 1.3830526471138,
      "epoch": 3.902053712480253,
      "grad_norm": 0.30658549070358276,
      "learning_rate": 9.769305485813872e-05,
      "loss": 0.17843928933143616,
      "mean_token_accuracy": 0.9258236736059189,
      "num_tokens": 91406676.0,
      "step": 7410
    },
    {
      "entropy": 1.3029418587684631,
      "epoch": 3.9025803054239074,
      "grad_norm": 0.26187410950660706,
      "learning_rate": 9.767133292079787e-05,
      "loss": 0.1668580174446106,
      "mean_token_accuracy": 0.9293637126684189,
      "num_tokens": 91419012.0,
      "step": 7411
    },
    {
      "entropy": 1.2614350020885468,
      "epoch": 3.903106898367562,
      "grad_norm": 0.2766633927822113,
      "learning_rate": 9.76496117153381e-05,
      "loss": 0.16704759001731873,
      "mean_token_accuracy": 0.9306674897670746,
      "num_tokens": 91431348.0,
      "step": 7412
    },
    {
      "entropy": 1.3585382103919983,
      "epoch": 3.9036334913112163,
      "grad_norm": 0.3252298831939697,
      "learning_rate": 9.762789124304894e-05,
      "loss": 0.196746826171875,
      "mean_token_accuracy": 0.9180266261100769,
      "num_tokens": 91443684.0,
      "step": 7413
    },
    {
      "entropy": 1.3883097469806671,
      "epoch": 3.9041600842548707,
      "grad_norm": 0.29954347014427185,
      "learning_rate": 9.760617150521976e-05,
      "loss": 0.19299542903900146,
      "mean_token_accuracy": 0.9197171628475189,
      "num_tokens": 91456020.0,
      "step": 7414
    },
    {
      "entropy": 1.31036975979805,
      "epoch": 3.9046866771985256,
      "grad_norm": 0.284512996673584,
      "learning_rate": 9.758445250313994e-05,
      "loss": 0.1800897717475891,
      "mean_token_accuracy": 0.9242556691169739,
      "num_tokens": 91468356.0,
      "step": 7415
    },
    {
      "entropy": 1.38027885556221,
      "epoch": 3.90521327014218,
      "grad_norm": 0.3107808828353882,
      "learning_rate": 9.756273423809883e-05,
      "loss": 0.18819507956504822,
      "mean_token_accuracy": 0.9225617349147797,
      "num_tokens": 91480692.0,
      "step": 7416
    },
    {
      "entropy": 1.3482477962970734,
      "epoch": 3.9057398630858344,
      "grad_norm": 0.28120309114456177,
      "learning_rate": 9.754101671138569e-05,
      "loss": 0.18053530156612396,
      "mean_token_accuracy": 0.9240847080945969,
      "num_tokens": 91493028.0,
      "step": 7417
    },
    {
      "entropy": 1.3876953423023224,
      "epoch": 3.9062664560294893,
      "grad_norm": 0.3125147223472595,
      "learning_rate": 9.751929992428976e-05,
      "loss": 0.20853109657764435,
      "mean_token_accuracy": 0.9141730666160583,
      "num_tokens": 91505364.0,
      "step": 7418
    },
    {
      "entropy": 1.3991923928260803,
      "epoch": 3.9067930489731437,
      "grad_norm": 0.2972346544265747,
      "learning_rate": 9.749758387810021e-05,
      "loss": 0.17821037769317627,
      "mean_token_accuracy": 0.9238766133785248,
      "num_tokens": 91517700.0,
      "step": 7419
    },
    {
      "entropy": 1.362459808588028,
      "epoch": 3.907319641916798,
      "grad_norm": 0.29278430342674255,
      "learning_rate": 9.747586857410629e-05,
      "loss": 0.1727725863456726,
      "mean_token_accuracy": 0.9270545095205307,
      "num_tokens": 91530036.0,
      "step": 7420
    },
    {
      "entropy": 1.3600943982601166,
      "epoch": 3.907846234860453,
      "grad_norm": 0.2770959138870239,
      "learning_rate": 9.745415401359706e-05,
      "loss": 0.17293483018875122,
      "mean_token_accuracy": 0.9269449561834335,
      "num_tokens": 91542372.0,
      "step": 7421
    },
    {
      "entropy": 1.363686740398407,
      "epoch": 3.9083728278041074,
      "grad_norm": 0.2894051969051361,
      "learning_rate": 9.743244019786157e-05,
      "loss": 0.18016286194324493,
      "mean_token_accuracy": 0.9236745089292526,
      "num_tokens": 91554708.0,
      "step": 7422
    },
    {
      "entropy": 1.3273597359657288,
      "epoch": 3.908899420747762,
      "grad_norm": 0.28462719917297363,
      "learning_rate": 9.741072712818885e-05,
      "loss": 0.16882094740867615,
      "mean_token_accuracy": 0.9264349341392517,
      "num_tokens": 91567044.0,
      "step": 7423
    },
    {
      "entropy": 1.4484579861164093,
      "epoch": 3.9094260136914167,
      "grad_norm": 0.31760042905807495,
      "learning_rate": 9.738901480586789e-05,
      "loss": 0.17673248052597046,
      "mean_token_accuracy": 0.9217555373907089,
      "num_tokens": 91579380.0,
      "step": 7424
    },
    {
      "entropy": 1.3021227717399597,
      "epoch": 3.909952606635071,
      "grad_norm": 0.29617711901664734,
      "learning_rate": 9.736730323218761e-05,
      "loss": 0.17839474976062775,
      "mean_token_accuracy": 0.9257320016622543,
      "num_tokens": 91591716.0,
      "step": 7425
    },
    {
      "entropy": 1.3438735902309418,
      "epoch": 3.9104791995787256,
      "grad_norm": 0.2710460126399994,
      "learning_rate": 9.73455924084369e-05,
      "loss": 0.1708822101354599,
      "mean_token_accuracy": 0.9289853870868683,
      "num_tokens": 91604052.0,
      "step": 7426
    },
    {
      "entropy": 1.3481594324111938,
      "epoch": 3.9110057925223805,
      "grad_norm": 0.3210081160068512,
      "learning_rate": 9.732388233590463e-05,
      "loss": 0.20448635518550873,
      "mean_token_accuracy": 0.9112216532230377,
      "num_tokens": 91616388.0,
      "step": 7427
    },
    {
      "entropy": 1.2938117980957031,
      "epoch": 3.911532385466035,
      "grad_norm": 0.26161590218544006,
      "learning_rate": 9.730217301587959e-05,
      "loss": 0.1704331338405609,
      "mean_token_accuracy": 0.9322764575481415,
      "num_tokens": 91628724.0,
      "step": 7428
    },
    {
      "entropy": 1.3291106820106506,
      "epoch": 3.9120589784096893,
      "grad_norm": 0.3009251058101654,
      "learning_rate": 9.728046444965051e-05,
      "loss": 0.17591515183448792,
      "mean_token_accuracy": 0.9239979386329651,
      "num_tokens": 91641060.0,
      "step": 7429
    },
    {
      "entropy": 1.3399436473846436,
      "epoch": 3.9125855713533437,
      "grad_norm": 0.2723159193992615,
      "learning_rate": 9.725875663850615e-05,
      "loss": 0.17526356875896454,
      "mean_token_accuracy": 0.9254540950059891,
      "num_tokens": 91653396.0,
      "step": 7430
    },
    {
      "entropy": 1.4018910825252533,
      "epoch": 3.913112164296998,
      "grad_norm": 0.295527845621109,
      "learning_rate": 9.723704958373511e-05,
      "loss": 0.17586205899715424,
      "mean_token_accuracy": 0.923242911696434,
      "num_tokens": 91665732.0,
      "step": 7431
    },
    {
      "entropy": 1.3623782396316528,
      "epoch": 3.913638757240653,
      "grad_norm": 0.30898094177246094,
      "learning_rate": 9.721534328662609e-05,
      "loss": 0.1770631968975067,
      "mean_token_accuracy": 0.9300126433372498,
      "num_tokens": 91678068.0,
      "step": 7432
    },
    {
      "entropy": 1.3543266355991364,
      "epoch": 3.9141653501843074,
      "grad_norm": 0.28661686182022095,
      "learning_rate": 9.719363774846756e-05,
      "loss": 0.1654135286808014,
      "mean_token_accuracy": 0.930344969034195,
      "num_tokens": 91690404.0,
      "step": 7433
    },
    {
      "entropy": 1.3302703201770782,
      "epoch": 3.914691943127962,
      "grad_norm": 0.29969698190689087,
      "learning_rate": 9.717193297054816e-05,
      "loss": 0.18720367550849915,
      "mean_token_accuracy": 0.9195693731307983,
      "num_tokens": 91702740.0,
      "step": 7434
    },
    {
      "entropy": 1.3499610722064972,
      "epoch": 3.9152185360716167,
      "grad_norm": 0.2968071401119232,
      "learning_rate": 9.715022895415634e-05,
      "loss": 0.18096867203712463,
      "mean_token_accuracy": 0.9255775958299637,
      "num_tokens": 91715076.0,
      "step": 7435
    },
    {
      "entropy": 1.3046875596046448,
      "epoch": 3.915745129015271,
      "grad_norm": 0.2661040723323822,
      "learning_rate": 9.712852570058053e-05,
      "loss": 0.16175732016563416,
      "mean_token_accuracy": 0.9289435148239136,
      "num_tokens": 91727412.0,
      "step": 7436
    },
    {
      "entropy": 1.2640199065208435,
      "epoch": 3.9162717219589256,
      "grad_norm": 0.26960012316703796,
      "learning_rate": 9.710682321110913e-05,
      "loss": 0.16636931896209717,
      "mean_token_accuracy": 0.9285288751125336,
      "num_tokens": 91739748.0,
      "step": 7437
    },
    {
      "entropy": 1.302351325750351,
      "epoch": 3.9167983149025805,
      "grad_norm": 0.26437661051750183,
      "learning_rate": 9.708512148703049e-05,
      "loss": 0.1572932004928589,
      "mean_token_accuracy": 0.9320330768823624,
      "num_tokens": 91752084.0,
      "step": 7438
    },
    {
      "entropy": 1.3512612879276276,
      "epoch": 3.917324907846235,
      "grad_norm": 0.3063144385814667,
      "learning_rate": 9.706342052963289e-05,
      "loss": 0.17273718118667603,
      "mean_token_accuracy": 0.9298366010189056,
      "num_tokens": 91764420.0,
      "step": 7439
    },
    {
      "entropy": 1.3201676607131958,
      "epoch": 3.9178515007898893,
      "grad_norm": 0.2926127314567566,
      "learning_rate": 9.70417203402046e-05,
      "loss": 0.17421020567417145,
      "mean_token_accuracy": 0.922990620136261,
      "num_tokens": 91776756.0,
      "step": 7440
    },
    {
      "entropy": 1.3591834604740143,
      "epoch": 3.918378093733544,
      "grad_norm": 0.3088493347167969,
      "learning_rate": 9.702002092003388e-05,
      "loss": 0.17949983477592468,
      "mean_token_accuracy": 0.9272099882364273,
      "num_tokens": 91789092.0,
      "step": 7441
    },
    {
      "entropy": 1.2428484857082367,
      "epoch": 3.9189046866771986,
      "grad_norm": 0.29576629400253296,
      "learning_rate": 9.699832227040885e-05,
      "loss": 0.17247886955738068,
      "mean_token_accuracy": 0.9230233579874039,
      "num_tokens": 91801428.0,
      "step": 7442
    },
    {
      "entropy": 1.266256421804428,
      "epoch": 3.919431279620853,
      "grad_norm": 0.28943321108818054,
      "learning_rate": 9.697662439261768e-05,
      "loss": 0.18631991744041443,
      "mean_token_accuracy": 0.9180447310209274,
      "num_tokens": 91813764.0,
      "step": 7443
    },
    {
      "entropy": 1.3520332872867584,
      "epoch": 3.919957872564508,
      "grad_norm": 0.3416178822517395,
      "learning_rate": 9.695492728794837e-05,
      "loss": 0.20197802782058716,
      "mean_token_accuracy": 0.9203462600708008,
      "num_tokens": 91826100.0,
      "step": 7444
    },
    {
      "entropy": 1.3061773180961609,
      "epoch": 3.9204844655081623,
      "grad_norm": 0.31303954124450684,
      "learning_rate": 9.693323095768902e-05,
      "loss": 0.1945120394229889,
      "mean_token_accuracy": 0.9196473807096481,
      "num_tokens": 91838436.0,
      "step": 7445
    },
    {
      "entropy": 1.2885983288288116,
      "epoch": 3.9210110584518167,
      "grad_norm": 0.31333503127098083,
      "learning_rate": 9.691153540312759e-05,
      "loss": 0.17745260894298553,
      "mean_token_accuracy": 0.9222434908151627,
      "num_tokens": 91850772.0,
      "step": 7446
    },
    {
      "entropy": 1.320429652929306,
      "epoch": 3.921537651395471,
      "grad_norm": 0.30087795853614807,
      "learning_rate": 9.688984062555197e-05,
      "loss": 0.18600033223628998,
      "mean_token_accuracy": 0.9226914793252945,
      "num_tokens": 91863108.0,
      "step": 7447
    },
    {
      "entropy": 1.2709153592586517,
      "epoch": 3.9220642443391256,
      "grad_norm": 0.325582891702652,
      "learning_rate": 9.686814662625017e-05,
      "loss": 0.189141646027565,
      "mean_token_accuracy": 0.9198153913021088,
      "num_tokens": 91875444.0,
      "step": 7448
    },
    {
      "entropy": 1.3018944561481476,
      "epoch": 3.9225908372827805,
      "grad_norm": 0.3107694685459137,
      "learning_rate": 9.684645340650993e-05,
      "loss": 0.1884014904499054,
      "mean_token_accuracy": 0.9218618273735046,
      "num_tokens": 91887780.0,
      "step": 7449
    },
    {
      "entropy": 1.3278419375419617,
      "epoch": 3.923117430226435,
      "grad_norm": 0.28061795234680176,
      "learning_rate": 9.682476096761907e-05,
      "loss": 0.17283833026885986,
      "mean_token_accuracy": 0.9281010776758194,
      "num_tokens": 91900116.0,
      "step": 7450
    },
    {
      "entropy": 1.2768584489822388,
      "epoch": 3.9236440231700893,
      "grad_norm": 0.266977459192276,
      "learning_rate": 9.680306931086543e-05,
      "loss": 0.16279029846191406,
      "mean_token_accuracy": 0.926258385181427,
      "num_tokens": 91912452.0,
      "step": 7451
    },
    {
      "entropy": 1.2815205752849579,
      "epoch": 3.924170616113744,
      "grad_norm": 0.2692156136035919,
      "learning_rate": 9.678137843753659e-05,
      "loss": 0.16042062640190125,
      "mean_token_accuracy": 0.9334524720907211,
      "num_tokens": 91924788.0,
      "step": 7452
    },
    {
      "entropy": 1.341881901025772,
      "epoch": 3.9246972090573986,
      "grad_norm": 0.292605459690094,
      "learning_rate": 9.675968834892029e-05,
      "loss": 0.17764052748680115,
      "mean_token_accuracy": 0.9222121387720108,
      "num_tokens": 91937124.0,
      "step": 7453
    },
    {
      "entropy": 1.3380673825740814,
      "epoch": 3.925223802001053,
      "grad_norm": 0.27861467003822327,
      "learning_rate": 9.673799904630408e-05,
      "loss": 0.17420029640197754,
      "mean_token_accuracy": 0.92889104783535,
      "num_tokens": 91949460.0,
      "step": 7454
    },
    {
      "entropy": 1.3240340948104858,
      "epoch": 3.925750394944708,
      "grad_norm": 0.29774072766304016,
      "learning_rate": 9.671631053097561e-05,
      "loss": 0.1879979372024536,
      "mean_token_accuracy": 0.9188611954450607,
      "num_tokens": 91961796.0,
      "step": 7455
    },
    {
      "entropy": 1.3600371181964874,
      "epoch": 3.9262769878883623,
      "grad_norm": 0.28510600328445435,
      "learning_rate": 9.669462280422234e-05,
      "loss": 0.17808692157268524,
      "mean_token_accuracy": 0.9247149080038071,
      "num_tokens": 91974132.0,
      "step": 7456
    },
    {
      "entropy": 1.336561381816864,
      "epoch": 3.9268035808320167,
      "grad_norm": 0.2943764626979828,
      "learning_rate": 9.667293586733179e-05,
      "loss": 0.1932375282049179,
      "mean_token_accuracy": 0.9196823388338089,
      "num_tokens": 91986468.0,
      "step": 7457
    },
    {
      "entropy": 1.275018721818924,
      "epoch": 3.9273301737756716,
      "grad_norm": 0.29437708854675293,
      "learning_rate": 9.665124972159134e-05,
      "loss": 0.16790060698986053,
      "mean_token_accuracy": 0.9270847886800766,
      "num_tokens": 91998804.0,
      "step": 7458
    },
    {
      "entropy": 1.29185551404953,
      "epoch": 3.927856766719326,
      "grad_norm": 0.30873504281044006,
      "learning_rate": 9.66295643682884e-05,
      "loss": 0.19838179647922516,
      "mean_token_accuracy": 0.9141746461391449,
      "num_tokens": 92011140.0,
      "step": 7459
    },
    {
      "entropy": 1.2768851220607758,
      "epoch": 3.9283833596629805,
      "grad_norm": 0.29629775881767273,
      "learning_rate": 9.66078798087103e-05,
      "loss": 0.1842423528432846,
      "mean_token_accuracy": 0.9186161458492279,
      "num_tokens": 92023476.0,
      "step": 7460
    },
    {
      "entropy": 1.3205865621566772,
      "epoch": 3.9289099526066353,
      "grad_norm": 0.2964547276496887,
      "learning_rate": 9.65861960441443e-05,
      "loss": 0.18825949728488922,
      "mean_token_accuracy": 0.9202955067157745,
      "num_tokens": 92035812.0,
      "step": 7461
    },
    {
      "entropy": 1.3484995067119598,
      "epoch": 3.9294365455502898,
      "grad_norm": 0.3002781271934509,
      "learning_rate": 9.656451307587769e-05,
      "loss": 0.1901530921459198,
      "mean_token_accuracy": 0.9150425940752029,
      "num_tokens": 92048148.0,
      "step": 7462
    },
    {
      "entropy": 1.3245843350887299,
      "epoch": 3.929963138493944,
      "grad_norm": 0.30892708897590637,
      "learning_rate": 9.65428309051976e-05,
      "loss": 0.1806984543800354,
      "mean_token_accuracy": 0.9242218285799026,
      "num_tokens": 92060484.0,
      "step": 7463
    },
    {
      "entropy": 1.2997255325317383,
      "epoch": 3.9304897314375986,
      "grad_norm": 0.28640472888946533,
      "learning_rate": 9.652114953339122e-05,
      "loss": 0.1763458102941513,
      "mean_token_accuracy": 0.9265084564685822,
      "num_tokens": 92072820.0,
      "step": 7464
    },
    {
      "entropy": 1.30248361825943,
      "epoch": 3.931016324381253,
      "grad_norm": 0.28299543261528015,
      "learning_rate": 9.649946896174564e-05,
      "loss": 0.17997360229492188,
      "mean_token_accuracy": 0.9239591509103775,
      "num_tokens": 92085156.0,
      "step": 7465
    },
    {
      "entropy": 1.3352027833461761,
      "epoch": 3.931542917324908,
      "grad_norm": 0.28869009017944336,
      "learning_rate": 9.64777891915479e-05,
      "loss": 0.16349303722381592,
      "mean_token_accuracy": 0.9304029792547226,
      "num_tokens": 92097492.0,
      "step": 7466
    },
    {
      "entropy": 1.3092913031578064,
      "epoch": 3.9320695102685623,
      "grad_norm": 0.28738799691200256,
      "learning_rate": 9.6456110224085e-05,
      "loss": 0.17659634351730347,
      "mean_token_accuracy": 0.924927607178688,
      "num_tokens": 92109828.0,
      "step": 7467
    },
    {
      "entropy": 1.3021188080310822,
      "epoch": 3.9325961032122168,
      "grad_norm": 0.28562963008880615,
      "learning_rate": 9.643443206064386e-05,
      "loss": 0.17467309534549713,
      "mean_token_accuracy": 0.9223776757717133,
      "num_tokens": 92122164.0,
      "step": 7468
    },
    {
      "entropy": 1.304491639137268,
      "epoch": 3.9331226961558716,
      "grad_norm": 0.29766130447387695,
      "learning_rate": 9.641275470251145e-05,
      "loss": 0.18060892820358276,
      "mean_token_accuracy": 0.9242375940084457,
      "num_tokens": 92134500.0,
      "step": 7469
    },
    {
      "entropy": 1.2207664251327515,
      "epoch": 3.933649289099526,
      "grad_norm": 0.2631109356880188,
      "learning_rate": 9.639107815097462e-05,
      "loss": 0.1687084436416626,
      "mean_token_accuracy": 0.928931936621666,
      "num_tokens": 92146836.0,
      "step": 7470
    },
    {
      "entropy": 1.2501631677150726,
      "epoch": 3.9341758820431805,
      "grad_norm": 0.2933122217655182,
      "learning_rate": 9.636940240732015e-05,
      "loss": 0.1884680986404419,
      "mean_token_accuracy": 0.9194168895483017,
      "num_tokens": 92159172.0,
      "step": 7471
    },
    {
      "entropy": 1.2987435460090637,
      "epoch": 3.9347024749868353,
      "grad_norm": 0.30980929732322693,
      "learning_rate": 9.63477274728348e-05,
      "loss": 0.18887969851493835,
      "mean_token_accuracy": 0.920253798365593,
      "num_tokens": 92171508.0,
      "step": 7472
    },
    {
      "entropy": 1.3175013065338135,
      "epoch": 3.9352290679304898,
      "grad_norm": 0.2880313992500305,
      "learning_rate": 9.632605334880531e-05,
      "loss": 0.16244207322597504,
      "mean_token_accuracy": 0.9357964843511581,
      "num_tokens": 92183844.0,
      "step": 7473
    },
    {
      "entropy": 1.249307543039322,
      "epoch": 3.935755660874144,
      "grad_norm": 0.3131891191005707,
      "learning_rate": 9.630438003651833e-05,
      "loss": 0.19666621088981628,
      "mean_token_accuracy": 0.9181525409221649,
      "num_tokens": 92196180.0,
      "step": 7474
    },
    {
      "entropy": 1.2992484867572784,
      "epoch": 3.936282253817799,
      "grad_norm": 0.29129382967948914,
      "learning_rate": 9.628270753726048e-05,
      "loss": 0.16414833068847656,
      "mean_token_accuracy": 0.9326082319021225,
      "num_tokens": 92208516.0,
      "step": 7475
    },
    {
      "entropy": 1.3063248097896576,
      "epoch": 3.9368088467614535,
      "grad_norm": 0.3242841064929962,
      "learning_rate": 9.626103585231834e-05,
      "loss": 0.18781878054141998,
      "mean_token_accuracy": 0.921628475189209,
      "num_tokens": 92220852.0,
      "step": 7476
    },
    {
      "entropy": 1.30518838763237,
      "epoch": 3.937335439705108,
      "grad_norm": 0.27939802408218384,
      "learning_rate": 9.623936498297841e-05,
      "loss": 0.1684660017490387,
      "mean_token_accuracy": 0.9288779944181442,
      "num_tokens": 92233188.0,
      "step": 7477
    },
    {
      "entropy": 1.2671433985233307,
      "epoch": 3.9378620326487628,
      "grad_norm": 0.28031885623931885,
      "learning_rate": 9.621769493052722e-05,
      "loss": 0.1693497598171234,
      "mean_token_accuracy": 0.9269088804721832,
      "num_tokens": 92245524.0,
      "step": 7478
    },
    {
      "entropy": 1.2910725772380829,
      "epoch": 3.938388625592417,
      "grad_norm": 0.26939520239830017,
      "learning_rate": 9.61960256962511e-05,
      "loss": 0.16636882722377777,
      "mean_token_accuracy": 0.9293422400951385,
      "num_tokens": 92257860.0,
      "step": 7479
    },
    {
      "entropy": 1.30040642619133,
      "epoch": 3.9389152185360716,
      "grad_norm": 0.2899145483970642,
      "learning_rate": 9.617435728143654e-05,
      "loss": 0.17569515109062195,
      "mean_token_accuracy": 0.9298725128173828,
      "num_tokens": 92270196.0,
      "step": 7480
    },
    {
      "entropy": 1.3440462350845337,
      "epoch": 3.939441811479726,
      "grad_norm": 0.3209908604621887,
      "learning_rate": 9.615268968736975e-05,
      "loss": 0.1894574612379074,
      "mean_token_accuracy": 0.9194609373807907,
      "num_tokens": 92282532.0,
      "step": 7481
    },
    {
      "entropy": 1.327031284570694,
      "epoch": 3.9399684044233805,
      "grad_norm": 0.3005867004394531,
      "learning_rate": 9.613102291533705e-05,
      "loss": 0.19162625074386597,
      "mean_token_accuracy": 0.9237973093986511,
      "num_tokens": 92294868.0,
      "step": 7482
    },
    {
      "entropy": 1.3306479156017303,
      "epoch": 3.9404949973670353,
      "grad_norm": 0.29519328474998474,
      "learning_rate": 9.610935696662473e-05,
      "loss": 0.19496075809001923,
      "mean_token_accuracy": 0.9158709049224854,
      "num_tokens": 92307204.0,
      "step": 7483
    },
    {
      "entropy": 1.3298758566379547,
      "epoch": 3.9410215903106898,
      "grad_norm": 0.2970565855503082,
      "learning_rate": 9.60876918425189e-05,
      "loss": 0.1941128969192505,
      "mean_token_accuracy": 0.9188930094242096,
      "num_tokens": 92319540.0,
      "step": 7484
    },
    {
      "entropy": 1.32647043466568,
      "epoch": 3.941548183254344,
      "grad_norm": 0.2894797623157501,
      "learning_rate": 9.606602754430576e-05,
      "loss": 0.18311838805675507,
      "mean_token_accuracy": 0.9203295111656189,
      "num_tokens": 92331876.0,
      "step": 7485
    },
    {
      "entropy": 1.3235417902469635,
      "epoch": 3.942074776197999,
      "grad_norm": 0.2994464933872223,
      "learning_rate": 9.60443640732713e-05,
      "loss": 0.18579910695552826,
      "mean_token_accuracy": 0.9255027323961258,
      "num_tokens": 92344212.0,
      "step": 7486
    },
    {
      "entropy": 1.2972019612789154,
      "epoch": 3.9426013691416535,
      "grad_norm": 0.28680962324142456,
      "learning_rate": 9.602270143070164e-05,
      "loss": 0.16868475079536438,
      "mean_token_accuracy": 0.9337401539087296,
      "num_tokens": 92356548.0,
      "step": 7487
    },
    {
      "entropy": 1.3408292531967163,
      "epoch": 3.943127962085308,
      "grad_norm": 0.2828841209411621,
      "learning_rate": 9.600103961788274e-05,
      "loss": 0.17516160011291504,
      "mean_token_accuracy": 0.9230733662843704,
      "num_tokens": 92368884.0,
      "step": 7488
    },
    {
      "entropy": 1.3030468225479126,
      "epoch": 3.943654555028963,
      "grad_norm": 0.3317798972129822,
      "learning_rate": 9.597937863610047e-05,
      "loss": 0.1834927648305893,
      "mean_token_accuracy": 0.9237143397331238,
      "num_tokens": 92381220.0,
      "step": 7489
    },
    {
      "entropy": 1.300564169883728,
      "epoch": 3.944181147972617,
      "grad_norm": 0.2664189338684082,
      "learning_rate": 9.595771848664082e-05,
      "loss": 0.16435495018959045,
      "mean_token_accuracy": 0.9301376342773438,
      "num_tokens": 92393556.0,
      "step": 7490
    },
    {
      "entropy": 1.3278022706508636,
      "epoch": 3.9447077409162716,
      "grad_norm": 0.32556602358818054,
      "learning_rate": 9.593605917078956e-05,
      "loss": 0.20098881423473358,
      "mean_token_accuracy": 0.9177375733852386,
      "num_tokens": 92405892.0,
      "step": 7491
    },
    {
      "entropy": 1.3283456563949585,
      "epoch": 3.9452343338599265,
      "grad_norm": 0.30426931381225586,
      "learning_rate": 9.59144006898325e-05,
      "loss": 0.18360796570777893,
      "mean_token_accuracy": 0.9221368879079819,
      "num_tokens": 92418228.0,
      "step": 7492
    },
    {
      "entropy": 1.2963765859603882,
      "epoch": 3.945760926803581,
      "grad_norm": 0.286146879196167,
      "learning_rate": 9.589274304505537e-05,
      "loss": 0.1922021359205246,
      "mean_token_accuracy": 0.9193236529827118,
      "num_tokens": 92430564.0,
      "step": 7493
    },
    {
      "entropy": 1.314402014017105,
      "epoch": 3.9462875197472354,
      "grad_norm": 0.2976517081260681,
      "learning_rate": 9.587108623774388e-05,
      "loss": 0.1778688281774521,
      "mean_token_accuracy": 0.9237999469041824,
      "num_tokens": 92442900.0,
      "step": 7494
    },
    {
      "entropy": 1.2591935694217682,
      "epoch": 3.94681411269089,
      "grad_norm": 0.278967946767807,
      "learning_rate": 9.584943026918365e-05,
      "loss": 0.1706763654947281,
      "mean_token_accuracy": 0.9272800981998444,
      "num_tokens": 92455236.0,
      "step": 7495
    },
    {
      "entropy": 1.250948041677475,
      "epoch": 3.9473407056345446,
      "grad_norm": 0.2639889121055603,
      "learning_rate": 9.582777514066024e-05,
      "loss": 0.16510087251663208,
      "mean_token_accuracy": 0.9280741214752197,
      "num_tokens": 92467572.0,
      "step": 7496
    },
    {
      "entropy": 1.235801875591278,
      "epoch": 3.947867298578199,
      "grad_norm": 0.27677178382873535,
      "learning_rate": 9.580612085345924e-05,
      "loss": 0.18638868629932404,
      "mean_token_accuracy": 0.9201086163520813,
      "num_tokens": 92479908.0,
      "step": 7497
    },
    {
      "entropy": 1.2722296118736267,
      "epoch": 3.9483938915218535,
      "grad_norm": 0.284332811832428,
      "learning_rate": 9.57844674088661e-05,
      "loss": 0.17478200793266296,
      "mean_token_accuracy": 0.9271548986434937,
      "num_tokens": 92492244.0,
      "step": 7498
    },
    {
      "entropy": 1.340325802564621,
      "epoch": 3.948920484465508,
      "grad_norm": 0.2919856011867523,
      "learning_rate": 9.576281480816631e-05,
      "loss": 0.18050037324428558,
      "mean_token_accuracy": 0.9251342415809631,
      "num_tokens": 92504580.0,
      "step": 7499
    },
    {
      "entropy": 1.3454786837100983,
      "epoch": 3.949447077409163,
      "grad_norm": 0.29903966188430786,
      "learning_rate": 9.57411630526452e-05,
      "loss": 0.1800300031900406,
      "mean_token_accuracy": 0.926009476184845,
      "num_tokens": 92516916.0,
      "step": 7500
    },
    {
      "entropy": 1.3552602529525757,
      "epoch": 3.949973670352817,
      "grad_norm": 0.3141869604587555,
      "learning_rate": 9.571951214358814e-05,
      "loss": 0.18520255386829376,
      "mean_token_accuracy": 0.9232632219791412,
      "num_tokens": 92529252.0,
      "step": 7501
    },
    {
      "entropy": 1.34744131565094,
      "epoch": 3.9505002632964716,
      "grad_norm": 0.2958497107028961,
      "learning_rate": 9.569786208228041e-05,
      "loss": 0.1796453446149826,
      "mean_token_accuracy": 0.9225263148546219,
      "num_tokens": 92541588.0,
      "step": 7502
    },
    {
      "entropy": 1.2951800525188446,
      "epoch": 3.9510268562401265,
      "grad_norm": 0.31804752349853516,
      "learning_rate": 9.567621287000724e-05,
      "loss": 0.19041743874549866,
      "mean_token_accuracy": 0.9186290353536606,
      "num_tokens": 92553924.0,
      "step": 7503
    },
    {
      "entropy": 1.2884739637374878,
      "epoch": 3.951553449183781,
      "grad_norm": 0.2896903455257416,
      "learning_rate": 9.565456450805382e-05,
      "loss": 0.17561444640159607,
      "mean_token_accuracy": 0.9240155518054962,
      "num_tokens": 92566260.0,
      "step": 7504
    },
    {
      "entropy": 1.2947382926940918,
      "epoch": 3.9520800421274354,
      "grad_norm": 0.2867039442062378,
      "learning_rate": 9.563291699770532e-05,
      "loss": 0.1776714026927948,
      "mean_token_accuracy": 0.9248031079769135,
      "num_tokens": 92578596.0,
      "step": 7505
    },
    {
      "entropy": 1.3121528923511505,
      "epoch": 3.9526066350710902,
      "grad_norm": 0.30147653818130493,
      "learning_rate": 9.561127034024677e-05,
      "loss": 0.17624038457870483,
      "mean_token_accuracy": 0.9245792776346207,
      "num_tokens": 92590932.0,
      "step": 7506
    },
    {
      "entropy": 1.3536790311336517,
      "epoch": 3.9531332280147446,
      "grad_norm": 0.27647480368614197,
      "learning_rate": 9.558962453696325e-05,
      "loss": 0.16962884366512299,
      "mean_token_accuracy": 0.9276664108037949,
      "num_tokens": 92603268.0,
      "step": 7507
    },
    {
      "entropy": 1.3156141638755798,
      "epoch": 3.953659820958399,
      "grad_norm": 0.28009626269340515,
      "learning_rate": 9.556797958913972e-05,
      "loss": 0.16551202535629272,
      "mean_token_accuracy": 0.9273913949728012,
      "num_tokens": 92615604.0,
      "step": 7508
    },
    {
      "entropy": 1.347121149301529,
      "epoch": 3.954186413902054,
      "grad_norm": 0.31527823209762573,
      "learning_rate": 9.554633549806116e-05,
      "loss": 0.16975153982639313,
      "mean_token_accuracy": 0.9277451187372208,
      "num_tokens": 92627940.0,
      "step": 7509
    },
    {
      "entropy": 1.3640653491020203,
      "epoch": 3.9547130068457084,
      "grad_norm": 0.3153354227542877,
      "learning_rate": 9.552469226501237e-05,
      "loss": 0.18295100331306458,
      "mean_token_accuracy": 0.9233006238937378,
      "num_tokens": 92640276.0,
      "step": 7510
    },
    {
      "entropy": 1.3119199573993683,
      "epoch": 3.955239599789363,
      "grad_norm": 0.2903278172016144,
      "learning_rate": 9.550304989127828e-05,
      "loss": 0.1801615059375763,
      "mean_token_accuracy": 0.9202624261379242,
      "num_tokens": 92652612.0,
      "step": 7511
    },
    {
      "entropy": 1.2877613604068756,
      "epoch": 3.955766192733017,
      "grad_norm": 0.2637787163257599,
      "learning_rate": 9.548140837814359e-05,
      "loss": 0.1738012731075287,
      "mean_token_accuracy": 0.9286294877529144,
      "num_tokens": 92664948.0,
      "step": 7512
    },
    {
      "entropy": 1.318818360567093,
      "epoch": 3.956292785676672,
      "grad_norm": 0.29511192440986633,
      "learning_rate": 9.545976772689307e-05,
      "loss": 0.18746647238731384,
      "mean_token_accuracy": 0.9240599721670151,
      "num_tokens": 92677284.0,
      "step": 7513
    },
    {
      "entropy": 1.312627375125885,
      "epoch": 3.9568193786203265,
      "grad_norm": 0.2809144854545593,
      "learning_rate": 9.543812793881142e-05,
      "loss": 0.1732306331396103,
      "mean_token_accuracy": 0.9288646876811981,
      "num_tokens": 92689620.0,
      "step": 7514
    },
    {
      "entropy": 1.3550857603549957,
      "epoch": 3.957345971563981,
      "grad_norm": 0.30199602246284485,
      "learning_rate": 9.541648901518323e-05,
      "loss": 0.18501755595207214,
      "mean_token_accuracy": 0.9217173457145691,
      "num_tokens": 92701956.0,
      "step": 7515
    },
    {
      "entropy": 1.3081051707267761,
      "epoch": 3.9578725645076354,
      "grad_norm": 0.2714572846889496,
      "learning_rate": 9.539485095729308e-05,
      "loss": 0.17930418252944946,
      "mean_token_accuracy": 0.9250104427337646,
      "num_tokens": 92714292.0,
      "step": 7516
    },
    {
      "entropy": 1.3340463936328888,
      "epoch": 3.9583991574512902,
      "grad_norm": 0.2664049565792084,
      "learning_rate": 9.537321376642546e-05,
      "loss": 0.17378565669059753,
      "mean_token_accuracy": 0.9263237565755844,
      "num_tokens": 92726628.0,
      "step": 7517
    },
    {
      "entropy": 1.321879804134369,
      "epoch": 3.9589257503949447,
      "grad_norm": 0.2919703722000122,
      "learning_rate": 9.535157744386494e-05,
      "loss": 0.1797531098127365,
      "mean_token_accuracy": 0.9261958748102188,
      "num_tokens": 92738964.0,
      "step": 7518
    },
    {
      "entropy": 1.3287270367145538,
      "epoch": 3.959452343338599,
      "grad_norm": 0.2807745039463043,
      "learning_rate": 9.532994199089588e-05,
      "loss": 0.1720634400844574,
      "mean_token_accuracy": 0.9296756684780121,
      "num_tokens": 92751300.0,
      "step": 7519
    },
    {
      "entropy": 1.3139150440692902,
      "epoch": 3.959978936282254,
      "grad_norm": 0.2645365595817566,
      "learning_rate": 9.530830740880268e-05,
      "loss": 0.16413986682891846,
      "mean_token_accuracy": 0.9297731667757034,
      "num_tokens": 92763636.0,
      "step": 7520
    },
    {
      "entropy": 1.2684156894683838,
      "epoch": 3.9605055292259084,
      "grad_norm": 0.28925225138664246,
      "learning_rate": 9.528667369886961e-05,
      "loss": 0.18101872503757477,
      "mean_token_accuracy": 0.9200309365987778,
      "num_tokens": 92775972.0,
      "step": 7521
    },
    {
      "entropy": 1.2909513413906097,
      "epoch": 3.961032122169563,
      "grad_norm": 0.2839042842388153,
      "learning_rate": 9.526504086238097e-05,
      "loss": 0.1760377734899521,
      "mean_token_accuracy": 0.9249915182590485,
      "num_tokens": 92788308.0,
      "step": 7522
    },
    {
      "entropy": 1.356577455997467,
      "epoch": 3.9615587151132177,
      "grad_norm": 0.3162682354450226,
      "learning_rate": 9.524340890062097e-05,
      "loss": 0.19605055451393127,
      "mean_token_accuracy": 0.9202203303575516,
      "num_tokens": 92800644.0,
      "step": 7523
    },
    {
      "entropy": 1.3656745851039886,
      "epoch": 3.962085308056872,
      "grad_norm": 0.31002482771873474,
      "learning_rate": 9.522177781487378e-05,
      "loss": 0.18034009635448456,
      "mean_token_accuracy": 0.9234649538993835,
      "num_tokens": 92812980.0,
      "step": 7524
    },
    {
      "entropy": 1.295270323753357,
      "epoch": 3.9626119010005265,
      "grad_norm": 0.2766196131706238,
      "learning_rate": 9.52001476064235e-05,
      "loss": 0.18228909373283386,
      "mean_token_accuracy": 0.9220178127288818,
      "num_tokens": 92825316.0,
      "step": 7525
    },
    {
      "entropy": 1.348963350057602,
      "epoch": 3.9631384939441814,
      "grad_norm": 0.3193948566913605,
      "learning_rate": 9.517851827655422e-05,
      "loss": 0.20217934250831604,
      "mean_token_accuracy": 0.9191700518131256,
      "num_tokens": 92837652.0,
      "step": 7526
    },
    {
      "entropy": 1.3162882328033447,
      "epoch": 3.963665086887836,
      "grad_norm": 0.27619749307632446,
      "learning_rate": 9.515688982654991e-05,
      "loss": 0.1549035906791687,
      "mean_token_accuracy": 0.9353099465370178,
      "num_tokens": 92849988.0,
      "step": 7527
    },
    {
      "entropy": 1.3428030014038086,
      "epoch": 3.9641916798314902,
      "grad_norm": 0.304727703332901,
      "learning_rate": 9.513526225769454e-05,
      "loss": 0.18223881721496582,
      "mean_token_accuracy": 0.9228622019290924,
      "num_tokens": 92862324.0,
      "step": 7528
    },
    {
      "entropy": 1.30735844373703,
      "epoch": 3.9647182727751447,
      "grad_norm": 0.2727219760417938,
      "learning_rate": 9.511363557127204e-05,
      "loss": 0.16652140021324158,
      "mean_token_accuracy": 0.9266299903392792,
      "num_tokens": 92874660.0,
      "step": 7529
    },
    {
      "entropy": 1.2992176413536072,
      "epoch": 3.9652448657187995,
      "grad_norm": 0.27040427923202515,
      "learning_rate": 9.509200976856621e-05,
      "loss": 0.1641957014799118,
      "mean_token_accuracy": 0.9336581975221634,
      "num_tokens": 92886996.0,
      "step": 7530
    },
    {
      "entropy": 1.3509510457515717,
      "epoch": 3.965771458662454,
      "grad_norm": 0.3059140741825104,
      "learning_rate": 9.507038485086081e-05,
      "loss": 0.18820726871490479,
      "mean_token_accuracy": 0.9169155806303024,
      "num_tokens": 92899332.0,
      "step": 7531
    },
    {
      "entropy": 1.3464178442955017,
      "epoch": 3.9662980516061084,
      "grad_norm": 0.29788780212402344,
      "learning_rate": 9.504876081943969e-05,
      "loss": 0.18500803411006927,
      "mean_token_accuracy": 0.9244650900363922,
      "num_tokens": 92911668.0,
      "step": 7532
    },
    {
      "entropy": 1.2732579112052917,
      "epoch": 3.966824644549763,
      "grad_norm": 0.2813820242881775,
      "learning_rate": 9.502713767558653e-05,
      "loss": 0.1802762895822525,
      "mean_token_accuracy": 0.9253692179918289,
      "num_tokens": 92924004.0,
      "step": 7533
    },
    {
      "entropy": 1.3419006168842316,
      "epoch": 3.9673512374934177,
      "grad_norm": 0.312103271484375,
      "learning_rate": 9.500551542058492e-05,
      "loss": 0.17921848595142365,
      "mean_token_accuracy": 0.9268561005592346,
      "num_tokens": 92936340.0,
      "step": 7534
    },
    {
      "entropy": 1.2903282344341278,
      "epoch": 3.967877830437072,
      "grad_norm": 0.29019781947135925,
      "learning_rate": 9.498389405571846e-05,
      "loss": 0.18339002132415771,
      "mean_token_accuracy": 0.9240312725305557,
      "num_tokens": 92948676.0,
      "step": 7535
    },
    {
      "entropy": 1.3509527444839478,
      "epoch": 3.9684044233807265,
      "grad_norm": 0.2800314128398895,
      "learning_rate": 9.49622735822707e-05,
      "loss": 0.1678372621536255,
      "mean_token_accuracy": 0.9301303774118423,
      "num_tokens": 92961012.0,
      "step": 7536
    },
    {
      "entropy": 1.299644261598587,
      "epoch": 3.9689310163243814,
      "grad_norm": 0.28727006912231445,
      "learning_rate": 9.494065400152508e-05,
      "loss": 0.17354363203048706,
      "mean_token_accuracy": 0.9214708656072617,
      "num_tokens": 92973348.0,
      "step": 7537
    },
    {
      "entropy": 1.331787645816803,
      "epoch": 3.969457609268036,
      "grad_norm": 0.29024335741996765,
      "learning_rate": 9.491903531476509e-05,
      "loss": 0.18393193185329437,
      "mean_token_accuracy": 0.920793890953064,
      "num_tokens": 92985684.0,
      "step": 7538
    },
    {
      "entropy": 1.311621993780136,
      "epoch": 3.9699842022116902,
      "grad_norm": 0.28642910718917847,
      "learning_rate": 9.489741752327403e-05,
      "loss": 0.1857396811246872,
      "mean_token_accuracy": 0.9191377907991409,
      "num_tokens": 92998020.0,
      "step": 7539
    },
    {
      "entropy": 1.390136182308197,
      "epoch": 3.970510795155345,
      "grad_norm": 0.32270437479019165,
      "learning_rate": 9.487580062833532e-05,
      "loss": 0.19870172441005707,
      "mean_token_accuracy": 0.9207863509654999,
      "num_tokens": 93010356.0,
      "step": 7540
    },
    {
      "entropy": 1.3202392160892487,
      "epoch": 3.9710373880989995,
      "grad_norm": 0.3119673430919647,
      "learning_rate": 9.485418463123215e-05,
      "loss": 0.18765456974506378,
      "mean_token_accuracy": 0.9160043448209763,
      "num_tokens": 93022692.0,
      "step": 7541
    },
    {
      "entropy": 1.3205336928367615,
      "epoch": 3.971563981042654,
      "grad_norm": 0.31557634472846985,
      "learning_rate": 9.483256953324777e-05,
      "loss": 0.19298219680786133,
      "mean_token_accuracy": 0.9155784547328949,
      "num_tokens": 93035028.0,
      "step": 7542
    },
    {
      "entropy": 1.3472093343734741,
      "epoch": 3.972090573986309,
      "grad_norm": 0.29996630549430847,
      "learning_rate": 9.481095533566533e-05,
      "loss": 0.18451613187789917,
      "mean_token_accuracy": 0.92573282122612,
      "num_tokens": 93047364.0,
      "step": 7543
    },
    {
      "entropy": 1.3062803745269775,
      "epoch": 3.9726171669299633,
      "grad_norm": 0.2833060622215271,
      "learning_rate": 9.478934203976797e-05,
      "loss": 0.1899745911359787,
      "mean_token_accuracy": 0.9236193895339966,
      "num_tokens": 93059700.0,
      "step": 7544
    },
    {
      "entropy": 1.3540114164352417,
      "epoch": 3.9731437598736177,
      "grad_norm": 0.2676866054534912,
      "learning_rate": 9.476772964683869e-05,
      "loss": 0.17042262852191925,
      "mean_token_accuracy": 0.9254991859197617,
      "num_tokens": 93072036.0,
      "step": 7545
    },
    {
      "entropy": 1.3275431990623474,
      "epoch": 3.973670352817272,
      "grad_norm": 0.26242774724960327,
      "learning_rate": 9.474611815816048e-05,
      "loss": 0.17449872195720673,
      "mean_token_accuracy": 0.9249527454376221,
      "num_tokens": 93084372.0,
      "step": 7546
    },
    {
      "entropy": 1.3328191637992859,
      "epoch": 3.974196945760927,
      "grad_norm": 0.2825845181941986,
      "learning_rate": 9.472450757501638e-05,
      "loss": 0.17568956315517426,
      "mean_token_accuracy": 0.9227619171142578,
      "num_tokens": 93096708.0,
      "step": 7547
    },
    {
      "entropy": 1.3070374727249146,
      "epoch": 3.9747235387045814,
      "grad_norm": 0.2826155424118042,
      "learning_rate": 9.470289789868924e-05,
      "loss": 0.17785626649856567,
      "mean_token_accuracy": 0.9231225848197937,
      "num_tokens": 93109044.0,
      "step": 7548
    },
    {
      "entropy": 1.3746168613433838,
      "epoch": 3.975250131648236,
      "grad_norm": 0.3055529296398163,
      "learning_rate": 9.468128913046188e-05,
      "loss": 0.1931118220090866,
      "mean_token_accuracy": 0.9192795604467392,
      "num_tokens": 93121380.0,
      "step": 7549
    },
    {
      "entropy": 1.344740629196167,
      "epoch": 3.9757767245918902,
      "grad_norm": 0.2886352241039276,
      "learning_rate": 9.465968127161711e-05,
      "loss": 0.18964360654354095,
      "mean_token_accuracy": 0.9224054515361786,
      "num_tokens": 93133716.0,
      "step": 7550
    },
    {
      "entropy": 1.3456671237945557,
      "epoch": 3.976303317535545,
      "grad_norm": 0.28195491433143616,
      "learning_rate": 9.463807432343763e-05,
      "loss": 0.17482024431228638,
      "mean_token_accuracy": 0.9267143756151199,
      "num_tokens": 93146052.0,
      "step": 7551
    },
    {
      "entropy": 1.3401457071304321,
      "epoch": 3.9768299104791995,
      "grad_norm": 0.27699512243270874,
      "learning_rate": 9.461646828720616e-05,
      "loss": 0.18693599104881287,
      "mean_token_accuracy": 0.9195725619792938,
      "num_tokens": 93158388.0,
      "step": 7552
    },
    {
      "entropy": 1.3148573637008667,
      "epoch": 3.977356503422854,
      "grad_norm": 0.2510269284248352,
      "learning_rate": 9.459486316420527e-05,
      "loss": 0.16350793838500977,
      "mean_token_accuracy": 0.9319835603237152,
      "num_tokens": 93170724.0,
      "step": 7553
    },
    {
      "entropy": 1.3479683995246887,
      "epoch": 3.977883096366509,
      "grad_norm": 0.2932959198951721,
      "learning_rate": 9.45732589557176e-05,
      "loss": 0.17321151494979858,
      "mean_token_accuracy": 0.9280009567737579,
      "num_tokens": 93183060.0,
      "step": 7554
    },
    {
      "entropy": 1.3800325095653534,
      "epoch": 3.9784096893101633,
      "grad_norm": 0.31480640172958374,
      "learning_rate": 9.455165566302565e-05,
      "loss": 0.1839628666639328,
      "mean_token_accuracy": 0.9236285239458084,
      "num_tokens": 93195396.0,
      "step": 7555
    },
    {
      "entropy": 1.3246859014034271,
      "epoch": 3.9789362822538177,
      "grad_norm": 0.2820781469345093,
      "learning_rate": 9.453005328741183e-05,
      "loss": 0.1780439019203186,
      "mean_token_accuracy": 0.9251691251993179,
      "num_tokens": 93207732.0,
      "step": 7556
    },
    {
      "entropy": 1.3023313879966736,
      "epoch": 3.9794628751974725,
      "grad_norm": 0.29186537861824036,
      "learning_rate": 9.450845183015857e-05,
      "loss": 0.18912522494792938,
      "mean_token_accuracy": 0.9216838777065277,
      "num_tokens": 93220068.0,
      "step": 7557
    },
    {
      "entropy": 1.3183483183383942,
      "epoch": 3.979989468141127,
      "grad_norm": 0.2793582081794739,
      "learning_rate": 9.448685129254828e-05,
      "loss": 0.1669609248638153,
      "mean_token_accuracy": 0.9317817240953445,
      "num_tokens": 93232404.0,
      "step": 7558
    },
    {
      "entropy": 1.3475527167320251,
      "epoch": 3.9805160610847814,
      "grad_norm": 0.27668139338493347,
      "learning_rate": 9.44652516758632e-05,
      "loss": 0.16608241200447083,
      "mean_token_accuracy": 0.9317691028118134,
      "num_tokens": 93244740.0,
      "step": 7559
    },
    {
      "entropy": 1.3177028894424438,
      "epoch": 3.9810426540284363,
      "grad_norm": 0.3113577663898468,
      "learning_rate": 9.444365298138553e-05,
      "loss": 0.18435797095298767,
      "mean_token_accuracy": 0.9207548052072525,
      "num_tokens": 93257076.0,
      "step": 7560
    },
    {
      "entropy": 1.2700896859169006,
      "epoch": 3.9815692469720907,
      "grad_norm": 0.2768622934818268,
      "learning_rate": 9.442205521039756e-05,
      "loss": 0.16947168111801147,
      "mean_token_accuracy": 0.92853944003582,
      "num_tokens": 93269412.0,
      "step": 7561
    },
    {
      "entropy": 1.2747684717178345,
      "epoch": 3.982095839915745,
      "grad_norm": 0.2852453291416168,
      "learning_rate": 9.440045836418137e-05,
      "loss": 0.16986969113349915,
      "mean_token_accuracy": 0.9299718290567398,
      "num_tokens": 93281748.0,
      "step": 7562
    },
    {
      "entropy": 1.319030523300171,
      "epoch": 3.9826224328593995,
      "grad_norm": 0.32257211208343506,
      "learning_rate": 9.437886244401906e-05,
      "loss": 0.18968331813812256,
      "mean_token_accuracy": 0.9188378304243088,
      "num_tokens": 93294084.0,
      "step": 7563
    },
    {
      "entropy": 1.3039062321186066,
      "epoch": 3.983149025803054,
      "grad_norm": 0.30198219418525696,
      "learning_rate": 9.435726745119264e-05,
      "loss": 0.1780025064945221,
      "mean_token_accuracy": 0.924556702375412,
      "num_tokens": 93306420.0,
      "step": 7564
    },
    {
      "entropy": 1.3020071685314178,
      "epoch": 3.983675618746709,
      "grad_norm": 0.3142753839492798,
      "learning_rate": 9.433567338698406e-05,
      "loss": 0.17521704733371735,
      "mean_token_accuracy": 0.9252939373254776,
      "num_tokens": 93318756.0,
      "step": 7565
    },
    {
      "entropy": 1.2601705193519592,
      "epoch": 3.9842022116903633,
      "grad_norm": 0.2831016480922699,
      "learning_rate": 9.431408025267525e-05,
      "loss": 0.19090916216373444,
      "mean_token_accuracy": 0.9184614717960358,
      "num_tokens": 93331092.0,
      "step": 7566
    },
    {
      "entropy": 1.2693793177604675,
      "epoch": 3.9847288046340177,
      "grad_norm": 0.2954610288143158,
      "learning_rate": 9.429248804954807e-05,
      "loss": 0.18366365134716034,
      "mean_token_accuracy": 0.9261559098958969,
      "num_tokens": 93343428.0,
      "step": 7567
    },
    {
      "entropy": 1.2716824412345886,
      "epoch": 3.9852553975776726,
      "grad_norm": 0.3004774749279022,
      "learning_rate": 9.427089677888432e-05,
      "loss": 0.18958234786987305,
      "mean_token_accuracy": 0.9214614927768707,
      "num_tokens": 93355764.0,
      "step": 7568
    },
    {
      "entropy": 1.2541852295398712,
      "epoch": 3.985781990521327,
      "grad_norm": 0.30183055996894836,
      "learning_rate": 9.42493064419658e-05,
      "loss": 0.17600035667419434,
      "mean_token_accuracy": 0.9274772852659225,
      "num_tokens": 93368100.0,
      "step": 7569
    },
    {
      "entropy": 1.237665981054306,
      "epoch": 3.9863085834649814,
      "grad_norm": 0.28173959255218506,
      "learning_rate": 9.422771704007409e-05,
      "loss": 0.1743810623884201,
      "mean_token_accuracy": 0.9307523667812347,
      "num_tokens": 93380436.0,
      "step": 7570
    },
    {
      "entropy": 1.2573004364967346,
      "epoch": 3.9868351764086363,
      "grad_norm": 0.2884520888328552,
      "learning_rate": 9.420612857449094e-05,
      "loss": 0.1682325303554535,
      "mean_token_accuracy": 0.9293691515922546,
      "num_tokens": 93392772.0,
      "step": 7571
    },
    {
      "entropy": 1.290991097688675,
      "epoch": 3.9873617693522907,
      "grad_norm": 0.29057928919792175,
      "learning_rate": 9.418454104649785e-05,
      "loss": 0.17541126906871796,
      "mean_token_accuracy": 0.9285570681095123,
      "num_tokens": 93405108.0,
      "step": 7572
    },
    {
      "entropy": 1.293534904718399,
      "epoch": 3.987888362295945,
      "grad_norm": 0.30405089259147644,
      "learning_rate": 9.41629544573764e-05,
      "loss": 0.17564499378204346,
      "mean_token_accuracy": 0.9292964190244675,
      "num_tokens": 93417444.0,
      "step": 7573
    },
    {
      "entropy": 1.3413696885108948,
      "epoch": 3.9884149552396,
      "grad_norm": 0.3326863944530487,
      "learning_rate": 9.4141368808408e-05,
      "loss": 0.1866447627544403,
      "mean_token_accuracy": 0.9208780378103256,
      "num_tokens": 93429780.0,
      "step": 7574
    },
    {
      "entropy": 1.3060389757156372,
      "epoch": 3.9889415481832544,
      "grad_norm": 0.27616894245147705,
      "learning_rate": 9.411978410087414e-05,
      "loss": 0.150806725025177,
      "mean_token_accuracy": 0.9384946525096893,
      "num_tokens": 93442116.0,
      "step": 7575
    },
    {
      "entropy": 1.2838166058063507,
      "epoch": 3.989468141126909,
      "grad_norm": 0.2959181070327759,
      "learning_rate": 9.409820033605614e-05,
      "loss": 0.18270926177501678,
      "mean_token_accuracy": 0.9219162911176682,
      "num_tokens": 93454452.0,
      "step": 7576
    },
    {
      "entropy": 1.3128345012664795,
      "epoch": 3.9899947340705637,
      "grad_norm": 0.28953227400779724,
      "learning_rate": 9.407661751523529e-05,
      "loss": 0.16359031200408936,
      "mean_token_accuracy": 0.9269338846206665,
      "num_tokens": 93466788.0,
      "step": 7577
    },
    {
      "entropy": 1.3239604234695435,
      "epoch": 3.990521327014218,
      "grad_norm": 0.2909933030605316,
      "learning_rate": 9.405503563969289e-05,
      "loss": 0.1780087947845459,
      "mean_token_accuracy": 0.9250047355890274,
      "num_tokens": 93479124.0,
      "step": 7578
    },
    {
      "entropy": 1.2926201224327087,
      "epoch": 3.9910479199578726,
      "grad_norm": 0.27546995878219604,
      "learning_rate": 9.403345471071006e-05,
      "loss": 0.1740032583475113,
      "mean_token_accuracy": 0.9257869869470596,
      "num_tokens": 93491460.0,
      "step": 7579
    },
    {
      "entropy": 1.3467094898223877,
      "epoch": 3.991574512901527,
      "grad_norm": 0.2991516590118408,
      "learning_rate": 9.401187472956796e-05,
      "loss": 0.19406607747077942,
      "mean_token_accuracy": 0.9218985587358475,
      "num_tokens": 93503796.0,
      "step": 7580
    },
    {
      "entropy": 1.2814543843269348,
      "epoch": 3.9921011058451814,
      "grad_norm": 0.28001993894577026,
      "learning_rate": 9.399029569754762e-05,
      "loss": 0.1770983785390854,
      "mean_token_accuracy": 0.9243462979793549,
      "num_tokens": 93516132.0,
      "step": 7581
    },
    {
      "entropy": 1.3417298197746277,
      "epoch": 3.9926276987888363,
      "grad_norm": 0.309066504240036,
      "learning_rate": 9.39687176159302e-05,
      "loss": 0.19472132623195648,
      "mean_token_accuracy": 0.9183698892593384,
      "num_tokens": 93528468.0,
      "step": 7582
    },
    {
      "entropy": 1.2971207201480865,
      "epoch": 3.9931542917324907,
      "grad_norm": 0.2736274003982544,
      "learning_rate": 9.394714048599653e-05,
      "loss": 0.1718100607395172,
      "mean_token_accuracy": 0.9273222237825394,
      "num_tokens": 93540804.0,
      "step": 7583
    },
    {
      "entropy": 1.3284357488155365,
      "epoch": 3.993680884676145,
      "grad_norm": 0.30235692858695984,
      "learning_rate": 9.392556430902759e-05,
      "loss": 0.18034395575523376,
      "mean_token_accuracy": 0.9234701097011566,
      "num_tokens": 93553140.0,
      "step": 7584
    },
    {
      "entropy": 1.2959426939487457,
      "epoch": 3.9942074776198,
      "grad_norm": 0.3051435649394989,
      "learning_rate": 9.39039890863042e-05,
      "loss": 0.19450972974300385,
      "mean_token_accuracy": 0.9161807149648666,
      "num_tokens": 93565476.0,
      "step": 7585
    },
    {
      "entropy": 1.3091471493244171,
      "epoch": 3.9947340705634544,
      "grad_norm": 0.27894070744514465,
      "learning_rate": 9.388241481910716e-05,
      "loss": 0.1739863008260727,
      "mean_token_accuracy": 0.9266135543584824,
      "num_tokens": 93577812.0,
      "step": 7586
    },
    {
      "entropy": 1.3096213042736053,
      "epoch": 3.995260663507109,
      "grad_norm": 0.30204686522483826,
      "learning_rate": 9.386084150871721e-05,
      "loss": 0.1900097280740738,
      "mean_token_accuracy": 0.9203787744045258,
      "num_tokens": 93590148.0,
      "step": 7587
    },
    {
      "entropy": 1.303184688091278,
      "epoch": 3.9957872564507637,
      "grad_norm": 0.28003254532814026,
      "learning_rate": 9.3839269156415e-05,
      "loss": 0.16158366203308105,
      "mean_token_accuracy": 0.9301228523254395,
      "num_tokens": 93602484.0,
      "step": 7588
    },
    {
      "entropy": 1.3454047739505768,
      "epoch": 3.996313849394418,
      "grad_norm": 0.31151407957077026,
      "learning_rate": 9.381769776348124e-05,
      "loss": 0.2038501352071762,
      "mean_token_accuracy": 0.9134414196014404,
      "num_tokens": 93614820.0,
      "step": 7589
    },
    {
      "entropy": 1.3581558465957642,
      "epoch": 3.9968404423380726,
      "grad_norm": 0.26742124557495117,
      "learning_rate": 9.379612733119643e-05,
      "loss": 0.1595117598772049,
      "mean_token_accuracy": 0.9353603273630142,
      "num_tokens": 93627156.0,
      "step": 7590
    },
    {
      "entropy": 1.305214762687683,
      "epoch": 3.9973670352817274,
      "grad_norm": 0.2921280562877655,
      "learning_rate": 9.37745578608411e-05,
      "loss": 0.18154413998126984,
      "mean_token_accuracy": 0.9237124472856522,
      "num_tokens": 93639492.0,
      "step": 7591
    },
    {
      "entropy": 1.304631233215332,
      "epoch": 3.997893628225382,
      "grad_norm": 0.2871142029762268,
      "learning_rate": 9.37529893536957e-05,
      "loss": 0.1777011752128601,
      "mean_token_accuracy": 0.9233687371015549,
      "num_tokens": 93651828.0,
      "step": 7592
    },
    {
      "entropy": 1.3112380504608154,
      "epoch": 3.9984202211690363,
      "grad_norm": 0.2854442000389099,
      "learning_rate": 9.373142181104065e-05,
      "loss": 0.18337658047676086,
      "mean_token_accuracy": 0.9219502508640289,
      "num_tokens": 93664164.0,
      "step": 7593
    },
    {
      "entropy": 1.2995306849479675,
      "epoch": 3.998946814112691,
      "grad_norm": 0.28682613372802734,
      "learning_rate": 9.370985523415623e-05,
      "loss": 0.18945598602294922,
      "mean_token_accuracy": 0.9239040315151215,
      "num_tokens": 93676500.0,
      "step": 7594
    },
    {
      "entropy": 1.3250145316123962,
      "epoch": 3.9994734070563456,
      "grad_norm": 0.3174082934856415,
      "learning_rate": 9.368828962432275e-05,
      "loss": 0.18585076928138733,
      "mean_token_accuracy": 0.9195867776870728,
      "num_tokens": 93688836.0,
      "step": 7595
    },
    {
      "entropy": 1.3290630280971527,
      "epoch": 4.0,
      "grad_norm": 0.29731038212776184,
      "learning_rate": 9.366672498282045e-05,
      "loss": 0.1742815226316452,
      "mean_token_accuracy": 0.9284393042325974,
      "num_tokens": 93700144.0,
      "step": 7596
    },
    {
      "epoch": 4.0,
      "eval_entropy": 1.3225353589617874,
      "eval_loss": 0.19561795890331268,
      "eval_mean_token_accuracy": 0.9197511109256192,
      "eval_num_tokens": 93700144.0,
      "eval_runtime": 665.4902,
      "eval_samples_per_second": 8.561,
      "eval_steps_per_second": 2.854,
      "step": 7596
    },
    {
      "entropy": 1.3247161507606506,
      "epoch": 4.000526592943655,
      "grad_norm": 0.28038930892944336,
      "learning_rate": 9.364516131092953e-05,
      "loss": 0.17094804346561432,
      "mean_token_accuracy": 0.933413103222847,
      "num_tokens": 93712480.0,
      "step": 7597
    },
    {
      "entropy": 1.3140203058719635,
      "epoch": 4.001053185887309,
      "grad_norm": 0.2799552381038666,
      "learning_rate": 9.362359860993004e-05,
      "loss": 0.1631288230419159,
      "mean_token_accuracy": 0.9272639006376266,
      "num_tokens": 93724816.0,
      "step": 7598
    },
    {
      "entropy": 1.315209150314331,
      "epoch": 4.001579778830964,
      "grad_norm": 0.2712099850177765,
      "learning_rate": 9.360203688110203e-05,
      "loss": 0.16098222136497498,
      "mean_token_accuracy": 0.9326698184013367,
      "num_tokens": 93737152.0,
      "step": 7599
    },
    {
      "entropy": 1.2736327052116394,
      "epoch": 4.002106371774619,
      "grad_norm": 0.2905905842781067,
      "learning_rate": 9.358047612572554e-05,
      "loss": 0.17343084514141083,
      "mean_token_accuracy": 0.9266268759965897,
      "num_tokens": 93749488.0,
      "step": 7600
    },
    {
      "entropy": 1.2937598526477814,
      "epoch": 4.002632964718273,
      "grad_norm": 0.2726406753063202,
      "learning_rate": 9.355891634508048e-05,
      "loss": 0.15777438879013062,
      "mean_token_accuracy": 0.9339054971933365,
      "num_tokens": 93761824.0,
      "step": 7601
    },
    {
      "entropy": 1.2746125161647797,
      "epoch": 4.003159557661927,
      "grad_norm": 0.2830272912979126,
      "learning_rate": 9.353735754044671e-05,
      "loss": 0.16077198088169098,
      "mean_token_accuracy": 0.928732693195343,
      "num_tokens": 93774160.0,
      "step": 7602
    },
    {
      "entropy": 1.2600156366825104,
      "epoch": 4.003686150605582,
      "grad_norm": 0.29081904888153076,
      "learning_rate": 9.351579971310406e-05,
      "loss": 0.1667558252811432,
      "mean_token_accuracy": 0.9284805953502655,
      "num_tokens": 93786496.0,
      "step": 7603
    },
    {
      "entropy": 1.2500266432762146,
      "epoch": 4.004212743549236,
      "grad_norm": 0.28229841589927673,
      "learning_rate": 9.349424286433235e-05,
      "loss": 0.1627058982849121,
      "mean_token_accuracy": 0.9333308339118958,
      "num_tokens": 93798832.0,
      "step": 7604
    },
    {
      "entropy": 1.2111939489841461,
      "epoch": 4.004739336492891,
      "grad_norm": 0.25802233815193176,
      "learning_rate": 9.34726869954112e-05,
      "loss": 0.15984860062599182,
      "mean_token_accuracy": 0.9333975464105606,
      "num_tokens": 93811168.0,
      "step": 7605
    },
    {
      "entropy": 1.279177486896515,
      "epoch": 4.005265929436545,
      "grad_norm": 0.28383538126945496,
      "learning_rate": 9.345113210762033e-05,
      "loss": 0.1689261943101883,
      "mean_token_accuracy": 0.9301945567131042,
      "num_tokens": 93823504.0,
      "step": 7606
    },
    {
      "entropy": 1.189927488565445,
      "epoch": 4.0057925223802,
      "grad_norm": 0.2803577780723572,
      "learning_rate": 9.342957820223928e-05,
      "loss": 0.16628915071487427,
      "mean_token_accuracy": 0.9307795763015747,
      "num_tokens": 93835840.0,
      "step": 7607
    },
    {
      "entropy": 1.2364709079265594,
      "epoch": 4.006319115323855,
      "grad_norm": 0.3095937967300415,
      "learning_rate": 9.340802528054759e-05,
      "loss": 0.1679101139307022,
      "mean_token_accuracy": 0.9346950799226761,
      "num_tokens": 93848176.0,
      "step": 7608
    },
    {
      "entropy": 1.2356630861759186,
      "epoch": 4.006845708267509,
      "grad_norm": 0.29372677206993103,
      "learning_rate": 9.338647334382468e-05,
      "loss": 0.17340870201587677,
      "mean_token_accuracy": 0.9274387061595917,
      "num_tokens": 93860512.0,
      "step": 7609
    },
    {
      "entropy": 1.1975701749324799,
      "epoch": 4.007372301211164,
      "grad_norm": 0.2953545153141022,
      "learning_rate": 9.336492239335009e-05,
      "loss": 0.16814272105693817,
      "mean_token_accuracy": 0.930532306432724,
      "num_tokens": 93872848.0,
      "step": 7610
    },
    {
      "entropy": 1.2367688715457916,
      "epoch": 4.007898894154819,
      "grad_norm": 0.32117795944213867,
      "learning_rate": 9.334337243040308e-05,
      "loss": 0.17463235557079315,
      "mean_token_accuracy": 0.9313275516033173,
      "num_tokens": 93885184.0,
      "step": 7611
    },
    {
      "entropy": 1.1813238263130188,
      "epoch": 4.008425487098473,
      "grad_norm": 0.3083871006965637,
      "learning_rate": 9.332182345626297e-05,
      "loss": 0.16935645043849945,
      "mean_token_accuracy": 0.9265637248754501,
      "num_tokens": 93897520.0,
      "step": 7612
    },
    {
      "entropy": 1.1819349527359009,
      "epoch": 4.008952080042127,
      "grad_norm": 0.2872036099433899,
      "learning_rate": 9.330027547220901e-05,
      "loss": 0.16818834841251373,
      "mean_token_accuracy": 0.9316693395376205,
      "num_tokens": 93909856.0,
      "step": 7613
    },
    {
      "entropy": 1.1806930601596832,
      "epoch": 4.009478672985782,
      "grad_norm": 0.30134302377700806,
      "learning_rate": 9.327872847952033e-05,
      "loss": 0.16665314137935638,
      "mean_token_accuracy": 0.9284096360206604,
      "num_tokens": 93922192.0,
      "step": 7614
    },
    {
      "entropy": 1.2107106745243073,
      "epoch": 4.010005265929436,
      "grad_norm": 0.3218785524368286,
      "learning_rate": 9.32571824794761e-05,
      "loss": 0.16514340043067932,
      "mean_token_accuracy": 0.9272081404924393,
      "num_tokens": 93934528.0,
      "step": 7615
    },
    {
      "entropy": 1.220152109861374,
      "epoch": 4.010531858873091,
      "grad_norm": 0.2992253601551056,
      "learning_rate": 9.323563747335534e-05,
      "loss": 0.15220335125923157,
      "mean_token_accuracy": 0.9355946183204651,
      "num_tokens": 93946864.0,
      "step": 7616
    },
    {
      "entropy": 1.1810352504253387,
      "epoch": 4.011058451816746,
      "grad_norm": 0.30068567395210266,
      "learning_rate": 9.321409346243708e-05,
      "loss": 0.17050591111183167,
      "mean_token_accuracy": 0.9311152696609497,
      "num_tokens": 93959200.0,
      "step": 7617
    },
    {
      "entropy": 1.2200407087802887,
      "epoch": 4.0115850447604,
      "grad_norm": 0.3415607511997223,
      "learning_rate": 9.319255044800026e-05,
      "loss": 0.19771385192871094,
      "mean_token_accuracy": 0.9130164235830307,
      "num_tokens": 93971536.0,
      "step": 7618
    },
    {
      "entropy": 1.22043839097023,
      "epoch": 4.012111637704055,
      "grad_norm": 0.29582586884498596,
      "learning_rate": 9.317100843132376e-05,
      "loss": 0.16629989445209503,
      "mean_token_accuracy": 0.9278903901576996,
      "num_tokens": 93983872.0,
      "step": 7619
    },
    {
      "entropy": 1.1786834001541138,
      "epoch": 4.01263823064771,
      "grad_norm": 0.27751415967941284,
      "learning_rate": 9.31494674136864e-05,
      "loss": 0.15829351544380188,
      "mean_token_accuracy": 0.9284335672855377,
      "num_tokens": 93996208.0,
      "step": 7620
    },
    {
      "entropy": 1.186080425977707,
      "epoch": 4.013164823591364,
      "grad_norm": 0.2798832654953003,
      "learning_rate": 9.312792739636695e-05,
      "loss": 0.16906145215034485,
      "mean_token_accuracy": 0.9283495247364044,
      "num_tokens": 94008544.0,
      "step": 7621
    },
    {
      "entropy": 1.2248172760009766,
      "epoch": 4.013691416535019,
      "grad_norm": 0.29892873764038086,
      "learning_rate": 9.310638838064412e-05,
      "loss": 0.16769753396511078,
      "mean_token_accuracy": 0.9286182522773743,
      "num_tokens": 94020880.0,
      "step": 7622
    },
    {
      "entropy": 1.207178771495819,
      "epoch": 4.014218009478673,
      "grad_norm": 0.3020927608013153,
      "learning_rate": 9.308485036779653e-05,
      "loss": 0.1729079931974411,
      "mean_token_accuracy": 0.9266164600849152,
      "num_tokens": 94033216.0,
      "step": 7623
    },
    {
      "entropy": 1.2358357906341553,
      "epoch": 4.0147446024223274,
      "grad_norm": 0.2990860939025879,
      "learning_rate": 9.306331335910279e-05,
      "loss": 0.16199710965156555,
      "mean_token_accuracy": 0.9346498548984528,
      "num_tokens": 94045552.0,
      "step": 7624
    },
    {
      "entropy": 1.2032581567764282,
      "epoch": 4.015271195365982,
      "grad_norm": 0.2720547616481781,
      "learning_rate": 9.30417773558414e-05,
      "loss": 0.15510684251785278,
      "mean_token_accuracy": 0.9321534186601639,
      "num_tokens": 94057888.0,
      "step": 7625
    },
    {
      "entropy": 1.1826758682727814,
      "epoch": 4.015797788309636,
      "grad_norm": 0.2762111723423004,
      "learning_rate": 9.30202423592909e-05,
      "loss": 0.1652383804321289,
      "mean_token_accuracy": 0.9285622835159302,
      "num_tokens": 94070224.0,
      "step": 7626
    },
    {
      "entropy": 1.2435753345489502,
      "epoch": 4.016324381253291,
      "grad_norm": 0.3012768030166626,
      "learning_rate": 9.299870837072963e-05,
      "loss": 0.1751018464565277,
      "mean_token_accuracy": 0.9221531003713608,
      "num_tokens": 94082560.0,
      "step": 7627
    },
    {
      "entropy": 1.2213923335075378,
      "epoch": 4.016850974196946,
      "grad_norm": 0.30483144521713257,
      "learning_rate": 9.297717539143595e-05,
      "loss": 0.177472323179245,
      "mean_token_accuracy": 0.9264393597841263,
      "num_tokens": 94094896.0,
      "step": 7628
    },
    {
      "entropy": 1.2358782887458801,
      "epoch": 4.0173775671406,
      "grad_norm": 0.2983492314815521,
      "learning_rate": 9.295564342268814e-05,
      "loss": 0.16957278549671173,
      "mean_token_accuracy": 0.9324619174003601,
      "num_tokens": 94107232.0,
      "step": 7629
    },
    {
      "entropy": 1.1866886913776398,
      "epoch": 4.017904160084255,
      "grad_norm": 0.2812538146972656,
      "learning_rate": 9.293411246576442e-05,
      "loss": 0.16486573219299316,
      "mean_token_accuracy": 0.9269909560680389,
      "num_tokens": 94119568.0,
      "step": 7630
    },
    {
      "entropy": 1.1595516502857208,
      "epoch": 4.01843075302791,
      "grad_norm": 0.27558693289756775,
      "learning_rate": 9.291258252194302e-05,
      "loss": 0.171391099691391,
      "mean_token_accuracy": 0.9275785386562347,
      "num_tokens": 94131904.0,
      "step": 7631
    },
    {
      "entropy": 1.2074815928936005,
      "epoch": 4.018957345971564,
      "grad_norm": 0.27395132184028625,
      "learning_rate": 9.289105359250197e-05,
      "loss": 0.1544472724199295,
      "mean_token_accuracy": 0.9366539567708969,
      "num_tokens": 94144240.0,
      "step": 7632
    },
    {
      "entropy": 1.2296015918254852,
      "epoch": 4.019483938915219,
      "grad_norm": 0.31429916620254517,
      "learning_rate": 9.286952567871937e-05,
      "loss": 0.18945646286010742,
      "mean_token_accuracy": 0.9228938817977905,
      "num_tokens": 94156576.0,
      "step": 7633
    },
    {
      "entropy": 1.2256301045417786,
      "epoch": 4.0200105318588735,
      "grad_norm": 0.3136304020881653,
      "learning_rate": 9.284799878187318e-05,
      "loss": 0.17368486523628235,
      "mean_token_accuracy": 0.9205718636512756,
      "num_tokens": 94168912.0,
      "step": 7634
    },
    {
      "entropy": 1.1827037334442139,
      "epoch": 4.0205371248025275,
      "grad_norm": 0.27767834067344666,
      "learning_rate": 9.282647290324131e-05,
      "loss": 0.15670537948608398,
      "mean_token_accuracy": 0.9334949851036072,
      "num_tokens": 94181248.0,
      "step": 7635
    },
    {
      "entropy": 1.160859763622284,
      "epoch": 4.021063717746182,
      "grad_norm": 0.28529998660087585,
      "learning_rate": 9.280494804410167e-05,
      "loss": 0.17210625112056732,
      "mean_token_accuracy": 0.92775097489357,
      "num_tokens": 94193584.0,
      "step": 7636
    },
    {
      "entropy": 1.234814703464508,
      "epoch": 4.021590310689837,
      "grad_norm": 0.3064624071121216,
      "learning_rate": 9.278342420573203e-05,
      "loss": 0.17227551341056824,
      "mean_token_accuracy": 0.9232846647500992,
      "num_tokens": 94205920.0,
      "step": 7637
    },
    {
      "entropy": 1.1790968775749207,
      "epoch": 4.022116903633491,
      "grad_norm": 0.2977254092693329,
      "learning_rate": 9.276190138941016e-05,
      "loss": 0.16933214664459229,
      "mean_token_accuracy": 0.9287389814853668,
      "num_tokens": 94218256.0,
      "step": 7638
    },
    {
      "entropy": 1.2044597268104553,
      "epoch": 4.022643496577146,
      "grad_norm": 0.31107184290885925,
      "learning_rate": 9.274037959641375e-05,
      "loss": 0.16668042540550232,
      "mean_token_accuracy": 0.9309228360652924,
      "num_tokens": 94230592.0,
      "step": 7639
    },
    {
      "entropy": 1.1837238669395447,
      "epoch": 4.0231700895208,
      "grad_norm": 0.28915640711784363,
      "learning_rate": 9.271885882802038e-05,
      "loss": 0.16029953956604004,
      "mean_token_accuracy": 0.9349488466978073,
      "num_tokens": 94242928.0,
      "step": 7640
    },
    {
      "entropy": 1.1911191940307617,
      "epoch": 4.023696682464455,
      "grad_norm": 0.2777077257633209,
      "learning_rate": 9.269733908550766e-05,
      "loss": 0.15964436531066895,
      "mean_token_accuracy": 0.9307241886854172,
      "num_tokens": 94255264.0,
      "step": 7641
    },
    {
      "entropy": 1.197167456150055,
      "epoch": 4.02422327540811,
      "grad_norm": 0.319174200296402,
      "learning_rate": 9.267582037015308e-05,
      "loss": 0.17807632684707642,
      "mean_token_accuracy": 0.9253161549568176,
      "num_tokens": 94267600.0,
      "step": 7642
    },
    {
      "entropy": 1.1747731864452362,
      "epoch": 4.024749868351764,
      "grad_norm": 0.3175990879535675,
      "learning_rate": 9.265430268323405e-05,
      "loss": 0.17977938055992126,
      "mean_token_accuracy": 0.9230655282735825,
      "num_tokens": 94279936.0,
      "step": 7643
    },
    {
      "entropy": 1.2191549837589264,
      "epoch": 4.025276461295419,
      "grad_norm": 0.30055975914001465,
      "learning_rate": 9.263278602602793e-05,
      "loss": 0.16622185707092285,
      "mean_token_accuracy": 0.9327243268489838,
      "num_tokens": 94292272.0,
      "step": 7644
    },
    {
      "entropy": 1.2362992465496063,
      "epoch": 4.0258030542390735,
      "grad_norm": 0.2924559414386749,
      "learning_rate": 9.261127039981212e-05,
      "loss": 0.15750610828399658,
      "mean_token_accuracy": 0.9389059543609619,
      "num_tokens": 94304608.0,
      "step": 7645
    },
    {
      "entropy": 1.2117838263511658,
      "epoch": 4.0263296471827275,
      "grad_norm": 0.2834373414516449,
      "learning_rate": 9.258975580586386e-05,
      "loss": 0.17272533476352692,
      "mean_token_accuracy": 0.9291851818561554,
      "num_tokens": 94316944.0,
      "step": 7646
    },
    {
      "entropy": 1.2353555560112,
      "epoch": 4.026856240126382,
      "grad_norm": 0.3317455053329468,
      "learning_rate": 9.256824224546033e-05,
      "loss": 0.16746200621128082,
      "mean_token_accuracy": 0.9323631823062897,
      "num_tokens": 94329280.0,
      "step": 7647
    },
    {
      "entropy": 1.2107363045215607,
      "epoch": 4.027382833070037,
      "grad_norm": 0.30810093879699707,
      "learning_rate": 9.254672971987863e-05,
      "loss": 0.16410067677497864,
      "mean_token_accuracy": 0.9268451780080795,
      "num_tokens": 94341616.0,
      "step": 7648
    },
    {
      "entropy": 1.2018360793590546,
      "epoch": 4.027909426013691,
      "grad_norm": 0.299425333738327,
      "learning_rate": 9.252521823039586e-05,
      "loss": 0.16118401288986206,
      "mean_token_accuracy": 0.9302340000867844,
      "num_tokens": 94353952.0,
      "step": 7649
    },
    {
      "entropy": 1.183328002691269,
      "epoch": 4.028436018957346,
      "grad_norm": 0.3065415322780609,
      "learning_rate": 9.250370777828903e-05,
      "loss": 0.17037290334701538,
      "mean_token_accuracy": 0.9274340569972992,
      "num_tokens": 94366288.0,
      "step": 7650
    },
    {
      "entropy": 1.2194631397724152,
      "epoch": 4.028962611901001,
      "grad_norm": 0.28854551911354065,
      "learning_rate": 9.24821983648351e-05,
      "loss": 0.16530030965805054,
      "mean_token_accuracy": 0.9310954511165619,
      "num_tokens": 94378624.0,
      "step": 7651
    },
    {
      "entropy": 1.1633674800395966,
      "epoch": 4.029489204844655,
      "grad_norm": 0.2949987053871155,
      "learning_rate": 9.246068999131095e-05,
      "loss": 0.16954991221427917,
      "mean_token_accuracy": 0.926768958568573,
      "num_tokens": 94390960.0,
      "step": 7652
    },
    {
      "entropy": 1.1591119170188904,
      "epoch": 4.03001579778831,
      "grad_norm": 0.30837175250053406,
      "learning_rate": 9.24391826589934e-05,
      "loss": 0.1687844842672348,
      "mean_token_accuracy": 0.9251392185688019,
      "num_tokens": 94403296.0,
      "step": 7653
    },
    {
      "entropy": 1.1513308882713318,
      "epoch": 4.030542390731965,
      "grad_norm": 0.27197301387786865,
      "learning_rate": 9.241767636915923e-05,
      "loss": 0.15422660112380981,
      "mean_token_accuracy": 0.9365798979997635,
      "num_tokens": 94415632.0,
      "step": 7654
    },
    {
      "entropy": 1.1872649192810059,
      "epoch": 4.031068983675619,
      "grad_norm": 0.3005768954753876,
      "learning_rate": 9.239617112308514e-05,
      "loss": 0.17201045155525208,
      "mean_token_accuracy": 0.9225270748138428,
      "num_tokens": 94427968.0,
      "step": 7655
    },
    {
      "entropy": 1.186728149652481,
      "epoch": 4.0315955766192735,
      "grad_norm": 0.3057633638381958,
      "learning_rate": 9.237466692204779e-05,
      "loss": 0.17150339484214783,
      "mean_token_accuracy": 0.9274269491434097,
      "num_tokens": 94440304.0,
      "step": 7656
    },
    {
      "entropy": 1.1701045334339142,
      "epoch": 4.0321221695629275,
      "grad_norm": 0.3205059766769409,
      "learning_rate": 9.235316376732372e-05,
      "loss": 0.17326562106609344,
      "mean_token_accuracy": 0.9279216825962067,
      "num_tokens": 94452640.0,
      "step": 7657
    },
    {
      "entropy": 1.1700592339038849,
      "epoch": 4.032648762506582,
      "grad_norm": 0.2768852710723877,
      "learning_rate": 9.23316616601894e-05,
      "loss": 0.1595529466867447,
      "mean_token_accuracy": 0.9313892722129822,
      "num_tokens": 94464976.0,
      "step": 7658
    },
    {
      "entropy": 1.1968066990375519,
      "epoch": 4.033175355450237,
      "grad_norm": 0.3190721869468689,
      "learning_rate": 9.231016060192144e-05,
      "loss": 0.17669986188411713,
      "mean_token_accuracy": 0.9311529695987701,
      "num_tokens": 94477312.0,
      "step": 7659
    },
    {
      "entropy": 1.1885199844837189,
      "epoch": 4.033701948393891,
      "grad_norm": 0.3203064799308777,
      "learning_rate": 9.22886605937961e-05,
      "loss": 0.19421350955963135,
      "mean_token_accuracy": 0.9194667488336563,
      "num_tokens": 94489648.0,
      "step": 7660
    },
    {
      "entropy": 1.1926580965518951,
      "epoch": 4.034228541337546,
      "grad_norm": 0.2843152582645416,
      "learning_rate": 9.226716163708977e-05,
      "loss": 0.17572253942489624,
      "mean_token_accuracy": 0.9240142405033112,
      "num_tokens": 94501984.0,
      "step": 7661
    },
    {
      "entropy": 1.1637931764125824,
      "epoch": 4.034755134281201,
      "grad_norm": 0.29263269901275635,
      "learning_rate": 9.224566373307868e-05,
      "loss": 0.1703728288412094,
      "mean_token_accuracy": 0.9271754175424576,
      "num_tokens": 94514320.0,
      "step": 7662
    },
    {
      "entropy": 1.1855376064777374,
      "epoch": 4.035281727224855,
      "grad_norm": 0.2865487337112427,
      "learning_rate": 9.222416688303908e-05,
      "loss": 0.16310441493988037,
      "mean_token_accuracy": 0.9321675002574921,
      "num_tokens": 94526656.0,
      "step": 7663
    },
    {
      "entropy": 1.2216345369815826,
      "epoch": 4.03580832016851,
      "grad_norm": 0.3266688883304596,
      "learning_rate": 9.220267108824705e-05,
      "loss": 0.18946443498134613,
      "mean_token_accuracy": 0.9214964807033539,
      "num_tokens": 94538992.0,
      "step": 7664
    },
    {
      "entropy": 1.2222075462341309,
      "epoch": 4.036334913112165,
      "grad_norm": 0.2968806028366089,
      "learning_rate": 9.218117634997872e-05,
      "loss": 0.1718311607837677,
      "mean_token_accuracy": 0.9296060502529144,
      "num_tokens": 94551328.0,
      "step": 7665
    },
    {
      "entropy": 1.2179431319236755,
      "epoch": 4.036861506055819,
      "grad_norm": 0.30279970169067383,
      "learning_rate": 9.21596826695101e-05,
      "loss": 0.16893449425697327,
      "mean_token_accuracy": 0.9293600171804428,
      "num_tokens": 94563664.0,
      "step": 7666
    },
    {
      "entropy": 1.1651307940483093,
      "epoch": 4.0373880989994735,
      "grad_norm": 0.29839229583740234,
      "learning_rate": 9.213819004811713e-05,
      "loss": 0.16824936866760254,
      "mean_token_accuracy": 0.9303262531757355,
      "num_tokens": 94576000.0,
      "step": 7667
    },
    {
      "entropy": 1.1775596141815186,
      "epoch": 4.037914691943128,
      "grad_norm": 0.2823132276535034,
      "learning_rate": 9.211669848707571e-05,
      "loss": 0.16800472140312195,
      "mean_token_accuracy": 0.9270628541707993,
      "num_tokens": 94588336.0,
      "step": 7668
    },
    {
      "entropy": 1.1981311738491058,
      "epoch": 4.038441284886782,
      "grad_norm": 0.2845676839351654,
      "learning_rate": 9.209520798766167e-05,
      "loss": 0.16031980514526367,
      "mean_token_accuracy": 0.930248573422432,
      "num_tokens": 94600672.0,
      "step": 7669
    },
    {
      "entropy": 1.1773464381694794,
      "epoch": 4.038967877830437,
      "grad_norm": 0.2765822410583496,
      "learning_rate": 9.207371855115079e-05,
      "loss": 0.1604149490594864,
      "mean_token_accuracy": 0.9327312558889389,
      "num_tokens": 94613008.0,
      "step": 7670
    },
    {
      "entropy": 1.1805593073368073,
      "epoch": 4.039494470774091,
      "grad_norm": 0.30690789222717285,
      "learning_rate": 9.205223017881876e-05,
      "loss": 0.16592201590538025,
      "mean_token_accuracy": 0.9334001690149307,
      "num_tokens": 94625344.0,
      "step": 7671
    },
    {
      "entropy": 1.2014459669589996,
      "epoch": 4.040021063717746,
      "grad_norm": 0.28434988856315613,
      "learning_rate": 9.203074287194118e-05,
      "loss": 0.1585979461669922,
      "mean_token_accuracy": 0.9333416223526001,
      "num_tokens": 94637680.0,
      "step": 7672
    },
    {
      "entropy": 1.161016583442688,
      "epoch": 4.040547656661401,
      "grad_norm": 0.29344049096107483,
      "learning_rate": 9.200925663179368e-05,
      "loss": 0.17373454570770264,
      "mean_token_accuracy": 0.9291245192289352,
      "num_tokens": 94650016.0,
      "step": 7673
    },
    {
      "entropy": 1.15232253074646,
      "epoch": 4.041074249605055,
      "grad_norm": 0.28777480125427246,
      "learning_rate": 9.198777145965176e-05,
      "loss": 0.1674036979675293,
      "mean_token_accuracy": 0.9266617000102997,
      "num_tokens": 94662352.0,
      "step": 7674
    },
    {
      "entropy": 1.1632183492183685,
      "epoch": 4.04160084254871,
      "grad_norm": 0.2981170117855072,
      "learning_rate": 9.196628735679089e-05,
      "loss": 0.16659529507160187,
      "mean_token_accuracy": 0.9293522536754608,
      "num_tokens": 94674688.0,
      "step": 7675
    },
    {
      "entropy": 1.1899208426475525,
      "epoch": 4.042127435492365,
      "grad_norm": 0.30841583013534546,
      "learning_rate": 9.194480432448642e-05,
      "loss": 0.16978931427001953,
      "mean_token_accuracy": 0.9258794486522675,
      "num_tokens": 94687024.0,
      "step": 7676
    },
    {
      "entropy": 1.1658337712287903,
      "epoch": 4.042654028436019,
      "grad_norm": 0.31560614705085754,
      "learning_rate": 9.192332236401366e-05,
      "loss": 0.174443319439888,
      "mean_token_accuracy": 0.9253016412258148,
      "num_tokens": 94699360.0,
      "step": 7677
    },
    {
      "entropy": 1.1231940388679504,
      "epoch": 4.0431806213796735,
      "grad_norm": 0.3076082766056061,
      "learning_rate": 9.190184147664791e-05,
      "loss": 0.18150557577610016,
      "mean_token_accuracy": 0.9241388291120529,
      "num_tokens": 94711696.0,
      "step": 7678
    },
    {
      "entropy": 1.1360673606395721,
      "epoch": 4.043707214323328,
      "grad_norm": 0.2781311273574829,
      "learning_rate": 9.188036166366434e-05,
      "loss": 0.14312730729579926,
      "mean_token_accuracy": 0.9410314112901688,
      "num_tokens": 94724032.0,
      "step": 7679
    },
    {
      "entropy": 1.1913892030715942,
      "epoch": 4.044233807266982,
      "grad_norm": 0.27803662419319153,
      "learning_rate": 9.185888292633808e-05,
      "loss": 0.15081775188446045,
      "mean_token_accuracy": 0.9355782866477966,
      "num_tokens": 94736368.0,
      "step": 7680
    },
    {
      "entropy": 1.231190025806427,
      "epoch": 4.044760400210637,
      "grad_norm": 0.31414955854415894,
      "learning_rate": 9.183740526594422e-05,
      "loss": 0.17506641149520874,
      "mean_token_accuracy": 0.9279109090566635,
      "num_tokens": 94748704.0,
      "step": 7681
    },
    {
      "entropy": 1.169010728597641,
      "epoch": 4.045286993154292,
      "grad_norm": 0.31384047865867615,
      "learning_rate": 9.181592868375774e-05,
      "loss": 0.16979721188545227,
      "mean_token_accuracy": 0.927495077252388,
      "num_tokens": 94761040.0,
      "step": 7682
    },
    {
      "entropy": 1.1776751279830933,
      "epoch": 4.045813586097946,
      "grad_norm": 0.30844661593437195,
      "learning_rate": 9.179445318105359e-05,
      "loss": 0.179082989692688,
      "mean_token_accuracy": 0.9232382923364639,
      "num_tokens": 94773376.0,
      "step": 7683
    },
    {
      "entropy": 1.1751681566238403,
      "epoch": 4.046340179041601,
      "grad_norm": 0.31246253848075867,
      "learning_rate": 9.177297875910667e-05,
      "loss": 0.17229889333248138,
      "mean_token_accuracy": 0.9299486726522446,
      "num_tokens": 94785712.0,
      "step": 7684
    },
    {
      "entropy": 1.1758607625961304,
      "epoch": 4.046866771985256,
      "grad_norm": 0.2907005250453949,
      "learning_rate": 9.175150541919175e-05,
      "loss": 0.17025050520896912,
      "mean_token_accuracy": 0.9275670498609543,
      "num_tokens": 94798048.0,
      "step": 7685
    },
    {
      "entropy": 1.175299733877182,
      "epoch": 4.04739336492891,
      "grad_norm": 0.3472633957862854,
      "learning_rate": 9.173003316258358e-05,
      "loss": 0.18637455999851227,
      "mean_token_accuracy": 0.9222482144832611,
      "num_tokens": 94810384.0,
      "step": 7686
    },
    {
      "entropy": 1.1338006854057312,
      "epoch": 4.047919957872565,
      "grad_norm": 0.31773582100868225,
      "learning_rate": 9.170856199055686e-05,
      "loss": 0.1717059314250946,
      "mean_token_accuracy": 0.9300213307142258,
      "num_tokens": 94822720.0,
      "step": 7687
    },
    {
      "entropy": 1.1469743847846985,
      "epoch": 4.048446550816219,
      "grad_norm": 0.30012696981430054,
      "learning_rate": 9.16870919043862e-05,
      "loss": 0.1653796136379242,
      "mean_token_accuracy": 0.9329687058925629,
      "num_tokens": 94835056.0,
      "step": 7688
    },
    {
      "entropy": 1.2179695963859558,
      "epoch": 4.0489731437598735,
      "grad_norm": 0.3160686790943146,
      "learning_rate": 9.166562290534617e-05,
      "loss": 0.1708589345216751,
      "mean_token_accuracy": 0.9323849081993103,
      "num_tokens": 94847392.0,
      "step": 7689
    },
    {
      "entropy": 1.1527775526046753,
      "epoch": 4.049499736703528,
      "grad_norm": 0.3100028932094574,
      "learning_rate": 9.164415499471126e-05,
      "loss": 0.16755883395671844,
      "mean_token_accuracy": 0.9321497231721878,
      "num_tokens": 94859728.0,
      "step": 7690
    },
    {
      "entropy": 1.1934599876403809,
      "epoch": 4.050026329647182,
      "grad_norm": 0.3215983211994171,
      "learning_rate": 9.162268817375588e-05,
      "loss": 0.17192670702934265,
      "mean_token_accuracy": 0.9277555197477341,
      "num_tokens": 94872064.0,
      "step": 7691
    },
    {
      "entropy": 1.1944094598293304,
      "epoch": 4.050552922590837,
      "grad_norm": 0.3270973265171051,
      "learning_rate": 9.160122244375438e-05,
      "loss": 0.17648163437843323,
      "mean_token_accuracy": 0.9270730316638947,
      "num_tokens": 94884400.0,
      "step": 7692
    },
    {
      "entropy": 1.085465520620346,
      "epoch": 4.051079515534492,
      "grad_norm": 0.2777954339981079,
      "learning_rate": 9.157975780598102e-05,
      "loss": 0.1513337343931198,
      "mean_token_accuracy": 0.9367313086986542,
      "num_tokens": 94896736.0,
      "step": 7693
    },
    {
      "entropy": 1.1473877131938934,
      "epoch": 4.051606108478146,
      "grad_norm": 0.28670284152030945,
      "learning_rate": 9.155829426171014e-05,
      "loss": 0.16905376315116882,
      "mean_token_accuracy": 0.9282949566841125,
      "num_tokens": 94909072.0,
      "step": 7694
    },
    {
      "entropy": 1.1510609090328217,
      "epoch": 4.052132701421801,
      "grad_norm": 0.29587459564208984,
      "learning_rate": 9.153683181221584e-05,
      "loss": 0.16510915756225586,
      "mean_token_accuracy": 0.9254784286022186,
      "num_tokens": 94921408.0,
      "step": 7695
    },
    {
      "entropy": 1.2123608887195587,
      "epoch": 4.052659294365456,
      "grad_norm": 0.30327266454696655,
      "learning_rate": 9.151537045877221e-05,
      "loss": 0.16675429046154022,
      "mean_token_accuracy": 0.9305920749902725,
      "num_tokens": 94933744.0,
      "step": 7696
    },
    {
      "entropy": 1.13215371966362,
      "epoch": 4.05318588730911,
      "grad_norm": 0.28354012966156006,
      "learning_rate": 9.149391020265329e-05,
      "loss": 0.1574067920446396,
      "mean_token_accuracy": 0.9331772923469543,
      "num_tokens": 94946080.0,
      "step": 7697
    },
    {
      "entropy": 1.1962259411811829,
      "epoch": 4.053712480252765,
      "grad_norm": 0.3534603714942932,
      "learning_rate": 9.14724510451331e-05,
      "loss": 0.17996813356876373,
      "mean_token_accuracy": 0.9196730256080627,
      "num_tokens": 94958416.0,
      "step": 7698
    },
    {
      "entropy": 1.1980598270893097,
      "epoch": 4.0542390731964195,
      "grad_norm": 0.31711629033088684,
      "learning_rate": 9.145099298748548e-05,
      "loss": 0.18711501359939575,
      "mean_token_accuracy": 0.9252251833677292,
      "num_tokens": 94970752.0,
      "step": 7699
    },
    {
      "entropy": 1.1515044569969177,
      "epoch": 4.0547656661400735,
      "grad_norm": 0.3168085515499115,
      "learning_rate": 9.142953603098426e-05,
      "loss": 0.1905432641506195,
      "mean_token_accuracy": 0.9175328314304352,
      "num_tokens": 94983088.0,
      "step": 7700
    },
    {
      "entropy": 1.2119370102882385,
      "epoch": 4.055292259083728,
      "grad_norm": 0.3189827501773834,
      "learning_rate": 9.140808017690327e-05,
      "loss": 0.17547422647476196,
      "mean_token_accuracy": 0.9245815277099609,
      "num_tokens": 94995424.0,
      "step": 7701
    },
    {
      "entropy": 1.1621211171150208,
      "epoch": 4.055818852027383,
      "grad_norm": 0.27919694781303406,
      "learning_rate": 9.138662542651621e-05,
      "loss": 0.1620962917804718,
      "mean_token_accuracy": 0.9340144544839859,
      "num_tokens": 95007760.0,
      "step": 7702
    },
    {
      "entropy": 1.1824475526809692,
      "epoch": 4.056345444971037,
      "grad_norm": 0.30076032876968384,
      "learning_rate": 9.136517178109672e-05,
      "loss": 0.16206148266792297,
      "mean_token_accuracy": 0.9283110499382019,
      "num_tokens": 95020096.0,
      "step": 7703
    },
    {
      "entropy": 1.1707939803600311,
      "epoch": 4.056872037914692,
      "grad_norm": 0.2912147641181946,
      "learning_rate": 9.134371924191835e-05,
      "loss": 0.16207239031791687,
      "mean_token_accuracy": 0.9271405339241028,
      "num_tokens": 95032432.0,
      "step": 7704
    },
    {
      "entropy": 1.1700113713741302,
      "epoch": 4.057398630858346,
      "grad_norm": 0.28976571559906006,
      "learning_rate": 9.132226781025466e-05,
      "loss": 0.1699160784482956,
      "mean_token_accuracy": 0.9242600500583649,
      "num_tokens": 95044768.0,
      "step": 7705
    },
    {
      "entropy": 1.1340029835700989,
      "epoch": 4.057925223802001,
      "grad_norm": 0.30861327052116394,
      "learning_rate": 9.130081748737904e-05,
      "loss": 0.1820501834154129,
      "mean_token_accuracy": 0.9228951781988144,
      "num_tokens": 95057104.0,
      "step": 7706
    },
    {
      "entropy": 1.2257311046123505,
      "epoch": 4.058451816745656,
      "grad_norm": 0.29901832342147827,
      "learning_rate": 9.127936827456487e-05,
      "loss": 0.1696225255727768,
      "mean_token_accuracy": 0.92838254570961,
      "num_tokens": 95069440.0,
      "step": 7707
    },
    {
      "entropy": 1.2083201706409454,
      "epoch": 4.05897840968931,
      "grad_norm": 0.3079107999801636,
      "learning_rate": 9.125792017308553e-05,
      "loss": 0.18569520115852356,
      "mean_token_accuracy": 0.9237589240074158,
      "num_tokens": 95081776.0,
      "step": 7708
    },
    {
      "entropy": 1.1831274926662445,
      "epoch": 4.059505002632965,
      "grad_norm": 0.32265934348106384,
      "learning_rate": 9.123647318421424e-05,
      "loss": 0.1859566867351532,
      "mean_token_accuracy": 0.9216562062501907,
      "num_tokens": 95094112.0,
      "step": 7709
    },
    {
      "entropy": 1.2142370343208313,
      "epoch": 4.0600315955766195,
      "grad_norm": 0.2909403443336487,
      "learning_rate": 9.121502730922418e-05,
      "loss": 0.14973674714565277,
      "mean_token_accuracy": 0.9379659593105316,
      "num_tokens": 95106448.0,
      "step": 7710
    },
    {
      "entropy": 1.199058085680008,
      "epoch": 4.0605581885202735,
      "grad_norm": 0.30686154961586,
      "learning_rate": 9.119358254938845e-05,
      "loss": 0.15622928738594055,
      "mean_token_accuracy": 0.9359981417655945,
      "num_tokens": 95118784.0,
      "step": 7711
    },
    {
      "entropy": 1.2475480735301971,
      "epoch": 4.061084781463928,
      "grad_norm": 0.3384522795677185,
      "learning_rate": 9.11721389059801e-05,
      "loss": 0.18361498415470123,
      "mean_token_accuracy": 0.9245945066213608,
      "num_tokens": 95131120.0,
      "step": 7712
    },
    {
      "entropy": 1.2219036519527435,
      "epoch": 4.061611374407583,
      "grad_norm": 0.3108859360218048,
      "learning_rate": 9.115069638027214e-05,
      "loss": 0.17015130817890167,
      "mean_token_accuracy": 0.9291042238473892,
      "num_tokens": 95143456.0,
      "step": 7713
    },
    {
      "entropy": 1.2195827066898346,
      "epoch": 4.062137967351237,
      "grad_norm": 0.319803386926651,
      "learning_rate": 9.112925497353746e-05,
      "loss": 0.19158239662647247,
      "mean_token_accuracy": 0.9158174991607666,
      "num_tokens": 95155792.0,
      "step": 7714
    },
    {
      "entropy": 1.237177848815918,
      "epoch": 4.062664560294892,
      "grad_norm": 0.2918310761451721,
      "learning_rate": 9.110781468704892e-05,
      "loss": 0.1691100299358368,
      "mean_token_accuracy": 0.9284612387418747,
      "num_tokens": 95168128.0,
      "step": 7715
    },
    {
      "entropy": 1.2189947366714478,
      "epoch": 4.063191153238547,
      "grad_norm": 0.326535165309906,
      "learning_rate": 9.108637552207935e-05,
      "loss": 0.18463623523712158,
      "mean_token_accuracy": 0.9204123914241791,
      "num_tokens": 95180464.0,
      "step": 7716
    },
    {
      "entropy": 1.1953076124191284,
      "epoch": 4.063717746182201,
      "grad_norm": 0.32633188366889954,
      "learning_rate": 9.106493747990139e-05,
      "loss": 0.17715126276016235,
      "mean_token_accuracy": 0.9264031648635864,
      "num_tokens": 95192800.0,
      "step": 7717
    },
    {
      "entropy": 1.2300822734832764,
      "epoch": 4.064244339125856,
      "grad_norm": 0.32896727323532104,
      "learning_rate": 9.104350056178774e-05,
      "loss": 0.17694605886936188,
      "mean_token_accuracy": 0.9269661009311676,
      "num_tokens": 95205136.0,
      "step": 7718
    },
    {
      "entropy": 1.2718767821788788,
      "epoch": 4.064770932069511,
      "grad_norm": 0.31029990315437317,
      "learning_rate": 9.102206476901098e-05,
      "loss": 0.14992089569568634,
      "mean_token_accuracy": 0.9372245967388153,
      "num_tokens": 95217472.0,
      "step": 7719
    },
    {
      "entropy": 1.1803152859210968,
      "epoch": 4.065297525013165,
      "grad_norm": 0.28587645292282104,
      "learning_rate": 9.100063010284366e-05,
      "loss": 0.15189969539642334,
      "mean_token_accuracy": 0.9393310397863388,
      "num_tokens": 95229808.0,
      "step": 7720
    },
    {
      "entropy": 1.2242991924285889,
      "epoch": 4.0658241179568195,
      "grad_norm": 0.2987886667251587,
      "learning_rate": 9.097919656455808e-05,
      "loss": 0.16917482018470764,
      "mean_token_accuracy": 0.9314169138669968,
      "num_tokens": 95242144.0,
      "step": 7721
    },
    {
      "entropy": 1.2272664308547974,
      "epoch": 4.0663507109004735,
      "grad_norm": 0.30778566002845764,
      "learning_rate": 9.095776415542683e-05,
      "loss": 0.17835679650306702,
      "mean_token_accuracy": 0.9243713170289993,
      "num_tokens": 95254480.0,
      "step": 7722
    },
    {
      "entropy": 1.185521274805069,
      "epoch": 4.066877303844128,
      "grad_norm": 0.3092713952064514,
      "learning_rate": 9.093633287672212e-05,
      "loss": 0.1805608719587326,
      "mean_token_accuracy": 0.922214686870575,
      "num_tokens": 95266816.0,
      "step": 7723
    },
    {
      "entropy": 1.2591804265975952,
      "epoch": 4.067403896787783,
      "grad_norm": 0.29992327094078064,
      "learning_rate": 9.091490272971623e-05,
      "loss": 0.16175399720668793,
      "mean_token_accuracy": 0.9302794933319092,
      "num_tokens": 95279152.0,
      "step": 7724
    },
    {
      "entropy": 1.2036868631839752,
      "epoch": 4.067930489731437,
      "grad_norm": 0.29974281787872314,
      "learning_rate": 9.08934737156813e-05,
      "loss": 0.16755646467208862,
      "mean_token_accuracy": 0.9305300414562225,
      "num_tokens": 95291488.0,
      "step": 7725
    },
    {
      "entropy": 1.2664932012557983,
      "epoch": 4.068457082675092,
      "grad_norm": 0.30560722947120667,
      "learning_rate": 9.087204583588951e-05,
      "loss": 0.1747075617313385,
      "mean_token_accuracy": 0.9243289828300476,
      "num_tokens": 95303824.0,
      "step": 7726
    },
    {
      "entropy": 1.2596045136451721,
      "epoch": 4.068983675618747,
      "grad_norm": 0.3133663833141327,
      "learning_rate": 9.085061909161285e-05,
      "loss": 0.16157250106334686,
      "mean_token_accuracy": 0.9298543632030487,
      "num_tokens": 95316160.0,
      "step": 7727
    },
    {
      "entropy": 1.2500620186328888,
      "epoch": 4.069510268562401,
      "grad_norm": 0.31191858649253845,
      "learning_rate": 9.082919348412329e-05,
      "loss": 0.18304812908172607,
      "mean_token_accuracy": 0.9186305999755859,
      "num_tokens": 95328496.0,
      "step": 7728
    },
    {
      "entropy": 1.20553919672966,
      "epoch": 4.070036861506056,
      "grad_norm": 0.3020358681678772,
      "learning_rate": 9.080776901469284e-05,
      "loss": 0.16491544246673584,
      "mean_token_accuracy": 0.9246530532836914,
      "num_tokens": 95340832.0,
      "step": 7729
    },
    {
      "entropy": 1.2194979190826416,
      "epoch": 4.070563454449711,
      "grad_norm": 0.2910735011100769,
      "learning_rate": 9.078634568459324e-05,
      "loss": 0.1644223928451538,
      "mean_token_accuracy": 0.9331174492835999,
      "num_tokens": 95353168.0,
      "step": 7730
    },
    {
      "entropy": 1.2480650544166565,
      "epoch": 4.071090047393365,
      "grad_norm": 0.30329522490501404,
      "learning_rate": 9.076492349509633e-05,
      "loss": 0.1677975356578827,
      "mean_token_accuracy": 0.927104040980339,
      "num_tokens": 95365504.0,
      "step": 7731
    },
    {
      "entropy": 1.2466902136802673,
      "epoch": 4.0716166403370195,
      "grad_norm": 0.28745365142822266,
      "learning_rate": 9.074350244747379e-05,
      "loss": 0.15721100568771362,
      "mean_token_accuracy": 0.9370080977678299,
      "num_tokens": 95377840.0,
      "step": 7732
    },
    {
      "entropy": 1.2403151988983154,
      "epoch": 4.072143233280674,
      "grad_norm": 0.2975686192512512,
      "learning_rate": 9.072208254299727e-05,
      "loss": 0.1805305778980255,
      "mean_token_accuracy": 0.9232339859008789,
      "num_tokens": 95390176.0,
      "step": 7733
    },
    {
      "entropy": 1.2152469158172607,
      "epoch": 4.072669826224328,
      "grad_norm": 0.2642306387424469,
      "learning_rate": 9.070066378293836e-05,
      "loss": 0.14989545941352844,
      "mean_token_accuracy": 0.9385390877723694,
      "num_tokens": 95402512.0,
      "step": 7734
    },
    {
      "entropy": 1.2145140171051025,
      "epoch": 4.073196419167983,
      "grad_norm": 0.29692116379737854,
      "learning_rate": 9.067924616856854e-05,
      "loss": 0.16534504294395447,
      "mean_token_accuracy": 0.9297915995121002,
      "num_tokens": 95414848.0,
      "step": 7735
    },
    {
      "entropy": 1.244487464427948,
      "epoch": 4.073723012111638,
      "grad_norm": 0.3275160491466522,
      "learning_rate": 9.065782970115928e-05,
      "loss": 0.18590737879276276,
      "mean_token_accuracy": 0.924352690577507,
      "num_tokens": 95427184.0,
      "step": 7736
    },
    {
      "entropy": 1.2427468597888947,
      "epoch": 4.074249605055292,
      "grad_norm": 0.3044041097164154,
      "learning_rate": 9.063641438198194e-05,
      "loss": 0.17704616487026215,
      "mean_token_accuracy": 0.9291693568229675,
      "num_tokens": 95439520.0,
      "step": 7737
    },
    {
      "entropy": 1.2756830751895905,
      "epoch": 4.074776197998947,
      "grad_norm": 0.35712915658950806,
      "learning_rate": 9.061500021230782e-05,
      "loss": 0.16697832942008972,
      "mean_token_accuracy": 0.930440753698349,
      "num_tokens": 95451856.0,
      "step": 7738
    },
    {
      "entropy": 1.2054972350597382,
      "epoch": 4.075302790942601,
      "grad_norm": 0.3086952269077301,
      "learning_rate": 9.059358719340816e-05,
      "loss": 0.16568411886692047,
      "mean_token_accuracy": 0.926430806517601,
      "num_tokens": 95464192.0,
      "step": 7739
    },
    {
      "entropy": 1.225044995546341,
      "epoch": 4.075829383886256,
      "grad_norm": 0.29447293281555176,
      "learning_rate": 9.057217532655411e-05,
      "loss": 0.15924039483070374,
      "mean_token_accuracy": 0.9344728738069534,
      "num_tokens": 95476528.0,
      "step": 7740
    },
    {
      "entropy": 1.2552023231983185,
      "epoch": 4.076355976829911,
      "grad_norm": 0.3137618601322174,
      "learning_rate": 9.055076461301677e-05,
      "loss": 0.17044930160045624,
      "mean_token_accuracy": 0.9279296547174454,
      "num_tokens": 95488864.0,
      "step": 7741
    },
    {
      "entropy": 1.2101345360279083,
      "epoch": 4.076882569773565,
      "grad_norm": 0.33105573058128357,
      "learning_rate": 9.052935505406715e-05,
      "loss": 0.17854714393615723,
      "mean_token_accuracy": 0.9252999573945999,
      "num_tokens": 95501200.0,
      "step": 7742
    },
    {
      "entropy": 1.2018029391765594,
      "epoch": 4.0774091627172195,
      "grad_norm": 0.3053717613220215,
      "learning_rate": 9.050794665097628e-05,
      "loss": 0.1497926563024521,
      "mean_token_accuracy": 0.9343288093805313,
      "num_tokens": 95513536.0,
      "step": 7743
    },
    {
      "entropy": 1.2654300034046173,
      "epoch": 4.077935755660874,
      "grad_norm": 0.35785624384880066,
      "learning_rate": 9.048653940501499e-05,
      "loss": 0.17250262200832367,
      "mean_token_accuracy": 0.9261425137519836,
      "num_tokens": 95525872.0,
      "step": 7744
    },
    {
      "entropy": 1.1997900307178497,
      "epoch": 4.078462348604528,
      "grad_norm": 0.32767707109451294,
      "learning_rate": 9.046513331745414e-05,
      "loss": 0.1583937704563141,
      "mean_token_accuracy": 0.9334874302148819,
      "num_tokens": 95538208.0,
      "step": 7745
    },
    {
      "entropy": 1.2555785775184631,
      "epoch": 4.078988941548183,
      "grad_norm": 0.3156816363334656,
      "learning_rate": 9.044372838956446e-05,
      "loss": 0.17291590571403503,
      "mean_token_accuracy": 0.9330429434776306,
      "num_tokens": 95550544.0,
      "step": 7746
    },
    {
      "entropy": 1.194413661956787,
      "epoch": 4.079515534491838,
      "grad_norm": 0.3167170584201813,
      "learning_rate": 9.042232462261664e-05,
      "loss": 0.1628277599811554,
      "mean_token_accuracy": 0.9321805536746979,
      "num_tokens": 95562880.0,
      "step": 7747
    },
    {
      "entropy": 1.2325744330883026,
      "epoch": 4.080042127435492,
      "grad_norm": 0.3167288303375244,
      "learning_rate": 9.040092201788129e-05,
      "loss": 0.16678446531295776,
      "mean_token_accuracy": 0.9272855222225189,
      "num_tokens": 95575216.0,
      "step": 7748
    },
    {
      "entropy": 1.23721382021904,
      "epoch": 4.080568720379147,
      "grad_norm": 0.33681100606918335,
      "learning_rate": 9.037952057662895e-05,
      "loss": 0.18301434814929962,
      "mean_token_accuracy": 0.9208259582519531,
      "num_tokens": 95587552.0,
      "step": 7749
    },
    {
      "entropy": 1.2467168271541595,
      "epoch": 4.081095313322802,
      "grad_norm": 0.32016125321388245,
      "learning_rate": 9.035812030013013e-05,
      "loss": 0.16686171293258667,
      "mean_token_accuracy": 0.9261872917413712,
      "num_tokens": 95599888.0,
      "step": 7750
    },
    {
      "entropy": 1.2390804886817932,
      "epoch": 4.081621906266456,
      "grad_norm": 0.31561383605003357,
      "learning_rate": 9.033672118965526e-05,
      "loss": 0.16775859892368317,
      "mean_token_accuracy": 0.9265642762184143,
      "num_tokens": 95612224.0,
      "step": 7751
    },
    {
      "entropy": 1.2519412338733673,
      "epoch": 4.082148499210111,
      "grad_norm": 0.32623958587646484,
      "learning_rate": 9.031532324647462e-05,
      "loss": 0.16160587966442108,
      "mean_token_accuracy": 0.936798706650734,
      "num_tokens": 95624560.0,
      "step": 7752
    },
    {
      "entropy": 1.2217656672000885,
      "epoch": 4.082675092153766,
      "grad_norm": 0.3501085042953491,
      "learning_rate": 9.02939264718585e-05,
      "loss": 0.18669341504573822,
      "mean_token_accuracy": 0.9153125435113907,
      "num_tokens": 95636896.0,
      "step": 7753
    },
    {
      "entropy": 1.2231608033180237,
      "epoch": 4.0832016850974195,
      "grad_norm": 0.295173317193985,
      "learning_rate": 9.027253086707716e-05,
      "loss": 0.15407906472682953,
      "mean_token_accuracy": 0.9321415275335312,
      "num_tokens": 95649232.0,
      "step": 7754
    },
    {
      "entropy": 1.2381119132041931,
      "epoch": 4.083728278041074,
      "grad_norm": 0.3035016655921936,
      "learning_rate": 9.025113643340064e-05,
      "loss": 0.16616517305374146,
      "mean_token_accuracy": 0.9286200404167175,
      "num_tokens": 95661568.0,
      "step": 7755
    },
    {
      "entropy": 1.2252253592014313,
      "epoch": 4.084254870984728,
      "grad_norm": 0.31341826915740967,
      "learning_rate": 9.022974317209902e-05,
      "loss": 0.15845082700252533,
      "mean_token_accuracy": 0.9357934594154358,
      "num_tokens": 95673904.0,
      "step": 7756
    },
    {
      "entropy": 1.1895354688167572,
      "epoch": 4.084781463928383,
      "grad_norm": 0.2934425473213196,
      "learning_rate": 9.020835108444235e-05,
      "loss": 0.15976643562316895,
      "mean_token_accuracy": 0.9344072192907333,
      "num_tokens": 95686240.0,
      "step": 7757
    },
    {
      "entropy": 1.238107591867447,
      "epoch": 4.085308056872038,
      "grad_norm": 0.30392828583717346,
      "learning_rate": 9.018696017170054e-05,
      "loss": 0.1551521122455597,
      "mean_token_accuracy": 0.9287526309490204,
      "num_tokens": 95698576.0,
      "step": 7758
    },
    {
      "entropy": 1.2240920960903168,
      "epoch": 4.085834649815692,
      "grad_norm": 0.3208746612071991,
      "learning_rate": 9.01655704351434e-05,
      "loss": 0.1767064332962036,
      "mean_token_accuracy": 0.9299392402172089,
      "num_tokens": 95710912.0,
      "step": 7759
    },
    {
      "entropy": 1.2479586005210876,
      "epoch": 4.086361242759347,
      "grad_norm": 0.31244781613349915,
      "learning_rate": 9.014418187604074e-05,
      "loss": 0.18113332986831665,
      "mean_token_accuracy": 0.9223150759935379,
      "num_tokens": 95723248.0,
      "step": 7760
    },
    {
      "entropy": 1.2941612601280212,
      "epoch": 4.086887835703002,
      "grad_norm": 0.32946693897247314,
      "learning_rate": 9.012279449566228e-05,
      "loss": 0.18078386783599854,
      "mean_token_accuracy": 0.9275902062654495,
      "num_tokens": 95735584.0,
      "step": 7761
    },
    {
      "entropy": 1.2265251874923706,
      "epoch": 4.087414428646656,
      "grad_norm": 0.2798350155353546,
      "learning_rate": 9.010140829527767e-05,
      "loss": 0.16829606890678406,
      "mean_token_accuracy": 0.9291946738958359,
      "num_tokens": 95747920.0,
      "step": 7762
    },
    {
      "entropy": 1.2662497758865356,
      "epoch": 4.087941021590311,
      "grad_norm": 0.30638816952705383,
      "learning_rate": 9.008002327615644e-05,
      "loss": 0.16932989656925201,
      "mean_token_accuracy": 0.9284424632787704,
      "num_tokens": 95760256.0,
      "step": 7763
    },
    {
      "entropy": 1.218248039484024,
      "epoch": 4.088467614533966,
      "grad_norm": 0.32308030128479004,
      "learning_rate": 9.005863943956812e-05,
      "loss": 0.16459035873413086,
      "mean_token_accuracy": 0.930854395031929,
      "num_tokens": 95772592.0,
      "step": 7764
    },
    {
      "entropy": 1.2341713905334473,
      "epoch": 4.0889942074776195,
      "grad_norm": 0.3084065020084381,
      "learning_rate": 9.003725678678219e-05,
      "loss": 0.17533165216445923,
      "mean_token_accuracy": 0.9273016005754471,
      "num_tokens": 95784928.0,
      "step": 7765
    },
    {
      "entropy": 1.2990855276584625,
      "epoch": 4.089520800421274,
      "grad_norm": 0.30591991543769836,
      "learning_rate": 9.001587531906794e-05,
      "loss": 0.15463925898075104,
      "mean_token_accuracy": 0.9317610263824463,
      "num_tokens": 95797264.0,
      "step": 7766
    },
    {
      "entropy": 1.225580245256424,
      "epoch": 4.090047393364929,
      "grad_norm": 0.2803047001361847,
      "learning_rate": 8.999449503769472e-05,
      "loss": 0.15674909949302673,
      "mean_token_accuracy": 0.935524508357048,
      "num_tokens": 95809600.0,
      "step": 7767
    },
    {
      "entropy": 1.2573725581169128,
      "epoch": 4.090573986308583,
      "grad_norm": 0.3105691075325012,
      "learning_rate": 8.997311594393172e-05,
      "loss": 0.16267746686935425,
      "mean_token_accuracy": 0.9329666048288345,
      "num_tokens": 95821936.0,
      "step": 7768
    },
    {
      "entropy": 1.2551613450050354,
      "epoch": 4.091100579252238,
      "grad_norm": 0.2923014461994171,
      "learning_rate": 8.995173803904811e-05,
      "loss": 0.16139426827430725,
      "mean_token_accuracy": 0.9278410375118256,
      "num_tokens": 95834272.0,
      "step": 7769
    },
    {
      "entropy": 1.239789605140686,
      "epoch": 4.091627172195893,
      "grad_norm": 0.2825500965118408,
      "learning_rate": 8.993036132431292e-05,
      "loss": 0.15421581268310547,
      "mean_token_accuracy": 0.9366891384124756,
      "num_tokens": 95846608.0,
      "step": 7770
    },
    {
      "entropy": 1.2593735456466675,
      "epoch": 4.092153765139547,
      "grad_norm": 0.2957497537136078,
      "learning_rate": 8.990898580099524e-05,
      "loss": 0.15918530523777008,
      "mean_token_accuracy": 0.9337555021047592,
      "num_tokens": 95858944.0,
      "step": 7771
    },
    {
      "entropy": 1.2203782498836517,
      "epoch": 4.092680358083202,
      "grad_norm": 0.2896980941295624,
      "learning_rate": 8.988761147036397e-05,
      "loss": 0.1596841812133789,
      "mean_token_accuracy": 0.9279924631118774,
      "num_tokens": 95871280.0,
      "step": 7772
    },
    {
      "entropy": 1.2221176326274872,
      "epoch": 4.093206951026856,
      "grad_norm": 0.2983795702457428,
      "learning_rate": 8.986623833368801e-05,
      "loss": 0.16424298286437988,
      "mean_token_accuracy": 0.9330798983573914,
      "num_tokens": 95883616.0,
      "step": 7773
    },
    {
      "entropy": 1.2075627148151398,
      "epoch": 4.093733543970511,
      "grad_norm": 0.30806633830070496,
      "learning_rate": 8.98448663922361e-05,
      "loss": 0.17070338129997253,
      "mean_token_accuracy": 0.9266282320022583,
      "num_tokens": 95895952.0,
      "step": 7774
    },
    {
      "entropy": 1.2562209963798523,
      "epoch": 4.094260136914166,
      "grad_norm": 0.3529546856880188,
      "learning_rate": 8.982349564727699e-05,
      "loss": 0.17133307456970215,
      "mean_token_accuracy": 0.9285664856433868,
      "num_tokens": 95908288.0,
      "step": 7775
    },
    {
      "entropy": 1.2491154372692108,
      "epoch": 4.0947867298578196,
      "grad_norm": 0.3027105927467346,
      "learning_rate": 8.980212610007939e-05,
      "loss": 0.15416307747364044,
      "mean_token_accuracy": 0.9330688118934631,
      "num_tokens": 95920624.0,
      "step": 7776
    },
    {
      "entropy": 1.2435694336891174,
      "epoch": 4.095313322801474,
      "grad_norm": 0.30684900283813477,
      "learning_rate": 8.978075775191177e-05,
      "loss": 0.1644163578748703,
      "mean_token_accuracy": 0.9304230064153671,
      "num_tokens": 95932960.0,
      "step": 7777
    },
    {
      "entropy": 1.220851480960846,
      "epoch": 4.095839915745129,
      "grad_norm": 0.32513976097106934,
      "learning_rate": 8.975939060404277e-05,
      "loss": 0.17666807770729065,
      "mean_token_accuracy": 0.9211707562208176,
      "num_tokens": 95945296.0,
      "step": 7778
    },
    {
      "entropy": 1.2059510946273804,
      "epoch": 4.096366508688783,
      "grad_norm": 0.28802773356437683,
      "learning_rate": 8.973802465774075e-05,
      "loss": 0.160865917801857,
      "mean_token_accuracy": 0.9303705394268036,
      "num_tokens": 95957632.0,
      "step": 7779
    },
    {
      "entropy": 1.2136945128440857,
      "epoch": 4.096893101632438,
      "grad_norm": 0.3125780522823334,
      "learning_rate": 8.971665991427414e-05,
      "loss": 0.16175074875354767,
      "mean_token_accuracy": 0.9307304322719574,
      "num_tokens": 95969968.0,
      "step": 7780
    },
    {
      "entropy": 1.1918500661849976,
      "epoch": 4.097419694576093,
      "grad_norm": 0.3186854422092438,
      "learning_rate": 8.969529637491119e-05,
      "loss": 0.16778559982776642,
      "mean_token_accuracy": 0.924680545926094,
      "num_tokens": 95982304.0,
      "step": 7781
    },
    {
      "entropy": 1.2299046516418457,
      "epoch": 4.097946287519747,
      "grad_norm": 0.333080530166626,
      "learning_rate": 8.967393404092018e-05,
      "loss": 0.1835906207561493,
      "mean_token_accuracy": 0.9239201098680496,
      "num_tokens": 95994640.0,
      "step": 7782
    },
    {
      "entropy": 1.1866918206214905,
      "epoch": 4.098472880463402,
      "grad_norm": 0.29609766602516174,
      "learning_rate": 8.96525729135692e-05,
      "loss": 0.16436761617660522,
      "mean_token_accuracy": 0.9310314506292343,
      "num_tokens": 96006976.0,
      "step": 7783
    },
    {
      "entropy": 1.1631359457969666,
      "epoch": 4.098999473407057,
      "grad_norm": 0.3122749328613281,
      "learning_rate": 8.963121299412636e-05,
      "loss": 0.18169131875038147,
      "mean_token_accuracy": 0.9222935438156128,
      "num_tokens": 96019312.0,
      "step": 7784
    },
    {
      "entropy": 1.2098788022994995,
      "epoch": 4.099526066350711,
      "grad_norm": 0.2992385923862457,
      "learning_rate": 8.96098542838597e-05,
      "loss": 0.16331830620765686,
      "mean_token_accuracy": 0.9328923672437668,
      "num_tokens": 96031648.0,
      "step": 7785
    },
    {
      "entropy": 1.2220342457294464,
      "epoch": 4.100052659294366,
      "grad_norm": 0.320968896150589,
      "learning_rate": 8.958849678403716e-05,
      "loss": 0.16907034814357758,
      "mean_token_accuracy": 0.9300814419984818,
      "num_tokens": 96043984.0,
      "step": 7786
    },
    {
      "entropy": 1.2747842371463776,
      "epoch": 4.1005792522380204,
      "grad_norm": 0.3546020984649658,
      "learning_rate": 8.956714049592661e-05,
      "loss": 0.19964481890201569,
      "mean_token_accuracy": 0.9119447022676468,
      "num_tokens": 96056320.0,
      "step": 7787
    },
    {
      "entropy": 1.244139701128006,
      "epoch": 4.101105845181674,
      "grad_norm": 0.3346027731895447,
      "learning_rate": 8.954578542079584e-05,
      "loss": 0.174786776304245,
      "mean_token_accuracy": 0.9248683005571365,
      "num_tokens": 96068656.0,
      "step": 7788
    },
    {
      "entropy": 1.2504622042179108,
      "epoch": 4.101632438125329,
      "grad_norm": 0.33030298352241516,
      "learning_rate": 8.952443155991257e-05,
      "loss": 0.1916545033454895,
      "mean_token_accuracy": 0.9185705184936523,
      "num_tokens": 96080992.0,
      "step": 7789
    },
    {
      "entropy": 1.2617883086204529,
      "epoch": 4.102159031068983,
      "grad_norm": 0.3044522702693939,
      "learning_rate": 8.950307891454444e-05,
      "loss": 0.17294572293758392,
      "mean_token_accuracy": 0.9226913154125214,
      "num_tokens": 96093328.0,
      "step": 7790
    },
    {
      "entropy": 1.2183136641979218,
      "epoch": 4.102685624012638,
      "grad_norm": 0.3035319447517395,
      "learning_rate": 8.9481727485959e-05,
      "loss": 0.1683570146560669,
      "mean_token_accuracy": 0.9234451353549957,
      "num_tokens": 96105664.0,
      "step": 7791
    },
    {
      "entropy": 1.1727025508880615,
      "epoch": 4.103212216956293,
      "grad_norm": 0.2806466221809387,
      "learning_rate": 8.946037727542389e-05,
      "loss": 0.16041648387908936,
      "mean_token_accuracy": 0.9294952601194382,
      "num_tokens": 96118000.0,
      "step": 7792
    },
    {
      "entropy": 1.2113775610923767,
      "epoch": 4.103738809899947,
      "grad_norm": 0.2875956892967224,
      "learning_rate": 8.943902828420645e-05,
      "loss": 0.170582577586174,
      "mean_token_accuracy": 0.9257608503103256,
      "num_tokens": 96130336.0,
      "step": 7793
    },
    {
      "entropy": 1.1674440205097198,
      "epoch": 4.104265402843602,
      "grad_norm": 0.27453169226646423,
      "learning_rate": 8.941768051357407e-05,
      "loss": 0.15505698323249817,
      "mean_token_accuracy": 0.9339439123868942,
      "num_tokens": 96142672.0,
      "step": 7794
    },
    {
      "entropy": 1.204866111278534,
      "epoch": 4.104791995787257,
      "grad_norm": 0.2872864902019501,
      "learning_rate": 8.939633396479401e-05,
      "loss": 0.1590895652770996,
      "mean_token_accuracy": 0.9344259947538376,
      "num_tokens": 96155008.0,
      "step": 7795
    },
    {
      "entropy": 1.1859584152698517,
      "epoch": 4.105318588730911,
      "grad_norm": 0.28996652364730835,
      "learning_rate": 8.937498863913352e-05,
      "loss": 0.1534387171268463,
      "mean_token_accuracy": 0.9312669485807419,
      "num_tokens": 96167344.0,
      "step": 7796
    },
    {
      "entropy": 1.2309530079364777,
      "epoch": 4.105845181674566,
      "grad_norm": 0.34393784403800964,
      "learning_rate": 8.935364453785975e-05,
      "loss": 0.18367165327072144,
      "mean_token_accuracy": 0.923644557595253,
      "num_tokens": 96179680.0,
      "step": 7797
    },
    {
      "entropy": 1.210770308971405,
      "epoch": 4.1063717746182204,
      "grad_norm": 0.2864152193069458,
      "learning_rate": 8.933230166223973e-05,
      "loss": 0.15165504813194275,
      "mean_token_accuracy": 0.9372637122869492,
      "num_tokens": 96192016.0,
      "step": 7798
    },
    {
      "entropy": 1.1396176218986511,
      "epoch": 4.106898367561874,
      "grad_norm": 0.3214961886405945,
      "learning_rate": 8.931096001354056e-05,
      "loss": 0.18969902396202087,
      "mean_token_accuracy": 0.9195330291986465,
      "num_tokens": 96204352.0,
      "step": 7799
    },
    {
      "entropy": 1.1593232154846191,
      "epoch": 4.107424960505529,
      "grad_norm": 0.3097698390483856,
      "learning_rate": 8.928961959302908e-05,
      "loss": 0.1748993694782257,
      "mean_token_accuracy": 0.9295019805431366,
      "num_tokens": 96216688.0,
      "step": 7800
    },
    {
      "entropy": 1.161198079586029,
      "epoch": 4.107951553449184,
      "grad_norm": 0.30992424488067627,
      "learning_rate": 8.926828040197218e-05,
      "loss": 0.1573079526424408,
      "mean_token_accuracy": 0.9285661578178406,
      "num_tokens": 96229024.0,
      "step": 7801
    },
    {
      "entropy": 1.1368975341320038,
      "epoch": 4.108478146392838,
      "grad_norm": 0.2966200113296509,
      "learning_rate": 8.924694244163665e-05,
      "loss": 0.16821102797985077,
      "mean_token_accuracy": 0.9266536235809326,
      "num_tokens": 96241360.0,
      "step": 7802
    },
    {
      "entropy": 1.1540457904338837,
      "epoch": 4.109004739336493,
      "grad_norm": 0.28433090448379517,
      "learning_rate": 8.922560571328918e-05,
      "loss": 0.1597246378660202,
      "mean_token_accuracy": 0.9324362426996231,
      "num_tokens": 96253696.0,
      "step": 7803
    },
    {
      "entropy": 1.1672990918159485,
      "epoch": 4.109531332280147,
      "grad_norm": 0.338194340467453,
      "learning_rate": 8.920427021819642e-05,
      "loss": 0.15494990348815918,
      "mean_token_accuracy": 0.936102956533432,
      "num_tokens": 96266032.0,
      "step": 7804
    },
    {
      "entropy": 1.0869658887386322,
      "epoch": 4.110057925223802,
      "grad_norm": 0.3005244731903076,
      "learning_rate": 8.918293595762488e-05,
      "loss": 0.16309787333011627,
      "mean_token_accuracy": 0.9356388449668884,
      "num_tokens": 96278368.0,
      "step": 7805
    },
    {
      "entropy": 1.1280148327350616,
      "epoch": 4.110584518167457,
      "grad_norm": 0.2926557660102844,
      "learning_rate": 8.916160293284115e-05,
      "loss": 0.15952609479427338,
      "mean_token_accuracy": 0.9271687120199203,
      "num_tokens": 96290704.0,
      "step": 7806
    },
    {
      "entropy": 1.12529456615448,
      "epoch": 4.111111111111111,
      "grad_norm": 0.35064613819122314,
      "learning_rate": 8.914027114511163e-05,
      "loss": 0.2010664939880371,
      "mean_token_accuracy": 0.914945438504219,
      "num_tokens": 96303040.0,
      "step": 7807
    },
    {
      "entropy": 1.1171996295452118,
      "epoch": 4.111637704054766,
      "grad_norm": 0.7903205752372742,
      "learning_rate": 8.911894059570259e-05,
      "loss": 0.16090406477451324,
      "mean_token_accuracy": 0.9351358860731125,
      "num_tokens": 96315376.0,
      "step": 7808
    },
    {
      "entropy": 1.0898367166519165,
      "epoch": 4.1121642969984205,
      "grad_norm": 0.2943427562713623,
      "learning_rate": 8.909761128588034e-05,
      "loss": 0.15463635325431824,
      "mean_token_accuracy": 0.9347275495529175,
      "num_tokens": 96327712.0,
      "step": 7809
    },
    {
      "entropy": 1.130621761083603,
      "epoch": 4.112690889942074,
      "grad_norm": 0.3493938744068146,
      "learning_rate": 8.90762832169111e-05,
      "loss": 0.1699773371219635,
      "mean_token_accuracy": 0.9224660992622375,
      "num_tokens": 96340048.0,
      "step": 7810
    },
    {
      "entropy": 1.1240656971931458,
      "epoch": 4.113217482885729,
      "grad_norm": 0.32472723722457886,
      "learning_rate": 8.905495639006096e-05,
      "loss": 0.16446636617183685,
      "mean_token_accuracy": 0.9324799329042435,
      "num_tokens": 96352384.0,
      "step": 7811
    },
    {
      "entropy": 1.0538891851902008,
      "epoch": 4.113744075829384,
      "grad_norm": 0.30771127343177795,
      "learning_rate": 8.903363080659596e-05,
      "loss": 0.1678232103586197,
      "mean_token_accuracy": 0.9295049905776978,
      "num_tokens": 96364720.0,
      "step": 7812
    },
    {
      "entropy": 1.1159376204013824,
      "epoch": 4.114270668773038,
      "grad_norm": 0.2992302179336548,
      "learning_rate": 8.901230646778214e-05,
      "loss": 0.16447745263576508,
      "mean_token_accuracy": 0.9330683052539825,
      "num_tokens": 96377056.0,
      "step": 7813
    },
    {
      "entropy": 1.0891575366258621,
      "epoch": 4.114797261716693,
      "grad_norm": 0.3098483979701996,
      "learning_rate": 8.899098337488533e-05,
      "loss": 0.17213174700737,
      "mean_token_accuracy": 0.925985649228096,
      "num_tokens": 96389392.0,
      "step": 7814
    },
    {
      "entropy": 1.0644939839839935,
      "epoch": 4.115323854660348,
      "grad_norm": 0.28281939029693604,
      "learning_rate": 8.896966152917142e-05,
      "loss": 0.15706969797611237,
      "mean_token_accuracy": 0.9341888576745987,
      "num_tokens": 96401728.0,
      "step": 7815
    },
    {
      "entropy": 1.084628701210022,
      "epoch": 4.115850447604002,
      "grad_norm": 0.27311787009239197,
      "learning_rate": 8.89483409319061e-05,
      "loss": 0.15404914319515228,
      "mean_token_accuracy": 0.9348626583814621,
      "num_tokens": 96414064.0,
      "step": 7816
    },
    {
      "entropy": 1.087325781583786,
      "epoch": 4.116377040547657,
      "grad_norm": 0.3202205300331116,
      "learning_rate": 8.892702158435508e-05,
      "loss": 0.176896333694458,
      "mean_token_accuracy": 0.9233476519584656,
      "num_tokens": 96426400.0,
      "step": 7817
    },
    {
      "entropy": 1.072919636964798,
      "epoch": 4.116903633491312,
      "grad_norm": 0.28495314717292786,
      "learning_rate": 8.8905703487784e-05,
      "loss": 0.16241121292114258,
      "mean_token_accuracy": 0.9331525415182114,
      "num_tokens": 96438736.0,
      "step": 7818
    },
    {
      "entropy": 1.0810973346233368,
      "epoch": 4.117430226434966,
      "grad_norm": 0.3251652419567108,
      "learning_rate": 8.888438664345828e-05,
      "loss": 0.1854064166545868,
      "mean_token_accuracy": 0.921946182847023,
      "num_tokens": 96451072.0,
      "step": 7819
    },
    {
      "entropy": 1.0986481010913849,
      "epoch": 4.1179568193786205,
      "grad_norm": 0.2982044816017151,
      "learning_rate": 8.886307105264354e-05,
      "loss": 0.15907225012779236,
      "mean_token_accuracy": 0.9335672706365585,
      "num_tokens": 96463408.0,
      "step": 7820
    },
    {
      "entropy": 1.1077307760715485,
      "epoch": 4.118483412322275,
      "grad_norm": 0.318979948759079,
      "learning_rate": 8.884175671660501e-05,
      "loss": 0.170145183801651,
      "mean_token_accuracy": 0.9240032136440277,
      "num_tokens": 96475744.0,
      "step": 7821
    },
    {
      "entropy": 1.098257452249527,
      "epoch": 4.119010005265929,
      "grad_norm": 0.2987666130065918,
      "learning_rate": 8.882044363660813e-05,
      "loss": 0.16981391608715057,
      "mean_token_accuracy": 0.928363248705864,
      "num_tokens": 96488080.0,
      "step": 7822
    },
    {
      "entropy": 1.0558017492294312,
      "epoch": 4.119536598209584,
      "grad_norm": 0.28144413232803345,
      "learning_rate": 8.879913181391803e-05,
      "loss": 0.15899887681007385,
      "mean_token_accuracy": 0.932498574256897,
      "num_tokens": 96500416.0,
      "step": 7823
    },
    {
      "entropy": 1.1114909052848816,
      "epoch": 4.120063191153238,
      "grad_norm": 0.32200029492378235,
      "learning_rate": 8.877782124979992e-05,
      "loss": 0.1900215595960617,
      "mean_token_accuracy": 0.9178457707166672,
      "num_tokens": 96512752.0,
      "step": 7824
    },
    {
      "entropy": 1.1083886921405792,
      "epoch": 4.120589784096893,
      "grad_norm": 0.30182939767837524,
      "learning_rate": 8.875651194551886e-05,
      "loss": 0.16824182868003845,
      "mean_token_accuracy": 0.9271001517772675,
      "num_tokens": 96525088.0,
      "step": 7825
    },
    {
      "entropy": 1.1638360023498535,
      "epoch": 4.121116377040548,
      "grad_norm": 0.29805734753608704,
      "learning_rate": 8.873520390233985e-05,
      "loss": 0.1673271805047989,
      "mean_token_accuracy": 0.9269016534090042,
      "num_tokens": 96537424.0,
      "step": 7826
    },
    {
      "entropy": 1.0961701571941376,
      "epoch": 4.121642969984202,
      "grad_norm": 0.2877432107925415,
      "learning_rate": 8.871389712152787e-05,
      "loss": 0.1591220498085022,
      "mean_token_accuracy": 0.9321826994419098,
      "num_tokens": 96549760.0,
      "step": 7827
    },
    {
      "entropy": 1.0818660408258438,
      "epoch": 4.122169562927857,
      "grad_norm": 0.29051029682159424,
      "learning_rate": 8.869259160434776e-05,
      "loss": 0.15846318006515503,
      "mean_token_accuracy": 0.9337704032659531,
      "num_tokens": 96562096.0,
      "step": 7828
    },
    {
      "entropy": 1.0876257717609406,
      "epoch": 4.122696155871512,
      "grad_norm": 0.2964712381362915,
      "learning_rate": 8.867128735206427e-05,
      "loss": 0.17084316909313202,
      "mean_token_accuracy": 0.9285953342914581,
      "num_tokens": 96574432.0,
      "step": 7829
    },
    {
      "entropy": 1.081564873456955,
      "epoch": 4.123222748815166,
      "grad_norm": 0.29430457949638367,
      "learning_rate": 8.864998436594214e-05,
      "loss": 0.16626888513565063,
      "mean_token_accuracy": 0.9285861700773239,
      "num_tokens": 96586768.0,
      "step": 7830
    },
    {
      "entropy": 1.1084083318710327,
      "epoch": 4.1237493417588205,
      "grad_norm": 0.2968306243419647,
      "learning_rate": 8.862868264724601e-05,
      "loss": 0.16054664552211761,
      "mean_token_accuracy": 0.930002897977829,
      "num_tokens": 96599104.0,
      "step": 7831
    },
    {
      "entropy": 1.0822164118289948,
      "epoch": 4.124275934702475,
      "grad_norm": 0.2795105576515198,
      "learning_rate": 8.860738219724045e-05,
      "loss": 0.1478923261165619,
      "mean_token_accuracy": 0.9389977902173996,
      "num_tokens": 96611440.0,
      "step": 7832
    },
    {
      "entropy": 1.1297181248664856,
      "epoch": 4.124802527646129,
      "grad_norm": 0.3377971351146698,
      "learning_rate": 8.858608301718987e-05,
      "loss": 0.17175334692001343,
      "mean_token_accuracy": 0.9285246282815933,
      "num_tokens": 96623776.0,
      "step": 7833
    },
    {
      "entropy": 1.069557100534439,
      "epoch": 4.125329120589784,
      "grad_norm": 0.29777979850769043,
      "learning_rate": 8.856478510835878e-05,
      "loss": 0.1625191867351532,
      "mean_token_accuracy": 0.9311179369688034,
      "num_tokens": 96636112.0,
      "step": 7834
    },
    {
      "entropy": 1.1241136193275452,
      "epoch": 4.125855713533439,
      "grad_norm": 0.30811306834220886,
      "learning_rate": 8.854348847201144e-05,
      "loss": 0.16325482726097107,
      "mean_token_accuracy": 0.9307897835969925,
      "num_tokens": 96648448.0,
      "step": 7835
    },
    {
      "entropy": 1.077067494392395,
      "epoch": 4.126382306477093,
      "grad_norm": 0.28740742802619934,
      "learning_rate": 8.852219310941217e-05,
      "loss": 0.15477372705936432,
      "mean_token_accuracy": 0.938041016459465,
      "num_tokens": 96660784.0,
      "step": 7836
    },
    {
      "entropy": 1.1043914556503296,
      "epoch": 4.126908899420748,
      "grad_norm": 0.3099571466445923,
      "learning_rate": 8.850089902182508e-05,
      "loss": 0.1632595658302307,
      "mean_token_accuracy": 0.9288739413022995,
      "num_tokens": 96673120.0,
      "step": 7837
    },
    {
      "entropy": 1.1408192813396454,
      "epoch": 4.127435492364402,
      "grad_norm": 0.29412955045700073,
      "learning_rate": 8.847960621051433e-05,
      "loss": 0.15991520881652832,
      "mean_token_accuracy": 0.9341659396886826,
      "num_tokens": 96685456.0,
      "step": 7838
    },
    {
      "entropy": 1.0586369335651398,
      "epoch": 4.127962085308057,
      "grad_norm": 0.3056924343109131,
      "learning_rate": 8.84583146767439e-05,
      "loss": 0.16030479967594147,
      "mean_token_accuracy": 0.933611661195755,
      "num_tokens": 96697792.0,
      "step": 7839
    },
    {
      "entropy": 1.0581893771886826,
      "epoch": 4.128488678251712,
      "grad_norm": 0.32126346230506897,
      "learning_rate": 8.843702442177777e-05,
      "loss": 0.1636396050453186,
      "mean_token_accuracy": 0.9323246031999588,
      "num_tokens": 96710128.0,
      "step": 7840
    },
    {
      "entropy": 1.0848739743232727,
      "epoch": 4.129015271195366,
      "grad_norm": 0.3324814736843109,
      "learning_rate": 8.841573544687984e-05,
      "loss": 0.1712619513273239,
      "mean_token_accuracy": 0.928612694144249,
      "num_tokens": 96722464.0,
      "step": 7841
    },
    {
      "entropy": 1.1315793097019196,
      "epoch": 4.1295418641390205,
      "grad_norm": 0.31529390811920166,
      "learning_rate": 8.839444775331389e-05,
      "loss": 0.1689848154783249,
      "mean_token_accuracy": 0.9284860342741013,
      "num_tokens": 96734800.0,
      "step": 7842
    },
    {
      "entropy": 1.1421118676662445,
      "epoch": 4.130068457082675,
      "grad_norm": 0.33357715606689453,
      "learning_rate": 8.837316134234366e-05,
      "loss": 0.17190055549144745,
      "mean_token_accuracy": 0.9273479878902435,
      "num_tokens": 96747136.0,
      "step": 7843
    },
    {
      "entropy": 1.1058302819728851,
      "epoch": 4.130595050026329,
      "grad_norm": 0.33913272619247437,
      "learning_rate": 8.835187621523277e-05,
      "loss": 0.18200349807739258,
      "mean_token_accuracy": 0.9255418479442596,
      "num_tokens": 96759472.0,
      "step": 7844
    },
    {
      "entropy": 1.1072732210159302,
      "epoch": 4.131121642969984,
      "grad_norm": 0.3066416382789612,
      "learning_rate": 8.833059237324483e-05,
      "loss": 0.17035739123821259,
      "mean_token_accuracy": 0.9298305064439774,
      "num_tokens": 96771808.0,
      "step": 7845
    },
    {
      "entropy": 1.1339519917964935,
      "epoch": 4.131648235913639,
      "grad_norm": 0.3012479543685913,
      "learning_rate": 8.830930981764331e-05,
      "loss": 0.15778815746307373,
      "mean_token_accuracy": 0.9348950833082199,
      "num_tokens": 96784144.0,
      "step": 7846
    },
    {
      "entropy": 1.0989419221878052,
      "epoch": 4.132174828857293,
      "grad_norm": 0.30705466866493225,
      "learning_rate": 8.828802854969164e-05,
      "loss": 0.17275996506214142,
      "mean_token_accuracy": 0.9243705123662949,
      "num_tokens": 96796480.0,
      "step": 7847
    },
    {
      "entropy": 1.1299847066402435,
      "epoch": 4.132701421800948,
      "grad_norm": 0.3280251920223236,
      "learning_rate": 8.826674857065318e-05,
      "loss": 0.1808452010154724,
      "mean_token_accuracy": 0.9271404445171356,
      "num_tokens": 96808816.0,
      "step": 7848
    },
    {
      "entropy": 1.1137258112430573,
      "epoch": 4.133228014744603,
      "grad_norm": 0.31740695238113403,
      "learning_rate": 8.824546988179118e-05,
      "loss": 0.16680635511875153,
      "mean_token_accuracy": 0.9240747690200806,
      "num_tokens": 96821152.0,
      "step": 7849
    },
    {
      "entropy": 1.1617943942546844,
      "epoch": 4.133754607688257,
      "grad_norm": 0.34091678261756897,
      "learning_rate": 8.822419248436886e-05,
      "loss": 0.1889715939760208,
      "mean_token_accuracy": 0.9218580722808838,
      "num_tokens": 96833488.0,
      "step": 7850
    },
    {
      "entropy": 1.1270453333854675,
      "epoch": 4.134281200631912,
      "grad_norm": 0.3339605927467346,
      "learning_rate": 8.820291637964931e-05,
      "loss": 0.17558015882968903,
      "mean_token_accuracy": 0.9282533079385757,
      "num_tokens": 96845824.0,
      "step": 7851
    },
    {
      "entropy": 1.1447749137878418,
      "epoch": 4.1348077935755665,
      "grad_norm": 0.31565946340560913,
      "learning_rate": 8.818164156889557e-05,
      "loss": 0.1743800938129425,
      "mean_token_accuracy": 0.9264811277389526,
      "num_tokens": 96858160.0,
      "step": 7852
    },
    {
      "entropy": 1.0944451987743378,
      "epoch": 4.1353343865192205,
      "grad_norm": 0.2948586046695709,
      "learning_rate": 8.816036805337057e-05,
      "loss": 0.1544438749551773,
      "mean_token_accuracy": 0.9375045746564865,
      "num_tokens": 96870496.0,
      "step": 7853
    },
    {
      "entropy": 1.140945941209793,
      "epoch": 4.135860979462875,
      "grad_norm": 0.2922118604183197,
      "learning_rate": 8.813909583433722e-05,
      "loss": 0.15705375373363495,
      "mean_token_accuracy": 0.9344226717948914,
      "num_tokens": 96882832.0,
      "step": 7854
    },
    {
      "entropy": 1.146077036857605,
      "epoch": 4.136387572406529,
      "grad_norm": 0.3004426658153534,
      "learning_rate": 8.81178249130584e-05,
      "loss": 0.15746715664863586,
      "mean_token_accuracy": 0.9333512932062149,
      "num_tokens": 96895168.0,
      "step": 7855
    },
    {
      "entropy": 1.113210380077362,
      "epoch": 4.136914165350184,
      "grad_norm": 0.2916608154773712,
      "learning_rate": 8.809655529079672e-05,
      "loss": 0.1683368980884552,
      "mean_token_accuracy": 0.9285342842340469,
      "num_tokens": 96907504.0,
      "step": 7856
    },
    {
      "entropy": 1.1452086865901947,
      "epoch": 4.137440758293839,
      "grad_norm": 0.33270788192749023,
      "learning_rate": 8.807528696881491e-05,
      "loss": 0.18464891612529755,
      "mean_token_accuracy": 0.9207575619220734,
      "num_tokens": 96919840.0,
      "step": 7857
    },
    {
      "entropy": 1.1652954816818237,
      "epoch": 4.137967351237493,
      "grad_norm": 0.33288833498954773,
      "learning_rate": 8.805401994837552e-05,
      "loss": 0.1774139106273651,
      "mean_token_accuracy": 0.926798090338707,
      "num_tokens": 96932176.0,
      "step": 7858
    },
    {
      "entropy": 1.146183341741562,
      "epoch": 4.138493944181148,
      "grad_norm": 0.287411630153656,
      "learning_rate": 8.803275423074108e-05,
      "loss": 0.1744053065776825,
      "mean_token_accuracy": 0.9268616735935211,
      "num_tokens": 96944512.0,
      "step": 7859
    },
    {
      "entropy": 1.2056533992290497,
      "epoch": 4.139020537124803,
      "grad_norm": 0.310466468334198,
      "learning_rate": 8.801148981717397e-05,
      "loss": 0.17282357811927795,
      "mean_token_accuracy": 0.9263655841350555,
      "num_tokens": 96956848.0,
      "step": 7860
    },
    {
      "entropy": 1.2036525011062622,
      "epoch": 4.139547130068457,
      "grad_norm": 0.32176509499549866,
      "learning_rate": 8.799022670893653e-05,
      "loss": 0.17282113432884216,
      "mean_token_accuracy": 0.9260594546794891,
      "num_tokens": 96969184.0,
      "step": 7861
    },
    {
      "entropy": 1.1906183660030365,
      "epoch": 4.140073723012112,
      "grad_norm": 0.321964830160141,
      "learning_rate": 8.796896490729108e-05,
      "loss": 0.1866556704044342,
      "mean_token_accuracy": 0.9168805927038193,
      "num_tokens": 96981520.0,
      "step": 7862
    },
    {
      "entropy": 1.1591020822525024,
      "epoch": 4.1406003159557665,
      "grad_norm": 0.29262489080429077,
      "learning_rate": 8.794770441349973e-05,
      "loss": 0.1749974638223648,
      "mean_token_accuracy": 0.923712745308876,
      "num_tokens": 96993856.0,
      "step": 7863
    },
    {
      "entropy": 1.169787347316742,
      "epoch": 4.1411269088994205,
      "grad_norm": 0.279094398021698,
      "learning_rate": 8.79264452288247e-05,
      "loss": 0.1488257199525833,
      "mean_token_accuracy": 0.9390291124582291,
      "num_tokens": 97006192.0,
      "step": 7864
    },
    {
      "entropy": 1.192745953798294,
      "epoch": 4.141653501843075,
      "grad_norm": 0.31311383843421936,
      "learning_rate": 8.790518735452791e-05,
      "loss": 0.1840343475341797,
      "mean_token_accuracy": 0.9203387498855591,
      "num_tokens": 97018528.0,
      "step": 7865
    },
    {
      "entropy": 1.2096452116966248,
      "epoch": 4.14218009478673,
      "grad_norm": 0.2992757558822632,
      "learning_rate": 8.788393079187143e-05,
      "loss": 0.18025711178779602,
      "mean_token_accuracy": 0.9227023422718048,
      "num_tokens": 97030864.0,
      "step": 7866
    },
    {
      "entropy": 1.2235386669635773,
      "epoch": 4.142706687730384,
      "grad_norm": 0.3282581567764282,
      "learning_rate": 8.786267554211702e-05,
      "loss": 0.17555806040763855,
      "mean_token_accuracy": 0.9277769923210144,
      "num_tokens": 97043200.0,
      "step": 7867
    },
    {
      "entropy": 1.1785718202590942,
      "epoch": 4.143233280674039,
      "grad_norm": 0.30684641003608704,
      "learning_rate": 8.784142160652653e-05,
      "loss": 0.1687370091676712,
      "mean_token_accuracy": 0.9266917407512665,
      "num_tokens": 97055536.0,
      "step": 7868
    },
    {
      "entropy": 1.1978355944156647,
      "epoch": 4.143759873617694,
      "grad_norm": 0.3126814067363739,
      "learning_rate": 8.782016898636171e-05,
      "loss": 0.1742529571056366,
      "mean_token_accuracy": 0.9298644214868546,
      "num_tokens": 97067872.0,
      "step": 7869
    },
    {
      "entropy": 1.1947324872016907,
      "epoch": 4.144286466561348,
      "grad_norm": 0.30397486686706543,
      "learning_rate": 8.77989176828842e-05,
      "loss": 0.1776277720928192,
      "mean_token_accuracy": 0.9236465692520142,
      "num_tokens": 97080208.0,
      "step": 7870
    },
    {
      "entropy": 1.1997438669204712,
      "epoch": 4.144813059505003,
      "grad_norm": 0.3125491142272949,
      "learning_rate": 8.777766769735553e-05,
      "loss": 0.16377533972263336,
      "mean_token_accuracy": 0.9303515553474426,
      "num_tokens": 97092544.0,
      "step": 7871
    },
    {
      "entropy": 1.1845740675926208,
      "epoch": 4.145339652448657,
      "grad_norm": 0.31364214420318604,
      "learning_rate": 8.775641903103722e-05,
      "loss": 0.1769748330116272,
      "mean_token_accuracy": 0.9256869256496429,
      "num_tokens": 97104880.0,
      "step": 7872
    },
    {
      "entropy": 1.2070526480674744,
      "epoch": 4.145866245392312,
      "grad_norm": 0.33908551931381226,
      "learning_rate": 8.773517168519068e-05,
      "loss": 0.19118264317512512,
      "mean_token_accuracy": 0.9204799234867096,
      "num_tokens": 97117216.0,
      "step": 7873
    },
    {
      "entropy": 1.1500370502471924,
      "epoch": 4.1463928383359665,
      "grad_norm": 0.30359727144241333,
      "learning_rate": 8.771392566107719e-05,
      "loss": 0.15972694754600525,
      "mean_token_accuracy": 0.9310502260923386,
      "num_tokens": 97129552.0,
      "step": 7874
    },
    {
      "entropy": 1.1818935573101044,
      "epoch": 4.1469194312796205,
      "grad_norm": 0.2914751172065735,
      "learning_rate": 8.769268095995804e-05,
      "loss": 0.15774892270565033,
      "mean_token_accuracy": 0.93589748442173,
      "num_tokens": 97141888.0,
      "step": 7875
    },
    {
      "entropy": 1.1127046644687653,
      "epoch": 4.147446024223275,
      "grad_norm": 0.3282979428768158,
      "learning_rate": 8.767143758309441e-05,
      "loss": 0.1835567206144333,
      "mean_token_accuracy": 0.9219317585229874,
      "num_tokens": 97154224.0,
      "step": 7876
    },
    {
      "entropy": 1.1789429485797882,
      "epoch": 4.14797261716693,
      "grad_norm": 0.3184622824192047,
      "learning_rate": 8.765019553174743e-05,
      "loss": 0.16821898519992828,
      "mean_token_accuracy": 0.9285916388034821,
      "num_tokens": 97166560.0,
      "step": 7877
    },
    {
      "entropy": 1.1564103364944458,
      "epoch": 4.148499210110584,
      "grad_norm": 0.3147515058517456,
      "learning_rate": 8.762895480717806e-05,
      "loss": 0.18371067941188812,
      "mean_token_accuracy": 0.9219361543655396,
      "num_tokens": 97178896.0,
      "step": 7878
    },
    {
      "entropy": 1.1647071540355682,
      "epoch": 4.149025803054239,
      "grad_norm": 0.30860114097595215,
      "learning_rate": 8.760771541064723e-05,
      "loss": 0.16761994361877441,
      "mean_token_accuracy": 0.9280609488487244,
      "num_tokens": 97191232.0,
      "step": 7879
    },
    {
      "entropy": 1.1906790435314178,
      "epoch": 4.149552395997894,
      "grad_norm": 0.3416535258293152,
      "learning_rate": 8.758647734341585e-05,
      "loss": 0.17877760529518127,
      "mean_token_accuracy": 0.9286190420389175,
      "num_tokens": 97203568.0,
      "step": 7880
    },
    {
      "entropy": 1.1658837795257568,
      "epoch": 4.150078988941548,
      "grad_norm": 0.31593748927116394,
      "learning_rate": 8.756524060674467e-05,
      "loss": 0.1759740263223648,
      "mean_token_accuracy": 0.9265497177839279,
      "num_tokens": 97215904.0,
      "step": 7881
    },
    {
      "entropy": 1.1537308394908905,
      "epoch": 4.150605581885203,
      "grad_norm": 0.29740700125694275,
      "learning_rate": 8.754400520189434e-05,
      "loss": 0.16418850421905518,
      "mean_token_accuracy": 0.9289576411247253,
      "num_tokens": 97228240.0,
      "step": 7882
    },
    {
      "entropy": 1.1212096512317657,
      "epoch": 4.151132174828858,
      "grad_norm": 0.3116762340068817,
      "learning_rate": 8.752277113012561e-05,
      "loss": 0.18620739877223969,
      "mean_token_accuracy": 0.9213346540927887,
      "num_tokens": 97240576.0,
      "step": 7883
    },
    {
      "entropy": 1.156780868768692,
      "epoch": 4.151658767772512,
      "grad_norm": 0.31119340658187866,
      "learning_rate": 8.750153839269893e-05,
      "loss": 0.17240305244922638,
      "mean_token_accuracy": 0.9259216040372849,
      "num_tokens": 97252912.0,
      "step": 7884
    },
    {
      "entropy": 1.1416524350643158,
      "epoch": 4.1521853607161665,
      "grad_norm": 0.3117210566997528,
      "learning_rate": 8.748030699087479e-05,
      "loss": 0.16234537959098816,
      "mean_token_accuracy": 0.926701694726944,
      "num_tokens": 97265248.0,
      "step": 7885
    },
    {
      "entropy": 1.1640398502349854,
      "epoch": 4.152711953659821,
      "grad_norm": 0.32278892397880554,
      "learning_rate": 8.745907692591355e-05,
      "loss": 0.18029090762138367,
      "mean_token_accuracy": 0.9224131107330322,
      "num_tokens": 97277584.0,
      "step": 7886
    },
    {
      "entropy": 1.1261101365089417,
      "epoch": 4.153238546603475,
      "grad_norm": 0.30563411116600037,
      "learning_rate": 8.743784819907555e-05,
      "loss": 0.1692604422569275,
      "mean_token_accuracy": 0.9304532557725906,
      "num_tokens": 97289920.0,
      "step": 7887
    },
    {
      "entropy": 1.0596873760223389,
      "epoch": 4.15376513954713,
      "grad_norm": 0.2899622321128845,
      "learning_rate": 8.741662081162101e-05,
      "loss": 0.16613377630710602,
      "mean_token_accuracy": 0.9299019277095795,
      "num_tokens": 97302256.0,
      "step": 7888
    },
    {
      "entropy": 1.1265613734722137,
      "epoch": 4.154291732490784,
      "grad_norm": 0.35447025299072266,
      "learning_rate": 8.739539476481002e-05,
      "loss": 0.18643632531166077,
      "mean_token_accuracy": 0.9222086817026138,
      "num_tokens": 97314592.0,
      "step": 7889
    },
    {
      "entropy": 1.1076932549476624,
      "epoch": 4.154818325434439,
      "grad_norm": 0.3150840997695923,
      "learning_rate": 8.737417005990273e-05,
      "loss": 0.18493393063545227,
      "mean_token_accuracy": 0.9213186651468277,
      "num_tokens": 97326928.0,
      "step": 7890
    },
    {
      "entropy": 1.1085457503795624,
      "epoch": 4.155344918378094,
      "grad_norm": 0.313933789730072,
      "learning_rate": 8.735294669815907e-05,
      "loss": 0.17420503497123718,
      "mean_token_accuracy": 0.9284477382898331,
      "num_tokens": 97339264.0,
      "step": 7891
    },
    {
      "entropy": 1.1525159180164337,
      "epoch": 4.155871511321748,
      "grad_norm": 0.3383539915084839,
      "learning_rate": 8.733172468083899e-05,
      "loss": 0.19659945368766785,
      "mean_token_accuracy": 0.9155688285827637,
      "num_tokens": 97351600.0,
      "step": 7892
    },
    {
      "entropy": 1.1956713497638702,
      "epoch": 4.156398104265403,
      "grad_norm": 0.3446492850780487,
      "learning_rate": 8.731050400920228e-05,
      "loss": 0.19122977554798126,
      "mean_token_accuracy": 0.919826090335846,
      "num_tokens": 97363936.0,
      "step": 7893
    },
    {
      "entropy": 1.133418470621109,
      "epoch": 4.156924697209058,
      "grad_norm": 0.3261299431324005,
      "learning_rate": 8.728928468450872e-05,
      "loss": 0.18709659576416016,
      "mean_token_accuracy": 0.9200118184089661,
      "num_tokens": 97376272.0,
      "step": 7894
    },
    {
      "entropy": 1.1456515491008759,
      "epoch": 4.157451290152712,
      "grad_norm": 0.3069509267807007,
      "learning_rate": 8.726806670801795e-05,
      "loss": 0.16389334201812744,
      "mean_token_accuracy": 0.9278152287006378,
      "num_tokens": 97388608.0,
      "step": 7895
    },
    {
      "entropy": 1.1480842530727386,
      "epoch": 4.1579778830963665,
      "grad_norm": 0.2948766350746155,
      "learning_rate": 8.724685008098957e-05,
      "loss": 0.16451896727085114,
      "mean_token_accuracy": 0.9315945357084274,
      "num_tokens": 97400944.0,
      "step": 7896
    },
    {
      "entropy": 1.1332952380180359,
      "epoch": 4.158504476040021,
      "grad_norm": 0.3097462058067322,
      "learning_rate": 8.722563480468307e-05,
      "loss": 0.17155258357524872,
      "mean_token_accuracy": 0.9271485656499863,
      "num_tokens": 97413280.0,
      "step": 7897
    },
    {
      "entropy": 1.1336309015750885,
      "epoch": 4.159031068983675,
      "grad_norm": 0.33266201615333557,
      "learning_rate": 8.720442088035792e-05,
      "loss": 0.18347817659378052,
      "mean_token_accuracy": 0.923959031701088,
      "num_tokens": 97425616.0,
      "step": 7898
    },
    {
      "entropy": 1.171826809644699,
      "epoch": 4.15955766192733,
      "grad_norm": 0.3139927387237549,
      "learning_rate": 8.718320830927342e-05,
      "loss": 0.18313026428222656,
      "mean_token_accuracy": 0.921869158744812,
      "num_tokens": 97437952.0,
      "step": 7899
    },
    {
      "entropy": 1.1280116140842438,
      "epoch": 4.160084254870985,
      "grad_norm": 0.3097572922706604,
      "learning_rate": 8.716199709268888e-05,
      "loss": 0.1742512583732605,
      "mean_token_accuracy": 0.9275535643100739,
      "num_tokens": 97450288.0,
      "step": 7900
    },
    {
      "entropy": 1.176198035478592,
      "epoch": 4.160610847814639,
      "grad_norm": 0.333377867937088,
      "learning_rate": 8.714078723186345e-05,
      "loss": 0.1746947467327118,
      "mean_token_accuracy": 0.922828882932663,
      "num_tokens": 97462624.0,
      "step": 7901
    },
    {
      "entropy": 1.0927571058273315,
      "epoch": 4.161137440758294,
      "grad_norm": 0.28481024503707886,
      "learning_rate": 8.711957872805623e-05,
      "loss": 0.1627894937992096,
      "mean_token_accuracy": 0.9318502992391586,
      "num_tokens": 97474960.0,
      "step": 7902
    },
    {
      "entropy": 1.1829233169555664,
      "epoch": 4.161664033701949,
      "grad_norm": 0.325119286775589,
      "learning_rate": 8.709837158252625e-05,
      "loss": 0.15235716104507446,
      "mean_token_accuracy": 0.9361916035413742,
      "num_tokens": 97487296.0,
      "step": 7903
    },
    {
      "entropy": 1.132764995098114,
      "epoch": 4.162190626645603,
      "grad_norm": 0.2946729063987732,
      "learning_rate": 8.707716579653252e-05,
      "loss": 0.1627139300107956,
      "mean_token_accuracy": 0.9323912262916565,
      "num_tokens": 97499632.0,
      "step": 7904
    },
    {
      "entropy": 1.1086375713348389,
      "epoch": 4.162717219589258,
      "grad_norm": 0.31512293219566345,
      "learning_rate": 8.705596137133383e-05,
      "loss": 0.17438560724258423,
      "mean_token_accuracy": 0.9283198565244675,
      "num_tokens": 97511968.0,
      "step": 7905
    },
    {
      "entropy": 1.1292003095149994,
      "epoch": 4.163243812532912,
      "grad_norm": 0.3263719379901886,
      "learning_rate": 8.703475830818897e-05,
      "loss": 0.18227732181549072,
      "mean_token_accuracy": 0.9306434243917465,
      "num_tokens": 97524304.0,
      "step": 7906
    },
    {
      "entropy": 1.1621695756912231,
      "epoch": 4.1637704054765665,
      "grad_norm": 0.33717191219329834,
      "learning_rate": 8.701355660835668e-05,
      "loss": 0.18825942277908325,
      "mean_token_accuracy": 0.9183655381202698,
      "num_tokens": 97536640.0,
      "step": 7907
    },
    {
      "entropy": 1.1358915865421295,
      "epoch": 4.164296998420221,
      "grad_norm": 0.2987043559551239,
      "learning_rate": 8.699235627309555e-05,
      "loss": 0.16116175055503845,
      "mean_token_accuracy": 0.9334104210138321,
      "num_tokens": 97548976.0,
      "step": 7908
    },
    {
      "entropy": 1.1565150618553162,
      "epoch": 4.164823591363875,
      "grad_norm": 0.3205986022949219,
      "learning_rate": 8.697115730366412e-05,
      "loss": 0.16675369441509247,
      "mean_token_accuracy": 0.9308912009000778,
      "num_tokens": 97561312.0,
      "step": 7909
    },
    {
      "entropy": 1.0812119841575623,
      "epoch": 4.16535018430753,
      "grad_norm": 0.2919435501098633,
      "learning_rate": 8.694995970132085e-05,
      "loss": 0.16248270869255066,
      "mean_token_accuracy": 0.9337244182825089,
      "num_tokens": 97573648.0,
      "step": 7910
    },
    {
      "entropy": 1.13819220662117,
      "epoch": 4.165876777251185,
      "grad_norm": 0.30222269892692566,
      "learning_rate": 8.692876346732414e-05,
      "loss": 0.16139960289001465,
      "mean_token_accuracy": 0.9313693344593048,
      "num_tokens": 97585984.0,
      "step": 7911
    },
    {
      "entropy": 1.1090191304683685,
      "epoch": 4.166403370194839,
      "grad_norm": 0.30965638160705566,
      "learning_rate": 8.690756860293228e-05,
      "loss": 0.17174068093299866,
      "mean_token_accuracy": 0.9258616119623184,
      "num_tokens": 97598320.0,
      "step": 7912
    },
    {
      "entropy": 1.1270416676998138,
      "epoch": 4.166929963138494,
      "grad_norm": 0.3268696963787079,
      "learning_rate": 8.688637510940347e-05,
      "loss": 0.18002545833587646,
      "mean_token_accuracy": 0.9255077242851257,
      "num_tokens": 97610656.0,
      "step": 7913
    },
    {
      "entropy": 1.0708400905132294,
      "epoch": 4.167456556082149,
      "grad_norm": 0.305288702249527,
      "learning_rate": 8.686518298799588e-05,
      "loss": 0.17188933491706848,
      "mean_token_accuracy": 0.929124191403389,
      "num_tokens": 97622992.0,
      "step": 7914
    },
    {
      "entropy": 1.150255173444748,
      "epoch": 4.167983149025803,
      "grad_norm": 0.319894015789032,
      "learning_rate": 8.684399223996753e-05,
      "loss": 0.18005041778087616,
      "mean_token_accuracy": 0.9222705960273743,
      "num_tokens": 97635328.0,
      "step": 7915
    },
    {
      "entropy": 1.1469556391239166,
      "epoch": 4.168509741969458,
      "grad_norm": 0.2839876413345337,
      "learning_rate": 8.682280286657638e-05,
      "loss": 0.15683573484420776,
      "mean_token_accuracy": 0.9359724968671799,
      "num_tokens": 97647664.0,
      "step": 7916
    },
    {
      "entropy": 1.130246102809906,
      "epoch": 4.1690363349131125,
      "grad_norm": 0.2888103723526001,
      "learning_rate": 8.680161486908029e-05,
      "loss": 0.17167188227176666,
      "mean_token_accuracy": 0.9267813563346863,
      "num_tokens": 97660000.0,
      "step": 7917
    },
    {
      "entropy": 1.1806441247463226,
      "epoch": 4.1695629278567665,
      "grad_norm": 0.31467166543006897,
      "learning_rate": 8.678042824873718e-05,
      "loss": 0.16514365375041962,
      "mean_token_accuracy": 0.9340773224830627,
      "num_tokens": 97672336.0,
      "step": 7918
    },
    {
      "entropy": 1.1494162380695343,
      "epoch": 4.170089520800421,
      "grad_norm": 0.29949796199798584,
      "learning_rate": 8.675924300680472e-05,
      "loss": 0.16776269674301147,
      "mean_token_accuracy": 0.9300021529197693,
      "num_tokens": 97684672.0,
      "step": 7919
    },
    {
      "entropy": 1.1326565742492676,
      "epoch": 4.170616113744076,
      "grad_norm": 0.28281745314598083,
      "learning_rate": 8.673805914454052e-05,
      "loss": 0.15195021033287048,
      "mean_token_accuracy": 0.93742336332798,
      "num_tokens": 97697008.0,
      "step": 7920
    },
    {
      "entropy": 1.155946284532547,
      "epoch": 4.17114270668773,
      "grad_norm": 0.3132133483886719,
      "learning_rate": 8.671687666320216e-05,
      "loss": 0.16573591530323029,
      "mean_token_accuracy": 0.9311816990375519,
      "num_tokens": 97709344.0,
      "step": 7921
    },
    {
      "entropy": 1.1932643055915833,
      "epoch": 4.171669299631385,
      "grad_norm": 0.30656367540359497,
      "learning_rate": 8.669569556404715e-05,
      "loss": 0.17846107482910156,
      "mean_token_accuracy": 0.9251795262098312,
      "num_tokens": 97721680.0,
      "step": 7922
    },
    {
      "entropy": 1.1108348965644836,
      "epoch": 4.172195892575039,
      "grad_norm": 0.28596436977386475,
      "learning_rate": 8.667451584833286e-05,
      "loss": 0.16041474044322968,
      "mean_token_accuracy": 0.932331770658493,
      "num_tokens": 97734016.0,
      "step": 7923
    },
    {
      "entropy": 1.1430757641792297,
      "epoch": 4.172722485518694,
      "grad_norm": 0.29431983828544617,
      "learning_rate": 8.665333751731657e-05,
      "loss": 0.16474981606006622,
      "mean_token_accuracy": 0.9339012801647186,
      "num_tokens": 97746352.0,
      "step": 7924
    },
    {
      "entropy": 1.1723743379116058,
      "epoch": 4.173249078462349,
      "grad_norm": 0.2873007655143738,
      "learning_rate": 8.66321605722556e-05,
      "loss": 0.17336371541023254,
      "mean_token_accuracy": 0.9289100468158722,
      "num_tokens": 97758688.0,
      "step": 7925
    },
    {
      "entropy": 1.1681491136550903,
      "epoch": 4.173775671406003,
      "grad_norm": 0.30883681774139404,
      "learning_rate": 8.661098501440705e-05,
      "loss": 0.16132764518260956,
      "mean_token_accuracy": 0.9322926998138428,
      "num_tokens": 97771024.0,
      "step": 7926
    },
    {
      "entropy": 1.1442429721355438,
      "epoch": 4.174302264349658,
      "grad_norm": 0.2862742841243744,
      "learning_rate": 8.6589810845028e-05,
      "loss": 0.1608617901802063,
      "mean_token_accuracy": 0.9310795217752457,
      "num_tokens": 97783360.0,
      "step": 7927
    },
    {
      "entropy": 1.1497341990470886,
      "epoch": 4.1748288572933125,
      "grad_norm": 0.317093163728714,
      "learning_rate": 8.656863806537545e-05,
      "loss": 0.1696247011423111,
      "mean_token_accuracy": 0.928160235285759,
      "num_tokens": 97795696.0,
      "step": 7928
    },
    {
      "entropy": 1.1339037120342255,
      "epoch": 4.1753554502369665,
      "grad_norm": 0.29926052689552307,
      "learning_rate": 8.654746667670625e-05,
      "loss": 0.17137229442596436,
      "mean_token_accuracy": 0.9275287538766861,
      "num_tokens": 97808032.0,
      "step": 7929
    },
    {
      "entropy": 1.12783482670784,
      "epoch": 4.175882043180621,
      "grad_norm": 0.32092663645744324,
      "learning_rate": 8.652629668027731e-05,
      "loss": 0.1702403575181961,
      "mean_token_accuracy": 0.9298358708620071,
      "num_tokens": 97820368.0,
      "step": 7930
    },
    {
      "entropy": 1.07560133934021,
      "epoch": 4.176408636124276,
      "grad_norm": 0.2888548970222473,
      "learning_rate": 8.650512807734528e-05,
      "loss": 0.1581571102142334,
      "mean_token_accuracy": 0.9314680397510529,
      "num_tokens": 97832704.0,
      "step": 7931
    },
    {
      "entropy": 1.1324105262756348,
      "epoch": 4.17693522906793,
      "grad_norm": 0.3102474808692932,
      "learning_rate": 8.648396086916689e-05,
      "loss": 0.16579195857048035,
      "mean_token_accuracy": 0.9326518923044205,
      "num_tokens": 97845040.0,
      "step": 7932
    },
    {
      "entropy": 1.1441057622432709,
      "epoch": 4.177461822011585,
      "grad_norm": 0.3496611714363098,
      "learning_rate": 8.646279505699866e-05,
      "loss": 0.16715586185455322,
      "mean_token_accuracy": 0.9297567307949066,
      "num_tokens": 97857376.0,
      "step": 7933
    },
    {
      "entropy": 1.1016460955142975,
      "epoch": 4.17798841495524,
      "grad_norm": 0.3026537001132965,
      "learning_rate": 8.644163064209712e-05,
      "loss": 0.16574016213417053,
      "mean_token_accuracy": 0.9266877919435501,
      "num_tokens": 97869712.0,
      "step": 7934
    },
    {
      "entropy": 1.0845454335212708,
      "epoch": 4.178515007898894,
      "grad_norm": 0.3035183250904083,
      "learning_rate": 8.642046762571865e-05,
      "loss": 0.1641194224357605,
      "mean_token_accuracy": 0.9290327280759811,
      "num_tokens": 97882048.0,
      "step": 7935
    },
    {
      "entropy": 1.1284317672252655,
      "epoch": 4.179041600842549,
      "grad_norm": 0.32802003622055054,
      "learning_rate": 8.639930600911958e-05,
      "loss": 0.17272982001304626,
      "mean_token_accuracy": 0.9265357404947281,
      "num_tokens": 97894384.0,
      "step": 7936
    },
    {
      "entropy": 1.116320550441742,
      "epoch": 4.179568193786203,
      "grad_norm": 0.30705663561820984,
      "learning_rate": 8.637814579355616e-05,
      "loss": 0.17422252893447876,
      "mean_token_accuracy": 0.9294481426477432,
      "num_tokens": 97906720.0,
      "step": 7937
    },
    {
      "entropy": 1.0654830634593964,
      "epoch": 4.180094786729858,
      "grad_norm": 0.2974829375743866,
      "learning_rate": 8.635698698028454e-05,
      "loss": 0.1572597324848175,
      "mean_token_accuracy": 0.931130439043045,
      "num_tokens": 97919056.0,
      "step": 7938
    },
    {
      "entropy": 1.038515031337738,
      "epoch": 4.1806213796735125,
      "grad_norm": 0.30022743344306946,
      "learning_rate": 8.633582957056078e-05,
      "loss": 0.1798889935016632,
      "mean_token_accuracy": 0.923603281378746,
      "num_tokens": 97931392.0,
      "step": 7939
    },
    {
      "entropy": 1.1388559639453888,
      "epoch": 4.1811479726171665,
      "grad_norm": 0.3145715594291687,
      "learning_rate": 8.631467356564094e-05,
      "loss": 0.179629847407341,
      "mean_token_accuracy": 0.9220999926328659,
      "num_tokens": 97943728.0,
      "step": 7940
    },
    {
      "entropy": 1.0842942893505096,
      "epoch": 4.181674565560821,
      "grad_norm": 0.3120422959327698,
      "learning_rate": 8.629351896678087e-05,
      "loss": 0.16004841029644012,
      "mean_token_accuracy": 0.9332457929849625,
      "num_tokens": 97956064.0,
      "step": 7941
    },
    {
      "entropy": 1.117704689502716,
      "epoch": 4.182201158504476,
      "grad_norm": 0.3080293834209442,
      "learning_rate": 8.627236577523638e-05,
      "loss": 0.17224644124507904,
      "mean_token_accuracy": 0.9285424053668976,
      "num_tokens": 97968400.0,
      "step": 7942
    },
    {
      "entropy": 1.092921793460846,
      "epoch": 4.18272775144813,
      "grad_norm": 0.2939336597919464,
      "learning_rate": 8.625121399226325e-05,
      "loss": 0.1657520979642868,
      "mean_token_accuracy": 0.9322685897350311,
      "num_tokens": 97980736.0,
      "step": 7943
    },
    {
      "entropy": 1.1526983082294464,
      "epoch": 4.183254344391785,
      "grad_norm": 0.3342089056968689,
      "learning_rate": 8.623006361911717e-05,
      "loss": 0.16991101205348969,
      "mean_token_accuracy": 0.9307911396026611,
      "num_tokens": 97993072.0,
      "step": 7944
    },
    {
      "entropy": 1.1522125899791718,
      "epoch": 4.18378093733544,
      "grad_norm": 0.32566115260124207,
      "learning_rate": 8.620891465705361e-05,
      "loss": 0.18175408244132996,
      "mean_token_accuracy": 0.9232334196567535,
      "num_tokens": 98005408.0,
      "step": 7945
    },
    {
      "entropy": 1.090649127960205,
      "epoch": 4.184307530279094,
      "grad_norm": 0.29311081767082214,
      "learning_rate": 8.618776710732815e-05,
      "loss": 0.17214401066303253,
      "mean_token_accuracy": 0.9269599318504333,
      "num_tokens": 98017744.0,
      "step": 7946
    },
    {
      "entropy": 1.063426434993744,
      "epoch": 4.184834123222749,
      "grad_norm": 0.3156338334083557,
      "learning_rate": 8.616662097119616e-05,
      "loss": 0.1816682517528534,
      "mean_token_accuracy": 0.9186233878135681,
      "num_tokens": 98030080.0,
      "step": 7947
    },
    {
      "entropy": 1.161031723022461,
      "epoch": 4.185360716166404,
      "grad_norm": 0.32171863317489624,
      "learning_rate": 8.614547624991298e-05,
      "loss": 0.1648269146680832,
      "mean_token_accuracy": 0.9334552884101868,
      "num_tokens": 98042416.0,
      "step": 7948
    },
    {
      "entropy": 1.1177985966205597,
      "epoch": 4.185887309110058,
      "grad_norm": 0.2832168936729431,
      "learning_rate": 8.612433294473388e-05,
      "loss": 0.15666519105434418,
      "mean_token_accuracy": 0.9333730190992355,
      "num_tokens": 98054752.0,
      "step": 7949
    },
    {
      "entropy": 1.1391778588294983,
      "epoch": 4.1864139020537126,
      "grad_norm": 0.3128279149532318,
      "learning_rate": 8.610319105691391e-05,
      "loss": 0.18297505378723145,
      "mean_token_accuracy": 0.9255256950855255,
      "num_tokens": 98067088.0,
      "step": 7950
    },
    {
      "entropy": 1.1098828315734863,
      "epoch": 4.186940494997367,
      "grad_norm": 0.2972641587257385,
      "learning_rate": 8.608205058770824e-05,
      "loss": 0.17479528486728668,
      "mean_token_accuracy": 0.9239056408405304,
      "num_tokens": 98079424.0,
      "step": 7951
    },
    {
      "entropy": 1.090119868516922,
      "epoch": 4.187467087941021,
      "grad_norm": 0.2823604941368103,
      "learning_rate": 8.606091153837177e-05,
      "loss": 0.17192570865154266,
      "mean_token_accuracy": 0.926673173904419,
      "num_tokens": 98091760.0,
      "step": 7952
    },
    {
      "entropy": 1.1354528367519379,
      "epoch": 4.187993680884676,
      "grad_norm": 0.3025682270526886,
      "learning_rate": 8.603977391015952e-05,
      "loss": 0.1576496660709381,
      "mean_token_accuracy": 0.9324574172496796,
      "num_tokens": 98104096.0,
      "step": 7953
    },
    {
      "entropy": 1.0947018563747406,
      "epoch": 4.188520273828331,
      "grad_norm": 0.29626360535621643,
      "learning_rate": 8.601863770432621e-05,
      "loss": 0.15951457619667053,
      "mean_token_accuracy": 0.9352555125951767,
      "num_tokens": 98116432.0,
      "step": 7954
    },
    {
      "entropy": 1.0987676084041595,
      "epoch": 4.189046866771985,
      "grad_norm": 0.31504783034324646,
      "learning_rate": 8.599750292212663e-05,
      "loss": 0.1648494005203247,
      "mean_token_accuracy": 0.9316914975643158,
      "num_tokens": 98128768.0,
      "step": 7955
    },
    {
      "entropy": 1.1356965005397797,
      "epoch": 4.18957345971564,
      "grad_norm": 0.33778253197669983,
      "learning_rate": 8.597636956481539e-05,
      "loss": 0.1846567839384079,
      "mean_token_accuracy": 0.9174581170082092,
      "num_tokens": 98141104.0,
      "step": 7956
    },
    {
      "entropy": 1.1384792923927307,
      "epoch": 4.190100052659294,
      "grad_norm": 0.31808388233184814,
      "learning_rate": 8.595523763364705e-05,
      "loss": 0.17047390341758728,
      "mean_token_accuracy": 0.9234932065010071,
      "num_tokens": 98153440.0,
      "step": 7957
    },
    {
      "entropy": 1.1021815836429596,
      "epoch": 4.190626645602949,
      "grad_norm": 0.30722370743751526,
      "learning_rate": 8.593410712987613e-05,
      "loss": 0.1636321246623993,
      "mean_token_accuracy": 0.9300334602594376,
      "num_tokens": 98165776.0,
      "step": 7958
    },
    {
      "entropy": 1.11932834982872,
      "epoch": 4.191153238546604,
      "grad_norm": 0.2899395823478699,
      "learning_rate": 8.591297805475698e-05,
      "loss": 0.1672748178243637,
      "mean_token_accuracy": 0.9299319982528687,
      "num_tokens": 98178112.0,
      "step": 7959
    },
    {
      "entropy": 1.1545650362968445,
      "epoch": 4.191679831490258,
      "grad_norm": 0.4581027030944824,
      "learning_rate": 8.589185040954397e-05,
      "loss": 0.17158180475234985,
      "mean_token_accuracy": 0.926212728023529,
      "num_tokens": 98190448.0,
      "step": 7960
    },
    {
      "entropy": 1.06837597489357,
      "epoch": 4.192206424433913,
      "grad_norm": 0.2880743443965912,
      "learning_rate": 8.587072419549124e-05,
      "loss": 0.15795296430587769,
      "mean_token_accuracy": 0.932521715760231,
      "num_tokens": 98202784.0,
      "step": 7961
    },
    {
      "entropy": 1.1178251504898071,
      "epoch": 4.192733017377567,
      "grad_norm": 0.32974228262901306,
      "learning_rate": 8.584959941385302e-05,
      "loss": 0.16619572043418884,
      "mean_token_accuracy": 0.9276171326637268,
      "num_tokens": 98215120.0,
      "step": 7962
    },
    {
      "entropy": 1.0960520803928375,
      "epoch": 4.193259610321221,
      "grad_norm": 0.2987158000469208,
      "learning_rate": 8.582847606588331e-05,
      "loss": 0.1672661453485489,
      "mean_token_accuracy": 0.934171587228775,
      "num_tokens": 98227456.0,
      "step": 7963
    },
    {
      "entropy": 1.1045244336128235,
      "epoch": 4.193786203264876,
      "grad_norm": 0.3211537003517151,
      "learning_rate": 8.580735415283611e-05,
      "loss": 0.16704663634300232,
      "mean_token_accuracy": 0.9283624291419983,
      "num_tokens": 98239792.0,
      "step": 7964
    },
    {
      "entropy": 1.110570877790451,
      "epoch": 4.194312796208531,
      "grad_norm": 0.2917163074016571,
      "learning_rate": 8.578623367596528e-05,
      "loss": 0.15598426759243011,
      "mean_token_accuracy": 0.9323853105306625,
      "num_tokens": 98252128.0,
      "step": 7965
    },
    {
      "entropy": 1.1223257184028625,
      "epoch": 4.194839389152185,
      "grad_norm": 0.31256887316703796,
      "learning_rate": 8.576511463652459e-05,
      "loss": 0.17231056094169617,
      "mean_token_accuracy": 0.9279206097126007,
      "num_tokens": 98264464.0,
      "step": 7966
    },
    {
      "entropy": 1.140623539686203,
      "epoch": 4.19536598209584,
      "grad_norm": 0.3077540397644043,
      "learning_rate": 8.574399703576783e-05,
      "loss": 0.16856276988983154,
      "mean_token_accuracy": 0.9312508851289749,
      "num_tokens": 98276800.0,
      "step": 7967
    },
    {
      "entropy": 1.063914641737938,
      "epoch": 4.195892575039495,
      "grad_norm": 0.3021930754184723,
      "learning_rate": 8.57228808749486e-05,
      "loss": 0.1651664525270462,
      "mean_token_accuracy": 0.9296476095914841,
      "num_tokens": 98289136.0,
      "step": 7968
    },
    {
      "entropy": 1.0992920398712158,
      "epoch": 4.196419167983149,
      "grad_norm": 0.3029404282569885,
      "learning_rate": 8.570176615532042e-05,
      "loss": 0.1661408394575119,
      "mean_token_accuracy": 0.9296275675296783,
      "num_tokens": 98301472.0,
      "step": 7969
    },
    {
      "entropy": 1.1151938438415527,
      "epoch": 4.196945760926804,
      "grad_norm": 0.3258514106273651,
      "learning_rate": 8.568065287813676e-05,
      "loss": 0.17900605499744415,
      "mean_token_accuracy": 0.9241693019866943,
      "num_tokens": 98313808.0,
      "step": 7970
    },
    {
      "entropy": 1.0781232714653015,
      "epoch": 4.197472353870458,
      "grad_norm": 0.30975598096847534,
      "learning_rate": 8.565954104465099e-05,
      "loss": 0.1776178479194641,
      "mean_token_accuracy": 0.9261597841978073,
      "num_tokens": 98326144.0,
      "step": 7971
    },
    {
      "entropy": 1.06985804438591,
      "epoch": 4.197998946814113,
      "grad_norm": 0.3054265081882477,
      "learning_rate": 8.563843065611644e-05,
      "loss": 0.16655519604682922,
      "mean_token_accuracy": 0.9298802763223648,
      "num_tokens": 98338480.0,
      "step": 7972
    },
    {
      "entropy": 1.1291494071483612,
      "epoch": 4.198525539757767,
      "grad_norm": 0.34682711958885193,
      "learning_rate": 8.56173217137862e-05,
      "loss": 0.19172894954681396,
      "mean_token_accuracy": 0.9177665114402771,
      "num_tokens": 98350816.0,
      "step": 7973
    },
    {
      "entropy": 1.1112886667251587,
      "epoch": 4.199052132701421,
      "grad_norm": 0.3117178678512573,
      "learning_rate": 8.559621421891351e-05,
      "loss": 0.17443540692329407,
      "mean_token_accuracy": 0.9223020076751709,
      "num_tokens": 98363152.0,
      "step": 7974
    },
    {
      "entropy": 1.0742622911930084,
      "epoch": 4.199578725645076,
      "grad_norm": 0.34515872597694397,
      "learning_rate": 8.557510817275132e-05,
      "loss": 0.19704975187778473,
      "mean_token_accuracy": 0.9175226539373398,
      "num_tokens": 98375488.0,
      "step": 7975
    },
    {
      "entropy": 1.1295852661132812,
      "epoch": 4.200105318588731,
      "grad_norm": 0.36591923236846924,
      "learning_rate": 8.555400357655263e-05,
      "loss": 0.18784211575984955,
      "mean_token_accuracy": 0.9206234365701675,
      "num_tokens": 98387824.0,
      "step": 7976
    },
    {
      "entropy": 1.0534974783658981,
      "epoch": 4.200631911532385,
      "grad_norm": 0.3111979365348816,
      "learning_rate": 8.553290043157023e-05,
      "loss": 0.16622695326805115,
      "mean_token_accuracy": 0.929931178689003,
      "num_tokens": 98400160.0,
      "step": 7977
    },
    {
      "entropy": 1.115527868270874,
      "epoch": 4.20115850447604,
      "grad_norm": 0.3002520799636841,
      "learning_rate": 8.551179873905695e-05,
      "loss": 0.15955239534378052,
      "mean_token_accuracy": 0.9324867129325867,
      "num_tokens": 98412496.0,
      "step": 7978
    },
    {
      "entropy": 1.0896766781806946,
      "epoch": 4.201685097419695,
      "grad_norm": 0.2898309826850891,
      "learning_rate": 8.549069850026548e-05,
      "loss": 0.16048483550548553,
      "mean_token_accuracy": 0.9341316372156143,
      "num_tokens": 98424832.0,
      "step": 7979
    },
    {
      "entropy": 1.0702438354492188,
      "epoch": 4.202211690363349,
      "grad_norm": 0.29576775431632996,
      "learning_rate": 8.54695997164483e-05,
      "loss": 0.17099790275096893,
      "mean_token_accuracy": 0.9304496645927429,
      "num_tokens": 98437168.0,
      "step": 7980
    },
    {
      "entropy": 1.1288506090641022,
      "epoch": 4.202738283307004,
      "grad_norm": 0.30057501792907715,
      "learning_rate": 8.544850238885807e-05,
      "loss": 0.1682431399822235,
      "mean_token_accuracy": 0.9278569519519806,
      "num_tokens": 98449504.0,
      "step": 7981
    },
    {
      "entropy": 1.069754958152771,
      "epoch": 4.203264876250659,
      "grad_norm": 0.2968423366546631,
      "learning_rate": 8.542740651874714e-05,
      "loss": 0.15609124302864075,
      "mean_token_accuracy": 0.9299767464399338,
      "num_tokens": 98461840.0,
      "step": 7982
    },
    {
      "entropy": 1.1045666337013245,
      "epoch": 4.203791469194313,
      "grad_norm": 0.31645500659942627,
      "learning_rate": 8.540631210736792e-05,
      "loss": 0.1641203612089157,
      "mean_token_accuracy": 0.9307035803794861,
      "num_tokens": 98474176.0,
      "step": 7983
    },
    {
      "entropy": 1.0776027143001556,
      "epoch": 4.204318062137967,
      "grad_norm": 0.3043416142463684,
      "learning_rate": 8.538521915597255e-05,
      "loss": 0.17317499220371246,
      "mean_token_accuracy": 0.92805016040802,
      "num_tokens": 98486512.0,
      "step": 7984
    },
    {
      "entropy": 1.0777670294046402,
      "epoch": 4.204844655081622,
      "grad_norm": 0.3128795623779297,
      "learning_rate": 8.536412766581326e-05,
      "loss": 0.17743663489818573,
      "mean_token_accuracy": 0.9234709143638611,
      "num_tokens": 98498848.0,
      "step": 7985
    },
    {
      "entropy": 1.1268915832042694,
      "epoch": 4.205371248025276,
      "grad_norm": 0.3350537419319153,
      "learning_rate": 8.534303763814213e-05,
      "loss": 0.18315504491329193,
      "mean_token_accuracy": 0.9224580526351929,
      "num_tokens": 98511184.0,
      "step": 7986
    },
    {
      "entropy": 1.0878113806247711,
      "epoch": 4.205897840968931,
      "grad_norm": 0.30929380655288696,
      "learning_rate": 8.532194907421111e-05,
      "loss": 0.1497439742088318,
      "mean_token_accuracy": 0.9385425448417664,
      "num_tokens": 98523520.0,
      "step": 7987
    },
    {
      "entropy": 1.1011975407600403,
      "epoch": 4.206424433912586,
      "grad_norm": 0.3478199541568756,
      "learning_rate": 8.530086197527218e-05,
      "loss": 0.17903770506381989,
      "mean_token_accuracy": 0.9228616505861282,
      "num_tokens": 98535856.0,
      "step": 7988
    },
    {
      "entropy": 1.099312573671341,
      "epoch": 4.20695102685624,
      "grad_norm": 0.30666959285736084,
      "learning_rate": 8.52797763425771e-05,
      "loss": 0.15633097290992737,
      "mean_token_accuracy": 0.9343074262142181,
      "num_tokens": 98548192.0,
      "step": 7989
    },
    {
      "entropy": 1.1367004215717316,
      "epoch": 4.207477619799895,
      "grad_norm": 0.3231365382671356,
      "learning_rate": 8.525869217737765e-05,
      "loss": 0.17650002241134644,
      "mean_token_accuracy": 0.9277526140213013,
      "num_tokens": 98560528.0,
      "step": 7990
    },
    {
      "entropy": 1.1399459838867188,
      "epoch": 4.208004212743549,
      "grad_norm": 0.3330667316913605,
      "learning_rate": 8.52376094809254e-05,
      "loss": 0.16690731048583984,
      "mean_token_accuracy": 0.9290760457515717,
      "num_tokens": 98572864.0,
      "step": 7991
    },
    {
      "entropy": 1.1063806414604187,
      "epoch": 4.208530805687204,
      "grad_norm": 0.3353106677532196,
      "learning_rate": 8.521652825447198e-05,
      "loss": 0.1873796135187149,
      "mean_token_accuracy": 0.9198074191808701,
      "num_tokens": 98585200.0,
      "step": 7992
    },
    {
      "entropy": 1.1440241932868958,
      "epoch": 4.209057398630859,
      "grad_norm": 0.28894269466400146,
      "learning_rate": 8.519544849926881e-05,
      "loss": 0.1548701375722885,
      "mean_token_accuracy": 0.9302964210510254,
      "num_tokens": 98597536.0,
      "step": 7993
    },
    {
      "entropy": 1.0891906917095184,
      "epoch": 4.209583991574513,
      "grad_norm": 0.3314685523509979,
      "learning_rate": 8.517437021656729e-05,
      "loss": 0.18251857161521912,
      "mean_token_accuracy": 0.9253616034984589,
      "num_tokens": 98609872.0,
      "step": 7994
    },
    {
      "entropy": 1.1146202385425568,
      "epoch": 4.210110584518167,
      "grad_norm": 0.3131825029850006,
      "learning_rate": 8.51532934076187e-05,
      "loss": 0.17522290349006653,
      "mean_token_accuracy": 0.924506425857544,
      "num_tokens": 98622208.0,
      "step": 7995
    },
    {
      "entropy": 1.1207568645477295,
      "epoch": 4.210637177461822,
      "grad_norm": 0.3039415180683136,
      "learning_rate": 8.513221807367431e-05,
      "loss": 0.16855962574481964,
      "mean_token_accuracy": 0.9287214279174805,
      "num_tokens": 98634544.0,
      "step": 7996
    },
    {
      "entropy": 1.0943831205368042,
      "epoch": 4.211163770405476,
      "grad_norm": 0.3006591200828552,
      "learning_rate": 8.511114421598513e-05,
      "loss": 0.16578052937984467,
      "mean_token_accuracy": 0.9278803467750549,
      "num_tokens": 98646880.0,
      "step": 7997
    },
    {
      "entropy": 1.1078577935695648,
      "epoch": 4.211690363349131,
      "grad_norm": 0.3214150667190552,
      "learning_rate": 8.509007183580233e-05,
      "loss": 0.17923012375831604,
      "mean_token_accuracy": 0.9264998435974121,
      "num_tokens": 98659216.0,
      "step": 7998
    },
    {
      "entropy": 1.1035672426223755,
      "epoch": 4.212216956292786,
      "grad_norm": 0.3220691978931427,
      "learning_rate": 8.506900093437676e-05,
      "loss": 0.19403526186943054,
      "mean_token_accuracy": 0.9187631011009216,
      "num_tokens": 98671552.0,
      "step": 7999
    },
    {
      "entropy": 1.165832370519638,
      "epoch": 4.21274354923644,
      "grad_norm": 0.32074758410453796,
      "learning_rate": 8.504793151295924e-05,
      "loss": 0.16181930899620056,
      "mean_token_accuracy": 0.930884599685669,
      "num_tokens": 98683888.0,
      "step": 8000
    },
    {
      "entropy": 1.1062804758548737,
      "epoch": 4.213270142180095,
      "grad_norm": 0.3058002293109894,
      "learning_rate": 8.502686357280065e-05,
      "loss": 0.15885786712169647,
      "mean_token_accuracy": 0.9301334470510483,
      "num_tokens": 98696224.0,
      "step": 8001
    },
    {
      "entropy": 1.086621880531311,
      "epoch": 4.21379673512375,
      "grad_norm": 0.2891179621219635,
      "learning_rate": 8.500579711515157e-05,
      "loss": 0.16997601091861725,
      "mean_token_accuracy": 0.9267712384462357,
      "num_tokens": 98708560.0,
      "step": 8002
    },
    {
      "entropy": 1.072465568780899,
      "epoch": 4.214323328067404,
      "grad_norm": 0.2842663824558258,
      "learning_rate": 8.498473214126265e-05,
      "loss": 0.16176854074001312,
      "mean_token_accuracy": 0.9316033124923706,
      "num_tokens": 98720896.0,
      "step": 8003
    },
    {
      "entropy": 1.1301456093788147,
      "epoch": 4.214849921011059,
      "grad_norm": 0.31222987174987793,
      "learning_rate": 8.49636686523844e-05,
      "loss": 0.18115736544132233,
      "mean_token_accuracy": 0.9255324006080627,
      "num_tokens": 98733232.0,
      "step": 8004
    },
    {
      "entropy": 1.1585054695606232,
      "epoch": 4.215376513954713,
      "grad_norm": 0.29984283447265625,
      "learning_rate": 8.494260664976723e-05,
      "loss": 0.16444486379623413,
      "mean_token_accuracy": 0.9284003525972366,
      "num_tokens": 98745568.0,
      "step": 8005
    },
    {
      "entropy": 1.1902699172496796,
      "epoch": 4.2159031068983674,
      "grad_norm": 0.31386664509773254,
      "learning_rate": 8.492154613466143e-05,
      "loss": 0.18429820239543915,
      "mean_token_accuracy": 0.9183901697397232,
      "num_tokens": 98757904.0,
      "step": 8006
    },
    {
      "entropy": 1.126481145620346,
      "epoch": 4.216429699842022,
      "grad_norm": 0.3164695203304291,
      "learning_rate": 8.490048710831731e-05,
      "loss": 0.16379857063293457,
      "mean_token_accuracy": 0.9324172735214233,
      "num_tokens": 98770240.0,
      "step": 8007
    },
    {
      "entropy": 1.1338652968406677,
      "epoch": 4.216956292785676,
      "grad_norm": 0.2901149392127991,
      "learning_rate": 8.487942957198494e-05,
      "loss": 0.16023913025856018,
      "mean_token_accuracy": 0.9293277263641357,
      "num_tokens": 98782576.0,
      "step": 8008
    },
    {
      "entropy": 1.1322993040084839,
      "epoch": 4.217482885729331,
      "grad_norm": 0.2852417826652527,
      "learning_rate": 8.485837352691442e-05,
      "loss": 0.16661527752876282,
      "mean_token_accuracy": 0.9281364530324936,
      "num_tokens": 98794912.0,
      "step": 8009
    },
    {
      "entropy": 1.1744911968708038,
      "epoch": 4.218009478672986,
      "grad_norm": 0.3385275602340698,
      "learning_rate": 8.483731897435575e-05,
      "loss": 0.18559469282627106,
      "mean_token_accuracy": 0.9225542694330215,
      "num_tokens": 98807248.0,
      "step": 8010
    },
    {
      "entropy": 1.1092327535152435,
      "epoch": 4.21853607161664,
      "grad_norm": 0.3092364966869354,
      "learning_rate": 8.481626591555876e-05,
      "loss": 0.16419336199760437,
      "mean_token_accuracy": 0.9309809654951096,
      "num_tokens": 98819584.0,
      "step": 8011
    },
    {
      "entropy": 1.1482544243335724,
      "epoch": 4.219062664560295,
      "grad_norm": 0.2930870056152344,
      "learning_rate": 8.479521435177333e-05,
      "loss": 0.14658129215240479,
      "mean_token_accuracy": 0.9398646652698517,
      "num_tokens": 98831920.0,
      "step": 8012
    },
    {
      "entropy": 1.1356305181980133,
      "epoch": 4.21958925750395,
      "grad_norm": 0.2949535846710205,
      "learning_rate": 8.477416428424909e-05,
      "loss": 0.15849418938159943,
      "mean_token_accuracy": 0.9352830946445465,
      "num_tokens": 98844256.0,
      "step": 8013
    },
    {
      "entropy": 1.110932320356369,
      "epoch": 4.220115850447604,
      "grad_norm": 0.3075162470340729,
      "learning_rate": 8.47531157142357e-05,
      "loss": 0.1856692135334015,
      "mean_token_accuracy": 0.9179325848817825,
      "num_tokens": 98856592.0,
      "step": 8014
    },
    {
      "entropy": 1.0929087102413177,
      "epoch": 4.220642443391259,
      "grad_norm": 0.28426045179367065,
      "learning_rate": 8.473206864298267e-05,
      "loss": 0.15878359973430634,
      "mean_token_accuracy": 0.9336794316768646,
      "num_tokens": 98868928.0,
      "step": 8015
    },
    {
      "entropy": 1.1330317258834839,
      "epoch": 4.2211690363349135,
      "grad_norm": 0.3047355115413666,
      "learning_rate": 8.471102307173943e-05,
      "loss": 0.16607818007469177,
      "mean_token_accuracy": 0.9291487038135529,
      "num_tokens": 98881264.0,
      "step": 8016
    },
    {
      "entropy": 1.124397486448288,
      "epoch": 4.2216956292785675,
      "grad_norm": 0.3267490565776825,
      "learning_rate": 8.468997900175537e-05,
      "loss": 0.17738410830497742,
      "mean_token_accuracy": 0.9273585677146912,
      "num_tokens": 98893600.0,
      "step": 8017
    },
    {
      "entropy": 1.083697721362114,
      "epoch": 4.222222222222222,
      "grad_norm": 0.3247063457965851,
      "learning_rate": 8.466893643427975e-05,
      "loss": 0.17643986642360687,
      "mean_token_accuracy": 0.923137441277504,
      "num_tokens": 98905936.0,
      "step": 8018
    },
    {
      "entropy": 1.0832822024822235,
      "epoch": 4.222748815165877,
      "grad_norm": 0.2928093671798706,
      "learning_rate": 8.464789537056173e-05,
      "loss": 0.1664019078016281,
      "mean_token_accuracy": 0.9295320510864258,
      "num_tokens": 98918272.0,
      "step": 8019
    },
    {
      "entropy": 1.0755034387111664,
      "epoch": 4.223275408109531,
      "grad_norm": 0.3058106303215027,
      "learning_rate": 8.462685581185041e-05,
      "loss": 0.1633222997188568,
      "mean_token_accuracy": 0.9281518012285233,
      "num_tokens": 98930608.0,
      "step": 8020
    },
    {
      "entropy": 1.1339251101016998,
      "epoch": 4.223802001053186,
      "grad_norm": 0.35978588461875916,
      "learning_rate": 8.460581775939475e-05,
      "loss": 0.18212288618087769,
      "mean_token_accuracy": 0.9245026856660843,
      "num_tokens": 98942944.0,
      "step": 8021
    },
    {
      "entropy": 1.0699157416820526,
      "epoch": 4.22432859399684,
      "grad_norm": 0.30391931533813477,
      "learning_rate": 8.458478121444371e-05,
      "loss": 0.1710571050643921,
      "mean_token_accuracy": 0.9311480075120926,
      "num_tokens": 98955280.0,
      "step": 8022
    },
    {
      "entropy": 1.1171269714832306,
      "epoch": 4.224855186940495,
      "grad_norm": 0.32767462730407715,
      "learning_rate": 8.456374617824605e-05,
      "loss": 0.1657298356294632,
      "mean_token_accuracy": 0.9313566833734512,
      "num_tokens": 98967616.0,
      "step": 8023
    },
    {
      "entropy": 1.0930550694465637,
      "epoch": 4.22538177988415,
      "grad_norm": 0.31651630997657776,
      "learning_rate": 8.454271265205054e-05,
      "loss": 0.15753570199012756,
      "mean_token_accuracy": 0.9354836642742157,
      "num_tokens": 98979952.0,
      "step": 8024
    },
    {
      "entropy": 1.0365755259990692,
      "epoch": 4.225908372827804,
      "grad_norm": 0.2817818820476532,
      "learning_rate": 8.45216806371058e-05,
      "loss": 0.1531212478876114,
      "mean_token_accuracy": 0.9320075809955597,
      "num_tokens": 98992288.0,
      "step": 8025
    },
    {
      "entropy": 1.0220824927091599,
      "epoch": 4.226434965771459,
      "grad_norm": 0.3157874345779419,
      "learning_rate": 8.450065013466038e-05,
      "loss": 0.16431528329849243,
      "mean_token_accuracy": 0.9330697357654572,
      "num_tokens": 99004624.0,
      "step": 8026
    },
    {
      "entropy": 1.069034457206726,
      "epoch": 4.2269615587151135,
      "grad_norm": 0.3280310034751892,
      "learning_rate": 8.447962114596278e-05,
      "loss": 0.16932621598243713,
      "mean_token_accuracy": 0.9302221536636353,
      "num_tokens": 99016960.0,
      "step": 8027
    },
    {
      "entropy": 1.084026724100113,
      "epoch": 4.2274881516587675,
      "grad_norm": 0.34443116188049316,
      "learning_rate": 8.445859367226128e-05,
      "loss": 0.17805546522140503,
      "mean_token_accuracy": 0.9209295958280563,
      "num_tokens": 99029296.0,
      "step": 8028
    },
    {
      "entropy": 1.0402539670467377,
      "epoch": 4.228014744602422,
      "grad_norm": 0.3030589520931244,
      "learning_rate": 8.443756771480422e-05,
      "loss": 0.15634694695472717,
      "mean_token_accuracy": 0.9327075630426407,
      "num_tokens": 99041632.0,
      "step": 8029
    },
    {
      "entropy": 1.0674857199192047,
      "epoch": 4.228541337546077,
      "grad_norm": 0.3273454010486603,
      "learning_rate": 8.441654327483975e-05,
      "loss": 0.16213634610176086,
      "mean_token_accuracy": 0.9327954798936844,
      "num_tokens": 99053968.0,
      "step": 8030
    },
    {
      "entropy": 1.0766152143478394,
      "epoch": 4.229067930489731,
      "grad_norm": 0.32836732268333435,
      "learning_rate": 8.439552035361602e-05,
      "loss": 0.17243799567222595,
      "mean_token_accuracy": 0.9306440651416779,
      "num_tokens": 99066304.0,
      "step": 8031
    },
    {
      "entropy": 1.0273121744394302,
      "epoch": 4.229594523433386,
      "grad_norm": 0.29390573501586914,
      "learning_rate": 8.437449895238103e-05,
      "loss": 0.16573050618171692,
      "mean_token_accuracy": 0.9292726665735245,
      "num_tokens": 99078640.0,
      "step": 8032
    },
    {
      "entropy": 1.0838201642036438,
      "epoch": 4.230121116377041,
      "grad_norm": 0.3103295564651489,
      "learning_rate": 8.435347907238265e-05,
      "loss": 0.16615623235702515,
      "mean_token_accuracy": 0.9277691394090652,
      "num_tokens": 99090976.0,
      "step": 8033
    },
    {
      "entropy": 1.0215496718883514,
      "epoch": 4.230647709320695,
      "grad_norm": 0.3264978528022766,
      "learning_rate": 8.433246071486875e-05,
      "loss": 0.17719753086566925,
      "mean_token_accuracy": 0.924858808517456,
      "num_tokens": 99103312.0,
      "step": 8034
    },
    {
      "entropy": 1.075823962688446,
      "epoch": 4.23117430226435,
      "grad_norm": 0.3413081765174866,
      "learning_rate": 8.431144388108705e-05,
      "loss": 0.16930459439754486,
      "mean_token_accuracy": 0.924733966588974,
      "num_tokens": 99115648.0,
      "step": 8035
    },
    {
      "entropy": 1.0820096135139465,
      "epoch": 4.231700895208005,
      "grad_norm": 0.29625174403190613,
      "learning_rate": 8.429042857228521e-05,
      "loss": 0.15780115127563477,
      "mean_token_accuracy": 0.933433324098587,
      "num_tokens": 99127984.0,
      "step": 8036
    },
    {
      "entropy": 1.0696033835411072,
      "epoch": 4.232227488151659,
      "grad_norm": 0.34695589542388916,
      "learning_rate": 8.426941478971076e-05,
      "loss": 0.18729788064956665,
      "mean_token_accuracy": 0.920822873711586,
      "num_tokens": 99140320.0,
      "step": 8037
    },
    {
      "entropy": 1.0454368889331818,
      "epoch": 4.2327540810953135,
      "grad_norm": 0.30334046483039856,
      "learning_rate": 8.424840253461122e-05,
      "loss": 0.16217342019081116,
      "mean_token_accuracy": 0.9307625442743301,
      "num_tokens": 99152656.0,
      "step": 8038
    },
    {
      "entropy": 1.082186222076416,
      "epoch": 4.2332806740389675,
      "grad_norm": 0.3212319314479828,
      "learning_rate": 8.42273918082339e-05,
      "loss": 0.17911629378795624,
      "mean_token_accuracy": 0.9254394620656967,
      "num_tokens": 99164992.0,
      "step": 8039
    },
    {
      "entropy": 1.0837086737155914,
      "epoch": 4.233807266982622,
      "grad_norm": 0.30776384472846985,
      "learning_rate": 8.420638261182615e-05,
      "loss": 0.1589154452085495,
      "mean_token_accuracy": 0.9329381138086319,
      "num_tokens": 99177328.0,
      "step": 8040
    },
    {
      "entropy": 1.1049553155899048,
      "epoch": 4.234333859926277,
      "grad_norm": 0.33793139457702637,
      "learning_rate": 8.418537494663508e-05,
      "loss": 0.16878412663936615,
      "mean_token_accuracy": 0.9275045096874237,
      "num_tokens": 99189664.0,
      "step": 8041
    },
    {
      "entropy": 1.0484260320663452,
      "epoch": 4.234860452869931,
      "grad_norm": 0.30055221915245056,
      "learning_rate": 8.416436881390788e-05,
      "loss": 0.1704391986131668,
      "mean_token_accuracy": 0.927143931388855,
      "num_tokens": 99202000.0,
      "step": 8042
    },
    {
      "entropy": 1.091255709528923,
      "epoch": 4.235387045813586,
      "grad_norm": 0.28448906540870667,
      "learning_rate": 8.41433642148915e-05,
      "loss": 0.16143342852592468,
      "mean_token_accuracy": 0.9285534769296646,
      "num_tokens": 99214336.0,
      "step": 8043
    },
    {
      "entropy": 1.0518599450588226,
      "epoch": 4.235913638757241,
      "grad_norm": 0.3260214328765869,
      "learning_rate": 8.412236115083285e-05,
      "loss": 0.17668919265270233,
      "mean_token_accuracy": 0.9262640476226807,
      "num_tokens": 99226672.0,
      "step": 8044
    },
    {
      "entropy": 1.1799680888652802,
      "epoch": 4.236440231700895,
      "grad_norm": 0.3557574450969696,
      "learning_rate": 8.410135962297884e-05,
      "loss": 0.1807963252067566,
      "mean_token_accuracy": 0.9215933084487915,
      "num_tokens": 99239008.0,
      "step": 8045
    },
    {
      "entropy": 1.1260333359241486,
      "epoch": 4.23696682464455,
      "grad_norm": 0.30928295850753784,
      "learning_rate": 8.408035963257617e-05,
      "loss": 0.17098373174667358,
      "mean_token_accuracy": 0.930481031537056,
      "num_tokens": 99251344.0,
      "step": 8046
    },
    {
      "entropy": 1.053674966096878,
      "epoch": 4.237493417588205,
      "grad_norm": 0.29614847898483276,
      "learning_rate": 8.405936118087146e-05,
      "loss": 0.16931068897247314,
      "mean_token_accuracy": 0.9240099638700485,
      "num_tokens": 99263680.0,
      "step": 8047
    },
    {
      "entropy": 1.1174336969852448,
      "epoch": 4.238020010531859,
      "grad_norm": 0.2886521816253662,
      "learning_rate": 8.403836426911129e-05,
      "loss": 0.15806442499160767,
      "mean_token_accuracy": 0.9285331666469574,
      "num_tokens": 99276016.0,
      "step": 8048
    },
    {
      "entropy": 1.0821426808834076,
      "epoch": 4.2385466034755135,
      "grad_norm": 0.2885926067829132,
      "learning_rate": 8.40173688985421e-05,
      "loss": 0.16343951225280762,
      "mean_token_accuracy": 0.9332862049341202,
      "num_tokens": 99288352.0,
      "step": 8049
    },
    {
      "entropy": 1.108601987361908,
      "epoch": 4.239073196419168,
      "grad_norm": 0.3193191587924957,
      "learning_rate": 8.399637507041029e-05,
      "loss": 0.18125522136688232,
      "mean_token_accuracy": 0.9237243086099625,
      "num_tokens": 99300688.0,
      "step": 8050
    },
    {
      "entropy": 1.147266149520874,
      "epoch": 4.239599789362822,
      "grad_norm": 0.34496814012527466,
      "learning_rate": 8.397538278596213e-05,
      "loss": 0.17239923775196075,
      "mean_token_accuracy": 0.9295861274003983,
      "num_tokens": 99313024.0,
      "step": 8051
    },
    {
      "entropy": 1.1571033895015717,
      "epoch": 4.240126382306477,
      "grad_norm": 0.3378940224647522,
      "learning_rate": 8.39543920464438e-05,
      "loss": 0.164166659116745,
      "mean_token_accuracy": 0.9328539669513702,
      "num_tokens": 99325360.0,
      "step": 8052
    },
    {
      "entropy": 1.1600407361984253,
      "epoch": 4.240652975250132,
      "grad_norm": 0.30191361904144287,
      "learning_rate": 8.393340285310143e-05,
      "loss": 0.15520229935646057,
      "mean_token_accuracy": 0.9355641454458237,
      "num_tokens": 99337696.0,
      "step": 8053
    },
    {
      "entropy": 1.1555682718753815,
      "epoch": 4.241179568193786,
      "grad_norm": 0.30903249979019165,
      "learning_rate": 8.391241520718098e-05,
      "loss": 0.1642795205116272,
      "mean_token_accuracy": 0.9308182299137115,
      "num_tokens": 99350032.0,
      "step": 8054
    },
    {
      "entropy": 1.1596508026123047,
      "epoch": 4.241706161137441,
      "grad_norm": 0.3145478069782257,
      "learning_rate": 8.389142910992842e-05,
      "loss": 0.1676783561706543,
      "mean_token_accuracy": 0.9271293580532074,
      "num_tokens": 99362368.0,
      "step": 8055
    },
    {
      "entropy": 1.098535805940628,
      "epoch": 4.242232754081095,
      "grad_norm": 0.2904662489891052,
      "learning_rate": 8.387044456258952e-05,
      "loss": 0.15562503039836884,
      "mean_token_accuracy": 0.9329851716756821,
      "num_tokens": 99374704.0,
      "step": 8056
    },
    {
      "entropy": 1.0891389846801758,
      "epoch": 4.24275934702475,
      "grad_norm": 0.29856348037719727,
      "learning_rate": 8.384946156641007e-05,
      "loss": 0.17358572781085968,
      "mean_token_accuracy": 0.9263082295656204,
      "num_tokens": 99387040.0,
      "step": 8057
    },
    {
      "entropy": 1.0838370621204376,
      "epoch": 4.243285939968405,
      "grad_norm": 0.2928997576236725,
      "learning_rate": 8.38284801226356e-05,
      "loss": 0.16847765445709229,
      "mean_token_accuracy": 0.9249438941478729,
      "num_tokens": 99399376.0,
      "step": 8058
    },
    {
      "entropy": 1.1125162243843079,
      "epoch": 4.243812532912059,
      "grad_norm": 0.3397181034088135,
      "learning_rate": 8.380750023251176e-05,
      "loss": 0.191731259226799,
      "mean_token_accuracy": 0.9217439591884613,
      "num_tokens": 99411712.0,
      "step": 8059
    },
    {
      "entropy": 1.1244372129440308,
      "epoch": 4.2443391258557135,
      "grad_norm": 0.3366466462612152,
      "learning_rate": 8.378652189728398e-05,
      "loss": 0.16498181223869324,
      "mean_token_accuracy": 0.9316605627536774,
      "num_tokens": 99424048.0,
      "step": 8060
    },
    {
      "entropy": 1.0650639832019806,
      "epoch": 4.244865718799368,
      "grad_norm": 0.31246834993362427,
      "learning_rate": 8.376554511819763e-05,
      "loss": 0.1643528938293457,
      "mean_token_accuracy": 0.9288981109857559,
      "num_tokens": 99436384.0,
      "step": 8061
    },
    {
      "entropy": 1.1070029437541962,
      "epoch": 4.245392311743022,
      "grad_norm": 0.3298671543598175,
      "learning_rate": 8.37445698964979e-05,
      "loss": 0.18249450623989105,
      "mean_token_accuracy": 0.9227327108383179,
      "num_tokens": 99448720.0,
      "step": 8062
    },
    {
      "entropy": 1.0613518357276917,
      "epoch": 4.245918904686677,
      "grad_norm": 0.29906949400901794,
      "learning_rate": 8.372359623343008e-05,
      "loss": 0.17232637107372284,
      "mean_token_accuracy": 0.9269876629114151,
      "num_tokens": 99461056.0,
      "step": 8063
    },
    {
      "entropy": 1.0885744094848633,
      "epoch": 4.246445497630332,
      "grad_norm": 0.3588619828224182,
      "learning_rate": 8.37026241302392e-05,
      "loss": 0.17881016433238983,
      "mean_token_accuracy": 0.9230170994997025,
      "num_tokens": 99473392.0,
      "step": 8064
    },
    {
      "entropy": 1.0489404201507568,
      "epoch": 4.246972090573986,
      "grad_norm": 0.3277314305305481,
      "learning_rate": 8.36816535881702e-05,
      "loss": 0.17448216676712036,
      "mean_token_accuracy": 0.9284185767173767,
      "num_tokens": 99485728.0,
      "step": 8065
    },
    {
      "entropy": 1.1063831150531769,
      "epoch": 4.247498683517641,
      "grad_norm": 0.31005457043647766,
      "learning_rate": 8.366068460846808e-05,
      "loss": 0.16176143288612366,
      "mean_token_accuracy": 0.9315316379070282,
      "num_tokens": 99498064.0,
      "step": 8066
    },
    {
      "entropy": 1.0681479573249817,
      "epoch": 4.248025276461296,
      "grad_norm": 0.28904494643211365,
      "learning_rate": 8.363971719237758e-05,
      "loss": 0.1548696756362915,
      "mean_token_accuracy": 0.933369517326355,
      "num_tokens": 99510400.0,
      "step": 8067
    },
    {
      "entropy": 1.0995591580867767,
      "epoch": 4.24855186940495,
      "grad_norm": 0.31316521763801575,
      "learning_rate": 8.361875134114343e-05,
      "loss": 0.17337429523468018,
      "mean_token_accuracy": 0.9250330924987793,
      "num_tokens": 99522736.0,
      "step": 8068
    },
    {
      "entropy": 1.0685006529092789,
      "epoch": 4.249078462348605,
      "grad_norm": 0.2747979760169983,
      "learning_rate": 8.359778705601027e-05,
      "loss": 0.15470798313617706,
      "mean_token_accuracy": 0.9313899427652359,
      "num_tokens": 99535072.0,
      "step": 8069
    },
    {
      "entropy": 1.0539186894893646,
      "epoch": 4.249605055292259,
      "grad_norm": 0.27866604924201965,
      "learning_rate": 8.357682433822257e-05,
      "loss": 0.17389869689941406,
      "mean_token_accuracy": 0.9213372319936752,
      "num_tokens": 99547408.0,
      "step": 8070
    },
    {
      "entropy": 1.0815083235502243,
      "epoch": 4.2501316482359135,
      "grad_norm": 0.33380910754203796,
      "learning_rate": 8.355586318902484e-05,
      "loss": 0.18013812601566315,
      "mean_token_accuracy": 0.9280896037817001,
      "num_tokens": 99559744.0,
      "step": 8071
    },
    {
      "entropy": 1.066886693239212,
      "epoch": 4.250658241179568,
      "grad_norm": 0.3224329352378845,
      "learning_rate": 8.353490360966137e-05,
      "loss": 0.17354170978069305,
      "mean_token_accuracy": 0.9260164499282837,
      "num_tokens": 99572080.0,
      "step": 8072
    },
    {
      "entropy": 1.1515626013278961,
      "epoch": 4.251184834123222,
      "grad_norm": 0.295924574136734,
      "learning_rate": 8.351394560137641e-05,
      "loss": 0.16896162927150726,
      "mean_token_accuracy": 0.9296974241733551,
      "num_tokens": 99584416.0,
      "step": 8073
    },
    {
      "entropy": 1.1497910618782043,
      "epoch": 4.251711427066877,
      "grad_norm": 0.29935047030448914,
      "learning_rate": 8.349298916541415e-05,
      "loss": 0.1688576340675354,
      "mean_token_accuracy": 0.9272406250238419,
      "num_tokens": 99596752.0,
      "step": 8074
    },
    {
      "entropy": 1.1204953491687775,
      "epoch": 4.252238020010532,
      "grad_norm": 0.33380720019340515,
      "learning_rate": 8.347203430301863e-05,
      "loss": 0.17503707110881805,
      "mean_token_accuracy": 0.9305144399404526,
      "num_tokens": 99609088.0,
      "step": 8075
    },
    {
      "entropy": 1.1706304848194122,
      "epoch": 4.252764612954186,
      "grad_norm": 0.3090486526489258,
      "learning_rate": 8.345108101543382e-05,
      "loss": 0.15380382537841797,
      "mean_token_accuracy": 0.9374998956918716,
      "num_tokens": 99621424.0,
      "step": 8076
    },
    {
      "entropy": 1.07712322473526,
      "epoch": 4.253291205897841,
      "grad_norm": 0.31502121686935425,
      "learning_rate": 8.34301293039036e-05,
      "loss": 0.17999747395515442,
      "mean_token_accuracy": 0.9227079004049301,
      "num_tokens": 99633760.0,
      "step": 8077
    },
    {
      "entropy": 1.0868211090564728,
      "epoch": 4.253817798841496,
      "grad_norm": 0.31252890825271606,
      "learning_rate": 8.340917916967171e-05,
      "loss": 0.17292320728302002,
      "mean_token_accuracy": 0.9276772290468216,
      "num_tokens": 99646096.0,
      "step": 8078
    },
    {
      "entropy": 1.1564848721027374,
      "epoch": 4.25434439178515,
      "grad_norm": 0.32048600912094116,
      "learning_rate": 8.338823061398184e-05,
      "loss": 0.16539078950881958,
      "mean_token_accuracy": 0.9319958090782166,
      "num_tokens": 99658432.0,
      "step": 8079
    },
    {
      "entropy": 1.172350913286209,
      "epoch": 4.254870984728805,
      "grad_norm": 0.3275053799152374,
      "learning_rate": 8.336728363807767e-05,
      "loss": 0.1759481132030487,
      "mean_token_accuracy": 0.9236879795789719,
      "num_tokens": 99670768.0,
      "step": 8080
    },
    {
      "entropy": 1.1566478908061981,
      "epoch": 4.2553975776724595,
      "grad_norm": 0.3351489305496216,
      "learning_rate": 8.334633824320264e-05,
      "loss": 0.18259534239768982,
      "mean_token_accuracy": 0.9200667142868042,
      "num_tokens": 99683104.0,
      "step": 8081
    },
    {
      "entropy": 1.1785503327846527,
      "epoch": 4.2559241706161135,
      "grad_norm": 0.30122020840644836,
      "learning_rate": 8.332539443060014e-05,
      "loss": 0.16306838393211365,
      "mean_token_accuracy": 0.9354003071784973,
      "num_tokens": 99695440.0,
      "step": 8082
    },
    {
      "entropy": 1.1457667648792267,
      "epoch": 4.256450763559768,
      "grad_norm": 0.3123452365398407,
      "learning_rate": 8.33044522015135e-05,
      "loss": 0.1754731982946396,
      "mean_token_accuracy": 0.9264841824769974,
      "num_tokens": 99707776.0,
      "step": 8083
    },
    {
      "entropy": 1.118004322052002,
      "epoch": 4.256977356503423,
      "grad_norm": 0.29696592688560486,
      "learning_rate": 8.328351155718597e-05,
      "loss": 0.16492369771003723,
      "mean_token_accuracy": 0.9289798140525818,
      "num_tokens": 99720112.0,
      "step": 8084
    },
    {
      "entropy": 1.1619296073913574,
      "epoch": 4.257503949447077,
      "grad_norm": 0.3061583936214447,
      "learning_rate": 8.326257249886062e-05,
      "loss": 0.17469191551208496,
      "mean_token_accuracy": 0.9250694811344147,
      "num_tokens": 99732448.0,
      "step": 8085
    },
    {
      "entropy": 1.1556996703147888,
      "epoch": 4.258030542390732,
      "grad_norm": 0.32471799850463867,
      "learning_rate": 8.324163502778048e-05,
      "loss": 0.17089048027992249,
      "mean_token_accuracy": 0.9294106811285019,
      "num_tokens": 99744784.0,
      "step": 8086
    },
    {
      "entropy": 1.0968319475650787,
      "epoch": 4.258557135334387,
      "grad_norm": 0.29185327887535095,
      "learning_rate": 8.322069914518853e-05,
      "loss": 0.1605537384748459,
      "mean_token_accuracy": 0.933579221367836,
      "num_tokens": 99757120.0,
      "step": 8087
    },
    {
      "entropy": 1.1535795331001282,
      "epoch": 4.259083728278041,
      "grad_norm": 0.32085347175598145,
      "learning_rate": 8.31997648523276e-05,
      "loss": 0.16581113636493683,
      "mean_token_accuracy": 0.9338376224040985,
      "num_tokens": 99769456.0,
      "step": 8088
    },
    {
      "entropy": 1.1065370440483093,
      "epoch": 4.259610321221696,
      "grad_norm": 0.3301770091056824,
      "learning_rate": 8.317883215044043e-05,
      "loss": 0.18444573879241943,
      "mean_token_accuracy": 0.9218945354223251,
      "num_tokens": 99781792.0,
      "step": 8089
    },
    {
      "entropy": 1.128524899482727,
      "epoch": 4.26013691416535,
      "grad_norm": 0.32063019275665283,
      "learning_rate": 8.315790104076966e-05,
      "loss": 0.17200259864330292,
      "mean_token_accuracy": 0.9251203238964081,
      "num_tokens": 99794128.0,
      "step": 8090
    },
    {
      "entropy": 1.1268632411956787,
      "epoch": 4.260663507109005,
      "grad_norm": 0.31877580285072327,
      "learning_rate": 8.313697152455786e-05,
      "loss": 0.16964519023895264,
      "mean_token_accuracy": 0.928652435541153,
      "num_tokens": 99806464.0,
      "step": 8091
    },
    {
      "entropy": 1.0946198105812073,
      "epoch": 4.2611901000526595,
      "grad_norm": 0.3191579282283783,
      "learning_rate": 8.31160436030475e-05,
      "loss": 0.15536580979824066,
      "mean_token_accuracy": 0.936287522315979,
      "num_tokens": 99818800.0,
      "step": 8092
    },
    {
      "entropy": 1.1416193544864655,
      "epoch": 4.2617166929963135,
      "grad_norm": 0.32858961820602417,
      "learning_rate": 8.309511727748089e-05,
      "loss": 0.1654105931520462,
      "mean_token_accuracy": 0.927264392375946,
      "num_tokens": 99831136.0,
      "step": 8093
    },
    {
      "entropy": 1.1381915807724,
      "epoch": 4.262243285939968,
      "grad_norm": 0.33304300904273987,
      "learning_rate": 8.307419254910039e-05,
      "loss": 0.17604365944862366,
      "mean_token_accuracy": 0.9258278459310532,
      "num_tokens": 99843472.0,
      "step": 8094
    },
    {
      "entropy": 1.1308756172657013,
      "epoch": 4.262769878883623,
      "grad_norm": 0.31295496225357056,
      "learning_rate": 8.305326941914817e-05,
      "loss": 0.17429137229919434,
      "mean_token_accuracy": 0.9237045347690582,
      "num_tokens": 99855808.0,
      "step": 8095
    },
    {
      "entropy": 1.0989869236946106,
      "epoch": 4.263296471827277,
      "grad_norm": 0.34125518798828125,
      "learning_rate": 8.303234788886625e-05,
      "loss": 0.19118639826774597,
      "mean_token_accuracy": 0.9206728041172028,
      "num_tokens": 99868144.0,
      "step": 8096
    },
    {
      "entropy": 1.1231334209442139,
      "epoch": 4.263823064770932,
      "grad_norm": 0.31131285429000854,
      "learning_rate": 8.301142795949668e-05,
      "loss": 0.16339470446109772,
      "mean_token_accuracy": 0.933625802397728,
      "num_tokens": 99880480.0,
      "step": 8097
    },
    {
      "entropy": 1.0876518487930298,
      "epoch": 4.264349657714587,
      "grad_norm": 0.3088369369506836,
      "learning_rate": 8.299050963228133e-05,
      "loss": 0.17668671905994415,
      "mean_token_accuracy": 0.9236136674880981,
      "num_tokens": 99892816.0,
      "step": 8098
    },
    {
      "entropy": 1.0807808935642242,
      "epoch": 4.264876250658241,
      "grad_norm": 0.3156672418117523,
      "learning_rate": 8.296959290846196e-05,
      "loss": 0.18399551510810852,
      "mean_token_accuracy": 0.9209917485713959,
      "num_tokens": 99905152.0,
      "step": 8099
    },
    {
      "entropy": 1.1325292885303497,
      "epoch": 4.265402843601896,
      "grad_norm": 0.3002985715866089,
      "learning_rate": 8.294867778928036e-05,
      "loss": 0.1714736372232437,
      "mean_token_accuracy": 0.9298235177993774,
      "num_tokens": 99917488.0,
      "step": 8100
    },
    {
      "entropy": 1.1102986931800842,
      "epoch": 4.265929436545551,
      "grad_norm": 0.31801483035087585,
      "learning_rate": 8.292776427597806e-05,
      "loss": 0.17188766598701477,
      "mean_token_accuracy": 0.925754576921463,
      "num_tokens": 99929824.0,
      "step": 8101
    },
    {
      "entropy": 1.104709655046463,
      "epoch": 4.266456029489205,
      "grad_norm": 0.34049540758132935,
      "learning_rate": 8.29068523697966e-05,
      "loss": 0.17808237671852112,
      "mean_token_accuracy": 0.9203847646713257,
      "num_tokens": 99942160.0,
      "step": 8102
    },
    {
      "entropy": 1.0856394171714783,
      "epoch": 4.2669826224328595,
      "grad_norm": 0.30023178458213806,
      "learning_rate": 8.288594207197744e-05,
      "loss": 0.16197536885738373,
      "mean_token_accuracy": 0.9287407249212265,
      "num_tokens": 99954496.0,
      "step": 8103
    },
    {
      "entropy": 1.1135685741901398,
      "epoch": 4.2675092153765135,
      "grad_norm": 0.3104518949985504,
      "learning_rate": 8.286503338376186e-05,
      "loss": 0.1633913815021515,
      "mean_token_accuracy": 0.9310812950134277,
      "num_tokens": 99966832.0,
      "step": 8104
    },
    {
      "entropy": 1.1092984676361084,
      "epoch": 4.268035808320168,
      "grad_norm": 0.31064674258232117,
      "learning_rate": 8.28441263063911e-05,
      "loss": 0.17475390434265137,
      "mean_token_accuracy": 0.9235182404518127,
      "num_tokens": 99979168.0,
      "step": 8105
    },
    {
      "entropy": 1.1336276531219482,
      "epoch": 4.268562401263823,
      "grad_norm": 0.3191753923892975,
      "learning_rate": 8.282322084110627e-05,
      "loss": 0.17750518023967743,
      "mean_token_accuracy": 0.929021954536438,
      "num_tokens": 99991504.0,
      "step": 8106
    },
    {
      "entropy": 1.0961753129959106,
      "epoch": 4.269088994207477,
      "grad_norm": 0.31571850180625916,
      "learning_rate": 8.280231698914837e-05,
      "loss": 0.17665252089500427,
      "mean_token_accuracy": 0.9285895675420761,
      "num_tokens": 100003840.0,
      "step": 8107
    },
    {
      "entropy": 1.0950091779232025,
      "epoch": 4.269615587151132,
      "grad_norm": 0.27700111269950867,
      "learning_rate": 8.278141475175844e-05,
      "loss": 0.15171709656715393,
      "mean_token_accuracy": 0.936496689915657,
      "num_tokens": 100016176.0,
      "step": 8108
    },
    {
      "entropy": 1.102284163236618,
      "epoch": 4.270142180094787,
      "grad_norm": 0.27984926104545593,
      "learning_rate": 8.276051413017729e-05,
      "loss": 0.15890219807624817,
      "mean_token_accuracy": 0.9340721517801285,
      "num_tokens": 100028512.0,
      "step": 8109
    },
    {
      "entropy": 1.0916558504104614,
      "epoch": 4.270668773038441,
      "grad_norm": 0.32002735137939453,
      "learning_rate": 8.273961512564566e-05,
      "loss": 0.1711985170841217,
      "mean_token_accuracy": 0.9211053550243378,
      "num_tokens": 100040848.0,
      "step": 8110
    },
    {
      "entropy": 1.1369658708572388,
      "epoch": 4.271195365982096,
      "grad_norm": 0.3338232636451721,
      "learning_rate": 8.271871773940415e-05,
      "loss": 0.16055558621883392,
      "mean_token_accuracy": 0.9368612915277481,
      "num_tokens": 100053184.0,
      "step": 8111
    },
    {
      "entropy": 1.1092744767665863,
      "epoch": 4.271721958925751,
      "grad_norm": 0.330318421125412,
      "learning_rate": 8.26978219726934e-05,
      "loss": 0.1709112673997879,
      "mean_token_accuracy": 0.9239360392093658,
      "num_tokens": 100065520.0,
      "step": 8112
    },
    {
      "entropy": 1.1184650361537933,
      "epoch": 4.272248551869405,
      "grad_norm": 0.32604461908340454,
      "learning_rate": 8.267692782675383e-05,
      "loss": 0.18035155534744263,
      "mean_token_accuracy": 0.9268295615911484,
      "num_tokens": 100077856.0,
      "step": 8113
    },
    {
      "entropy": 1.074916884303093,
      "epoch": 4.2727751448130595,
      "grad_norm": 0.3004838228225708,
      "learning_rate": 8.265603530282576e-05,
      "loss": 0.16890446841716766,
      "mean_token_accuracy": 0.9294091016054153,
      "num_tokens": 100090192.0,
      "step": 8114
    },
    {
      "entropy": 1.0736864805221558,
      "epoch": 4.273301737756714,
      "grad_norm": 0.336443156003952,
      "learning_rate": 8.263514440214952e-05,
      "loss": 0.1787278950214386,
      "mean_token_accuracy": 0.9264860451221466,
      "num_tokens": 100102528.0,
      "step": 8115
    },
    {
      "entropy": 1.1153449714183807,
      "epoch": 4.273828330700368,
      "grad_norm": 0.32195696234703064,
      "learning_rate": 8.261425512596525e-05,
      "loss": 0.17047405242919922,
      "mean_token_accuracy": 0.9277778267860413,
      "num_tokens": 100114864.0,
      "step": 8116
    },
    {
      "entropy": 1.0776326656341553,
      "epoch": 4.274354923644023,
      "grad_norm": 0.3388669192790985,
      "learning_rate": 8.259336747551307e-05,
      "loss": 0.17405882477760315,
      "mean_token_accuracy": 0.9185441732406616,
      "num_tokens": 100127200.0,
      "step": 8117
    },
    {
      "entropy": 1.1092575788497925,
      "epoch": 4.274881516587678,
      "grad_norm": 0.3515068292617798,
      "learning_rate": 8.257248145203291e-05,
      "loss": 0.18974429368972778,
      "mean_token_accuracy": 0.918947622179985,
      "num_tokens": 100139536.0,
      "step": 8118
    },
    {
      "entropy": 1.1126601994037628,
      "epoch": 4.275408109531332,
      "grad_norm": 0.33457013964653015,
      "learning_rate": 8.255159705676467e-05,
      "loss": 0.17667625844478607,
      "mean_token_accuracy": 0.9245069324970245,
      "num_tokens": 100151872.0,
      "step": 8119
    },
    {
      "entropy": 1.0818258821964264,
      "epoch": 4.275934702474987,
      "grad_norm": 0.30092358589172363,
      "learning_rate": 8.253071429094814e-05,
      "loss": 0.17009970545768738,
      "mean_token_accuracy": 0.9274181872606277,
      "num_tokens": 100164208.0,
      "step": 8120
    },
    {
      "entropy": 1.093058466911316,
      "epoch": 4.276461295418642,
      "grad_norm": 0.3126024603843689,
      "learning_rate": 8.250983315582297e-05,
      "loss": 0.16962504386901855,
      "mean_token_accuracy": 0.9259705096483231,
      "num_tokens": 100176544.0,
      "step": 8121
    },
    {
      "entropy": 1.0457747280597687,
      "epoch": 4.276987888362296,
      "grad_norm": 0.3177169859409332,
      "learning_rate": 8.24889536526288e-05,
      "loss": 0.18231360614299774,
      "mean_token_accuracy": 0.9204312413930893,
      "num_tokens": 100188880.0,
      "step": 8122
    },
    {
      "entropy": 1.089835911989212,
      "epoch": 4.277514481305951,
      "grad_norm": 0.28656700253486633,
      "learning_rate": 8.246807578260511e-05,
      "loss": 0.15455403923988342,
      "mean_token_accuracy": 0.9351130276918411,
      "num_tokens": 100201216.0,
      "step": 8123
    },
    {
      "entropy": 1.0800141394138336,
      "epoch": 4.278041074249605,
      "grad_norm": 0.31688860058784485,
      "learning_rate": 8.24471995469913e-05,
      "loss": 0.17136967182159424,
      "mean_token_accuracy": 0.9254559725522995,
      "num_tokens": 100213552.0,
      "step": 8124
    },
    {
      "entropy": 1.0572218894958496,
      "epoch": 4.2785676671932595,
      "grad_norm": 0.2976747155189514,
      "learning_rate": 8.242632494702664e-05,
      "loss": 0.16170823574066162,
      "mean_token_accuracy": 0.9315948784351349,
      "num_tokens": 100225888.0,
      "step": 8125
    },
    {
      "entropy": 1.0977745354175568,
      "epoch": 4.279094260136914,
      "grad_norm": 0.31714725494384766,
      "learning_rate": 8.240545198395038e-05,
      "loss": 0.1668914258480072,
      "mean_token_accuracy": 0.9267875403165817,
      "num_tokens": 100238224.0,
      "step": 8126
    },
    {
      "entropy": 1.0821348428726196,
      "epoch": 4.279620853080568,
      "grad_norm": 0.3463470935821533,
      "learning_rate": 8.238458065900158e-05,
      "loss": 0.18148234486579895,
      "mean_token_accuracy": 0.9224915206432343,
      "num_tokens": 100250560.0,
      "step": 8127
    },
    {
      "entropy": 1.0404681116342545,
      "epoch": 4.280147446024223,
      "grad_norm": 0.3089762032032013,
      "learning_rate": 8.236371097341925e-05,
      "loss": 0.1620589643716812,
      "mean_token_accuracy": 0.9293773472309113,
      "num_tokens": 100262896.0,
      "step": 8128
    },
    {
      "entropy": 1.0546789467334747,
      "epoch": 4.280674038967878,
      "grad_norm": 0.311089426279068,
      "learning_rate": 8.234284292844237e-05,
      "loss": 0.1701953411102295,
      "mean_token_accuracy": 0.9255341589450836,
      "num_tokens": 100275232.0,
      "step": 8129
    },
    {
      "entropy": 1.0441174358129501,
      "epoch": 4.281200631911532,
      "grad_norm": 0.297513484954834,
      "learning_rate": 8.232197652530971e-05,
      "loss": 0.160035640001297,
      "mean_token_accuracy": 0.9331898093223572,
      "num_tokens": 100287568.0,
      "step": 8130
    },
    {
      "entropy": 1.047688141465187,
      "epoch": 4.281727224855187,
      "grad_norm": 0.3260231614112854,
      "learning_rate": 8.230111176525994e-05,
      "loss": 0.17879530787467957,
      "mean_token_accuracy": 0.9271712005138397,
      "num_tokens": 100299904.0,
      "step": 8131
    },
    {
      "entropy": 1.0368105620145798,
      "epoch": 4.282253817798842,
      "grad_norm": 0.33203935623168945,
      "learning_rate": 8.228024864953177e-05,
      "loss": 0.1704830527305603,
      "mean_token_accuracy": 0.924525648355484,
      "num_tokens": 100312240.0,
      "step": 8132
    },
    {
      "entropy": 1.0623641312122345,
      "epoch": 4.282780410742496,
      "grad_norm": 0.28985536098480225,
      "learning_rate": 8.225938717936363e-05,
      "loss": 0.16252949833869934,
      "mean_token_accuracy": 0.9282232373952866,
      "num_tokens": 100324576.0,
      "step": 8133
    },
    {
      "entropy": 1.0668547749519348,
      "epoch": 4.283307003686151,
      "grad_norm": 0.3509078621864319,
      "learning_rate": 8.223852735599402e-05,
      "loss": 0.18604707717895508,
      "mean_token_accuracy": 0.9223463833332062,
      "num_tokens": 100336912.0,
      "step": 8134
    },
    {
      "entropy": 1.0122692286968231,
      "epoch": 4.283833596629806,
      "grad_norm": 0.33317187428474426,
      "learning_rate": 8.22176691806612e-05,
      "loss": 0.1752258837223053,
      "mean_token_accuracy": 0.9266980290412903,
      "num_tokens": 100349248.0,
      "step": 8135
    },
    {
      "entropy": 1.0725390017032623,
      "epoch": 4.2843601895734595,
      "grad_norm": 0.3354742228984833,
      "learning_rate": 8.219681265460347e-05,
      "loss": 0.17461788654327393,
      "mean_token_accuracy": 0.9295875281095505,
      "num_tokens": 100361584.0,
      "step": 8136
    },
    {
      "entropy": 1.0339740365743637,
      "epoch": 4.284886782517114,
      "grad_norm": 0.33746543526649475,
      "learning_rate": 8.217595777905891e-05,
      "loss": 0.17826615273952484,
      "mean_token_accuracy": 0.9233043342828751,
      "num_tokens": 100373920.0,
      "step": 8137
    },
    {
      "entropy": 1.0048495382070541,
      "epoch": 4.285413375460768,
      "grad_norm": 0.3237973153591156,
      "learning_rate": 8.215510455526558e-05,
      "loss": 0.1711120903491974,
      "mean_token_accuracy": 0.9247938394546509,
      "num_tokens": 100386256.0,
      "step": 8138
    },
    {
      "entropy": 1.0556179583072662,
      "epoch": 4.285939968404423,
      "grad_norm": 0.329245924949646,
      "learning_rate": 8.213425298446138e-05,
      "loss": 0.18409737944602966,
      "mean_token_accuracy": 0.9237063080072403,
      "num_tokens": 100398592.0,
      "step": 8139
    },
    {
      "entropy": 0.998849630355835,
      "epoch": 4.286466561348078,
      "grad_norm": 0.2911140024662018,
      "learning_rate": 8.21134030678842e-05,
      "loss": 0.166805699467659,
      "mean_token_accuracy": 0.9282380193471909,
      "num_tokens": 100410928.0,
      "step": 8140
    },
    {
      "entropy": 1.0545998513698578,
      "epoch": 4.286993154291732,
      "grad_norm": 0.31316500902175903,
      "learning_rate": 8.209255480677172e-05,
      "loss": 0.16142991185188293,
      "mean_token_accuracy": 0.9323518574237823,
      "num_tokens": 100423264.0,
      "step": 8141
    },
    {
      "entropy": 1.0759258568286896,
      "epoch": 4.287519747235387,
      "grad_norm": 0.3297795057296753,
      "learning_rate": 8.207170820236156e-05,
      "loss": 0.16118834912776947,
      "mean_token_accuracy": 0.930723175406456,
      "num_tokens": 100435600.0,
      "step": 8142
    },
    {
      "entropy": 1.059980496764183,
      "epoch": 4.288046340179042,
      "grad_norm": 0.33654120564460754,
      "learning_rate": 8.205086325589135e-05,
      "loss": 0.1766843944787979,
      "mean_token_accuracy": 0.9282799512147903,
      "num_tokens": 100447936.0,
      "step": 8143
    },
    {
      "entropy": 1.0725449323654175,
      "epoch": 4.288572933122696,
      "grad_norm": 0.2979128956794739,
      "learning_rate": 8.203001996859851e-05,
      "loss": 0.17614522576332092,
      "mean_token_accuracy": 0.9281874150037766,
      "num_tokens": 100460272.0,
      "step": 8144
    },
    {
      "entropy": 1.0823680460453033,
      "epoch": 4.289099526066351,
      "grad_norm": 0.3079119026660919,
      "learning_rate": 8.200917834172033e-05,
      "loss": 0.1636887639760971,
      "mean_token_accuracy": 0.9274661689996719,
      "num_tokens": 100472608.0,
      "step": 8145
    },
    {
      "entropy": 1.086853265762329,
      "epoch": 4.289626119010006,
      "grad_norm": 0.2915719151496887,
      "learning_rate": 8.198833837649412e-05,
      "loss": 0.1665974259376526,
      "mean_token_accuracy": 0.9324109107255936,
      "num_tokens": 100484944.0,
      "step": 8146
    },
    {
      "entropy": 1.0810329616069794,
      "epoch": 4.2901527119536595,
      "grad_norm": 0.35588109493255615,
      "learning_rate": 8.196750007415697e-05,
      "loss": 0.2031797468662262,
      "mean_token_accuracy": 0.9193731099367142,
      "num_tokens": 100497280.0,
      "step": 8147
    },
    {
      "entropy": 1.1038683354854584,
      "epoch": 4.290679304897314,
      "grad_norm": 0.2943333387374878,
      "learning_rate": 8.194666343594597e-05,
      "loss": 0.15437810122966766,
      "mean_token_accuracy": 0.9334327727556229,
      "num_tokens": 100509616.0,
      "step": 8148
    },
    {
      "entropy": 1.0905348658561707,
      "epoch": 4.291205897840969,
      "grad_norm": 0.2971420884132385,
      "learning_rate": 8.192582846309802e-05,
      "loss": 0.16804301738739014,
      "mean_token_accuracy": 0.9282435178756714,
      "num_tokens": 100521952.0,
      "step": 8149
    },
    {
      "entropy": 1.0975285172462463,
      "epoch": 4.291732490784623,
      "grad_norm": 0.30539795756340027,
      "learning_rate": 8.190499515685001e-05,
      "loss": 0.1692083477973938,
      "mean_token_accuracy": 0.928924560546875,
      "num_tokens": 100534288.0,
      "step": 8150
    },
    {
      "entropy": 1.1452202200889587,
      "epoch": 4.292259083728278,
      "grad_norm": 0.326580286026001,
      "learning_rate": 8.188416351843873e-05,
      "loss": 0.1752394288778305,
      "mean_token_accuracy": 0.9249264299869537,
      "num_tokens": 100546624.0,
      "step": 8151
    },
    {
      "entropy": 1.098305195569992,
      "epoch": 4.292785676671933,
      "grad_norm": 0.30398303270339966,
      "learning_rate": 8.186333354910076e-05,
      "loss": 0.16921615600585938,
      "mean_token_accuracy": 0.9286245256662369,
      "num_tokens": 100558960.0,
      "step": 8152
    },
    {
      "entropy": 1.1302293241024017,
      "epoch": 4.293312269615587,
      "grad_norm": 0.3367408812046051,
      "learning_rate": 8.18425052500727e-05,
      "loss": 0.17959974706172943,
      "mean_token_accuracy": 0.9232142269611359,
      "num_tokens": 100571296.0,
      "step": 8153
    },
    {
      "entropy": 1.132985681295395,
      "epoch": 4.293838862559242,
      "grad_norm": 0.3377591669559479,
      "learning_rate": 8.182167862259098e-05,
      "loss": 0.18186485767364502,
      "mean_token_accuracy": 0.9227332919836044,
      "num_tokens": 100583632.0,
      "step": 8154
    },
    {
      "entropy": 1.086346834897995,
      "epoch": 4.294365455502897,
      "grad_norm": 0.29264533519744873,
      "learning_rate": 8.180085366789199e-05,
      "loss": 0.1761588454246521,
      "mean_token_accuracy": 0.9259068518877029,
      "num_tokens": 100595968.0,
      "step": 8155
    },
    {
      "entropy": 1.1045507341623306,
      "epoch": 4.294892048446551,
      "grad_norm": 0.3215351402759552,
      "learning_rate": 8.178003038721189e-05,
      "loss": 0.17329712212085724,
      "mean_token_accuracy": 0.9306823462247849,
      "num_tokens": 100608304.0,
      "step": 8156
    },
    {
      "entropy": 1.120214581489563,
      "epoch": 4.295418641390206,
      "grad_norm": 0.30607500672340393,
      "learning_rate": 8.175920878178696e-05,
      "loss": 0.16385003924369812,
      "mean_token_accuracy": 0.9280714094638824,
      "num_tokens": 100620640.0,
      "step": 8157
    },
    {
      "entropy": 1.1060186922550201,
      "epoch": 4.2959452343338596,
      "grad_norm": 0.3463023900985718,
      "learning_rate": 8.17383888528532e-05,
      "loss": 0.18140186369419098,
      "mean_token_accuracy": 0.9217744767665863,
      "num_tokens": 100632976.0,
      "step": 8158
    },
    {
      "entropy": 1.11289843916893,
      "epoch": 4.296471827277514,
      "grad_norm": 0.3010319471359253,
      "learning_rate": 8.17175706016466e-05,
      "loss": 0.17120137810707092,
      "mean_token_accuracy": 0.9283831566572189,
      "num_tokens": 100645312.0,
      "step": 8159
    },
    {
      "entropy": 1.0737940967082977,
      "epoch": 4.296998420221169,
      "grad_norm": 0.29794272780418396,
      "learning_rate": 8.169675402940296e-05,
      "loss": 0.1645357608795166,
      "mean_token_accuracy": 0.9277155101299286,
      "num_tokens": 100657648.0,
      "step": 8160
    },
    {
      "entropy": 1.10798978805542,
      "epoch": 4.297525013164823,
      "grad_norm": 0.3176315724849701,
      "learning_rate": 8.167593913735807e-05,
      "loss": 0.17072540521621704,
      "mean_token_accuracy": 0.9245764017105103,
      "num_tokens": 100669984.0,
      "step": 8161
    },
    {
      "entropy": 1.0653933882713318,
      "epoch": 4.298051606108478,
      "grad_norm": 0.2844933867454529,
      "learning_rate": 8.16551259267476e-05,
      "loss": 0.162329763174057,
      "mean_token_accuracy": 0.9278407990932465,
      "num_tokens": 100682320.0,
      "step": 8162
    },
    {
      "entropy": 1.126930981874466,
      "epoch": 4.298578199052133,
      "grad_norm": 0.3100861608982086,
      "learning_rate": 8.163431439880706e-05,
      "loss": 0.1709498018026352,
      "mean_token_accuracy": 0.9280589818954468,
      "num_tokens": 100694656.0,
      "step": 8163
    },
    {
      "entropy": 1.1281602084636688,
      "epoch": 4.299104791995787,
      "grad_norm": 0.3464886546134949,
      "learning_rate": 8.161350455477197e-05,
      "loss": 0.17898613214492798,
      "mean_token_accuracy": 0.9198716878890991,
      "num_tokens": 100706992.0,
      "step": 8164
    },
    {
      "entropy": 1.0705135464668274,
      "epoch": 4.299631384939442,
      "grad_norm": 0.292278915643692,
      "learning_rate": 8.159269639587768e-05,
      "loss": 0.16062796115875244,
      "mean_token_accuracy": 0.9355687946081161,
      "num_tokens": 100719328.0,
      "step": 8165
    },
    {
      "entropy": 1.1133157908916473,
      "epoch": 4.300157977883097,
      "grad_norm": 0.30878230929374695,
      "learning_rate": 8.157188992335942e-05,
      "loss": 0.1541077047586441,
      "mean_token_accuracy": 0.9316331744194031,
      "num_tokens": 100731664.0,
      "step": 8166
    },
    {
      "entropy": 1.1160552203655243,
      "epoch": 4.300684570826751,
      "grad_norm": 0.33631134033203125,
      "learning_rate": 8.155108513845238e-05,
      "loss": 0.16401691734790802,
      "mean_token_accuracy": 0.9292161464691162,
      "num_tokens": 100744000.0,
      "step": 8167
    },
    {
      "entropy": 1.106523036956787,
      "epoch": 4.301211163770406,
      "grad_norm": 0.368388831615448,
      "learning_rate": 8.153028204239157e-05,
      "loss": 0.1936802864074707,
      "mean_token_accuracy": 0.9172158390283585,
      "num_tokens": 100756336.0,
      "step": 8168
    },
    {
      "entropy": 1.1062554717063904,
      "epoch": 4.3017377567140604,
      "grad_norm": 0.3216768503189087,
      "learning_rate": 8.150948063641201e-05,
      "loss": 0.16468966007232666,
      "mean_token_accuracy": 0.9298742860555649,
      "num_tokens": 100768672.0,
      "step": 8169
    },
    {
      "entropy": 1.0903112590312958,
      "epoch": 4.302264349657714,
      "grad_norm": 0.3083604872226715,
      "learning_rate": 8.14886809217485e-05,
      "loss": 0.16503220796585083,
      "mean_token_accuracy": 0.9277413338422775,
      "num_tokens": 100781008.0,
      "step": 8170
    },
    {
      "entropy": 1.1245929300785065,
      "epoch": 4.302790942601369,
      "grad_norm": 0.3205699324607849,
      "learning_rate": 8.146788289963585e-05,
      "loss": 0.16636714339256287,
      "mean_token_accuracy": 0.9310216158628464,
      "num_tokens": 100793344.0,
      "step": 8171
    },
    {
      "entropy": 1.1266671121120453,
      "epoch": 4.303317535545023,
      "grad_norm": 0.33314967155456543,
      "learning_rate": 8.144708657130868e-05,
      "loss": 0.16889874637126923,
      "mean_token_accuracy": 0.933164969086647,
      "num_tokens": 100805680.0,
      "step": 8172
    },
    {
      "entropy": 1.1593082547187805,
      "epoch": 4.303844128488678,
      "grad_norm": 0.3262190818786621,
      "learning_rate": 8.142629193800158e-05,
      "loss": 0.17436131834983826,
      "mean_token_accuracy": 0.9216962605714798,
      "num_tokens": 100818016.0,
      "step": 8173
    },
    {
      "entropy": 1.1116544306278229,
      "epoch": 4.304370721432333,
      "grad_norm": 0.31282833218574524,
      "learning_rate": 8.1405499000949e-05,
      "loss": 0.1572895646095276,
      "mean_token_accuracy": 0.9315769672393799,
      "num_tokens": 100830352.0,
      "step": 8174
    },
    {
      "entropy": 1.1222815811634064,
      "epoch": 4.304897314375987,
      "grad_norm": 0.2925593852996826,
      "learning_rate": 8.138470776138527e-05,
      "loss": 0.16084107756614685,
      "mean_token_accuracy": 0.933529332280159,
      "num_tokens": 100842688.0,
      "step": 8175
    },
    {
      "entropy": 1.0700201392173767,
      "epoch": 4.305423907319642,
      "grad_norm": 0.3212798535823822,
      "learning_rate": 8.136391822054466e-05,
      "loss": 0.17239071428775787,
      "mean_token_accuracy": 0.928302064538002,
      "num_tokens": 100855024.0,
      "step": 8176
    },
    {
      "entropy": 1.171550303697586,
      "epoch": 4.305950500263297,
      "grad_norm": 0.3465503752231598,
      "learning_rate": 8.134313037966126e-05,
      "loss": 0.16142232716083527,
      "mean_token_accuracy": 0.9277971833944321,
      "num_tokens": 100867360.0,
      "step": 8177
    },
    {
      "entropy": 1.1796109676361084,
      "epoch": 4.306477093206951,
      "grad_norm": 0.31742334365844727,
      "learning_rate": 8.132234423996924e-05,
      "loss": 0.1670718491077423,
      "mean_token_accuracy": 0.927848294377327,
      "num_tokens": 100879696.0,
      "step": 8178
    },
    {
      "entropy": 1.130541980266571,
      "epoch": 4.307003686150606,
      "grad_norm": 0.31697362661361694,
      "learning_rate": 8.130155980270251e-05,
      "loss": 0.16957148909568787,
      "mean_token_accuracy": 0.9272715300321579,
      "num_tokens": 100892032.0,
      "step": 8179
    },
    {
      "entropy": 1.090952455997467,
      "epoch": 4.3075302790942605,
      "grad_norm": 0.3405947983264923,
      "learning_rate": 8.12807770690949e-05,
      "loss": 0.18022146821022034,
      "mean_token_accuracy": 0.9248402118682861,
      "num_tokens": 100904368.0,
      "step": 8180
    },
    {
      "entropy": 1.123077630996704,
      "epoch": 4.308056872037914,
      "grad_norm": 0.32419994473457336,
      "learning_rate": 8.125999604038015e-05,
      "loss": 0.16609665751457214,
      "mean_token_accuracy": 0.9302123636007309,
      "num_tokens": 100916704.0,
      "step": 8181
    },
    {
      "entropy": 1.1256820559501648,
      "epoch": 4.308583464981569,
      "grad_norm": 0.3087804615497589,
      "learning_rate": 8.123921671779193e-05,
      "loss": 0.17110829055309296,
      "mean_token_accuracy": 0.9277935028076172,
      "num_tokens": 100929040.0,
      "step": 8182
    },
    {
      "entropy": 1.1385431289672852,
      "epoch": 4.309110057925224,
      "grad_norm": 0.3376697599887848,
      "learning_rate": 8.121843910256379e-05,
      "loss": 0.17482459545135498,
      "mean_token_accuracy": 0.924680083990097,
      "num_tokens": 100941376.0,
      "step": 8183
    },
    {
      "entropy": 1.1289255023002625,
      "epoch": 4.309636650868878,
      "grad_norm": 0.29189300537109375,
      "learning_rate": 8.119766319592917e-05,
      "loss": 0.17363496124744415,
      "mean_token_accuracy": 0.9293947070837021,
      "num_tokens": 100953712.0,
      "step": 8184
    },
    {
      "entropy": 1.1921924948692322,
      "epoch": 4.310163243812533,
      "grad_norm": 0.3469204008579254,
      "learning_rate": 8.11768889991214e-05,
      "loss": 0.195139542222023,
      "mean_token_accuracy": 0.9169978052377701,
      "num_tokens": 100966048.0,
      "step": 8185
    },
    {
      "entropy": 1.1126163601875305,
      "epoch": 4.310689836756188,
      "grad_norm": 0.27752256393432617,
      "learning_rate": 8.115611651337373e-05,
      "loss": 0.1593462973833084,
      "mean_token_accuracy": 0.931410163640976,
      "num_tokens": 100978384.0,
      "step": 8186
    },
    {
      "entropy": 1.1427457630634308,
      "epoch": 4.311216429699842,
      "grad_norm": 0.3243776559829712,
      "learning_rate": 8.113534573991932e-05,
      "loss": 0.17774705588817596,
      "mean_token_accuracy": 0.9292377531528473,
      "num_tokens": 100990720.0,
      "step": 8187
    },
    {
      "entropy": 1.1437013745307922,
      "epoch": 4.311743022643497,
      "grad_norm": 0.3289145529270172,
      "learning_rate": 8.111457667999123e-05,
      "loss": 0.18479706346988678,
      "mean_token_accuracy": 0.9234186708927155,
      "num_tokens": 101003056.0,
      "step": 8188
    },
    {
      "entropy": 1.1382064819335938,
      "epoch": 4.312269615587151,
      "grad_norm": 0.3195880651473999,
      "learning_rate": 8.109380933482234e-05,
      "loss": 0.18417364358901978,
      "mean_token_accuracy": 0.9193008244037628,
      "num_tokens": 101015392.0,
      "step": 8189
    },
    {
      "entropy": 1.1460935473442078,
      "epoch": 4.312796208530806,
      "grad_norm": 0.313053160905838,
      "learning_rate": 8.10730437056455e-05,
      "loss": 0.169764444231987,
      "mean_token_accuracy": 0.9266730397939682,
      "num_tokens": 101027728.0,
      "step": 8190
    },
    {
      "entropy": 1.1244223415851593,
      "epoch": 4.3133228014744605,
      "grad_norm": 0.3006483018398285,
      "learning_rate": 8.105227979369343e-05,
      "loss": 0.1739342212677002,
      "mean_token_accuracy": 0.9234243780374527,
      "num_tokens": 101040064.0,
      "step": 8191
    },
    {
      "entropy": 1.0908695459365845,
      "epoch": 4.313849394418114,
      "grad_norm": 0.28290775418281555,
      "learning_rate": 8.103151760019883e-05,
      "loss": 0.1659098118543625,
      "mean_token_accuracy": 0.9311094582080841,
      "num_tokens": 101052400.0,
      "step": 8192
    },
    {
      "entropy": 1.1070260107517242,
      "epoch": 4.314375987361769,
      "grad_norm": 0.2765958309173584,
      "learning_rate": 8.101075712639418e-05,
      "loss": 0.15549322962760925,
      "mean_token_accuracy": 0.9331227988004684,
      "num_tokens": 101064736.0,
      "step": 8193
    },
    {
      "entropy": 1.1252348721027374,
      "epoch": 4.314902580305424,
      "grad_norm": 0.31335318088531494,
      "learning_rate": 8.098999837351193e-05,
      "loss": 0.17366810142993927,
      "mean_token_accuracy": 0.9225934594869614,
      "num_tokens": 101077072.0,
      "step": 8194
    },
    {
      "entropy": 1.1199999153614044,
      "epoch": 4.315429173249078,
      "grad_norm": 0.32513147592544556,
      "learning_rate": 8.096924134278438e-05,
      "loss": 0.18315058946609497,
      "mean_token_accuracy": 0.9197819530963898,
      "num_tokens": 101089408.0,
      "step": 8195
    },
    {
      "entropy": 1.0830354988574982,
      "epoch": 4.315955766192733,
      "grad_norm": 0.3034358322620392,
      "learning_rate": 8.094848603544376e-05,
      "loss": 0.17146772146224976,
      "mean_token_accuracy": 0.9262642115354538,
      "num_tokens": 101101744.0,
      "step": 8196
    },
    {
      "entropy": 1.1177780628204346,
      "epoch": 4.316482359136388,
      "grad_norm": 0.3099781274795532,
      "learning_rate": 8.092773245272223e-05,
      "loss": 0.16914069652557373,
      "mean_token_accuracy": 0.92646524310112,
      "num_tokens": 101114080.0,
      "step": 8197
    },
    {
      "entropy": 1.1332667469978333,
      "epoch": 4.317008952080042,
      "grad_norm": 0.32865744829177856,
      "learning_rate": 8.090698059585174e-05,
      "loss": 0.17182783782482147,
      "mean_token_accuracy": 0.9264388978481293,
      "num_tokens": 101126416.0,
      "step": 8198
    },
    {
      "entropy": 1.1120539903640747,
      "epoch": 4.317535545023697,
      "grad_norm": 0.3232680857181549,
      "learning_rate": 8.088623046606425e-05,
      "loss": 0.18395736813545227,
      "mean_token_accuracy": 0.9195074439048767,
      "num_tokens": 101138752.0,
      "step": 8199
    },
    {
      "entropy": 1.100280076265335,
      "epoch": 4.318062137967352,
      "grad_norm": 0.31531259417533875,
      "learning_rate": 8.086548206459157e-05,
      "loss": 0.17667675018310547,
      "mean_token_accuracy": 0.9256656914949417,
      "num_tokens": 101151088.0,
      "step": 8200
    },
    {
      "entropy": 1.0641197562217712,
      "epoch": 4.318588730911006,
      "grad_norm": 0.2955949306488037,
      "learning_rate": 8.084473539266544e-05,
      "loss": 0.18040215969085693,
      "mean_token_accuracy": 0.9200959354639053,
      "num_tokens": 101163424.0,
      "step": 8201
    },
    {
      "entropy": 1.1014371812343597,
      "epoch": 4.3191153238546605,
      "grad_norm": 0.3408636748790741,
      "learning_rate": 8.08239904515174e-05,
      "loss": 0.1776238977909088,
      "mean_token_accuracy": 0.9219309389591217,
      "num_tokens": 101175760.0,
      "step": 8202
    },
    {
      "entropy": 1.0921312868595123,
      "epoch": 4.319641916798314,
      "grad_norm": 0.301546186208725,
      "learning_rate": 8.080324724237904e-05,
      "loss": 0.16484412550926208,
      "mean_token_accuracy": 0.9259563833475113,
      "num_tokens": 101188096.0,
      "step": 8203
    },
    {
      "entropy": 1.1107077896595001,
      "epoch": 4.320168509741969,
      "grad_norm": 0.3042460083961487,
      "learning_rate": 8.078250576648171e-05,
      "loss": 0.15505820512771606,
      "mean_token_accuracy": 0.9346891492605209,
      "num_tokens": 101200432.0,
      "step": 8204
    },
    {
      "entropy": 1.110642284154892,
      "epoch": 4.320695102685624,
      "grad_norm": 0.2894485890865326,
      "learning_rate": 8.076176602505667e-05,
      "loss": 0.16648268699645996,
      "mean_token_accuracy": 0.9282403737306595,
      "num_tokens": 101212768.0,
      "step": 8205
    },
    {
      "entropy": 1.0758769810199738,
      "epoch": 4.321221695629278,
      "grad_norm": 0.29397353529930115,
      "learning_rate": 8.07410280193352e-05,
      "loss": 0.15893948078155518,
      "mean_token_accuracy": 0.936321809887886,
      "num_tokens": 101225104.0,
      "step": 8206
    },
    {
      "entropy": 1.1222665905952454,
      "epoch": 4.321748288572933,
      "grad_norm": 0.3513292968273163,
      "learning_rate": 8.07202917505484e-05,
      "loss": 0.18077202141284943,
      "mean_token_accuracy": 0.924056202173233,
      "num_tokens": 101237440.0,
      "step": 8207
    },
    {
      "entropy": 1.130362629890442,
      "epoch": 4.322274881516588,
      "grad_norm": 0.3298923671245575,
      "learning_rate": 8.069955721992719e-05,
      "loss": 0.18404030799865723,
      "mean_token_accuracy": 0.9224449396133423,
      "num_tokens": 101249776.0,
      "step": 8208
    },
    {
      "entropy": 1.096226155757904,
      "epoch": 4.322801474460242,
      "grad_norm": 0.32115793228149414,
      "learning_rate": 8.067882442870249e-05,
      "loss": 0.18709102272987366,
      "mean_token_accuracy": 0.9188259690999985,
      "num_tokens": 101262112.0,
      "step": 8209
    },
    {
      "entropy": 1.0604323893785477,
      "epoch": 4.323328067403897,
      "grad_norm": 0.30055803060531616,
      "learning_rate": 8.065809337810508e-05,
      "loss": 0.15584015846252441,
      "mean_token_accuracy": 0.9344631284475327,
      "num_tokens": 101274448.0,
      "step": 8210
    },
    {
      "entropy": 1.1120015680789948,
      "epoch": 4.323854660347552,
      "grad_norm": 0.3131413459777832,
      "learning_rate": 8.063736406936566e-05,
      "loss": 0.17670822143554688,
      "mean_token_accuracy": 0.9243551790714264,
      "num_tokens": 101286784.0,
      "step": 8211
    },
    {
      "entropy": 1.0935775637626648,
      "epoch": 4.324381253291206,
      "grad_norm": 0.3220430016517639,
      "learning_rate": 8.061663650371478e-05,
      "loss": 0.17674963176250458,
      "mean_token_accuracy": 0.9223094582557678,
      "num_tokens": 101299120.0,
      "step": 8212
    },
    {
      "entropy": 1.0746557414531708,
      "epoch": 4.3249078462348605,
      "grad_norm": 0.32474976778030396,
      "learning_rate": 8.059591068238295e-05,
      "loss": 0.16875208914279938,
      "mean_token_accuracy": 0.9258905351161957,
      "num_tokens": 101311456.0,
      "step": 8213
    },
    {
      "entropy": 1.0265684574842453,
      "epoch": 4.325434439178515,
      "grad_norm": 0.29002878069877625,
      "learning_rate": 8.05751866066005e-05,
      "loss": 0.16856175661087036,
      "mean_token_accuracy": 0.931471049785614,
      "num_tokens": 101323792.0,
      "step": 8214
    },
    {
      "entropy": 1.0913830697536469,
      "epoch": 4.325961032122169,
      "grad_norm": 0.31670016050338745,
      "learning_rate": 8.055446427759776e-05,
      "loss": 0.17259293794631958,
      "mean_token_accuracy": 0.9279854148626328,
      "num_tokens": 101336128.0,
      "step": 8215
    },
    {
      "entropy": 1.1229613423347473,
      "epoch": 4.326487625065824,
      "grad_norm": 0.3105393946170807,
      "learning_rate": 8.053374369660485e-05,
      "loss": 0.17131748795509338,
      "mean_token_accuracy": 0.9297657907009125,
      "num_tokens": 101348464.0,
      "step": 8216
    },
    {
      "entropy": 1.1170935332775116,
      "epoch": 4.327014218009479,
      "grad_norm": 0.3121127784252167,
      "learning_rate": 8.051302486485179e-05,
      "loss": 0.16797514259815216,
      "mean_token_accuracy": 0.9313868284225464,
      "num_tokens": 101360800.0,
      "step": 8217
    },
    {
      "entropy": 1.1099775731563568,
      "epoch": 4.327540810953133,
      "grad_norm": 0.3253271281719208,
      "learning_rate": 8.049230778356864e-05,
      "loss": 0.18405406177043915,
      "mean_token_accuracy": 0.9198083132505417,
      "num_tokens": 101373136.0,
      "step": 8218
    },
    {
      "entropy": 1.0884588360786438,
      "epoch": 4.328067403896788,
      "grad_norm": 0.316718190908432,
      "learning_rate": 8.047159245398512e-05,
      "loss": 0.16955435276031494,
      "mean_token_accuracy": 0.9275417774915695,
      "num_tokens": 101385472.0,
      "step": 8219
    },
    {
      "entropy": 1.0748783946037292,
      "epoch": 4.328593996840443,
      "grad_norm": 0.29072773456573486,
      "learning_rate": 8.045087887733109e-05,
      "loss": 0.1633455902338028,
      "mean_token_accuracy": 0.9305248558521271,
      "num_tokens": 101397808.0,
      "step": 8220
    },
    {
      "entropy": 1.1298742890357971,
      "epoch": 4.329120589784097,
      "grad_norm": 0.3651246428489685,
      "learning_rate": 8.043016705483617e-05,
      "loss": 0.18680410087108612,
      "mean_token_accuracy": 0.9223460257053375,
      "num_tokens": 101410144.0,
      "step": 8221
    },
    {
      "entropy": 1.1025915145874023,
      "epoch": 4.329647182727752,
      "grad_norm": 0.33131837844848633,
      "learning_rate": 8.04094569877299e-05,
      "loss": 0.18216900527477264,
      "mean_token_accuracy": 0.9213276356458664,
      "num_tokens": 101422480.0,
      "step": 8222
    },
    {
      "entropy": 1.1199238896369934,
      "epoch": 4.330173775671406,
      "grad_norm": 0.30368417501449585,
      "learning_rate": 8.038874867724169e-05,
      "loss": 0.16693207621574402,
      "mean_token_accuracy": 0.9274206012487411,
      "num_tokens": 101434816.0,
      "step": 8223
    },
    {
      "entropy": 1.0599284619092941,
      "epoch": 4.3307003686150605,
      "grad_norm": 0.2972218990325928,
      "learning_rate": 8.036804212460085e-05,
      "loss": 0.1491440236568451,
      "mean_token_accuracy": 0.939417839050293,
      "num_tokens": 101447152.0,
      "step": 8224
    },
    {
      "entropy": 1.1089601516723633,
      "epoch": 4.331226961558715,
      "grad_norm": 0.42216140031814575,
      "learning_rate": 8.034733733103667e-05,
      "loss": 0.17708426713943481,
      "mean_token_accuracy": 0.922296941280365,
      "num_tokens": 101459488.0,
      "step": 8225
    },
    {
      "entropy": 1.1045544743537903,
      "epoch": 4.331753554502369,
      "grad_norm": 0.33001288771629333,
      "learning_rate": 8.03266342977782e-05,
      "loss": 0.16404478251934052,
      "mean_token_accuracy": 0.9315997511148453,
      "num_tokens": 101471824.0,
      "step": 8226
    },
    {
      "entropy": 1.053360790014267,
      "epoch": 4.332280147446024,
      "grad_norm": 0.2918814420700073,
      "learning_rate": 8.030593302605455e-05,
      "loss": 0.1530412882566452,
      "mean_token_accuracy": 0.9319270551204681,
      "num_tokens": 101484160.0,
      "step": 8227
    },
    {
      "entropy": 1.090760126709938,
      "epoch": 4.332806740389679,
      "grad_norm": 0.32294756174087524,
      "learning_rate": 8.028523351709459e-05,
      "loss": 0.1628977656364441,
      "mean_token_accuracy": 0.9272487610578537,
      "num_tokens": 101496496.0,
      "step": 8228
    },
    {
      "entropy": 1.0790897905826569,
      "epoch": 4.333333333333333,
      "grad_norm": 0.28238362073898315,
      "learning_rate": 8.026453577212709e-05,
      "loss": 0.14742904901504517,
      "mean_token_accuracy": 0.9376995116472244,
      "num_tokens": 101508832.0,
      "step": 8229
    },
    {
      "entropy": 1.1392150819301605,
      "epoch": 4.333859926276988,
      "grad_norm": 0.30798256397247314,
      "learning_rate": 8.024383979238082e-05,
      "loss": 0.1569814682006836,
      "mean_token_accuracy": 0.9336879998445511,
      "num_tokens": 101521168.0,
      "step": 8230
    },
    {
      "entropy": 1.0673526525497437,
      "epoch": 4.334386519220643,
      "grad_norm": 0.31136372685432434,
      "learning_rate": 8.022314557908433e-05,
      "loss": 0.17883369326591492,
      "mean_token_accuracy": 0.9270809441804886,
      "num_tokens": 101533504.0,
      "step": 8231
    },
    {
      "entropy": 1.0603607296943665,
      "epoch": 4.334913112164297,
      "grad_norm": 0.31112179160118103,
      "learning_rate": 8.020245313346615e-05,
      "loss": 0.16724520921707153,
      "mean_token_accuracy": 0.9250359982252121,
      "num_tokens": 101545840.0,
      "step": 8232
    },
    {
      "entropy": 1.132779985666275,
      "epoch": 4.335439705107952,
      "grad_norm": 0.3429107069969177,
      "learning_rate": 8.018176245675462e-05,
      "loss": 0.1782541573047638,
      "mean_token_accuracy": 0.9295335859060287,
      "num_tokens": 101558176.0,
      "step": 8233
    },
    {
      "entropy": 1.0894978046417236,
      "epoch": 4.3359662980516065,
      "grad_norm": 0.3060559630393982,
      "learning_rate": 8.016107355017808e-05,
      "loss": 0.16454628109931946,
      "mean_token_accuracy": 0.933361291885376,
      "num_tokens": 101570512.0,
      "step": 8234
    },
    {
      "entropy": 1.0907858908176422,
      "epoch": 4.3364928909952605,
      "grad_norm": 0.29935285449028015,
      "learning_rate": 8.014038641496471e-05,
      "loss": 0.1586533635854721,
      "mean_token_accuracy": 0.9328591227531433,
      "num_tokens": 101582848.0,
      "step": 8235
    },
    {
      "entropy": 1.1055465936660767,
      "epoch": 4.337019483938915,
      "grad_norm": 0.28609541058540344,
      "learning_rate": 8.011970105234254e-05,
      "loss": 0.164537712931633,
      "mean_token_accuracy": 0.9320068061351776,
      "num_tokens": 101595184.0,
      "step": 8236
    },
    {
      "entropy": 1.0481821596622467,
      "epoch": 4.337546076882569,
      "grad_norm": 0.2869083881378174,
      "learning_rate": 8.00990174635396e-05,
      "loss": 0.16271764039993286,
      "mean_token_accuracy": 0.9291568249464035,
      "num_tokens": 101607520.0,
      "step": 8237
    },
    {
      "entropy": 1.0833254307508469,
      "epoch": 4.338072669826224,
      "grad_norm": 0.30742931365966797,
      "learning_rate": 8.007833564978368e-05,
      "loss": 0.16880999505519867,
      "mean_token_accuracy": 0.9294129312038422,
      "num_tokens": 101619856.0,
      "step": 8238
    },
    {
      "entropy": 1.0962770581245422,
      "epoch": 4.338599262769879,
      "grad_norm": 0.3189244568347931,
      "learning_rate": 8.005765561230259e-05,
      "loss": 0.1695052683353424,
      "mean_token_accuracy": 0.9264257550239563,
      "num_tokens": 101632192.0,
      "step": 8239
    },
    {
      "entropy": 1.1270262598991394,
      "epoch": 4.339125855713533,
      "grad_norm": 0.29804810881614685,
      "learning_rate": 8.003697735232391e-05,
      "loss": 0.17976202070713043,
      "mean_token_accuracy": 0.9248080998659134,
      "num_tokens": 101644528.0,
      "step": 8240
    },
    {
      "entropy": 1.0901038348674774,
      "epoch": 4.339652448657188,
      "grad_norm": 0.290440171957016,
      "learning_rate": 8.001630087107533e-05,
      "loss": 0.15589408576488495,
      "mean_token_accuracy": 0.9316317737102509,
      "num_tokens": 101656864.0,
      "step": 8241
    },
    {
      "entropy": 1.0958742499351501,
      "epoch": 4.340179041600843,
      "grad_norm": 0.29598626494407654,
      "learning_rate": 7.999562616978418e-05,
      "loss": 0.16408175230026245,
      "mean_token_accuracy": 0.9289140403270721,
      "num_tokens": 101669200.0,
      "step": 8242
    },
    {
      "entropy": 1.1210950911045074,
      "epoch": 4.340705634544497,
      "grad_norm": 0.27935031056404114,
      "learning_rate": 7.99749532496778e-05,
      "loss": 0.14974689483642578,
      "mean_token_accuracy": 0.9354144334793091,
      "num_tokens": 101681536.0,
      "step": 8243
    },
    {
      "entropy": 1.1266154646873474,
      "epoch": 4.341232227488152,
      "grad_norm": 0.3143881559371948,
      "learning_rate": 7.99542821119835e-05,
      "loss": 0.1778939664363861,
      "mean_token_accuracy": 0.9241852015256882,
      "num_tokens": 101693872.0,
      "step": 8244
    },
    {
      "entropy": 1.1384577751159668,
      "epoch": 4.3417588204318065,
      "grad_norm": 0.342058002948761,
      "learning_rate": 7.993361275792832e-05,
      "loss": 0.179581880569458,
      "mean_token_accuracy": 0.9234207570552826,
      "num_tokens": 101706208.0,
      "step": 8245
    },
    {
      "entropy": 1.139938086271286,
      "epoch": 4.3422854133754605,
      "grad_norm": 0.3223731815814972,
      "learning_rate": 7.99129451887393e-05,
      "loss": 0.1684388965368271,
      "mean_token_accuracy": 0.927556499838829,
      "num_tokens": 101718544.0,
      "step": 8246
    },
    {
      "entropy": 1.1553017795085907,
      "epoch": 4.342812006319115,
      "grad_norm": 0.3298601508140564,
      "learning_rate": 7.989227940564337e-05,
      "loss": 0.17502903938293457,
      "mean_token_accuracy": 0.9261777848005295,
      "num_tokens": 101730880.0,
      "step": 8247
    },
    {
      "entropy": 1.1250576972961426,
      "epoch": 4.34333859926277,
      "grad_norm": 0.30573534965515137,
      "learning_rate": 7.987161540986733e-05,
      "loss": 0.1717490255832672,
      "mean_token_accuracy": 0.9294047057628632,
      "num_tokens": 101743216.0,
      "step": 8248
    },
    {
      "entropy": 1.1117684543132782,
      "epoch": 4.343865192206424,
      "grad_norm": 0.3257971704006195,
      "learning_rate": 7.985095320263789e-05,
      "loss": 0.17920038104057312,
      "mean_token_accuracy": 0.9236772358417511,
      "num_tokens": 101755552.0,
      "step": 8249
    },
    {
      "entropy": 1.082334816455841,
      "epoch": 4.344391785150079,
      "grad_norm": 0.31119608879089355,
      "learning_rate": 7.983029278518164e-05,
      "loss": 0.17083615064620972,
      "mean_token_accuracy": 0.9256705492734909,
      "num_tokens": 101767888.0,
      "step": 8250
    },
    {
      "entropy": 1.1682907044887543,
      "epoch": 4.344918378093734,
      "grad_norm": 0.3243691623210907,
      "learning_rate": 7.980963415872504e-05,
      "loss": 0.16943079233169556,
      "mean_token_accuracy": 0.9284241944551468,
      "num_tokens": 101780224.0,
      "step": 8251
    },
    {
      "entropy": 1.0876426696777344,
      "epoch": 4.345444971037388,
      "grad_norm": 0.26470980048179626,
      "learning_rate": 7.978897732449456e-05,
      "loss": 0.14852547645568848,
      "mean_token_accuracy": 0.9360248148441315,
      "num_tokens": 101792560.0,
      "step": 8252
    },
    {
      "entropy": 1.0622229874134064,
      "epoch": 4.345971563981043,
      "grad_norm": 0.2839398682117462,
      "learning_rate": 7.976832228371636e-05,
      "loss": 0.1672789752483368,
      "mean_token_accuracy": 0.9297099262475967,
      "num_tokens": 101804896.0,
      "step": 8253
    },
    {
      "entropy": 1.151972770690918,
      "epoch": 4.346498156924698,
      "grad_norm": 0.29570528864860535,
      "learning_rate": 7.974766903761663e-05,
      "loss": 0.15970942378044128,
      "mean_token_accuracy": 0.9289858639240265,
      "num_tokens": 101817232.0,
      "step": 8254
    },
    {
      "entropy": 1.1545989215373993,
      "epoch": 4.347024749868352,
      "grad_norm": 0.3340533971786499,
      "learning_rate": 7.972701758742151e-05,
      "loss": 0.1708054095506668,
      "mean_token_accuracy": 0.9274019151926041,
      "num_tokens": 101829568.0,
      "step": 8255
    },
    {
      "entropy": 1.1824675798416138,
      "epoch": 4.3475513428120065,
      "grad_norm": 0.3400794267654419,
      "learning_rate": 7.970636793435693e-05,
      "loss": 0.17325887084007263,
      "mean_token_accuracy": 0.9251816272735596,
      "num_tokens": 101841904.0,
      "step": 8256
    },
    {
      "entropy": 1.0942457616329193,
      "epoch": 4.3480779357556605,
      "grad_norm": 0.28624141216278076,
      "learning_rate": 7.968572007964869e-05,
      "loss": 0.15125054121017456,
      "mean_token_accuracy": 0.9370819926261902,
      "num_tokens": 101854240.0,
      "step": 8257
    },
    {
      "entropy": 1.1075288206338882,
      "epoch": 4.348604528699315,
      "grad_norm": 0.2798440158367157,
      "learning_rate": 7.966507402452259e-05,
      "loss": 0.15465722978115082,
      "mean_token_accuracy": 0.9323037266731262,
      "num_tokens": 101866576.0,
      "step": 8258
    },
    {
      "entropy": 1.1464895009994507,
      "epoch": 4.34913112164297,
      "grad_norm": 0.31809863448143005,
      "learning_rate": 7.964442977020423e-05,
      "loss": 0.16774927079677582,
      "mean_token_accuracy": 0.9289859533309937,
      "num_tokens": 101878912.0,
      "step": 8259
    },
    {
      "entropy": 1.073552131652832,
      "epoch": 4.349657714586624,
      "grad_norm": 0.32322588562965393,
      "learning_rate": 7.962378731791913e-05,
      "loss": 0.18572726845741272,
      "mean_token_accuracy": 0.9255398958921432,
      "num_tokens": 101891248.0,
      "step": 8260
    },
    {
      "entropy": 1.1035448908805847,
      "epoch": 4.350184307530279,
      "grad_norm": 0.29818570613861084,
      "learning_rate": 7.960314666889272e-05,
      "loss": 0.16822704672813416,
      "mean_token_accuracy": 0.9293882697820663,
      "num_tokens": 101903584.0,
      "step": 8261
    },
    {
      "entropy": 1.1297992765903473,
      "epoch": 4.350710900473934,
      "grad_norm": 0.3098200857639313,
      "learning_rate": 7.958250782435035e-05,
      "loss": 0.16754473745822906,
      "mean_token_accuracy": 0.9317881762981415,
      "num_tokens": 101915920.0,
      "step": 8262
    },
    {
      "entropy": 1.1439328491687775,
      "epoch": 4.351237493417588,
      "grad_norm": 0.3593650460243225,
      "learning_rate": 7.95618707855172e-05,
      "loss": 0.19793128967285156,
      "mean_token_accuracy": 0.9188781529664993,
      "num_tokens": 101928256.0,
      "step": 8263
    },
    {
      "entropy": 1.0824010372161865,
      "epoch": 4.351764086361243,
      "grad_norm": 0.29909494519233704,
      "learning_rate": 7.954123555361839e-05,
      "loss": 0.1682833433151245,
      "mean_token_accuracy": 0.9274113923311234,
      "num_tokens": 101940592.0,
      "step": 8264
    },
    {
      "entropy": 1.1380861103534698,
      "epoch": 4.352290679304898,
      "grad_norm": 0.34476616978645325,
      "learning_rate": 7.952060212987889e-05,
      "loss": 0.1827372908592224,
      "mean_token_accuracy": 0.9255722165107727,
      "num_tokens": 101952928.0,
      "step": 8265
    },
    {
      "entropy": 1.1156515777111053,
      "epoch": 4.352817272248552,
      "grad_norm": 0.3097212612628937,
      "learning_rate": 7.949997051552358e-05,
      "loss": 0.17167949676513672,
      "mean_token_accuracy": 0.928065225481987,
      "num_tokens": 101965264.0,
      "step": 8266
    },
    {
      "entropy": 1.1728405356407166,
      "epoch": 4.3533438651922065,
      "grad_norm": 0.31493303179740906,
      "learning_rate": 7.947934071177729e-05,
      "loss": 0.16536009311676025,
      "mean_token_accuracy": 0.9326711148023605,
      "num_tokens": 101977600.0,
      "step": 8267
    },
    {
      "entropy": 1.0769576132297516,
      "epoch": 4.353870458135861,
      "grad_norm": 0.28808754682540894,
      "learning_rate": 7.945871271986458e-05,
      "loss": 0.16274510324001312,
      "mean_token_accuracy": 0.9312023371458054,
      "num_tokens": 101989936.0,
      "step": 8268
    },
    {
      "entropy": 1.1095938682556152,
      "epoch": 4.354397051079515,
      "grad_norm": 0.32415324449539185,
      "learning_rate": 7.943808654101012e-05,
      "loss": 0.18838092684745789,
      "mean_token_accuracy": 0.9194755703210831,
      "num_tokens": 102002272.0,
      "step": 8269
    },
    {
      "entropy": 1.1336401998996735,
      "epoch": 4.35492364402317,
      "grad_norm": 0.29520735144615173,
      "learning_rate": 7.941746217643837e-05,
      "loss": 0.1495727300643921,
      "mean_token_accuracy": 0.9346309304237366,
      "num_tokens": 102014608.0,
      "step": 8270
    },
    {
      "entropy": 1.150920331478119,
      "epoch": 4.355450236966824,
      "grad_norm": 0.3009397089481354,
      "learning_rate": 7.939683962737363e-05,
      "loss": 0.176979199051857,
      "mean_token_accuracy": 0.921332985162735,
      "num_tokens": 102026944.0,
      "step": 8271
    },
    {
      "entropy": 1.090204119682312,
      "epoch": 4.355976829910479,
      "grad_norm": 0.31908345222473145,
      "learning_rate": 7.937621889504015e-05,
      "loss": 0.1687454879283905,
      "mean_token_accuracy": 0.928538590669632,
      "num_tokens": 102039280.0,
      "step": 8272
    },
    {
      "entropy": 1.0926564633846283,
      "epoch": 4.356503422854134,
      "grad_norm": 0.33169350028038025,
      "learning_rate": 7.935559998066206e-05,
      "loss": 0.18119175732135773,
      "mean_token_accuracy": 0.9230940490961075,
      "num_tokens": 102051616.0,
      "step": 8273
    },
    {
      "entropy": 1.0917340517044067,
      "epoch": 4.357030015797788,
      "grad_norm": 0.33342838287353516,
      "learning_rate": 7.933498288546339e-05,
      "loss": 0.18393750488758087,
      "mean_token_accuracy": 0.9222940355539322,
      "num_tokens": 102063952.0,
      "step": 8274
    },
    {
      "entropy": 1.145456463098526,
      "epoch": 4.357556608741443,
      "grad_norm": 0.3058793842792511,
      "learning_rate": 7.931436761066801e-05,
      "loss": 0.16043967008590698,
      "mean_token_accuracy": 0.9309965074062347,
      "num_tokens": 102076288.0,
      "step": 8275
    },
    {
      "entropy": 1.1027463674545288,
      "epoch": 4.358083201685098,
      "grad_norm": 0.32194235920906067,
      "learning_rate": 7.929375415749984e-05,
      "loss": 0.1844458281993866,
      "mean_token_accuracy": 0.9184555113315582,
      "num_tokens": 102088624.0,
      "step": 8276
    },
    {
      "entropy": 1.0956955552101135,
      "epoch": 4.358609794628752,
      "grad_norm": 0.3411625623703003,
      "learning_rate": 7.927314252718249e-05,
      "loss": 0.18235260248184204,
      "mean_token_accuracy": 0.9206727594137192,
      "num_tokens": 102100960.0,
      "step": 8277
    },
    {
      "entropy": 1.095288634300232,
      "epoch": 4.3591363875724065,
      "grad_norm": 0.2878154516220093,
      "learning_rate": 7.925253272093959e-05,
      "loss": 0.15658070147037506,
      "mean_token_accuracy": 0.933790922164917,
      "num_tokens": 102113296.0,
      "step": 8278
    },
    {
      "entropy": 1.1430086195468903,
      "epoch": 4.359662980516061,
      "grad_norm": 0.3287704885005951,
      "learning_rate": 7.923192473999461e-05,
      "loss": 0.16294170916080475,
      "mean_token_accuracy": 0.9294786006212234,
      "num_tokens": 102125632.0,
      "step": 8279
    },
    {
      "entropy": 1.0570924878120422,
      "epoch": 4.360189573459715,
      "grad_norm": 0.29489120841026306,
      "learning_rate": 7.921131858557091e-05,
      "loss": 0.1653260886669159,
      "mean_token_accuracy": 0.9298093169927597,
      "num_tokens": 102137968.0,
      "step": 8280
    },
    {
      "entropy": 1.1138690412044525,
      "epoch": 4.36071616640337,
      "grad_norm": 0.3456246256828308,
      "learning_rate": 7.91907142588918e-05,
      "loss": 0.18220512568950653,
      "mean_token_accuracy": 0.92311991751194,
      "num_tokens": 102150304.0,
      "step": 8281
    },
    {
      "entropy": 1.0722036063671112,
      "epoch": 4.361242759347025,
      "grad_norm": 0.31399351358413696,
      "learning_rate": 7.917011176118039e-05,
      "loss": 0.17284181714057922,
      "mean_token_accuracy": 0.9255773425102234,
      "num_tokens": 102162640.0,
      "step": 8282
    },
    {
      "entropy": 1.0678641498088837,
      "epoch": 4.361769352290679,
      "grad_norm": 0.2958000600337982,
      "learning_rate": 7.914951109365978e-05,
      "loss": 0.1648251712322235,
      "mean_token_accuracy": 0.9297206252813339,
      "num_tokens": 102174976.0,
      "step": 8283
    },
    {
      "entropy": 1.087259441614151,
      "epoch": 4.362295945234334,
      "grad_norm": 0.3215687870979309,
      "learning_rate": 7.912891225755288e-05,
      "loss": 0.1759418547153473,
      "mean_token_accuracy": 0.9221944957971573,
      "num_tokens": 102187312.0,
      "step": 8284
    },
    {
      "entropy": 1.0881843864917755,
      "epoch": 4.362822538177989,
      "grad_norm": 0.29326653480529785,
      "learning_rate": 7.910831525408251e-05,
      "loss": 0.17087194323539734,
      "mean_token_accuracy": 0.9310245662927628,
      "num_tokens": 102199648.0,
      "step": 8285
    },
    {
      "entropy": 1.029636561870575,
      "epoch": 4.363349131121643,
      "grad_norm": 0.29191556572914124,
      "learning_rate": 7.908772008447144e-05,
      "loss": 0.15503615140914917,
      "mean_token_accuracy": 0.9320820719003677,
      "num_tokens": 102211984.0,
      "step": 8286
    },
    {
      "entropy": 1.024407148361206,
      "epoch": 4.363875724065298,
      "grad_norm": 0.2988927960395813,
      "learning_rate": 7.906712674994225e-05,
      "loss": 0.16153010725975037,
      "mean_token_accuracy": 0.9317641705274582,
      "num_tokens": 102224320.0,
      "step": 8287
    },
    {
      "entropy": 1.0935994684696198,
      "epoch": 4.3644023170089525,
      "grad_norm": 0.32042306661605835,
      "learning_rate": 7.904653525171746e-05,
      "loss": 0.17825911939144135,
      "mean_token_accuracy": 0.9292528629302979,
      "num_tokens": 102236656.0,
      "step": 8288
    },
    {
      "entropy": 1.036267563700676,
      "epoch": 4.3649289099526065,
      "grad_norm": 0.2919551432132721,
      "learning_rate": 7.902594559101941e-05,
      "loss": 0.15347644686698914,
      "mean_token_accuracy": 0.9331680238246918,
      "num_tokens": 102248992.0,
      "step": 8289
    },
    {
      "entropy": 1.0453420579433441,
      "epoch": 4.365455502896261,
      "grad_norm": 0.30875593423843384,
      "learning_rate": 7.900535776907049e-05,
      "loss": 0.17227022349834442,
      "mean_token_accuracy": 0.9263574779033661,
      "num_tokens": 102261328.0,
      "step": 8290
    },
    {
      "entropy": 0.9784771054983139,
      "epoch": 4.365982095839915,
      "grad_norm": 0.31372925639152527,
      "learning_rate": 7.898477178709283e-05,
      "loss": 0.17568835616111755,
      "mean_token_accuracy": 0.922653004527092,
      "num_tokens": 102273664.0,
      "step": 8291
    },
    {
      "entropy": 1.0353483855724335,
      "epoch": 4.36650868878357,
      "grad_norm": 0.3407802879810333,
      "learning_rate": 7.896418764630848e-05,
      "loss": 0.16763520240783691,
      "mean_token_accuracy": 0.9367599487304688,
      "num_tokens": 102286000.0,
      "step": 8292
    },
    {
      "entropy": 1.0111843794584274,
      "epoch": 4.367035281727225,
      "grad_norm": 0.3053271174430847,
      "learning_rate": 7.894360534793942e-05,
      "loss": 0.163558229804039,
      "mean_token_accuracy": 0.9313251227140427,
      "num_tokens": 102298336.0,
      "step": 8293
    },
    {
      "entropy": 0.990428626537323,
      "epoch": 4.367561874670879,
      "grad_norm": 0.3073551654815674,
      "learning_rate": 7.892302489320753e-05,
      "loss": 0.16031494736671448,
      "mean_token_accuracy": 0.9327969402074814,
      "num_tokens": 102310672.0,
      "step": 8294
    },
    {
      "entropy": 1.025489255785942,
      "epoch": 4.368088467614534,
      "grad_norm": 0.30879753828048706,
      "learning_rate": 7.890244628333449e-05,
      "loss": 0.16243739426136017,
      "mean_token_accuracy": 0.9296454042196274,
      "num_tokens": 102323008.0,
      "step": 8295
    },
    {
      "entropy": 1.015230417251587,
      "epoch": 4.368615060558189,
      "grad_norm": 0.3144312798976898,
      "learning_rate": 7.888186951954197e-05,
      "loss": 0.185271754860878,
      "mean_token_accuracy": 0.9228046387434006,
      "num_tokens": 102335344.0,
      "step": 8296
    },
    {
      "entropy": 0.9996123164892197,
      "epoch": 4.369141653501843,
      "grad_norm": 0.3275398313999176,
      "learning_rate": 7.886129460305147e-05,
      "loss": 0.17242178320884705,
      "mean_token_accuracy": 0.9246769696474075,
      "num_tokens": 102347680.0,
      "step": 8297
    },
    {
      "entropy": 1.019558385014534,
      "epoch": 4.369668246445498,
      "grad_norm": 0.3341851234436035,
      "learning_rate": 7.884072153508444e-05,
      "loss": 0.17046117782592773,
      "mean_token_accuracy": 0.9291224926710129,
      "num_tokens": 102360016.0,
      "step": 8298
    },
    {
      "entropy": 1.0215636640787125,
      "epoch": 4.3701948393891525,
      "grad_norm": 0.3042353689670563,
      "learning_rate": 7.882015031686218e-05,
      "loss": 0.17030411958694458,
      "mean_token_accuracy": 0.9261429458856583,
      "num_tokens": 102372352.0,
      "step": 8299
    },
    {
      "entropy": 1.0183587074279785,
      "epoch": 4.3707214323328065,
      "grad_norm": 0.3405551314353943,
      "learning_rate": 7.879958094960585e-05,
      "loss": 0.1821388304233551,
      "mean_token_accuracy": 0.9244387149810791,
      "num_tokens": 102384688.0,
      "step": 8300
    },
    {
      "entropy": 1.0054609030485153,
      "epoch": 4.371248025276461,
      "grad_norm": 0.3741503357887268,
      "learning_rate": 7.877901343453655e-05,
      "loss": 0.1840052753686905,
      "mean_token_accuracy": 0.9217223525047302,
      "num_tokens": 102397024.0,
      "step": 8301
    },
    {
      "entropy": 1.0249035060405731,
      "epoch": 4.371774618220116,
      "grad_norm": 0.31064465641975403,
      "learning_rate": 7.875844777287526e-05,
      "loss": 0.16266967356204987,
      "mean_token_accuracy": 0.9317588061094284,
      "num_tokens": 102409360.0,
      "step": 8302
    },
    {
      "entropy": 1.0520051419734955,
      "epoch": 4.37230121116377,
      "grad_norm": 0.35928305983543396,
      "learning_rate": 7.873788396584279e-05,
      "loss": 0.1857140064239502,
      "mean_token_accuracy": 0.922241285443306,
      "num_tokens": 102421696.0,
      "step": 8303
    },
    {
      "entropy": 0.9905842542648315,
      "epoch": 4.372827804107425,
      "grad_norm": 0.2932959198951721,
      "learning_rate": 7.871732201465997e-05,
      "loss": 0.16448016464710236,
      "mean_token_accuracy": 0.9280882775783539,
      "num_tokens": 102434032.0,
      "step": 8304
    },
    {
      "entropy": 1.0434667617082596,
      "epoch": 4.373354397051079,
      "grad_norm": 0.3276841342449188,
      "learning_rate": 7.869676192054744e-05,
      "loss": 0.16333431005477905,
      "mean_token_accuracy": 0.9330431222915649,
      "num_tokens": 102446368.0,
      "step": 8305
    },
    {
      "entropy": 1.0375141203403473,
      "epoch": 4.373880989994734,
      "grad_norm": 0.30908146500587463,
      "learning_rate": 7.867620368472568e-05,
      "loss": 0.17605969309806824,
      "mean_token_accuracy": 0.9249200969934464,
      "num_tokens": 102458704.0,
      "step": 8306
    },
    {
      "entropy": 1.0875357389450073,
      "epoch": 4.374407582938389,
      "grad_norm": 0.3515836000442505,
      "learning_rate": 7.865564730841515e-05,
      "loss": 0.19391240179538727,
      "mean_token_accuracy": 0.9192430227994919,
      "num_tokens": 102471040.0,
      "step": 8307
    },
    {
      "entropy": 1.0471609085798264,
      "epoch": 4.374934175882043,
      "grad_norm": 0.3264077305793762,
      "learning_rate": 7.863509279283614e-05,
      "loss": 0.17311157286167145,
      "mean_token_accuracy": 0.9325474202632904,
      "num_tokens": 102483376.0,
      "step": 8308
    },
    {
      "entropy": 1.0320720374584198,
      "epoch": 4.375460768825698,
      "grad_norm": 0.29720941185951233,
      "learning_rate": 7.861454013920886e-05,
      "loss": 0.16066929697990417,
      "mean_token_accuracy": 0.9330228120088577,
      "num_tokens": 102495712.0,
      "step": 8309
    },
    {
      "entropy": 1.0482531040906906,
      "epoch": 4.3759873617693525,
      "grad_norm": 0.337213397026062,
      "learning_rate": 7.859398934875339e-05,
      "loss": 0.1838356852531433,
      "mean_token_accuracy": 0.9218758940696716,
      "num_tokens": 102508048.0,
      "step": 8310
    },
    {
      "entropy": 1.0781738758087158,
      "epoch": 4.3765139547130065,
      "grad_norm": 0.30988427996635437,
      "learning_rate": 7.857344042268975e-05,
      "loss": 0.16771355271339417,
      "mean_token_accuracy": 0.9302680045366287,
      "num_tokens": 102520384.0,
      "step": 8311
    },
    {
      "entropy": 1.0523044168949127,
      "epoch": 4.377040547656661,
      "grad_norm": 0.31549984216690063,
      "learning_rate": 7.855289336223777e-05,
      "loss": 0.1637999564409256,
      "mean_token_accuracy": 0.9321995824575424,
      "num_tokens": 102532720.0,
      "step": 8312
    },
    {
      "entropy": 1.1031942069530487,
      "epoch": 4.377567140600316,
      "grad_norm": 0.3448771834373474,
      "learning_rate": 7.853234816861723e-05,
      "loss": 0.18763162195682526,
      "mean_token_accuracy": 0.9227224886417389,
      "num_tokens": 102545056.0,
      "step": 8313
    },
    {
      "entropy": 1.0884676575660706,
      "epoch": 4.37809373354397,
      "grad_norm": 0.2905152440071106,
      "learning_rate": 7.851180484304779e-05,
      "loss": 0.1556159406900406,
      "mean_token_accuracy": 0.9334323853254318,
      "num_tokens": 102557392.0,
      "step": 8314
    },
    {
      "entropy": 1.0691013932228088,
      "epoch": 4.378620326487625,
      "grad_norm": 0.2874990403652191,
      "learning_rate": 7.849126338674893e-05,
      "loss": 0.15471886098384857,
      "mean_token_accuracy": 0.936391144990921,
      "num_tokens": 102569728.0,
      "step": 8315
    },
    {
      "entropy": 1.1152063310146332,
      "epoch": 4.37914691943128,
      "grad_norm": 0.34200170636177063,
      "learning_rate": 7.847072380094016e-05,
      "loss": 0.1857263594865799,
      "mean_token_accuracy": 0.924913078546524,
      "num_tokens": 102582064.0,
      "step": 8316
    },
    {
      "entropy": 1.084595412015915,
      "epoch": 4.379673512374934,
      "grad_norm": 0.3384974002838135,
      "learning_rate": 7.845018608684069e-05,
      "loss": 0.1742813140153885,
      "mean_token_accuracy": 0.9260141849517822,
      "num_tokens": 102594400.0,
      "step": 8317
    },
    {
      "entropy": 1.0446248650550842,
      "epoch": 4.380200105318589,
      "grad_norm": 0.3072529137134552,
      "learning_rate": 7.842965024566981e-05,
      "loss": 0.16779743134975433,
      "mean_token_accuracy": 0.9303091764450073,
      "num_tokens": 102606736.0,
      "step": 8318
    },
    {
      "entropy": 1.0834141671657562,
      "epoch": 4.380726698262244,
      "grad_norm": 0.2832176387310028,
      "learning_rate": 7.840911627864662e-05,
      "loss": 0.14648564159870148,
      "mean_token_accuracy": 0.9374445974826813,
      "num_tokens": 102619072.0,
      "step": 8319
    },
    {
      "entropy": 1.0966499745845795,
      "epoch": 4.381253291205898,
      "grad_norm": 0.34519094228744507,
      "learning_rate": 7.838858418699006e-05,
      "loss": 0.18011869490146637,
      "mean_token_accuracy": 0.9262439161539078,
      "num_tokens": 102631408.0,
      "step": 8320
    },
    {
      "entropy": 1.0953200459480286,
      "epoch": 4.3817798841495526,
      "grad_norm": 0.32482707500457764,
      "learning_rate": 7.836805397191902e-05,
      "loss": 0.17221830785274506,
      "mean_token_accuracy": 0.9294404089450836,
      "num_tokens": 102643744.0,
      "step": 8321
    },
    {
      "entropy": 1.1350574493408203,
      "epoch": 4.382306477093207,
      "grad_norm": 0.35024866461753845,
      "learning_rate": 7.834752563465225e-05,
      "loss": 0.18126574158668518,
      "mean_token_accuracy": 0.9207083135843277,
      "num_tokens": 102656080.0,
      "step": 8322
    },
    {
      "entropy": 1.0730578005313873,
      "epoch": 4.382833070036861,
      "grad_norm": 0.2861269414424896,
      "learning_rate": 7.832699917640838e-05,
      "loss": 0.15728303790092468,
      "mean_token_accuracy": 0.9339731335639954,
      "num_tokens": 102668416.0,
      "step": 8323
    },
    {
      "entropy": 1.1431684494018555,
      "epoch": 4.383359662980516,
      "grad_norm": 0.33823126554489136,
      "learning_rate": 7.830647459840595e-05,
      "loss": 0.17363259196281433,
      "mean_token_accuracy": 0.9255153089761734,
      "num_tokens": 102680752.0,
      "step": 8324
    },
    {
      "entropy": 1.0835270881652832,
      "epoch": 4.38388625592417,
      "grad_norm": 0.3068356215953827,
      "learning_rate": 7.82859519018634e-05,
      "loss": 0.16449736058712006,
      "mean_token_accuracy": 0.9274246692657471,
      "num_tokens": 102693088.0,
      "step": 8325
    },
    {
      "entropy": 1.0368002653121948,
      "epoch": 4.384412848867825,
      "grad_norm": 0.28711849451065063,
      "learning_rate": 7.826543108799909e-05,
      "loss": 0.16489332914352417,
      "mean_token_accuracy": 0.9274584352970123,
      "num_tokens": 102705424.0,
      "step": 8326
    },
    {
      "entropy": 1.0520197004079819,
      "epoch": 4.38493944181148,
      "grad_norm": 0.31600284576416016,
      "learning_rate": 7.824491215803113e-05,
      "loss": 0.1721746176481247,
      "mean_token_accuracy": 0.9221912920475006,
      "num_tokens": 102717760.0,
      "step": 8327
    },
    {
      "entropy": 1.070871889591217,
      "epoch": 4.385466034755134,
      "grad_norm": 0.29123905301094055,
      "learning_rate": 7.822439511317768e-05,
      "loss": 0.15711231529712677,
      "mean_token_accuracy": 0.9360679537057877,
      "num_tokens": 102730096.0,
      "step": 8328
    },
    {
      "entropy": 1.1568743288516998,
      "epoch": 4.385992627698789,
      "grad_norm": 0.3370809853076935,
      "learning_rate": 7.820387995465665e-05,
      "loss": 0.1795637607574463,
      "mean_token_accuracy": 0.9237714856863022,
      "num_tokens": 102742432.0,
      "step": 8329
    },
    {
      "entropy": 1.118773490190506,
      "epoch": 4.386519220642444,
      "grad_norm": 0.3151535987854004,
      "learning_rate": 7.818336668368597e-05,
      "loss": 0.1708720177412033,
      "mean_token_accuracy": 0.9265831559896469,
      "num_tokens": 102754768.0,
      "step": 8330
    },
    {
      "entropy": 1.1478121280670166,
      "epoch": 4.387045813586098,
      "grad_norm": 0.31625819206237793,
      "learning_rate": 7.816285530148335e-05,
      "loss": 0.15450188517570496,
      "mean_token_accuracy": 0.9355681091547012,
      "num_tokens": 102767104.0,
      "step": 8331
    },
    {
      "entropy": 1.0916865170001984,
      "epoch": 4.3875724065297526,
      "grad_norm": 0.3406420648097992,
      "learning_rate": 7.814234580926649e-05,
      "loss": 0.186141699552536,
      "mean_token_accuracy": 0.9204560667276382,
      "num_tokens": 102779440.0,
      "step": 8332
    },
    {
      "entropy": 1.067175418138504,
      "epoch": 4.388098999473407,
      "grad_norm": 0.29393482208251953,
      "learning_rate": 7.812183820825285e-05,
      "loss": 0.163357213139534,
      "mean_token_accuracy": 0.9288803040981293,
      "num_tokens": 102791776.0,
      "step": 8333
    },
    {
      "entropy": 1.130958765745163,
      "epoch": 4.388625592417061,
      "grad_norm": 0.31902971863746643,
      "learning_rate": 7.81013324996599e-05,
      "loss": 0.1739107370376587,
      "mean_token_accuracy": 0.9246774911880493,
      "num_tokens": 102804112.0,
      "step": 8334
    },
    {
      "entropy": 1.1164607107639313,
      "epoch": 4.389152185360716,
      "grad_norm": 0.32076266407966614,
      "learning_rate": 7.808082868470489e-05,
      "loss": 0.16847071051597595,
      "mean_token_accuracy": 0.9282487034797668,
      "num_tokens": 102816448.0,
      "step": 8335
    },
    {
      "entropy": 1.1388525664806366,
      "epoch": 4.38967877830437,
      "grad_norm": 0.33620843291282654,
      "learning_rate": 7.806032676460506e-05,
      "loss": 0.1777355670928955,
      "mean_token_accuracy": 0.923837274312973,
      "num_tokens": 102828784.0,
      "step": 8336
    },
    {
      "entropy": 1.114677220582962,
      "epoch": 4.390205371248025,
      "grad_norm": 0.3161875903606415,
      "learning_rate": 7.803982674057748e-05,
      "loss": 0.16435915231704712,
      "mean_token_accuracy": 0.9287358820438385,
      "num_tokens": 102841120.0,
      "step": 8337
    },
    {
      "entropy": 1.1199637949466705,
      "epoch": 4.39073196419168,
      "grad_norm": 0.3084072470664978,
      "learning_rate": 7.801932861383907e-05,
      "loss": 0.1659913957118988,
      "mean_token_accuracy": 0.9315567314624786,
      "num_tokens": 102853456.0,
      "step": 8338
    },
    {
      "entropy": 1.1161784827709198,
      "epoch": 4.391258557135334,
      "grad_norm": 0.3483307659626007,
      "learning_rate": 7.799883238560677e-05,
      "loss": 0.18561196327209473,
      "mean_token_accuracy": 0.9188537895679474,
      "num_tokens": 102865792.0,
      "step": 8339
    },
    {
      "entropy": 1.1186930239200592,
      "epoch": 4.391785150078989,
      "grad_norm": 0.3259585201740265,
      "learning_rate": 7.797833805709724e-05,
      "loss": 0.18975147604942322,
      "mean_token_accuracy": 0.9193664044141769,
      "num_tokens": 102878128.0,
      "step": 8340
    },
    {
      "entropy": 1.132372885942459,
      "epoch": 4.392311743022644,
      "grad_norm": 0.3266378939151764,
      "learning_rate": 7.795784562952718e-05,
      "loss": 0.16240766644477844,
      "mean_token_accuracy": 0.9322902411222458,
      "num_tokens": 102890464.0,
      "step": 8341
    },
    {
      "entropy": 1.0790059268474579,
      "epoch": 4.392838335966298,
      "grad_norm": 0.28562331199645996,
      "learning_rate": 7.793735510411302e-05,
      "loss": 0.1667494922876358,
      "mean_token_accuracy": 0.926630973815918,
      "num_tokens": 102902800.0,
      "step": 8342
    },
    {
      "entropy": 1.0901443362236023,
      "epoch": 4.393364928909953,
      "grad_norm": 0.31055060029029846,
      "learning_rate": 7.791686648207124e-05,
      "loss": 0.17224203050136566,
      "mean_token_accuracy": 0.9274156838655472,
      "num_tokens": 102915136.0,
      "step": 8343
    },
    {
      "entropy": 1.1036354005336761,
      "epoch": 4.393891521853607,
      "grad_norm": 0.3095247149467468,
      "learning_rate": 7.78963797646181e-05,
      "loss": 0.17547622323036194,
      "mean_token_accuracy": 0.9240781366825104,
      "num_tokens": 102927472.0,
      "step": 8344
    },
    {
      "entropy": 1.1046380698680878,
      "epoch": 4.394418114797261,
      "grad_norm": 0.2952825725078583,
      "learning_rate": 7.787589495296976e-05,
      "loss": 0.17367370426654816,
      "mean_token_accuracy": 0.9269566237926483,
      "num_tokens": 102939808.0,
      "step": 8345
    },
    {
      "entropy": 1.0728683471679688,
      "epoch": 4.394944707740916,
      "grad_norm": 0.313921719789505,
      "learning_rate": 7.785541204834232e-05,
      "loss": 0.17117363214492798,
      "mean_token_accuracy": 0.9291615635156631,
      "num_tokens": 102952144.0,
      "step": 8346
    },
    {
      "entropy": 1.1034887731075287,
      "epoch": 4.395471300684571,
      "grad_norm": 0.2988021671772003,
      "learning_rate": 7.783493105195172e-05,
      "loss": 0.16765370965003967,
      "mean_token_accuracy": 0.9321981817483902,
      "num_tokens": 102964480.0,
      "step": 8347
    },
    {
      "entropy": 1.0934976935386658,
      "epoch": 4.395997893628225,
      "grad_norm": 0.285640150308609,
      "learning_rate": 7.781445196501378e-05,
      "loss": 0.16364088654518127,
      "mean_token_accuracy": 0.929412230849266,
      "num_tokens": 102976816.0,
      "step": 8348
    },
    {
      "entropy": 1.1029529571533203,
      "epoch": 4.39652448657188,
      "grad_norm": 0.3293870985507965,
      "learning_rate": 7.779397478874424e-05,
      "loss": 0.17802828550338745,
      "mean_token_accuracy": 0.9208266884088516,
      "num_tokens": 102989152.0,
      "step": 8349
    },
    {
      "entropy": 1.1420040726661682,
      "epoch": 4.397051079515535,
      "grad_norm": 0.3254111409187317,
      "learning_rate": 7.777349952435873e-05,
      "loss": 0.1666165590286255,
      "mean_token_accuracy": 0.9263311624526978,
      "num_tokens": 103001488.0,
      "step": 8350
    },
    {
      "entropy": 1.0624985098838806,
      "epoch": 4.397577672459189,
      "grad_norm": 0.28192704916000366,
      "learning_rate": 7.775302617307271e-05,
      "loss": 0.16333523392677307,
      "mean_token_accuracy": 0.9288174957036972,
      "num_tokens": 103013824.0,
      "step": 8351
    },
    {
      "entropy": 1.101064771413803,
      "epoch": 4.398104265402844,
      "grad_norm": 0.2985323965549469,
      "learning_rate": 7.773255473610153e-05,
      "loss": 0.16419664025306702,
      "mean_token_accuracy": 0.9274439066648483,
      "num_tokens": 103026160.0,
      "step": 8352
    },
    {
      "entropy": 1.1078698635101318,
      "epoch": 4.398630858346499,
      "grad_norm": 0.29670053720474243,
      "learning_rate": 7.771208521466054e-05,
      "loss": 0.15546825528144836,
      "mean_token_accuracy": 0.9353156536817551,
      "num_tokens": 103038496.0,
      "step": 8353
    },
    {
      "entropy": 1.0689546167850494,
      "epoch": 4.399157451290153,
      "grad_norm": 0.29663437604904175,
      "learning_rate": 7.769161760996488e-05,
      "loss": 0.1630391925573349,
      "mean_token_accuracy": 0.9319390803575516,
      "num_tokens": 103050832.0,
      "step": 8354
    },
    {
      "entropy": 1.0939607620239258,
      "epoch": 4.399684044233807,
      "grad_norm": 0.3043624758720398,
      "learning_rate": 7.767115192322958e-05,
      "loss": 0.17216776311397552,
      "mean_token_accuracy": 0.9238497316837311,
      "num_tokens": 103063168.0,
      "step": 8355
    },
    {
      "entropy": 1.0829295814037323,
      "epoch": 4.400210637177461,
      "grad_norm": 0.2891440689563751,
      "learning_rate": 7.765068815566954e-05,
      "loss": 0.1584639698266983,
      "mean_token_accuracy": 0.9335407316684723,
      "num_tokens": 103075504.0,
      "step": 8356
    },
    {
      "entropy": 1.102199226617813,
      "epoch": 4.400737230121116,
      "grad_norm": 0.2820277810096741,
      "learning_rate": 7.763022630849958e-05,
      "loss": 0.1639498770236969,
      "mean_token_accuracy": 0.9334799498319626,
      "num_tokens": 103087840.0,
      "step": 8357
    },
    {
      "entropy": 1.14246666431427,
      "epoch": 4.401263823064771,
      "grad_norm": 0.322818785905838,
      "learning_rate": 7.760976638293447e-05,
      "loss": 0.17627064883708954,
      "mean_token_accuracy": 0.9249568432569504,
      "num_tokens": 103100176.0,
      "step": 8358
    },
    {
      "entropy": 1.1458443701267242,
      "epoch": 4.401790416008425,
      "grad_norm": 0.314328134059906,
      "learning_rate": 7.75893083801887e-05,
      "loss": 0.17229615151882172,
      "mean_token_accuracy": 0.9282985627651215,
      "num_tokens": 103112512.0,
      "step": 8359
    },
    {
      "entropy": 1.0990630984306335,
      "epoch": 4.40231700895208,
      "grad_norm": 0.3130625784397125,
      "learning_rate": 7.756885230147679e-05,
      "loss": 0.16665244102478027,
      "mean_token_accuracy": 0.931442990899086,
      "num_tokens": 103124848.0,
      "step": 8360
    },
    {
      "entropy": 1.187071055173874,
      "epoch": 4.402843601895735,
      "grad_norm": 0.3278498947620392,
      "learning_rate": 7.754839814801312e-05,
      "loss": 0.16480353474617004,
      "mean_token_accuracy": 0.9343954473733902,
      "num_tokens": 103137184.0,
      "step": 8361
    },
    {
      "entropy": 1.0789116024971008,
      "epoch": 4.403370194839389,
      "grad_norm": 0.31040650606155396,
      "learning_rate": 7.75279459210119e-05,
      "loss": 0.16955912113189697,
      "mean_token_accuracy": 0.9281894862651825,
      "num_tokens": 103149520.0,
      "step": 8362
    },
    {
      "entropy": 1.1937116384506226,
      "epoch": 4.403896787783044,
      "grad_norm": 0.39075520634651184,
      "learning_rate": 7.750749562168726e-05,
      "loss": 0.19120678305625916,
      "mean_token_accuracy": 0.9256240129470825,
      "num_tokens": 103161856.0,
      "step": 8363
    },
    {
      "entropy": 1.1735577285289764,
      "epoch": 4.404423380726699,
      "grad_norm": 0.37804561853408813,
      "learning_rate": 7.748704725125324e-05,
      "loss": 0.19023069739341736,
      "mean_token_accuracy": 0.9222802668809891,
      "num_tokens": 103174192.0,
      "step": 8364
    },
    {
      "entropy": 1.11763334274292,
      "epoch": 4.404949973670353,
      "grad_norm": 0.3547898530960083,
      "learning_rate": 7.746660081092374e-05,
      "loss": 0.18030297756195068,
      "mean_token_accuracy": 0.9217110425233841,
      "num_tokens": 103186528.0,
      "step": 8365
    },
    {
      "entropy": 1.1391336917877197,
      "epoch": 4.405476566614007,
      "grad_norm": 0.33012184500694275,
      "learning_rate": 7.744615630191246e-05,
      "loss": 0.1677279770374298,
      "mean_token_accuracy": 0.928357407450676,
      "num_tokens": 103198864.0,
      "step": 8366
    },
    {
      "entropy": 1.1174531877040863,
      "epoch": 4.406003159557662,
      "grad_norm": 0.32291877269744873,
      "learning_rate": 7.742571372543319e-05,
      "loss": 0.16517490148544312,
      "mean_token_accuracy": 0.9265066087245941,
      "num_tokens": 103211200.0,
      "step": 8367
    },
    {
      "entropy": 1.1210643649101257,
      "epoch": 4.406529752501316,
      "grad_norm": 0.3204652965068817,
      "learning_rate": 7.740527308269945e-05,
      "loss": 0.16921760141849518,
      "mean_token_accuracy": 0.9277530610561371,
      "num_tokens": 103223536.0,
      "step": 8368
    },
    {
      "entropy": 1.1069200932979584,
      "epoch": 4.407056345444971,
      "grad_norm": 0.31456953287124634,
      "learning_rate": 7.738483437492468e-05,
      "loss": 0.1544417142868042,
      "mean_token_accuracy": 0.9324262142181396,
      "num_tokens": 103235872.0,
      "step": 8369
    },
    {
      "entropy": 1.12002232670784,
      "epoch": 4.407582938388625,
      "grad_norm": 0.30929791927337646,
      "learning_rate": 7.736439760332217e-05,
      "loss": 0.16462525725364685,
      "mean_token_accuracy": 0.9314171969890594,
      "num_tokens": 103248208.0,
      "step": 8370
    },
    {
      "entropy": 1.1748793423175812,
      "epoch": 4.40810953133228,
      "grad_norm": 0.2985473573207855,
      "learning_rate": 7.734396276910518e-05,
      "loss": 0.1532450169324875,
      "mean_token_accuracy": 0.9394589960575104,
      "num_tokens": 103260544.0,
      "step": 8371
    },
    {
      "entropy": 1.1170945763587952,
      "epoch": 4.408636124275935,
      "grad_norm": 0.3181641399860382,
      "learning_rate": 7.732352987348677e-05,
      "loss": 0.15791934728622437,
      "mean_token_accuracy": 0.9288433194160461,
      "num_tokens": 103272880.0,
      "step": 8372
    },
    {
      "entropy": 1.187763661146164,
      "epoch": 4.409162717219589,
      "grad_norm": 0.34279096126556396,
      "learning_rate": 7.730309891767993e-05,
      "loss": 0.18462227284908295,
      "mean_token_accuracy": 0.9185385406017303,
      "num_tokens": 103285216.0,
      "step": 8373
    },
    {
      "entropy": 1.1743906438350677,
      "epoch": 4.409689310163244,
      "grad_norm": 0.3374393582344055,
      "learning_rate": 7.728266990289754e-05,
      "loss": 0.17026188969612122,
      "mean_token_accuracy": 0.9263552576303482,
      "num_tokens": 103297552.0,
      "step": 8374
    },
    {
      "entropy": 1.15870600938797,
      "epoch": 4.410215903106899,
      "grad_norm": 0.34642383456230164,
      "learning_rate": 7.726224283035235e-05,
      "loss": 0.17768001556396484,
      "mean_token_accuracy": 0.9258678555488586,
      "num_tokens": 103309888.0,
      "step": 8375
    },
    {
      "entropy": 1.179932415485382,
      "epoch": 4.410742496050553,
      "grad_norm": 0.3416505753993988,
      "learning_rate": 7.724181770125701e-05,
      "loss": 0.1815774142742157,
      "mean_token_accuracy": 0.9244155734777451,
      "num_tokens": 103322224.0,
      "step": 8376
    },
    {
      "entropy": 1.1441168785095215,
      "epoch": 4.411269088994207,
      "grad_norm": 0.34104859828948975,
      "learning_rate": 7.722139451682399e-05,
      "loss": 0.1751016527414322,
      "mean_token_accuracy": 0.9238454103469849,
      "num_tokens": 103334560.0,
      "step": 8377
    },
    {
      "entropy": 1.150140255689621,
      "epoch": 4.411795681937862,
      "grad_norm": 0.33411258459091187,
      "learning_rate": 7.720097327826577e-05,
      "loss": 0.169435054063797,
      "mean_token_accuracy": 0.9205000847578049,
      "num_tokens": 103346896.0,
      "step": 8378
    },
    {
      "entropy": 1.1098092794418335,
      "epoch": 4.412322274881516,
      "grad_norm": 0.28074219822883606,
      "learning_rate": 7.71805539867946e-05,
      "loss": 0.1499934196472168,
      "mean_token_accuracy": 0.9380492717027664,
      "num_tokens": 103359232.0,
      "step": 8379
    },
    {
      "entropy": 1.1838358938694,
      "epoch": 4.412848867825171,
      "grad_norm": 0.3145032823085785,
      "learning_rate": 7.716013664362258e-05,
      "loss": 0.16694800555706024,
      "mean_token_accuracy": 0.9277818500995636,
      "num_tokens": 103371568.0,
      "step": 8380
    },
    {
      "entropy": 1.1603033542633057,
      "epoch": 4.413375460768826,
      "grad_norm": 0.3271830081939697,
      "learning_rate": 7.713972124996192e-05,
      "loss": 0.17177407443523407,
      "mean_token_accuracy": 0.9279496967792511,
      "num_tokens": 103383904.0,
      "step": 8381
    },
    {
      "entropy": 1.1457706093788147,
      "epoch": 4.41390205371248,
      "grad_norm": 0.31502124667167664,
      "learning_rate": 7.711930780702445e-05,
      "loss": 0.16537320613861084,
      "mean_token_accuracy": 0.9297925382852554,
      "num_tokens": 103396240.0,
      "step": 8382
    },
    {
      "entropy": 1.1601044535636902,
      "epoch": 4.414428646656135,
      "grad_norm": 0.322947233915329,
      "learning_rate": 7.709889631602205e-05,
      "loss": 0.17435915768146515,
      "mean_token_accuracy": 0.9245762825012207,
      "num_tokens": 103408576.0,
      "step": 8383
    },
    {
      "entropy": 1.125311404466629,
      "epoch": 4.41495523959979,
      "grad_norm": 0.3117965757846832,
      "learning_rate": 7.707848677816642e-05,
      "loss": 0.1723887175321579,
      "mean_token_accuracy": 0.9227554500102997,
      "num_tokens": 103420912.0,
      "step": 8384
    },
    {
      "entropy": 1.1340227723121643,
      "epoch": 4.415481832543444,
      "grad_norm": 0.32244864106178284,
      "learning_rate": 7.705807919466914e-05,
      "loss": 0.1690177619457245,
      "mean_token_accuracy": 0.9253842532634735,
      "num_tokens": 103433248.0,
      "step": 8385
    },
    {
      "entropy": 1.14969140291214,
      "epoch": 4.416008425487099,
      "grad_norm": 0.32291510701179504,
      "learning_rate": 7.70376735667417e-05,
      "loss": 0.16696715354919434,
      "mean_token_accuracy": 0.9303669035434723,
      "num_tokens": 103445584.0,
      "step": 8386
    },
    {
      "entropy": 1.1150422990322113,
      "epoch": 4.4165350184307535,
      "grad_norm": 0.33403468132019043,
      "learning_rate": 7.701726989559544e-05,
      "loss": 0.1756865680217743,
      "mean_token_accuracy": 0.9245531111955643,
      "num_tokens": 103457920.0,
      "step": 8387
    },
    {
      "entropy": 1.1605649888515472,
      "epoch": 4.4170616113744074,
      "grad_norm": 0.3420243263244629,
      "learning_rate": 7.699686818244163e-05,
      "loss": 0.17954641580581665,
      "mean_token_accuracy": 0.9211717247962952,
      "num_tokens": 103470256.0,
      "step": 8388
    },
    {
      "entropy": 1.1440449059009552,
      "epoch": 4.417588204318062,
      "grad_norm": 0.335814505815506,
      "learning_rate": 7.697646842849142e-05,
      "loss": 0.17243674397468567,
      "mean_token_accuracy": 0.9276414960622787,
      "num_tokens": 103482592.0,
      "step": 8389
    },
    {
      "entropy": 1.1576780080795288,
      "epoch": 4.418114797261716,
      "grad_norm": 0.3149440586566925,
      "learning_rate": 7.695607063495578e-05,
      "loss": 0.16912463307380676,
      "mean_token_accuracy": 0.92839415371418,
      "num_tokens": 103494928.0,
      "step": 8390
    },
    {
      "entropy": 1.1885141134262085,
      "epoch": 4.418641390205371,
      "grad_norm": 0.3676983118057251,
      "learning_rate": 7.693567480304567e-05,
      "loss": 0.19174298644065857,
      "mean_token_accuracy": 0.9174454212188721,
      "num_tokens": 103507264.0,
      "step": 8391
    },
    {
      "entropy": 1.171632468700409,
      "epoch": 4.419167983149026,
      "grad_norm": 0.34532836079597473,
      "learning_rate": 7.69152809339718e-05,
      "loss": 0.1755070984363556,
      "mean_token_accuracy": 0.9248087704181671,
      "num_tokens": 103519600.0,
      "step": 8392
    },
    {
      "entropy": 1.1931438148021698,
      "epoch": 4.41969457609268,
      "grad_norm": 0.3367908298969269,
      "learning_rate": 7.689488902894489e-05,
      "loss": 0.17951609194278717,
      "mean_token_accuracy": 0.9275099188089371,
      "num_tokens": 103531936.0,
      "step": 8393
    },
    {
      "entropy": 1.1273884773254395,
      "epoch": 4.420221169036335,
      "grad_norm": 0.29583027958869934,
      "learning_rate": 7.687449908917543e-05,
      "loss": 0.15583457052707672,
      "mean_token_accuracy": 0.9299625754356384,
      "num_tokens": 103544272.0,
      "step": 8394
    },
    {
      "entropy": 1.1927360594272614,
      "epoch": 4.42074776197999,
      "grad_norm": 0.32044270634651184,
      "learning_rate": 7.68541111158739e-05,
      "loss": 0.17775748670101166,
      "mean_token_accuracy": 0.9299187660217285,
      "num_tokens": 103556608.0,
      "step": 8395
    },
    {
      "entropy": 1.1787251234054565,
      "epoch": 4.421274354923644,
      "grad_norm": 0.3348028063774109,
      "learning_rate": 7.683372511025064e-05,
      "loss": 0.17427965998649597,
      "mean_token_accuracy": 0.9278364479541779,
      "num_tokens": 103568944.0,
      "step": 8396
    },
    {
      "entropy": 1.1640098690986633,
      "epoch": 4.421800947867299,
      "grad_norm": 0.2946585416793823,
      "learning_rate": 7.681334107351578e-05,
      "loss": 0.1568325161933899,
      "mean_token_accuracy": 0.933339461684227,
      "num_tokens": 103581280.0,
      "step": 8397
    },
    {
      "entropy": 1.2109858095645905,
      "epoch": 4.4223275408109535,
      "grad_norm": 0.3569537103176117,
      "learning_rate": 7.679295900687946e-05,
      "loss": 0.19348572194576263,
      "mean_token_accuracy": 0.9151138514280319,
      "num_tokens": 103593616.0,
      "step": 8398
    },
    {
      "entropy": 1.183549165725708,
      "epoch": 4.4228541337546075,
      "grad_norm": 0.3093925416469574,
      "learning_rate": 7.677257891155163e-05,
      "loss": 0.15875285863876343,
      "mean_token_accuracy": 0.9322007298469543,
      "num_tokens": 103605952.0,
      "step": 8399
    },
    {
      "entropy": 1.1850107312202454,
      "epoch": 4.423380726698262,
      "grad_norm": 0.29908472299575806,
      "learning_rate": 7.67522007887421e-05,
      "loss": 0.16990849375724792,
      "mean_token_accuracy": 0.9256811887025833,
      "num_tokens": 103618288.0,
      "step": 8400
    },
    {
      "entropy": 1.1876828074455261,
      "epoch": 4.423907319641917,
      "grad_norm": 0.31594136357307434,
      "learning_rate": 7.673182463966057e-05,
      "loss": 0.16882580518722534,
      "mean_token_accuracy": 0.930813804268837,
      "num_tokens": 103630624.0,
      "step": 8401
    },
    {
      "entropy": 1.1995342373847961,
      "epoch": 4.424433912585571,
      "grad_norm": 0.30118241906166077,
      "learning_rate": 7.67114504655168e-05,
      "loss": 0.17157086730003357,
      "mean_token_accuracy": 0.9306039363145828,
      "num_tokens": 103642960.0,
      "step": 8402
    },
    {
      "entropy": 1.2159958481788635,
      "epoch": 4.424960505529226,
      "grad_norm": 0.3153426945209503,
      "learning_rate": 7.669107826752015e-05,
      "loss": 0.16894468665122986,
      "mean_token_accuracy": 0.9226904064416885,
      "num_tokens": 103655296.0,
      "step": 8403
    },
    {
      "entropy": 1.1882687211036682,
      "epoch": 4.42548709847288,
      "grad_norm": 0.3680673837661743,
      "learning_rate": 7.667070804688006e-05,
      "loss": 0.19156639277935028,
      "mean_token_accuracy": 0.9129916727542877,
      "num_tokens": 103667632.0,
      "step": 8404
    },
    {
      "entropy": 1.1941133439540863,
      "epoch": 4.426013691416535,
      "grad_norm": 0.3307591676712036,
      "learning_rate": 7.665033980480579e-05,
      "loss": 0.18654422461986542,
      "mean_token_accuracy": 0.923293262720108,
      "num_tokens": 103679968.0,
      "step": 8405
    },
    {
      "entropy": 1.186953067779541,
      "epoch": 4.42654028436019,
      "grad_norm": 0.3209822475910187,
      "learning_rate": 7.662997354250645e-05,
      "loss": 0.18281899392604828,
      "mean_token_accuracy": 0.9238188564777374,
      "num_tokens": 103692304.0,
      "step": 8406
    },
    {
      "entropy": 1.1703291535377502,
      "epoch": 4.427066877303844,
      "grad_norm": 0.31133076548576355,
      "learning_rate": 7.660960926119107e-05,
      "loss": 0.17949330806732178,
      "mean_token_accuracy": 0.919961005449295,
      "num_tokens": 103704640.0,
      "step": 8407
    },
    {
      "entropy": 1.1952329874038696,
      "epoch": 4.427593470247499,
      "grad_norm": 0.32679304480552673,
      "learning_rate": 7.65892469620686e-05,
      "loss": 0.1816975623369217,
      "mean_token_accuracy": 0.9183577597141266,
      "num_tokens": 103716976.0,
      "step": 8408
    },
    {
      "entropy": 1.1395338773727417,
      "epoch": 4.4281200631911535,
      "grad_norm": 0.28885093331336975,
      "learning_rate": 7.656888664634778e-05,
      "loss": 0.17022892832756042,
      "mean_token_accuracy": 0.9255540519952774,
      "num_tokens": 103729312.0,
      "step": 8409
    },
    {
      "entropy": 1.18154376745224,
      "epoch": 4.4286466561348075,
      "grad_norm": 0.30899184942245483,
      "learning_rate": 7.654852831523732e-05,
      "loss": 0.17996257543563843,
      "mean_token_accuracy": 0.924812838435173,
      "num_tokens": 103741648.0,
      "step": 8410
    },
    {
      "entropy": 1.1718483865261078,
      "epoch": 4.429173249078462,
      "grad_norm": 0.3028598725795746,
      "learning_rate": 7.652817196994577e-05,
      "loss": 0.15958929061889648,
      "mean_token_accuracy": 0.9312357604503632,
      "num_tokens": 103753984.0,
      "step": 8411
    },
    {
      "entropy": 1.2056899666786194,
      "epoch": 4.429699842022117,
      "grad_norm": 0.34338873624801636,
      "learning_rate": 7.650781761168157e-05,
      "loss": 0.18257784843444824,
      "mean_token_accuracy": 0.9223008453845978,
      "num_tokens": 103766320.0,
      "step": 8412
    },
    {
      "entropy": 1.1947410106658936,
      "epoch": 4.430226434965771,
      "grad_norm": 0.33735391497612,
      "learning_rate": 7.648746524165302e-05,
      "loss": 0.1745343655347824,
      "mean_token_accuracy": 0.927484929561615,
      "num_tokens": 103778656.0,
      "step": 8413
    },
    {
      "entropy": 1.1994907557964325,
      "epoch": 4.430753027909426,
      "grad_norm": 0.31284675002098083,
      "learning_rate": 7.646711486106834e-05,
      "loss": 0.1723497062921524,
      "mean_token_accuracy": 0.9282457530498505,
      "num_tokens": 103790992.0,
      "step": 8414
    },
    {
      "entropy": 1.173362910747528,
      "epoch": 4.431279620853081,
      "grad_norm": 0.2921752333641052,
      "learning_rate": 7.644676647113558e-05,
      "loss": 0.15854358673095703,
      "mean_token_accuracy": 0.9307354688644409,
      "num_tokens": 103803328.0,
      "step": 8415
    },
    {
      "entropy": 1.1649015545845032,
      "epoch": 4.431806213796735,
      "grad_norm": 0.3033360540866852,
      "learning_rate": 7.642642007306275e-05,
      "loss": 0.17482881247997284,
      "mean_token_accuracy": 0.9215189963579178,
      "num_tokens": 103815664.0,
      "step": 8416
    },
    {
      "entropy": 1.1332653760910034,
      "epoch": 4.43233280674039,
      "grad_norm": 0.294369637966156,
      "learning_rate": 7.640607566805769e-05,
      "loss": 0.16559576988220215,
      "mean_token_accuracy": 0.9328326433897018,
      "num_tokens": 103828000.0,
      "step": 8417
    },
    {
      "entropy": 1.1266019940376282,
      "epoch": 4.432859399684045,
      "grad_norm": 0.2806363105773926,
      "learning_rate": 7.638573325732811e-05,
      "loss": 0.16257208585739136,
      "mean_token_accuracy": 0.9307931661605835,
      "num_tokens": 103840336.0,
      "step": 8418
    },
    {
      "entropy": 1.1631579399108887,
      "epoch": 4.433385992627699,
      "grad_norm": 0.3195684254169464,
      "learning_rate": 7.636539284208162e-05,
      "loss": 0.18540135025978088,
      "mean_token_accuracy": 0.9234038889408112,
      "num_tokens": 103852672.0,
      "step": 8419
    },
    {
      "entropy": 1.1342191100120544,
      "epoch": 4.4339125855713535,
      "grad_norm": 0.30643901228904724,
      "learning_rate": 7.634505442352573e-05,
      "loss": 0.162317156791687,
      "mean_token_accuracy": 0.9350567162036896,
      "num_tokens": 103865008.0,
      "step": 8420
    },
    {
      "entropy": 1.1070061326026917,
      "epoch": 4.434439178515008,
      "grad_norm": 0.2911212742328644,
      "learning_rate": 7.63247180028678e-05,
      "loss": 0.16851764917373657,
      "mean_token_accuracy": 0.930103987455368,
      "num_tokens": 103877344.0,
      "step": 8421
    },
    {
      "entropy": 1.1462253034114838,
      "epoch": 4.434965771458662,
      "grad_norm": 0.29653191566467285,
      "learning_rate": 7.630438358131506e-05,
      "loss": 0.1634315401315689,
      "mean_token_accuracy": 0.9309469014406204,
      "num_tokens": 103889680.0,
      "step": 8422
    },
    {
      "entropy": 1.1170426607131958,
      "epoch": 4.435492364402317,
      "grad_norm": 0.28300195932388306,
      "learning_rate": 7.628405116007472e-05,
      "loss": 0.15759967267513275,
      "mean_token_accuracy": 0.9321088343858719,
      "num_tokens": 103902016.0,
      "step": 8423
    },
    {
      "entropy": 1.1150572001934052,
      "epoch": 4.436018957345971,
      "grad_norm": 0.31304875016212463,
      "learning_rate": 7.626372074035371e-05,
      "loss": 0.17278333008289337,
      "mean_token_accuracy": 0.9233694821596146,
      "num_tokens": 103914352.0,
      "step": 8424
    },
    {
      "entropy": 1.1611599922180176,
      "epoch": 4.436545550289626,
      "grad_norm": 0.31712397933006287,
      "learning_rate": 7.6243392323359e-05,
      "loss": 0.16291208565235138,
      "mean_token_accuracy": 0.9327135235071182,
      "num_tokens": 103926688.0,
      "step": 8425
    },
    {
      "entropy": 1.1345867216587067,
      "epoch": 4.437072143233281,
      "grad_norm": 0.32005640864372253,
      "learning_rate": 7.622306591029733e-05,
      "loss": 0.17034733295440674,
      "mean_token_accuracy": 0.9302045255899429,
      "num_tokens": 103939024.0,
      "step": 8426
    },
    {
      "entropy": 1.130208432674408,
      "epoch": 4.437598736176935,
      "grad_norm": 0.32117676734924316,
      "learning_rate": 7.620274150237537e-05,
      "loss": 0.1721271127462387,
      "mean_token_accuracy": 0.9290124922990799,
      "num_tokens": 103951360.0,
      "step": 8427
    },
    {
      "entropy": 1.1899872124195099,
      "epoch": 4.43812532912059,
      "grad_norm": 0.3016345202922821,
      "learning_rate": 7.618241910079969e-05,
      "loss": 0.15409058332443237,
      "mean_token_accuracy": 0.936784103512764,
      "num_tokens": 103963696.0,
      "step": 8428
    },
    {
      "entropy": 1.1210709512233734,
      "epoch": 4.438651922064245,
      "grad_norm": 0.32231977581977844,
      "learning_rate": 7.616209870677659e-05,
      "loss": 0.1683426797389984,
      "mean_token_accuracy": 0.926812157034874,
      "num_tokens": 103976032.0,
      "step": 8429
    },
    {
      "entropy": 1.1640322506427765,
      "epoch": 4.439178515007899,
      "grad_norm": 0.3329298198223114,
      "learning_rate": 7.614178032151255e-05,
      "loss": 0.17446720600128174,
      "mean_token_accuracy": 0.924039363861084,
      "num_tokens": 103988368.0,
      "step": 8430
    },
    {
      "entropy": 1.1792564988136292,
      "epoch": 4.4397051079515535,
      "grad_norm": 0.3173338770866394,
      "learning_rate": 7.612146394621364e-05,
      "loss": 0.17538343369960785,
      "mean_token_accuracy": 0.9262731224298477,
      "num_tokens": 104000704.0,
      "step": 8431
    },
    {
      "entropy": 1.1290611922740936,
      "epoch": 4.440231700895208,
      "grad_norm": 0.3201102912425995,
      "learning_rate": 7.610114958208601e-05,
      "loss": 0.17354167997837067,
      "mean_token_accuracy": 0.9299298077821732,
      "num_tokens": 104013040.0,
      "step": 8432
    },
    {
      "entropy": 1.169890135526657,
      "epoch": 4.440758293838862,
      "grad_norm": 0.42167460918426514,
      "learning_rate": 7.608083723033552e-05,
      "loss": 0.19123922288417816,
      "mean_token_accuracy": 0.9196564853191376,
      "num_tokens": 104025376.0,
      "step": 8433
    },
    {
      "entropy": 1.1458514034748077,
      "epoch": 4.441284886782517,
      "grad_norm": 0.31748509407043457,
      "learning_rate": 7.6060526892168e-05,
      "loss": 0.1763705015182495,
      "mean_token_accuracy": 0.9211138784885406,
      "num_tokens": 104037712.0,
      "step": 8434
    },
    {
      "entropy": 1.1005293428897858,
      "epoch": 4.441811479726172,
      "grad_norm": 0.29628878831863403,
      "learning_rate": 7.604021856878922e-05,
      "loss": 0.15931612253189087,
      "mean_token_accuracy": 0.9339651465415955,
      "num_tokens": 104050048.0,
      "step": 8435
    },
    {
      "entropy": 1.152970939874649,
      "epoch": 4.442338072669826,
      "grad_norm": 0.3029753863811493,
      "learning_rate": 7.601991226140467e-05,
      "loss": 0.1684556007385254,
      "mean_token_accuracy": 0.927347719669342,
      "num_tokens": 104062384.0,
      "step": 8436
    },
    {
      "entropy": 1.1114393770694733,
      "epoch": 4.442864665613481,
      "grad_norm": 0.34585103392601013,
      "learning_rate": 7.599960797121995e-05,
      "loss": 0.17668956518173218,
      "mean_token_accuracy": 0.9263940453529358,
      "num_tokens": 104074720.0,
      "step": 8437
    },
    {
      "entropy": 1.1421588361263275,
      "epoch": 4.443391258557135,
      "grad_norm": 0.30870088934898376,
      "learning_rate": 7.597930569944029e-05,
      "loss": 0.1706380546092987,
      "mean_token_accuracy": 0.9274469017982483,
      "num_tokens": 104087056.0,
      "step": 8438
    },
    {
      "entropy": 1.1632334291934967,
      "epoch": 4.44391785150079,
      "grad_norm": 0.323515921831131,
      "learning_rate": 7.5959005447271e-05,
      "loss": 0.1692359447479248,
      "mean_token_accuracy": 0.9250432997941971,
      "num_tokens": 104099392.0,
      "step": 8439
    },
    {
      "entropy": 1.1219305992126465,
      "epoch": 4.444444444444445,
      "grad_norm": 0.30658406019210815,
      "learning_rate": 7.593870721591714e-05,
      "loss": 0.15897363424301147,
      "mean_token_accuracy": 0.9280763119459152,
      "num_tokens": 104111728.0,
      "step": 8440
    },
    {
      "entropy": 1.1651221811771393,
      "epoch": 4.444971037388099,
      "grad_norm": 0.31844380497932434,
      "learning_rate": 7.591841100658371e-05,
      "loss": 0.16563773155212402,
      "mean_token_accuracy": 0.9349388629198074,
      "num_tokens": 104124064.0,
      "step": 8441
    },
    {
      "entropy": 1.1618649363517761,
      "epoch": 4.4454976303317535,
      "grad_norm": 0.33157169818878174,
      "learning_rate": 7.589811682047556e-05,
      "loss": 0.17135582864284515,
      "mean_token_accuracy": 0.9291810691356659,
      "num_tokens": 104136400.0,
      "step": 8442
    },
    {
      "entropy": 1.146396517753601,
      "epoch": 4.446024223275408,
      "grad_norm": 0.30438268184661865,
      "learning_rate": 7.587782465879743e-05,
      "loss": 0.16211694478988647,
      "mean_token_accuracy": 0.9298663437366486,
      "num_tokens": 104148736.0,
      "step": 8443
    },
    {
      "entropy": 1.1388017237186432,
      "epoch": 4.446550816219062,
      "grad_norm": 0.3000304400920868,
      "learning_rate": 7.585753452275401e-05,
      "loss": 0.170515239238739,
      "mean_token_accuracy": 0.9270701557397842,
      "num_tokens": 104161072.0,
      "step": 8444
    },
    {
      "entropy": 1.117406189441681,
      "epoch": 4.447077409162717,
      "grad_norm": 0.2809015214443207,
      "learning_rate": 7.583724641354975e-05,
      "loss": 0.1588825285434723,
      "mean_token_accuracy": 0.9333102852106094,
      "num_tokens": 104173408.0,
      "step": 8445
    },
    {
      "entropy": 1.1422425508499146,
      "epoch": 4.447604002106372,
      "grad_norm": 0.2744351029396057,
      "learning_rate": 7.581696033238907e-05,
      "loss": 0.14139851927757263,
      "mean_token_accuracy": 0.9374795705080032,
      "num_tokens": 104185744.0,
      "step": 8446
    },
    {
      "entropy": 1.1064694821834564,
      "epoch": 4.448130595050026,
      "grad_norm": 0.31977495551109314,
      "learning_rate": 7.579667628047622e-05,
      "loss": 0.17230455577373505,
      "mean_token_accuracy": 0.9253697097301483,
      "num_tokens": 104198080.0,
      "step": 8447
    },
    {
      "entropy": 1.1942217946052551,
      "epoch": 4.448657187993681,
      "grad_norm": 0.3635144531726837,
      "learning_rate": 7.577639425901534e-05,
      "loss": 0.17706610262393951,
      "mean_token_accuracy": 0.926745280623436,
      "num_tokens": 104210416.0,
      "step": 8448
    },
    {
      "entropy": 1.1594175398349762,
      "epoch": 4.449183780937336,
      "grad_norm": 0.33133333921432495,
      "learning_rate": 7.575611426921048e-05,
      "loss": 0.16024044156074524,
      "mean_token_accuracy": 0.9303994327783585,
      "num_tokens": 104222752.0,
      "step": 8449
    },
    {
      "entropy": 1.1542670130729675,
      "epoch": 4.44971037388099,
      "grad_norm": 0.3534603416919708,
      "learning_rate": 7.573583631226546e-05,
      "loss": 0.19002632796764374,
      "mean_token_accuracy": 0.9212928414344788,
      "num_tokens": 104235088.0,
      "step": 8450
    },
    {
      "entropy": 1.10170778632164,
      "epoch": 4.450236966824645,
      "grad_norm": 0.32598719000816345,
      "learning_rate": 7.57155603893842e-05,
      "loss": 0.17832794785499573,
      "mean_token_accuracy": 0.9255920499563217,
      "num_tokens": 104247424.0,
      "step": 8451
    },
    {
      "entropy": 1.148159921169281,
      "epoch": 4.4507635597682995,
      "grad_norm": 0.33278730511665344,
      "learning_rate": 7.569528650177029e-05,
      "loss": 0.1623687446117401,
      "mean_token_accuracy": 0.932820126414299,
      "num_tokens": 104259760.0,
      "step": 8452
    },
    {
      "entropy": 1.1053113639354706,
      "epoch": 4.4512901527119535,
      "grad_norm": 0.3084471523761749,
      "learning_rate": 7.567501465062728e-05,
      "loss": 0.16041335463523865,
      "mean_token_accuracy": 0.9310987293720245,
      "num_tokens": 104272096.0,
      "step": 8453
    },
    {
      "entropy": 1.2006906270980835,
      "epoch": 4.451816745655608,
      "grad_norm": 0.4248177707195282,
      "learning_rate": 7.565474483715858e-05,
      "loss": 0.17785029113292694,
      "mean_token_accuracy": 0.9251182526350021,
      "num_tokens": 104284432.0,
      "step": 8454
    },
    {
      "entropy": 1.0966787934303284,
      "epoch": 4.452343338599263,
      "grad_norm": 0.3300796449184418,
      "learning_rate": 7.563447706256747e-05,
      "loss": 0.1696946769952774,
      "mean_token_accuracy": 0.9282518625259399,
      "num_tokens": 104296768.0,
      "step": 8455
    },
    {
      "entropy": 1.1123622953891754,
      "epoch": 4.452869931542917,
      "grad_norm": 0.3217577040195465,
      "learning_rate": 7.56142113280572e-05,
      "loss": 0.17596352100372314,
      "mean_token_accuracy": 0.9228823781013489,
      "num_tokens": 104309104.0,
      "step": 8456
    },
    {
      "entropy": 1.1256005465984344,
      "epoch": 4.453396524486572,
      "grad_norm": 0.3379938304424286,
      "learning_rate": 7.559394763483075e-05,
      "loss": 0.16395236551761627,
      "mean_token_accuracy": 0.9304747730493546,
      "num_tokens": 104321440.0,
      "step": 8457
    },
    {
      "entropy": 1.1497615575790405,
      "epoch": 4.453923117430226,
      "grad_norm": 0.33467888832092285,
      "learning_rate": 7.557368598409111e-05,
      "loss": 0.17233118414878845,
      "mean_token_accuracy": 0.9285077452659607,
      "num_tokens": 104333776.0,
      "step": 8458
    },
    {
      "entropy": 1.1492519974708557,
      "epoch": 4.454449710373881,
      "grad_norm": 0.32112306356430054,
      "learning_rate": 7.555342637704108e-05,
      "loss": 0.178175151348114,
      "mean_token_accuracy": 0.9256938993930817,
      "num_tokens": 104346112.0,
      "step": 8459
    },
    {
      "entropy": 1.1114400625228882,
      "epoch": 4.454976303317536,
      "grad_norm": 0.31285762786865234,
      "learning_rate": 7.553316881488337e-05,
      "loss": 0.16353988647460938,
      "mean_token_accuracy": 0.9300347864627838,
      "num_tokens": 104358448.0,
      "step": 8460
    },
    {
      "entropy": 1.1686404645442963,
      "epoch": 4.45550289626119,
      "grad_norm": 0.3369942307472229,
      "learning_rate": 7.551291329882053e-05,
      "loss": 0.17610526084899902,
      "mean_token_accuracy": 0.923306405544281,
      "num_tokens": 104370784.0,
      "step": 8461
    },
    {
      "entropy": 1.1252988278865814,
      "epoch": 4.456029489204845,
      "grad_norm": 0.3011777997016907,
      "learning_rate": 7.549265983005504e-05,
      "loss": 0.16479860246181488,
      "mean_token_accuracy": 0.9301632195711136,
      "num_tokens": 104383120.0,
      "step": 8462
    },
    {
      "entropy": 1.2057690620422363,
      "epoch": 4.4565560821484995,
      "grad_norm": 0.33670660853385925,
      "learning_rate": 7.547240840978922e-05,
      "loss": 0.1779303252696991,
      "mean_token_accuracy": 0.9197793751955032,
      "num_tokens": 104395456.0,
      "step": 8463
    },
    {
      "entropy": 1.1532905399799347,
      "epoch": 4.4570826750921535,
      "grad_norm": 0.315647691488266,
      "learning_rate": 7.545215903922524e-05,
      "loss": 0.1743641197681427,
      "mean_token_accuracy": 0.9221219271421432,
      "num_tokens": 104407792.0,
      "step": 8464
    },
    {
      "entropy": 1.1454414129257202,
      "epoch": 4.457609268035808,
      "grad_norm": 0.3356110155582428,
      "learning_rate": 7.543191171956519e-05,
      "loss": 0.18045343458652496,
      "mean_token_accuracy": 0.9212951511144638,
      "num_tokens": 104420128.0,
      "step": 8465
    },
    {
      "entropy": 1.1666088700294495,
      "epoch": 4.458135860979463,
      "grad_norm": 0.31954318284988403,
      "learning_rate": 7.541166645201112e-05,
      "loss": 0.1708226501941681,
      "mean_token_accuracy": 0.9309587478637695,
      "num_tokens": 104432464.0,
      "step": 8466
    },
    {
      "entropy": 1.1739060282707214,
      "epoch": 4.458662453923117,
      "grad_norm": 0.32615551352500916,
      "learning_rate": 7.539142323776478e-05,
      "loss": 0.1824362874031067,
      "mean_token_accuracy": 0.9265118688344955,
      "num_tokens": 104444800.0,
      "step": 8467
    },
    {
      "entropy": 1.1715387403964996,
      "epoch": 4.459189046866772,
      "grad_norm": 0.3394647538661957,
      "learning_rate": 7.537118207802796e-05,
      "loss": 0.18361705541610718,
      "mean_token_accuracy": 0.9243173897266388,
      "num_tokens": 104457136.0,
      "step": 8468
    },
    {
      "entropy": 1.125695139169693,
      "epoch": 4.459715639810426,
      "grad_norm": 0.336599200963974,
      "learning_rate": 7.53509429740022e-05,
      "loss": 0.17182287573814392,
      "mean_token_accuracy": 0.9275820404291153,
      "num_tokens": 104469472.0,
      "step": 8469
    },
    {
      "entropy": 1.167458951473236,
      "epoch": 4.460242232754081,
      "grad_norm": 0.30065464973449707,
      "learning_rate": 7.533070592688902e-05,
      "loss": 0.1620563268661499,
      "mean_token_accuracy": 0.9307129085063934,
      "num_tokens": 104481808.0,
      "step": 8470
    },
    {
      "entropy": 1.1041797399520874,
      "epoch": 4.460768825697736,
      "grad_norm": 0.3129633665084839,
      "learning_rate": 7.531047093788977e-05,
      "loss": 0.16948609054088593,
      "mean_token_accuracy": 0.931243821978569,
      "num_tokens": 104494144.0,
      "step": 8471
    },
    {
      "entropy": 1.201309859752655,
      "epoch": 4.46129541864139,
      "grad_norm": 0.3466167151927948,
      "learning_rate": 7.529023800820563e-05,
      "loss": 0.18264959752559662,
      "mean_token_accuracy": 0.9177737534046173,
      "num_tokens": 104506480.0,
      "step": 8472
    },
    {
      "entropy": 1.173547476530075,
      "epoch": 4.461822011585045,
      "grad_norm": 0.31830111145973206,
      "learning_rate": 7.52700071390378e-05,
      "loss": 0.18476280570030212,
      "mean_token_accuracy": 0.9239428043365479,
      "num_tokens": 104518816.0,
      "step": 8473
    },
    {
      "entropy": 1.1615335941314697,
      "epoch": 4.4623486045286995,
      "grad_norm": 0.31656646728515625,
      "learning_rate": 7.524977833158722e-05,
      "loss": 0.1678553819656372,
      "mean_token_accuracy": 0.9285225719213486,
      "num_tokens": 104531152.0,
      "step": 8474
    },
    {
      "entropy": 1.1629353165626526,
      "epoch": 4.4628751974723535,
      "grad_norm": 0.29458194971084595,
      "learning_rate": 7.522955158705474e-05,
      "loss": 0.1683681160211563,
      "mean_token_accuracy": 0.9332570284605026,
      "num_tokens": 104543488.0,
      "step": 8475
    },
    {
      "entropy": 1.1883186399936676,
      "epoch": 4.463401790416008,
      "grad_norm": 0.34101203083992004,
      "learning_rate": 7.520932690664115e-05,
      "loss": 0.16952675580978394,
      "mean_token_accuracy": 0.9276389330625534,
      "num_tokens": 104555824.0,
      "step": 8476
    },
    {
      "entropy": 1.1620250940322876,
      "epoch": 4.463928383359663,
      "grad_norm": 0.2982504069805145,
      "learning_rate": 7.518910429154704e-05,
      "loss": 0.15610679984092712,
      "mean_token_accuracy": 0.9357586652040482,
      "num_tokens": 104568160.0,
      "step": 8477
    },
    {
      "entropy": 1.197614997625351,
      "epoch": 4.464454976303317,
      "grad_norm": 0.30682411789894104,
      "learning_rate": 7.516888374297291e-05,
      "loss": 0.1719253957271576,
      "mean_token_accuracy": 0.9263016879558563,
      "num_tokens": 104580496.0,
      "step": 8478
    },
    {
      "entropy": 1.1783290803432465,
      "epoch": 4.464981569246972,
      "grad_norm": 0.3327757716178894,
      "learning_rate": 7.514866526211912e-05,
      "loss": 0.17140227556228638,
      "mean_token_accuracy": 0.9275024682283401,
      "num_tokens": 104592832.0,
      "step": 8479
    },
    {
      "entropy": 1.162148356437683,
      "epoch": 4.465508162190627,
      "grad_norm": 0.3191398084163666,
      "learning_rate": 7.512844885018594e-05,
      "loss": 0.1670476347208023,
      "mean_token_accuracy": 0.9323391020298004,
      "num_tokens": 104605168.0,
      "step": 8480
    },
    {
      "entropy": 1.1739769876003265,
      "epoch": 4.466034755134281,
      "grad_norm": 0.3413907587528229,
      "learning_rate": 7.510823450837356e-05,
      "loss": 0.18024437129497528,
      "mean_token_accuracy": 0.9216061383485794,
      "num_tokens": 104617504.0,
      "step": 8481
    },
    {
      "entropy": 1.162779986858368,
      "epoch": 4.466561348077936,
      "grad_norm": 0.3138342797756195,
      "learning_rate": 7.508802223788189e-05,
      "loss": 0.16494205594062805,
      "mean_token_accuracy": 0.9311954081058502,
      "num_tokens": 104629840.0,
      "step": 8482
    },
    {
      "entropy": 1.1326565444469452,
      "epoch": 4.467087941021591,
      "grad_norm": 0.3279440999031067,
      "learning_rate": 7.506781203991089e-05,
      "loss": 0.17870019376277924,
      "mean_token_accuracy": 0.9224427193403244,
      "num_tokens": 104642176.0,
      "step": 8483
    },
    {
      "entropy": 1.1368190050125122,
      "epoch": 4.467614533965245,
      "grad_norm": 0.3391387462615967,
      "learning_rate": 7.504760391566025e-05,
      "loss": 0.17431044578552246,
      "mean_token_accuracy": 0.9252381324768066,
      "num_tokens": 104654512.0,
      "step": 8484
    },
    {
      "entropy": 1.1795203983783722,
      "epoch": 4.4681411269088995,
      "grad_norm": 0.3195769488811493,
      "learning_rate": 7.502739786632968e-05,
      "loss": 0.1850837767124176,
      "mean_token_accuracy": 0.9250011593103409,
      "num_tokens": 104666848.0,
      "step": 8485
    },
    {
      "entropy": 1.0992188155651093,
      "epoch": 4.468667719852554,
      "grad_norm": 0.2884109318256378,
      "learning_rate": 7.500719389311864e-05,
      "loss": 0.16247399151325226,
      "mean_token_accuracy": 0.9310379177331924,
      "num_tokens": 104679184.0,
      "step": 8486
    },
    {
      "entropy": 1.1694341599941254,
      "epoch": 4.469194312796208,
      "grad_norm": 0.35028311610221863,
      "learning_rate": 7.498699199722654e-05,
      "loss": 0.1847420036792755,
      "mean_token_accuracy": 0.9226665943861008,
      "num_tokens": 104691520.0,
      "step": 8487
    },
    {
      "entropy": 1.1405926644802094,
      "epoch": 4.469720905739863,
      "grad_norm": 0.2975562810897827,
      "learning_rate": 7.496679217985267e-05,
      "loss": 0.1603151559829712,
      "mean_token_accuracy": 0.9311546981334686,
      "num_tokens": 104703856.0,
      "step": 8488
    },
    {
      "entropy": 1.1884474158287048,
      "epoch": 4.470247498683517,
      "grad_norm": 0.3177323341369629,
      "learning_rate": 7.494659444219616e-05,
      "loss": 0.15342968702316284,
      "mean_token_accuracy": 0.9374910742044449,
      "num_tokens": 104716192.0,
      "step": 8489
    },
    {
      "entropy": 1.1844165325164795,
      "epoch": 4.470774091627172,
      "grad_norm": 0.32195863127708435,
      "learning_rate": 7.492639878545604e-05,
      "loss": 0.17014195024967194,
      "mean_token_accuracy": 0.9291600584983826,
      "num_tokens": 104728528.0,
      "step": 8490
    },
    {
      "entropy": 1.158256620168686,
      "epoch": 4.471300684570827,
      "grad_norm": 0.29549679160118103,
      "learning_rate": 7.490620521083119e-05,
      "loss": 0.1600731909275055,
      "mean_token_accuracy": 0.9307269752025604,
      "num_tokens": 104740864.0,
      "step": 8491
    },
    {
      "entropy": 1.1640253365039825,
      "epoch": 4.471827277514481,
      "grad_norm": 0.3163624703884125,
      "learning_rate": 7.48860137195204e-05,
      "loss": 0.1576489955186844,
      "mean_token_accuracy": 0.9304119348526001,
      "num_tokens": 104753200.0,
      "step": 8492
    },
    {
      "entropy": 1.1176630556583405,
      "epoch": 4.472353870458136,
      "grad_norm": 0.3096276521682739,
      "learning_rate": 7.486582431272227e-05,
      "loss": 0.17143644392490387,
      "mean_token_accuracy": 0.9221581369638443,
      "num_tokens": 104765536.0,
      "step": 8493
    },
    {
      "entropy": 1.1509735584259033,
      "epoch": 4.472880463401791,
      "grad_norm": 0.31744223833084106,
      "learning_rate": 7.48456369916354e-05,
      "loss": 0.18509355187416077,
      "mean_token_accuracy": 0.9221777617931366,
      "num_tokens": 104777872.0,
      "step": 8494
    },
    {
      "entropy": 1.1565062999725342,
      "epoch": 4.473407056345445,
      "grad_norm": 0.32583150267601013,
      "learning_rate": 7.482545175745819e-05,
      "loss": 0.16323620080947876,
      "mean_token_accuracy": 0.9322451949119568,
      "num_tokens": 104790208.0,
      "step": 8495
    },
    {
      "entropy": 1.1247378885746002,
      "epoch": 4.4739336492890995,
      "grad_norm": 0.28839007019996643,
      "learning_rate": 7.48052686113889e-05,
      "loss": 0.16859430074691772,
      "mean_token_accuracy": 0.926827684044838,
      "num_tokens": 104802544.0,
      "step": 8496
    },
    {
      "entropy": 1.1516145467758179,
      "epoch": 4.474460242232754,
      "grad_norm": 0.32340723276138306,
      "learning_rate": 7.478508755462568e-05,
      "loss": 0.168523907661438,
      "mean_token_accuracy": 0.925041526556015,
      "num_tokens": 104814880.0,
      "step": 8497
    },
    {
      "entropy": 1.1144302487373352,
      "epoch": 4.474986835176408,
      "grad_norm": 0.29585084319114685,
      "learning_rate": 7.476490858836655e-05,
      "loss": 0.1568192094564438,
      "mean_token_accuracy": 0.9378913789987564,
      "num_tokens": 104827216.0,
      "step": 8498
    },
    {
      "entropy": 1.124032586812973,
      "epoch": 4.475513428120063,
      "grad_norm": 0.3131750226020813,
      "learning_rate": 7.474473171380942e-05,
      "loss": 0.17422203719615936,
      "mean_token_accuracy": 0.9270296394824982,
      "num_tokens": 104839552.0,
      "step": 8499
    },
    {
      "entropy": 1.170444518327713,
      "epoch": 4.476040021063718,
      "grad_norm": 0.3095943331718445,
      "learning_rate": 7.472455693215207e-05,
      "loss": 0.1751677393913269,
      "mean_token_accuracy": 0.9230714440345764,
      "num_tokens": 104851888.0,
      "step": 8500
    },
    {
      "entropy": 1.1354945302009583,
      "epoch": 4.476566614007372,
      "grad_norm": 0.3480255603790283,
      "learning_rate": 7.470438424459221e-05,
      "loss": 0.17664237320423126,
      "mean_token_accuracy": 0.9248726069927216,
      "num_tokens": 104864224.0,
      "step": 8501
    },
    {
      "entropy": 1.1977665424346924,
      "epoch": 4.477093206951027,
      "grad_norm": 0.3040744662284851,
      "learning_rate": 7.468421365232733e-05,
      "loss": 0.1657549887895584,
      "mean_token_accuracy": 0.9289834946393967,
      "num_tokens": 104876560.0,
      "step": 8502
    },
    {
      "entropy": 1.111277014017105,
      "epoch": 4.477619799894681,
      "grad_norm": 0.30074837803840637,
      "learning_rate": 7.466404515655485e-05,
      "loss": 0.1681838482618332,
      "mean_token_accuracy": 0.9249631613492966,
      "num_tokens": 104888896.0,
      "step": 8503
    },
    {
      "entropy": 1.0993533730506897,
      "epoch": 4.478146392838336,
      "grad_norm": 0.3021378517150879,
      "learning_rate": 7.464387875847205e-05,
      "loss": 0.16494373977184296,
      "mean_token_accuracy": 0.9283095747232437,
      "num_tokens": 104901232.0,
      "step": 8504
    },
    {
      "entropy": 1.1367763876914978,
      "epoch": 4.478672985781991,
      "grad_norm": 0.3028930127620697,
      "learning_rate": 7.462371445927612e-05,
      "loss": 0.16392675042152405,
      "mean_token_accuracy": 0.9312854111194611,
      "num_tokens": 104913568.0,
      "step": 8505
    },
    {
      "entropy": 1.1491846442222595,
      "epoch": 4.479199578725645,
      "grad_norm": 0.3246096670627594,
      "learning_rate": 7.460355226016405e-05,
      "loss": 0.18212057650089264,
      "mean_token_accuracy": 0.9192562699317932,
      "num_tokens": 104925904.0,
      "step": 8506
    },
    {
      "entropy": 1.1474373042583466,
      "epoch": 4.4797261716692995,
      "grad_norm": 0.31376343965530396,
      "learning_rate": 7.458339216233276e-05,
      "loss": 0.159938246011734,
      "mean_token_accuracy": 0.9344170987606049,
      "num_tokens": 104938240.0,
      "step": 8507
    },
    {
      "entropy": 1.1231917142868042,
      "epoch": 4.480252764612954,
      "grad_norm": 0.30059564113616943,
      "learning_rate": 7.456323416697909e-05,
      "loss": 0.16879117488861084,
      "mean_token_accuracy": 0.9291982054710388,
      "num_tokens": 104950576.0,
      "step": 8508
    },
    {
      "entropy": 1.1523529291152954,
      "epoch": 4.480779357556608,
      "grad_norm": 0.3439018726348877,
      "learning_rate": 7.454307827529962e-05,
      "loss": 0.18372313678264618,
      "mean_token_accuracy": 0.9249345660209656,
      "num_tokens": 104962912.0,
      "step": 8509
    },
    {
      "entropy": 1.124235212802887,
      "epoch": 4.481305950500263,
      "grad_norm": 0.31315478682518005,
      "learning_rate": 7.452292448849099e-05,
      "loss": 0.15979686379432678,
      "mean_token_accuracy": 0.9314919114112854,
      "num_tokens": 104975248.0,
      "step": 8510
    },
    {
      "entropy": 1.118671476840973,
      "epoch": 4.481832543443918,
      "grad_norm": 0.30271604657173157,
      "learning_rate": 7.450277280774953e-05,
      "loss": 0.16365578770637512,
      "mean_token_accuracy": 0.9299983978271484,
      "num_tokens": 104987584.0,
      "step": 8511
    },
    {
      "entropy": 1.1343393325805664,
      "epoch": 4.482359136387572,
      "grad_norm": 0.34993165731430054,
      "learning_rate": 7.448262323427157e-05,
      "loss": 0.1759578436613083,
      "mean_token_accuracy": 0.9217237085103989,
      "num_tokens": 104999920.0,
      "step": 8512
    },
    {
      "entropy": 1.144093632698059,
      "epoch": 4.482885729331227,
      "grad_norm": 0.3411248028278351,
      "learning_rate": 7.446247576925325e-05,
      "loss": 0.18619801104068756,
      "mean_token_accuracy": 0.9250534325838089,
      "num_tokens": 105012256.0,
      "step": 8513
    },
    {
      "entropy": 1.1944924592971802,
      "epoch": 4.483412322274882,
      "grad_norm": 0.33487844467163086,
      "learning_rate": 7.444233041389057e-05,
      "loss": 0.16586452722549438,
      "mean_token_accuracy": 0.9306043982505798,
      "num_tokens": 105024592.0,
      "step": 8514
    },
    {
      "entropy": 1.1001003682613373,
      "epoch": 4.483938915218536,
      "grad_norm": 0.3087322413921356,
      "learning_rate": 7.442218716937954e-05,
      "loss": 0.16738297045230865,
      "mean_token_accuracy": 0.9300091564655304,
      "num_tokens": 105036928.0,
      "step": 8515
    },
    {
      "entropy": 1.1125937104225159,
      "epoch": 4.484465508162191,
      "grad_norm": 0.302702397108078,
      "learning_rate": 7.44020460369159e-05,
      "loss": 0.15746758878231049,
      "mean_token_accuracy": 0.9346013367176056,
      "num_tokens": 105049264.0,
      "step": 8516
    },
    {
      "entropy": 1.1246448755264282,
      "epoch": 4.484992101105846,
      "grad_norm": 0.3068924844264984,
      "learning_rate": 7.438190701769528e-05,
      "loss": 0.15912923216819763,
      "mean_token_accuracy": 0.9279821664094925,
      "num_tokens": 105061600.0,
      "step": 8517
    },
    {
      "entropy": 1.1221618354320526,
      "epoch": 4.4855186940494995,
      "grad_norm": 0.3064675033092499,
      "learning_rate": 7.436177011291323e-05,
      "loss": 0.1689017415046692,
      "mean_token_accuracy": 0.9291050136089325,
      "num_tokens": 105073936.0,
      "step": 8518
    },
    {
      "entropy": 1.0712625682353973,
      "epoch": 4.486045286993154,
      "grad_norm": 0.2907482087612152,
      "learning_rate": 7.434163532376521e-05,
      "loss": 0.15641829371452332,
      "mean_token_accuracy": 0.9321083426475525,
      "num_tokens": 105086272.0,
      "step": 8519
    },
    {
      "entropy": 1.0565262734889984,
      "epoch": 4.486571879936809,
      "grad_norm": 0.30794206261634827,
      "learning_rate": 7.432150265144643e-05,
      "loss": 0.16304053366184235,
      "mean_token_accuracy": 0.9296928197145462,
      "num_tokens": 105098608.0,
      "step": 8520
    },
    {
      "entropy": 1.104030430316925,
      "epoch": 4.487098472880463,
      "grad_norm": 0.485408753156662,
      "learning_rate": 7.430137209715211e-05,
      "loss": 0.18189111351966858,
      "mean_token_accuracy": 0.924937978386879,
      "num_tokens": 105110944.0,
      "step": 8521
    },
    {
      "entropy": 1.0996279120445251,
      "epoch": 4.487625065824118,
      "grad_norm": 0.312265008687973,
      "learning_rate": 7.428124366207727e-05,
      "loss": 0.1506020873785019,
      "mean_token_accuracy": 0.933230385184288,
      "num_tokens": 105123280.0,
      "step": 8522
    },
    {
      "entropy": 1.1094778776168823,
      "epoch": 4.488151658767772,
      "grad_norm": 0.3368557393550873,
      "learning_rate": 7.426111734741679e-05,
      "loss": 0.17881613969802856,
      "mean_token_accuracy": 0.9285390824079514,
      "num_tokens": 105135616.0,
      "step": 8523
    },
    {
      "entropy": 1.140645056962967,
      "epoch": 4.488678251711427,
      "grad_norm": 0.3306402266025543,
      "learning_rate": 7.424099315436548e-05,
      "loss": 0.17833511531352997,
      "mean_token_accuracy": 0.9260087609291077,
      "num_tokens": 105147952.0,
      "step": 8524
    },
    {
      "entropy": 1.1619056463241577,
      "epoch": 4.489204844655082,
      "grad_norm": 0.33646610379219055,
      "learning_rate": 7.4220871084118e-05,
      "loss": 0.16644509136676788,
      "mean_token_accuracy": 0.9322192668914795,
      "num_tokens": 105160288.0,
      "step": 8525
    },
    {
      "entropy": 1.085435301065445,
      "epoch": 4.489731437598736,
      "grad_norm": 0.30755895376205444,
      "learning_rate": 7.420075113786886e-05,
      "loss": 0.16950206458568573,
      "mean_token_accuracy": 0.9276347011327744,
      "num_tokens": 105172624.0,
      "step": 8526
    },
    {
      "entropy": 1.086694598197937,
      "epoch": 4.490258030542391,
      "grad_norm": 0.31825414299964905,
      "learning_rate": 7.418063331681247e-05,
      "loss": 0.16300199925899506,
      "mean_token_accuracy": 0.9320763200521469,
      "num_tokens": 105184960.0,
      "step": 8527
    },
    {
      "entropy": 1.153379887342453,
      "epoch": 4.490784623486046,
      "grad_norm": 0.3786374628543854,
      "learning_rate": 7.416051762214307e-05,
      "loss": 0.1907190978527069,
      "mean_token_accuracy": 0.9148112684488297,
      "num_tokens": 105197296.0,
      "step": 8528
    },
    {
      "entropy": 1.0967216789722443,
      "epoch": 4.4913112164296995,
      "grad_norm": 0.32600608468055725,
      "learning_rate": 7.41404040550549e-05,
      "loss": 0.1884557604789734,
      "mean_token_accuracy": 0.9216152876615524,
      "num_tokens": 105209632.0,
      "step": 8529
    },
    {
      "entropy": 1.1392711102962494,
      "epoch": 4.491837809373354,
      "grad_norm": 0.33852773904800415,
      "learning_rate": 7.412029261674191e-05,
      "loss": 0.17196306586265564,
      "mean_token_accuracy": 0.9291158318519592,
      "num_tokens": 105221968.0,
      "step": 8530
    },
    {
      "entropy": 1.1099463403224945,
      "epoch": 4.492364402317009,
      "grad_norm": 0.30658501386642456,
      "learning_rate": 7.410018330839805e-05,
      "loss": 0.1679382622241974,
      "mean_token_accuracy": 0.9314810335636139,
      "num_tokens": 105234304.0,
      "step": 8531
    },
    {
      "entropy": 1.1185071468353271,
      "epoch": 4.492890995260663,
      "grad_norm": 0.3201732337474823,
      "learning_rate": 7.408007613121706e-05,
      "loss": 0.17588239908218384,
      "mean_token_accuracy": 0.9251694977283478,
      "num_tokens": 105246640.0,
      "step": 8532
    },
    {
      "entropy": 1.0944017171859741,
      "epoch": 4.493417588204318,
      "grad_norm": 0.3167847692966461,
      "learning_rate": 7.405997108639258e-05,
      "loss": 0.17677423357963562,
      "mean_token_accuracy": 0.924406886100769,
      "num_tokens": 105258976.0,
      "step": 8533
    },
    {
      "entropy": 1.1053787171840668,
      "epoch": 4.493944181147973,
      "grad_norm": 0.30526003241539,
      "learning_rate": 7.403986817511814e-05,
      "loss": 0.16546492278575897,
      "mean_token_accuracy": 0.9274102002382278,
      "num_tokens": 105271312.0,
      "step": 8534
    },
    {
      "entropy": 1.087305873632431,
      "epoch": 4.494470774091627,
      "grad_norm": 0.32260844111442566,
      "learning_rate": 7.401976739858714e-05,
      "loss": 0.1721532940864563,
      "mean_token_accuracy": 0.9245280623435974,
      "num_tokens": 105283648.0,
      "step": 8535
    },
    {
      "entropy": 1.149102807044983,
      "epoch": 4.494997367035282,
      "grad_norm": 0.3301563858985901,
      "learning_rate": 7.399966875799282e-05,
      "loss": 0.1612205058336258,
      "mean_token_accuracy": 0.930897980928421,
      "num_tokens": 105295984.0,
      "step": 8536
    },
    {
      "entropy": 1.1381434500217438,
      "epoch": 4.495523959978936,
      "grad_norm": 0.3319721817970276,
      "learning_rate": 7.397957225452838e-05,
      "loss": 0.17157316207885742,
      "mean_token_accuracy": 0.9277976751327515,
      "num_tokens": 105308320.0,
      "step": 8537
    },
    {
      "entropy": 1.0968379527330399,
      "epoch": 4.496050552922591,
      "grad_norm": 0.3020884394645691,
      "learning_rate": 7.395947788938677e-05,
      "loss": 0.16621725261211395,
      "mean_token_accuracy": 0.9268390387296677,
      "num_tokens": 105320656.0,
      "step": 8538
    },
    {
      "entropy": 1.1605389714241028,
      "epoch": 4.496577145866246,
      "grad_norm": 0.3507307469844818,
      "learning_rate": 7.393938566376091e-05,
      "loss": 0.16039294004440308,
      "mean_token_accuracy": 0.9318993240594864,
      "num_tokens": 105332992.0,
      "step": 8539
    },
    {
      "entropy": 1.137021780014038,
      "epoch": 4.4971037388098996,
      "grad_norm": 0.32756632566452026,
      "learning_rate": 7.391929557884354e-05,
      "loss": 0.16134996712207794,
      "mean_token_accuracy": 0.9335807263851166,
      "num_tokens": 105345328.0,
      "step": 8540
    },
    {
      "entropy": 1.0737043917179108,
      "epoch": 4.497630331753554,
      "grad_norm": 0.3067854940891266,
      "learning_rate": 7.389920763582729e-05,
      "loss": 0.16501881182193756,
      "mean_token_accuracy": 0.9256889075040817,
      "num_tokens": 105357664.0,
      "step": 8541
    },
    {
      "entropy": 1.1259743571281433,
      "epoch": 4.498156924697209,
      "grad_norm": 0.3393458127975464,
      "learning_rate": 7.387912183590466e-05,
      "loss": 0.17241157591342926,
      "mean_token_accuracy": 0.9270386695861816,
      "num_tokens": 105370000.0,
      "step": 8542
    },
    {
      "entropy": 1.117572009563446,
      "epoch": 4.498683517640863,
      "grad_norm": 0.3278200030326843,
      "learning_rate": 7.385903818026804e-05,
      "loss": 0.16908778250217438,
      "mean_token_accuracy": 0.9263511300086975,
      "num_tokens": 105382336.0,
      "step": 8543
    },
    {
      "entropy": 1.0400933921337128,
      "epoch": 4.499210110584518,
      "grad_norm": 0.31926998496055603,
      "learning_rate": 7.383895667010969e-05,
      "loss": 0.17449776828289032,
      "mean_token_accuracy": 0.9278433322906494,
      "num_tokens": 105394672.0,
      "step": 8544
    },
    {
      "entropy": 1.1213461458683014,
      "epoch": 4.499736703528173,
      "grad_norm": 0.3107103705406189,
      "learning_rate": 7.381887730662176e-05,
      "loss": 0.15966379642486572,
      "mean_token_accuracy": 0.9357942491769791,
      "num_tokens": 105407008.0,
      "step": 8545
    },
    {
      "entropy": 1.088557094335556,
      "epoch": 4.500263296471827,
      "grad_norm": 0.33047419786453247,
      "learning_rate": 7.379880009099616e-05,
      "loss": 0.16721244156360626,
      "mean_token_accuracy": 0.9257785826921463,
      "num_tokens": 105419344.0,
      "step": 8546
    },
    {
      "entropy": 1.118956983089447,
      "epoch": 4.500789889415482,
      "grad_norm": 0.31375518441200256,
      "learning_rate": 7.377872502442483e-05,
      "loss": 0.15484580397605896,
      "mean_token_accuracy": 0.936336100101471,
      "num_tokens": 105431680.0,
      "step": 8547
    },
    {
      "entropy": 1.080495923757553,
      "epoch": 4.501316482359137,
      "grad_norm": 0.27866771817207336,
      "learning_rate": 7.375865210809945e-05,
      "loss": 0.1486835777759552,
      "mean_token_accuracy": 0.9381964206695557,
      "num_tokens": 105444016.0,
      "step": 8548
    },
    {
      "entropy": 1.0975594818592072,
      "epoch": 4.501843075302791,
      "grad_norm": 0.3373928368091583,
      "learning_rate": 7.373858134321168e-05,
      "loss": 0.17517077922821045,
      "mean_token_accuracy": 0.9238870292901993,
      "num_tokens": 105456352.0,
      "step": 8549
    },
    {
      "entropy": 1.136119544506073,
      "epoch": 4.502369668246446,
      "grad_norm": 0.3324989080429077,
      "learning_rate": 7.371851273095297e-05,
      "loss": 0.16266067326068878,
      "mean_token_accuracy": 0.9347175359725952,
      "num_tokens": 105468688.0,
      "step": 8550
    },
    {
      "entropy": 1.0626960098743439,
      "epoch": 4.5028962611901004,
      "grad_norm": 0.2993353307247162,
      "learning_rate": 7.36984462725147e-05,
      "loss": 0.15661530196666718,
      "mean_token_accuracy": 0.9310466200113297,
      "num_tokens": 105481024.0,
      "step": 8551
    },
    {
      "entropy": 1.086951494216919,
      "epoch": 4.503422854133754,
      "grad_norm": 0.33778616786003113,
      "learning_rate": 7.36783819690881e-05,
      "loss": 0.1886535882949829,
      "mean_token_accuracy": 0.9179743528366089,
      "num_tokens": 105493360.0,
      "step": 8552
    },
    {
      "entropy": 1.052822306752205,
      "epoch": 4.503949447077409,
      "grad_norm": 0.30305930972099304,
      "learning_rate": 7.365831982186427e-05,
      "loss": 0.1595156341791153,
      "mean_token_accuracy": 0.9331249743700027,
      "num_tokens": 105505696.0,
      "step": 8553
    },
    {
      "entropy": 1.0562293082475662,
      "epoch": 4.504476040021064,
      "grad_norm": 0.2993280291557312,
      "learning_rate": 7.363825983203416e-05,
      "loss": 0.15765467286109924,
      "mean_token_accuracy": 0.933802455663681,
      "num_tokens": 105518032.0,
      "step": 8554
    },
    {
      "entropy": 1.0926483571529388,
      "epoch": 4.505002632964718,
      "grad_norm": 0.304964542388916,
      "learning_rate": 7.361820200078867e-05,
      "loss": 0.1578100025653839,
      "mean_token_accuracy": 0.9362167567014694,
      "num_tokens": 105530368.0,
      "step": 8555
    },
    {
      "entropy": 1.0794464945793152,
      "epoch": 4.505529225908373,
      "grad_norm": 0.3184886872768402,
      "learning_rate": 7.359814632931842e-05,
      "loss": 0.16962140798568726,
      "mean_token_accuracy": 0.9269872158765793,
      "num_tokens": 105542704.0,
      "step": 8556
    },
    {
      "entropy": 1.0823484659194946,
      "epoch": 4.506055818852027,
      "grad_norm": 0.3482179641723633,
      "learning_rate": 7.35780928188141e-05,
      "loss": 0.16899374127388,
      "mean_token_accuracy": 0.9297892153263092,
      "num_tokens": 105555040.0,
      "step": 8557
    },
    {
      "entropy": 1.0916792750358582,
      "epoch": 4.506582411795682,
      "grad_norm": 0.30531033873558044,
      "learning_rate": 7.35580414704661e-05,
      "loss": 0.1647178828716278,
      "mean_token_accuracy": 0.9307073503732681,
      "num_tokens": 105567376.0,
      "step": 8558
    },
    {
      "entropy": 1.0854963660240173,
      "epoch": 4.507109004739337,
      "grad_norm": 0.3209065794944763,
      "learning_rate": 7.353799228546481e-05,
      "loss": 0.17538440227508545,
      "mean_token_accuracy": 0.924390122294426,
      "num_tokens": 105579712.0,
      "step": 8559
    },
    {
      "entropy": 1.061386227607727,
      "epoch": 4.507635597682991,
      "grad_norm": 0.3283348083496094,
      "learning_rate": 7.351794526500038e-05,
      "loss": 0.1658862680196762,
      "mean_token_accuracy": 0.9272024780511856,
      "num_tokens": 105592048.0,
      "step": 8560
    },
    {
      "entropy": 1.0981746166944504,
      "epoch": 4.508162190626646,
      "grad_norm": 0.3161288797855377,
      "learning_rate": 7.349790041026292e-05,
      "loss": 0.16507594287395477,
      "mean_token_accuracy": 0.937386766076088,
      "num_tokens": 105604384.0,
      "step": 8561
    },
    {
      "entropy": 1.1265353560447693,
      "epoch": 4.5086887835703005,
      "grad_norm": 0.3288407027721405,
      "learning_rate": 7.347785772244236e-05,
      "loss": 0.18043377995491028,
      "mean_token_accuracy": 0.924061119556427,
      "num_tokens": 105616720.0,
      "step": 8562
    },
    {
      "entropy": 1.140741229057312,
      "epoch": 4.509215376513954,
      "grad_norm": 0.3171781003475189,
      "learning_rate": 7.345781720272849e-05,
      "loss": 0.16179314255714417,
      "mean_token_accuracy": 0.9300349056720734,
      "num_tokens": 105629056.0,
      "step": 8563
    },
    {
      "entropy": 1.077030524611473,
      "epoch": 4.509741969457609,
      "grad_norm": 0.3282569348812103,
      "learning_rate": 7.343777885231105e-05,
      "loss": 0.16976416110992432,
      "mean_token_accuracy": 0.9266364425420761,
      "num_tokens": 105641392.0,
      "step": 8564
    },
    {
      "entropy": 1.0798559784889221,
      "epoch": 4.510268562401264,
      "grad_norm": 0.33072084188461304,
      "learning_rate": 7.341774267237957e-05,
      "loss": 0.1782056838274002,
      "mean_token_accuracy": 0.9258958399295807,
      "num_tokens": 105653728.0,
      "step": 8565
    },
    {
      "entropy": 1.122019350528717,
      "epoch": 4.510795155344918,
      "grad_norm": 0.3206283748149872,
      "learning_rate": 7.339770866412349e-05,
      "loss": 0.15470805764198303,
      "mean_token_accuracy": 0.9337612390518188,
      "num_tokens": 105666064.0,
      "step": 8566
    },
    {
      "entropy": 1.1067062616348267,
      "epoch": 4.511321748288573,
      "grad_norm": 0.3272755742073059,
      "learning_rate": 7.337767682873214e-05,
      "loss": 0.178415447473526,
      "mean_token_accuracy": 0.9239523708820343,
      "num_tokens": 105678400.0,
      "step": 8567
    },
    {
      "entropy": 1.0962112545967102,
      "epoch": 4.511848341232227,
      "grad_norm": 0.31566762924194336,
      "learning_rate": 7.335764716739463e-05,
      "loss": 0.16647085547447205,
      "mean_token_accuracy": 0.9280145019292831,
      "num_tokens": 105690736.0,
      "step": 8568
    },
    {
      "entropy": 1.070769488811493,
      "epoch": 4.512374934175882,
      "grad_norm": 0.29418423771858215,
      "learning_rate": 7.333761968130007e-05,
      "loss": 0.16237285733222961,
      "mean_token_accuracy": 0.9303525388240814,
      "num_tokens": 105703072.0,
      "step": 8569
    },
    {
      "entropy": 1.154471606016159,
      "epoch": 4.512901527119537,
      "grad_norm": 0.35960978269577026,
      "learning_rate": 7.33175943716373e-05,
      "loss": 0.1800576001405716,
      "mean_token_accuracy": 0.9220800697803497,
      "num_tokens": 105715408.0,
      "step": 8570
    },
    {
      "entropy": 1.1531608998775482,
      "epoch": 4.513428120063191,
      "grad_norm": 0.327642560005188,
      "learning_rate": 7.32975712395952e-05,
      "loss": 0.17608018219470978,
      "mean_token_accuracy": 0.9217364192008972,
      "num_tokens": 105727744.0,
      "step": 8571
    },
    {
      "entropy": 1.0929425954818726,
      "epoch": 4.513954713006846,
      "grad_norm": 0.3168841302394867,
      "learning_rate": 7.327755028636236e-05,
      "loss": 0.16714897751808167,
      "mean_token_accuracy": 0.9331997483968735,
      "num_tokens": 105740080.0,
      "step": 8572
    },
    {
      "entropy": 1.1154712438583374,
      "epoch": 4.5144813059505005,
      "grad_norm": 0.3164999783039093,
      "learning_rate": 7.325753151312731e-05,
      "loss": 0.17461040616035461,
      "mean_token_accuracy": 0.9234779328107834,
      "num_tokens": 105752416.0,
      "step": 8573
    },
    {
      "entropy": 1.1246903836727142,
      "epoch": 4.515007898894154,
      "grad_norm": 0.31671062111854553,
      "learning_rate": 7.32375149210785e-05,
      "loss": 0.1774444580078125,
      "mean_token_accuracy": 0.9238055646419525,
      "num_tokens": 105764752.0,
      "step": 8574
    },
    {
      "entropy": 1.1531439423561096,
      "epoch": 4.515534491837809,
      "grad_norm": 0.32286152243614197,
      "learning_rate": 7.321750051140415e-05,
      "loss": 0.1749352514743805,
      "mean_token_accuracy": 0.9257891327142715,
      "num_tokens": 105777088.0,
      "step": 8575
    },
    {
      "entropy": 1.1073280274868011,
      "epoch": 4.516061084781464,
      "grad_norm": 0.29248252511024475,
      "learning_rate": 7.319748828529238e-05,
      "loss": 0.16649188101291656,
      "mean_token_accuracy": 0.9266532808542252,
      "num_tokens": 105789424.0,
      "step": 8576
    },
    {
      "entropy": 1.0988223552703857,
      "epoch": 4.516587677725118,
      "grad_norm": 0.296479731798172,
      "learning_rate": 7.317747824393121e-05,
      "loss": 0.17069338262081146,
      "mean_token_accuracy": 0.9276268184185028,
      "num_tokens": 105801760.0,
      "step": 8577
    },
    {
      "entropy": 1.1450729966163635,
      "epoch": 4.517114270668773,
      "grad_norm": 0.3118171989917755,
      "learning_rate": 7.315747038850863e-05,
      "loss": 0.15814495086669922,
      "mean_token_accuracy": 0.9315481334924698,
      "num_tokens": 105814096.0,
      "step": 8578
    },
    {
      "entropy": 1.1546815037727356,
      "epoch": 4.517640863612428,
      "grad_norm": 0.33646687865257263,
      "learning_rate": 7.313746472021224e-05,
      "loss": 0.17359524965286255,
      "mean_token_accuracy": 0.9252714812755585,
      "num_tokens": 105826432.0,
      "step": 8579
    },
    {
      "entropy": 1.139982283115387,
      "epoch": 4.518167456556082,
      "grad_norm": 0.3143118619918823,
      "learning_rate": 7.311746124022971e-05,
      "loss": 0.16783250868320465,
      "mean_token_accuracy": 0.9310547113418579,
      "num_tokens": 105838768.0,
      "step": 8580
    },
    {
      "entropy": 1.132720172405243,
      "epoch": 4.518694049499737,
      "grad_norm": 0.33125928044319153,
      "learning_rate": 7.309745994974854e-05,
      "loss": 0.1766769289970398,
      "mean_token_accuracy": 0.9286185204982758,
      "num_tokens": 105851104.0,
      "step": 8581
    },
    {
      "entropy": 1.131675124168396,
      "epoch": 4.519220642443392,
      "grad_norm": 0.31773433089256287,
      "learning_rate": 7.307746084995613e-05,
      "loss": 0.1767442226409912,
      "mean_token_accuracy": 0.9222896993160248,
      "num_tokens": 105863440.0,
      "step": 8582
    },
    {
      "entropy": 1.0698230564594269,
      "epoch": 4.519747235387046,
      "grad_norm": 0.2915615141391754,
      "learning_rate": 7.305746394203963e-05,
      "loss": 0.1695348471403122,
      "mean_token_accuracy": 0.9268006384372711,
      "num_tokens": 105875776.0,
      "step": 8583
    },
    {
      "entropy": 1.1281887292861938,
      "epoch": 4.5202738283307005,
      "grad_norm": 0.31251072883605957,
      "learning_rate": 7.303746922718617e-05,
      "loss": 0.17473962903022766,
      "mean_token_accuracy": 0.9253868460655212,
      "num_tokens": 105888112.0,
      "step": 8584
    },
    {
      "entropy": 1.122416466474533,
      "epoch": 4.520800421274355,
      "grad_norm": 0.3060236871242523,
      "learning_rate": 7.301747670658275e-05,
      "loss": 0.16630981862545013,
      "mean_token_accuracy": 0.922129675745964,
      "num_tokens": 105900448.0,
      "step": 8585
    },
    {
      "entropy": 1.0944432616233826,
      "epoch": 4.521327014218009,
      "grad_norm": 0.3030654788017273,
      "learning_rate": 7.299748638141617e-05,
      "loss": 0.15867483615875244,
      "mean_token_accuracy": 0.9311196506023407,
      "num_tokens": 105912784.0,
      "step": 8586
    },
    {
      "entropy": 1.1073821485042572,
      "epoch": 4.521853607161664,
      "grad_norm": 0.3436656594276428,
      "learning_rate": 7.297749825287317e-05,
      "loss": 0.1834273338317871,
      "mean_token_accuracy": 0.9229354411363602,
      "num_tokens": 105925120.0,
      "step": 8587
    },
    {
      "entropy": 1.1482012867927551,
      "epoch": 4.522380200105319,
      "grad_norm": 0.3048330843448639,
      "learning_rate": 7.29575123221403e-05,
      "loss": 0.1720239818096161,
      "mean_token_accuracy": 0.926640197634697,
      "num_tokens": 105937456.0,
      "step": 8588
    },
    {
      "entropy": 1.1464100182056427,
      "epoch": 4.522906793048973,
      "grad_norm": 0.29938042163848877,
      "learning_rate": 7.293752859040408e-05,
      "loss": 0.15156853199005127,
      "mean_token_accuracy": 0.9363648891448975,
      "num_tokens": 105949792.0,
      "step": 8589
    },
    {
      "entropy": 1.1625692248344421,
      "epoch": 4.523433385992628,
      "grad_norm": 0.404100626707077,
      "learning_rate": 7.291754705885073e-05,
      "loss": 0.20120716094970703,
      "mean_token_accuracy": 0.9150242507457733,
      "num_tokens": 105962128.0,
      "step": 8590
    },
    {
      "entropy": 1.1558034420013428,
      "epoch": 4.523959978936282,
      "grad_norm": 0.32560110092163086,
      "learning_rate": 7.289756772866642e-05,
      "loss": 0.16128265857696533,
      "mean_token_accuracy": 0.9327718168497086,
      "num_tokens": 105974464.0,
      "step": 8591
    },
    {
      "entropy": 1.121302604675293,
      "epoch": 4.524486571879937,
      "grad_norm": 0.314691424369812,
      "learning_rate": 7.287759060103734e-05,
      "loss": 0.16107745468616486,
      "mean_token_accuracy": 0.9332721680402756,
      "num_tokens": 105986800.0,
      "step": 8592
    },
    {
      "entropy": 1.1109824180603027,
      "epoch": 4.525013164823592,
      "grad_norm": 0.30954957008361816,
      "learning_rate": 7.285761567714933e-05,
      "loss": 0.17070379853248596,
      "mean_token_accuracy": 0.9300545454025269,
      "num_tokens": 105999136.0,
      "step": 8593
    },
    {
      "entropy": 1.1232934296131134,
      "epoch": 4.525539757767246,
      "grad_norm": 0.31915655732154846,
      "learning_rate": 7.283764295818818e-05,
      "loss": 0.15884867310523987,
      "mean_token_accuracy": 0.9324177652597427,
      "num_tokens": 106011472.0,
      "step": 8594
    },
    {
      "entropy": 1.0580084770917892,
      "epoch": 4.5260663507109005,
      "grad_norm": 0.3162199556827545,
      "learning_rate": 7.28176724453396e-05,
      "loss": 0.17864418029785156,
      "mean_token_accuracy": 0.9264102429151535,
      "num_tokens": 106023808.0,
      "step": 8595
    },
    {
      "entropy": 1.0718005001544952,
      "epoch": 4.526592943654555,
      "grad_norm": 0.302757203578949,
      "learning_rate": 7.279770413978905e-05,
      "loss": 0.17261910438537598,
      "mean_token_accuracy": 0.927869588136673,
      "num_tokens": 106036144.0,
      "step": 8596
    },
    {
      "entropy": 1.0653924643993378,
      "epoch": 4.527119536598209,
      "grad_norm": 0.3032852113246918,
      "learning_rate": 7.277773804272199e-05,
      "loss": 0.1635109782218933,
      "mean_token_accuracy": 0.9316293299198151,
      "num_tokens": 106048480.0,
      "step": 8597
    },
    {
      "entropy": 1.0054285675287247,
      "epoch": 4.527646129541864,
      "grad_norm": 0.29458484053611755,
      "learning_rate": 7.275777415532368e-05,
      "loss": 0.16193795204162598,
      "mean_token_accuracy": 0.9285301566123962,
      "num_tokens": 106060816.0,
      "step": 8598
    },
    {
      "entropy": 1.0754818618297577,
      "epoch": 4.528172722485519,
      "grad_norm": 0.3533173203468323,
      "learning_rate": 7.273781247877922e-05,
      "loss": 0.1745351403951645,
      "mean_token_accuracy": 0.9249437004327774,
      "num_tokens": 106073152.0,
      "step": 8599
    },
    {
      "entropy": 1.0397689640522003,
      "epoch": 4.528699315429173,
      "grad_norm": 0.3107565939426422,
      "learning_rate": 7.271785301427369e-05,
      "loss": 0.17134985327720642,
      "mean_token_accuracy": 0.9243182092905045,
      "num_tokens": 106085488.0,
      "step": 8600
    },
    {
      "entropy": 1.055548071861267,
      "epoch": 4.529225908372828,
      "grad_norm": 0.31921958923339844,
      "learning_rate": 7.269789576299192e-05,
      "loss": 0.17372101545333862,
      "mean_token_accuracy": 0.9243157655000687,
      "num_tokens": 106097824.0,
      "step": 8601
    },
    {
      "entropy": 1.0808067321777344,
      "epoch": 4.529752501316482,
      "grad_norm": 0.2958208918571472,
      "learning_rate": 7.267794072611869e-05,
      "loss": 0.1557004600763321,
      "mean_token_accuracy": 0.9333107024431229,
      "num_tokens": 106110160.0,
      "step": 8602
    },
    {
      "entropy": 1.0594066977500916,
      "epoch": 4.530279094260137,
      "grad_norm": 0.3041718304157257,
      "learning_rate": 7.265798790483855e-05,
      "loss": 0.16468429565429688,
      "mean_token_accuracy": 0.9312732368707657,
      "num_tokens": 106122496.0,
      "step": 8603
    },
    {
      "entropy": 1.1187460720539093,
      "epoch": 4.530805687203792,
      "grad_norm": 0.29810383915901184,
      "learning_rate": 7.263803730033607e-05,
      "loss": 0.17059026658535004,
      "mean_token_accuracy": 0.9303388297557831,
      "num_tokens": 106134832.0,
      "step": 8604
    },
    {
      "entropy": 1.1017462611198425,
      "epoch": 4.531332280147446,
      "grad_norm": 0.33466941118240356,
      "learning_rate": 7.261808891379549e-05,
      "loss": 0.17381291091442108,
      "mean_token_accuracy": 0.9265600442886353,
      "num_tokens": 106147168.0,
      "step": 8605
    },
    {
      "entropy": 1.0778556764125824,
      "epoch": 4.5318588730911005,
      "grad_norm": 0.328126460313797,
      "learning_rate": 7.259814274640114e-05,
      "loss": 0.17598795890808105,
      "mean_token_accuracy": 0.9262767136096954,
      "num_tokens": 106159504.0,
      "step": 8606
    },
    {
      "entropy": 1.1326199769973755,
      "epoch": 4.532385466034755,
      "grad_norm": 0.3385554850101471,
      "learning_rate": 7.257819879933703e-05,
      "loss": 0.18585970997810364,
      "mean_token_accuracy": 0.9193847179412842,
      "num_tokens": 106171840.0,
      "step": 8607
    },
    {
      "entropy": 1.0936433672904968,
      "epoch": 4.532912058978409,
      "grad_norm": 0.3459876477718353,
      "learning_rate": 7.255825707378719e-05,
      "loss": 0.17953082919120789,
      "mean_token_accuracy": 0.9288031160831451,
      "num_tokens": 106184176.0,
      "step": 8608
    },
    {
      "entropy": 1.1019591391086578,
      "epoch": 4.533438651922064,
      "grad_norm": 0.331623375415802,
      "learning_rate": 7.253831757093534e-05,
      "loss": 0.18319639563560486,
      "mean_token_accuracy": 0.9238258898258209,
      "num_tokens": 106196512.0,
      "step": 8609
    },
    {
      "entropy": 1.0682066679000854,
      "epoch": 4.533965244865719,
      "grad_norm": 0.326761394739151,
      "learning_rate": 7.251838029196528e-05,
      "loss": 0.17491675913333893,
      "mean_token_accuracy": 0.9245114624500275,
      "num_tokens": 106208848.0,
      "step": 8610
    },
    {
      "entropy": 1.1281436085700989,
      "epoch": 4.534491837809373,
      "grad_norm": 0.311954528093338,
      "learning_rate": 7.249844523806049e-05,
      "loss": 0.16504040360450745,
      "mean_token_accuracy": 0.9301846921443939,
      "num_tokens": 106221184.0,
      "step": 8611
    },
    {
      "entropy": 1.0995433032512665,
      "epoch": 4.535018430753028,
      "grad_norm": 0.3156920075416565,
      "learning_rate": 7.247851241040446e-05,
      "loss": 0.16396453976631165,
      "mean_token_accuracy": 0.9300717115402222,
      "num_tokens": 106233520.0,
      "step": 8612
    },
    {
      "entropy": 1.060952678322792,
      "epoch": 4.535545023696683,
      "grad_norm": 0.30757302045822144,
      "learning_rate": 7.245858181018044e-05,
      "loss": 0.1719992756843567,
      "mean_token_accuracy": 0.929116889834404,
      "num_tokens": 106245856.0,
      "step": 8613
    },
    {
      "entropy": 1.097676694393158,
      "epoch": 4.536071616640337,
      "grad_norm": 0.3066072463989258,
      "learning_rate": 7.243865343857163e-05,
      "loss": 0.16438062489032745,
      "mean_token_accuracy": 0.9286708384752274,
      "num_tokens": 106258192.0,
      "step": 8614
    },
    {
      "entropy": 1.1292257606983185,
      "epoch": 4.536598209583992,
      "grad_norm": 0.3034124970436096,
      "learning_rate": 7.241872729676102e-05,
      "loss": 0.1616891324520111,
      "mean_token_accuracy": 0.9323288798332214,
      "num_tokens": 106270528.0,
      "step": 8615
    },
    {
      "entropy": 1.1192235052585602,
      "epoch": 4.5371248025276465,
      "grad_norm": 0.3250950574874878,
      "learning_rate": 7.239880338593152e-05,
      "loss": 0.1760547012090683,
      "mean_token_accuracy": 0.9272078275680542,
      "num_tokens": 106282864.0,
      "step": 8616
    },
    {
      "entropy": 1.1232163608074188,
      "epoch": 4.5376513954713005,
      "grad_norm": 0.3211779296398163,
      "learning_rate": 7.237888170726595e-05,
      "loss": 0.17828568816184998,
      "mean_token_accuracy": 0.9265819787979126,
      "num_tokens": 106295200.0,
      "step": 8617
    },
    {
      "entropy": 1.1110897064208984,
      "epoch": 4.538177988414955,
      "grad_norm": 0.31655165553092957,
      "learning_rate": 7.235896226194687e-05,
      "loss": 0.16581428050994873,
      "mean_token_accuracy": 0.9282494932413101,
      "num_tokens": 106307536.0,
      "step": 8618
    },
    {
      "entropy": 1.059042513370514,
      "epoch": 4.53870458135861,
      "grad_norm": 0.2922663688659668,
      "learning_rate": 7.233904505115676e-05,
      "loss": 0.16460026800632477,
      "mean_token_accuracy": 0.9286544919013977,
      "num_tokens": 106319872.0,
      "step": 8619
    },
    {
      "entropy": 1.0846833884716034,
      "epoch": 4.539231174302264,
      "grad_norm": 0.29960334300994873,
      "learning_rate": 7.231913007607813e-05,
      "loss": 0.16176140308380127,
      "mean_token_accuracy": 0.9320085644721985,
      "num_tokens": 106332208.0,
      "step": 8620
    },
    {
      "entropy": 1.1031128764152527,
      "epoch": 4.539757767245919,
      "grad_norm": 0.32958582043647766,
      "learning_rate": 7.22992173378931e-05,
      "loss": 0.18814656138420105,
      "mean_token_accuracy": 0.9222319424152374,
      "num_tokens": 106344544.0,
      "step": 8621
    },
    {
      "entropy": 1.0944068729877472,
      "epoch": 4.540284360189574,
      "grad_norm": 0.32673224806785583,
      "learning_rate": 7.227930683778381e-05,
      "loss": 0.1563873142004013,
      "mean_token_accuracy": 0.934342235326767,
      "num_tokens": 106356880.0,
      "step": 8622
    },
    {
      "entropy": 1.1245568096637726,
      "epoch": 4.540810953133228,
      "grad_norm": 0.32896438241004944,
      "learning_rate": 7.22593985769322e-05,
      "loss": 0.16392138600349426,
      "mean_token_accuracy": 0.9313053041696548,
      "num_tokens": 106369216.0,
      "step": 8623
    },
    {
      "entropy": 1.1108445227146149,
      "epoch": 4.541337546076883,
      "grad_norm": 0.3246438503265381,
      "learning_rate": 7.223949255652014e-05,
      "loss": 0.16873814165592194,
      "mean_token_accuracy": 0.9327792525291443,
      "num_tokens": 106381552.0,
      "step": 8624
    },
    {
      "entropy": 1.1474404633045197,
      "epoch": 4.541864139020537,
      "grad_norm": 0.3198279142379761,
      "learning_rate": 7.221958877772932e-05,
      "loss": 0.17041805386543274,
      "mean_token_accuracy": 0.9279144108295441,
      "num_tokens": 106393888.0,
      "step": 8625
    },
    {
      "entropy": 1.0965960025787354,
      "epoch": 4.542390731964192,
      "grad_norm": 0.34083613753318787,
      "learning_rate": 7.219968724174125e-05,
      "loss": 0.16393917798995972,
      "mean_token_accuracy": 0.9325901865959167,
      "num_tokens": 106406224.0,
      "step": 8626
    },
    {
      "entropy": 1.141867220401764,
      "epoch": 4.5429173249078465,
      "grad_norm": 0.3511935770511627,
      "learning_rate": 7.21797879497375e-05,
      "loss": 0.1670934408903122,
      "mean_token_accuracy": 0.9295719414949417,
      "num_tokens": 106418560.0,
      "step": 8627
    },
    {
      "entropy": 1.0650420784950256,
      "epoch": 4.5434439178515005,
      "grad_norm": 0.29902124404907227,
      "learning_rate": 7.215989090289928e-05,
      "loss": 0.16046491265296936,
      "mean_token_accuracy": 0.9318942427635193,
      "num_tokens": 106430896.0,
      "step": 8628
    },
    {
      "entropy": 1.1215861141681671,
      "epoch": 4.543970510795155,
      "grad_norm": 0.3233654201030731,
      "learning_rate": 7.21399961024078e-05,
      "loss": 0.16879481077194214,
      "mean_token_accuracy": 0.927575096487999,
      "num_tokens": 106443232.0,
      "step": 8629
    },
    {
      "entropy": 1.1291065514087677,
      "epoch": 4.54449710373881,
      "grad_norm": 0.32695838809013367,
      "learning_rate": 7.212010354944409e-05,
      "loss": 0.17683281004428864,
      "mean_token_accuracy": 0.9240487813949585,
      "num_tokens": 106455568.0,
      "step": 8630
    },
    {
      "entropy": 1.073424756526947,
      "epoch": 4.545023696682464,
      "grad_norm": 0.3119884729385376,
      "learning_rate": 7.21002132451891e-05,
      "loss": 0.16512653231620789,
      "mean_token_accuracy": 0.9337073564529419,
      "num_tokens": 106467904.0,
      "step": 8631
    },
    {
      "entropy": 1.1196120381355286,
      "epoch": 4.545550289626119,
      "grad_norm": 0.35574230551719666,
      "learning_rate": 7.20803251908235e-05,
      "loss": 0.17509186267852783,
      "mean_token_accuracy": 0.9257360398769379,
      "num_tokens": 106480240.0,
      "step": 8632
    },
    {
      "entropy": 1.0766751170158386,
      "epoch": 4.546076882569774,
      "grad_norm": 0.31140777468681335,
      "learning_rate": 7.206043938752796e-05,
      "loss": 0.16393479704856873,
      "mean_token_accuracy": 0.9272533506155014,
      "num_tokens": 106492576.0,
      "step": 8633
    },
    {
      "entropy": 1.1171353459358215,
      "epoch": 4.546603475513428,
      "grad_norm": 0.3182012140750885,
      "learning_rate": 7.204055583648301e-05,
      "loss": 0.15968182682991028,
      "mean_token_accuracy": 0.9303576350212097,
      "num_tokens": 106504912.0,
      "step": 8634
    },
    {
      "entropy": 1.0699645578861237,
      "epoch": 4.547130068457083,
      "grad_norm": 0.3135298192501068,
      "learning_rate": 7.202067453886902e-05,
      "loss": 0.16891828179359436,
      "mean_token_accuracy": 0.9271212965250015,
      "num_tokens": 106517248.0,
      "step": 8635
    },
    {
      "entropy": 1.1310140788555145,
      "epoch": 4.547656661400737,
      "grad_norm": 0.3326754570007324,
      "learning_rate": 7.200079549586627e-05,
      "loss": 0.15686854720115662,
      "mean_token_accuracy": 0.933062732219696,
      "num_tokens": 106529584.0,
      "step": 8636
    },
    {
      "entropy": 1.087046355009079,
      "epoch": 4.548183254344392,
      "grad_norm": 0.30275726318359375,
      "learning_rate": 7.198091870865475e-05,
      "loss": 0.16297847032546997,
      "mean_token_accuracy": 0.9343578368425369,
      "num_tokens": 106541920.0,
      "step": 8637
    },
    {
      "entropy": 1.1067259907722473,
      "epoch": 4.5487098472880465,
      "grad_norm": 0.32432666420936584,
      "learning_rate": 7.196104417841456e-05,
      "loss": 0.1783636063337326,
      "mean_token_accuracy": 0.9246051460504532,
      "num_tokens": 106554256.0,
      "step": 8638
    },
    {
      "entropy": 1.0900868475437164,
      "epoch": 4.5492364402317005,
      "grad_norm": 0.3207528293132782,
      "learning_rate": 7.19411719063254e-05,
      "loss": 0.17123816907405853,
      "mean_token_accuracy": 0.9290398508310318,
      "num_tokens": 106566592.0,
      "step": 8639
    },
    {
      "entropy": 1.061654806137085,
      "epoch": 4.549763033175355,
      "grad_norm": 0.282748818397522,
      "learning_rate": 7.192130189356709e-05,
      "loss": 0.15271571278572083,
      "mean_token_accuracy": 0.9335532784461975,
      "num_tokens": 106578928.0,
      "step": 8640
    },
    {
      "entropy": 1.0427027940750122,
      "epoch": 4.55028962611901,
      "grad_norm": 0.28999966382980347,
      "learning_rate": 7.190143414131916e-05,
      "loss": 0.15280868113040924,
      "mean_token_accuracy": 0.9352436363697052,
      "num_tokens": 106591264.0,
      "step": 8641
    },
    {
      "entropy": 1.0661968886852264,
      "epoch": 4.550816219062664,
      "grad_norm": 0.3222913444042206,
      "learning_rate": 7.188156865076099e-05,
      "loss": 0.16964468359947205,
      "mean_token_accuracy": 0.9289585947990417,
      "num_tokens": 106603600.0,
      "step": 8642
    },
    {
      "entropy": 1.0490186661481857,
      "epoch": 4.551342812006319,
      "grad_norm": 0.2893037796020508,
      "learning_rate": 7.186170542307193e-05,
      "loss": 0.1633986383676529,
      "mean_token_accuracy": 0.9309844374656677,
      "num_tokens": 106615936.0,
      "step": 8643
    },
    {
      "entropy": 1.129934698343277,
      "epoch": 4.551869404949974,
      "grad_norm": 0.3267245590686798,
      "learning_rate": 7.184184445943111e-05,
      "loss": 0.1662716567516327,
      "mean_token_accuracy": 0.9279222786426544,
      "num_tokens": 106628272.0,
      "step": 8644
    },
    {
      "entropy": 1.0623022317886353,
      "epoch": 4.552395997893628,
      "grad_norm": 0.3231813311576843,
      "learning_rate": 7.182198576101762e-05,
      "loss": 0.1643461138010025,
      "mean_token_accuracy": 0.9299049526453018,
      "num_tokens": 106640608.0,
      "step": 8645
    },
    {
      "entropy": 1.1140716075897217,
      "epoch": 4.552922590837283,
      "grad_norm": 0.31757470965385437,
      "learning_rate": 7.180212932901031e-05,
      "loss": 0.16377797722816467,
      "mean_token_accuracy": 0.9325571209192276,
      "num_tokens": 106652944.0,
      "step": 8646
    },
    {
      "entropy": 1.0587803423404694,
      "epoch": 4.553449183780938,
      "grad_norm": 0.31495264172554016,
      "learning_rate": 7.178227516458789e-05,
      "loss": 0.16437000036239624,
      "mean_token_accuracy": 0.9316940456628799,
      "num_tokens": 106665280.0,
      "step": 8647
    },
    {
      "entropy": 1.0641800165176392,
      "epoch": 4.553975776724592,
      "grad_norm": 0.3161664307117462,
      "learning_rate": 7.17624232689291e-05,
      "loss": 0.16967880725860596,
      "mean_token_accuracy": 0.9243513494729996,
      "num_tokens": 106677616.0,
      "step": 8648
    },
    {
      "entropy": 1.064298003911972,
      "epoch": 4.5545023696682465,
      "grad_norm": 0.3035197854042053,
      "learning_rate": 7.174257364321239e-05,
      "loss": 0.1707356572151184,
      "mean_token_accuracy": 0.9280406534671783,
      "num_tokens": 106689952.0,
      "step": 8649
    },
    {
      "entropy": 1.0933890044689178,
      "epoch": 4.555028962611901,
      "grad_norm": 0.33313021063804626,
      "learning_rate": 7.17227262886161e-05,
      "loss": 0.1591399610042572,
      "mean_token_accuracy": 0.9340995401144028,
      "num_tokens": 106702288.0,
      "step": 8650
    },
    {
      "entropy": 1.0891567170619965,
      "epoch": 4.555555555555555,
      "grad_norm": 0.3368584215641022,
      "learning_rate": 7.170288120631843e-05,
      "loss": 0.18220099806785583,
      "mean_token_accuracy": 0.9209961295127869,
      "num_tokens": 106714624.0,
      "step": 8651
    },
    {
      "entropy": 1.1064401865005493,
      "epoch": 4.55608214849921,
      "grad_norm": 0.36571386456489563,
      "learning_rate": 7.16830383974975e-05,
      "loss": 0.1797991842031479,
      "mean_token_accuracy": 0.9172768592834473,
      "num_tokens": 106726960.0,
      "step": 8652
    },
    {
      "entropy": 1.0690472424030304,
      "epoch": 4.556608741442865,
      "grad_norm": 0.3507973849773407,
      "learning_rate": 7.166319786333125e-05,
      "loss": 0.1866639405488968,
      "mean_token_accuracy": 0.9236337691545486,
      "num_tokens": 106739296.0,
      "step": 8653
    },
    {
      "entropy": 0.9976196885108948,
      "epoch": 4.557135334386519,
      "grad_norm": 0.31047511100769043,
      "learning_rate": 7.164335960499744e-05,
      "loss": 0.16871517896652222,
      "mean_token_accuracy": 0.9263797998428345,
      "num_tokens": 106751632.0,
      "step": 8654
    },
    {
      "entropy": 1.0752581059932709,
      "epoch": 4.557661927330174,
      "grad_norm": 0.32285797595977783,
      "learning_rate": 7.162352362367385e-05,
      "loss": 0.16266748309135437,
      "mean_token_accuracy": 0.9306293576955795,
      "num_tokens": 106763968.0,
      "step": 8655
    },
    {
      "entropy": 1.1051332354545593,
      "epoch": 4.558188520273829,
      "grad_norm": 0.3700147867202759,
      "learning_rate": 7.160368992053803e-05,
      "loss": 0.19394490122795105,
      "mean_token_accuracy": 0.9186254739761353,
      "num_tokens": 106776304.0,
      "step": 8656
    },
    {
      "entropy": 1.0695521384477615,
      "epoch": 4.558715113217483,
      "grad_norm": 0.31823670864105225,
      "learning_rate": 7.158385849676732e-05,
      "loss": 0.16147494316101074,
      "mean_token_accuracy": 0.9299113750457764,
      "num_tokens": 106788640.0,
      "step": 8657
    },
    {
      "entropy": 1.0720454156398773,
      "epoch": 4.559241706161138,
      "grad_norm": 0.3523479700088501,
      "learning_rate": 7.156402935353903e-05,
      "loss": 0.16577114164829254,
      "mean_token_accuracy": 0.9317851662635803,
      "num_tokens": 106800976.0,
      "step": 8658
    },
    {
      "entropy": 1.092285007238388,
      "epoch": 4.559768299104792,
      "grad_norm": 0.35989680886268616,
      "learning_rate": 7.154420249203028e-05,
      "loss": 0.1683732271194458,
      "mean_token_accuracy": 0.9251654595136642,
      "num_tokens": 106813312.0,
      "step": 8659
    },
    {
      "entropy": 1.048419177532196,
      "epoch": 4.5602948920484465,
      "grad_norm": 0.2887498438358307,
      "learning_rate": 7.152437791341814e-05,
      "loss": 0.1596927046775818,
      "mean_token_accuracy": 0.9350206553936005,
      "num_tokens": 106825648.0,
      "step": 8660
    },
    {
      "entropy": 1.0721503496170044,
      "epoch": 4.560821484992101,
      "grad_norm": 0.32527533173561096,
      "learning_rate": 7.150455561887937e-05,
      "loss": 0.16587302088737488,
      "mean_token_accuracy": 0.9299864172935486,
      "num_tokens": 106837984.0,
      "step": 8661
    },
    {
      "entropy": 1.1039204895496368,
      "epoch": 4.561348077935755,
      "grad_norm": 0.3423987329006195,
      "learning_rate": 7.148473560959077e-05,
      "loss": 0.1740478277206421,
      "mean_token_accuracy": 0.9244489520788193,
      "num_tokens": 106850320.0,
      "step": 8662
    },
    {
      "entropy": 1.0168361812829971,
      "epoch": 4.56187467087941,
      "grad_norm": 0.32267218828201294,
      "learning_rate": 7.146491788672895e-05,
      "loss": 0.1730199158191681,
      "mean_token_accuracy": 0.9277898073196411,
      "num_tokens": 106862656.0,
      "step": 8663
    },
    {
      "entropy": 1.0239315032958984,
      "epoch": 4.562401263823065,
      "grad_norm": 0.311694860458374,
      "learning_rate": 7.144510245147038e-05,
      "loss": 0.1736176609992981,
      "mean_token_accuracy": 0.9258221387863159,
      "num_tokens": 106874992.0,
      "step": 8664
    },
    {
      "entropy": 1.0748440623283386,
      "epoch": 4.562927856766719,
      "grad_norm": 0.328838974237442,
      "learning_rate": 7.142528930499133e-05,
      "loss": 0.17330893874168396,
      "mean_token_accuracy": 0.9279798716306686,
      "num_tokens": 106887328.0,
      "step": 8665
    },
    {
      "entropy": 1.0589421093463898,
      "epoch": 4.563454449710374,
      "grad_norm": 0.31295597553253174,
      "learning_rate": 7.140547844846803e-05,
      "loss": 0.15671877562999725,
      "mean_token_accuracy": 0.9362427443265915,
      "num_tokens": 106899664.0,
      "step": 8666
    },
    {
      "entropy": 1.0276417583227158,
      "epoch": 4.563981042654029,
      "grad_norm": 0.32179734110832214,
      "learning_rate": 7.138566988307654e-05,
      "loss": 0.16797716915607452,
      "mean_token_accuracy": 0.9260842502117157,
      "num_tokens": 106912000.0,
      "step": 8667
    },
    {
      "entropy": 1.118698626756668,
      "epoch": 4.564507635597683,
      "grad_norm": 0.3249455690383911,
      "learning_rate": 7.136586360999275e-05,
      "loss": 0.19012728333473206,
      "mean_token_accuracy": 0.9203059673309326,
      "num_tokens": 106924336.0,
      "step": 8668
    },
    {
      "entropy": 1.0857931673526764,
      "epoch": 4.565034228541338,
      "grad_norm": 0.32796764373779297,
      "learning_rate": 7.134605963039245e-05,
      "loss": 0.17321154475212097,
      "mean_token_accuracy": 0.9267846494913101,
      "num_tokens": 106936672.0,
      "step": 8669
    },
    {
      "entropy": 1.0889154821634293,
      "epoch": 4.565560821484992,
      "grad_norm": 0.3461117446422577,
      "learning_rate": 7.132625794545134e-05,
      "loss": 0.18601475656032562,
      "mean_token_accuracy": 0.9206071197986603,
      "num_tokens": 106949008.0,
      "step": 8670
    },
    {
      "entropy": 1.1496550142765045,
      "epoch": 4.5660874144286465,
      "grad_norm": 0.33955106139183044,
      "learning_rate": 7.130645855634486e-05,
      "loss": 0.18470525741577148,
      "mean_token_accuracy": 0.9172615110874176,
      "num_tokens": 106961344.0,
      "step": 8671
    },
    {
      "entropy": 1.1040054261684418,
      "epoch": 4.566614007372301,
      "grad_norm": 0.3226144015789032,
      "learning_rate": 7.12866614642484e-05,
      "loss": 0.1825535148382187,
      "mean_token_accuracy": 0.922840341925621,
      "num_tokens": 106973680.0,
      "step": 8672
    },
    {
      "entropy": 1.1121874451637268,
      "epoch": 4.567140600315955,
      "grad_norm": 0.31177595257759094,
      "learning_rate": 7.126686667033724e-05,
      "loss": 0.18004962801933289,
      "mean_token_accuracy": 0.9256362766027451,
      "num_tokens": 106986016.0,
      "step": 8673
    },
    {
      "entropy": 1.1306947767734528,
      "epoch": 4.56766719325961,
      "grad_norm": 0.3360869586467743,
      "learning_rate": 7.124707417578645e-05,
      "loss": 0.18021290004253387,
      "mean_token_accuracy": 0.9254167079925537,
      "num_tokens": 106998352.0,
      "step": 8674
    },
    {
      "entropy": 1.0924843549728394,
      "epoch": 4.568193786203265,
      "grad_norm": 0.2975752055644989,
      "learning_rate": 7.122728398177095e-05,
      "loss": 0.15803509950637817,
      "mean_token_accuracy": 0.9346482157707214,
      "num_tokens": 107010688.0,
      "step": 8675
    },
    {
      "entropy": 1.0998488664627075,
      "epoch": 4.568720379146919,
      "grad_norm": 0.30178508162498474,
      "learning_rate": 7.120749608946565e-05,
      "loss": 0.17362216114997864,
      "mean_token_accuracy": 0.9280165731906891,
      "num_tokens": 107023024.0,
      "step": 8676
    },
    {
      "entropy": 1.1455123126506805,
      "epoch": 4.569246972090574,
      "grad_norm": 0.3140106499195099,
      "learning_rate": 7.11877105000452e-05,
      "loss": 0.18174360692501068,
      "mean_token_accuracy": 0.9210865646600723,
      "num_tokens": 107035360.0,
      "step": 8677
    },
    {
      "entropy": 1.045901894569397,
      "epoch": 4.569773565034229,
      "grad_norm": 0.29214707016944885,
      "learning_rate": 7.116792721468417e-05,
      "loss": 0.1577940732240677,
      "mean_token_accuracy": 0.936051145195961,
      "num_tokens": 107047696.0,
      "step": 8678
    },
    {
      "entropy": 1.044377788901329,
      "epoch": 4.570300157977883,
      "grad_norm": 0.2773629128932953,
      "learning_rate": 7.1148146234557e-05,
      "loss": 0.15500973165035248,
      "mean_token_accuracy": 0.9314758628606796,
      "num_tokens": 107060032.0,
      "step": 8679
    },
    {
      "entropy": 1.1480259597301483,
      "epoch": 4.570826750921538,
      "grad_norm": 0.3102291226387024,
      "learning_rate": 7.112836756083791e-05,
      "loss": 0.16833597421646118,
      "mean_token_accuracy": 0.9297477602958679,
      "num_tokens": 107072368.0,
      "step": 8680
    },
    {
      "entropy": 1.1289499998092651,
      "epoch": 4.5713533438651925,
      "grad_norm": 0.31731173396110535,
      "learning_rate": 7.110859119470107e-05,
      "loss": 0.16683533787727356,
      "mean_token_accuracy": 0.9299915730953217,
      "num_tokens": 107084704.0,
      "step": 8681
    },
    {
      "entropy": 1.1283690333366394,
      "epoch": 4.5718799368088465,
      "grad_norm": 0.317582905292511,
      "learning_rate": 7.108881713732049e-05,
      "loss": 0.17901158332824707,
      "mean_token_accuracy": 0.9264525324106216,
      "num_tokens": 107097040.0,
      "step": 8682
    },
    {
      "entropy": 1.1011227369308472,
      "epoch": 4.572406529752501,
      "grad_norm": 0.2942031919956207,
      "learning_rate": 7.106904538987006e-05,
      "loss": 0.17026060819625854,
      "mean_token_accuracy": 0.9274697154760361,
      "num_tokens": 107109376.0,
      "step": 8683
    },
    {
      "entropy": 1.1156458854675293,
      "epoch": 4.572933122696156,
      "grad_norm": 0.30721092224121094,
      "learning_rate": 7.10492759535235e-05,
      "loss": 0.18025022745132446,
      "mean_token_accuracy": 0.9253346472978592,
      "num_tokens": 107121712.0,
      "step": 8684
    },
    {
      "entropy": 1.0282520651817322,
      "epoch": 4.57345971563981,
      "grad_norm": 0.29607510566711426,
      "learning_rate": 7.10295088294544e-05,
      "loss": 0.16659298539161682,
      "mean_token_accuracy": 0.9305984079837799,
      "num_tokens": 107134048.0,
      "step": 8685
    },
    {
      "entropy": 1.1497196555137634,
      "epoch": 4.573986308583465,
      "grad_norm": 0.3343541622161865,
      "learning_rate": 7.100974401883622e-05,
      "loss": 0.17614148557186127,
      "mean_token_accuracy": 0.9200554937124252,
      "num_tokens": 107146384.0,
      "step": 8686
    },
    {
      "entropy": 1.1118314564228058,
      "epoch": 4.57451290152712,
      "grad_norm": 0.31019845604896545,
      "learning_rate": 7.098998152284229e-05,
      "loss": 0.17119279503822327,
      "mean_token_accuracy": 0.923244521021843,
      "num_tokens": 107158720.0,
      "step": 8687
    },
    {
      "entropy": 1.0423270016908646,
      "epoch": 4.575039494470774,
      "grad_norm": 0.2982739210128784,
      "learning_rate": 7.09702213426458e-05,
      "loss": 0.1640046238899231,
      "mean_token_accuracy": 0.9256590902805328,
      "num_tokens": 107171056.0,
      "step": 8688
    },
    {
      "entropy": 1.08947092294693,
      "epoch": 4.575566087414429,
      "grad_norm": 0.3556402921676636,
      "learning_rate": 7.095046347941978e-05,
      "loss": 0.18411381542682648,
      "mean_token_accuracy": 0.9257789105176926,
      "num_tokens": 107183392.0,
      "step": 8689
    },
    {
      "entropy": 1.0914960205554962,
      "epoch": 4.576092680358084,
      "grad_norm": 0.321100652217865,
      "learning_rate": 7.093070793433714e-05,
      "loss": 0.1736738234758377,
      "mean_token_accuracy": 0.9281085878610611,
      "num_tokens": 107195728.0,
      "step": 8690
    },
    {
      "entropy": 1.0638926923274994,
      "epoch": 4.576619273301738,
      "grad_norm": 0.3066602051258087,
      "learning_rate": 7.091095470857066e-05,
      "loss": 0.15998324751853943,
      "mean_token_accuracy": 0.9323301017284393,
      "num_tokens": 107208064.0,
      "step": 8691
    },
    {
      "entropy": 1.0748650431632996,
      "epoch": 4.5771458662453925,
      "grad_norm": 0.32310837507247925,
      "learning_rate": 7.089120380329298e-05,
      "loss": 0.1661323606967926,
      "mean_token_accuracy": 0.9274686872959137,
      "num_tokens": 107220400.0,
      "step": 8692
    },
    {
      "entropy": 1.0882180035114288,
      "epoch": 4.5776724591890465,
      "grad_norm": 0.3006623089313507,
      "learning_rate": 7.087145521967658e-05,
      "loss": 0.16791214048862457,
      "mean_token_accuracy": 0.9302281737327576,
      "num_tokens": 107232736.0,
      "step": 8693
    },
    {
      "entropy": 1.0967628061771393,
      "epoch": 4.578199052132701,
      "grad_norm": 0.31030550599098206,
      "learning_rate": 7.085170895889382e-05,
      "loss": 0.1743021011352539,
      "mean_token_accuracy": 0.9264057874679565,
      "num_tokens": 107245072.0,
      "step": 8694
    },
    {
      "entropy": 1.0810537934303284,
      "epoch": 4.578725645076356,
      "grad_norm": 0.3004099130630493,
      "learning_rate": 7.083196502211693e-05,
      "loss": 0.16865485906600952,
      "mean_token_accuracy": 0.9283222258090973,
      "num_tokens": 107257408.0,
      "step": 8695
    },
    {
      "entropy": 1.064576044678688,
      "epoch": 4.57925223802001,
      "grad_norm": 0.3168226480484009,
      "learning_rate": 7.081222341051794e-05,
      "loss": 0.17242692410945892,
      "mean_token_accuracy": 0.9287631809711456,
      "num_tokens": 107269744.0,
      "step": 8696
    },
    {
      "entropy": 0.9771599024534225,
      "epoch": 4.579778830963665,
      "grad_norm": 0.2934156656265259,
      "learning_rate": 7.079248412526888e-05,
      "loss": 0.16696205735206604,
      "mean_token_accuracy": 0.9282114207744598,
      "num_tokens": 107282080.0,
      "step": 8697
    },
    {
      "entropy": 1.0334157943725586,
      "epoch": 4.58030542390732,
      "grad_norm": 0.3131554424762726,
      "learning_rate": 7.077274716754156e-05,
      "loss": 0.16662999987602234,
      "mean_token_accuracy": 0.9287344515323639,
      "num_tokens": 107294416.0,
      "step": 8698
    },
    {
      "entropy": 0.9986395239830017,
      "epoch": 4.580832016850974,
      "grad_norm": 0.32664307951927185,
      "learning_rate": 7.075301253850759e-05,
      "loss": 0.17536792159080505,
      "mean_token_accuracy": 0.9243778139352798,
      "num_tokens": 107306752.0,
      "step": 8699
    },
    {
      "entropy": 1.0363193899393082,
      "epoch": 4.581358609794629,
      "grad_norm": 0.33753758668899536,
      "learning_rate": 7.073328023933849e-05,
      "loss": 0.17446276545524597,
      "mean_token_accuracy": 0.9281366765499115,
      "num_tokens": 107319088.0,
      "step": 8700
    },
    {
      "entropy": 1.0517511069774628,
      "epoch": 4.581885202738283,
      "grad_norm": 0.3835679590702057,
      "learning_rate": 7.07135502712057e-05,
      "loss": 0.1867254376411438,
      "mean_token_accuracy": 0.9165402054786682,
      "num_tokens": 107331424.0,
      "step": 8701
    },
    {
      "entropy": 1.0205737203359604,
      "epoch": 4.582411795681938,
      "grad_norm": 0.31008943915367126,
      "learning_rate": 7.069382263528043e-05,
      "loss": 0.16276168823242188,
      "mean_token_accuracy": 0.9319659471511841,
      "num_tokens": 107343760.0,
      "step": 8702
    },
    {
      "entropy": 0.9978193044662476,
      "epoch": 4.5829383886255926,
      "grad_norm": 0.329265296459198,
      "learning_rate": 7.06740973327338e-05,
      "loss": 0.16471801698207855,
      "mean_token_accuracy": 0.925895556807518,
      "num_tokens": 107356096.0,
      "step": 8703
    },
    {
      "entropy": 1.0166545361280441,
      "epoch": 4.5834649815692465,
      "grad_norm": 0.31310710310935974,
      "learning_rate": 7.065437436473683e-05,
      "loss": 0.1673336625099182,
      "mean_token_accuracy": 0.9306180328130722,
      "num_tokens": 107368432.0,
      "step": 8704
    },
    {
      "entropy": 0.9913676977157593,
      "epoch": 4.583991574512901,
      "grad_norm": 0.30291885137557983,
      "learning_rate": 7.063465373246032e-05,
      "loss": 0.1598743349313736,
      "mean_token_accuracy": 0.9280618876218796,
      "num_tokens": 107380768.0,
      "step": 8705
    },
    {
      "entropy": 0.9993337392807007,
      "epoch": 4.584518167456556,
      "grad_norm": 0.3154926300048828,
      "learning_rate": 7.061493543707501e-05,
      "loss": 0.16369374096393585,
      "mean_token_accuracy": 0.9313799887895584,
      "num_tokens": 107393104.0,
      "step": 8706
    },
    {
      "entropy": 1.044825702905655,
      "epoch": 4.58504476040021,
      "grad_norm": 0.33816179633140564,
      "learning_rate": 7.059521947975139e-05,
      "loss": 0.19750162959098816,
      "mean_token_accuracy": 0.9153089374303818,
      "num_tokens": 107405440.0,
      "step": 8707
    },
    {
      "entropy": 1.047785148024559,
      "epoch": 4.585571353343865,
      "grad_norm": 0.3284870684146881,
      "learning_rate": 7.057550586165993e-05,
      "loss": 0.17213356494903564,
      "mean_token_accuracy": 0.9218959957361221,
      "num_tokens": 107417776.0,
      "step": 8708
    },
    {
      "entropy": 0.9950270652770996,
      "epoch": 4.58609794628752,
      "grad_norm": 0.4453790783882141,
      "learning_rate": 7.055579458397097e-05,
      "loss": 0.16260553896427155,
      "mean_token_accuracy": 0.9297661185264587,
      "num_tokens": 107430112.0,
      "step": 8709
    },
    {
      "entropy": 1.0443234890699387,
      "epoch": 4.586624539231174,
      "grad_norm": 0.3269834518432617,
      "learning_rate": 7.05360856478545e-05,
      "loss": 0.17089609801769257,
      "mean_token_accuracy": 0.9307948648929596,
      "num_tokens": 107442448.0,
      "step": 8710
    },
    {
      "entropy": 1.0410026907920837,
      "epoch": 4.587151132174829,
      "grad_norm": 0.2909794747829437,
      "learning_rate": 7.051637905448063e-05,
      "loss": 0.15479862689971924,
      "mean_token_accuracy": 0.9330165535211563,
      "num_tokens": 107454784.0,
      "step": 8711
    },
    {
      "entropy": 1.0876724123954773,
      "epoch": 4.587677725118484,
      "grad_norm": 0.3122361898422241,
      "learning_rate": 7.049667480501922e-05,
      "loss": 0.1811986118555069,
      "mean_token_accuracy": 0.9227711260318756,
      "num_tokens": 107467120.0,
      "step": 8712
    },
    {
      "entropy": 1.0514351725578308,
      "epoch": 4.588204318062138,
      "grad_norm": 0.33499598503112793,
      "learning_rate": 7.047697290063997e-05,
      "loss": 0.1878758668899536,
      "mean_token_accuracy": 0.9214672744274139,
      "num_tokens": 107479456.0,
      "step": 8713
    },
    {
      "entropy": 0.9934461116790771,
      "epoch": 4.588730911005793,
      "grad_norm": 0.3381115198135376,
      "learning_rate": 7.04572733425125e-05,
      "loss": 0.18743182718753815,
      "mean_token_accuracy": 0.9202657341957092,
      "num_tokens": 107491792.0,
      "step": 8714
    },
    {
      "entropy": 1.0088402181863785,
      "epoch": 4.589257503949447,
      "grad_norm": 0.2974151074886322,
      "learning_rate": 7.043757613180623e-05,
      "loss": 0.16457468271255493,
      "mean_token_accuracy": 0.926583006978035,
      "num_tokens": 107504128.0,
      "step": 8715
    },
    {
      "entropy": 1.0807583034038544,
      "epoch": 4.589784096893101,
      "grad_norm": 0.4040805697441101,
      "learning_rate": 7.041788126969048e-05,
      "loss": 0.18308116495609283,
      "mean_token_accuracy": 0.9198445230722427,
      "num_tokens": 107516464.0,
      "step": 8716
    },
    {
      "entropy": 0.9993950575590134,
      "epoch": 4.590310689836756,
      "grad_norm": 0.3046441376209259,
      "learning_rate": 7.039818875733444e-05,
      "loss": 0.15982067584991455,
      "mean_token_accuracy": 0.9315261244773865,
      "num_tokens": 107528800.0,
      "step": 8717
    },
    {
      "entropy": 1.0157035887241364,
      "epoch": 4.590837282780411,
      "grad_norm": 0.2929827868938446,
      "learning_rate": 7.037849859590708e-05,
      "loss": 0.15987715125083923,
      "mean_token_accuracy": 0.9329131245613098,
      "num_tokens": 107541136.0,
      "step": 8718
    },
    {
      "entropy": 1.053165391087532,
      "epoch": 4.591363875724065,
      "grad_norm": 0.3398328125476837,
      "learning_rate": 7.035881078657736e-05,
      "loss": 0.17398418486118317,
      "mean_token_accuracy": 0.9235986769199371,
      "num_tokens": 107553472.0,
      "step": 8719
    },
    {
      "entropy": 1.1066934168338776,
      "epoch": 4.59189046866772,
      "grad_norm": 0.3254387080669403,
      "learning_rate": 7.033912533051398e-05,
      "loss": 0.1771440953016281,
      "mean_token_accuracy": 0.9198944419622421,
      "num_tokens": 107565808.0,
      "step": 8720
    },
    {
      "entropy": 1.0543211102485657,
      "epoch": 4.592417061611375,
      "grad_norm": 0.3355952799320221,
      "learning_rate": 7.031944222888555e-05,
      "loss": 0.15953224897384644,
      "mean_token_accuracy": 0.9345828890800476,
      "num_tokens": 107578144.0,
      "step": 8721
    },
    {
      "entropy": 1.033324733376503,
      "epoch": 4.592943654555029,
      "grad_norm": 0.335302472114563,
      "learning_rate": 7.02997614828606e-05,
      "loss": 0.18134362995624542,
      "mean_token_accuracy": 0.9263257384300232,
      "num_tokens": 107590480.0,
      "step": 8722
    },
    {
      "entropy": 1.0930776596069336,
      "epoch": 4.593470247498684,
      "grad_norm": 0.3225703835487366,
      "learning_rate": 7.028008309360739e-05,
      "loss": 0.17418448626995087,
      "mean_token_accuracy": 0.9256875365972519,
      "num_tokens": 107602816.0,
      "step": 8723
    },
    {
      "entropy": 1.021464854478836,
      "epoch": 4.593996840442338,
      "grad_norm": 0.3606475293636322,
      "learning_rate": 7.02604070622941e-05,
      "loss": 0.1569170355796814,
      "mean_token_accuracy": 0.9325631409883499,
      "num_tokens": 107615152.0,
      "step": 8724
    },
    {
      "entropy": 1.0355309695005417,
      "epoch": 4.594523433385993,
      "grad_norm": 0.29902201890945435,
      "learning_rate": 7.024073339008887e-05,
      "loss": 0.1473335176706314,
      "mean_token_accuracy": 0.9369276762008667,
      "num_tokens": 107627488.0,
      "step": 8725
    },
    {
      "entropy": 1.0017937123775482,
      "epoch": 4.595050026329647,
      "grad_norm": 0.31027984619140625,
      "learning_rate": 7.022106207815957e-05,
      "loss": 0.1686737835407257,
      "mean_token_accuracy": 0.9254660606384277,
      "num_tokens": 107639824.0,
      "step": 8726
    },
    {
      "entropy": 0.9633382558822632,
      "epoch": 4.595576619273301,
      "grad_norm": 0.31015104055404663,
      "learning_rate": 7.020139312767396e-05,
      "loss": 0.1631452441215515,
      "mean_token_accuracy": 0.9335481226444244,
      "num_tokens": 107652160.0,
      "step": 8727
    },
    {
      "entropy": 0.9984524995088577,
      "epoch": 4.596103212216956,
      "grad_norm": 0.32390734553337097,
      "learning_rate": 7.01817265397997e-05,
      "loss": 0.1819961965084076,
      "mean_token_accuracy": 0.9274021834135056,
      "num_tokens": 107664496.0,
      "step": 8728
    },
    {
      "entropy": 0.9658143371343613,
      "epoch": 4.596629805160611,
      "grad_norm": 0.2848398685455322,
      "learning_rate": 7.016206231570426e-05,
      "loss": 0.16318723559379578,
      "mean_token_accuracy": 0.929384633898735,
      "num_tokens": 107676832.0,
      "step": 8729
    },
    {
      "entropy": 1.0240425169467926,
      "epoch": 4.597156398104265,
      "grad_norm": 0.310256689786911,
      "learning_rate": 7.014240045655494e-05,
      "loss": 0.17658701539039612,
      "mean_token_accuracy": 0.9227228760719299,
      "num_tokens": 107689168.0,
      "step": 8730
    },
    {
      "entropy": 1.0089117288589478,
      "epoch": 4.59768299104792,
      "grad_norm": 0.3132132589817047,
      "learning_rate": 7.012274096351898e-05,
      "loss": 0.18013951182365417,
      "mean_token_accuracy": 0.9222273379564285,
      "num_tokens": 107701504.0,
      "step": 8731
    },
    {
      "entropy": 1.0355899930000305,
      "epoch": 4.598209583991575,
      "grad_norm": 0.3282855451107025,
      "learning_rate": 7.010308383776352e-05,
      "loss": 0.1806597262620926,
      "mean_token_accuracy": 0.923142671585083,
      "num_tokens": 107713840.0,
      "step": 8732
    },
    {
      "entropy": 1.0216601938009262,
      "epoch": 4.598736176935229,
      "grad_norm": 0.3035755455493927,
      "learning_rate": 7.008342908045542e-05,
      "loss": 0.17160706222057343,
      "mean_token_accuracy": 0.9253741949796677,
      "num_tokens": 107726176.0,
      "step": 8733
    },
    {
      "entropy": 1.0056361258029938,
      "epoch": 4.599262769878884,
      "grad_norm": 0.28573930263519287,
      "learning_rate": 7.006377669276147e-05,
      "loss": 0.15307191014289856,
      "mean_token_accuracy": 0.9353896081447601,
      "num_tokens": 107738512.0,
      "step": 8734
    },
    {
      "entropy": 1.0611897706985474,
      "epoch": 4.599789362822538,
      "grad_norm": 0.32753387093544006,
      "learning_rate": 7.004412667584834e-05,
      "loss": 0.17335793375968933,
      "mean_token_accuracy": 0.9274525344371796,
      "num_tokens": 107750848.0,
      "step": 8735
    },
    {
      "entropy": 1.003476321697235,
      "epoch": 4.600315955766193,
      "grad_norm": 0.3149878978729248,
      "learning_rate": 7.002447903088254e-05,
      "loss": 0.16786393523216248,
      "mean_token_accuracy": 0.928516298532486,
      "num_tokens": 107763184.0,
      "step": 8736
    },
    {
      "entropy": 1.010802537202835,
      "epoch": 4.600842548709847,
      "grad_norm": 0.2895418107509613,
      "learning_rate": 7.000483375903045e-05,
      "loss": 0.15088866651058197,
      "mean_token_accuracy": 0.9328432381153107,
      "num_tokens": 107775520.0,
      "step": 8737
    },
    {
      "entropy": 0.9834362864494324,
      "epoch": 4.601369141653501,
      "grad_norm": 0.3100012242794037,
      "learning_rate": 6.998519086145822e-05,
      "loss": 0.16654789447784424,
      "mean_token_accuracy": 0.9275851845741272,
      "num_tokens": 107787856.0,
      "step": 8738
    },
    {
      "entropy": 0.9788924306631088,
      "epoch": 4.601895734597156,
      "grad_norm": 0.33185940980911255,
      "learning_rate": 6.9965550339332e-05,
      "loss": 0.1668771505355835,
      "mean_token_accuracy": 0.9286605268716812,
      "num_tokens": 107800192.0,
      "step": 8739
    },
    {
      "entropy": 0.9902924746274948,
      "epoch": 4.602422327540811,
      "grad_norm": 0.3223089873790741,
      "learning_rate": 6.994591219381774e-05,
      "loss": 0.1609772890806198,
      "mean_token_accuracy": 0.9281881749629974,
      "num_tokens": 107812528.0,
      "step": 8740
    },
    {
      "entropy": 0.9932299852371216,
      "epoch": 4.602948920484465,
      "grad_norm": 0.31837165355682373,
      "learning_rate": 6.992627642608118e-05,
      "loss": 0.1766543835401535,
      "mean_token_accuracy": 0.9214552193880081,
      "num_tokens": 107824864.0,
      "step": 8741
    },
    {
      "entropy": 1.0190290212631226,
      "epoch": 4.60347551342812,
      "grad_norm": 0.35585761070251465,
      "learning_rate": 6.990664303728804e-05,
      "loss": 0.18412742018699646,
      "mean_token_accuracy": 0.923671618103981,
      "num_tokens": 107837200.0,
      "step": 8742
    },
    {
      "entropy": 0.9401917308568954,
      "epoch": 4.604002106371775,
      "grad_norm": 0.2916022539138794,
      "learning_rate": 6.988701202860384e-05,
      "loss": 0.15726128220558167,
      "mean_token_accuracy": 0.931135967373848,
      "num_tokens": 107849536.0,
      "step": 8743
    },
    {
      "entropy": 0.9833636432886124,
      "epoch": 4.604528699315429,
      "grad_norm": 0.35866621136665344,
      "learning_rate": 6.986738340119392e-05,
      "loss": 0.18105976283550262,
      "mean_token_accuracy": 0.9182242453098297,
      "num_tokens": 107861872.0,
      "step": 8744
    },
    {
      "entropy": 0.9567048400640488,
      "epoch": 4.605055292259084,
      "grad_norm": 0.3104209899902344,
      "learning_rate": 6.98477571562235e-05,
      "loss": 0.17179551720619202,
      "mean_token_accuracy": 0.9252661168575287,
      "num_tokens": 107874208.0,
      "step": 8745
    },
    {
      "entropy": 0.9775879085063934,
      "epoch": 4.605581885202739,
      "grad_norm": 0.3023075759410858,
      "learning_rate": 6.982813329485776e-05,
      "loss": 0.15816596150398254,
      "mean_token_accuracy": 0.9306709617376328,
      "num_tokens": 107886544.0,
      "step": 8746
    },
    {
      "entropy": 0.9464512020349503,
      "epoch": 4.606108478146393,
      "grad_norm": 0.30875155329704285,
      "learning_rate": 6.980851181826162e-05,
      "loss": 0.1684526652097702,
      "mean_token_accuracy": 0.9325686395168304,
      "num_tokens": 107898880.0,
      "step": 8747
    },
    {
      "entropy": 0.9552883207798004,
      "epoch": 4.606635071090047,
      "grad_norm": 0.3210124671459198,
      "learning_rate": 6.978889272759987e-05,
      "loss": 0.17382028698921204,
      "mean_token_accuracy": 0.9314020872116089,
      "num_tokens": 107911216.0,
      "step": 8748
    },
    {
      "entropy": 0.9750551283359528,
      "epoch": 4.607161664033702,
      "grad_norm": 0.2832907736301422,
      "learning_rate": 6.976927602403718e-05,
      "loss": 0.14793327450752258,
      "mean_token_accuracy": 0.9364615827798843,
      "num_tokens": 107923552.0,
      "step": 8749
    },
    {
      "entropy": 0.9648535847663879,
      "epoch": 4.607688256977356,
      "grad_norm": 0.30289632081985474,
      "learning_rate": 6.97496617087381e-05,
      "loss": 0.1678515374660492,
      "mean_token_accuracy": 0.9286443442106247,
      "num_tokens": 107935888.0,
      "step": 8750
    },
    {
      "entropy": 1.0221524238586426,
      "epoch": 4.608214849921011,
      "grad_norm": 0.3171069920063019,
      "learning_rate": 6.9730049782867e-05,
      "loss": 0.1789323091506958,
      "mean_token_accuracy": 0.9245229810476303,
      "num_tokens": 107948224.0,
      "step": 8751
    },
    {
      "entropy": 1.0002122968435287,
      "epoch": 4.608741442864666,
      "grad_norm": 0.3247002065181732,
      "learning_rate": 6.97104402475881e-05,
      "loss": 0.1693766862154007,
      "mean_token_accuracy": 0.9273964613676071,
      "num_tokens": 107960560.0,
      "step": 8752
    },
    {
      "entropy": 1.0298629105091095,
      "epoch": 4.60926803580832,
      "grad_norm": 0.31016674637794495,
      "learning_rate": 6.969083310406554e-05,
      "loss": 0.15787182748317719,
      "mean_token_accuracy": 0.935954675078392,
      "num_tokens": 107972896.0,
      "step": 8753
    },
    {
      "entropy": 0.9574241787195206,
      "epoch": 4.609794628751975,
      "grad_norm": 0.2948024570941925,
      "learning_rate": 6.967122835346334e-05,
      "loss": 0.15867428481578827,
      "mean_token_accuracy": 0.9365124106407166,
      "num_tokens": 107985232.0,
      "step": 8754
    },
    {
      "entropy": 0.996378943324089,
      "epoch": 4.61032122169563,
      "grad_norm": 0.2876661419868469,
      "learning_rate": 6.965162599694521e-05,
      "loss": 0.15610000491142273,
      "mean_token_accuracy": 0.9314628392457962,
      "num_tokens": 107997568.0,
      "step": 8755
    },
    {
      "entropy": 0.9925418645143509,
      "epoch": 4.610847814639284,
      "grad_norm": 0.32017311453819275,
      "learning_rate": 6.963202603567489e-05,
      "loss": 0.1787632256746292,
      "mean_token_accuracy": 0.9199665635824203,
      "num_tokens": 108009904.0,
      "step": 8756
    },
    {
      "entropy": 1.0136579424142838,
      "epoch": 4.611374407582939,
      "grad_norm": 0.3359333574771881,
      "learning_rate": 6.96124284708159e-05,
      "loss": 0.18175891041755676,
      "mean_token_accuracy": 0.9239500015974045,
      "num_tokens": 108022240.0,
      "step": 8757
    },
    {
      "entropy": 1.0320144146680832,
      "epoch": 4.611901000526593,
      "grad_norm": 0.33800309896469116,
      "learning_rate": 6.959283330353168e-05,
      "loss": 0.18834570050239563,
      "mean_token_accuracy": 0.9215432703495026,
      "num_tokens": 108034576.0,
      "step": 8758
    },
    {
      "entropy": 0.9942899644374847,
      "epoch": 4.6124275934702474,
      "grad_norm": 0.3465676009654999,
      "learning_rate": 6.957324053498535e-05,
      "loss": 0.18794047832489014,
      "mean_token_accuracy": 0.9213247150182724,
      "num_tokens": 108046912.0,
      "step": 8759
    },
    {
      "entropy": 0.9882669001817703,
      "epoch": 4.612954186413902,
      "grad_norm": 0.3110610842704773,
      "learning_rate": 6.955365016634016e-05,
      "loss": 0.16755543649196625,
      "mean_token_accuracy": 0.9325612932443619,
      "num_tokens": 108059248.0,
      "step": 8760
    },
    {
      "entropy": 1.0631378889083862,
      "epoch": 4.613480779357556,
      "grad_norm": 0.33182603120803833,
      "learning_rate": 6.953406219875902e-05,
      "loss": 0.18475112318992615,
      "mean_token_accuracy": 0.9203294217586517,
      "num_tokens": 108071584.0,
      "step": 8761
    },
    {
      "entropy": 1.0360340029001236,
      "epoch": 4.614007372301211,
      "grad_norm": 0.3316214382648468,
      "learning_rate": 6.951447663340474e-05,
      "loss": 0.17253397405147552,
      "mean_token_accuracy": 0.9253484904766083,
      "num_tokens": 108083920.0,
      "step": 8762
    },
    {
      "entropy": 0.9969138652086258,
      "epoch": 4.614533965244866,
      "grad_norm": 0.28674861788749695,
      "learning_rate": 6.949489347144002e-05,
      "loss": 0.15221770107746124,
      "mean_token_accuracy": 0.9338681846857071,
      "num_tokens": 108096256.0,
      "step": 8763
    },
    {
      "entropy": 0.9754995256662369,
      "epoch": 4.61506055818852,
      "grad_norm": 0.3082277476787567,
      "learning_rate": 6.947531271402738e-05,
      "loss": 0.16877172887325287,
      "mean_token_accuracy": 0.9282251596450806,
      "num_tokens": 108108592.0,
      "step": 8764
    },
    {
      "entropy": 1.0329546630382538,
      "epoch": 4.615587151132175,
      "grad_norm": 0.3272295594215393,
      "learning_rate": 6.945573436232924e-05,
      "loss": 0.17032533884048462,
      "mean_token_accuracy": 0.9320805817842484,
      "num_tokens": 108120928.0,
      "step": 8765
    },
    {
      "entropy": 1.024133563041687,
      "epoch": 4.61611374407583,
      "grad_norm": 0.3026040196418762,
      "learning_rate": 6.943615841750783e-05,
      "loss": 0.16396477818489075,
      "mean_token_accuracy": 0.9272926896810532,
      "num_tokens": 108133264.0,
      "step": 8766
    },
    {
      "entropy": 1.0271407514810562,
      "epoch": 4.616640337019484,
      "grad_norm": 0.32383832335472107,
      "learning_rate": 6.94165848807253e-05,
      "loss": 0.1656980812549591,
      "mean_token_accuracy": 0.9300740510225296,
      "num_tokens": 108145600.0,
      "step": 8767
    },
    {
      "entropy": 1.054329827427864,
      "epoch": 4.617166929963139,
      "grad_norm": 0.3429238796234131,
      "learning_rate": 6.939701375314358e-05,
      "loss": 0.17477953433990479,
      "mean_token_accuracy": 0.9288169890642166,
      "num_tokens": 108157936.0,
      "step": 8768
    },
    {
      "entropy": 1.020611047744751,
      "epoch": 4.617693522906793,
      "grad_norm": 0.32155418395996094,
      "learning_rate": 6.937744503592448e-05,
      "loss": 0.16102604568004608,
      "mean_token_accuracy": 0.9318195581436157,
      "num_tokens": 108170272.0,
      "step": 8769
    },
    {
      "entropy": 0.9947672635316849,
      "epoch": 4.6182201158504474,
      "grad_norm": 0.3039378821849823,
      "learning_rate": 6.935787873022971e-05,
      "loss": 0.16173730790615082,
      "mean_token_accuracy": 0.9321751296520233,
      "num_tokens": 108182608.0,
      "step": 8770
    },
    {
      "entropy": 0.993101179599762,
      "epoch": 4.618746708794102,
      "grad_norm": 0.3157470226287842,
      "learning_rate": 6.93383148372208e-05,
      "loss": 0.17102232575416565,
      "mean_token_accuracy": 0.9239119738340378,
      "num_tokens": 108194944.0,
      "step": 8771
    },
    {
      "entropy": 1.037236139178276,
      "epoch": 4.619273301737756,
      "grad_norm": 0.30419865250587463,
      "learning_rate": 6.931875335805915e-05,
      "loss": 0.17111724615097046,
      "mean_token_accuracy": 0.9261699765920639,
      "num_tokens": 108207280.0,
      "step": 8772
    },
    {
      "entropy": 1.0121787637472153,
      "epoch": 4.619799894681411,
      "grad_norm": 0.30403435230255127,
      "learning_rate": 6.929919429390597e-05,
      "loss": 0.1617126315832138,
      "mean_token_accuracy": 0.9305363595485687,
      "num_tokens": 108219616.0,
      "step": 8773
    },
    {
      "entropy": 0.9898070394992828,
      "epoch": 4.620326487625066,
      "grad_norm": 0.3091951608657837,
      "learning_rate": 6.927963764592242e-05,
      "loss": 0.17442327737808228,
      "mean_token_accuracy": 0.9246090650558472,
      "num_tokens": 108231952.0,
      "step": 8774
    },
    {
      "entropy": 1.0206804275512695,
      "epoch": 4.62085308056872,
      "grad_norm": 0.3396073281764984,
      "learning_rate": 6.926008341526947e-05,
      "loss": 0.17265398800373077,
      "mean_token_accuracy": 0.930071085691452,
      "num_tokens": 108244288.0,
      "step": 8775
    },
    {
      "entropy": 1.0402859449386597,
      "epoch": 4.621379673512375,
      "grad_norm": 0.31713059544563293,
      "learning_rate": 6.924053160310791e-05,
      "loss": 0.16856266558170319,
      "mean_token_accuracy": 0.9330925494432449,
      "num_tokens": 108256624.0,
      "step": 8776
    },
    {
      "entropy": 0.9876909554004669,
      "epoch": 4.62190626645603,
      "grad_norm": 0.31122639775276184,
      "learning_rate": 6.922098221059845e-05,
      "loss": 0.16525253653526306,
      "mean_token_accuracy": 0.9351381659507751,
      "num_tokens": 108268960.0,
      "step": 8777
    },
    {
      "entropy": 0.9891609102487564,
      "epoch": 4.622432859399684,
      "grad_norm": 0.31381669640541077,
      "learning_rate": 6.920143523890156e-05,
      "loss": 0.17055228352546692,
      "mean_token_accuracy": 0.9297476559877396,
      "num_tokens": 108281296.0,
      "step": 8778
    },
    {
      "entropy": 0.9974513947963715,
      "epoch": 4.622959452343339,
      "grad_norm": 0.3451111316680908,
      "learning_rate": 6.918189068917769e-05,
      "loss": 0.17237506806850433,
      "mean_token_accuracy": 0.9254035949707031,
      "num_tokens": 108293632.0,
      "step": 8779
    },
    {
      "entropy": 1.0044319182634354,
      "epoch": 4.6234860452869935,
      "grad_norm": 0.29564011096954346,
      "learning_rate": 6.916234856258701e-05,
      "loss": 0.15609613060951233,
      "mean_token_accuracy": 0.9366973489522934,
      "num_tokens": 108305968.0,
      "step": 8780
    },
    {
      "entropy": 0.9985024929046631,
      "epoch": 4.6240126382306475,
      "grad_norm": 0.33756110072135925,
      "learning_rate": 6.914280886028974e-05,
      "loss": 0.17060284316539764,
      "mean_token_accuracy": 0.9281386137008667,
      "num_tokens": 108318304.0,
      "step": 8781
    },
    {
      "entropy": 1.0304796695709229,
      "epoch": 4.624539231174302,
      "grad_norm": 0.31444182991981506,
      "learning_rate": 6.912327158344574e-05,
      "loss": 0.1649145632982254,
      "mean_token_accuracy": 0.9314026683568954,
      "num_tokens": 108330640.0,
      "step": 8782
    },
    {
      "entropy": 1.002663105726242,
      "epoch": 4.625065824117957,
      "grad_norm": 0.3058187961578369,
      "learning_rate": 6.910373673321486e-05,
      "loss": 0.15711860358715057,
      "mean_token_accuracy": 0.9331463426351547,
      "num_tokens": 108342976.0,
      "step": 8783
    },
    {
      "entropy": 0.9933810830116272,
      "epoch": 4.625592417061611,
      "grad_norm": 0.3585160970687866,
      "learning_rate": 6.908420431075678e-05,
      "loss": 0.18280671536922455,
      "mean_token_accuracy": 0.9210616052150726,
      "num_tokens": 108355312.0,
      "step": 8784
    },
    {
      "entropy": 0.9912279844284058,
      "epoch": 4.626119010005266,
      "grad_norm": 0.3315320611000061,
      "learning_rate": 6.906467431723103e-05,
      "loss": 0.17114242911338806,
      "mean_token_accuracy": 0.9239901006221771,
      "num_tokens": 108367648.0,
      "step": 8785
    },
    {
      "entropy": 0.97652168571949,
      "epoch": 4.626645602948921,
      "grad_norm": 0.34236273169517517,
      "learning_rate": 6.904514675379695e-05,
      "loss": 0.17682960629463196,
      "mean_token_accuracy": 0.9264667481184006,
      "num_tokens": 108379984.0,
      "step": 8786
    },
    {
      "entropy": 0.9817781150341034,
      "epoch": 4.627172195892575,
      "grad_norm": 0.32331737875938416,
      "learning_rate": 6.902562162161381e-05,
      "loss": 0.17068400979042053,
      "mean_token_accuracy": 0.9266734570264816,
      "num_tokens": 108392320.0,
      "step": 8787
    },
    {
      "entropy": 1.0087388455867767,
      "epoch": 4.62769878883623,
      "grad_norm": 0.35012075304985046,
      "learning_rate": 6.900609892184071e-05,
      "loss": 0.17089639604091644,
      "mean_token_accuracy": 0.9257703721523285,
      "num_tokens": 108404656.0,
      "step": 8788
    },
    {
      "entropy": 0.946434274315834,
      "epoch": 4.628225381779885,
      "grad_norm": 0.2971382737159729,
      "learning_rate": 6.898657865563657e-05,
      "loss": 0.1607464849948883,
      "mean_token_accuracy": 0.9312983602285385,
      "num_tokens": 108416992.0,
      "step": 8789
    },
    {
      "entropy": 0.9486940801143646,
      "epoch": 4.628751974723539,
      "grad_norm": 0.31900492310523987,
      "learning_rate": 6.896706082416022e-05,
      "loss": 0.16454634070396423,
      "mean_token_accuracy": 0.9287359267473221,
      "num_tokens": 108429328.0,
      "step": 8790
    },
    {
      "entropy": 0.9518442749977112,
      "epoch": 4.6292785676671935,
      "grad_norm": 0.32260754704475403,
      "learning_rate": 6.894754542857031e-05,
      "loss": 0.17235061526298523,
      "mean_token_accuracy": 0.9254087507724762,
      "num_tokens": 108441664.0,
      "step": 8791
    },
    {
      "entropy": 0.981766864657402,
      "epoch": 4.6298051606108475,
      "grad_norm": 0.3477701246738434,
      "learning_rate": 6.892803247002535e-05,
      "loss": 0.18203023076057434,
      "mean_token_accuracy": 0.9227351993322372,
      "num_tokens": 108454000.0,
      "step": 8792
    },
    {
      "entropy": 0.9812089949846268,
      "epoch": 4.630331753554502,
      "grad_norm": 0.34699487686157227,
      "learning_rate": 6.890852194968374e-05,
      "loss": 0.17964249849319458,
      "mean_token_accuracy": 0.918544813990593,
      "num_tokens": 108466336.0,
      "step": 8793
    },
    {
      "entropy": 1.016109973192215,
      "epoch": 4.630858346498157,
      "grad_norm": 0.33892375230789185,
      "learning_rate": 6.888901386870361e-05,
      "loss": 0.1662784069776535,
      "mean_token_accuracy": 0.931085929274559,
      "num_tokens": 108478672.0,
      "step": 8794
    },
    {
      "entropy": 0.9515203684568405,
      "epoch": 4.631384939441811,
      "grad_norm": 0.2908926010131836,
      "learning_rate": 6.886950822824319e-05,
      "loss": 0.1586223542690277,
      "mean_token_accuracy": 0.932471439242363,
      "num_tokens": 108491008.0,
      "step": 8795
    },
    {
      "entropy": 0.9727869927883148,
      "epoch": 4.631911532385466,
      "grad_norm": 0.3081713020801544,
      "learning_rate": 6.885000502946035e-05,
      "loss": 0.1696777045726776,
      "mean_token_accuracy": 0.9311618357896805,
      "num_tokens": 108503344.0,
      "step": 8796
    },
    {
      "entropy": 0.9695296436548233,
      "epoch": 4.632438125329121,
      "grad_norm": 0.3235263228416443,
      "learning_rate": 6.883050427351283e-05,
      "loss": 0.16967125236988068,
      "mean_token_accuracy": 0.9278686344623566,
      "num_tokens": 108515680.0,
      "step": 8797
    },
    {
      "entropy": 0.9924297034740448,
      "epoch": 4.632964718272775,
      "grad_norm": 0.33361580967903137,
      "learning_rate": 6.881100596155833e-05,
      "loss": 0.16959601640701294,
      "mean_token_accuracy": 0.9281372725963593,
      "num_tokens": 108528016.0,
      "step": 8798
    },
    {
      "entropy": 0.9642326831817627,
      "epoch": 4.63349131121643,
      "grad_norm": 0.3119698166847229,
      "learning_rate": 6.879151009475433e-05,
      "loss": 0.18061037361621857,
      "mean_token_accuracy": 0.922494113445282,
      "num_tokens": 108540352.0,
      "step": 8799
    },
    {
      "entropy": 0.9996363669633865,
      "epoch": 4.634017904160085,
      "grad_norm": 0.3007625341415405,
      "learning_rate": 6.87720166742582e-05,
      "loss": 0.1670391857624054,
      "mean_token_accuracy": 0.9293374568223953,
      "num_tokens": 108552688.0,
      "step": 8800
    },
    {
      "entropy": 1.0450481325387955,
      "epoch": 4.634544497103739,
      "grad_norm": 0.3554733693599701,
      "learning_rate": 6.875252570122712e-05,
      "loss": 0.1709325611591339,
      "mean_token_accuracy": 0.9246032536029816,
      "num_tokens": 108565024.0,
      "step": 8801
    },
    {
      "entropy": 1.0150825679302216,
      "epoch": 4.6350710900473935,
      "grad_norm": 0.325372576713562,
      "learning_rate": 6.87330371768182e-05,
      "loss": 0.17957626283168793,
      "mean_token_accuracy": 0.9251673370599747,
      "num_tokens": 108577360.0,
      "step": 8802
    },
    {
      "entropy": 1.0115544945001602,
      "epoch": 4.6355976829910475,
      "grad_norm": 0.328909307718277,
      "learning_rate": 6.871355110218834e-05,
      "loss": 0.17813214659690857,
      "mean_token_accuracy": 0.9213647395372391,
      "num_tokens": 108589696.0,
      "step": 8803
    },
    {
      "entropy": 1.015116199851036,
      "epoch": 4.636124275934702,
      "grad_norm": 0.31800439953804016,
      "learning_rate": 6.869406747849433e-05,
      "loss": 0.17300941050052643,
      "mean_token_accuracy": 0.9271139949560165,
      "num_tokens": 108602032.0,
      "step": 8804
    },
    {
      "entropy": 0.9945334792137146,
      "epoch": 4.636650868878357,
      "grad_norm": 0.31376081705093384,
      "learning_rate": 6.867458630689277e-05,
      "loss": 0.18528945744037628,
      "mean_token_accuracy": 0.920944482088089,
      "num_tokens": 108614368.0,
      "step": 8805
    },
    {
      "entropy": 1.0312229245901108,
      "epoch": 4.637177461822011,
      "grad_norm": 0.30054622888565063,
      "learning_rate": 6.865510758854017e-05,
      "loss": 0.16745856404304504,
      "mean_token_accuracy": 0.9292907565832138,
      "num_tokens": 108626704.0,
      "step": 8806
    },
    {
      "entropy": 1.053056538105011,
      "epoch": 4.637704054765666,
      "grad_norm": 0.357748806476593,
      "learning_rate": 6.863563132459287e-05,
      "loss": 0.18521693348884583,
      "mean_token_accuracy": 0.9175962060689926,
      "num_tokens": 108639040.0,
      "step": 8807
    },
    {
      "entropy": 1.073351889848709,
      "epoch": 4.638230647709321,
      "grad_norm": 0.32469242811203003,
      "learning_rate": 6.861615751620701e-05,
      "loss": 0.16757342219352722,
      "mean_token_accuracy": 0.927339494228363,
      "num_tokens": 108651376.0,
      "step": 8808
    },
    {
      "entropy": 0.9891101121902466,
      "epoch": 4.638757240652975,
      "grad_norm": 0.27564358711242676,
      "learning_rate": 6.859668616453869e-05,
      "loss": 0.1580924689769745,
      "mean_token_accuracy": 0.9324152171611786,
      "num_tokens": 108663712.0,
      "step": 8809
    },
    {
      "entropy": 1.0437711328268051,
      "epoch": 4.63928383359663,
      "grad_norm": 0.3324485421180725,
      "learning_rate": 6.85772172707438e-05,
      "loss": 0.18268510699272156,
      "mean_token_accuracy": 0.9208597838878632,
      "num_tokens": 108676048.0,
      "step": 8810
    },
    {
      "entropy": 1.0445916056632996,
      "epoch": 4.639810426540285,
      "grad_norm": 0.31860390305519104,
      "learning_rate": 6.855775083597808e-05,
      "loss": 0.16359257698059082,
      "mean_token_accuracy": 0.9334599822759628,
      "num_tokens": 108688384.0,
      "step": 8811
    },
    {
      "entropy": 1.0297775864601135,
      "epoch": 4.640337019483939,
      "grad_norm": 0.3371516764163971,
      "learning_rate": 6.853828686139716e-05,
      "loss": 0.18046995997428894,
      "mean_token_accuracy": 0.9293525815010071,
      "num_tokens": 108700720.0,
      "step": 8812
    },
    {
      "entropy": 1.0049115121364594,
      "epoch": 4.6408636124275935,
      "grad_norm": 0.29611310362815857,
      "learning_rate": 6.851882534815649e-05,
      "loss": 0.15877759456634521,
      "mean_token_accuracy": 0.9356518685817719,
      "num_tokens": 108713056.0,
      "step": 8813
    },
    {
      "entropy": 1.0410502701997757,
      "epoch": 4.641390205371248,
      "grad_norm": 0.27554795145988464,
      "learning_rate": 6.849936629741138e-05,
      "loss": 0.15286138653755188,
      "mean_token_accuracy": 0.933782085776329,
      "num_tokens": 108725392.0,
      "step": 8814
    },
    {
      "entropy": 1.0450260192155838,
      "epoch": 4.641916798314902,
      "grad_norm": 0.29855507612228394,
      "learning_rate": 6.847990971031704e-05,
      "loss": 0.1780097931623459,
      "mean_token_accuracy": 0.9241447150707245,
      "num_tokens": 108737728.0,
      "step": 8815
    },
    {
      "entropy": 1.0352587401866913,
      "epoch": 4.642443391258557,
      "grad_norm": 0.30585408210754395,
      "learning_rate": 6.846045558802842e-05,
      "loss": 0.1605367362499237,
      "mean_token_accuracy": 0.9307635575532913,
      "num_tokens": 108750064.0,
      "step": 8816
    },
    {
      "entropy": 0.9806309789419174,
      "epoch": 4.642969984202212,
      "grad_norm": 0.2845015525817871,
      "learning_rate": 6.844100393170047e-05,
      "loss": 0.1599006950855255,
      "mean_token_accuracy": 0.9303335249423981,
      "num_tokens": 108762400.0,
      "step": 8817
    },
    {
      "entropy": 1.057199478149414,
      "epoch": 4.643496577145866,
      "grad_norm": 0.3356047570705414,
      "learning_rate": 6.84215547424879e-05,
      "loss": 0.15979963541030884,
      "mean_token_accuracy": 0.9306587725877762,
      "num_tokens": 108774736.0,
      "step": 8818
    },
    {
      "entropy": 1.0132673531770706,
      "epoch": 4.644023170089521,
      "grad_norm": 0.2916070520877838,
      "learning_rate": 6.840210802154525e-05,
      "loss": 0.16350027918815613,
      "mean_token_accuracy": 0.9252615720033646,
      "num_tokens": 108787072.0,
      "step": 8819
    },
    {
      "entropy": 1.0183844417333603,
      "epoch": 4.644549763033176,
      "grad_norm": 0.2789771854877472,
      "learning_rate": 6.838266377002702e-05,
      "loss": 0.14993609488010406,
      "mean_token_accuracy": 0.9347283989191055,
      "num_tokens": 108799408.0,
      "step": 8820
    },
    {
      "entropy": 0.9974929839372635,
      "epoch": 4.64507635597683,
      "grad_norm": 0.3179318308830261,
      "learning_rate": 6.83632219890875e-05,
      "loss": 0.17396214604377747,
      "mean_token_accuracy": 0.9251264482736588,
      "num_tokens": 108811744.0,
      "step": 8821
    },
    {
      "entropy": 1.0086408704519272,
      "epoch": 4.645602948920485,
      "grad_norm": 0.3315167725086212,
      "learning_rate": 6.834378267988072e-05,
      "loss": 0.1686999499797821,
      "mean_token_accuracy": 0.9279903620481491,
      "num_tokens": 108824080.0,
      "step": 8822
    },
    {
      "entropy": 1.0120555758476257,
      "epoch": 4.6461295418641395,
      "grad_norm": 0.31629952788352966,
      "learning_rate": 6.832434584356084e-05,
      "loss": 0.16480788588523865,
      "mean_token_accuracy": 0.930375799536705,
      "num_tokens": 108836416.0,
      "step": 8823
    },
    {
      "entropy": 1.036504328250885,
      "epoch": 4.6466561348077935,
      "grad_norm": 0.33338403701782227,
      "learning_rate": 6.830491148128166e-05,
      "loss": 0.16853323578834534,
      "mean_token_accuracy": 0.9285017400979996,
      "num_tokens": 108848752.0,
      "step": 8824
    },
    {
      "entropy": 1.0557902306318283,
      "epoch": 4.647182727751448,
      "grad_norm": 0.3519696593284607,
      "learning_rate": 6.828547959419688e-05,
      "loss": 0.18914440274238586,
      "mean_token_accuracy": 0.9214043617248535,
      "num_tokens": 108861088.0,
      "step": 8825
    },
    {
      "entropy": 1.0277077853679657,
      "epoch": 4.647709320695102,
      "grad_norm": 0.3393276631832123,
      "learning_rate": 6.826605018346004e-05,
      "loss": 0.16981624066829681,
      "mean_token_accuracy": 0.9295682162046432,
      "num_tokens": 108873424.0,
      "step": 8826
    },
    {
      "entropy": 0.9834050834178925,
      "epoch": 4.648235913638757,
      "grad_norm": 0.3136768639087677,
      "learning_rate": 6.824662325022451e-05,
      "loss": 0.15382970869541168,
      "mean_token_accuracy": 0.9335749596357346,
      "num_tokens": 108885760.0,
      "step": 8827
    },
    {
      "entropy": 1.0283687263727188,
      "epoch": 4.648762506582412,
      "grad_norm": 0.31119608879089355,
      "learning_rate": 6.822719879564364e-05,
      "loss": 0.14895489811897278,
      "mean_token_accuracy": 0.935601681470871,
      "num_tokens": 108898096.0,
      "step": 8828
    },
    {
      "entropy": 0.9725286513566971,
      "epoch": 4.649289099526066,
      "grad_norm": 0.30130234360694885,
      "learning_rate": 6.820777682087044e-05,
      "loss": 0.1579904854297638,
      "mean_token_accuracy": 0.9328674077987671,
      "num_tokens": 108910432.0,
      "step": 8829
    },
    {
      "entropy": 0.9630727767944336,
      "epoch": 4.649815692469721,
      "grad_norm": 0.2871607840061188,
      "learning_rate": 6.818835732705799e-05,
      "loss": 0.1505204439163208,
      "mean_token_accuracy": 0.938169002532959,
      "num_tokens": 108922768.0,
      "step": 8830
    },
    {
      "entropy": 1.0013564974069595,
      "epoch": 4.650342285413376,
      "grad_norm": 0.29978296160697937,
      "learning_rate": 6.816894031535908e-05,
      "loss": 0.17103968560695648,
      "mean_token_accuracy": 0.9303154200315475,
      "num_tokens": 108935104.0,
      "step": 8831
    },
    {
      "entropy": 1.035158023238182,
      "epoch": 4.65086887835703,
      "grad_norm": 0.3239688575267792,
      "learning_rate": 6.814952578692637e-05,
      "loss": 0.1669498085975647,
      "mean_token_accuracy": 0.9296908974647522,
      "num_tokens": 108947440.0,
      "step": 8832
    },
    {
      "entropy": 0.9917216300964355,
      "epoch": 4.651395471300685,
      "grad_norm": 0.3029177188873291,
      "learning_rate": 6.81301137429124e-05,
      "loss": 0.17024460434913635,
      "mean_token_accuracy": 0.9244478642940521,
      "num_tokens": 108959776.0,
      "step": 8833
    },
    {
      "entropy": 0.9994790703058243,
      "epoch": 4.651922064244339,
      "grad_norm": 0.3118470013141632,
      "learning_rate": 6.811070418446953e-05,
      "loss": 0.16566264629364014,
      "mean_token_accuracy": 0.9256698489189148,
      "num_tokens": 108972112.0,
      "step": 8834
    },
    {
      "entropy": 0.9939791262149811,
      "epoch": 4.6524486571879935,
      "grad_norm": 0.3334689736366272,
      "learning_rate": 6.809129711275e-05,
      "loss": 0.17950989305973053,
      "mean_token_accuracy": 0.9242898970842361,
      "num_tokens": 108984448.0,
      "step": 8835
    },
    {
      "entropy": 0.9304020404815674,
      "epoch": 4.652975250131648,
      "grad_norm": 0.33595287799835205,
      "learning_rate": 6.80718925289059e-05,
      "loss": 0.17988379299640656,
      "mean_token_accuracy": 0.9204196929931641,
      "num_tokens": 108996784.0,
      "step": 8836
    },
    {
      "entropy": 1.0219557583332062,
      "epoch": 4.653501843075302,
      "grad_norm": 0.3548906743526459,
      "learning_rate": 6.805249043408918e-05,
      "loss": 0.1703854650259018,
      "mean_token_accuracy": 0.9284821301698685,
      "num_tokens": 109009120.0,
      "step": 8837
    },
    {
      "entropy": 0.9631175994873047,
      "epoch": 4.654028436018957,
      "grad_norm": 0.3329707980155945,
      "learning_rate": 6.803309082945159e-05,
      "loss": 0.1741081178188324,
      "mean_token_accuracy": 0.9273754209280014,
      "num_tokens": 109021456.0,
      "step": 8838
    },
    {
      "entropy": 1.0109445601701736,
      "epoch": 4.654555028962612,
      "grad_norm": 0.34718072414398193,
      "learning_rate": 6.80136937161448e-05,
      "loss": 0.17463737726211548,
      "mean_token_accuracy": 0.9279660135507584,
      "num_tokens": 109033792.0,
      "step": 8839
    },
    {
      "entropy": 0.97547547519207,
      "epoch": 4.655081621906266,
      "grad_norm": 0.3017706871032715,
      "learning_rate": 6.799429909532032e-05,
      "loss": 0.15317855775356293,
      "mean_token_accuracy": 0.9332242012023926,
      "num_tokens": 109046128.0,
      "step": 8840
    },
    {
      "entropy": 0.9765388518571854,
      "epoch": 4.655608214849921,
      "grad_norm": 0.3442109525203705,
      "learning_rate": 6.797490696812948e-05,
      "loss": 0.17506183683872223,
      "mean_token_accuracy": 0.9238266944885254,
      "num_tokens": 109058464.0,
      "step": 8841
    },
    {
      "entropy": 0.992646723985672,
      "epoch": 4.656134807793576,
      "grad_norm": 0.33706432580947876,
      "learning_rate": 6.795551733572342e-05,
      "loss": 0.1667737364768982,
      "mean_token_accuracy": 0.9309960454702377,
      "num_tokens": 109070800.0,
      "step": 8842
    },
    {
      "entropy": 0.9908954352140427,
      "epoch": 4.65666140073723,
      "grad_norm": 0.31130319833755493,
      "learning_rate": 6.793613019925326e-05,
      "loss": 0.17460176348686218,
      "mean_token_accuracy": 0.9256035834550858,
      "num_tokens": 109083136.0,
      "step": 8843
    },
    {
      "entropy": 1.02932608127594,
      "epoch": 4.657187993680885,
      "grad_norm": 0.3063564598560333,
      "learning_rate": 6.791674555986994e-05,
      "loss": 0.1669730693101883,
      "mean_token_accuracy": 0.9289689511060715,
      "num_tokens": 109095472.0,
      "step": 8844
    },
    {
      "entropy": 1.0227219462394714,
      "epoch": 4.6577145866245395,
      "grad_norm": 0.3334781527519226,
      "learning_rate": 6.78973634187241e-05,
      "loss": 0.16546808183193207,
      "mean_token_accuracy": 0.9333768039941788,
      "num_tokens": 109107808.0,
      "step": 8845
    },
    {
      "entropy": 0.9633027613162994,
      "epoch": 4.6582411795681935,
      "grad_norm": 0.30114901065826416,
      "learning_rate": 6.787798377696641e-05,
      "loss": 0.1666855365037918,
      "mean_token_accuracy": 0.9321528971195221,
      "num_tokens": 109120144.0,
      "step": 8846
    },
    {
      "entropy": 1.0001439154148102,
      "epoch": 4.658767772511848,
      "grad_norm": 0.3661600351333618,
      "learning_rate": 6.785860663574735e-05,
      "loss": 0.16610540449619293,
      "mean_token_accuracy": 0.9269569218158722,
      "num_tokens": 109132480.0,
      "step": 8847
    },
    {
      "entropy": 0.9895338863134384,
      "epoch": 4.659294365455503,
      "grad_norm": 0.3270148038864136,
      "learning_rate": 6.783923199621716e-05,
      "loss": 0.17534351348876953,
      "mean_token_accuracy": 0.9320991188287735,
      "num_tokens": 109144816.0,
      "step": 8848
    },
    {
      "entropy": 0.9660409092903137,
      "epoch": 4.659820958399157,
      "grad_norm": 0.2960972785949707,
      "learning_rate": 6.781985985952604e-05,
      "loss": 0.157662034034729,
      "mean_token_accuracy": 0.9349133521318436,
      "num_tokens": 109157152.0,
      "step": 8849
    },
    {
      "entropy": 0.9822867512702942,
      "epoch": 4.660347551342812,
      "grad_norm": 0.3005823493003845,
      "learning_rate": 6.780049022682396e-05,
      "loss": 0.15429461002349854,
      "mean_token_accuracy": 0.9349739104509354,
      "num_tokens": 109169488.0,
      "step": 8850
    },
    {
      "entropy": 0.9865624308586121,
      "epoch": 4.660874144286467,
      "grad_norm": 0.3327685296535492,
      "learning_rate": 6.778112309926085e-05,
      "loss": 0.1758040487766266,
      "mean_token_accuracy": 0.9246583431959152,
      "num_tokens": 109181824.0,
      "step": 8851
    },
    {
      "entropy": 0.977432519197464,
      "epoch": 4.661400737230121,
      "grad_norm": 0.28886678814888,
      "learning_rate": 6.776175847798641e-05,
      "loss": 0.16094927489757538,
      "mean_token_accuracy": 0.9336694926023483,
      "num_tokens": 109194160.0,
      "step": 8852
    },
    {
      "entropy": 1.0480350255966187,
      "epoch": 4.661927330173776,
      "grad_norm": 0.32053396105766296,
      "learning_rate": 6.774239636415016e-05,
      "loss": 0.170475035905838,
      "mean_token_accuracy": 0.9263121783733368,
      "num_tokens": 109206496.0,
      "step": 8853
    },
    {
      "entropy": 0.9531082808971405,
      "epoch": 4.662453923117431,
      "grad_norm": 0.3278583884239197,
      "learning_rate": 6.772303675890158e-05,
      "loss": 0.17197012901306152,
      "mean_token_accuracy": 0.9243444502353668,
      "num_tokens": 109218832.0,
      "step": 8854
    },
    {
      "entropy": 0.9761087149381638,
      "epoch": 4.662980516061085,
      "grad_norm": 0.3090464174747467,
      "learning_rate": 6.77036796633899e-05,
      "loss": 0.17255406081676483,
      "mean_token_accuracy": 0.9278616309165955,
      "num_tokens": 109231168.0,
      "step": 8855
    },
    {
      "entropy": 1.0074574649333954,
      "epoch": 4.6635071090047395,
      "grad_norm": 0.3336094319820404,
      "learning_rate": 6.768432507876421e-05,
      "loss": 0.17501670122146606,
      "mean_token_accuracy": 0.9241137206554413,
      "num_tokens": 109243504.0,
      "step": 8856
    },
    {
      "entropy": 1.0223061740398407,
      "epoch": 4.6640337019483935,
      "grad_norm": 0.3564497232437134,
      "learning_rate": 6.766497300617347e-05,
      "loss": 0.1735677570104599,
      "mean_token_accuracy": 0.9331323206424713,
      "num_tokens": 109255840.0,
      "step": 8857
    },
    {
      "entropy": 0.9463786333799362,
      "epoch": 4.664560294892048,
      "grad_norm": 0.30366936326026917,
      "learning_rate": 6.764562344676658e-05,
      "loss": 0.1649957001209259,
      "mean_token_accuracy": 0.9283649027347565,
      "num_tokens": 109268176.0,
      "step": 8858
    },
    {
      "entropy": 0.9768461138010025,
      "epoch": 4.665086887835703,
      "grad_norm": 0.30928438901901245,
      "learning_rate": 6.762627640169216e-05,
      "loss": 0.16720689833164215,
      "mean_token_accuracy": 0.9308343380689621,
      "num_tokens": 109280512.0,
      "step": 8859
    },
    {
      "entropy": 1.016520380973816,
      "epoch": 4.665613480779357,
      "grad_norm": 0.32247668504714966,
      "learning_rate": 6.760693187209875e-05,
      "loss": 0.17250953614711761,
      "mean_token_accuracy": 0.9311464577913284,
      "num_tokens": 109292848.0,
      "step": 8860
    },
    {
      "entropy": 0.9843768626451492,
      "epoch": 4.666140073723012,
      "grad_norm": 0.3198558986186981,
      "learning_rate": 6.758758985913468e-05,
      "loss": 0.1559639275074005,
      "mean_token_accuracy": 0.9342668056488037,
      "num_tokens": 109305184.0,
      "step": 8861
    },
    {
      "entropy": 0.9267884641885757,
      "epoch": 4.666666666666667,
      "grad_norm": 0.2985943853855133,
      "learning_rate": 6.756825036394822e-05,
      "loss": 0.16593988239765167,
      "mean_token_accuracy": 0.9271457940340042,
      "num_tokens": 109317520.0,
      "step": 8862
    },
    {
      "entropy": 0.9875479191541672,
      "epoch": 4.667193259610321,
      "grad_norm": 0.3236728310585022,
      "learning_rate": 6.754891338768742e-05,
      "loss": 0.15824377536773682,
      "mean_token_accuracy": 0.9307315498590469,
      "num_tokens": 109329856.0,
      "step": 8863
    },
    {
      "entropy": 0.9694172888994217,
      "epoch": 4.667719852553976,
      "grad_norm": 0.33678045868873596,
      "learning_rate": 6.75295789315002e-05,
      "loss": 0.1783120334148407,
      "mean_token_accuracy": 0.9196342825889587,
      "num_tokens": 109342192.0,
      "step": 8864
    },
    {
      "entropy": 0.9831208437681198,
      "epoch": 4.668246445497631,
      "grad_norm": 0.3103920519351959,
      "learning_rate": 6.751024699653436e-05,
      "loss": 0.17995822429656982,
      "mean_token_accuracy": 0.9253063350915909,
      "num_tokens": 109354528.0,
      "step": 8865
    },
    {
      "entropy": 0.9898233413696289,
      "epoch": 4.668773038441285,
      "grad_norm": 0.34138306975364685,
      "learning_rate": 6.749091758393751e-05,
      "loss": 0.17750176787376404,
      "mean_token_accuracy": 0.9203987568616867,
      "num_tokens": 109366864.0,
      "step": 8866
    },
    {
      "entropy": 1.002127543091774,
      "epoch": 4.6692996313849395,
      "grad_norm": 0.33900532126426697,
      "learning_rate": 6.747159069485711e-05,
      "loss": 0.1801394820213318,
      "mean_token_accuracy": 0.9212726056575775,
      "num_tokens": 109379200.0,
      "step": 8867
    },
    {
      "entropy": 0.9915066063404083,
      "epoch": 4.6698262243285935,
      "grad_norm": 0.3155921697616577,
      "learning_rate": 6.74522663304405e-05,
      "loss": 0.1677819937467575,
      "mean_token_accuracy": 0.9282025843858719,
      "num_tokens": 109391536.0,
      "step": 8868
    },
    {
      "entropy": 0.9988605678081512,
      "epoch": 4.670352817272248,
      "grad_norm": 0.34599870443344116,
      "learning_rate": 6.743294449183484e-05,
      "loss": 0.18143510818481445,
      "mean_token_accuracy": 0.9232761263847351,
      "num_tokens": 109403872.0,
      "step": 8869
    },
    {
      "entropy": 0.9647913128137589,
      "epoch": 4.670879410215903,
      "grad_norm": 0.32590943574905396,
      "learning_rate": 6.741362518018717e-05,
      "loss": 0.17814356088638306,
      "mean_token_accuracy": 0.9267314821481705,
      "num_tokens": 109416208.0,
      "step": 8870
    },
    {
      "entropy": 1.0397229194641113,
      "epoch": 4.671406003159557,
      "grad_norm": 0.37251850962638855,
      "learning_rate": 6.739430839664433e-05,
      "loss": 0.1976044923067093,
      "mean_token_accuracy": 0.9167015850543976,
      "num_tokens": 109428544.0,
      "step": 8871
    },
    {
      "entropy": 0.9867613464593887,
      "epoch": 4.671932596103212,
      "grad_norm": 0.33050283789634705,
      "learning_rate": 6.737499414235311e-05,
      "loss": 0.17148789763450623,
      "mean_token_accuracy": 0.9267898797988892,
      "num_tokens": 109440880.0,
      "step": 8872
    },
    {
      "entropy": 1.0076529681682587,
      "epoch": 4.672459189046867,
      "grad_norm": 0.3122973144054413,
      "learning_rate": 6.735568241846006e-05,
      "loss": 0.17840178310871124,
      "mean_token_accuracy": 0.9209193140268326,
      "num_tokens": 109453216.0,
      "step": 8873
    },
    {
      "entropy": 1.0246639102697372,
      "epoch": 4.672985781990521,
      "grad_norm": 0.36017894744873047,
      "learning_rate": 6.733637322611162e-05,
      "loss": 0.17562445998191833,
      "mean_token_accuracy": 0.9206981956958771,
      "num_tokens": 109465552.0,
      "step": 8874
    },
    {
      "entropy": 1.0051601976156235,
      "epoch": 4.673512374934176,
      "grad_norm": 0.3045935034751892,
      "learning_rate": 6.7317066566454e-05,
      "loss": 0.17342790961265564,
      "mean_token_accuracy": 0.9259002655744553,
      "num_tokens": 109477888.0,
      "step": 8875
    },
    {
      "entropy": 1.0023014098405838,
      "epoch": 4.674038967877831,
      "grad_norm": 0.30328163504600525,
      "learning_rate": 6.729776244063338e-05,
      "loss": 0.17087529599666595,
      "mean_token_accuracy": 0.9310528039932251,
      "num_tokens": 109490224.0,
      "step": 8876
    },
    {
      "entropy": 1.069144368171692,
      "epoch": 4.674565560821485,
      "grad_norm": 0.34419316053390503,
      "learning_rate": 6.727846084979569e-05,
      "loss": 0.18067161738872528,
      "mean_token_accuracy": 0.9256099611520767,
      "num_tokens": 109502560.0,
      "step": 8877
    },
    {
      "entropy": 1.006365492939949,
      "epoch": 4.6750921537651395,
      "grad_norm": 0.31279006600379944,
      "learning_rate": 6.725916179508677e-05,
      "loss": 0.170567125082016,
      "mean_token_accuracy": 0.9316010475158691,
      "num_tokens": 109514896.0,
      "step": 8878
    },
    {
      "entropy": 1.003068968653679,
      "epoch": 4.675618746708794,
      "grad_norm": 0.3209126591682434,
      "learning_rate": 6.723986527765232e-05,
      "loss": 0.1880316585302353,
      "mean_token_accuracy": 0.9221562743186951,
      "num_tokens": 109527232.0,
      "step": 8879
    },
    {
      "entropy": 0.9958730340003967,
      "epoch": 4.676145339652448,
      "grad_norm": 0.34401512145996094,
      "learning_rate": 6.722057129863784e-05,
      "loss": 0.18341436982154846,
      "mean_token_accuracy": 0.9197598397731781,
      "num_tokens": 109539568.0,
      "step": 8880
    },
    {
      "entropy": 1.0369246304035187,
      "epoch": 4.676671932596103,
      "grad_norm": 0.315122127532959,
      "learning_rate": 6.720127985918872e-05,
      "loss": 0.16368912160396576,
      "mean_token_accuracy": 0.93260657787323,
      "num_tokens": 109551904.0,
      "step": 8881
    },
    {
      "entropy": 0.9783090353012085,
      "epoch": 4.677198525539758,
      "grad_norm": 0.2675493061542511,
      "learning_rate": 6.718199096045017e-05,
      "loss": 0.14673762023448944,
      "mean_token_accuracy": 0.9364724457263947,
      "num_tokens": 109564240.0,
      "step": 8882
    },
    {
      "entropy": 0.9681363850831985,
      "epoch": 4.677725118483412,
      "grad_norm": 0.30054959654808044,
      "learning_rate": 6.716270460356724e-05,
      "loss": 0.1793966293334961,
      "mean_token_accuracy": 0.9206178188323975,
      "num_tokens": 109576576.0,
      "step": 8883
    },
    {
      "entropy": 0.9800532460212708,
      "epoch": 4.678251711427067,
      "grad_norm": 0.3233085572719574,
      "learning_rate": 6.714342078968485e-05,
      "loss": 0.17788507044315338,
      "mean_token_accuracy": 0.9231614917516708,
      "num_tokens": 109588912.0,
      "step": 8884
    },
    {
      "entropy": 0.9717035740613937,
      "epoch": 4.678778304370722,
      "grad_norm": 0.30318590998649597,
      "learning_rate": 6.71241395199478e-05,
      "loss": 0.1675342321395874,
      "mean_token_accuracy": 0.9272963255643845,
      "num_tokens": 109601248.0,
      "step": 8885
    },
    {
      "entropy": 0.9713033586740494,
      "epoch": 4.679304897314376,
      "grad_norm": 0.30094268918037415,
      "learning_rate": 6.710486079550067e-05,
      "loss": 0.15500126779079437,
      "mean_token_accuracy": 0.9303532838821411,
      "num_tokens": 109613584.0,
      "step": 8886
    },
    {
      "entropy": 0.9672758132219315,
      "epoch": 4.679831490258031,
      "grad_norm": 0.35199296474456787,
      "learning_rate": 6.708558461748798e-05,
      "loss": 0.18320858478546143,
      "mean_token_accuracy": 0.9226799160242081,
      "num_tokens": 109625920.0,
      "step": 8887
    },
    {
      "entropy": 0.9596314132213593,
      "epoch": 4.680358083201686,
      "grad_norm": 0.3195297420024872,
      "learning_rate": 6.706631098705398e-05,
      "loss": 0.17051655054092407,
      "mean_token_accuracy": 0.9298547506332397,
      "num_tokens": 109638256.0,
      "step": 8888
    },
    {
      "entropy": 0.9257749915122986,
      "epoch": 4.6808846761453395,
      "grad_norm": 0.3066015839576721,
      "learning_rate": 6.704703990534286e-05,
      "loss": 0.1655864417552948,
      "mean_token_accuracy": 0.9320216923952103,
      "num_tokens": 109650592.0,
      "step": 8889
    },
    {
      "entropy": 0.9602459818124771,
      "epoch": 4.681411269088994,
      "grad_norm": 0.30135196447372437,
      "learning_rate": 6.702777137349863e-05,
      "loss": 0.15446868538856506,
      "mean_token_accuracy": 0.934455156326294,
      "num_tokens": 109662928.0,
      "step": 8890
    },
    {
      "entropy": 0.9682223051786423,
      "epoch": 4.681937862032648,
      "grad_norm": 0.3189322352409363,
      "learning_rate": 6.700850539266517e-05,
      "loss": 0.16403675079345703,
      "mean_token_accuracy": 0.9304446130990982,
      "num_tokens": 109675264.0,
      "step": 8891
    },
    {
      "entropy": 0.9647337943315506,
      "epoch": 4.682464454976303,
      "grad_norm": 0.36465713381767273,
      "learning_rate": 6.698924196398611e-05,
      "loss": 0.1772146224975586,
      "mean_token_accuracy": 0.9272486418485641,
      "num_tokens": 109687600.0,
      "step": 8892
    },
    {
      "entropy": 0.9535428136587143,
      "epoch": 4.682991047919958,
      "grad_norm": 0.34932729601860046,
      "learning_rate": 6.696998108860515e-05,
      "loss": 0.18551796674728394,
      "mean_token_accuracy": 0.9200732558965683,
      "num_tokens": 109699936.0,
      "step": 8893
    },
    {
      "entropy": 0.9440125077962875,
      "epoch": 4.683517640863612,
      "grad_norm": 0.31598344445228577,
      "learning_rate": 6.695072276766558e-05,
      "loss": 0.1678483486175537,
      "mean_token_accuracy": 0.9258812069892883,
      "num_tokens": 109712272.0,
      "step": 8894
    },
    {
      "entropy": 0.926901325583458,
      "epoch": 4.684044233807267,
      "grad_norm": 0.29600900411605835,
      "learning_rate": 6.693146700231071e-05,
      "loss": 0.15550197660923004,
      "mean_token_accuracy": 0.9305247664451599,
      "num_tokens": 109724608.0,
      "step": 8895
    },
    {
      "entropy": 0.9396387040615082,
      "epoch": 4.684570826750922,
      "grad_norm": 0.2947285771369934,
      "learning_rate": 6.69122137936836e-05,
      "loss": 0.1678035855293274,
      "mean_token_accuracy": 0.9310602843761444,
      "num_tokens": 109736944.0,
      "step": 8896
    },
    {
      "entropy": 0.9233750700950623,
      "epoch": 4.685097419694576,
      "grad_norm": 0.307919442653656,
      "learning_rate": 6.689296314292723e-05,
      "loss": 0.17454837262630463,
      "mean_token_accuracy": 0.9254536330699921,
      "num_tokens": 109749280.0,
      "step": 8897
    },
    {
      "entropy": 0.9896024316549301,
      "epoch": 4.685624012638231,
      "grad_norm": 0.33286166191101074,
      "learning_rate": 6.687371505118438e-05,
      "loss": 0.17996010184288025,
      "mean_token_accuracy": 0.9221208244562149,
      "num_tokens": 109761616.0,
      "step": 8898
    },
    {
      "entropy": 0.936344787478447,
      "epoch": 4.686150605581886,
      "grad_norm": 0.31250613927841187,
      "learning_rate": 6.685446951959769e-05,
      "loss": 0.175594300031662,
      "mean_token_accuracy": 0.9231078326702118,
      "num_tokens": 109773952.0,
      "step": 8899
    },
    {
      "entropy": 0.9215246587991714,
      "epoch": 4.6866771985255395,
      "grad_norm": 0.31440743803977966,
      "learning_rate": 6.683522654930969e-05,
      "loss": 0.16845035552978516,
      "mean_token_accuracy": 0.9260395914316177,
      "num_tokens": 109786288.0,
      "step": 8900
    },
    {
      "entropy": 0.950299933552742,
      "epoch": 4.687203791469194,
      "grad_norm": 0.3220463991165161,
      "learning_rate": 6.681598614146272e-05,
      "loss": 0.17246970534324646,
      "mean_token_accuracy": 0.9272022545337677,
      "num_tokens": 109798624.0,
      "step": 8901
    },
    {
      "entropy": 0.9349493682384491,
      "epoch": 4.687730384412848,
      "grad_norm": 0.30960309505462646,
      "learning_rate": 6.679674829719895e-05,
      "loss": 0.16517576575279236,
      "mean_token_accuracy": 0.926237940788269,
      "num_tokens": 109810960.0,
      "step": 8902
    },
    {
      "entropy": 0.9120568186044693,
      "epoch": 4.688256977356503,
      "grad_norm": 0.32024478912353516,
      "learning_rate": 6.677751301766043e-05,
      "loss": 0.17596381902694702,
      "mean_token_accuracy": 0.9267518371343613,
      "num_tokens": 109823296.0,
      "step": 8903
    },
    {
      "entropy": 0.9350269734859467,
      "epoch": 4.688783570300158,
      "grad_norm": 0.3189983665943146,
      "learning_rate": 6.675828030398907e-05,
      "loss": 0.1608733981847763,
      "mean_token_accuracy": 0.9338311553001404,
      "num_tokens": 109835632.0,
      "step": 8904
    },
    {
      "entropy": 0.931309312582016,
      "epoch": 4.689310163243812,
      "grad_norm": 0.3168869614601135,
      "learning_rate": 6.673905015732656e-05,
      "loss": 0.17592725157737732,
      "mean_token_accuracy": 0.9236612617969513,
      "num_tokens": 109847968.0,
      "step": 8905
    },
    {
      "entropy": 0.9076910465955734,
      "epoch": 4.689836756187467,
      "grad_norm": 0.31664350628852844,
      "learning_rate": 6.671982257881444e-05,
      "loss": 0.18773403763771057,
      "mean_token_accuracy": 0.9247518628835678,
      "num_tokens": 109860304.0,
      "step": 8906
    },
    {
      "entropy": 0.8888674080371857,
      "epoch": 4.690363349131122,
      "grad_norm": 0.28577545285224915,
      "learning_rate": 6.670059756959424e-05,
      "loss": 0.15422120690345764,
      "mean_token_accuracy": 0.9305146485567093,
      "num_tokens": 109872640.0,
      "step": 8907
    },
    {
      "entropy": 0.9113831967115402,
      "epoch": 4.690889942074776,
      "grad_norm": 0.30585935711860657,
      "learning_rate": 6.668137513080722e-05,
      "loss": 0.17351096868515015,
      "mean_token_accuracy": 0.9245708882808685,
      "num_tokens": 109884976.0,
      "step": 8908
    },
    {
      "entropy": 0.9621010273694992,
      "epoch": 4.691416535018431,
      "grad_norm": 0.35061115026474,
      "learning_rate": 6.666215526359443e-05,
      "loss": 0.19477160274982452,
      "mean_token_accuracy": 0.9192596524953842,
      "num_tokens": 109897312.0,
      "step": 8909
    },
    {
      "entropy": 0.8926963806152344,
      "epoch": 4.691943127962086,
      "grad_norm": 0.27797988057136536,
      "learning_rate": 6.664293796909693e-05,
      "loss": 0.16123740375041962,
      "mean_token_accuracy": 0.9285058081150055,
      "num_tokens": 109909648.0,
      "step": 8910
    },
    {
      "entropy": 0.9432859420776367,
      "epoch": 4.6924697209057395,
      "grad_norm": 0.3051757514476776,
      "learning_rate": 6.662372324845547e-05,
      "loss": 0.16980287432670593,
      "mean_token_accuracy": 0.927488312125206,
      "num_tokens": 109921984.0,
      "step": 8911
    },
    {
      "entropy": 0.8997198194265366,
      "epoch": 4.692996313849394,
      "grad_norm": 0.3046535551548004,
      "learning_rate": 6.660451110281076e-05,
      "loss": 0.1756792515516281,
      "mean_token_accuracy": 0.9270216971635818,
      "num_tokens": 109934320.0,
      "step": 8912
    },
    {
      "entropy": 0.9536482244729996,
      "epoch": 4.693522906793049,
      "grad_norm": 0.3259578347206116,
      "learning_rate": 6.65853015333033e-05,
      "loss": 0.17409871518611908,
      "mean_token_accuracy": 0.9254413843154907,
      "num_tokens": 109946656.0,
      "step": 8913
    },
    {
      "entropy": 0.9118822365999222,
      "epoch": 4.694049499736703,
      "grad_norm": 0.28276440501213074,
      "learning_rate": 6.656609454107344e-05,
      "loss": 0.15779581665992737,
      "mean_token_accuracy": 0.9322579503059387,
      "num_tokens": 109958992.0,
      "step": 8914
    },
    {
      "entropy": 0.9214333295822144,
      "epoch": 4.694576092680358,
      "grad_norm": 0.3079218864440918,
      "learning_rate": 6.65468901272614e-05,
      "loss": 0.16822515428066254,
      "mean_token_accuracy": 0.9315666556358337,
      "num_tokens": 109971328.0,
      "step": 8915
    },
    {
      "entropy": 0.869530126452446,
      "epoch": 4.695102685624013,
      "grad_norm": 0.2858998477458954,
      "learning_rate": 6.652768829300723e-05,
      "loss": 0.14545699954032898,
      "mean_token_accuracy": 0.939204603433609,
      "num_tokens": 109983664.0,
      "step": 8916
    },
    {
      "entropy": 0.8881965726613998,
      "epoch": 4.695629278567667,
      "grad_norm": 0.3210907280445099,
      "learning_rate": 6.650848903945083e-05,
      "loss": 0.17202621698379517,
      "mean_token_accuracy": 0.9288241863250732,
      "num_tokens": 109996000.0,
      "step": 8917
    },
    {
      "entropy": 0.8845930248498917,
      "epoch": 4.696155871511322,
      "grad_norm": 0.29390251636505127,
      "learning_rate": 6.648929236773196e-05,
      "loss": 0.16244374215602875,
      "mean_token_accuracy": 0.9327209442853928,
      "num_tokens": 110008336.0,
      "step": 8918
    },
    {
      "entropy": 0.8755250722169876,
      "epoch": 4.696682464454977,
      "grad_norm": 0.315922349691391,
      "learning_rate": 6.647009827899016e-05,
      "loss": 0.1748252511024475,
      "mean_token_accuracy": 0.9285299479961395,
      "num_tokens": 110020672.0,
      "step": 8919
    },
    {
      "entropy": 0.8949416726827621,
      "epoch": 4.697209057398631,
      "grad_norm": 0.3261052668094635,
      "learning_rate": 6.645090677436496e-05,
      "loss": 0.18132631480693817,
      "mean_token_accuracy": 0.9244782775640488,
      "num_tokens": 110033008.0,
      "step": 8920
    },
    {
      "entropy": 0.9308057427406311,
      "epoch": 4.697735650342286,
      "grad_norm": 0.3599795401096344,
      "learning_rate": 6.643171785499555e-05,
      "loss": 0.1775147169828415,
      "mean_token_accuracy": 0.924231082201004,
      "num_tokens": 110045344.0,
      "step": 8921
    },
    {
      "entropy": 0.8848343342542648,
      "epoch": 4.69826224328594,
      "grad_norm": 0.31419074535369873,
      "learning_rate": 6.641253152202115e-05,
      "loss": 0.1590585708618164,
      "mean_token_accuracy": 0.930869922041893,
      "num_tokens": 110057680.0,
      "step": 8922
    },
    {
      "entropy": 0.8654780089855194,
      "epoch": 4.698788836229594,
      "grad_norm": 0.32970088720321655,
      "learning_rate": 6.639334777658075e-05,
      "loss": 0.16258922219276428,
      "mean_token_accuracy": 0.933304637670517,
      "num_tokens": 110070016.0,
      "step": 8923
    },
    {
      "entropy": 0.8566028475761414,
      "epoch": 4.699315429173249,
      "grad_norm": 0.2976810932159424,
      "learning_rate": 6.637416661981308e-05,
      "loss": 0.1670779585838318,
      "mean_token_accuracy": 0.9328662753105164,
      "num_tokens": 110082352.0,
      "step": 8924
    },
    {
      "entropy": 0.8485848903656006,
      "epoch": 4.699842022116903,
      "grad_norm": 0.3209967017173767,
      "learning_rate": 6.635498805285687e-05,
      "loss": 0.1709061563014984,
      "mean_token_accuracy": 0.9276284873485565,
      "num_tokens": 110094688.0,
      "step": 8925
    },
    {
      "entropy": 0.8869950175285339,
      "epoch": 4.700368615060558,
      "grad_norm": 0.31358617544174194,
      "learning_rate": 6.633581207685065e-05,
      "loss": 0.16815422475337982,
      "mean_token_accuracy": 0.9248718768358231,
      "num_tokens": 110107024.0,
      "step": 8926
    },
    {
      "entropy": 0.8591439723968506,
      "epoch": 4.700895208004213,
      "grad_norm": 0.3258786201477051,
      "learning_rate": 6.631663869293278e-05,
      "loss": 0.15920323133468628,
      "mean_token_accuracy": 0.933823436498642,
      "num_tokens": 110119360.0,
      "step": 8927
    },
    {
      "entropy": 0.894145056605339,
      "epoch": 4.701421800947867,
      "grad_norm": 0.33094823360443115,
      "learning_rate": 6.62974679022414e-05,
      "loss": 0.17090025544166565,
      "mean_token_accuracy": 0.9283482134342194,
      "num_tokens": 110131696.0,
      "step": 8928
    },
    {
      "entropy": 0.9194226413965225,
      "epoch": 4.701948393891522,
      "grad_norm": 0.3311615586280823,
      "learning_rate": 6.627829970591469e-05,
      "loss": 0.1723853498697281,
      "mean_token_accuracy": 0.9209810793399811,
      "num_tokens": 110144032.0,
      "step": 8929
    },
    {
      "entropy": 0.8927403390407562,
      "epoch": 4.702474986835177,
      "grad_norm": 0.32033735513687134,
      "learning_rate": 6.62591341050905e-05,
      "loss": 0.17580394446849823,
      "mean_token_accuracy": 0.9258732199668884,
      "num_tokens": 110156368.0,
      "step": 8930
    },
    {
      "entropy": 0.8567298799753189,
      "epoch": 4.703001579778831,
      "grad_norm": 0.32278972864151,
      "learning_rate": 6.623997110090656e-05,
      "loss": 0.1617378145456314,
      "mean_token_accuracy": 0.9279116094112396,
      "num_tokens": 110168704.0,
      "step": 8931
    },
    {
      "entropy": 0.9135367274284363,
      "epoch": 4.703528172722486,
      "grad_norm": 0.3466006815433502,
      "learning_rate": 6.62208106945005e-05,
      "loss": 0.17141148447990417,
      "mean_token_accuracy": 0.9257617443799973,
      "num_tokens": 110181040.0,
      "step": 8932
    },
    {
      "entropy": 0.9058708250522614,
      "epoch": 4.7040547656661404,
      "grad_norm": 0.3600664436817169,
      "learning_rate": 6.620165288700974e-05,
      "loss": 0.1806974560022354,
      "mean_token_accuracy": 0.9254732429981232,
      "num_tokens": 110193376.0,
      "step": 8933
    },
    {
      "entropy": 0.8960120677947998,
      "epoch": 4.704581358609794,
      "grad_norm": 0.31437167525291443,
      "learning_rate": 6.618249767957158e-05,
      "loss": 0.16793763637542725,
      "mean_token_accuracy": 0.9310059249401093,
      "num_tokens": 110205712.0,
      "step": 8934
    },
    {
      "entropy": 0.9051641076803207,
      "epoch": 4.705107951553449,
      "grad_norm": 0.2790035605430603,
      "learning_rate": 6.616334507332312e-05,
      "loss": 0.14643120765686035,
      "mean_token_accuracy": 0.9365227818489075,
      "num_tokens": 110218048.0,
      "step": 8935
    },
    {
      "entropy": 0.8888637870550156,
      "epoch": 4.705634544497103,
      "grad_norm": 0.29089266061782837,
      "learning_rate": 6.614419506940137e-05,
      "loss": 0.1581224799156189,
      "mean_token_accuracy": 0.9336483627557755,
      "num_tokens": 110230384.0,
      "step": 8936
    },
    {
      "entropy": 0.8958435654640198,
      "epoch": 4.706161137440758,
      "grad_norm": 0.3086511790752411,
      "learning_rate": 6.612504766894316e-05,
      "loss": 0.15827912092208862,
      "mean_token_accuracy": 0.9326409995555878,
      "num_tokens": 110242720.0,
      "step": 8937
    },
    {
      "entropy": 0.9101904034614563,
      "epoch": 4.706687730384413,
      "grad_norm": 0.3037155568599701,
      "learning_rate": 6.610590287308514e-05,
      "loss": 0.1719973385334015,
      "mean_token_accuracy": 0.9325263053178787,
      "num_tokens": 110255056.0,
      "step": 8938
    },
    {
      "entropy": 0.8873958140611649,
      "epoch": 4.707214323328067,
      "grad_norm": 0.32435721158981323,
      "learning_rate": 6.608676068296382e-05,
      "loss": 0.17124015092849731,
      "mean_token_accuracy": 0.9277086555957794,
      "num_tokens": 110267392.0,
      "step": 8939
    },
    {
      "entropy": 0.9239481985569,
      "epoch": 4.707740916271722,
      "grad_norm": 0.33015745878219604,
      "learning_rate": 6.606762109971561e-05,
      "loss": 0.1701054871082306,
      "mean_token_accuracy": 0.9306849241256714,
      "num_tokens": 110279728.0,
      "step": 8940
    },
    {
      "entropy": 0.9267433881759644,
      "epoch": 4.708267509215377,
      "grad_norm": 0.34128713607788086,
      "learning_rate": 6.604848412447663e-05,
      "loss": 0.18052354454994202,
      "mean_token_accuracy": 0.9229462891817093,
      "num_tokens": 110292064.0,
      "step": 8941
    },
    {
      "entropy": 0.914429098367691,
      "epoch": 4.708794102159031,
      "grad_norm": 0.35832804441452026,
      "learning_rate": 6.602934975838301e-05,
      "loss": 0.18640364706516266,
      "mean_token_accuracy": 0.9196774363517761,
      "num_tokens": 110304400.0,
      "step": 8942
    },
    {
      "entropy": 0.9277569949626923,
      "epoch": 4.709320695102686,
      "grad_norm": 0.33268260955810547,
      "learning_rate": 6.601021800257061e-05,
      "loss": 0.1887909471988678,
      "mean_token_accuracy": 0.9203357994556427,
      "num_tokens": 110316736.0,
      "step": 8943
    },
    {
      "entropy": 0.9191337078809738,
      "epoch": 4.7098472880463405,
      "grad_norm": 0.31906694173812866,
      "learning_rate": 6.599108885817517e-05,
      "loss": 0.18403413891792297,
      "mean_token_accuracy": 0.9191030859947205,
      "num_tokens": 110329072.0,
      "step": 8944
    },
    {
      "entropy": 0.9383554905653,
      "epoch": 4.710373880989994,
      "grad_norm": 0.34196263551712036,
      "learning_rate": 6.597196232633229e-05,
      "loss": 0.17426656186580658,
      "mean_token_accuracy": 0.9295345395803452,
      "num_tokens": 110341408.0,
      "step": 8945
    },
    {
      "entropy": 0.9343491792678833,
      "epoch": 4.710900473933649,
      "grad_norm": 0.31792646646499634,
      "learning_rate": 6.595283840817739e-05,
      "loss": 0.16848713159561157,
      "mean_token_accuracy": 0.9252421706914902,
      "num_tokens": 110353744.0,
      "step": 8946
    },
    {
      "entropy": 0.9486443549394608,
      "epoch": 4.711427066877304,
      "grad_norm": 0.32646486163139343,
      "learning_rate": 6.593371710484574e-05,
      "loss": 0.1752864569425583,
      "mean_token_accuracy": 0.9279747307300568,
      "num_tokens": 110366080.0,
      "step": 8947
    },
    {
      "entropy": 0.9161777049303055,
      "epoch": 4.711953659820958,
      "grad_norm": 0.3029508590698242,
      "learning_rate": 6.591459841747247e-05,
      "loss": 0.15937493741512299,
      "mean_token_accuracy": 0.9336031377315521,
      "num_tokens": 110378416.0,
      "step": 8948
    },
    {
      "entropy": 0.9352288544178009,
      "epoch": 4.712480252764613,
      "grad_norm": 0.34143126010894775,
      "learning_rate": 6.58954823471925e-05,
      "loss": 0.174641415476799,
      "mean_token_accuracy": 0.9258183240890503,
      "num_tokens": 110390752.0,
      "step": 8949
    },
    {
      "entropy": 0.9488516598939896,
      "epoch": 4.713006845708268,
      "grad_norm": 0.32799118757247925,
      "learning_rate": 6.587636889514073e-05,
      "loss": 0.1729208379983902,
      "mean_token_accuracy": 0.9303274303674698,
      "num_tokens": 110403088.0,
      "step": 8950
    },
    {
      "entropy": 0.9332128018140793,
      "epoch": 4.713533438651922,
      "grad_norm": 0.30908095836639404,
      "learning_rate": 6.585725806245178e-05,
      "loss": 0.1610102653503418,
      "mean_token_accuracy": 0.9325119704008102,
      "num_tokens": 110415424.0,
      "step": 8951
    },
    {
      "entropy": 0.9808033108711243,
      "epoch": 4.714060031595577,
      "grad_norm": 0.3266369104385376,
      "learning_rate": 6.58381498502601e-05,
      "loss": 0.1728314310312271,
      "mean_token_accuracy": 0.9239996820688248,
      "num_tokens": 110427760.0,
      "step": 8952
    },
    {
      "entropy": 0.9401232153177261,
      "epoch": 4.714586624539232,
      "grad_norm": 0.34261366724967957,
      "learning_rate": 6.581904425970011e-05,
      "loss": 0.187231183052063,
      "mean_token_accuracy": 0.9219857454299927,
      "num_tokens": 110440096.0,
      "step": 8953
    },
    {
      "entropy": 0.9308113306760788,
      "epoch": 4.715113217482886,
      "grad_norm": 0.3259502947330475,
      "learning_rate": 6.579994129190594e-05,
      "loss": 0.17536687850952148,
      "mean_token_accuracy": 0.9281197637319565,
      "num_tokens": 110452432.0,
      "step": 8954
    },
    {
      "entropy": 0.9176206141710281,
      "epoch": 4.7156398104265405,
      "grad_norm": 0.3092325031757355,
      "learning_rate": 6.578084094801165e-05,
      "loss": 0.16749367117881775,
      "mean_token_accuracy": 0.9292961359024048,
      "num_tokens": 110464768.0,
      "step": 8955
    },
    {
      "entropy": 0.932731881737709,
      "epoch": 4.716166403370195,
      "grad_norm": 0.32101449370384216,
      "learning_rate": 6.576174322915104e-05,
      "loss": 0.1671086847782135,
      "mean_token_accuracy": 0.9271100908517838,
      "num_tokens": 110477104.0,
      "step": 8956
    },
    {
      "entropy": 0.9705034643411636,
      "epoch": 4.716692996313849,
      "grad_norm": 0.34419867396354675,
      "learning_rate": 6.574264813645794e-05,
      "loss": 0.18329662084579468,
      "mean_token_accuracy": 0.9214195013046265,
      "num_tokens": 110489440.0,
      "step": 8957
    },
    {
      "entropy": 0.9514918476343155,
      "epoch": 4.717219589257504,
      "grad_norm": 0.31465935707092285,
      "learning_rate": 6.572355567106588e-05,
      "loss": 0.1708117127418518,
      "mean_token_accuracy": 0.9297067523002625,
      "num_tokens": 110501776.0,
      "step": 8958
    },
    {
      "entropy": 0.9665895402431488,
      "epoch": 4.717746182201158,
      "grad_norm": 0.32876408100128174,
      "learning_rate": 6.570446583410823e-05,
      "loss": 0.17325656116008759,
      "mean_token_accuracy": 0.9287900924682617,
      "num_tokens": 110514112.0,
      "step": 8959
    },
    {
      "entropy": 0.9653350561857224,
      "epoch": 4.718272775144813,
      "grad_norm": 0.2947167754173279,
      "learning_rate": 6.568537862671828e-05,
      "loss": 0.14871157705783844,
      "mean_token_accuracy": 0.9373853355646133,
      "num_tokens": 110526448.0,
      "step": 8960
    },
    {
      "entropy": 0.9705002754926682,
      "epoch": 4.718799368088468,
      "grad_norm": 0.33642247319221497,
      "learning_rate": 6.566629405002914e-05,
      "loss": 0.17432236671447754,
      "mean_token_accuracy": 0.9288565516471863,
      "num_tokens": 110538784.0,
      "step": 8961
    },
    {
      "entropy": 0.9390337914228439,
      "epoch": 4.719325961032122,
      "grad_norm": 0.27955207228660583,
      "learning_rate": 6.564721210517369e-05,
      "loss": 0.16462326049804688,
      "mean_token_accuracy": 0.927482396364212,
      "num_tokens": 110551120.0,
      "step": 8962
    },
    {
      "entropy": 0.9465681612491608,
      "epoch": 4.719852553975777,
      "grad_norm": 0.33623427152633667,
      "learning_rate": 6.562813279328477e-05,
      "loss": 0.173857644200325,
      "mean_token_accuracy": 0.9263188689947128,
      "num_tokens": 110563456.0,
      "step": 8963
    },
    {
      "entropy": 0.9599556177854538,
      "epoch": 4.720379146919432,
      "grad_norm": 0.335986852645874,
      "learning_rate": 6.5609056115495e-05,
      "loss": 0.1687793731689453,
      "mean_token_accuracy": 0.9257626086473465,
      "num_tokens": 110575792.0,
      "step": 8964
    },
    {
      "entropy": 0.9443381130695343,
      "epoch": 4.720905739863086,
      "grad_norm": 0.3181096613407135,
      "learning_rate": 6.55899820729368e-05,
      "loss": 0.18053269386291504,
      "mean_token_accuracy": 0.9224587678909302,
      "num_tokens": 110588128.0,
      "step": 8965
    },
    {
      "entropy": 0.9383203834295273,
      "epoch": 4.7214323328067405,
      "grad_norm": 0.33349719643592834,
      "learning_rate": 6.557091066674255e-05,
      "loss": 0.18003800511360168,
      "mean_token_accuracy": 0.922263890504837,
      "num_tokens": 110600464.0,
      "step": 8966
    },
    {
      "entropy": 0.932285413146019,
      "epoch": 4.721958925750395,
      "grad_norm": 0.32349494099617004,
      "learning_rate": 6.555184189804437e-05,
      "loss": 0.15404555201530457,
      "mean_token_accuracy": 0.9335178434848785,
      "num_tokens": 110612800.0,
      "step": 8967
    },
    {
      "entropy": 0.9489399194717407,
      "epoch": 4.722485518694049,
      "grad_norm": 0.32240861654281616,
      "learning_rate": 6.553277576797427e-05,
      "loss": 0.1671348512172699,
      "mean_token_accuracy": 0.9278327971696854,
      "num_tokens": 110625136.0,
      "step": 8968
    },
    {
      "entropy": 0.9735501110553741,
      "epoch": 4.723012111637704,
      "grad_norm": 0.3254861831665039,
      "learning_rate": 6.551371227766412e-05,
      "loss": 0.16706396639347076,
      "mean_token_accuracy": 0.929393544793129,
      "num_tokens": 110637472.0,
      "step": 8969
    },
    {
      "entropy": 0.934670940041542,
      "epoch": 4.723538704581358,
      "grad_norm": 0.314018577337265,
      "learning_rate": 6.549465142824554e-05,
      "loss": 0.16582143306732178,
      "mean_token_accuracy": 0.9346720725297928,
      "num_tokens": 110649808.0,
      "step": 8970
    },
    {
      "entropy": 0.9515503495931625,
      "epoch": 4.724065297525013,
      "grad_norm": 0.35136815905570984,
      "learning_rate": 6.547559322085013e-05,
      "loss": 0.1879626214504242,
      "mean_token_accuracy": 0.9235444515943527,
      "num_tokens": 110662144.0,
      "step": 8971
    },
    {
      "entropy": 0.9929835647344589,
      "epoch": 4.724591890468668,
      "grad_norm": 0.3528837263584137,
      "learning_rate": 6.545653765660929e-05,
      "loss": 0.17926721274852753,
      "mean_token_accuracy": 0.9293420314788818,
      "num_tokens": 110674480.0,
      "step": 8972
    },
    {
      "entropy": 0.9760604649782181,
      "epoch": 4.725118483412322,
      "grad_norm": 0.32862380146980286,
      "learning_rate": 6.543748473665417e-05,
      "loss": 0.16842186450958252,
      "mean_token_accuracy": 0.9320703595876694,
      "num_tokens": 110686816.0,
      "step": 8973
    },
    {
      "entropy": 0.9648343026638031,
      "epoch": 4.725645076355977,
      "grad_norm": 0.33077824115753174,
      "learning_rate": 6.541843446211584e-05,
      "loss": 0.17277027666568756,
      "mean_token_accuracy": 0.9269263595342636,
      "num_tokens": 110699152.0,
      "step": 8974
    },
    {
      "entropy": 0.9570888429880142,
      "epoch": 4.726171669299632,
      "grad_norm": 0.30975818634033203,
      "learning_rate": 6.539938683412519e-05,
      "loss": 0.17477139830589294,
      "mean_token_accuracy": 0.9248654693365097,
      "num_tokens": 110711488.0,
      "step": 8975
    },
    {
      "entropy": 0.972335085272789,
      "epoch": 4.726698262243286,
      "grad_norm": 0.3457125425338745,
      "learning_rate": 6.538034185381301e-05,
      "loss": 0.17971360683441162,
      "mean_token_accuracy": 0.9200064539909363,
      "num_tokens": 110723824.0,
      "step": 8976
    },
    {
      "entropy": 0.943258211016655,
      "epoch": 4.7272248551869405,
      "grad_norm": 0.3156683146953583,
      "learning_rate": 6.536129952230985e-05,
      "loss": 0.1600583791732788,
      "mean_token_accuracy": 0.9336526691913605,
      "num_tokens": 110736160.0,
      "step": 8977
    },
    {
      "entropy": 0.9720461964607239,
      "epoch": 4.727751448130595,
      "grad_norm": 0.35263144969940186,
      "learning_rate": 6.534225984074618e-05,
      "loss": 0.18373900651931763,
      "mean_token_accuracy": 0.9243020862340927,
      "num_tokens": 110748496.0,
      "step": 8978
    },
    {
      "entropy": 0.955377921462059,
      "epoch": 4.728278041074249,
      "grad_norm": 0.29776355624198914,
      "learning_rate": 6.532322281025226e-05,
      "loss": 0.16605469584465027,
      "mean_token_accuracy": 0.9235288351774216,
      "num_tokens": 110760832.0,
      "step": 8979
    },
    {
      "entropy": 0.9818824827671051,
      "epoch": 4.728804634017904,
      "grad_norm": 0.3629790246486664,
      "learning_rate": 6.530418843195821e-05,
      "loss": 0.1792069971561432,
      "mean_token_accuracy": 0.9269376397132874,
      "num_tokens": 110773168.0,
      "step": 8980
    },
    {
      "entropy": 0.940108060836792,
      "epoch": 4.729331226961559,
      "grad_norm": 0.29533788561820984,
      "learning_rate": 6.5285156706994e-05,
      "loss": 0.1598343700170517,
      "mean_token_accuracy": 0.9303691387176514,
      "num_tokens": 110785504.0,
      "step": 8981
    },
    {
      "entropy": 0.9668853133916855,
      "epoch": 4.729857819905213,
      "grad_norm": 0.3225063383579254,
      "learning_rate": 6.526612763648939e-05,
      "loss": 0.17831671237945557,
      "mean_token_accuracy": 0.9203702360391617,
      "num_tokens": 110797840.0,
      "step": 8982
    },
    {
      "entropy": 0.976748988032341,
      "epoch": 4.730384412848868,
      "grad_norm": 0.36254167556762695,
      "learning_rate": 6.524710122157411e-05,
      "loss": 0.181860089302063,
      "mean_token_accuracy": 0.9245222508907318,
      "num_tokens": 110810176.0,
      "step": 8983
    },
    {
      "entropy": 0.9795864969491959,
      "epoch": 4.730911005792523,
      "grad_norm": 0.3496047258377075,
      "learning_rate": 6.522807746337752e-05,
      "loss": 0.17530706524848938,
      "mean_token_accuracy": 0.9220332503318787,
      "num_tokens": 110822512.0,
      "step": 8984
    },
    {
      "entropy": 0.9837092906236649,
      "epoch": 4.731437598736177,
      "grad_norm": 0.28007742762565613,
      "learning_rate": 6.520905636302904e-05,
      "loss": 0.15277674794197083,
      "mean_token_accuracy": 0.9338098019361496,
      "num_tokens": 110834848.0,
      "step": 8985
    },
    {
      "entropy": 1.0120174884796143,
      "epoch": 4.731964191679832,
      "grad_norm": 0.3262519836425781,
      "learning_rate": 6.519003792165783e-05,
      "loss": 0.16972126066684723,
      "mean_token_accuracy": 0.9269426912069321,
      "num_tokens": 110847184.0,
      "step": 8986
    },
    {
      "entropy": 0.9969107508659363,
      "epoch": 4.7324907846234865,
      "grad_norm": 0.3321049213409424,
      "learning_rate": 6.51710221403929e-05,
      "loss": 0.16228239238262177,
      "mean_token_accuracy": 0.9264969974756241,
      "num_tokens": 110859520.0,
      "step": 8987
    },
    {
      "entropy": 0.928468719124794,
      "epoch": 4.7330173775671405,
      "grad_norm": 0.32601937651634216,
      "learning_rate": 6.515200902036309e-05,
      "loss": 0.16544225811958313,
      "mean_token_accuracy": 0.9298947602510452,
      "num_tokens": 110871856.0,
      "step": 8988
    },
    {
      "entropy": 0.9660092294216156,
      "epoch": 4.733543970510795,
      "grad_norm": 0.4767933785915375,
      "learning_rate": 6.513299856269713e-05,
      "loss": 0.16338896751403809,
      "mean_token_accuracy": 0.9304181486368179,
      "num_tokens": 110884192.0,
      "step": 8989
    },
    {
      "entropy": 0.9938349425792694,
      "epoch": 4.73407056345445,
      "grad_norm": 0.33547815680503845,
      "learning_rate": 6.511399076852354e-05,
      "loss": 0.17066051065921783,
      "mean_token_accuracy": 0.9304036498069763,
      "num_tokens": 110896528.0,
      "step": 8990
    },
    {
      "entropy": 0.9267756044864655,
      "epoch": 4.734597156398104,
      "grad_norm": 0.2973441481590271,
      "learning_rate": 6.509498563897068e-05,
      "loss": 0.1597248911857605,
      "mean_token_accuracy": 0.9334191530942917,
      "num_tokens": 110908864.0,
      "step": 8991
    },
    {
      "entropy": 0.9660896360874176,
      "epoch": 4.735123749341759,
      "grad_norm": 0.30963265895843506,
      "learning_rate": 6.50759831751668e-05,
      "loss": 0.173097163438797,
      "mean_token_accuracy": 0.9288729876279831,
      "num_tokens": 110921200.0,
      "step": 8992
    },
    {
      "entropy": 0.9519427716732025,
      "epoch": 4.735650342285413,
      "grad_norm": 0.3269384503364563,
      "learning_rate": 6.505698337823997e-05,
      "loss": 0.17985959351062775,
      "mean_token_accuracy": 0.9229910671710968,
      "num_tokens": 110933536.0,
      "step": 8993
    },
    {
      "entropy": 0.9285813122987747,
      "epoch": 4.736176935229068,
      "grad_norm": 0.2939811646938324,
      "learning_rate": 6.503798624931809e-05,
      "loss": 0.16949841380119324,
      "mean_token_accuracy": 0.92819844186306,
      "num_tokens": 110945872.0,
      "step": 8994
    },
    {
      "entropy": 0.9772068858146667,
      "epoch": 4.736703528172723,
      "grad_norm": 0.31877076625823975,
      "learning_rate": 6.501899178952891e-05,
      "loss": 0.167781800031662,
      "mean_token_accuracy": 0.9289694726467133,
      "num_tokens": 110958208.0,
      "step": 8995
    },
    {
      "entropy": 0.9644787162542343,
      "epoch": 4.737230121116377,
      "grad_norm": 0.35433515906333923,
      "learning_rate": 6.500000000000002e-05,
      "loss": 0.170842707157135,
      "mean_token_accuracy": 0.9287001937627792,
      "num_tokens": 110970544.0,
      "step": 8996
    },
    {
      "entropy": 0.9160168617963791,
      "epoch": 4.737756714060032,
      "grad_norm": 0.3172443211078644,
      "learning_rate": 6.498101088185886e-05,
      "loss": 0.16134260594844818,
      "mean_token_accuracy": 0.9317062944173813,
      "num_tokens": 110982880.0,
      "step": 8997
    },
    {
      "entropy": 0.961934044957161,
      "epoch": 4.7382833070036865,
      "grad_norm": 0.3247859477996826,
      "learning_rate": 6.496202443623262e-05,
      "loss": 0.17156106233596802,
      "mean_token_accuracy": 0.9250394701957703,
      "num_tokens": 110995216.0,
      "step": 8998
    },
    {
      "entropy": 0.9096788316965103,
      "epoch": 4.7388098999473405,
      "grad_norm": 0.30898916721343994,
      "learning_rate": 6.494304066424857e-05,
      "loss": 0.1646384447813034,
      "mean_token_accuracy": 0.9309989959001541,
      "num_tokens": 111007552.0,
      "step": 8999
    },
    {
      "entropy": 0.947771430015564,
      "epoch": 4.739336492890995,
      "grad_norm": 0.3102348744869232,
      "learning_rate": 6.492405956703356e-05,
      "loss": 0.15817317366600037,
      "mean_token_accuracy": 0.9337723106145859,
      "num_tokens": 111019888.0,
      "step": 9000
    },
    {
      "entropy": 0.9348080158233643,
      "epoch": 4.739863085834649,
      "grad_norm": 0.32881397008895874,
      "learning_rate": 6.490508114571445e-05,
      "loss": 0.16392850875854492,
      "mean_token_accuracy": 0.9309188723564148,
      "num_tokens": 111032224.0,
      "step": 9001
    },
    {
      "entropy": 0.9109790027141571,
      "epoch": 4.740389678778304,
      "grad_norm": 0.35365355014801025,
      "learning_rate": 6.488610540141779e-05,
      "loss": 0.1841769814491272,
      "mean_token_accuracy": 0.9249490648508072,
      "num_tokens": 111044560.0,
      "step": 9002
    },
    {
      "entropy": 0.8916398137807846,
      "epoch": 4.740916271721959,
      "grad_norm": 0.29434317350387573,
      "learning_rate": 6.486713233527013e-05,
      "loss": 0.1550382375717163,
      "mean_token_accuracy": 0.9333101511001587,
      "num_tokens": 111056896.0,
      "step": 9003
    },
    {
      "entropy": 0.97572922706604,
      "epoch": 4.741442864665613,
      "grad_norm": 0.36800870299339294,
      "learning_rate": 6.484816194839775e-05,
      "loss": 0.1771194338798523,
      "mean_token_accuracy": 0.9240408539772034,
      "num_tokens": 111069232.0,
      "step": 9004
    },
    {
      "entropy": 0.9397173225879669,
      "epoch": 4.741969457609268,
      "grad_norm": 0.36061525344848633,
      "learning_rate": 6.482919424192681e-05,
      "loss": 0.16837380826473236,
      "mean_token_accuracy": 0.9230096936225891,
      "num_tokens": 111081568.0,
      "step": 9005
    },
    {
      "entropy": 0.9172497689723969,
      "epoch": 4.742496050552923,
      "grad_norm": 0.3479461967945099,
      "learning_rate": 6.481022921698336e-05,
      "loss": 0.18055102229118347,
      "mean_token_accuracy": 0.9213292896747589,
      "num_tokens": 111093904.0,
      "step": 9006
    },
    {
      "entropy": 0.9395283907651901,
      "epoch": 4.743022643496577,
      "grad_norm": 0.33629798889160156,
      "learning_rate": 6.479126687469321e-05,
      "loss": 0.16937856376171112,
      "mean_token_accuracy": 0.9283296167850494,
      "num_tokens": 111106240.0,
      "step": 9007
    },
    {
      "entropy": 0.9321241974830627,
      "epoch": 4.743549236440232,
      "grad_norm": 0.31431910395622253,
      "learning_rate": 6.477230721618206e-05,
      "loss": 0.17476430535316467,
      "mean_token_accuracy": 0.9284104853868484,
      "num_tokens": 111118576.0,
      "step": 9008
    },
    {
      "entropy": 0.9074824899435043,
      "epoch": 4.7440758293838865,
      "grad_norm": 0.3159473240375519,
      "learning_rate": 6.475335024257543e-05,
      "loss": 0.17281442880630493,
      "mean_token_accuracy": 0.932155430316925,
      "num_tokens": 111130912.0,
      "step": 9009
    },
    {
      "entropy": 0.9505660831928253,
      "epoch": 4.7446024223275405,
      "grad_norm": 0.3291624188423157,
      "learning_rate": 6.473439595499867e-05,
      "loss": 0.1661178171634674,
      "mean_token_accuracy": 0.9278183430433273,
      "num_tokens": 111143248.0,
      "step": 9010
    },
    {
      "entropy": 0.899037092924118,
      "epoch": 4.745129015271195,
      "grad_norm": 0.31241390109062195,
      "learning_rate": 6.4715444354577e-05,
      "loss": 0.16685061156749725,
      "mean_token_accuracy": 0.9238582998514175,
      "num_tokens": 111155584.0,
      "step": 9011
    },
    {
      "entropy": 0.955901101231575,
      "epoch": 4.74565560821485,
      "grad_norm": 0.3158174157142639,
      "learning_rate": 6.469649544243543e-05,
      "loss": 0.16738471388816833,
      "mean_token_accuracy": 0.9265148341655731,
      "num_tokens": 111167920.0,
      "step": 9012
    },
    {
      "entropy": 0.9404389709234238,
      "epoch": 4.746182201158504,
      "grad_norm": 0.3212829530239105,
      "learning_rate": 6.467754921969891e-05,
      "loss": 0.17458640038967133,
      "mean_token_accuracy": 0.9287828207015991,
      "num_tokens": 111180256.0,
      "step": 9013
    },
    {
      "entropy": 0.9119441509246826,
      "epoch": 4.746708794102159,
      "grad_norm": 0.3021114766597748,
      "learning_rate": 6.465860568749212e-05,
      "loss": 0.16936270892620087,
      "mean_token_accuracy": 0.9257699996232986,
      "num_tokens": 111192592.0,
      "step": 9014
    },
    {
      "entropy": 0.9254357069730759,
      "epoch": 4.747235387045814,
      "grad_norm": 0.30540579557418823,
      "learning_rate": 6.463966484693966e-05,
      "loss": 0.16780777275562286,
      "mean_token_accuracy": 0.931959256529808,
      "num_tokens": 111204928.0,
      "step": 9015
    },
    {
      "entropy": 0.9446334689855576,
      "epoch": 4.747761979989468,
      "grad_norm": 0.325285941362381,
      "learning_rate": 6.46207266991659e-05,
      "loss": 0.17704302072525024,
      "mean_token_accuracy": 0.9249174743890762,
      "num_tokens": 111217264.0,
      "step": 9016
    },
    {
      "entropy": 0.9317489713430405,
      "epoch": 4.748288572933123,
      "grad_norm": 0.29853665828704834,
      "learning_rate": 6.46017912452951e-05,
      "loss": 0.16018593311309814,
      "mean_token_accuracy": 0.92707958817482,
      "num_tokens": 111229600.0,
      "step": 9017
    },
    {
      "entropy": 0.9343649744987488,
      "epoch": 4.748815165876778,
      "grad_norm": 0.31143051385879517,
      "learning_rate": 6.458285848645136e-05,
      "loss": 0.16280323266983032,
      "mean_token_accuracy": 0.9271020591259003,
      "num_tokens": 111241936.0,
      "step": 9018
    },
    {
      "entropy": 0.9551049172878265,
      "epoch": 4.749341758820432,
      "grad_norm": 0.28899648785591125,
      "learning_rate": 6.456392842375858e-05,
      "loss": 0.1555640995502472,
      "mean_token_accuracy": 0.9314574897289276,
      "num_tokens": 111254272.0,
      "step": 9019
    },
    {
      "entropy": 0.9203036278486252,
      "epoch": 4.7498683517640865,
      "grad_norm": 0.31474632024765015,
      "learning_rate": 6.454500105834058e-05,
      "loss": 0.16402553021907806,
      "mean_token_accuracy": 0.9308990687131882,
      "num_tokens": 111266608.0,
      "step": 9020
    },
    {
      "entropy": 0.9286926239728928,
      "epoch": 4.750394944707741,
      "grad_norm": 0.33371585607528687,
      "learning_rate": 6.452607639132093e-05,
      "loss": 0.18327629566192627,
      "mean_token_accuracy": 0.921772688627243,
      "num_tokens": 111278944.0,
      "step": 9021
    },
    {
      "entropy": 0.946032926440239,
      "epoch": 4.750921537651395,
      "grad_norm": 0.3166511654853821,
      "learning_rate": 6.450715442382308e-05,
      "loss": 0.1802721470594406,
      "mean_token_accuracy": 0.9210911840200424,
      "num_tokens": 111291280.0,
      "step": 9022
    },
    {
      "entropy": 0.9355271309614182,
      "epoch": 4.75144813059505,
      "grad_norm": 0.3164330720901489,
      "learning_rate": 6.448823515697031e-05,
      "loss": 0.17939233779907227,
      "mean_token_accuracy": 0.9239465445280075,
      "num_tokens": 111303616.0,
      "step": 9023
    },
    {
      "entropy": 0.9391446709632874,
      "epoch": 4.751974723538704,
      "grad_norm": 0.34126725792884827,
      "learning_rate": 6.446931859188576e-05,
      "loss": 0.17269359529018402,
      "mean_token_accuracy": 0.9273637980222702,
      "num_tokens": 111315952.0,
      "step": 9024
    },
    {
      "entropy": 0.9569701850414276,
      "epoch": 4.752501316482359,
      "grad_norm": 0.34143349528312683,
      "learning_rate": 6.445040472969241e-05,
      "loss": 0.17444859445095062,
      "mean_token_accuracy": 0.9257920831441879,
      "num_tokens": 111328288.0,
      "step": 9025
    },
    {
      "entropy": 0.9466724842786789,
      "epoch": 4.753027909426014,
      "grad_norm": 0.3243260681629181,
      "learning_rate": 6.443149357151299e-05,
      "loss": 0.18756012618541718,
      "mean_token_accuracy": 0.9216291159391403,
      "num_tokens": 111340624.0,
      "step": 9026
    },
    {
      "entropy": 0.9434637129306793,
      "epoch": 4.753554502369668,
      "grad_norm": 0.3246390223503113,
      "learning_rate": 6.441258511847024e-05,
      "loss": 0.17142389714717865,
      "mean_token_accuracy": 0.9252357184886932,
      "num_tokens": 111352960.0,
      "step": 9027
    },
    {
      "entropy": 0.9123575240373611,
      "epoch": 4.754081095313323,
      "grad_norm": 0.2930547297000885,
      "learning_rate": 6.439367937168662e-05,
      "loss": 0.15911197662353516,
      "mean_token_accuracy": 0.933475911617279,
      "num_tokens": 111365296.0,
      "step": 9028
    },
    {
      "entropy": 0.928584486246109,
      "epoch": 4.754607688256978,
      "grad_norm": 0.27964845299720764,
      "learning_rate": 6.437477633228443e-05,
      "loss": 0.15053358674049377,
      "mean_token_accuracy": 0.9340831637382507,
      "num_tokens": 111377632.0,
      "step": 9029
    },
    {
      "entropy": 0.9594197422266006,
      "epoch": 4.755134281200632,
      "grad_norm": 0.3099207282066345,
      "learning_rate": 6.435587600138587e-05,
      "loss": 0.1709914356470108,
      "mean_token_accuracy": 0.9311748147010803,
      "num_tokens": 111389968.0,
      "step": 9030
    },
    {
      "entropy": 0.9574659168720245,
      "epoch": 4.7556608741442865,
      "grad_norm": 0.3323180377483368,
      "learning_rate": 6.433697838011292e-05,
      "loss": 0.17146876454353333,
      "mean_token_accuracy": 0.9289543330669403,
      "num_tokens": 111402304.0,
      "step": 9031
    },
    {
      "entropy": 0.9474862664937973,
      "epoch": 4.756187467087941,
      "grad_norm": 0.30546411871910095,
      "learning_rate": 6.43180834695874e-05,
      "loss": 0.15925800800323486,
      "mean_token_accuracy": 0.9380460530519485,
      "num_tokens": 111414640.0,
      "step": 9032
    },
    {
      "entropy": 0.9810915142297745,
      "epoch": 4.756714060031595,
      "grad_norm": 0.3353631794452667,
      "learning_rate": 6.429919127093098e-05,
      "loss": 0.17721283435821533,
      "mean_token_accuracy": 0.9227162003517151,
      "num_tokens": 111426976.0,
      "step": 9033
    },
    {
      "entropy": 1.0023733377456665,
      "epoch": 4.75724065297525,
      "grad_norm": 0.3292175531387329,
      "learning_rate": 6.428030178526522e-05,
      "loss": 0.17148852348327637,
      "mean_token_accuracy": 0.9293312579393387,
      "num_tokens": 111439312.0,
      "step": 9034
    },
    {
      "entropy": 0.9760813117027283,
      "epoch": 4.757767245918904,
      "grad_norm": 0.32482001185417175,
      "learning_rate": 6.426141501371151e-05,
      "loss": 0.17811252176761627,
      "mean_token_accuracy": 0.9272367805242538,
      "num_tokens": 111451648.0,
      "step": 9035
    },
    {
      "entropy": 0.901419147849083,
      "epoch": 4.758293838862559,
      "grad_norm": 0.3026784062385559,
      "learning_rate": 6.424253095739097e-05,
      "loss": 0.1650366485118866,
      "mean_token_accuracy": 0.9297737032175064,
      "num_tokens": 111463984.0,
      "step": 9036
    },
    {
      "entropy": 0.9676057994365692,
      "epoch": 4.758820431806214,
      "grad_norm": 0.32503530383110046,
      "learning_rate": 6.422364961742468e-05,
      "loss": 0.16921848058700562,
      "mean_token_accuracy": 0.928037241101265,
      "num_tokens": 111476320.0,
      "step": 9037
    },
    {
      "entropy": 0.9711627960205078,
      "epoch": 4.759347024749868,
      "grad_norm": 0.3061572015285492,
      "learning_rate": 6.420477099493352e-05,
      "loss": 0.1675848662853241,
      "mean_token_accuracy": 0.9253803789615631,
      "num_tokens": 111488656.0,
      "step": 9038
    },
    {
      "entropy": 0.9869196563959122,
      "epoch": 4.759873617693523,
      "grad_norm": 0.3351416289806366,
      "learning_rate": 6.418589509103818e-05,
      "loss": 0.18046732246875763,
      "mean_token_accuracy": 0.922397568821907,
      "num_tokens": 111500992.0,
      "step": 9039
    },
    {
      "entropy": 0.9444341063499451,
      "epoch": 4.760400210637178,
      "grad_norm": 0.33065369725227356,
      "learning_rate": 6.416702190685922e-05,
      "loss": 0.18055486679077148,
      "mean_token_accuracy": 0.9234813004732132,
      "num_tokens": 111513328.0,
      "step": 9040
    },
    {
      "entropy": 0.9259567856788635,
      "epoch": 4.760926803580832,
      "grad_norm": 0.3338969647884369,
      "learning_rate": 6.414815144351705e-05,
      "loss": 0.18156734108924866,
      "mean_token_accuracy": 0.9217091202735901,
      "num_tokens": 111525664.0,
      "step": 9041
    },
    {
      "entropy": 0.9558723270893097,
      "epoch": 4.7614533965244865,
      "grad_norm": 0.35250309109687805,
      "learning_rate": 6.412928370213186e-05,
      "loss": 0.17602044343948364,
      "mean_token_accuracy": 0.9272420257329941,
      "num_tokens": 111538000.0,
      "step": 9042
    },
    {
      "entropy": 0.9206525087356567,
      "epoch": 4.761979989468141,
      "grad_norm": 0.31426510214805603,
      "learning_rate": 6.411041868382376e-05,
      "loss": 0.1601458340883255,
      "mean_token_accuracy": 0.9292854964733124,
      "num_tokens": 111550336.0,
      "step": 9043
    },
    {
      "entropy": 0.970268115401268,
      "epoch": 4.762506582411795,
      "grad_norm": 0.32940199971199036,
      "learning_rate": 6.409155638971264e-05,
      "loss": 0.15830670297145844,
      "mean_token_accuracy": 0.931570902466774,
      "num_tokens": 111562672.0,
      "step": 9044
    },
    {
      "entropy": 0.9746520668268204,
      "epoch": 4.76303317535545,
      "grad_norm": 0.3464902937412262,
      "learning_rate": 6.407269682091823e-05,
      "loss": 0.1703244149684906,
      "mean_token_accuracy": 0.9258856028318405,
      "num_tokens": 111575008.0,
      "step": 9045
    },
    {
      "entropy": 0.9502895623445511,
      "epoch": 4.763559768299105,
      "grad_norm": 0.30303528904914856,
      "learning_rate": 6.405383997856014e-05,
      "loss": 0.16198208928108215,
      "mean_token_accuracy": 0.9311043471097946,
      "num_tokens": 111587344.0,
      "step": 9046
    },
    {
      "entropy": 0.9064134359359741,
      "epoch": 4.764086361242759,
      "grad_norm": 0.30101168155670166,
      "learning_rate": 6.403498586375774e-05,
      "loss": 0.16409745812416077,
      "mean_token_accuracy": 0.9261920154094696,
      "num_tokens": 111599680.0,
      "step": 9047
    },
    {
      "entropy": 0.9261251986026764,
      "epoch": 4.764612954186414,
      "grad_norm": 0.28687286376953125,
      "learning_rate": 6.401613447763036e-05,
      "loss": 0.16393136978149414,
      "mean_token_accuracy": 0.929685965180397,
      "num_tokens": 111612016.0,
      "step": 9048
    },
    {
      "entropy": 0.9612890779972076,
      "epoch": 4.765139547130069,
      "grad_norm": 0.3660378158092499,
      "learning_rate": 6.399728582129704e-05,
      "loss": 0.18570277094841003,
      "mean_token_accuracy": 0.9182675331830978,
      "num_tokens": 111624352.0,
      "step": 9049
    },
    {
      "entropy": 0.9497331380844116,
      "epoch": 4.765666140073723,
      "grad_norm": 0.3135928511619568,
      "learning_rate": 6.397843989587677e-05,
      "loss": 0.16727733612060547,
      "mean_token_accuracy": 0.9272850453853607,
      "num_tokens": 111636688.0,
      "step": 9050
    },
    {
      "entropy": 0.9651635140180588,
      "epoch": 4.766192733017378,
      "grad_norm": 0.2958713173866272,
      "learning_rate": 6.395959670248827e-05,
      "loss": 0.15490230917930603,
      "mean_token_accuracy": 0.9338932633399963,
      "num_tokens": 111649024.0,
      "step": 9051
    },
    {
      "entropy": 0.9416981935501099,
      "epoch": 4.7667193259610325,
      "grad_norm": 0.33557194471359253,
      "learning_rate": 6.394075624225019e-05,
      "loss": 0.1833014041185379,
      "mean_token_accuracy": 0.923352062702179,
      "num_tokens": 111661360.0,
      "step": 9052
    },
    {
      "entropy": 0.924249678850174,
      "epoch": 4.7672459189046865,
      "grad_norm": 0.32141128182411194,
      "learning_rate": 6.392191851628094e-05,
      "loss": 0.17952987551689148,
      "mean_token_accuracy": 0.9223910272121429,
      "num_tokens": 111673696.0,
      "step": 9053
    },
    {
      "entropy": 0.9650433510541916,
      "epoch": 4.767772511848341,
      "grad_norm": 0.34635400772094727,
      "learning_rate": 6.390308352569878e-05,
      "loss": 0.1668970137834549,
      "mean_token_accuracy": 0.9306865334510803,
      "num_tokens": 111686032.0,
      "step": 9054
    },
    {
      "entropy": 0.9683272689580917,
      "epoch": 4.768299104791996,
      "grad_norm": 0.32868492603302,
      "learning_rate": 6.388425127162192e-05,
      "loss": 0.17090605199337006,
      "mean_token_accuracy": 0.9256576895713806,
      "num_tokens": 111698368.0,
      "step": 9055
    },
    {
      "entropy": 0.9745743274688721,
      "epoch": 4.76882569773565,
      "grad_norm": 0.34252041578292847,
      "learning_rate": 6.386542175516828e-05,
      "loss": 0.183536559343338,
      "mean_token_accuracy": 0.9262122213840485,
      "num_tokens": 111710704.0,
      "step": 9056
    },
    {
      "entropy": 0.95796699821949,
      "epoch": 4.769352290679305,
      "grad_norm": 0.3901304304599762,
      "learning_rate": 6.384659497745566e-05,
      "loss": 0.1622026264667511,
      "mean_token_accuracy": 0.9298348277807236,
      "num_tokens": 111723040.0,
      "step": 9057
    },
    {
      "entropy": 0.964602530002594,
      "epoch": 4.769878883622959,
      "grad_norm": 0.3439449071884155,
      "learning_rate": 6.382777093960169e-05,
      "loss": 0.1925317943096161,
      "mean_token_accuracy": 0.920476108789444,
      "num_tokens": 111735376.0,
      "step": 9058
    },
    {
      "entropy": 0.9869271069765091,
      "epoch": 4.770405476566614,
      "grad_norm": 0.316421777009964,
      "learning_rate": 6.380894964272384e-05,
      "loss": 0.1708654910326004,
      "mean_token_accuracy": 0.9293173998594284,
      "num_tokens": 111747712.0,
      "step": 9059
    },
    {
      "entropy": 0.9626688212156296,
      "epoch": 4.770932069510269,
      "grad_norm": 0.29398012161254883,
      "learning_rate": 6.379013108793945e-05,
      "loss": 0.1604207158088684,
      "mean_token_accuracy": 0.9344052374362946,
      "num_tokens": 111760048.0,
      "step": 9060
    },
    {
      "entropy": 0.9678095579147339,
      "epoch": 4.771458662453923,
      "grad_norm": 0.3171744644641876,
      "learning_rate": 6.377131527636564e-05,
      "loss": 0.1560574471950531,
      "mean_token_accuracy": 0.9331692159175873,
      "num_tokens": 111772384.0,
      "step": 9061
    },
    {
      "entropy": 1.0159512609243393,
      "epoch": 4.771985255397578,
      "grad_norm": 0.33914652466773987,
      "learning_rate": 6.375250220911939e-05,
      "loss": 0.16417792439460754,
      "mean_token_accuracy": 0.9308343529701233,
      "num_tokens": 111784720.0,
      "step": 9062
    },
    {
      "entropy": 0.959560289978981,
      "epoch": 4.7725118483412325,
      "grad_norm": 0.30077576637268066,
      "learning_rate": 6.373369188731752e-05,
      "loss": 0.16021844744682312,
      "mean_token_accuracy": 0.932145431637764,
      "num_tokens": 111797056.0,
      "step": 9063
    },
    {
      "entropy": 0.981648176908493,
      "epoch": 4.7730384412848865,
      "grad_norm": 0.3147124648094177,
      "learning_rate": 6.371488431207671e-05,
      "loss": 0.16345477104187012,
      "mean_token_accuracy": 0.9277648627758026,
      "num_tokens": 111809392.0,
      "step": 9064
    },
    {
      "entropy": 0.9431145936250687,
      "epoch": 4.773565034228541,
      "grad_norm": 0.3094829320907593,
      "learning_rate": 6.369607948451346e-05,
      "loss": 0.15901650488376617,
      "mean_token_accuracy": 0.9305784106254578,
      "num_tokens": 111821728.0,
      "step": 9065
    },
    {
      "entropy": 0.9495803564786911,
      "epoch": 4.774091627172196,
      "grad_norm": 0.33323946595191956,
      "learning_rate": 6.36772774057441e-05,
      "loss": 0.17987127602100372,
      "mean_token_accuracy": 0.9216539263725281,
      "num_tokens": 111834064.0,
      "step": 9066
    },
    {
      "entropy": 0.9451373517513275,
      "epoch": 4.77461822011585,
      "grad_norm": 0.31161990761756897,
      "learning_rate": 6.365847807688478e-05,
      "loss": 0.1630258560180664,
      "mean_token_accuracy": 0.9296188056468964,
      "num_tokens": 111846400.0,
      "step": 9067
    },
    {
      "entropy": 0.9265732914209366,
      "epoch": 4.775144813059505,
      "grad_norm": 0.2965737283229828,
      "learning_rate": 6.363968149905148e-05,
      "loss": 0.16398824751377106,
      "mean_token_accuracy": 0.9289033710956573,
      "num_tokens": 111858736.0,
      "step": 9068
    },
    {
      "entropy": 0.9546487033367157,
      "epoch": 4.775671406003159,
      "grad_norm": 0.3014756739139557,
      "learning_rate": 6.362088767336014e-05,
      "loss": 0.16929973661899567,
      "mean_token_accuracy": 0.9295613765716553,
      "num_tokens": 111871072.0,
      "step": 9069
    },
    {
      "entropy": 0.9728638082742691,
      "epoch": 4.776197998946814,
      "grad_norm": 0.3430885672569275,
      "learning_rate": 6.360209660092637e-05,
      "loss": 0.18163266777992249,
      "mean_token_accuracy": 0.9220084995031357,
      "num_tokens": 111883408.0,
      "step": 9070
    },
    {
      "entropy": 0.9284171313047409,
      "epoch": 4.776724591890469,
      "grad_norm": 0.31685933470726013,
      "learning_rate": 6.358330828286567e-05,
      "loss": 0.16648146510124207,
      "mean_token_accuracy": 0.9305678457021713,
      "num_tokens": 111895744.0,
      "step": 9071
    },
    {
      "entropy": 0.9554715007543564,
      "epoch": 4.777251184834123,
      "grad_norm": 0.3176212012767792,
      "learning_rate": 6.356452272029344e-05,
      "loss": 0.1680004894733429,
      "mean_token_accuracy": 0.9303200393915176,
      "num_tokens": 111908080.0,
      "step": 9072
    },
    {
      "entropy": 0.9570530503988266,
      "epoch": 4.777777777777778,
      "grad_norm": 0.35659483075141907,
      "learning_rate": 6.354573991432485e-05,
      "loss": 0.1893249899148941,
      "mean_token_accuracy": 0.9239123463630676,
      "num_tokens": 111920416.0,
      "step": 9073
    },
    {
      "entropy": 0.9419193267822266,
      "epoch": 4.7783043707214325,
      "grad_norm": 0.3028180003166199,
      "learning_rate": 6.352695986607493e-05,
      "loss": 0.16518285870552063,
      "mean_token_accuracy": 0.9283025413751602,
      "num_tokens": 111932752.0,
      "step": 9074
    },
    {
      "entropy": 0.9351107627153397,
      "epoch": 4.7788309636650865,
      "grad_norm": 0.29300156235694885,
      "learning_rate": 6.350818257665849e-05,
      "loss": 0.16043254733085632,
      "mean_token_accuracy": 0.9351049810647964,
      "num_tokens": 111945088.0,
      "step": 9075
    },
    {
      "entropy": 0.9588852375745773,
      "epoch": 4.779357556608741,
      "grad_norm": 0.30463680624961853,
      "learning_rate": 6.34894080471903e-05,
      "loss": 0.1685238480567932,
      "mean_token_accuracy": 0.9276732206344604,
      "num_tokens": 111957424.0,
      "step": 9076
    },
    {
      "entropy": 0.9524371176958084,
      "epoch": 4.779884149552396,
      "grad_norm": 0.31636473536491394,
      "learning_rate": 6.34706362787849e-05,
      "loss": 0.16936713457107544,
      "mean_token_accuracy": 0.9265929758548737,
      "num_tokens": 111969760.0,
      "step": 9077
    },
    {
      "entropy": 0.9338591247797012,
      "epoch": 4.78041074249605,
      "grad_norm": 0.31156182289123535,
      "learning_rate": 6.345186727255661e-05,
      "loss": 0.17318032681941986,
      "mean_token_accuracy": 0.9255554527044296,
      "num_tokens": 111982096.0,
      "step": 9078
    },
    {
      "entropy": 0.9713913351297379,
      "epoch": 4.780937335439705,
      "grad_norm": 0.3218269944190979,
      "learning_rate": 6.343310102961966e-05,
      "loss": 0.1645289808511734,
      "mean_token_accuracy": 0.9315515011548996,
      "num_tokens": 111994432.0,
      "step": 9079
    },
    {
      "entropy": 0.9637110829353333,
      "epoch": 4.78146392838336,
      "grad_norm": 0.33749082684516907,
      "learning_rate": 6.341433755108813e-05,
      "loss": 0.17996618151664734,
      "mean_token_accuracy": 0.9206294417381287,
      "num_tokens": 112006768.0,
      "step": 9080
    },
    {
      "entropy": 0.9192416071891785,
      "epoch": 4.781990521327014,
      "grad_norm": 0.28524866700172424,
      "learning_rate": 6.339557683807581e-05,
      "loss": 0.15543463826179504,
      "mean_token_accuracy": 0.934823215007782,
      "num_tokens": 112019104.0,
      "step": 9081
    },
    {
      "entropy": 0.9043999761343002,
      "epoch": 4.782517114270669,
      "grad_norm": 0.3133680820465088,
      "learning_rate": 6.337681889169643e-05,
      "loss": 0.16104987263679504,
      "mean_token_accuracy": 0.929581955075264,
      "num_tokens": 112031440.0,
      "step": 9082
    },
    {
      "entropy": 0.9359278827905655,
      "epoch": 4.783043707214324,
      "grad_norm": 0.30294913053512573,
      "learning_rate": 6.33580637130636e-05,
      "loss": 0.16375668346881866,
      "mean_token_accuracy": 0.9279943853616714,
      "num_tokens": 112043776.0,
      "step": 9083
    },
    {
      "entropy": 0.9404270648956299,
      "epoch": 4.783570300157978,
      "grad_norm": 0.3146786093711853,
      "learning_rate": 6.333931130329069e-05,
      "loss": 0.17909596860408783,
      "mean_token_accuracy": 0.9296393245458603,
      "num_tokens": 112056112.0,
      "step": 9084
    },
    {
      "entropy": 0.9564862996339798,
      "epoch": 4.7840968931016326,
      "grad_norm": 0.34017446637153625,
      "learning_rate": 6.332056166349091e-05,
      "loss": 0.1690351665019989,
      "mean_token_accuracy": 0.9247809499502182,
      "num_tokens": 112068448.0,
      "step": 9085
    },
    {
      "entropy": 0.9514518976211548,
      "epoch": 4.784623486045287,
      "grad_norm": 0.30882126092910767,
      "learning_rate": 6.330181479477733e-05,
      "loss": 0.16744601726531982,
      "mean_token_accuracy": 0.9311587661504745,
      "num_tokens": 112080784.0,
      "step": 9086
    },
    {
      "entropy": 0.9229488521814346,
      "epoch": 4.785150078988941,
      "grad_norm": 0.33422407507896423,
      "learning_rate": 6.32830706982628e-05,
      "loss": 0.17872753739356995,
      "mean_token_accuracy": 0.9267108738422394,
      "num_tokens": 112093120.0,
      "step": 9087
    },
    {
      "entropy": 0.9272540509700775,
      "epoch": 4.785676671932596,
      "grad_norm": 0.3081918954849243,
      "learning_rate": 6.326432937506009e-05,
      "loss": 0.17642512917518616,
      "mean_token_accuracy": 0.9253614991903305,
      "num_tokens": 112105456.0,
      "step": 9088
    },
    {
      "entropy": 0.9433393478393555,
      "epoch": 4.786203264876251,
      "grad_norm": 0.3381654620170593,
      "learning_rate": 6.324559082628175e-05,
      "loss": 0.18298709392547607,
      "mean_token_accuracy": 0.9250814318656921,
      "num_tokens": 112117792.0,
      "step": 9089
    },
    {
      "entropy": 0.9583987295627594,
      "epoch": 4.786729857819905,
      "grad_norm": 0.29826247692108154,
      "learning_rate": 6.322685505304018e-05,
      "loss": 0.15026958286762238,
      "mean_token_accuracy": 0.936936229467392,
      "num_tokens": 112130128.0,
      "step": 9090
    },
    {
      "entropy": 0.9661408960819244,
      "epoch": 4.78725645076356,
      "grad_norm": 0.32385575771331787,
      "learning_rate": 6.32081220564476e-05,
      "loss": 0.17515888810157776,
      "mean_token_accuracy": 0.9241830855607986,
      "num_tokens": 112142464.0,
      "step": 9091
    },
    {
      "entropy": 0.8939239084720612,
      "epoch": 4.787783043707214,
      "grad_norm": 0.3073534369468689,
      "learning_rate": 6.31893918376161e-05,
      "loss": 0.15793702006340027,
      "mean_token_accuracy": 0.9343563467264175,
      "num_tokens": 112154800.0,
      "step": 9092
    },
    {
      "entropy": 0.906022310256958,
      "epoch": 4.788309636650869,
      "grad_norm": 0.3248325288295746,
      "learning_rate": 6.317066439765756e-05,
      "loss": 0.18267102539539337,
      "mean_token_accuracy": 0.9195359200239182,
      "num_tokens": 112167136.0,
      "step": 9093
    },
    {
      "entropy": 0.9159866869449615,
      "epoch": 4.788836229594524,
      "grad_norm": 0.30191734433174133,
      "learning_rate": 6.315193973768374e-05,
      "loss": 0.17407076060771942,
      "mean_token_accuracy": 0.9262288957834244,
      "num_tokens": 112179472.0,
      "step": 9094
    },
    {
      "entropy": 0.9487686604261398,
      "epoch": 4.789362822538178,
      "grad_norm": 0.31542494893074036,
      "learning_rate": 6.313321785880621e-05,
      "loss": 0.1765207052230835,
      "mean_token_accuracy": 0.9244837909936905,
      "num_tokens": 112191808.0,
      "step": 9095
    },
    {
      "entropy": 0.9285544902086258,
      "epoch": 4.789889415481833,
      "grad_norm": 0.318965345621109,
      "learning_rate": 6.311449876213633e-05,
      "loss": 0.1790979951620102,
      "mean_token_accuracy": 0.9226861447095871,
      "num_tokens": 112204144.0,
      "step": 9096
    },
    {
      "entropy": 0.9164725095033646,
      "epoch": 4.790416008425487,
      "grad_norm": 0.31515586376190186,
      "learning_rate": 6.309578244878544e-05,
      "loss": 0.1683732271194458,
      "mean_token_accuracy": 0.9258377999067307,
      "num_tokens": 112216480.0,
      "step": 9097
    },
    {
      "entropy": 0.9288828819990158,
      "epoch": 4.790942601369141,
      "grad_norm": 0.3244929015636444,
      "learning_rate": 6.307706891986456e-05,
      "loss": 0.1691761016845703,
      "mean_token_accuracy": 0.9247510433197021,
      "num_tokens": 112228816.0,
      "step": 9098
    },
    {
      "entropy": 0.9310002326965332,
      "epoch": 4.791469194312796,
      "grad_norm": 0.3144392669200897,
      "learning_rate": 6.305835817648464e-05,
      "loss": 0.1585865616798401,
      "mean_token_accuracy": 0.9319644272327423,
      "num_tokens": 112241152.0,
      "step": 9099
    },
    {
      "entropy": 0.9388798475265503,
      "epoch": 4.791995787256451,
      "grad_norm": 0.32102328538894653,
      "learning_rate": 6.303965021975636e-05,
      "loss": 0.17396274209022522,
      "mean_token_accuracy": 0.9284583330154419,
      "num_tokens": 112253488.0,
      "step": 9100
    },
    {
      "entropy": 0.9336607456207275,
      "epoch": 4.792522380200105,
      "grad_norm": 0.2871127128601074,
      "learning_rate": 6.302094505079035e-05,
      "loss": 0.15618155896663666,
      "mean_token_accuracy": 0.9364693909883499,
      "num_tokens": 112265824.0,
      "step": 9101
    },
    {
      "entropy": 0.938468337059021,
      "epoch": 4.79304897314376,
      "grad_norm": 0.29081955552101135,
      "learning_rate": 6.300224267069702e-05,
      "loss": 0.15577498078346252,
      "mean_token_accuracy": 0.9322134852409363,
      "num_tokens": 112278160.0,
      "step": 9102
    },
    {
      "entropy": 0.9274968355894089,
      "epoch": 4.793575566087414,
      "grad_norm": 0.33888524770736694,
      "learning_rate": 6.298354308058658e-05,
      "loss": 0.17000727355480194,
      "mean_token_accuracy": 0.9261871576309204,
      "num_tokens": 112290496.0,
      "step": 9103
    },
    {
      "entropy": 0.9365573227405548,
      "epoch": 4.794102159031069,
      "grad_norm": 0.31890496611595154,
      "learning_rate": 6.296484628156918e-05,
      "loss": 0.16275769472122192,
      "mean_token_accuracy": 0.9281330853700638,
      "num_tokens": 112302832.0,
      "step": 9104
    },
    {
      "entropy": 0.9545449614524841,
      "epoch": 4.794628751974724,
      "grad_norm": 0.3689518868923187,
      "learning_rate": 6.294615227475474e-05,
      "loss": 0.18185900151729584,
      "mean_token_accuracy": 0.9226024895906448,
      "num_tokens": 112315168.0,
      "step": 9105
    },
    {
      "entropy": 0.9652837216854095,
      "epoch": 4.795155344918378,
      "grad_norm": 0.34969696402549744,
      "learning_rate": 6.292746106125297e-05,
      "loss": 0.18228575587272644,
      "mean_token_accuracy": 0.921035885810852,
      "num_tokens": 112327504.0,
      "step": 9106
    },
    {
      "entropy": 0.92081518471241,
      "epoch": 4.795681937862033,
      "grad_norm": 0.3451554775238037,
      "learning_rate": 6.290877264217347e-05,
      "loss": 0.19245527684688568,
      "mean_token_accuracy": 0.9229613840579987,
      "num_tokens": 112339840.0,
      "step": 9107
    },
    {
      "entropy": 0.9458408206701279,
      "epoch": 4.796208530805687,
      "grad_norm": 0.32558906078338623,
      "learning_rate": 6.289008701862569e-05,
      "loss": 0.16560578346252441,
      "mean_token_accuracy": 0.9308914095163345,
      "num_tokens": 112352176.0,
      "step": 9108
    },
    {
      "entropy": 0.9107428342103958,
      "epoch": 4.796735123749341,
      "grad_norm": 0.30025514960289,
      "learning_rate": 6.287140419171885e-05,
      "loss": 0.1743806004524231,
      "mean_token_accuracy": 0.9275696724653244,
      "num_tokens": 112364512.0,
      "step": 9109
    },
    {
      "entropy": 0.9333881288766861,
      "epoch": 4.797261716692996,
      "grad_norm": 0.31256020069122314,
      "learning_rate": 6.285272416256207e-05,
      "loss": 0.16271354258060455,
      "mean_token_accuracy": 0.9293209463357925,
      "num_tokens": 112376848.0,
      "step": 9110
    },
    {
      "entropy": 0.9459578990936279,
      "epoch": 4.797788309636651,
      "grad_norm": 0.32335054874420166,
      "learning_rate": 6.283404693226424e-05,
      "loss": 0.17523100972175598,
      "mean_token_accuracy": 0.9279122352600098,
      "num_tokens": 112389184.0,
      "step": 9111
    },
    {
      "entropy": 0.9072926044464111,
      "epoch": 4.798314902580305,
      "grad_norm": 0.2974752187728882,
      "learning_rate": 6.281537250193417e-05,
      "loss": 0.15954649448394775,
      "mean_token_accuracy": 0.930755153298378,
      "num_tokens": 112401520.0,
      "step": 9112
    },
    {
      "entropy": 0.970455676317215,
      "epoch": 4.79884149552396,
      "grad_norm": 0.3365978002548218,
      "learning_rate": 6.27967008726804e-05,
      "loss": 0.17074136435985565,
      "mean_token_accuracy": 0.9294116944074631,
      "num_tokens": 112413856.0,
      "step": 9113
    },
    {
      "entropy": 0.9673090875148773,
      "epoch": 4.799368088467615,
      "grad_norm": 0.3307856023311615,
      "learning_rate": 6.277803204561138e-05,
      "loss": 0.16063982248306274,
      "mean_token_accuracy": 0.9351110458374023,
      "num_tokens": 112426192.0,
      "step": 9114
    },
    {
      "entropy": 0.9420793056488037,
      "epoch": 4.799894681411269,
      "grad_norm": 0.3055140972137451,
      "learning_rate": 6.275936602183537e-05,
      "loss": 0.16511839628219604,
      "mean_token_accuracy": 0.9317631870508194,
      "num_tokens": 112438528.0,
      "step": 9115
    },
    {
      "entropy": 0.965690553188324,
      "epoch": 4.800421274354924,
      "grad_norm": 0.335700124502182,
      "learning_rate": 6.274070280246047e-05,
      "loss": 0.18881773948669434,
      "mean_token_accuracy": 0.9202330857515335,
      "num_tokens": 112450864.0,
      "step": 9116
    },
    {
      "entropy": 0.9024168848991394,
      "epoch": 4.800947867298579,
      "grad_norm": 0.31919345259666443,
      "learning_rate": 6.272204238859453e-05,
      "loss": 0.17541448771953583,
      "mean_token_accuracy": 0.9276711642742157,
      "num_tokens": 112463200.0,
      "step": 9117
    },
    {
      "entropy": 0.9271547645330429,
      "epoch": 4.801474460242233,
      "grad_norm": 0.3327969014644623,
      "learning_rate": 6.270338478134547e-05,
      "loss": 0.1676233410835266,
      "mean_token_accuracy": 0.9297514259815216,
      "num_tokens": 112475536.0,
      "step": 9118
    },
    {
      "entropy": 0.9096930176019669,
      "epoch": 4.802001053185887,
      "grad_norm": 0.290777325630188,
      "learning_rate": 6.268472998182076e-05,
      "loss": 0.1540253609418869,
      "mean_token_accuracy": 0.9340391308069229,
      "num_tokens": 112487872.0,
      "step": 9119
    },
    {
      "entropy": 0.9078056812286377,
      "epoch": 4.802527646129542,
      "grad_norm": 0.2970122992992401,
      "learning_rate": 6.266607799112786e-05,
      "loss": 0.16196846961975098,
      "mean_token_accuracy": 0.9300400167703629,
      "num_tokens": 112500208.0,
      "step": 9120
    },
    {
      "entropy": 0.9710754305124283,
      "epoch": 4.803054239073196,
      "grad_norm": 0.3165084719657898,
      "learning_rate": 6.264742881037401e-05,
      "loss": 0.17036688327789307,
      "mean_token_accuracy": 0.9278738349676132,
      "num_tokens": 112512544.0,
      "step": 9121
    },
    {
      "entropy": 0.8850185126066208,
      "epoch": 4.803580832016851,
      "grad_norm": 0.320492148399353,
      "learning_rate": 6.262878244066632e-05,
      "loss": 0.1639227271080017,
      "mean_token_accuracy": 0.931957870721817,
      "num_tokens": 112524880.0,
      "step": 9122
    },
    {
      "entropy": 0.8749157786369324,
      "epoch": 4.804107424960506,
      "grad_norm": 0.30157896876335144,
      "learning_rate": 6.261013888311173e-05,
      "loss": 0.1490166187286377,
      "mean_token_accuracy": 0.9377554953098297,
      "num_tokens": 112537216.0,
      "step": 9123
    },
    {
      "entropy": 0.9255085736513138,
      "epoch": 4.80463401790416,
      "grad_norm": 0.32794201374053955,
      "learning_rate": 6.259149813881695e-05,
      "loss": 0.17577221989631653,
      "mean_token_accuracy": 0.9207169562578201,
      "num_tokens": 112549552.0,
      "step": 9124
    },
    {
      "entropy": 0.9130100309848785,
      "epoch": 4.805160610847815,
      "grad_norm": 0.29243433475494385,
      "learning_rate": 6.257286020888863e-05,
      "loss": 0.17095312476158142,
      "mean_token_accuracy": 0.9262748509645462,
      "num_tokens": 112561888.0,
      "step": 9125
    },
    {
      "entropy": 0.9452150911092758,
      "epoch": 4.805687203791469,
      "grad_norm": 0.30524101853370667,
      "learning_rate": 6.255422509443318e-05,
      "loss": 0.1610173135995865,
      "mean_token_accuracy": 0.9299993216991425,
      "num_tokens": 112574224.0,
      "step": 9126
    },
    {
      "entropy": 0.9425390809774399,
      "epoch": 4.806213796735124,
      "grad_norm": 0.30706319212913513,
      "learning_rate": 6.253559279655685e-05,
      "loss": 0.16968831419944763,
      "mean_token_accuracy": 0.924667090177536,
      "num_tokens": 112586560.0,
      "step": 9127
    },
    {
      "entropy": 0.919682040810585,
      "epoch": 4.806740389678779,
      "grad_norm": 0.2983159124851227,
      "learning_rate": 6.251696331636572e-05,
      "loss": 0.1572701781988144,
      "mean_token_accuracy": 0.9339618682861328,
      "num_tokens": 112598896.0,
      "step": 9128
    },
    {
      "entropy": 0.8853818327188492,
      "epoch": 4.807266982622433,
      "grad_norm": 0.3081471025943756,
      "learning_rate": 6.249833665496578e-05,
      "loss": 0.16573485732078552,
      "mean_token_accuracy": 0.9274046123027802,
      "num_tokens": 112611232.0,
      "step": 9129
    },
    {
      "entropy": 0.9164557904005051,
      "epoch": 4.807793575566087,
      "grad_norm": 0.3027510941028595,
      "learning_rate": 6.24797128134627e-05,
      "loss": 0.17423248291015625,
      "mean_token_accuracy": 0.9266597628593445,
      "num_tokens": 112623568.0,
      "step": 9130
    },
    {
      "entropy": 0.9424787163734436,
      "epoch": 4.808320168509742,
      "grad_norm": 0.32579028606414795,
      "learning_rate": 6.246109179296205e-05,
      "loss": 0.18080949783325195,
      "mean_token_accuracy": 0.922933965921402,
      "num_tokens": 112635904.0,
      "step": 9131
    },
    {
      "entropy": 0.9282560646533966,
      "epoch": 4.808846761453396,
      "grad_norm": 0.3106735944747925,
      "learning_rate": 6.244247359456935e-05,
      "loss": 0.16808384656906128,
      "mean_token_accuracy": 0.9309414029121399,
      "num_tokens": 112648240.0,
      "step": 9132
    },
    {
      "entropy": 0.94905224442482,
      "epoch": 4.809373354397051,
      "grad_norm": 0.3005760908126831,
      "learning_rate": 6.242385821938978e-05,
      "loss": 0.15837016701698303,
      "mean_token_accuracy": 0.9334435909986496,
      "num_tokens": 112660576.0,
      "step": 9133
    },
    {
      "entropy": 0.9732349216938019,
      "epoch": 4.809899947340705,
      "grad_norm": 0.3310624361038208,
      "learning_rate": 6.240524566852847e-05,
      "loss": 0.17375272512435913,
      "mean_token_accuracy": 0.9251791536808014,
      "num_tokens": 112672912.0,
      "step": 9134
    },
    {
      "entropy": 0.9393926411867142,
      "epoch": 4.81042654028436,
      "grad_norm": 0.32279565930366516,
      "learning_rate": 6.238663594309034e-05,
      "loss": 0.18074797093868256,
      "mean_token_accuracy": 0.923829197883606,
      "num_tokens": 112685248.0,
      "step": 9135
    },
    {
      "entropy": 0.8982023596763611,
      "epoch": 4.810953133228015,
      "grad_norm": 0.29356688261032104,
      "learning_rate": 6.23680290441801e-05,
      "loss": 0.16551603376865387,
      "mean_token_accuracy": 0.927782729268074,
      "num_tokens": 112697584.0,
      "step": 9136
    },
    {
      "entropy": 0.9252886474132538,
      "epoch": 4.811479726171669,
      "grad_norm": 0.3066103756427765,
      "learning_rate": 6.234942497290236e-05,
      "loss": 0.16104555130004883,
      "mean_token_accuracy": 0.931676521897316,
      "num_tokens": 112709920.0,
      "step": 9137
    },
    {
      "entropy": 0.9413758218288422,
      "epoch": 4.812006319115324,
      "grad_norm": 0.31761816143989563,
      "learning_rate": 6.233082373036153e-05,
      "loss": 0.1703198105096817,
      "mean_token_accuracy": 0.9248000532388687,
      "num_tokens": 112722256.0,
      "step": 9138
    },
    {
      "entropy": 0.941258043050766,
      "epoch": 4.812532912058979,
      "grad_norm": 0.35805585980415344,
      "learning_rate": 6.231222531766185e-05,
      "loss": 0.1860889196395874,
      "mean_token_accuracy": 0.9257703423500061,
      "num_tokens": 112734592.0,
      "step": 9139
    },
    {
      "entropy": 0.9381390064954758,
      "epoch": 4.813059505002633,
      "grad_norm": 0.2907848358154297,
      "learning_rate": 6.229362973590742e-05,
      "loss": 0.14537650346755981,
      "mean_token_accuracy": 0.9363595247268677,
      "num_tokens": 112746928.0,
      "step": 9140
    },
    {
      "entropy": 0.9168024659156799,
      "epoch": 4.8135860979462874,
      "grad_norm": 0.3004358410835266,
      "learning_rate": 6.227503698620216e-05,
      "loss": 0.15760137140750885,
      "mean_token_accuracy": 0.933418795466423,
      "num_tokens": 112759264.0,
      "step": 9141
    },
    {
      "entropy": 0.9295596182346344,
      "epoch": 4.814112690889942,
      "grad_norm": 0.3019280433654785,
      "learning_rate": 6.225644706964974e-05,
      "loss": 0.16697147488594055,
      "mean_token_accuracy": 0.929103896021843,
      "num_tokens": 112771600.0,
      "step": 9142
    },
    {
      "entropy": 0.9131826758384705,
      "epoch": 4.814639283833596,
      "grad_norm": 0.3130643665790558,
      "learning_rate": 6.223785998735382e-05,
      "loss": 0.1685619056224823,
      "mean_token_accuracy": 0.9274730831384659,
      "num_tokens": 112783936.0,
      "step": 9143
    },
    {
      "entropy": 0.923699751496315,
      "epoch": 4.815165876777251,
      "grad_norm": 0.31038033962249756,
      "learning_rate": 6.221927574041779e-05,
      "loss": 0.1619996428489685,
      "mean_token_accuracy": 0.9313331842422485,
      "num_tokens": 112796272.0,
      "step": 9144
    },
    {
      "entropy": 0.9440527558326721,
      "epoch": 4.815692469720906,
      "grad_norm": 0.3447670638561249,
      "learning_rate": 6.220069432994481e-05,
      "loss": 0.16763903200626373,
      "mean_token_accuracy": 0.9285595715045929,
      "num_tokens": 112808608.0,
      "step": 9145
    },
    {
      "entropy": 0.8979533016681671,
      "epoch": 4.81621906266456,
      "grad_norm": 0.29874899983406067,
      "learning_rate": 6.218211575703805e-05,
      "loss": 0.17092090845108032,
      "mean_token_accuracy": 0.9283223152160645,
      "num_tokens": 112820944.0,
      "step": 9146
    },
    {
      "entropy": 0.9298407733440399,
      "epoch": 4.816745655608215,
      "grad_norm": 0.3201086223125458,
      "learning_rate": 6.216354002280037e-05,
      "loss": 0.1501120924949646,
      "mean_token_accuracy": 0.9355747103691101,
      "num_tokens": 112833280.0,
      "step": 9147
    },
    {
      "entropy": 0.9380588829517365,
      "epoch": 4.81727224855187,
      "grad_norm": 0.3264695107936859,
      "learning_rate": 6.214496712833455e-05,
      "loss": 0.15935201942920685,
      "mean_token_accuracy": 0.9295739084482193,
      "num_tokens": 112845616.0,
      "step": 9148
    },
    {
      "entropy": 0.9423897117376328,
      "epoch": 4.817798841495524,
      "grad_norm": 0.36109310388565063,
      "learning_rate": 6.212639707474308e-05,
      "loss": 0.18861493468284607,
      "mean_token_accuracy": 0.916970893740654,
      "num_tokens": 112857952.0,
      "step": 9149
    },
    {
      "entropy": 0.8985049575567245,
      "epoch": 4.818325434439179,
      "grad_norm": 0.30845433473587036,
      "learning_rate": 6.210782986312838e-05,
      "loss": 0.16502808034420013,
      "mean_token_accuracy": 0.9298705011606216,
      "num_tokens": 112870288.0,
      "step": 9150
    },
    {
      "entropy": 0.9010489583015442,
      "epoch": 4.8188520273828335,
      "grad_norm": 0.34794074296951294,
      "learning_rate": 6.208926549459268e-05,
      "loss": 0.18474933505058289,
      "mean_token_accuracy": 0.921579048037529,
      "num_tokens": 112882624.0,
      "step": 9151
    },
    {
      "entropy": 0.9137450605630875,
      "epoch": 4.8193786203264875,
      "grad_norm": 0.30454927682876587,
      "learning_rate": 6.207070397023803e-05,
      "loss": 0.16628552973270416,
      "mean_token_accuracy": 0.925239771604538,
      "num_tokens": 112894960.0,
      "step": 9152
    },
    {
      "entropy": 0.9247958213090897,
      "epoch": 4.819905213270142,
      "grad_norm": 0.30902260541915894,
      "learning_rate": 6.205214529116634e-05,
      "loss": 0.16501721739768982,
      "mean_token_accuracy": 0.9277013838291168,
      "num_tokens": 112907296.0,
      "step": 9153
    },
    {
      "entropy": 0.9595179408788681,
      "epoch": 4.820431806213797,
      "grad_norm": 0.34334465861320496,
      "learning_rate": 6.203358945847935e-05,
      "loss": 0.18191400170326233,
      "mean_token_accuracy": 0.9258522987365723,
      "num_tokens": 112919632.0,
      "step": 9154
    },
    {
      "entropy": 0.9212562143802643,
      "epoch": 4.820958399157451,
      "grad_norm": 0.31868496537208557,
      "learning_rate": 6.201503647327855e-05,
      "loss": 0.17826935648918152,
      "mean_token_accuracy": 0.9231353104114532,
      "num_tokens": 112931968.0,
      "step": 9155
    },
    {
      "entropy": 0.9315204322338104,
      "epoch": 4.821484992101106,
      "grad_norm": 0.29482823610305786,
      "learning_rate": 6.199648633666538e-05,
      "loss": 0.16608521342277527,
      "mean_token_accuracy": 0.9294846951961517,
      "num_tokens": 112944304.0,
      "step": 9156
    },
    {
      "entropy": 0.9712758958339691,
      "epoch": 4.82201158504476,
      "grad_norm": 0.31313827633857727,
      "learning_rate": 6.197793904974104e-05,
      "loss": 0.17441165447235107,
      "mean_token_accuracy": 0.9243512004613876,
      "num_tokens": 112956640.0,
      "step": 9157
    },
    {
      "entropy": 0.9236172586679459,
      "epoch": 4.822538177988415,
      "grad_norm": 0.29153314232826233,
      "learning_rate": 6.195939461360654e-05,
      "loss": 0.1600339561700821,
      "mean_token_accuracy": 0.9282103329896927,
      "num_tokens": 112968976.0,
      "step": 9158
    },
    {
      "entropy": 0.9531195908784866,
      "epoch": 4.82306477093207,
      "grad_norm": 0.33459538221359253,
      "learning_rate": 6.194085302936278e-05,
      "loss": 0.18222567439079285,
      "mean_token_accuracy": 0.9229590743780136,
      "num_tokens": 112981312.0,
      "step": 9159
    },
    {
      "entropy": 0.9638189673423767,
      "epoch": 4.823591363875724,
      "grad_norm": 0.3411768674850464,
      "learning_rate": 6.192231429811048e-05,
      "loss": 0.18856140971183777,
      "mean_token_accuracy": 0.9173770844936371,
      "num_tokens": 112993648.0,
      "step": 9160
    },
    {
      "entropy": 0.9931163936853409,
      "epoch": 4.824117956819379,
      "grad_norm": 0.3286246061325073,
      "learning_rate": 6.190377842095017e-05,
      "loss": 0.18306970596313477,
      "mean_token_accuracy": 0.922531470656395,
      "num_tokens": 113005984.0,
      "step": 9161
    },
    {
      "entropy": 0.87714883685112,
      "epoch": 4.8246445497630335,
      "grad_norm": 0.2968125641345978,
      "learning_rate": 6.188524539898219e-05,
      "loss": 0.15188772976398468,
      "mean_token_accuracy": 0.9343786090612411,
      "num_tokens": 113018320.0,
      "step": 9162
    },
    {
      "entropy": 0.9333531558513641,
      "epoch": 4.8251711427066875,
      "grad_norm": 0.31941041350364685,
      "learning_rate": 6.186671523330676e-05,
      "loss": 0.17567633092403412,
      "mean_token_accuracy": 0.9214246571063995,
      "num_tokens": 113030656.0,
      "step": 9163
    },
    {
      "entropy": 0.9728813916444778,
      "epoch": 4.825697735650342,
      "grad_norm": 0.32849058508872986,
      "learning_rate": 6.184818792502391e-05,
      "loss": 0.1731397807598114,
      "mean_token_accuracy": 0.9273866266012192,
      "num_tokens": 113042992.0,
      "step": 9164
    },
    {
      "entropy": 0.9401708841323853,
      "epoch": 4.826224328593997,
      "grad_norm": 0.30741560459136963,
      "learning_rate": 6.182966347523349e-05,
      "loss": 0.17288266122341156,
      "mean_token_accuracy": 0.9284588396549225,
      "num_tokens": 113055328.0,
      "step": 9165
    },
    {
      "entropy": 0.9812659025192261,
      "epoch": 4.826750921537651,
      "grad_norm": 0.3393278121948242,
      "learning_rate": 6.181114188503516e-05,
      "loss": 0.1769835650920868,
      "mean_token_accuracy": 0.9237487763166428,
      "num_tokens": 113067664.0,
      "step": 9166
    },
    {
      "entropy": 0.9631835669279099,
      "epoch": 4.827277514481306,
      "grad_norm": 0.31807562708854675,
      "learning_rate": 6.179262315552853e-05,
      "loss": 0.17839747667312622,
      "mean_token_accuracy": 0.9289010018110275,
      "num_tokens": 113080000.0,
      "step": 9167
    },
    {
      "entropy": 0.9522756934165955,
      "epoch": 4.82780410742496,
      "grad_norm": 0.3397252559661865,
      "learning_rate": 6.177410728781287e-05,
      "loss": 0.1738055795431137,
      "mean_token_accuracy": 0.9253265112638474,
      "num_tokens": 113092336.0,
      "step": 9168
    },
    {
      "entropy": 0.9829897880554199,
      "epoch": 4.828330700368615,
      "grad_norm": 0.3597947955131531,
      "learning_rate": 6.175559428298737e-05,
      "loss": 0.18888182938098907,
      "mean_token_accuracy": 0.9227168262004852,
      "num_tokens": 113104672.0,
      "step": 9169
    },
    {
      "entropy": 0.9550391584634781,
      "epoch": 4.82885729331227,
      "grad_norm": 0.2912638783454895,
      "learning_rate": 6.173708414215105e-05,
      "loss": 0.1649709939956665,
      "mean_token_accuracy": 0.9291027784347534,
      "num_tokens": 113117008.0,
      "step": 9170
    },
    {
      "entropy": 0.8955940455198288,
      "epoch": 4.829383886255924,
      "grad_norm": 0.28251731395721436,
      "learning_rate": 6.171857686640274e-05,
      "loss": 0.14913088083267212,
      "mean_token_accuracy": 0.9360662996768951,
      "num_tokens": 113129344.0,
      "step": 9171
    },
    {
      "entropy": 0.9660614132881165,
      "epoch": 4.829910479199579,
      "grad_norm": 0.3158622682094574,
      "learning_rate": 6.170007245684113e-05,
      "loss": 0.16393935680389404,
      "mean_token_accuracy": 0.9336477220058441,
      "num_tokens": 113141680.0,
      "step": 9172
    },
    {
      "entropy": 0.9383984357118607,
      "epoch": 4.8304370721432335,
      "grad_norm": 0.32235682010650635,
      "learning_rate": 6.168157091456466e-05,
      "loss": 0.17803451418876648,
      "mean_token_accuracy": 0.9236116111278534,
      "num_tokens": 113154016.0,
      "step": 9173
    },
    {
      "entropy": 0.9070072770118713,
      "epoch": 4.8309636650868875,
      "grad_norm": 0.29480841755867004,
      "learning_rate": 6.166307224067173e-05,
      "loss": 0.164914071559906,
      "mean_token_accuracy": 0.9274289160966873,
      "num_tokens": 113166352.0,
      "step": 9174
    },
    {
      "entropy": 0.945151075720787,
      "epoch": 4.831490258030542,
      "grad_norm": 0.3259155750274658,
      "learning_rate": 6.16445764362605e-05,
      "loss": 0.17770960927009583,
      "mean_token_accuracy": 0.9249714463949203,
      "num_tokens": 113178688.0,
      "step": 9175
    },
    {
      "entropy": 0.9563282877206802,
      "epoch": 4.832016850974197,
      "grad_norm": 0.31394654512405396,
      "learning_rate": 6.162608350242888e-05,
      "loss": 0.16683940589427948,
      "mean_token_accuracy": 0.9329259991645813,
      "num_tokens": 113191024.0,
      "step": 9176
    },
    {
      "entropy": 0.9816624969244003,
      "epoch": 4.832543443917851,
      "grad_norm": 0.34521687030792236,
      "learning_rate": 6.160759344027482e-05,
      "loss": 0.1855495572090149,
      "mean_token_accuracy": 0.9137884080410004,
      "num_tokens": 113203360.0,
      "step": 9177
    },
    {
      "entropy": 0.9625774472951889,
      "epoch": 4.833070036861506,
      "grad_norm": 0.33744141459465027,
      "learning_rate": 6.158910625089584e-05,
      "loss": 0.1721527874469757,
      "mean_token_accuracy": 0.9258444309234619,
      "num_tokens": 113215696.0,
      "step": 9178
    },
    {
      "entropy": 0.9747310727834702,
      "epoch": 4.833596629805161,
      "grad_norm": 0.3287394642829895,
      "learning_rate": 6.157062193538945e-05,
      "loss": 0.1844010353088379,
      "mean_token_accuracy": 0.9236720949411392,
      "num_tokens": 113228032.0,
      "step": 9179
    },
    {
      "entropy": 0.9509319812059402,
      "epoch": 4.834123222748815,
      "grad_norm": 0.3232387900352478,
      "learning_rate": 6.155214049485294e-05,
      "loss": 0.1758861392736435,
      "mean_token_accuracy": 0.9263638705015182,
      "num_tokens": 113240368.0,
      "step": 9180
    },
    {
      "entropy": 0.9433962553739548,
      "epoch": 4.83464981569247,
      "grad_norm": 0.32960569858551025,
      "learning_rate": 6.15336619303835e-05,
      "loss": 0.17984148859977722,
      "mean_token_accuracy": 0.9253224283456802,
      "num_tokens": 113252704.0,
      "step": 9181
    },
    {
      "entropy": 0.9309116154909134,
      "epoch": 4.835176408636125,
      "grad_norm": 0.31888312101364136,
      "learning_rate": 6.151518624307808e-05,
      "loss": 0.1658877581357956,
      "mean_token_accuracy": 0.9319590628147125,
      "num_tokens": 113265040.0,
      "step": 9182
    },
    {
      "entropy": 0.9187856763601303,
      "epoch": 4.835703001579779,
      "grad_norm": 0.3147662878036499,
      "learning_rate": 6.149671343403347e-05,
      "loss": 0.17353439331054688,
      "mean_token_accuracy": 0.9259817600250244,
      "num_tokens": 113277376.0,
      "step": 9183
    },
    {
      "entropy": 0.9754699617624283,
      "epoch": 4.8362295945234335,
      "grad_norm": 0.32784178853034973,
      "learning_rate": 6.147824350434628e-05,
      "loss": 0.1768968254327774,
      "mean_token_accuracy": 0.9237811714410782,
      "num_tokens": 113289712.0,
      "step": 9184
    },
    {
      "entropy": 0.9289213418960571,
      "epoch": 4.836756187467088,
      "grad_norm": 0.30313992500305176,
      "learning_rate": 6.145977645511296e-05,
      "loss": 0.1519276350736618,
      "mean_token_accuracy": 0.9341972172260284,
      "num_tokens": 113302048.0,
      "step": 9185
    },
    {
      "entropy": 0.9625039547681808,
      "epoch": 4.837282780410742,
      "grad_norm": 0.3229649066925049,
      "learning_rate": 6.144131228742979e-05,
      "loss": 0.17990517616271973,
      "mean_token_accuracy": 0.9261969178915024,
      "num_tokens": 113314384.0,
      "step": 9186
    },
    {
      "entropy": 0.9287969917058945,
      "epoch": 4.837809373354397,
      "grad_norm": 0.3233555853366852,
      "learning_rate": 6.142285100239289e-05,
      "loss": 0.1754463016986847,
      "mean_token_accuracy": 0.9251480847597122,
      "num_tokens": 113326720.0,
      "step": 9187
    },
    {
      "entropy": 0.9506050199270248,
      "epoch": 4.838335966298052,
      "grad_norm": 0.3280300796031952,
      "learning_rate": 6.140439260109821e-05,
      "loss": 0.17056047916412354,
      "mean_token_accuracy": 0.9295740723609924,
      "num_tokens": 113339056.0,
      "step": 9188
    },
    {
      "entropy": 0.9726871848106384,
      "epoch": 4.838862559241706,
      "grad_norm": 0.3280206024646759,
      "learning_rate": 6.138593708464149e-05,
      "loss": 0.1634208709001541,
      "mean_token_accuracy": 0.932182252407074,
      "num_tokens": 113351392.0,
      "step": 9189
    },
    {
      "entropy": 0.949202686548233,
      "epoch": 4.839389152185361,
      "grad_norm": 0.31370067596435547,
      "learning_rate": 6.136748445411835e-05,
      "loss": 0.17531146109104156,
      "mean_token_accuracy": 0.9205900281667709,
      "num_tokens": 113363728.0,
      "step": 9190
    },
    {
      "entropy": 0.9522519260644913,
      "epoch": 4.839915745129015,
      "grad_norm": 0.31930145621299744,
      "learning_rate": 6.134903471062421e-05,
      "loss": 0.16543881595134735,
      "mean_token_accuracy": 0.9292179048061371,
      "num_tokens": 113376064.0,
      "step": 9191
    },
    {
      "entropy": 0.949477881193161,
      "epoch": 4.84044233807267,
      "grad_norm": 0.2924252450466156,
      "learning_rate": 6.133058785525431e-05,
      "loss": 0.1596398502588272,
      "mean_token_accuracy": 0.9307438880205154,
      "num_tokens": 113388400.0,
      "step": 9192
    },
    {
      "entropy": 0.9458207190036774,
      "epoch": 4.840968931016325,
      "grad_norm": 0.3245888650417328,
      "learning_rate": 6.131214388910374e-05,
      "loss": 0.17272739112377167,
      "mean_token_accuracy": 0.929176926612854,
      "num_tokens": 113400736.0,
      "step": 9193
    },
    {
      "entropy": 0.9470962882041931,
      "epoch": 4.841495523959979,
      "grad_norm": 0.334036648273468,
      "learning_rate": 6.129370281326738e-05,
      "loss": 0.18349584937095642,
      "mean_token_accuracy": 0.9212116599082947,
      "num_tokens": 113413072.0,
      "step": 9194
    },
    {
      "entropy": 0.963790088891983,
      "epoch": 4.8420221169036335,
      "grad_norm": 0.305763304233551,
      "learning_rate": 6.127526462884005e-05,
      "loss": 0.17330719530582428,
      "mean_token_accuracy": 0.9286128431558609,
      "num_tokens": 113425408.0,
      "step": 9195
    },
    {
      "entropy": 0.9783694744110107,
      "epoch": 4.842548709847288,
      "grad_norm": 0.3513582944869995,
      "learning_rate": 6.125682933691629e-05,
      "loss": 0.18046358227729797,
      "mean_token_accuracy": 0.9242364913225174,
      "num_tokens": 113437744.0,
      "step": 9196
    },
    {
      "entropy": 0.9755796492099762,
      "epoch": 4.843075302790942,
      "grad_norm": 0.32847991585731506,
      "learning_rate": 6.123839693859045e-05,
      "loss": 0.16616854071617126,
      "mean_token_accuracy": 0.9285303205251694,
      "num_tokens": 113450080.0,
      "step": 9197
    },
    {
      "entropy": 0.9715100824832916,
      "epoch": 4.843601895734597,
      "grad_norm": 0.32819437980651855,
      "learning_rate": 6.121996743495678e-05,
      "loss": 0.18121594190597534,
      "mean_token_accuracy": 0.9218470752239227,
      "num_tokens": 113462416.0,
      "step": 9198
    },
    {
      "entropy": 0.9456192255020142,
      "epoch": 4.844128488678252,
      "grad_norm": 0.3316733241081238,
      "learning_rate": 6.120154082710932e-05,
      "loss": 0.17713892459869385,
      "mean_token_accuracy": 0.9244439750909805,
      "num_tokens": 113474752.0,
      "step": 9199
    },
    {
      "entropy": 0.9350262731313705,
      "epoch": 4.844655081621906,
      "grad_norm": 0.3105098605155945,
      "learning_rate": 6.118311711614198e-05,
      "loss": 0.171644926071167,
      "mean_token_accuracy": 0.9325538128614426,
      "num_tokens": 113487088.0,
      "step": 9200
    },
    {
      "entropy": 0.9254423081874847,
      "epoch": 4.845181674565561,
      "grad_norm": 0.3185507655143738,
      "learning_rate": 6.11646963031484e-05,
      "loss": 0.15326949954032898,
      "mean_token_accuracy": 0.935181513428688,
      "num_tokens": 113499424.0,
      "step": 9201
    },
    {
      "entropy": 0.9758787006139755,
      "epoch": 4.845708267509215,
      "grad_norm": 0.34665438532829285,
      "learning_rate": 6.11462783892222e-05,
      "loss": 0.1998121738433838,
      "mean_token_accuracy": 0.9098737090826035,
      "num_tokens": 113511760.0,
      "step": 9202
    },
    {
      "entropy": 0.9731664806604385,
      "epoch": 4.84623486045287,
      "grad_norm": 0.30735206604003906,
      "learning_rate": 6.112786337545671e-05,
      "loss": 0.16850310564041138,
      "mean_token_accuracy": 0.9308719336986542,
      "num_tokens": 113524096.0,
      "step": 9203
    },
    {
      "entropy": 0.9057483524084091,
      "epoch": 4.846761453396525,
      "grad_norm": 0.2996695935726166,
      "learning_rate": 6.110945126294514e-05,
      "loss": 0.16708335280418396,
      "mean_token_accuracy": 0.929191917181015,
      "num_tokens": 113536432.0,
      "step": 9204
    },
    {
      "entropy": 0.9213566184043884,
      "epoch": 4.847288046340179,
      "grad_norm": 0.2910383641719818,
      "learning_rate": 6.109104205278049e-05,
      "loss": 0.1677648276090622,
      "mean_token_accuracy": 0.93411485850811,
      "num_tokens": 113548768.0,
      "step": 9205
    },
    {
      "entropy": 0.8875181376934052,
      "epoch": 4.8478146392838335,
      "grad_norm": 0.3171277642250061,
      "learning_rate": 6.10726357460556e-05,
      "loss": 0.17253826558589935,
      "mean_token_accuracy": 0.9264390170574188,
      "num_tokens": 113561104.0,
      "step": 9206
    },
    {
      "entropy": 0.9762510508298874,
      "epoch": 4.848341232227488,
      "grad_norm": 0.33811476826667786,
      "learning_rate": 6.10542323438632e-05,
      "loss": 0.17586681246757507,
      "mean_token_accuracy": 0.9236846566200256,
      "num_tokens": 113573440.0,
      "step": 9207
    },
    {
      "entropy": 0.978374108672142,
      "epoch": 4.848867825171142,
      "grad_norm": 0.31238728761672974,
      "learning_rate": 6.103583184729566e-05,
      "loss": 0.16370560228824615,
      "mean_token_accuracy": 0.9349195510149002,
      "num_tokens": 113585776.0,
      "step": 9208
    },
    {
      "entropy": 0.9273306876420975,
      "epoch": 4.849394418114797,
      "grad_norm": 0.30583444237709045,
      "learning_rate": 6.101743425744546e-05,
      "loss": 0.15903574228286743,
      "mean_token_accuracy": 0.9286919683218002,
      "num_tokens": 113598112.0,
      "step": 9209
    },
    {
      "entropy": 0.9907515197992325,
      "epoch": 4.849921011058452,
      "grad_norm": 0.3668973743915558,
      "learning_rate": 6.099903957540467e-05,
      "loss": 0.17592556774616241,
      "mean_token_accuracy": 0.9220422655344009,
      "num_tokens": 113610448.0,
      "step": 9210
    },
    {
      "entropy": 0.9322537630796432,
      "epoch": 4.850447604002106,
      "grad_norm": 0.33347219228744507,
      "learning_rate": 6.098064780226531e-05,
      "loss": 0.17654946446418762,
      "mean_token_accuracy": 0.9239524006843567,
      "num_tokens": 113622784.0,
      "step": 9211
    },
    {
      "entropy": 0.9455699473619461,
      "epoch": 4.850974196945761,
      "grad_norm": 0.290907621383667,
      "learning_rate": 6.09622589391192e-05,
      "loss": 0.16271138191223145,
      "mean_token_accuracy": 0.9278527349233627,
      "num_tokens": 113635120.0,
      "step": 9212
    },
    {
      "entropy": 0.912358745932579,
      "epoch": 4.851500789889416,
      "grad_norm": 0.28543636202812195,
      "learning_rate": 6.094387298705795e-05,
      "loss": 0.16902172565460205,
      "mean_token_accuracy": 0.9293874204158783,
      "num_tokens": 113647456.0,
      "step": 9213
    },
    {
      "entropy": 0.9107921272516251,
      "epoch": 4.85202738283307,
      "grad_norm": 0.28984534740448,
      "learning_rate": 6.092548994717303e-05,
      "loss": 0.16308516263961792,
      "mean_token_accuracy": 0.9295253306627274,
      "num_tokens": 113659792.0,
      "step": 9214
    },
    {
      "entropy": 0.9483311921358109,
      "epoch": 4.852553975776725,
      "grad_norm": 0.3155252933502197,
      "learning_rate": 6.0907109820555764e-05,
      "loss": 0.16149048507213593,
      "mean_token_accuracy": 0.9389007091522217,
      "num_tokens": 113672128.0,
      "step": 9215
    },
    {
      "entropy": 0.9115438163280487,
      "epoch": 4.8530805687203795,
      "grad_norm": 0.3140206038951874,
      "learning_rate": 6.088873260829724e-05,
      "loss": 0.16763517260551453,
      "mean_token_accuracy": 0.9338799566030502,
      "num_tokens": 113684464.0,
      "step": 9216
    },
    {
      "entropy": 0.9212208986282349,
      "epoch": 4.8536071616640335,
      "grad_norm": 0.3106561601161957,
      "learning_rate": 6.0870358311488415e-05,
      "loss": 0.16700613498687744,
      "mean_token_accuracy": 0.9301815181970596,
      "num_tokens": 113696800.0,
      "step": 9217
    },
    {
      "entropy": 0.9526770710945129,
      "epoch": 4.854133754607688,
      "grad_norm": 0.34409135580062866,
      "learning_rate": 6.0851986931220075e-05,
      "loss": 0.20234370231628418,
      "mean_token_accuracy": 0.9196347147226334,
      "num_tokens": 113709136.0,
      "step": 9218
    },
    {
      "entropy": 0.9145548343658447,
      "epoch": 4.854660347551343,
      "grad_norm": 0.3165554106235504,
      "learning_rate": 6.083361846858281e-05,
      "loss": 0.16674527525901794,
      "mean_token_accuracy": 0.9236103445291519,
      "num_tokens": 113721472.0,
      "step": 9219
    },
    {
      "entropy": 0.9819770008325577,
      "epoch": 4.855186940494997,
      "grad_norm": 0.3707553446292877,
      "learning_rate": 6.081525292466706e-05,
      "loss": 0.19176122546195984,
      "mean_token_accuracy": 0.9174077808856964,
      "num_tokens": 113733808.0,
      "step": 9220
    },
    {
      "entropy": 0.997324988245964,
      "epoch": 4.855713533438652,
      "grad_norm": 0.3498297333717346,
      "learning_rate": 6.079689030056308e-05,
      "loss": 0.19261667132377625,
      "mean_token_accuracy": 0.9157433807849884,
      "num_tokens": 113746144.0,
      "step": 9221
    },
    {
      "entropy": 0.944717675447464,
      "epoch": 4.856240126382307,
      "grad_norm": 0.2912974953651428,
      "learning_rate": 6.0778530597360895e-05,
      "loss": 0.1591337025165558,
      "mean_token_accuracy": 0.9307748228311539,
      "num_tokens": 113758480.0,
      "step": 9222
    },
    {
      "entropy": 0.9538642019033432,
      "epoch": 4.856766719325961,
      "grad_norm": 0.3185594379901886,
      "learning_rate": 6.07601738161505e-05,
      "loss": 0.1545395702123642,
      "mean_token_accuracy": 0.9348485469818115,
      "num_tokens": 113770816.0,
      "step": 9223
    },
    {
      "entropy": 0.9914166331291199,
      "epoch": 4.857293312269616,
      "grad_norm": 0.3420868515968323,
      "learning_rate": 6.074181995802159e-05,
      "loss": 0.18674421310424805,
      "mean_token_accuracy": 0.9197226315736771,
      "num_tokens": 113783152.0,
      "step": 9224
    },
    {
      "entropy": 0.9152809083461761,
      "epoch": 4.85781990521327,
      "grad_norm": 0.30477678775787354,
      "learning_rate": 6.072346902406375e-05,
      "loss": 0.1771145910024643,
      "mean_token_accuracy": 0.9249708354473114,
      "num_tokens": 113795488.0,
      "step": 9225
    },
    {
      "entropy": 0.9725821763277054,
      "epoch": 4.858346498156925,
      "grad_norm": 0.32783952355384827,
      "learning_rate": 6.070512101536637e-05,
      "loss": 0.16451381146907806,
      "mean_token_accuracy": 0.9245587736368179,
      "num_tokens": 113807824.0,
      "step": 9226
    },
    {
      "entropy": 0.9453549236059189,
      "epoch": 4.8588730911005795,
      "grad_norm": 0.3178243041038513,
      "learning_rate": 6.068677593301861e-05,
      "loss": 0.16712918877601624,
      "mean_token_accuracy": 0.9319606870412827,
      "num_tokens": 113820160.0,
      "step": 9227
    },
    {
      "entropy": 0.9540058225393295,
      "epoch": 4.8593996840442335,
      "grad_norm": 0.316835880279541,
      "learning_rate": 6.0668433778109545e-05,
      "loss": 0.16519233584403992,
      "mean_token_accuracy": 0.9322793334722519,
      "num_tokens": 113832496.0,
      "step": 9228
    },
    {
      "entropy": 1.006719246506691,
      "epoch": 4.859926276987888,
      "grad_norm": 0.358338326215744,
      "learning_rate": 6.065009455172801e-05,
      "loss": 0.17126135528087616,
      "mean_token_accuracy": 0.9240893870592117,
      "num_tokens": 113844832.0,
      "step": 9229
    },
    {
      "entropy": 1.0083835273981094,
      "epoch": 4.860452869931543,
      "grad_norm": 0.2907813489437103,
      "learning_rate": 6.0631758254962754e-05,
      "loss": 0.15603706240653992,
      "mean_token_accuracy": 0.9350569695234299,
      "num_tokens": 113857168.0,
      "step": 9230
    },
    {
      "entropy": 0.9859584420919418,
      "epoch": 4.860979462875197,
      "grad_norm": 0.3402427136898041,
      "learning_rate": 6.061342488890227e-05,
      "loss": 0.17818543314933777,
      "mean_token_accuracy": 0.9255679994821548,
      "num_tokens": 113869504.0,
      "step": 9231
    },
    {
      "entropy": 1.0371097922325134,
      "epoch": 4.861506055818852,
      "grad_norm": 0.32408174872398376,
      "learning_rate": 6.0595094454634915e-05,
      "loss": 0.16809692978858948,
      "mean_token_accuracy": 0.922844797372818,
      "num_tokens": 113881840.0,
      "step": 9232
    },
    {
      "entropy": 0.9761839061975479,
      "epoch": 4.862032648762507,
      "grad_norm": 0.3168609142303467,
      "learning_rate": 6.057676695324883e-05,
      "loss": 0.16999582946300507,
      "mean_token_accuracy": 0.9267615675926208,
      "num_tokens": 113894176.0,
      "step": 9233
    },
    {
      "entropy": 0.9961191415786743,
      "epoch": 4.862559241706161,
      "grad_norm": 0.31248804926872253,
      "learning_rate": 6.0558442385832047e-05,
      "loss": 0.16822709143161774,
      "mean_token_accuracy": 0.9302322715520859,
      "num_tokens": 113906512.0,
      "step": 9234
    },
    {
      "entropy": 0.9907224774360657,
      "epoch": 4.863085834649816,
      "grad_norm": 0.32921966910362244,
      "learning_rate": 6.054012075347237e-05,
      "loss": 0.1776476502418518,
      "mean_token_accuracy": 0.9271745085716248,
      "num_tokens": 113918848.0,
      "step": 9235
    },
    {
      "entropy": 0.9873936325311661,
      "epoch": 4.86361242759347,
      "grad_norm": 0.3174605071544647,
      "learning_rate": 6.052180205725745e-05,
      "loss": 0.1837022453546524,
      "mean_token_accuracy": 0.9242022931575775,
      "num_tokens": 113931184.0,
      "step": 9236
    },
    {
      "entropy": 0.9924033582210541,
      "epoch": 4.864139020537125,
      "grad_norm": 0.33693379163742065,
      "learning_rate": 6.050348629827476e-05,
      "loss": 0.18274997174739838,
      "mean_token_accuracy": 0.9257642328739166,
      "num_tokens": 113943520.0,
      "step": 9237
    },
    {
      "entropy": 0.9944822788238525,
      "epoch": 4.8646656134807795,
      "grad_norm": 0.32839271426200867,
      "learning_rate": 6.0485173477611604e-05,
      "loss": 0.17032325267791748,
      "mean_token_accuracy": 0.9244238436222076,
      "num_tokens": 113955856.0,
      "step": 9238
    },
    {
      "entropy": 0.9427923858165741,
      "epoch": 4.8651922064244335,
      "grad_norm": 0.2775003910064697,
      "learning_rate": 6.046686359635509e-05,
      "loss": 0.15462589263916016,
      "mean_token_accuracy": 0.9338415265083313,
      "num_tokens": 113968192.0,
      "step": 9239
    },
    {
      "entropy": 0.981944665312767,
      "epoch": 4.865718799368088,
      "grad_norm": 0.3248477280139923,
      "learning_rate": 6.0448556655592194e-05,
      "loss": 0.17406399548053741,
      "mean_token_accuracy": 0.9296783953905106,
      "num_tokens": 113980528.0,
      "step": 9240
    },
    {
      "entropy": 0.9421994984149933,
      "epoch": 4.866245392311743,
      "grad_norm": 0.33592692017555237,
      "learning_rate": 6.043025265640969e-05,
      "loss": 0.17528413236141205,
      "mean_token_accuracy": 0.9240370690822601,
      "num_tokens": 113992864.0,
      "step": 9241
    },
    {
      "entropy": 0.9783306866884232,
      "epoch": 4.866771985255397,
      "grad_norm": 0.31253552436828613,
      "learning_rate": 6.041195159989415e-05,
      "loss": 0.17901182174682617,
      "mean_token_accuracy": 0.9209093749523163,
      "num_tokens": 114005200.0,
      "step": 9242
    },
    {
      "entropy": 0.9683476388454437,
      "epoch": 4.867298578199052,
      "grad_norm": 0.3116768002510071,
      "learning_rate": 6.0393653487132004e-05,
      "loss": 0.17576105892658234,
      "mean_token_accuracy": 0.9291523545980453,
      "num_tokens": 114017536.0,
      "step": 9243
    },
    {
      "entropy": 0.9654994159936905,
      "epoch": 4.867825171142707,
      "grad_norm": 0.3077673316001892,
      "learning_rate": 6.037535831920955e-05,
      "loss": 0.17306140065193176,
      "mean_token_accuracy": 0.9240833222866058,
      "num_tokens": 114029872.0,
      "step": 9244
    },
    {
      "entropy": 0.9287323951721191,
      "epoch": 4.868351764086361,
      "grad_norm": 0.30974891781806946,
      "learning_rate": 6.035706609721288e-05,
      "loss": 0.1807565838098526,
      "mean_token_accuracy": 0.9256379753351212,
      "num_tokens": 114042208.0,
      "step": 9245
    },
    {
      "entropy": 0.986062690615654,
      "epoch": 4.868878357030016,
      "grad_norm": 0.2962946891784668,
      "learning_rate": 6.033877682222782e-05,
      "loss": 0.17127250134944916,
      "mean_token_accuracy": 0.9312177151441574,
      "num_tokens": 114054544.0,
      "step": 9246
    },
    {
      "entropy": 0.9835542589426041,
      "epoch": 4.869404949973671,
      "grad_norm": 0.30986228585243225,
      "learning_rate": 6.032049049534012e-05,
      "loss": 0.16538189351558685,
      "mean_token_accuracy": 0.9298783987760544,
      "num_tokens": 114066880.0,
      "step": 9247
    },
    {
      "entropy": 0.9464733898639679,
      "epoch": 4.869931542917325,
      "grad_norm": 0.3071298599243164,
      "learning_rate": 6.030220711763536e-05,
      "loss": 0.16907981038093567,
      "mean_token_accuracy": 0.9315258413553238,
      "num_tokens": 114079216.0,
      "step": 9248
    },
    {
      "entropy": 0.9497257322072983,
      "epoch": 4.8704581358609795,
      "grad_norm": 0.31876394152641296,
      "learning_rate": 6.0283926690198887e-05,
      "loss": 0.16890285909175873,
      "mean_token_accuracy": 0.9297789037227631,
      "num_tokens": 114091552.0,
      "step": 9249
    },
    {
      "entropy": 0.9604268223047256,
      "epoch": 4.870984728804634,
      "grad_norm": 0.3041291832923889,
      "learning_rate": 6.0265649214115904e-05,
      "loss": 0.16234137117862701,
      "mean_token_accuracy": 0.9319205433130264,
      "num_tokens": 114103888.0,
      "step": 9250
    },
    {
      "entropy": 0.9527145028114319,
      "epoch": 4.871511321748288,
      "grad_norm": 0.2935594916343689,
      "learning_rate": 6.024737469047147e-05,
      "loss": 0.15775327384471893,
      "mean_token_accuracy": 0.9312833100557327,
      "num_tokens": 114116224.0,
      "step": 9251
    },
    {
      "entropy": 0.9280911386013031,
      "epoch": 4.872037914691943,
      "grad_norm": 0.29950031638145447,
      "learning_rate": 6.0229103120350436e-05,
      "loss": 0.16385899484157562,
      "mean_token_accuracy": 0.9287741333246231,
      "num_tokens": 114128560.0,
      "step": 9252
    },
    {
      "entropy": 0.9573824107646942,
      "epoch": 4.872564507635598,
      "grad_norm": 0.3072300851345062,
      "learning_rate": 6.021083450483745e-05,
      "loss": 0.16058257222175598,
      "mean_token_accuracy": 0.9300986230373383,
      "num_tokens": 114140896.0,
      "step": 9253
    },
    {
      "entropy": 0.9535968005657196,
      "epoch": 4.873091100579252,
      "grad_norm": 0.3601013720035553,
      "learning_rate": 6.0192568845017025e-05,
      "loss": 0.19131764769554138,
      "mean_token_accuracy": 0.9160570949316025,
      "num_tokens": 114153232.0,
      "step": 9254
    },
    {
      "entropy": 0.9700489640235901,
      "epoch": 4.873617693522907,
      "grad_norm": 0.3587206304073334,
      "learning_rate": 6.0174306141973504e-05,
      "loss": 0.18195636570453644,
      "mean_token_accuracy": 0.9237683862447739,
      "num_tokens": 114165568.0,
      "step": 9255
    },
    {
      "entropy": 0.9935056418180466,
      "epoch": 4.874144286466562,
      "grad_norm": 0.31666064262390137,
      "learning_rate": 6.015604639679104e-05,
      "loss": 0.16849590837955475,
      "mean_token_accuracy": 0.9301074743270874,
      "num_tokens": 114177904.0,
      "step": 9256
    },
    {
      "entropy": 0.9813852608203888,
      "epoch": 4.874670879410216,
      "grad_norm": 0.34814703464508057,
      "learning_rate": 6.013778961055352e-05,
      "loss": 0.1790965497493744,
      "mean_token_accuracy": 0.926289826631546,
      "num_tokens": 114190240.0,
      "step": 9257
    },
    {
      "entropy": 0.9306141138076782,
      "epoch": 4.875197472353871,
      "grad_norm": 0.3174811899662018,
      "learning_rate": 6.011953578434486e-05,
      "loss": 0.1767577826976776,
      "mean_token_accuracy": 0.9272280782461166,
      "num_tokens": 114202576.0,
      "step": 9258
    },
    {
      "entropy": 0.932680755853653,
      "epoch": 4.875724065297525,
      "grad_norm": 0.30817291140556335,
      "learning_rate": 6.010128491924862e-05,
      "loss": 0.15768198668956757,
      "mean_token_accuracy": 0.9350852966308594,
      "num_tokens": 114214912.0,
      "step": 9259
    },
    {
      "entropy": 0.9436093419790268,
      "epoch": 4.8762506582411795,
      "grad_norm": 0.44346994161605835,
      "learning_rate": 6.008303701634827e-05,
      "loss": 0.19806651771068573,
      "mean_token_accuracy": 0.9160618036985397,
      "num_tokens": 114227248.0,
      "step": 9260
    },
    {
      "entropy": 0.9407719373703003,
      "epoch": 4.876777251184834,
      "grad_norm": 0.3240104615688324,
      "learning_rate": 6.006479207672708e-05,
      "loss": 0.17336773872375488,
      "mean_token_accuracy": 0.9298413693904877,
      "num_tokens": 114239584.0,
      "step": 9261
    },
    {
      "entropy": 0.9401187151670456,
      "epoch": 4.877303844128488,
      "grad_norm": 0.3415816128253937,
      "learning_rate": 6.0046550101468135e-05,
      "loss": 0.17950160801410675,
      "mean_token_accuracy": 0.925504520535469,
      "num_tokens": 114251920.0,
      "step": 9262
    },
    {
      "entropy": 0.9379282146692276,
      "epoch": 4.877830437072143,
      "grad_norm": 0.3367367684841156,
      "learning_rate": 6.002831109165435e-05,
      "loss": 0.17652186751365662,
      "mean_token_accuracy": 0.9250429421663284,
      "num_tokens": 114264256.0,
      "step": 9263
    },
    {
      "entropy": 0.9016122072935104,
      "epoch": 4.878357030015798,
      "grad_norm": 0.2723939120769501,
      "learning_rate": 6.001007504836848e-05,
      "loss": 0.13771790266036987,
      "mean_token_accuracy": 0.9420016556978226,
      "num_tokens": 114276592.0,
      "step": 9264
    },
    {
      "entropy": 0.916037529706955,
      "epoch": 4.878883622959452,
      "grad_norm": 0.32042205333709717,
      "learning_rate": 5.9991841972693107e-05,
      "loss": 0.17149311304092407,
      "mean_token_accuracy": 0.9224643856287003,
      "num_tokens": 114288928.0,
      "step": 9265
    },
    {
      "entropy": 0.9284915775060654,
      "epoch": 4.879410215903107,
      "grad_norm": 0.3282214403152466,
      "learning_rate": 5.99736118657106e-05,
      "loss": 0.17957088351249695,
      "mean_token_accuracy": 0.9195956438779831,
      "num_tokens": 114301264.0,
      "step": 9266
    },
    {
      "entropy": 0.9315703809261322,
      "epoch": 4.879936808846761,
      "grad_norm": 0.3315577507019043,
      "learning_rate": 5.995538472850317e-05,
      "loss": 0.17412766814231873,
      "mean_token_accuracy": 0.9253064095973969,
      "num_tokens": 114313600.0,
      "step": 9267
    },
    {
      "entropy": 0.915994867682457,
      "epoch": 4.880463401790416,
      "grad_norm": 0.30813804268836975,
      "learning_rate": 5.993716056215288e-05,
      "loss": 0.16539320349693298,
      "mean_token_accuracy": 0.9329143762588501,
      "num_tokens": 114325936.0,
      "step": 9268
    },
    {
      "entropy": 0.9077543020248413,
      "epoch": 4.880989994734071,
      "grad_norm": 0.30946749448776245,
      "learning_rate": 5.9918939367741577e-05,
      "loss": 0.16349415481090546,
      "mean_token_accuracy": 0.9320286363363266,
      "num_tokens": 114338272.0,
      "step": 9269
    },
    {
      "entropy": 0.9378580302000046,
      "epoch": 4.881516587677725,
      "grad_norm": 0.3118128478527069,
      "learning_rate": 5.990072114635095e-05,
      "loss": 0.1628214418888092,
      "mean_token_accuracy": 0.92903071641922,
      "num_tokens": 114350608.0,
      "step": 9270
    },
    {
      "entropy": 0.915115088224411,
      "epoch": 4.8820431806213795,
      "grad_norm": 0.3057735562324524,
      "learning_rate": 5.9882505899062456e-05,
      "loss": 0.14752550423145294,
      "mean_token_accuracy": 0.9344717413187027,
      "num_tokens": 114362944.0,
      "step": 9271
    },
    {
      "entropy": 0.9090307652950287,
      "epoch": 4.882569773565034,
      "grad_norm": 0.30963030457496643,
      "learning_rate": 5.9864293626957534e-05,
      "loss": 0.15232141315937042,
      "mean_token_accuracy": 0.9346283227205276,
      "num_tokens": 114375280.0,
      "step": 9272
    },
    {
      "entropy": 0.9352599084377289,
      "epoch": 4.883096366508688,
      "grad_norm": 0.3679341673851013,
      "learning_rate": 5.984608433111728e-05,
      "loss": 0.17839781939983368,
      "mean_token_accuracy": 0.9244855046272278,
      "num_tokens": 114387616.0,
      "step": 9273
    },
    {
      "entropy": 0.8894796818494797,
      "epoch": 4.883622959452343,
      "grad_norm": 0.2971315383911133,
      "learning_rate": 5.982787801262267e-05,
      "loss": 0.16099128127098083,
      "mean_token_accuracy": 0.9308123588562012,
      "num_tokens": 114399952.0,
      "step": 9274
    },
    {
      "entropy": 0.915225937962532,
      "epoch": 4.884149552395998,
      "grad_norm": 0.32512086629867554,
      "learning_rate": 5.9809674672554575e-05,
      "loss": 0.16589084267616272,
      "mean_token_accuracy": 0.9319024235010147,
      "num_tokens": 114412288.0,
      "step": 9275
    },
    {
      "entropy": 0.8668423146009445,
      "epoch": 4.884676145339652,
      "grad_norm": 0.3217252790927887,
      "learning_rate": 5.9791474311993514e-05,
      "loss": 0.16443023085594177,
      "mean_token_accuracy": 0.9309308081865311,
      "num_tokens": 114424624.0,
      "step": 9276
    },
    {
      "entropy": 0.906637892127037,
      "epoch": 4.885202738283307,
      "grad_norm": 0.35782021284103394,
      "learning_rate": 5.977327693201997e-05,
      "loss": 0.1807124763727188,
      "mean_token_accuracy": 0.9191261976957321,
      "num_tokens": 114436960.0,
      "step": 9277
    },
    {
      "entropy": 0.907283827662468,
      "epoch": 4.885729331226962,
      "grad_norm": 0.30407586693763733,
      "learning_rate": 5.9755082533714224e-05,
      "loss": 0.17008894681930542,
      "mean_token_accuracy": 0.9289335012435913,
      "num_tokens": 114449296.0,
      "step": 9278
    },
    {
      "entropy": 0.9228288233280182,
      "epoch": 4.886255924170616,
      "grad_norm": 0.3472314178943634,
      "learning_rate": 5.973689111815641e-05,
      "loss": 0.17782555520534515,
      "mean_token_accuracy": 0.9241014868021011,
      "num_tokens": 114461632.0,
      "step": 9279
    },
    {
      "entropy": 0.8673435002565384,
      "epoch": 4.886782517114271,
      "grad_norm": 0.29873529076576233,
      "learning_rate": 5.97187026864264e-05,
      "loss": 0.1670108437538147,
      "mean_token_accuracy": 0.9309113919734955,
      "num_tokens": 114473968.0,
      "step": 9280
    },
    {
      "entropy": 0.885276660323143,
      "epoch": 4.887309110057926,
      "grad_norm": 0.3258194029331207,
      "learning_rate": 5.9700517239603945e-05,
      "loss": 0.16862691938877106,
      "mean_token_accuracy": 0.9275283068418503,
      "num_tokens": 114486304.0,
      "step": 9281
    },
    {
      "entropy": 0.9040714204311371,
      "epoch": 4.8878357030015795,
      "grad_norm": 0.317599892616272,
      "learning_rate": 5.968233477876863e-05,
      "loss": 0.16430729627609253,
      "mean_token_accuracy": 0.9300103485584259,
      "num_tokens": 114498640.0,
      "step": 9282
    },
    {
      "entropy": 0.9044928401708603,
      "epoch": 4.888362295945234,
      "grad_norm": 0.3313135802745819,
      "learning_rate": 5.966415530499983e-05,
      "loss": 0.1630789339542389,
      "mean_token_accuracy": 0.9265755712985992,
      "num_tokens": 114510976.0,
      "step": 9283
    },
    {
      "entropy": 0.8958253413438797,
      "epoch": 4.888888888888889,
      "grad_norm": 0.32161232829093933,
      "learning_rate": 5.9645978819376736e-05,
      "loss": 0.1587880551815033,
      "mean_token_accuracy": 0.935462549328804,
      "num_tokens": 114523312.0,
      "step": 9284
    },
    {
      "entropy": 0.8946435302495956,
      "epoch": 4.889415481832543,
      "grad_norm": 0.3051639199256897,
      "learning_rate": 5.9627805322978405e-05,
      "loss": 0.16082033514976501,
      "mean_token_accuracy": 0.9311655312776566,
      "num_tokens": 114535648.0,
      "step": 9285
    },
    {
      "entropy": 0.8826522082090378,
      "epoch": 4.889942074776198,
      "grad_norm": 0.32270345091819763,
      "learning_rate": 5.960963481688366e-05,
      "loss": 0.16354745626449585,
      "mean_token_accuracy": 0.929367408156395,
      "num_tokens": 114547984.0,
      "step": 9286
    },
    {
      "entropy": 0.8944514393806458,
      "epoch": 4.890468667719853,
      "grad_norm": 0.3517097532749176,
      "learning_rate": 5.959146730217122e-05,
      "loss": 0.19123712182044983,
      "mean_token_accuracy": 0.9202098995447159,
      "num_tokens": 114560320.0,
      "step": 9287
    },
    {
      "entropy": 0.878686398267746,
      "epoch": 4.890995260663507,
      "grad_norm": 0.31580930948257446,
      "learning_rate": 5.957330277991956e-05,
      "loss": 0.16787363588809967,
      "mean_token_accuracy": 0.9280426502227783,
      "num_tokens": 114572656.0,
      "step": 9288
    },
    {
      "entropy": 0.9311894774436951,
      "epoch": 4.891521853607162,
      "grad_norm": 0.32805290818214417,
      "learning_rate": 5.9555141251207e-05,
      "loss": 0.17731031775474548,
      "mean_token_accuracy": 0.9237784296274185,
      "num_tokens": 114584992.0,
      "step": 9289
    },
    {
      "entropy": 0.9274815768003464,
      "epoch": 4.892048446550816,
      "grad_norm": 0.32746708393096924,
      "learning_rate": 5.9536982717111686e-05,
      "loss": 0.17079785466194153,
      "mean_token_accuracy": 0.9303460121154785,
      "num_tokens": 114597328.0,
      "step": 9290
    },
    {
      "entropy": 0.8722960203886032,
      "epoch": 4.892575039494471,
      "grad_norm": 0.31166478991508484,
      "learning_rate": 5.951882717871159e-05,
      "loss": 0.1584705412387848,
      "mean_token_accuracy": 0.9311783015727997,
      "num_tokens": 114609664.0,
      "step": 9291
    },
    {
      "entropy": 0.8913536667823792,
      "epoch": 4.893101632438126,
      "grad_norm": 0.33785316348075867,
      "learning_rate": 5.950067463708446e-05,
      "loss": 0.1690533310174942,
      "mean_token_accuracy": 0.9239986538887024,
      "num_tokens": 114622000.0,
      "step": 9292
    },
    {
      "entropy": 0.9379396438598633,
      "epoch": 4.8936282253817796,
      "grad_norm": 0.3421578109264374,
      "learning_rate": 5.9482525093307984e-05,
      "loss": 0.17159180343151093,
      "mean_token_accuracy": 0.9270720034837723,
      "num_tokens": 114634336.0,
      "step": 9293
    },
    {
      "entropy": 0.9428474754095078,
      "epoch": 4.894154818325434,
      "grad_norm": 0.3197522461414337,
      "learning_rate": 5.94643785484596e-05,
      "loss": 0.1563556045293808,
      "mean_token_accuracy": 0.9327787607908249,
      "num_tokens": 114646672.0,
      "step": 9294
    },
    {
      "entropy": 0.9069431126117706,
      "epoch": 4.894681411269089,
      "grad_norm": 0.2848811447620392,
      "learning_rate": 5.944623500361647e-05,
      "loss": 0.15425236523151398,
      "mean_token_accuracy": 0.9341136962175369,
      "num_tokens": 114659008.0,
      "step": 9295
    },
    {
      "entropy": 0.9342851489782333,
      "epoch": 4.895208004212743,
      "grad_norm": 0.3387324810028076,
      "learning_rate": 5.942809445985572e-05,
      "loss": 0.1770627349615097,
      "mean_token_accuracy": 0.9250018894672394,
      "num_tokens": 114671344.0,
      "step": 9296
    },
    {
      "entropy": 0.9041187316179276,
      "epoch": 4.895734597156398,
      "grad_norm": 0.3387044668197632,
      "learning_rate": 5.940995691825424e-05,
      "loss": 0.17872977256774902,
      "mean_token_accuracy": 0.9281503856182098,
      "num_tokens": 114683680.0,
      "step": 9297
    },
    {
      "entropy": 0.9258536696434021,
      "epoch": 4.896261190100053,
      "grad_norm": 0.3150176405906677,
      "learning_rate": 5.9391822379888764e-05,
      "loss": 0.16436132788658142,
      "mean_token_accuracy": 0.9273700267076492,
      "num_tokens": 114696016.0,
      "step": 9298
    },
    {
      "entropy": 0.9141188710927963,
      "epoch": 4.896787783043707,
      "grad_norm": 0.3235010802745819,
      "learning_rate": 5.9373690845835795e-05,
      "loss": 0.17278021574020386,
      "mean_token_accuracy": 0.9272926151752472,
      "num_tokens": 114708352.0,
      "step": 9299
    },
    {
      "entropy": 0.9123374670743942,
      "epoch": 4.897314375987362,
      "grad_norm": 0.3486504852771759,
      "learning_rate": 5.9355562317171766e-05,
      "loss": 0.17186430096626282,
      "mean_token_accuracy": 0.9291986376047134,
      "num_tokens": 114720688.0,
      "step": 9300
    },
    {
      "entropy": 0.8895768821239471,
      "epoch": 4.897840968931016,
      "grad_norm": 0.3237471878528595,
      "learning_rate": 5.933743679497281e-05,
      "loss": 0.1654815971851349,
      "mean_token_accuracy": 0.9295766353607178,
      "num_tokens": 114733024.0,
      "step": 9301
    },
    {
      "entropy": 0.9308163225650787,
      "epoch": 4.898367561874671,
      "grad_norm": 0.3276287913322449,
      "learning_rate": 5.9319314280314954e-05,
      "loss": 0.17924335598945618,
      "mean_token_accuracy": 0.9244420826435089,
      "num_tokens": 114745360.0,
      "step": 9302
    },
    {
      "entropy": 0.8628402054309845,
      "epoch": 4.898894154818326,
      "grad_norm": 0.2934078872203827,
      "learning_rate": 5.930119477427403e-05,
      "loss": 0.15336529910564423,
      "mean_token_accuracy": 0.9310443699359894,
      "num_tokens": 114757696.0,
      "step": 9303
    },
    {
      "entropy": 0.913166880607605,
      "epoch": 4.8994207477619796,
      "grad_norm": 0.2994149625301361,
      "learning_rate": 5.9283078277925654e-05,
      "loss": 0.1591852307319641,
      "mean_token_accuracy": 0.9312534183263779,
      "num_tokens": 114770032.0,
      "step": 9304
    },
    {
      "entropy": 0.9578994661569595,
      "epoch": 4.899947340705634,
      "grad_norm": 0.3304392397403717,
      "learning_rate": 5.9264964792345353e-05,
      "loss": 0.1748179942369461,
      "mean_token_accuracy": 0.921977162361145,
      "num_tokens": 114782368.0,
      "step": 9305
    },
    {
      "entropy": 0.9136923551559448,
      "epoch": 4.900473933649289,
      "grad_norm": 0.3101983368396759,
      "learning_rate": 5.9246854318608344e-05,
      "loss": 0.1654844731092453,
      "mean_token_accuracy": 0.9304972290992737,
      "num_tokens": 114794704.0,
      "step": 9306
    },
    {
      "entropy": 0.8925996571779251,
      "epoch": 4.901000526592943,
      "grad_norm": 0.33955803513526917,
      "learning_rate": 5.922874685778981e-05,
      "loss": 0.17222380638122559,
      "mean_token_accuracy": 0.9257993996143341,
      "num_tokens": 114807040.0,
      "step": 9307
    },
    {
      "entropy": 0.8864023834466934,
      "epoch": 4.901527119536598,
      "grad_norm": 0.3239210546016693,
      "learning_rate": 5.9210642410964645e-05,
      "loss": 0.17786379158496857,
      "mean_token_accuracy": 0.925503745675087,
      "num_tokens": 114819376.0,
      "step": 9308
    },
    {
      "entropy": 0.8842435479164124,
      "epoch": 4.902053712480253,
      "grad_norm": 0.33383238315582275,
      "learning_rate": 5.9192540979207637e-05,
      "loss": 0.1707066297531128,
      "mean_token_accuracy": 0.9289665967226028,
      "num_tokens": 114831712.0,
      "step": 9309
    },
    {
      "entropy": 0.9077869206666946,
      "epoch": 4.902580305423907,
      "grad_norm": 0.29188287258148193,
      "learning_rate": 5.917444256359332e-05,
      "loss": 0.16449929773807526,
      "mean_token_accuracy": 0.9325157999992371,
      "num_tokens": 114844048.0,
      "step": 9310
    },
    {
      "entropy": 0.9428171515464783,
      "epoch": 4.903106898367562,
      "grad_norm": 0.3549969792366028,
      "learning_rate": 5.9156347165196134e-05,
      "loss": 0.18001793324947357,
      "mean_token_accuracy": 0.9223020523786545,
      "num_tokens": 114856384.0,
      "step": 9311
    },
    {
      "entropy": 0.9449672996997833,
      "epoch": 4.903633491311217,
      "grad_norm": 0.31514841318130493,
      "learning_rate": 5.913825478509025e-05,
      "loss": 0.16791656613349915,
      "mean_token_accuracy": 0.9291617125272751,
      "num_tokens": 114868720.0,
      "step": 9312
    },
    {
      "entropy": 0.9265694469213486,
      "epoch": 4.904160084254871,
      "grad_norm": 0.30857884883880615,
      "learning_rate": 5.912016542434976e-05,
      "loss": 0.16086161136627197,
      "mean_token_accuracy": 0.9276047050952911,
      "num_tokens": 114881056.0,
      "step": 9313
    },
    {
      "entropy": 0.8907498270273209,
      "epoch": 4.904686677198526,
      "grad_norm": 0.32661426067352295,
      "learning_rate": 5.91020790840485e-05,
      "loss": 0.16189296543598175,
      "mean_token_accuracy": 0.932487428188324,
      "num_tokens": 114893392.0,
      "step": 9314
    },
    {
      "entropy": 0.9333143979310989,
      "epoch": 4.9052132701421804,
      "grad_norm": 0.32313498854637146,
      "learning_rate": 5.9083995765260134e-05,
      "loss": 0.174815833568573,
      "mean_token_accuracy": 0.9240923523902893,
      "num_tokens": 114905728.0,
      "step": 9315
    },
    {
      "entropy": 0.9181988686323166,
      "epoch": 4.905739863085834,
      "grad_norm": 0.3207535147666931,
      "learning_rate": 5.9065915469058177e-05,
      "loss": 0.1609046757221222,
      "mean_token_accuracy": 0.9294731467962265,
      "num_tokens": 114918064.0,
      "step": 9316
    },
    {
      "entropy": 0.9402085095643997,
      "epoch": 4.906266456029489,
      "grad_norm": 0.33442455530166626,
      "learning_rate": 5.904783819651597e-05,
      "loss": 0.17129158973693848,
      "mean_token_accuracy": 0.928265392780304,
      "num_tokens": 114930400.0,
      "step": 9317
    },
    {
      "entropy": 0.9600818157196045,
      "epoch": 4.906793048973144,
      "grad_norm": 0.3433302640914917,
      "learning_rate": 5.902976394870663e-05,
      "loss": 0.173824280500412,
      "mean_token_accuracy": 0.9248096644878387,
      "num_tokens": 114942736.0,
      "step": 9318
    },
    {
      "entropy": 0.9292252659797668,
      "epoch": 4.907319641916798,
      "grad_norm": 0.321300745010376,
      "learning_rate": 5.9011692726703116e-05,
      "loss": 0.17256402969360352,
      "mean_token_accuracy": 0.9269725978374481,
      "num_tokens": 114955072.0,
      "step": 9319
    },
    {
      "entropy": 0.8997122347354889,
      "epoch": 4.907846234860453,
      "grad_norm": 0.2938885986804962,
      "learning_rate": 5.8993624531578195e-05,
      "loss": 0.16067785024642944,
      "mean_token_accuracy": 0.932793065905571,
      "num_tokens": 114967408.0,
      "step": 9320
    },
    {
      "entropy": 0.9118629992008209,
      "epoch": 4.908372827804108,
      "grad_norm": 0.29671382904052734,
      "learning_rate": 5.897555936440453e-05,
      "loss": 0.1600441336631775,
      "mean_token_accuracy": 0.9317707270383835,
      "num_tokens": 114979744.0,
      "step": 9321
    },
    {
      "entropy": 0.9408626407384872,
      "epoch": 4.908899420747762,
      "grad_norm": 0.3658956289291382,
      "learning_rate": 5.895749722625452e-05,
      "loss": 0.1817052662372589,
      "mean_token_accuracy": 0.9241613149642944,
      "num_tokens": 114992080.0,
      "step": 9322
    },
    {
      "entropy": 0.877708300948143,
      "epoch": 4.909426013691417,
      "grad_norm": 0.33023250102996826,
      "learning_rate": 5.893943811820041e-05,
      "loss": 0.1727953851222992,
      "mean_token_accuracy": 0.9252429157495499,
      "num_tokens": 115004416.0,
      "step": 9323
    },
    {
      "entropy": 0.9067173004150391,
      "epoch": 4.909952606635071,
      "grad_norm": 0.35036468505859375,
      "learning_rate": 5.8921382041314275e-05,
      "loss": 0.16596205532550812,
      "mean_token_accuracy": 0.930809497833252,
      "num_tokens": 115016752.0,
      "step": 9324
    },
    {
      "entropy": 0.9250523298978806,
      "epoch": 4.910479199578726,
      "grad_norm": 0.32527902722358704,
      "learning_rate": 5.890332899666794e-05,
      "loss": 0.16674506664276123,
      "mean_token_accuracy": 0.9266581237316132,
      "num_tokens": 115029088.0,
      "step": 9325
    },
    {
      "entropy": 0.9385227710008621,
      "epoch": 4.9110057925223805,
      "grad_norm": 0.30357611179351807,
      "learning_rate": 5.8885278985333156e-05,
      "loss": 0.15314704179763794,
      "mean_token_accuracy": 0.9332702606916428,
      "num_tokens": 115041424.0,
      "step": 9326
    },
    {
      "entropy": 0.9614623934030533,
      "epoch": 4.911532385466034,
      "grad_norm": 0.35474637150764465,
      "learning_rate": 5.886723200838141e-05,
      "loss": 0.1641876995563507,
      "mean_token_accuracy": 0.9266858845949173,
      "num_tokens": 115053760.0,
      "step": 9327
    },
    {
      "entropy": 0.9489174485206604,
      "epoch": 4.912058978409689,
      "grad_norm": 0.3305496871471405,
      "learning_rate": 5.8849188066884106e-05,
      "loss": 0.15400606393814087,
      "mean_token_accuracy": 0.936336025595665,
      "num_tokens": 115066096.0,
      "step": 9328
    },
    {
      "entropy": 0.9381827712059021,
      "epoch": 4.912585571353344,
      "grad_norm": 0.34921711683273315,
      "learning_rate": 5.883114716191238e-05,
      "loss": 0.17779318988323212,
      "mean_token_accuracy": 0.9265210777521133,
      "num_tokens": 115078432.0,
      "step": 9329
    },
    {
      "entropy": 0.9520251005887985,
      "epoch": 4.913112164296998,
      "grad_norm": 0.3424340784549713,
      "learning_rate": 5.881310929453721e-05,
      "loss": 0.18406268954277039,
      "mean_token_accuracy": 0.9192226380109787,
      "num_tokens": 115090768.0,
      "step": 9330
    },
    {
      "entropy": 0.904927521944046,
      "epoch": 4.913638757240653,
      "grad_norm": 0.30872711539268494,
      "learning_rate": 5.879507446582939e-05,
      "loss": 0.15748125314712524,
      "mean_token_accuracy": 0.9327780157327652,
      "num_tokens": 115103104.0,
      "step": 9331
    },
    {
      "entropy": 0.9044003784656525,
      "epoch": 4.914165350184308,
      "grad_norm": 0.3096369504928589,
      "learning_rate": 5.877704267685956e-05,
      "loss": 0.16523298621177673,
      "mean_token_accuracy": 0.9302665144205093,
      "num_tokens": 115115440.0,
      "step": 9332
    },
    {
      "entropy": 0.9648252129554749,
      "epoch": 4.914691943127962,
      "grad_norm": 0.34394344687461853,
      "learning_rate": 5.875901392869816e-05,
      "loss": 0.18030229210853577,
      "mean_token_accuracy": 0.9267686456441879,
      "num_tokens": 115127776.0,
      "step": 9333
    },
    {
      "entropy": 0.9282388240098953,
      "epoch": 4.915218536071617,
      "grad_norm": 0.2991066575050354,
      "learning_rate": 5.874098822241545e-05,
      "loss": 0.14410465955734253,
      "mean_token_accuracy": 0.9370485544204712,
      "num_tokens": 115140112.0,
      "step": 9334
    },
    {
      "entropy": 0.9554052501916885,
      "epoch": 4.915745129015271,
      "grad_norm": 0.30297887325286865,
      "learning_rate": 5.87229655590815e-05,
      "loss": 0.1640850305557251,
      "mean_token_accuracy": 0.9307016283273697,
      "num_tokens": 115152448.0,
      "step": 9335
    },
    {
      "entropy": 0.9423719644546509,
      "epoch": 4.916271721958926,
      "grad_norm": 0.3117614984512329,
      "learning_rate": 5.8704945939766234e-05,
      "loss": 0.1675664633512497,
      "mean_token_accuracy": 0.9258502274751663,
      "num_tokens": 115164784.0,
      "step": 9336
    },
    {
      "entropy": 0.9139404147863388,
      "epoch": 4.9167983149025805,
      "grad_norm": 0.30298396944999695,
      "learning_rate": 5.868692936553935e-05,
      "loss": 0.16482360661029816,
      "mean_token_accuracy": 0.9292973428964615,
      "num_tokens": 115177120.0,
      "step": 9337
    },
    {
      "entropy": 0.9500172883272171,
      "epoch": 4.917324907846234,
      "grad_norm": 0.316505491733551,
      "learning_rate": 5.866891583747038e-05,
      "loss": 0.15865392982959747,
      "mean_token_accuracy": 0.9338560700416565,
      "num_tokens": 115189456.0,
      "step": 9338
    },
    {
      "entropy": 0.91290383040905,
      "epoch": 4.917851500789889,
      "grad_norm": 0.2912936806678772,
      "learning_rate": 5.865090535662871e-05,
      "loss": 0.15202707052230835,
      "mean_token_accuracy": 0.9352873414754868,
      "num_tokens": 115201792.0,
      "step": 9339
    },
    {
      "entropy": 0.9219886958599091,
      "epoch": 4.918378093733544,
      "grad_norm": 0.33818334341049194,
      "learning_rate": 5.863289792408351e-05,
      "loss": 0.17483165860176086,
      "mean_token_accuracy": 0.9254404604434967,
      "num_tokens": 115214128.0,
      "step": 9340
    },
    {
      "entropy": 0.9819406718015671,
      "epoch": 4.918904686677198,
      "grad_norm": 0.35899922251701355,
      "learning_rate": 5.861489354090373e-05,
      "loss": 0.19326823949813843,
      "mean_token_accuracy": 0.9183540940284729,
      "num_tokens": 115226464.0,
      "step": 9341
    },
    {
      "entropy": 0.9774647802114487,
      "epoch": 4.919431279620853,
      "grad_norm": 0.3061813712120056,
      "learning_rate": 5.859689220815826e-05,
      "loss": 0.15899212658405304,
      "mean_token_accuracy": 0.9301170408725739,
      "num_tokens": 115238800.0,
      "step": 9342
    },
    {
      "entropy": 0.9548447728157043,
      "epoch": 4.919957872564508,
      "grad_norm": 0.37087732553482056,
      "learning_rate": 5.8578893926915733e-05,
      "loss": 0.1835373044013977,
      "mean_token_accuracy": 0.9208451062440872,
      "num_tokens": 115251136.0,
      "step": 9343
    },
    {
      "entropy": 0.9868229180574417,
      "epoch": 4.920484465508162,
      "grad_norm": 0.4365648329257965,
      "learning_rate": 5.856089869824453e-05,
      "loss": 0.16759006679058075,
      "mean_token_accuracy": 0.9274678826332092,
      "num_tokens": 115263472.0,
      "step": 9344
    },
    {
      "entropy": 0.9716550260782242,
      "epoch": 4.921011058451817,
      "grad_norm": 0.3230785131454468,
      "learning_rate": 5.8542906523212985e-05,
      "loss": 0.15922094881534576,
      "mean_token_accuracy": 0.9334453046321869,
      "num_tokens": 115275808.0,
      "step": 9345
    },
    {
      "entropy": 0.9856964349746704,
      "epoch": 4.921537651395472,
      "grad_norm": 0.3701082766056061,
      "learning_rate": 5.852491740288913e-05,
      "loss": 0.19076339900493622,
      "mean_token_accuracy": 0.9177150577306747,
      "num_tokens": 115288144.0,
      "step": 9346
    },
    {
      "entropy": 0.9459339380264282,
      "epoch": 4.922064244339126,
      "grad_norm": 0.3227887451648712,
      "learning_rate": 5.850693133834093e-05,
      "loss": 0.1690850853919983,
      "mean_token_accuracy": 0.9287611097097397,
      "num_tokens": 115300480.0,
      "step": 9347
    },
    {
      "entropy": 0.9395811706781387,
      "epoch": 4.9225908372827805,
      "grad_norm": 0.3116438090801239,
      "learning_rate": 5.848894833063605e-05,
      "loss": 0.17614488303661346,
      "mean_token_accuracy": 0.9265125244855881,
      "num_tokens": 115312816.0,
      "step": 9348
    },
    {
      "entropy": 0.9723171889781952,
      "epoch": 4.923117430226435,
      "grad_norm": 0.3122919499874115,
      "learning_rate": 5.8470968380842116e-05,
      "loss": 0.16692239046096802,
      "mean_token_accuracy": 0.9280216097831726,
      "num_tokens": 115325152.0,
      "step": 9349
    },
    {
      "entropy": 0.9282762259244919,
      "epoch": 4.923644023170089,
      "grad_norm": 0.319529265165329,
      "learning_rate": 5.8452991490026456e-05,
      "loss": 0.17438848316669464,
      "mean_token_accuracy": 0.9270227700471878,
      "num_tokens": 115337488.0,
      "step": 9350
    },
    {
      "entropy": 0.9387772977352142,
      "epoch": 4.924170616113744,
      "grad_norm": 0.2963738441467285,
      "learning_rate": 5.843501765925625e-05,
      "loss": 0.14648286998271942,
      "mean_token_accuracy": 0.9406982809305191,
      "num_tokens": 115349824.0,
      "step": 9351
    },
    {
      "entropy": 0.9783714413642883,
      "epoch": 4.924697209057399,
      "grad_norm": 0.33284756541252136,
      "learning_rate": 5.841704688959851e-05,
      "loss": 0.1824008822441101,
      "mean_token_accuracy": 0.9226804673671722,
      "num_tokens": 115362160.0,
      "step": 9352
    },
    {
      "entropy": 1.0005029290914536,
      "epoch": 4.925223802001053,
      "grad_norm": 0.3474387228488922,
      "learning_rate": 5.839907918212004e-05,
      "loss": 0.17754733562469482,
      "mean_token_accuracy": 0.9215953201055527,
      "num_tokens": 115374496.0,
      "step": 9353
    },
    {
      "entropy": 0.9703384190797806,
      "epoch": 4.925750394944708,
      "grad_norm": 0.3040783107280731,
      "learning_rate": 5.8381114537887525e-05,
      "loss": 0.161107137799263,
      "mean_token_accuracy": 0.9309516549110413,
      "num_tokens": 115386832.0,
      "step": 9354
    },
    {
      "entropy": 0.9523343741893768,
      "epoch": 4.926276987888363,
      "grad_norm": 0.3469637334346771,
      "learning_rate": 5.8363152957967307e-05,
      "loss": 0.17863255739212036,
      "mean_token_accuracy": 0.9232510924339294,
      "num_tokens": 115399168.0,
      "step": 9355
    },
    {
      "entropy": 0.9407844543457031,
      "epoch": 4.926803580832017,
      "grad_norm": 0.30398619174957275,
      "learning_rate": 5.834519444342577e-05,
      "loss": 0.16317002475261688,
      "mean_token_accuracy": 0.9284565448760986,
      "num_tokens": 115411504.0,
      "step": 9356
    },
    {
      "entropy": 0.9656679928302765,
      "epoch": 4.927330173775672,
      "grad_norm": 0.30141156911849976,
      "learning_rate": 5.832723899532897e-05,
      "loss": 0.14263194799423218,
      "mean_token_accuracy": 0.9430341720581055,
      "num_tokens": 115423840.0,
      "step": 9357
    },
    {
      "entropy": 0.9619395136833191,
      "epoch": 4.927856766719326,
      "grad_norm": 0.32897594571113586,
      "learning_rate": 5.830928661474282e-05,
      "loss": 0.187732994556427,
      "mean_token_accuracy": 0.9198961853981018,
      "num_tokens": 115436176.0,
      "step": 9358
    },
    {
      "entropy": 0.9617651253938675,
      "epoch": 4.9283833596629805,
      "grad_norm": 0.31268131732940674,
      "learning_rate": 5.829133730273306e-05,
      "loss": 0.17076878249645233,
      "mean_token_accuracy": 0.9300666600465775,
      "num_tokens": 115448512.0,
      "step": 9359
    },
    {
      "entropy": 1.0081706196069717,
      "epoch": 4.928909952606635,
      "grad_norm": 0.3417055308818817,
      "learning_rate": 5.827339106036521e-05,
      "loss": 0.18210983276367188,
      "mean_token_accuracy": 0.9194571822881699,
      "num_tokens": 115460848.0,
      "step": 9360
    },
    {
      "entropy": 0.9907486289739609,
      "epoch": 4.929436545550289,
      "grad_norm": 0.3143501579761505,
      "learning_rate": 5.825544788870465e-05,
      "loss": 0.1710372418165207,
      "mean_token_accuracy": 0.9332590699195862,
      "num_tokens": 115473184.0,
      "step": 9361
    },
    {
      "entropy": 0.9614686220884323,
      "epoch": 4.929963138493944,
      "grad_norm": 0.3213489353656769,
      "learning_rate": 5.823750778881657e-05,
      "loss": 0.16459202766418457,
      "mean_token_accuracy": 0.9330161213874817,
      "num_tokens": 115485520.0,
      "step": 9362
    },
    {
      "entropy": 0.9386411756277084,
      "epoch": 4.930489731437599,
      "grad_norm": 0.29499050974845886,
      "learning_rate": 5.821957076176594e-05,
      "loss": 0.16460591554641724,
      "mean_token_accuracy": 0.9271390736103058,
      "num_tokens": 115497856.0,
      "step": 9363
    },
    {
      "entropy": 0.967206209897995,
      "epoch": 4.931016324381253,
      "grad_norm": 0.32410702109336853,
      "learning_rate": 5.820163680861761e-05,
      "loss": 0.17515471577644348,
      "mean_token_accuracy": 0.9262986332178116,
      "num_tokens": 115510192.0,
      "step": 9364
    },
    {
      "entropy": 0.997588574886322,
      "epoch": 4.931542917324908,
      "grad_norm": 0.32928773760795593,
      "learning_rate": 5.8183705930436186e-05,
      "loss": 0.172166645526886,
      "mean_token_accuracy": 0.9269043803215027,
      "num_tokens": 115522528.0,
      "step": 9365
    },
    {
      "entropy": 0.9945474416017532,
      "epoch": 4.932069510268563,
      "grad_norm": 0.3223489820957184,
      "learning_rate": 5.816577812828614e-05,
      "loss": 0.1568552851676941,
      "mean_token_accuracy": 0.9318364411592484,
      "num_tokens": 115534864.0,
      "step": 9366
    },
    {
      "entropy": 0.9689517766237259,
      "epoch": 4.932596103212217,
      "grad_norm": 0.3623938262462616,
      "learning_rate": 5.814785340323173e-05,
      "loss": 0.179118812084198,
      "mean_token_accuracy": 0.9249956458806992,
      "num_tokens": 115547200.0,
      "step": 9367
    },
    {
      "entropy": 0.9781549423933029,
      "epoch": 4.933122696155872,
      "grad_norm": 0.3357399106025696,
      "learning_rate": 5.8129931756337056e-05,
      "loss": 0.16774648427963257,
      "mean_token_accuracy": 0.9244172424077988,
      "num_tokens": 115559536.0,
      "step": 9368
    },
    {
      "entropy": 0.9704612046480179,
      "epoch": 4.933649289099526,
      "grad_norm": 0.32363811135292053,
      "learning_rate": 5.811201318866597e-05,
      "loss": 0.1788511425256729,
      "mean_token_accuracy": 0.9203287065029144,
      "num_tokens": 115571872.0,
      "step": 9369
    },
    {
      "entropy": 0.9416558891534805,
      "epoch": 4.9341758820431805,
      "grad_norm": 0.2980855405330658,
      "learning_rate": 5.809409770128227e-05,
      "loss": 0.16315722465515137,
      "mean_token_accuracy": 0.9311324954032898,
      "num_tokens": 115584208.0,
      "step": 9370
    },
    {
      "entropy": 0.9185937494039536,
      "epoch": 4.934702474986835,
      "grad_norm": 0.31508633494377136,
      "learning_rate": 5.807618529524945e-05,
      "loss": 0.15790027379989624,
      "mean_token_accuracy": 0.9321376830339432,
      "num_tokens": 115596544.0,
      "step": 9371
    },
    {
      "entropy": 0.9664212763309479,
      "epoch": 4.935229067930489,
      "grad_norm": 0.3180954158306122,
      "learning_rate": 5.8058275971630915e-05,
      "loss": 0.17068061232566833,
      "mean_token_accuracy": 0.9232131242752075,
      "num_tokens": 115608880.0,
      "step": 9372
    },
    {
      "entropy": 0.9768130630254745,
      "epoch": 4.935755660874144,
      "grad_norm": 0.34208452701568604,
      "learning_rate": 5.804036973148976e-05,
      "loss": 0.162981778383255,
      "mean_token_accuracy": 0.9295215010643005,
      "num_tokens": 115621216.0,
      "step": 9373
    },
    {
      "entropy": 0.9430342763662338,
      "epoch": 4.936282253817799,
      "grad_norm": 0.3556082546710968,
      "learning_rate": 5.802246657588901e-05,
      "loss": 0.17354175448417664,
      "mean_token_accuracy": 0.928137943148613,
      "num_tokens": 115633552.0,
      "step": 9374
    },
    {
      "entropy": 0.881765827536583,
      "epoch": 4.936808846761453,
      "grad_norm": 0.29666128754615784,
      "learning_rate": 5.8004566505891475e-05,
      "loss": 0.15805462002754211,
      "mean_token_accuracy": 0.9287291318178177,
      "num_tokens": 115645888.0,
      "step": 9375
    },
    {
      "entropy": 0.9361718595027924,
      "epoch": 4.937335439705108,
      "grad_norm": 0.3010590970516205,
      "learning_rate": 5.798666952255976e-05,
      "loss": 0.16552048921585083,
      "mean_token_accuracy": 0.9276218116283417,
      "num_tokens": 115658224.0,
      "step": 9376
    },
    {
      "entropy": 0.9072093516588211,
      "epoch": 4.937862032648763,
      "grad_norm": 0.3231160640716553,
      "learning_rate": 5.796877562695633e-05,
      "loss": 0.16154517233371735,
      "mean_token_accuracy": 0.9288052469491959,
      "num_tokens": 115670560.0,
      "step": 9377
    },
    {
      "entropy": 0.9381129592657089,
      "epoch": 4.938388625592417,
      "grad_norm": 0.35191112756729126,
      "learning_rate": 5.795088482014342e-05,
      "loss": 0.17764237523078918,
      "mean_token_accuracy": 0.9273613542318344,
      "num_tokens": 115682896.0,
      "step": 9378
    },
    {
      "entropy": 0.9498243778944016,
      "epoch": 4.938915218536072,
      "grad_norm": 0.3514261245727539,
      "learning_rate": 5.793299710318314e-05,
      "loss": 0.18616883456707,
      "mean_token_accuracy": 0.9251926690340042,
      "num_tokens": 115695232.0,
      "step": 9379
    },
    {
      "entropy": 0.9347743093967438,
      "epoch": 4.9394418114797265,
      "grad_norm": 0.33045297861099243,
      "learning_rate": 5.7915112477137335e-05,
      "loss": 0.17543768882751465,
      "mean_token_accuracy": 0.9261060506105423,
      "num_tokens": 115707568.0,
      "step": 9380
    },
    {
      "entropy": 0.9594391584396362,
      "epoch": 4.9399684044233805,
      "grad_norm": 0.3461679518222809,
      "learning_rate": 5.7897230943067734e-05,
      "loss": 0.1716955453157425,
      "mean_token_accuracy": 0.9220835864543915,
      "num_tokens": 115719904.0,
      "step": 9381
    },
    {
      "entropy": 0.912274107336998,
      "epoch": 4.940494997367035,
      "grad_norm": 0.3083917498588562,
      "learning_rate": 5.787935250203585e-05,
      "loss": 0.1601388305425644,
      "mean_token_accuracy": 0.9297442883253098,
      "num_tokens": 115732240.0,
      "step": 9382
    },
    {
      "entropy": 0.913392573595047,
      "epoch": 4.94102159031069,
      "grad_norm": 0.3031776547431946,
      "learning_rate": 5.786147715510306e-05,
      "loss": 0.1611706018447876,
      "mean_token_accuracy": 0.9329336583614349,
      "num_tokens": 115744576.0,
      "step": 9383
    },
    {
      "entropy": 0.9634020626544952,
      "epoch": 4.941548183254344,
      "grad_norm": 0.3595740497112274,
      "learning_rate": 5.784360490333042e-05,
      "loss": 0.1803797483444214,
      "mean_token_accuracy": 0.9214609563350677,
      "num_tokens": 115756912.0,
      "step": 9384
    },
    {
      "entropy": 0.9271872788667679,
      "epoch": 4.942074776197999,
      "grad_norm": 0.33162611722946167,
      "learning_rate": 5.782573574777899e-05,
      "loss": 0.16804532706737518,
      "mean_token_accuracy": 0.9263215363025665,
      "num_tokens": 115769248.0,
      "step": 9385
    },
    {
      "entropy": 0.92726831138134,
      "epoch": 4.942601369141654,
      "grad_norm": 0.32566916942596436,
      "learning_rate": 5.7807869689509554e-05,
      "loss": 0.18525749444961548,
      "mean_token_accuracy": 0.9209079444408417,
      "num_tokens": 115781584.0,
      "step": 9386
    },
    {
      "entropy": 0.9415363818407059,
      "epoch": 4.943127962085308,
      "grad_norm": 0.3226025104522705,
      "learning_rate": 5.779000672958268e-05,
      "loss": 0.16038167476654053,
      "mean_token_accuracy": 0.9302356839179993,
      "num_tokens": 115793920.0,
      "step": 9387
    },
    {
      "entropy": 0.9243748486042023,
      "epoch": 4.943654555028963,
      "grad_norm": 0.32364335656166077,
      "learning_rate": 5.777214686905881e-05,
      "loss": 0.157021626830101,
      "mean_token_accuracy": 0.9371575564146042,
      "num_tokens": 115806256.0,
      "step": 9388
    },
    {
      "entropy": 0.9075966626405716,
      "epoch": 4.944181147972618,
      "grad_norm": 0.30498507618904114,
      "learning_rate": 5.7754290108998175e-05,
      "loss": 0.15704648196697235,
      "mean_token_accuracy": 0.9347554445266724,
      "num_tokens": 115818592.0,
      "step": 9389
    },
    {
      "entropy": 0.9089589267969131,
      "epoch": 4.944707740916272,
      "grad_norm": 0.31486794352531433,
      "learning_rate": 5.773643645046083e-05,
      "loss": 0.17221665382385254,
      "mean_token_accuracy": 0.9295891523361206,
      "num_tokens": 115830928.0,
      "step": 9390
    },
    {
      "entropy": 0.922472670674324,
      "epoch": 4.9452343338599265,
      "grad_norm": 0.3181609511375427,
      "learning_rate": 5.771858589450663e-05,
      "loss": 0.16150923073291779,
      "mean_token_accuracy": 0.9333121627569199,
      "num_tokens": 115843264.0,
      "step": 9391
    },
    {
      "entropy": 0.915450856089592,
      "epoch": 4.9457609268035805,
      "grad_norm": 0.28799742460250854,
      "learning_rate": 5.7700738442195256e-05,
      "loss": 0.15151140093803406,
      "mean_token_accuracy": 0.9380517452955246,
      "num_tokens": 115855600.0,
      "step": 9392
    },
    {
      "entropy": 0.9430654495954514,
      "epoch": 4.946287519747235,
      "grad_norm": 0.3145394027233124,
      "learning_rate": 5.7682894094586224e-05,
      "loss": 0.16287337243556976,
      "mean_token_accuracy": 0.9282755255699158,
      "num_tokens": 115867936.0,
      "step": 9393
    },
    {
      "entropy": 0.9298929274082184,
      "epoch": 4.94681411269089,
      "grad_norm": 0.3063369691371918,
      "learning_rate": 5.7665052852738846e-05,
      "loss": 0.16224738955497742,
      "mean_token_accuracy": 0.93083256483078,
      "num_tokens": 115880272.0,
      "step": 9394
    },
    {
      "entropy": 0.8956407159566879,
      "epoch": 4.947340705634544,
      "grad_norm": 0.32702919840812683,
      "learning_rate": 5.764721471771224e-05,
      "loss": 0.1812189370393753,
      "mean_token_accuracy": 0.923660010099411,
      "num_tokens": 115892608.0,
      "step": 9395
    },
    {
      "entropy": 0.9093689769506454,
      "epoch": 4.947867298578199,
      "grad_norm": 0.3080686032772064,
      "learning_rate": 5.762937969056538e-05,
      "loss": 0.15826645493507385,
      "mean_token_accuracy": 0.9359744638204575,
      "num_tokens": 115904944.0,
      "step": 9396
    },
    {
      "entropy": 0.9180983453989029,
      "epoch": 4.948393891521854,
      "grad_norm": 0.30890801548957825,
      "learning_rate": 5.7611547772357e-05,
      "loss": 0.17149624228477478,
      "mean_token_accuracy": 0.9241906851530075,
      "num_tokens": 115917280.0,
      "step": 9397
    },
    {
      "entropy": 0.8848812580108643,
      "epoch": 4.948920484465508,
      "grad_norm": 0.3318977952003479,
      "learning_rate": 5.759371896414565e-05,
      "loss": 0.18148359656333923,
      "mean_token_accuracy": 0.9215588569641113,
      "num_tokens": 115929616.0,
      "step": 9398
    },
    {
      "entropy": 0.9263090342283249,
      "epoch": 4.949447077409163,
      "grad_norm": 0.3431561589241028,
      "learning_rate": 5.7575893266989785e-05,
      "loss": 0.17720580101013184,
      "mean_token_accuracy": 0.9285667836666107,
      "num_tokens": 115941952.0,
      "step": 9399
    },
    {
      "entropy": 0.8653641939163208,
      "epoch": 4.949973670352818,
      "grad_norm": 0.3527938723564148,
      "learning_rate": 5.755807068194759e-05,
      "loss": 0.1529770791530609,
      "mean_token_accuracy": 0.9344857335090637,
      "num_tokens": 115954288.0,
      "step": 9400
    },
    {
      "entropy": 0.8993526250123978,
      "epoch": 4.950500263296472,
      "grad_norm": 0.3243945240974426,
      "learning_rate": 5.754025121007709e-05,
      "loss": 0.16913282871246338,
      "mean_token_accuracy": 0.9260164946317673,
      "num_tokens": 115966624.0,
      "step": 9401
    },
    {
      "entropy": 0.8799459934234619,
      "epoch": 4.9510268562401265,
      "grad_norm": 0.29280006885528564,
      "learning_rate": 5.7522434852436146e-05,
      "loss": 0.16644714772701263,
      "mean_token_accuracy": 0.9297489523887634,
      "num_tokens": 115978960.0,
      "step": 9402
    },
    {
      "entropy": 0.9246260076761246,
      "epoch": 4.9515534491837805,
      "grad_norm": 0.3270162045955658,
      "learning_rate": 5.7504621610082354e-05,
      "loss": 0.1706567257642746,
      "mean_token_accuracy": 0.9282507449388504,
      "num_tokens": 115991296.0,
      "step": 9403
    },
    {
      "entropy": 0.9047050178050995,
      "epoch": 4.952080042127435,
      "grad_norm": 0.32847970724105835,
      "learning_rate": 5.74868114840732e-05,
      "loss": 0.1750563681125641,
      "mean_token_accuracy": 0.9257060140371323,
      "num_tokens": 116003632.0,
      "step": 9404
    },
    {
      "entropy": 0.9128126055002213,
      "epoch": 4.95260663507109,
      "grad_norm": 0.31370076537132263,
      "learning_rate": 5.7469004475465945e-05,
      "loss": 0.16696500778198242,
      "mean_token_accuracy": 0.9307531118392944,
      "num_tokens": 116015968.0,
      "step": 9405
    },
    {
      "entropy": 0.9282513111829758,
      "epoch": 4.953133228014744,
      "grad_norm": 0.29254990816116333,
      "learning_rate": 5.745120058531777e-05,
      "loss": 0.14533621072769165,
      "mean_token_accuracy": 0.9334218353033066,
      "num_tokens": 116028304.0,
      "step": 9406
    },
    {
      "entropy": 0.8847913295030594,
      "epoch": 4.953659820958399,
      "grad_norm": 0.31924283504486084,
      "learning_rate": 5.743339981468553e-05,
      "loss": 0.168746680021286,
      "mean_token_accuracy": 0.930860698223114,
      "num_tokens": 116040640.0,
      "step": 9407
    },
    {
      "entropy": 0.8906414955854416,
      "epoch": 4.954186413902054,
      "grad_norm": 0.31640809774398804,
      "learning_rate": 5.741560216462596e-05,
      "loss": 0.17188094556331635,
      "mean_token_accuracy": 0.9263556897640228,
      "num_tokens": 116052976.0,
      "step": 9408
    },
    {
      "entropy": 0.8946892470121384,
      "epoch": 4.954713006845708,
      "grad_norm": 0.3183387815952301,
      "learning_rate": 5.73978076361956e-05,
      "loss": 0.16452908515930176,
      "mean_token_accuracy": 0.9298630207777023,
      "num_tokens": 116065312.0,
      "step": 9409
    },
    {
      "entropy": 0.9338386058807373,
      "epoch": 4.955239599789363,
      "grad_norm": 0.3612841069698334,
      "learning_rate": 5.738001623045082e-05,
      "loss": 0.18146231770515442,
      "mean_token_accuracy": 0.9221067875623703,
      "num_tokens": 116077648.0,
      "step": 9410
    },
    {
      "entropy": 0.9196339249610901,
      "epoch": 4.955766192733018,
      "grad_norm": 0.3265555500984192,
      "learning_rate": 5.736222794844779e-05,
      "loss": 0.17468717694282532,
      "mean_token_accuracy": 0.9233785420656204,
      "num_tokens": 116089984.0,
      "step": 9411
    },
    {
      "entropy": 0.9070274978876114,
      "epoch": 4.956292785676672,
      "grad_norm": 0.3427065908908844,
      "learning_rate": 5.734444279124247e-05,
      "loss": 0.1812526136636734,
      "mean_token_accuracy": 0.9283244758844376,
      "num_tokens": 116102320.0,
      "step": 9412
    },
    {
      "entropy": 0.9212836623191833,
      "epoch": 4.9568193786203265,
      "grad_norm": 0.33548495173454285,
      "learning_rate": 5.732666075989069e-05,
      "loss": 0.16622008383274078,
      "mean_token_accuracy": 0.9248946905136108,
      "num_tokens": 116114656.0,
      "step": 9413
    },
    {
      "entropy": 0.884115681052208,
      "epoch": 4.957345971563981,
      "grad_norm": 0.3081040680408478,
      "learning_rate": 5.730888185544807e-05,
      "loss": 0.1468859314918518,
      "mean_token_accuracy": 0.9315646439790726,
      "num_tokens": 116126992.0,
      "step": 9414
    },
    {
      "entropy": 0.9673860967159271,
      "epoch": 4.957872564507635,
      "grad_norm": 0.37031683325767517,
      "learning_rate": 5.7291106078970014e-05,
      "loss": 0.18177001178264618,
      "mean_token_accuracy": 0.9270854294300079,
      "num_tokens": 116139328.0,
      "step": 9415
    },
    {
      "entropy": 0.9081185907125473,
      "epoch": 4.95839915745129,
      "grad_norm": 0.32079094648361206,
      "learning_rate": 5.7273333431511776e-05,
      "loss": 0.16575801372528076,
      "mean_token_accuracy": 0.9265887588262558,
      "num_tokens": 116151664.0,
      "step": 9416
    },
    {
      "entropy": 0.8875092566013336,
      "epoch": 4.958925750394945,
      "grad_norm": 0.3274156153202057,
      "learning_rate": 5.725556391412843e-05,
      "loss": 0.1604728102684021,
      "mean_token_accuracy": 0.9308022409677505,
      "num_tokens": 116164000.0,
      "step": 9417
    },
    {
      "entropy": 0.9134371280670166,
      "epoch": 4.959452343338599,
      "grad_norm": 0.31842270493507385,
      "learning_rate": 5.7237797527874824e-05,
      "loss": 0.16845707595348358,
      "mean_token_accuracy": 0.9241103827953339,
      "num_tokens": 116176336.0,
      "step": 9418
    },
    {
      "entropy": 0.9297776818275452,
      "epoch": 4.959978936282254,
      "grad_norm": 0.31639397144317627,
      "learning_rate": 5.722003427380562e-05,
      "loss": 0.1701333224773407,
      "mean_token_accuracy": 0.92668816447258,
      "num_tokens": 116188672.0,
      "step": 9419
    },
    {
      "entropy": 0.9101113528013229,
      "epoch": 4.960505529225909,
      "grad_norm": 0.3030407726764679,
      "learning_rate": 5.720227415297541e-05,
      "loss": 0.1549828201532364,
      "mean_token_accuracy": 0.9333423227071762,
      "num_tokens": 116201008.0,
      "step": 9420
    },
    {
      "entropy": 0.9348535239696503,
      "epoch": 4.961032122169563,
      "grad_norm": 0.29994162917137146,
      "learning_rate": 5.718451716643849e-05,
      "loss": 0.15588951110839844,
      "mean_token_accuracy": 0.9349348545074463,
      "num_tokens": 116213344.0,
      "step": 9421
    },
    {
      "entropy": 0.9469076693058014,
      "epoch": 4.961558715113218,
      "grad_norm": 0.3399045765399933,
      "learning_rate": 5.716676331524891e-05,
      "loss": 0.17242883145809174,
      "mean_token_accuracy": 0.9294859617948532,
      "num_tokens": 116225680.0,
      "step": 9422
    },
    {
      "entropy": 0.9611606448888779,
      "epoch": 4.9620853080568725,
      "grad_norm": 0.3694673478603363,
      "learning_rate": 5.714901260046064e-05,
      "loss": 0.18450355529785156,
      "mean_token_accuracy": 0.918832927942276,
      "num_tokens": 116238016.0,
      "step": 9423
    },
    {
      "entropy": 0.9355585128068924,
      "epoch": 4.9626119010005265,
      "grad_norm": 0.29991161823272705,
      "learning_rate": 5.713126502312748e-05,
      "loss": 0.1599108874797821,
      "mean_token_accuracy": 0.9283605962991714,
      "num_tokens": 116250352.0,
      "step": 9424
    },
    {
      "entropy": 0.9244567900896072,
      "epoch": 4.963138493944181,
      "grad_norm": 0.3170728087425232,
      "learning_rate": 5.711352058430295e-05,
      "loss": 0.1609768271446228,
      "mean_token_accuracy": 0.9301342666149139,
      "num_tokens": 116262688.0,
      "step": 9425
    },
    {
      "entropy": 0.8927604705095291,
      "epoch": 4.963665086887835,
      "grad_norm": 0.3002975583076477,
      "learning_rate": 5.709577928504044e-05,
      "loss": 0.1746864914894104,
      "mean_token_accuracy": 0.9230528324842453,
      "num_tokens": 116275024.0,
      "step": 9426
    },
    {
      "entropy": 0.9116622805595398,
      "epoch": 4.96419167983149,
      "grad_norm": 0.33546826243400574,
      "learning_rate": 5.707804112639319e-05,
      "loss": 0.16620329022407532,
      "mean_token_accuracy": 0.9282449781894684,
      "num_tokens": 116287360.0,
      "step": 9427
    },
    {
      "entropy": 0.9299652725458145,
      "epoch": 4.964718272775145,
      "grad_norm": 0.31117022037506104,
      "learning_rate": 5.7060306109414196e-05,
      "loss": 0.16363778710365295,
      "mean_token_accuracy": 0.9270086139440536,
      "num_tokens": 116299696.0,
      "step": 9428
    },
    {
      "entropy": 0.9201690256595612,
      "epoch": 4.965244865718799,
      "grad_norm": 0.3294250965118408,
      "learning_rate": 5.7042574235156266e-05,
      "loss": 0.1854371726512909,
      "mean_token_accuracy": 0.9225277602672577,
      "num_tokens": 116312032.0,
      "step": 9429
    },
    {
      "entropy": 0.8899281024932861,
      "epoch": 4.965771458662454,
      "grad_norm": 0.2941740155220032,
      "learning_rate": 5.702484550467204e-05,
      "loss": 0.16896389424800873,
      "mean_token_accuracy": 0.922933891415596,
      "num_tokens": 116324368.0,
      "step": 9430
    },
    {
      "entropy": 0.9170496612787247,
      "epoch": 4.966298051606109,
      "grad_norm": 0.3286184072494507,
      "learning_rate": 5.7007119919013974e-05,
      "loss": 0.18167276680469513,
      "mean_token_accuracy": 0.9213146716356277,
      "num_tokens": 116336704.0,
      "step": 9431
    },
    {
      "entropy": 0.9412564933300018,
      "epoch": 4.966824644549763,
      "grad_norm": 0.36381611227989197,
      "learning_rate": 5.6989397479234374e-05,
      "loss": 0.18499529361724854,
      "mean_token_accuracy": 0.9203563630580902,
      "num_tokens": 116349040.0,
      "step": 9432
    },
    {
      "entropy": 0.8793033510446548,
      "epoch": 4.967351237493418,
      "grad_norm": 0.2695026993751526,
      "learning_rate": 5.6971678186385204e-05,
      "loss": 0.14900889992713928,
      "mean_token_accuracy": 0.9365333318710327,
      "num_tokens": 116361376.0,
      "step": 9433
    },
    {
      "entropy": 0.8892877399921417,
      "epoch": 4.967877830437072,
      "grad_norm": 0.3132728338241577,
      "learning_rate": 5.695396204151846e-05,
      "loss": 0.1662237048149109,
      "mean_token_accuracy": 0.930364578962326,
      "num_tokens": 116373712.0,
      "step": 9434
    },
    {
      "entropy": 0.9384078681468964,
      "epoch": 4.9684044233807265,
      "grad_norm": 0.3141588866710663,
      "learning_rate": 5.693624904568582e-05,
      "loss": 0.16861280798912048,
      "mean_token_accuracy": 0.9259446710348129,
      "num_tokens": 116386048.0,
      "step": 9435
    },
    {
      "entropy": 0.8881506472826004,
      "epoch": 4.968931016324381,
      "grad_norm": 0.29548725485801697,
      "learning_rate": 5.6918539199938794e-05,
      "loss": 0.1610819697380066,
      "mean_token_accuracy": 0.9345932602882385,
      "num_tokens": 116398384.0,
      "step": 9436
    },
    {
      "entropy": 0.920554518699646,
      "epoch": 4.969457609268035,
      "grad_norm": 0.33068493008613586,
      "learning_rate": 5.690083250532869e-05,
      "loss": 0.17839539051055908,
      "mean_token_accuracy": 0.9240716248750687,
      "num_tokens": 116410720.0,
      "step": 9437
    },
    {
      "entropy": 0.9246425628662109,
      "epoch": 4.96998420221169,
      "grad_norm": 0.3303989768028259,
      "learning_rate": 5.688312896290671e-05,
      "loss": 0.1687246561050415,
      "mean_token_accuracy": 0.9263402670621872,
      "num_tokens": 116423056.0,
      "step": 9438
    },
    {
      "entropy": 0.9381060898303986,
      "epoch": 4.970510795155345,
      "grad_norm": 0.3343820571899414,
      "learning_rate": 5.6865428573723756e-05,
      "loss": 0.18233567476272583,
      "mean_token_accuracy": 0.9212479740381241,
      "num_tokens": 116435392.0,
      "step": 9439
    },
    {
      "entropy": 0.9213667809963226,
      "epoch": 4.971037388098999,
      "grad_norm": 0.2734557092189789,
      "learning_rate": 5.6847731338830614e-05,
      "loss": 0.15063132345676422,
      "mean_token_accuracy": 0.9358962327241898,
      "num_tokens": 116447728.0,
      "step": 9440
    },
    {
      "entropy": 0.890294760465622,
      "epoch": 4.971563981042654,
      "grad_norm": 0.34502044320106506,
      "learning_rate": 5.683003725927788e-05,
      "loss": 0.18147170543670654,
      "mean_token_accuracy": 0.925027921795845,
      "num_tokens": 116460064.0,
      "step": 9441
    },
    {
      "entropy": 0.9339338093996048,
      "epoch": 4.972090573986309,
      "grad_norm": 0.33315202593803406,
      "learning_rate": 5.681234633611593e-05,
      "loss": 0.17026439309120178,
      "mean_token_accuracy": 0.923798143863678,
      "num_tokens": 116472400.0,
      "step": 9442
    },
    {
      "entropy": 0.9210115224123001,
      "epoch": 4.972617166929963,
      "grad_norm": 0.33159175515174866,
      "learning_rate": 5.679465857039497e-05,
      "loss": 0.1778094321489334,
      "mean_token_accuracy": 0.9244426190853119,
      "num_tokens": 116484736.0,
      "step": 9443
    },
    {
      "entropy": 0.9362052083015442,
      "epoch": 4.973143759873618,
      "grad_norm": 0.3639310598373413,
      "learning_rate": 5.6776973963165035e-05,
      "loss": 0.19108359515666962,
      "mean_token_accuracy": 0.9217455983161926,
      "num_tokens": 116497072.0,
      "step": 9444
    },
    {
      "entropy": 0.9329159706830978,
      "epoch": 4.9736703528172725,
      "grad_norm": 0.3310304582118988,
      "learning_rate": 5.6759292515475936e-05,
      "loss": 0.16485299170017242,
      "mean_token_accuracy": 0.9283944219350815,
      "num_tokens": 116509408.0,
      "step": 9445
    },
    {
      "entropy": 0.9604510962963104,
      "epoch": 4.9741969457609265,
      "grad_norm": 0.36326339840888977,
      "learning_rate": 5.674161422837734e-05,
      "loss": 0.18484851717948914,
      "mean_token_accuracy": 0.922455444931984,
      "num_tokens": 116521744.0,
      "step": 9446
    },
    {
      "entropy": 0.9065830409526825,
      "epoch": 4.974723538704581,
      "grad_norm": 0.3017370104789734,
      "learning_rate": 5.6723939102918665e-05,
      "loss": 0.14546704292297363,
      "mean_token_accuracy": 0.9372200965881348,
      "num_tokens": 116534080.0,
      "step": 9447
    },
    {
      "entropy": 0.9202283024787903,
      "epoch": 4.975250131648236,
      "grad_norm": 0.2993650436401367,
      "learning_rate": 5.670626714014925e-05,
      "loss": 0.15203902125358582,
      "mean_token_accuracy": 0.9337811768054962,
      "num_tokens": 116546416.0,
      "step": 9448
    },
    {
      "entropy": 0.9600692093372345,
      "epoch": 4.97577672459189,
      "grad_norm": 0.3401379883289337,
      "learning_rate": 5.6688598341118124e-05,
      "loss": 0.1808190643787384,
      "mean_token_accuracy": 0.9247889965772629,
      "num_tokens": 116558752.0,
      "step": 9449
    },
    {
      "entropy": 0.8913252055644989,
      "epoch": 4.976303317535545,
      "grad_norm": 0.3390532433986664,
      "learning_rate": 5.667093270687418e-05,
      "loss": 0.16844721138477325,
      "mean_token_accuracy": 0.9272169321775436,
      "num_tokens": 116571088.0,
      "step": 9450
    },
    {
      "entropy": 0.9043051451444626,
      "epoch": 4.9768299104792,
      "grad_norm": 0.333987832069397,
      "learning_rate": 5.665327023846618e-05,
      "loss": 0.16089186072349548,
      "mean_token_accuracy": 0.9296596944332123,
      "num_tokens": 116583424.0,
      "step": 9451
    },
    {
      "entropy": 0.8988510817289352,
      "epoch": 4.977356503422854,
      "grad_norm": 0.3078000545501709,
      "learning_rate": 5.6635610936942565e-05,
      "loss": 0.15991199016571045,
      "mean_token_accuracy": 0.9318134486675262,
      "num_tokens": 116595760.0,
      "step": 9452
    },
    {
      "entropy": 0.9012234061956406,
      "epoch": 4.977883096366509,
      "grad_norm": 0.3006126880645752,
      "learning_rate": 5.66179548033517e-05,
      "loss": 0.16226999461650848,
      "mean_token_accuracy": 0.9322975128889084,
      "num_tokens": 116608096.0,
      "step": 9453
    },
    {
      "entropy": 0.940825805068016,
      "epoch": 4.978409689310164,
      "grad_norm": 0.33715713024139404,
      "learning_rate": 5.660030183874168e-05,
      "loss": 0.1778339147567749,
      "mean_token_accuracy": 0.9256445169448853,
      "num_tokens": 116620432.0,
      "step": 9454
    },
    {
      "entropy": 0.9499392062425613,
      "epoch": 4.978936282253818,
      "grad_norm": 0.35110846161842346,
      "learning_rate": 5.6582652044160556e-05,
      "loss": 0.16285377740859985,
      "mean_token_accuracy": 0.9258364588022232,
      "num_tokens": 116632768.0,
      "step": 9455
    },
    {
      "entropy": 0.9032035619020462,
      "epoch": 4.9794628751974725,
      "grad_norm": 0.347312331199646,
      "learning_rate": 5.656500542065603e-05,
      "loss": 0.19477276504039764,
      "mean_token_accuracy": 0.9145398288965225,
      "num_tokens": 116645104.0,
      "step": 9456
    },
    {
      "entropy": 0.905480682849884,
      "epoch": 4.9799894681411265,
      "grad_norm": 0.31421568989753723,
      "learning_rate": 5.654736196927568e-05,
      "loss": 0.16706067323684692,
      "mean_token_accuracy": 0.9269927889108658,
      "num_tokens": 116657440.0,
      "step": 9457
    },
    {
      "entropy": 0.9130947887897491,
      "epoch": 4.980516061084781,
      "grad_norm": 0.316919207572937,
      "learning_rate": 5.652972169106692e-05,
      "loss": 0.172123521566391,
      "mean_token_accuracy": 0.9237474501132965,
      "num_tokens": 116669776.0,
      "step": 9458
    },
    {
      "entropy": 0.9133815616369247,
      "epoch": 4.981042654028436,
      "grad_norm": 0.3086302876472473,
      "learning_rate": 5.6512084587076916e-05,
      "loss": 0.16468968987464905,
      "mean_token_accuracy": 0.9326214045286179,
      "num_tokens": 116682112.0,
      "step": 9459
    },
    {
      "entropy": 0.9098835438489914,
      "epoch": 4.98156924697209,
      "grad_norm": 0.30872488021850586,
      "learning_rate": 5.64944506583527e-05,
      "loss": 0.17147888243198395,
      "mean_token_accuracy": 0.9259092956781387,
      "num_tokens": 116694448.0,
      "step": 9460
    },
    {
      "entropy": 0.9035718142986298,
      "epoch": 4.982095839915745,
      "grad_norm": 0.3328627347946167,
      "learning_rate": 5.6476819905941114e-05,
      "loss": 0.18064668774604797,
      "mean_token_accuracy": 0.9239556342363358,
      "num_tokens": 116706784.0,
      "step": 9461
    },
    {
      "entropy": 0.8842890411615372,
      "epoch": 4.9826224328594,
      "grad_norm": 0.294079452753067,
      "learning_rate": 5.645919233088877e-05,
      "loss": 0.15284819900989532,
      "mean_token_accuracy": 0.9346184283494949,
      "num_tokens": 116719120.0,
      "step": 9462
    },
    {
      "entropy": 0.9337745606899261,
      "epoch": 4.983149025803054,
      "grad_norm": 0.35992324352264404,
      "learning_rate": 5.64415679342421e-05,
      "loss": 0.18033114075660706,
      "mean_token_accuracy": 0.9219166487455368,
      "num_tokens": 116731456.0,
      "step": 9463
    },
    {
      "entropy": 0.8690318018198013,
      "epoch": 4.983675618746709,
      "grad_norm": 0.32860904932022095,
      "learning_rate": 5.6423946717047406e-05,
      "loss": 0.16652944684028625,
      "mean_token_accuracy": 0.9334118962287903,
      "num_tokens": 116743792.0,
      "step": 9464
    },
    {
      "entropy": 0.8971494138240814,
      "epoch": 4.984202211690364,
      "grad_norm": 0.29319801926612854,
      "learning_rate": 5.6406328680350706e-05,
      "loss": 0.1558331549167633,
      "mean_token_accuracy": 0.934053510427475,
      "num_tokens": 116756128.0,
      "step": 9465
    },
    {
      "entropy": 0.8544014245271683,
      "epoch": 4.984728804634018,
      "grad_norm": 0.33784765005111694,
      "learning_rate": 5.6388713825197915e-05,
      "loss": 0.1779947578907013,
      "mean_token_accuracy": 0.9300804436206818,
      "num_tokens": 116768464.0,
      "step": 9466
    },
    {
      "entropy": 0.9156675636768341,
      "epoch": 4.9852553975776726,
      "grad_norm": 0.3178240954875946,
      "learning_rate": 5.6371102152634705e-05,
      "loss": 0.15659943222999573,
      "mean_token_accuracy": 0.9356032758951187,
      "num_tokens": 116780800.0,
      "step": 9467
    },
    {
      "entropy": 0.8807658702135086,
      "epoch": 4.9857819905213265,
      "grad_norm": 0.3392799198627472,
      "learning_rate": 5.6353493663706567e-05,
      "loss": 0.17571623623371124,
      "mean_token_accuracy": 0.9243208914995193,
      "num_tokens": 116793136.0,
      "step": 9468
    },
    {
      "entropy": 0.9186313897371292,
      "epoch": 4.986308583464981,
      "grad_norm": 0.3488842844963074,
      "learning_rate": 5.6335888359458866e-05,
      "loss": 0.16939771175384521,
      "mean_token_accuracy": 0.926127165555954,
      "num_tokens": 116805472.0,
      "step": 9469
    },
    {
      "entropy": 0.9254487901926041,
      "epoch": 4.986835176408636,
      "grad_norm": 0.31008264422416687,
      "learning_rate": 5.6318286240936713e-05,
      "loss": 0.16387975215911865,
      "mean_token_accuracy": 0.9316154420375824,
      "num_tokens": 116817808.0,
      "step": 9470
    },
    {
      "entropy": 0.9040963053703308,
      "epoch": 4.98736176935229,
      "grad_norm": 0.32900792360305786,
      "learning_rate": 5.630068730918501e-05,
      "loss": 0.16915391385555267,
      "mean_token_accuracy": 0.928293913602829,
      "num_tokens": 116830144.0,
      "step": 9471
    },
    {
      "entropy": 0.9082885682582855,
      "epoch": 4.987888362295945,
      "grad_norm": 0.3768515884876251,
      "learning_rate": 5.6283091565248517e-05,
      "loss": 0.18894949555397034,
      "mean_token_accuracy": 0.9235462546348572,
      "num_tokens": 116842480.0,
      "step": 9472
    },
    {
      "entropy": 0.9041440486907959,
      "epoch": 4.9884149552396,
      "grad_norm": 0.3412855863571167,
      "learning_rate": 5.626549901017177e-05,
      "loss": 0.16429667174816132,
      "mean_token_accuracy": 0.9322096705436707,
      "num_tokens": 116854816.0,
      "step": 9473
    },
    {
      "entropy": 0.91483274102211,
      "epoch": 4.988941548183254,
      "grad_norm": 0.3402027487754822,
      "learning_rate": 5.6247909644999176e-05,
      "loss": 0.1718006283044815,
      "mean_token_accuracy": 0.9307664930820465,
      "num_tokens": 116867152.0,
      "step": 9474
    },
    {
      "entropy": 0.9492096453905106,
      "epoch": 4.989468141126909,
      "grad_norm": 0.3414507806301117,
      "learning_rate": 5.623032347077487e-05,
      "loss": 0.1570313274860382,
      "mean_token_accuracy": 0.9314261972904205,
      "num_tokens": 116879488.0,
      "step": 9475
    },
    {
      "entropy": 0.9157185554504395,
      "epoch": 4.989994734070564,
      "grad_norm": 0.31004396080970764,
      "learning_rate": 5.621274048854289e-05,
      "loss": 0.15961430966854095,
      "mean_token_accuracy": 0.9323617666959763,
      "num_tokens": 116891824.0,
      "step": 9476
    },
    {
      "entropy": 0.9423941224813461,
      "epoch": 4.990521327014218,
      "grad_norm": 0.35401451587677,
      "learning_rate": 5.6195160699347004e-05,
      "loss": 0.17788419127464294,
      "mean_token_accuracy": 0.9256136417388916,
      "num_tokens": 116904160.0,
      "step": 9477
    },
    {
      "entropy": 0.9306418746709824,
      "epoch": 4.991047919957873,
      "grad_norm": 0.32740211486816406,
      "learning_rate": 5.617758410423084e-05,
      "loss": 0.15795734524726868,
      "mean_token_accuracy": 0.9341577291488647,
      "num_tokens": 116916496.0,
      "step": 9478
    },
    {
      "entropy": 0.9015368968248367,
      "epoch": 4.991574512901527,
      "grad_norm": 0.36364123225212097,
      "learning_rate": 5.61600107042378e-05,
      "loss": 0.18496762216091156,
      "mean_token_accuracy": 0.9203061461448669,
      "num_tokens": 116928832.0,
      "step": 9479
    },
    {
      "entropy": 0.9284609705209732,
      "epoch": 4.992101105845181,
      "grad_norm": 0.31846433877944946,
      "learning_rate": 5.614244050041113e-05,
      "loss": 0.17823076248168945,
      "mean_token_accuracy": 0.9223919361829758,
      "num_tokens": 116941168.0,
      "step": 9480
    },
    {
      "entropy": 0.9006500244140625,
      "epoch": 4.992627698788836,
      "grad_norm": 0.34518757462501526,
      "learning_rate": 5.6124873493793884e-05,
      "loss": 0.17152202129364014,
      "mean_token_accuracy": 0.9256195574998856,
      "num_tokens": 116953504.0,
      "step": 9481
    },
    {
      "entropy": 0.933140829205513,
      "epoch": 4.993154291732491,
      "grad_norm": 0.3343666195869446,
      "learning_rate": 5.6107309685428834e-05,
      "loss": 0.17478488385677338,
      "mean_token_accuracy": 0.9261410385370255,
      "num_tokens": 116965840.0,
      "step": 9482
    },
    {
      "entropy": 0.9274118393659592,
      "epoch": 4.993680884676145,
      "grad_norm": 0.30102184414863586,
      "learning_rate": 5.608974907635872e-05,
      "loss": 0.15346240997314453,
      "mean_token_accuracy": 0.9328440576791763,
      "num_tokens": 116978176.0,
      "step": 9483
    },
    {
      "entropy": 0.9209228157997131,
      "epoch": 4.9942074776198,
      "grad_norm": 0.30850332975387573,
      "learning_rate": 5.6072191667626006e-05,
      "loss": 0.1682778298854828,
      "mean_token_accuracy": 0.9296342134475708,
      "num_tokens": 116990512.0,
      "step": 9484
    },
    {
      "entropy": 0.9228739440441132,
      "epoch": 4.994734070563455,
      "grad_norm": 0.33433517813682556,
      "learning_rate": 5.605463746027294e-05,
      "loss": 0.1794205904006958,
      "mean_token_accuracy": 0.9201393127441406,
      "num_tokens": 117002848.0,
      "step": 9485
    },
    {
      "entropy": 0.8917040377855301,
      "epoch": 4.995260663507109,
      "grad_norm": 0.30780935287475586,
      "learning_rate": 5.603708645534165e-05,
      "loss": 0.1697072684764862,
      "mean_token_accuracy": 0.9269426912069321,
      "num_tokens": 117015184.0,
      "step": 9486
    },
    {
      "entropy": 0.9531418681144714,
      "epoch": 4.995787256450764,
      "grad_norm": 0.3103603422641754,
      "learning_rate": 5.6019538653873994e-05,
      "loss": 0.17060068249702454,
      "mean_token_accuracy": 0.9268237501382828,
      "num_tokens": 117027520.0,
      "step": 9487
    },
    {
      "entropy": 0.9021676480770111,
      "epoch": 4.996313849394419,
      "grad_norm": 0.32364600896835327,
      "learning_rate": 5.600199405691171e-05,
      "loss": 0.16891726851463318,
      "mean_token_accuracy": 0.9244484603404999,
      "num_tokens": 117039856.0,
      "step": 9488
    },
    {
      "entropy": 0.9253650903701782,
      "epoch": 4.996840442338073,
      "grad_norm": 0.31460437178611755,
      "learning_rate": 5.5984452665496346e-05,
      "loss": 0.15997791290283203,
      "mean_token_accuracy": 0.9296991527080536,
      "num_tokens": 117052192.0,
      "step": 9489
    },
    {
      "entropy": 0.9130527228116989,
      "epoch": 4.997367035281727,
      "grad_norm": 0.3126857876777649,
      "learning_rate": 5.596691448066916e-05,
      "loss": 0.17209011316299438,
      "mean_token_accuracy": 0.9269836246967316,
      "num_tokens": 117064528.0,
      "step": 9490
    },
    {
      "entropy": 0.9370010644197464,
      "epoch": 4.997893628225381,
      "grad_norm": 0.314369797706604,
      "learning_rate": 5.594937950347137e-05,
      "loss": 0.16863462328910828,
      "mean_token_accuracy": 0.9340494722127914,
      "num_tokens": 117076864.0,
      "step": 9491
    },
    {
      "entropy": 0.9399919509887695,
      "epoch": 4.998420221169036,
      "grad_norm": 0.3283589780330658,
      "learning_rate": 5.5931847734943865e-05,
      "loss": 0.18404729664325714,
      "mean_token_accuracy": 0.9177678525447845,
      "num_tokens": 117089200.0,
      "step": 9492
    },
    {
      "entropy": 0.9053280353546143,
      "epoch": 4.998946814112691,
      "grad_norm": 0.31343698501586914,
      "learning_rate": 5.5914319176127416e-05,
      "loss": 0.16595865786075592,
      "mean_token_accuracy": 0.925357311964035,
      "num_tokens": 117101536.0,
      "step": 9493
    },
    {
      "entropy": 0.8777082115411758,
      "epoch": 4.999473407056345,
      "grad_norm": 0.28059449791908264,
      "learning_rate": 5.589679382806262e-05,
      "loss": 0.15285533666610718,
      "mean_token_accuracy": 0.9359864443540573,
      "num_tokens": 117113872.0,
      "step": 9494
    },
    {
      "entropy": 0.8763579279184341,
      "epoch": 5.0,
      "grad_norm": 0.3026115596294403,
      "learning_rate": 5.587927169178984e-05,
      "loss": 0.1645173877477646,
      "mean_token_accuracy": 0.9272801280021667,
      "num_tokens": 117125180.0,
      "step": 9495
    },
    {
      "epoch": 5.0,
      "eval_entropy": 0.8973508889327368,
      "eval_loss": 0.19531196355819702,
      "eval_mean_token_accuracy": 0.9202164551532539,
      "eval_num_tokens": 117125180.0,
      "eval_runtime": 665.3223,
      "eval_samples_per_second": 8.563,
      "eval_steps_per_second": 2.854,
      "step": 9495
    }
  ],
  "logging_steps": 1,
  "max_steps": 13293,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.173011364224306e+19,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}