{
  "best_global_step": 43000,
  "best_metric": 1.4149765968322754,
  "best_model_checkpoint": "/home/mkurman/gitlab/ai/model-trainer/results/convgpt_v2_long_2d/convgpt_v2_rope_nope_hybrid_clean_h512_l16_seq512_bs8_fullloss_20260523_161605/checkpoint-43000",
  "epoch": 0.26994285112322564,
  "eval_steps": 1000,
  "global_step": 114000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00013167943957230517,
      "grad_norm": 1.5129085779190063,
      "learning_rate": 1.8762343647136274e-05,
      "loss": 10.0451,
      "step": 20
    },
    {
      "epoch": 0.00026335887914461035,
      "grad_norm": 1.4276129007339478,
      "learning_rate": 3.851217906517446e-05,
      "loss": 9.11,
      "step": 40
    },
    {
      "epoch": 0.00039503831871691555,
      "grad_norm": 1.3658931255340576,
      "learning_rate": 5.8262014483212636e-05,
      "loss": 8.3876,
      "step": 60
    },
    {
      "epoch": 0.0005267177582892207,
      "grad_norm": 1.8227916955947876,
      "learning_rate": 7.801184990125083e-05,
      "loss": 7.7517,
      "step": 80
    },
    {
      "epoch": 0.0006583971978615259,
      "grad_norm": 0.8309405446052551,
      "learning_rate": 9.7761685319289e-05,
      "loss": 7.1243,
      "step": 100
    },
    {
      "epoch": 0.0007900766374338311,
      "grad_norm": 0.6047381162643433,
      "learning_rate": 0.00011751152073732718,
      "loss": 6.5642,
      "step": 120
    },
    {
      "epoch": 0.0009217560770061363,
      "grad_norm": 6.724730014801025,
      "learning_rate": 0.00013726135615536539,
      "loss": 6.1814,
      "step": 140
    },
    {
      "epoch": 0.0010534355165784414,
      "grad_norm": 0.4805045425891876,
      "learning_rate": 0.00015701119157340358,
      "loss": 5.8953,
      "step": 160
    },
    {
      "epoch": 0.0011851149561507466,
      "grad_norm": 0.4253923296928406,
      "learning_rate": 0.00017676102699144174,
      "loss": 5.656,
      "step": 180
    },
    {
      "epoch": 0.0013167943957230518,
      "grad_norm": 0.479189395904541,
      "learning_rate": 0.0001965108624094799,
      "loss": 5.453,
      "step": 200
    },
    {
      "epoch": 0.001448473835295357,
      "grad_norm": 0.4821586310863495,
      "learning_rate": 0.0002162606978275181,
      "loss": 5.2631,
      "step": 220
    },
    {
      "epoch": 0.0015801532748676622,
      "grad_norm": 0.5253534913063049,
      "learning_rate": 0.0002360105332455563,
      "loss": 5.1262,
      "step": 240
    },
    {
      "epoch": 0.0017118327144399674,
      "grad_norm": 0.3778499662876129,
      "learning_rate": 0.0002557603686635945,
      "loss": 4.9729,
      "step": 260
    },
    {
      "epoch": 0.0018435121540122726,
      "grad_norm": 0.40282198786735535,
      "learning_rate": 0.0002755102040816327,
      "loss": 4.8491,
      "step": 280
    },
    {
      "epoch": 0.0019751915935845776,
      "grad_norm": 0.40169110894203186,
      "learning_rate": 0.00029526003949967083,
      "loss": 4.7163,
      "step": 300
    },
    {
      "epoch": 0.002106871033156883,
      "grad_norm": 0.4091223180294037,
      "learning_rate": 0.000315009874917709,
      "loss": 4.6122,
      "step": 320
    },
    {
      "epoch": 0.002238550472729188,
      "grad_norm": 0.3836401402950287,
      "learning_rate": 0.0003347597103357472,
      "loss": 4.5315,
      "step": 340
    },
    {
      "epoch": 0.002370229912301493,
      "grad_norm": 0.37077072262763977,
      "learning_rate": 0.0003545095457537854,
      "loss": 4.4403,
      "step": 360
    },
    {
      "epoch": 0.0025019093518737984,
      "grad_norm": 0.3423950970172882,
      "learning_rate": 0.0003742593811718236,
      "loss": 4.3758,
      "step": 380
    },
    {
      "epoch": 0.0026335887914461036,
      "grad_norm": 0.3271167576313019,
      "learning_rate": 0.00039400921658986173,
      "loss": 4.2859,
      "step": 400
    },
    {
      "epoch": 0.002765268231018409,
      "grad_norm": 0.32271647453308105,
      "learning_rate": 0.0004137590520079,
      "loss": 4.2281,
      "step": 420
    },
    {
      "epoch": 0.002896947670590714,
      "grad_norm": 0.3268634080886841,
      "learning_rate": 0.00043350888742593816,
      "loss": 4.152,
      "step": 440
    },
    {
      "epoch": 0.003028627110163019,
      "grad_norm": 0.3095104992389679,
      "learning_rate": 0.00045325872284397636,
      "loss": 4.0894,
      "step": 460
    },
    {
      "epoch": 0.0031603065497353244,
      "grad_norm": 0.33799174427986145,
      "learning_rate": 0.00047300855826201455,
      "loss": 4.0365,
      "step": 480
    },
    {
      "epoch": 0.0032919859893076296,
      "grad_norm": 0.39443933963775635,
      "learning_rate": 0.0004927583936800527,
      "loss": 3.9868,
      "step": 500
    },
    {
      "epoch": 0.003423665428879935,
      "grad_norm": 0.32354313135147095,
      "learning_rate": 0.0005125082290980909,
      "loss": 3.9136,
      "step": 520
    },
    {
      "epoch": 0.00355534486845224,
      "grad_norm": 0.31924670934677124,
      "learning_rate": 0.0005322580645161291,
      "loss": 3.8859,
      "step": 540
    },
    {
      "epoch": 0.0036870243080245452,
      "grad_norm": 0.2883984446525574,
      "learning_rate": 0.0005520078999341672,
      "loss": 3.8321,
      "step": 560
    },
    {
      "epoch": 0.0038187037475968504,
      "grad_norm": 0.31237614154815674,
      "learning_rate": 0.0005717577353522054,
      "loss": 3.799,
      "step": 580
    },
    {
      "epoch": 0.003950383187169155,
      "grad_norm": 0.29551151394844055,
      "learning_rate": 0.0005915075707702436,
      "loss": 3.7745,
      "step": 600
    },
    {
      "epoch": 0.004082062626741461,
      "grad_norm": 0.2833958566188812,
      "learning_rate": 0.0006112574061882818,
      "loss": 3.7411,
      "step": 620
    },
    {
      "epoch": 0.004213742066313766,
      "grad_norm": 0.28037381172180176,
      "learning_rate": 0.00063100724160632,
      "loss": 3.6861,
      "step": 640
    },
    {
      "epoch": 0.004345421505886071,
      "grad_norm": 0.39652886986732483,
      "learning_rate": 0.0006507570770243582,
      "loss": 3.6568,
      "step": 660
    },
    {
      "epoch": 0.004477100945458376,
      "grad_norm": 0.2646957337856293,
      "learning_rate": 0.0006705069124423963,
      "loss": 3.6185,
      "step": 680
    },
    {
      "epoch": 0.004608780385030682,
      "grad_norm": 0.2843409478664398,
      "learning_rate": 0.0006902567478604345,
      "loss": 3.5845,
      "step": 700
    },
    {
      "epoch": 0.004740459824602986,
      "grad_norm": 0.27913448214530945,
      "learning_rate": 0.0007100065832784727,
      "loss": 3.5546,
      "step": 720
    },
    {
      "epoch": 0.004872139264175292,
      "grad_norm": 0.2710491418838501,
      "learning_rate": 0.0007297564186965109,
      "loss": 3.5176,
      "step": 740
    },
    {
      "epoch": 0.005003818703747597,
      "grad_norm": 0.28077027201652527,
      "learning_rate": 0.0007495062541145491,
      "loss": 3.497,
      "step": 760
    },
    {
      "epoch": 0.0051354981433199024,
      "grad_norm": 0.27605777978897095,
      "learning_rate": 0.0007692560895325872,
      "loss": 3.4649,
      "step": 780
    },
    {
      "epoch": 0.005267177582892207,
      "grad_norm": 0.29779553413391113,
      "learning_rate": 0.0007890059249506255,
      "loss": 3.4509,
      "step": 800
    },
    {
      "epoch": 0.005398857022464512,
      "grad_norm": 0.3055330812931061,
      "learning_rate": 0.0008087557603686636,
      "loss": 3.4085,
      "step": 820
    },
    {
      "epoch": 0.005530536462036818,
      "grad_norm": 0.26349249482154846,
      "learning_rate": 0.0008285055957867019,
      "loss": 3.3842,
      "step": 840
    },
    {
      "epoch": 0.005662215901609122,
      "grad_norm": 0.2677903175354004,
      "learning_rate": 0.00084825543120474,
      "loss": 3.3555,
      "step": 860
    },
    {
      "epoch": 0.005793895341181428,
      "grad_norm": 0.2856537699699402,
      "learning_rate": 0.0008680052666227783,
      "loss": 3.3358,
      "step": 880
    },
    {
      "epoch": 0.005925574780753733,
      "grad_norm": 0.2631045877933502,
      "learning_rate": 0.0008877551020408163,
      "loss": 3.3164,
      "step": 900
    },
    {
      "epoch": 0.006057254220326038,
      "grad_norm": 0.2708953619003296,
      "learning_rate": 0.0009075049374588546,
      "loss": 3.2944,
      "step": 920
    },
    {
      "epoch": 0.006188933659898343,
      "grad_norm": 0.24445629119873047,
      "learning_rate": 0.0009272547728768927,
      "loss": 3.2718,
      "step": 940
    },
    {
      "epoch": 0.006320613099470649,
      "grad_norm": 0.2704775035381317,
      "learning_rate": 0.0009470046082949309,
      "loss": 3.249,
      "step": 960
    },
    {
      "epoch": 0.006452292539042954,
      "grad_norm": 0.25395041704177856,
      "learning_rate": 0.000966754443712969,
      "loss": 3.2161,
      "step": 980
    },
    {
      "epoch": 0.006583971978615259,
      "grad_norm": 0.2632465064525604,
      "learning_rate": 0.0009865042791310074,
      "loss": 3.1973,
      "step": 1000
    },
    {
      "epoch": 0.006583971978615259,
      "eval_loss": 3.229322671890259,
      "eval_runtime": 68.051,
      "eval_samples_per_second": 14.695,
      "eval_steps_per_second": 14.695,
      "step": 1000
    },
    {
      "epoch": 0.006715651418187564,
      "grad_norm": 0.268993079662323,
      "learning_rate": 0.0010062541145490454,
      "loss": 3.1752,
      "step": 1020
    },
    {
      "epoch": 0.00684733085775987,
      "grad_norm": 0.2427850067615509,
      "learning_rate": 0.0010260039499670837,
      "loss": 3.1407,
      "step": 1040
    },
    {
      "epoch": 0.006979010297332174,
      "grad_norm": 0.2634700834751129,
      "learning_rate": 0.0010457537853851217,
      "loss": 3.122,
      "step": 1060
    },
    {
      "epoch": 0.00711068973690448,
      "grad_norm": 0.23194676637649536,
      "learning_rate": 0.00106550362080316,
      "loss": 3.0935,
      "step": 1080
    },
    {
      "epoch": 0.007242369176476785,
      "grad_norm": 0.25881442427635193,
      "learning_rate": 0.0010852534562211982,
      "loss": 3.0758,
      "step": 1100
    },
    {
      "epoch": 0.0073740486160490904,
      "grad_norm": 0.2772982716560364,
      "learning_rate": 0.0011050032916392364,
      "loss": 3.0621,
      "step": 1120
    },
    {
      "epoch": 0.007505728055621395,
      "grad_norm": 0.25508975982666016,
      "learning_rate": 0.0011247531270572745,
      "loss": 3.0338,
      "step": 1140
    },
    {
      "epoch": 0.007637407495193701,
      "grad_norm": 0.2273252010345459,
      "learning_rate": 0.0011445029624753127,
      "loss": 3.0122,
      "step": 1160
    },
    {
      "epoch": 0.007769086934766006,
      "grad_norm": 0.25050970911979675,
      "learning_rate": 0.001164252797893351,
      "loss": 3.0121,
      "step": 1180
    },
    {
      "epoch": 0.00790076637433831,
      "grad_norm": 0.24488165974617004,
      "learning_rate": 0.0011840026333113892,
      "loss": 2.9993,
      "step": 1200
    },
    {
      "epoch": 0.008032445813910615,
      "grad_norm": 0.23485834896564484,
      "learning_rate": 0.0012037524687294272,
      "loss": 2.9679,
      "step": 1220
    },
    {
      "epoch": 0.008164125253482922,
      "grad_norm": 0.23405972123146057,
      "learning_rate": 0.0012235023041474655,
      "loss": 2.9485,
      "step": 1240
    },
    {
      "epoch": 0.008295804693055226,
      "grad_norm": 0.20557770133018494,
      "learning_rate": 0.0012432521395655035,
      "loss": 2.931,
      "step": 1260
    },
    {
      "epoch": 0.008427484132627531,
      "grad_norm": 0.22637082636356354,
      "learning_rate": 0.001263001974983542,
      "loss": 2.9185,
      "step": 1280
    },
    {
      "epoch": 0.008559163572199836,
      "grad_norm": 0.21511614322662354,
      "learning_rate": 0.00128275181040158,
      "loss": 2.9059,
      "step": 1300
    },
    {
      "epoch": 0.008690843011772142,
      "grad_norm": 0.20529645681381226,
      "learning_rate": 0.0013025016458196182,
      "loss": 2.8874,
      "step": 1320
    },
    {
      "epoch": 0.008822522451344447,
      "grad_norm": 0.19764812290668488,
      "learning_rate": 0.0013222514812376563,
      "loss": 2.8646,
      "step": 1340
    },
    {
      "epoch": 0.008954201890916752,
      "grad_norm": 0.2055043876171112,
      "learning_rate": 0.0013420013166556945,
      "loss": 2.849,
      "step": 1360
    },
    {
      "epoch": 0.009085881330489057,
      "grad_norm": 0.1990618258714676,
      "learning_rate": 0.0013617511520737328,
      "loss": 2.8303,
      "step": 1380
    },
    {
      "epoch": 0.009217560770061363,
      "grad_norm": 0.19160950183868408,
      "learning_rate": 0.001381500987491771,
      "loss": 2.827,
      "step": 1400
    },
    {
      "epoch": 0.009349240209633668,
      "grad_norm": 0.19032776355743408,
      "learning_rate": 0.001401250822909809,
      "loss": 2.8059,
      "step": 1420
    },
    {
      "epoch": 0.009480919649205973,
      "grad_norm": 0.18810242414474487,
      "learning_rate": 0.0014210006583278473,
      "loss": 2.8056,
      "step": 1440
    },
    {
      "epoch": 0.009612599088778278,
      "grad_norm": 0.18992900848388672,
      "learning_rate": 0.0014407504937458853,
      "loss": 2.7846,
      "step": 1460
    },
    {
      "epoch": 0.009744278528350584,
      "grad_norm": 0.18800954520702362,
      "learning_rate": 0.0014605003291639238,
      "loss": 2.7688,
      "step": 1480
    },
    {
      "epoch": 0.009875957967922889,
      "grad_norm": 0.19007734954357147,
      "learning_rate": 0.0014802501645819618,
      "loss": 2.7639,
      "step": 1500
    },
    {
      "epoch": 0.010007637407495194,
      "grad_norm": 0.18421195447444916,
      "learning_rate": 0.0015,
      "loss": 2.7567,
      "step": 1520
    },
    {
      "epoch": 0.010139316847067498,
      "grad_norm": 0.182477667927742,
      "learning_rate": 0.001519749835418038,
      "loss": 2.7508,
      "step": 1540
    },
    {
      "epoch": 0.010270996286639805,
      "grad_norm": 0.18245883285999298,
      "learning_rate": 0.0015394996708360765,
      "loss": 2.7286,
      "step": 1560
    },
    {
      "epoch": 0.01040267572621211,
      "grad_norm": 0.18001046776771545,
      "learning_rate": 0.0015592495062541148,
      "loss": 2.7224,
      "step": 1580
    },
    {
      "epoch": 0.010534355165784414,
      "grad_norm": 0.17572300136089325,
      "learning_rate": 0.0015789993416721528,
      "loss": 2.7126,
      "step": 1600
    },
    {
      "epoch": 0.01066603460535672,
      "grad_norm": 0.17919962108135223,
      "learning_rate": 0.0015987491770901908,
      "loss": 2.7043,
      "step": 1620
    },
    {
      "epoch": 0.010797714044929024,
      "grad_norm": 0.17204663157463074,
      "learning_rate": 0.0016184990125082293,
      "loss": 2.6945,
      "step": 1640
    },
    {
      "epoch": 0.01092939348450133,
      "grad_norm": 0.17742666602134705,
      "learning_rate": 0.0016382488479262673,
      "loss": 2.6981,
      "step": 1660
    },
    {
      "epoch": 0.011061072924073635,
      "grad_norm": 0.18275167047977448,
      "learning_rate": 0.0016579986833443056,
      "loss": 2.6782,
      "step": 1680
    },
    {
      "epoch": 0.01119275236364594,
      "grad_norm": 0.18246972560882568,
      "learning_rate": 0.0016777485187623436,
      "loss": 2.6805,
      "step": 1700
    },
    {
      "epoch": 0.011324431803218245,
      "grad_norm": 0.17320232093334198,
      "learning_rate": 0.001697498354180382,
      "loss": 2.6698,
      "step": 1720
    },
    {
      "epoch": 0.011456111242790551,
      "grad_norm": 0.16316188871860504,
      "learning_rate": 0.00171724818959842,
      "loss": 2.6563,
      "step": 1740
    },
    {
      "epoch": 0.011587790682362856,
      "grad_norm": 0.17399418354034424,
      "learning_rate": 0.001736998025016458,
      "loss": 2.6639,
      "step": 1760
    },
    {
      "epoch": 0.01171947012193516,
      "grad_norm": 0.16352471709251404,
      "learning_rate": 0.0017567478604344963,
      "loss": 2.6778,
      "step": 1780
    },
    {
      "epoch": 0.011851149561507466,
      "grad_norm": 0.167058527469635,
      "learning_rate": 0.0017764976958525348,
      "loss": 2.6616,
      "step": 1800
    },
    {
      "epoch": 0.011982829001079772,
      "grad_norm": 0.18267591297626495,
      "learning_rate": 0.0017962475312705728,
      "loss": 2.6714,
      "step": 1820
    },
    {
      "epoch": 0.012114508440652077,
      "grad_norm": 0.17048963904380798,
      "learning_rate": 0.0018159973666886109,
      "loss": 2.6543,
      "step": 1840
    },
    {
      "epoch": 0.012246187880224382,
      "grad_norm": 0.16198118031024933,
      "learning_rate": 0.0018357472021066491,
      "loss": 2.6529,
      "step": 1860
    },
    {
      "epoch": 0.012377867319796686,
      "grad_norm": 0.1616767942905426,
      "learning_rate": 0.0018554970375246874,
      "loss": 2.6387,
      "step": 1880
    },
    {
      "epoch": 0.012509546759368993,
      "grad_norm": 0.1541590839624405,
      "learning_rate": 0.0018752468729427256,
      "loss": 2.6181,
      "step": 1900
    },
    {
      "epoch": 0.012641226198941298,
      "grad_norm": 0.15989308059215546,
      "learning_rate": 0.0018949967083607636,
      "loss": 2.6226,
      "step": 1920
    },
    {
      "epoch": 0.012772905638513602,
      "grad_norm": 0.1667192429304123,
      "learning_rate": 0.0019147465437788017,
      "loss": 2.6192,
      "step": 1940
    },
    {
      "epoch": 0.012904585078085907,
      "grad_norm": 0.15245971083641052,
      "learning_rate": 0.0019344963791968401,
      "loss": 2.6103,
      "step": 1960
    },
    {
      "epoch": 0.013036264517658214,
      "grad_norm": 0.16327407956123352,
      "learning_rate": 0.001954246214614878,
      "loss": 2.5959,
      "step": 1980
    },
    {
      "epoch": 0.013167943957230518,
      "grad_norm": 0.1521947681903839,
      "learning_rate": 0.001973996050032916,
      "loss": 2.5968,
      "step": 2000
    },
    {
      "epoch": 0.013167943957230518,
      "eval_loss": 2.627993106842041,
      "eval_runtime": 67.8897,
      "eval_samples_per_second": 14.73,
      "eval_steps_per_second": 14.73,
      "step": 2000
    },
    {
      "epoch": 0.013299623396802823,
      "grad_norm": 0.14737296104431152,
      "learning_rate": 0.0019937458854509546,
      "loss": 2.5906,
      "step": 2020
    },
    {
      "epoch": 0.013431302836375128,
      "grad_norm": 0.1449994295835495,
      "learning_rate": 0.002013495720868993,
      "loss": 2.5751,
      "step": 2040
    },
    {
      "epoch": 0.013562982275947433,
      "grad_norm": 0.15211527049541473,
      "learning_rate": 0.002033245556287031,
      "loss": 2.5802,
      "step": 2060
    },
    {
      "epoch": 0.01369466171551974,
      "grad_norm": 0.1509827822446823,
      "learning_rate": 0.002052995391705069,
      "loss": 2.5624,
      "step": 2080
    },
    {
      "epoch": 0.013826341155092044,
      "grad_norm": 0.1424606740474701,
      "learning_rate": 0.002072745227123107,
      "loss": 2.5616,
      "step": 2100
    },
    {
      "epoch": 0.013958020594664349,
      "grad_norm": 0.1456785947084427,
      "learning_rate": 0.0020924950625411456,
      "loss": 2.5614,
      "step": 2120
    },
    {
      "epoch": 0.014089700034236654,
      "grad_norm": 0.1454639583826065,
      "learning_rate": 0.0021122448979591837,
      "loss": 2.548,
      "step": 2140
    },
    {
      "epoch": 0.01422137947380896,
      "grad_norm": 0.15996553003787994,
      "learning_rate": 0.0021319947333772217,
      "loss": 2.5564,
      "step": 2160
    },
    {
      "epoch": 0.014353058913381265,
      "grad_norm": 0.14995744824409485,
      "learning_rate": 0.0021517445687952597,
      "loss": 2.537,
      "step": 2180
    },
    {
      "epoch": 0.01448473835295357,
      "grad_norm": 0.1442878395318985,
      "learning_rate": 0.002171494404213298,
      "loss": 2.5322,
      "step": 2200
    },
    {
      "epoch": 0.014616417792525874,
      "grad_norm": 0.14572259783744812,
      "learning_rate": 0.0021912442396313367,
      "loss": 2.5248,
      "step": 2220
    },
    {
      "epoch": 0.014748097232098181,
      "grad_norm": 0.14792700111865997,
      "learning_rate": 0.0022109940750493747,
      "loss": 2.5214,
      "step": 2240
    },
    {
      "epoch": 0.014879776671670486,
      "grad_norm": 0.13958069682121277,
      "learning_rate": 0.0022307439104674127,
      "loss": 2.5081,
      "step": 2260
    },
    {
      "epoch": 0.01501145611124279,
      "grad_norm": 0.1351982206106186,
      "learning_rate": 0.002250493745885451,
      "loss": 2.5116,
      "step": 2280
    },
    {
      "epoch": 0.015143135550815095,
      "grad_norm": 0.14376430213451385,
      "learning_rate": 0.002270243581303489,
      "loss": 2.4984,
      "step": 2300
    },
    {
      "epoch": 0.015274814990387402,
      "grad_norm": 0.15098074078559875,
      "learning_rate": 0.0022899934167215272,
      "loss": 2.4982,
      "step": 2320
    },
    {
      "epoch": 0.015406494429959706,
      "grad_norm": 0.13247129321098328,
      "learning_rate": 0.0023097432521395657,
      "loss": 2.4841,
      "step": 2340
    },
    {
      "epoch": 0.015538173869532011,
      "grad_norm": 0.14014942944049835,
      "learning_rate": 0.0023294930875576037,
      "loss": 2.4941,
      "step": 2360
    },
    {
      "epoch": 0.015669853309104316,
      "grad_norm": 0.13669346272945404,
      "learning_rate": 0.0023492429229756417,
      "loss": 2.5085,
      "step": 2380
    },
    {
      "epoch": 0.01580153274867662,
      "grad_norm": 0.12908074259757996,
      "learning_rate": 0.00236899275839368,
      "loss": 2.4982,
      "step": 2400
    },
    {
      "epoch": 0.015933212188248926,
      "grad_norm": 0.12701626121997833,
      "learning_rate": 0.0023887425938117182,
      "loss": 2.4868,
      "step": 2420
    },
    {
      "epoch": 0.01606489162782123,
      "grad_norm": 0.13361623883247375,
      "learning_rate": 0.0024084924292297567,
      "loss": 2.4894,
      "step": 2440
    },
    {
      "epoch": 0.01619657106739354,
      "grad_norm": 0.12999138236045837,
      "learning_rate": 0.0024282422646477947,
      "loss": 2.4742,
      "step": 2460
    },
    {
      "epoch": 0.016328250506965843,
      "grad_norm": 0.12473666667938232,
      "learning_rate": 0.0024479921000658328,
      "loss": 2.466,
      "step": 2480
    },
    {
      "epoch": 0.016459929946538148,
      "grad_norm": 0.12962856888771057,
      "learning_rate": 0.002467741935483871,
      "loss": 2.4635,
      "step": 2500
    },
    {
      "epoch": 0.016591609386110453,
      "grad_norm": 0.13669155538082123,
      "learning_rate": 0.0024874917709019092,
      "loss": 2.4574,
      "step": 2520
    },
    {
      "epoch": 0.016723288825682758,
      "grad_norm": 0.1261315792798996,
      "learning_rate": 0.0025072416063199473,
      "loss": 2.4559,
      "step": 2540
    },
    {
      "epoch": 0.016854968265255062,
      "grad_norm": 0.1260860413312912,
      "learning_rate": 0.0025269914417379853,
      "loss": 2.4436,
      "step": 2560
    },
    {
      "epoch": 0.016986647704827367,
      "grad_norm": 0.11925692856311798,
      "learning_rate": 0.0025467412771560238,
      "loss": 2.4467,
      "step": 2580
    },
    {
      "epoch": 0.017118327144399672,
      "grad_norm": 0.12199912220239639,
      "learning_rate": 0.002566491112574062,
      "loss": 2.4395,
      "step": 2600
    },
    {
      "epoch": 0.01725000658397198,
      "grad_norm": 0.13121625781059265,
      "learning_rate": 0.0025862409479921003,
      "loss": 2.4364,
      "step": 2620
    },
    {
      "epoch": 0.017381686023544285,
      "grad_norm": 0.1242750957608223,
      "learning_rate": 0.0026059907834101383,
      "loss": 2.425,
      "step": 2640
    },
    {
      "epoch": 0.01751336546311659,
      "grad_norm": 0.11995479464530945,
      "learning_rate": 0.0026257406188281767,
      "loss": 2.4223,
      "step": 2660
    },
    {
      "epoch": 0.017645044902688894,
      "grad_norm": 0.11676183342933655,
      "learning_rate": 0.0026454904542462148,
      "loss": 2.4124,
      "step": 2680
    },
    {
      "epoch": 0.0177767243422612,
      "grad_norm": 0.11701571941375732,
      "learning_rate": 0.002665240289664253,
      "loss": 2.4044,
      "step": 2700
    },
    {
      "epoch": 0.017908403781833504,
      "grad_norm": 0.1141498014330864,
      "learning_rate": 0.002684990125082291,
      "loss": 2.4086,
      "step": 2720
    },
    {
      "epoch": 0.01804008322140581,
      "grad_norm": 0.11457623541355133,
      "learning_rate": 0.0027047399605003293,
      "loss": 2.405,
      "step": 2740
    },
    {
      "epoch": 0.018171762660978114,
      "grad_norm": 0.11170320957899094,
      "learning_rate": 0.0027244897959183673,
      "loss": 2.4041,
      "step": 2760
    },
    {
      "epoch": 0.018303442100550422,
      "grad_norm": 0.11429501324892044,
      "learning_rate": 0.0027442396313364053,
      "loss": 2.3886,
      "step": 2780
    },
    {
      "epoch": 0.018435121540122727,
      "grad_norm": 0.11575177311897278,
      "learning_rate": 0.002763989466754444,
      "loss": 2.3928,
      "step": 2800
    },
    {
      "epoch": 0.01856680097969503,
      "grad_norm": 0.11174377053976059,
      "learning_rate": 0.0027837393021724823,
      "loss": 2.3923,
      "step": 2820
    },
    {
      "epoch": 0.018698480419267336,
      "grad_norm": 0.11565430462360382,
      "learning_rate": 0.0028034891375905203,
      "loss": 2.3876,
      "step": 2840
    },
    {
      "epoch": 0.01883015985883964,
      "grad_norm": 0.1113847866654396,
      "learning_rate": 0.0028232389730085583,
      "loss": 2.3715,
      "step": 2860
    },
    {
      "epoch": 0.018961839298411946,
      "grad_norm": 0.11279511451721191,
      "learning_rate": 0.0028429888084265964,
      "loss": 2.3877,
      "step": 2880
    },
    {
      "epoch": 0.01909351873798425,
      "grad_norm": 0.10945755243301392,
      "learning_rate": 0.002862738643844635,
      "loss": 2.3713,
      "step": 2900
    },
    {
      "epoch": 0.019225198177556555,
      "grad_norm": 0.11334076523780823,
      "learning_rate": 0.002882488479262673,
      "loss": 2.3714,
      "step": 2920
    },
    {
      "epoch": 0.01935687761712886,
      "grad_norm": 0.11575334519147873,
      "learning_rate": 0.002902238314680711,
      "loss": 2.3841,
      "step": 2940
    },
    {
      "epoch": 0.019488557056701168,
      "grad_norm": 0.10688841342926025,
      "learning_rate": 0.002921988150098749,
      "loss": 2.4039,
      "step": 2960
    },
    {
      "epoch": 0.019620236496273473,
      "grad_norm": 0.10941021889448166,
      "learning_rate": 0.0029417379855167874,
      "loss": 2.4104,
      "step": 2980
    },
    {
      "epoch": 0.019751915935845778,
      "grad_norm": 0.10736847668886185,
      "learning_rate": 0.002961487820934826,
      "loss": 2.4152,
      "step": 3000
    },
    {
      "epoch": 0.019751915935845778,
      "eval_loss": 2.4375340938568115,
      "eval_runtime": 68.5796,
      "eval_samples_per_second": 14.582,
      "eval_steps_per_second": 14.582,
      "step": 3000
    },
    {
      "epoch": 0.019883595375418082,
      "grad_norm": 0.10122061520814896,
      "learning_rate": 0.002981237656352864,
      "loss": 2.4102,
      "step": 3020
    },
    {
      "epoch": 0.020015274814990387,
      "grad_norm": 0.10470600426197052,
      "learning_rate": 0.002999999999665892,
      "loss": 2.4042,
      "step": 3040
    },
    {
      "epoch": 0.020146954254562692,
      "grad_norm": 0.10254276543855667,
      "learning_rate": 0.002999999852658443,
      "loss": 2.4062,
      "step": 3060
    },
    {
      "epoch": 0.020278633694134997,
      "grad_norm": 0.10125599801540375,
      "learning_rate": 0.0029999994383647483,
      "loss": 2.3965,
      "step": 3080
    },
    {
      "epoch": 0.0204103131337073,
      "grad_norm": 0.10337845981121063,
      "learning_rate": 0.0029999987567848824,
      "loss": 2.3924,
      "step": 3100
    },
    {
      "epoch": 0.02054199257327961,
      "grad_norm": 0.09914848208427429,
      "learning_rate": 0.002999997807918967,
      "loss": 2.3903,
      "step": 3120
    },
    {
      "epoch": 0.020673672012851915,
      "grad_norm": 0.10522714257240295,
      "learning_rate": 0.0029999965917671696,
      "loss": 2.3765,
      "step": 3140
    },
    {
      "epoch": 0.02080535145242422,
      "grad_norm": 0.10708487033843994,
      "learning_rate": 0.0029999951083297086,
      "loss": 2.3739,
      "step": 3160
    },
    {
      "epoch": 0.020937030891996524,
      "grad_norm": 0.09805380553007126,
      "learning_rate": 0.002999993357606847,
      "loss": 2.3596,
      "step": 3180
    },
    {
      "epoch": 0.02106871033156883,
      "grad_norm": 0.09819849580526352,
      "learning_rate": 0.002999991339598898,
      "loss": 2.3597,
      "step": 3200
    },
    {
      "epoch": 0.021200389771141134,
      "grad_norm": 0.09890013933181763,
      "learning_rate": 0.0029999890543062208,
      "loss": 2.3504,
      "step": 3220
    },
    {
      "epoch": 0.02133206921071344,
      "grad_norm": 0.09736858308315277,
      "learning_rate": 0.002999986501729222,
      "loss": 2.3472,
      "step": 3240
    },
    {
      "epoch": 0.021463748650285743,
      "grad_norm": 0.09881199896335602,
      "learning_rate": 0.002999983681868357,
      "loss": 2.3522,
      "step": 3260
    },
    {
      "epoch": 0.021595428089858048,
      "grad_norm": 0.09606959670782089,
      "learning_rate": 0.0029999805947241278,
      "loss": 2.3455,
      "step": 3280
    },
    {
      "epoch": 0.021727107529430356,
      "grad_norm": 0.09520818293094635,
      "learning_rate": 0.0029999772402970856,
      "loss": 2.3422,
      "step": 3300
    },
    {
      "epoch": 0.02185878696900266,
      "grad_norm": 0.09379695355892181,
      "learning_rate": 0.0029999736185878268,
      "loss": 2.3387,
      "step": 3320
    },
    {
      "epoch": 0.021990466408574966,
      "grad_norm": 0.0959760919213295,
      "learning_rate": 0.0029999697295969977,
      "loss": 2.3281,
      "step": 3340
    },
    {
      "epoch": 0.02212214584814727,
      "grad_norm": 0.0942387580871582,
      "learning_rate": 0.002999965573325291,
      "loss": 2.3259,
      "step": 3360
    },
    {
      "epoch": 0.022253825287719575,
      "grad_norm": 0.09221373498439789,
      "learning_rate": 0.002999961149773447,
      "loss": 2.3186,
      "step": 3380
    },
    {
      "epoch": 0.02238550472729188,
      "grad_norm": 0.08974644541740417,
      "learning_rate": 0.0029999564589422542,
      "loss": 2.3115,
      "step": 3400
    },
    {
      "epoch": 0.022517184166864185,
      "grad_norm": 0.09186951071023941,
      "learning_rate": 0.0029999515008325484,
      "loss": 2.3165,
      "step": 3420
    },
    {
      "epoch": 0.02264886360643649,
      "grad_norm": 0.08975733816623688,
      "learning_rate": 0.002999946275445213,
      "loss": 2.3027,
      "step": 3440
    },
    {
      "epoch": 0.022780543046008798,
      "grad_norm": 0.08940092474222183,
      "learning_rate": 0.0029999407827811797,
      "loss": 2.3092,
      "step": 3460
    },
    {
      "epoch": 0.022912222485581103,
      "grad_norm": 0.08854611217975616,
      "learning_rate": 0.0029999350228414262,
      "loss": 2.3091,
      "step": 3480
    },
    {
      "epoch": 0.023043901925153407,
      "grad_norm": 0.10679975897073746,
      "learning_rate": 0.0029999289956269804,
      "loss": 2.2952,
      "step": 3500
    },
    {
      "epoch": 0.023175581364725712,
      "grad_norm": 0.08808766305446625,
      "learning_rate": 0.002999922701138914,
      "loss": 2.294,
      "step": 3520
    },
    {
      "epoch": 0.023307260804298017,
      "grad_norm": 0.0875706821680069,
      "learning_rate": 0.0029999161393783513,
      "loss": 2.2998,
      "step": 3540
    },
    {
      "epoch": 0.02343894024387032,
      "grad_norm": 0.0930817723274231,
      "learning_rate": 0.0029999093103464593,
      "loss": 2.302,
      "step": 3560
    },
    {
      "epoch": 0.023570619683442626,
      "grad_norm": 0.08741556853055954,
      "learning_rate": 0.0029999022140444562,
      "loss": 2.2977,
      "step": 3580
    },
    {
      "epoch": 0.02370229912301493,
      "grad_norm": 0.08739639818668365,
      "learning_rate": 0.002999894850473606,
      "loss": 2.2979,
      "step": 3600
    },
    {
      "epoch": 0.023833978562587236,
      "grad_norm": 0.08801335841417313,
      "learning_rate": 0.0029998872196352207,
      "loss": 2.2972,
      "step": 3620
    },
    {
      "epoch": 0.023965658002159544,
      "grad_norm": 0.08532336354255676,
      "learning_rate": 0.0029998793215306606,
      "loss": 2.2853,
      "step": 3640
    },
    {
      "epoch": 0.02409733744173185,
      "grad_norm": 0.08780607581138611,
      "learning_rate": 0.0029998711561613324,
      "loss": 2.2747,
      "step": 3660
    },
    {
      "epoch": 0.024229016881304154,
      "grad_norm": 0.08572492003440857,
      "learning_rate": 0.0029998627235286917,
      "loss": 2.2817,
      "step": 3680
    },
    {
      "epoch": 0.02436069632087646,
      "grad_norm": 0.0868530198931694,
      "learning_rate": 0.0029998540236342405,
      "loss": 2.2666,
      "step": 3700
    },
    {
      "epoch": 0.024492375760448763,
      "grad_norm": 0.0859028622508049,
      "learning_rate": 0.00299984505647953,
      "loss": 2.2736,
      "step": 3720
    },
    {
      "epoch": 0.024624055200021068,
      "grad_norm": 0.08333039283752441,
      "learning_rate": 0.0029998358220661566,
      "loss": 2.264,
      "step": 3740
    },
    {
      "epoch": 0.024755734639593373,
      "grad_norm": 0.08496104925870895,
      "learning_rate": 0.002999826320395767,
      "loss": 2.2508,
      "step": 3760
    },
    {
      "epoch": 0.024887414079165678,
      "grad_norm": 0.08279310911893845,
      "learning_rate": 0.002999816551470054,
      "loss": 2.2428,
      "step": 3780
    },
    {
      "epoch": 0.025019093518737986,
      "grad_norm": 0.08103163540363312,
      "learning_rate": 0.0029998065152907582,
      "loss": 2.2534,
      "step": 3800
    },
    {
      "epoch": 0.02515077295831029,
      "grad_norm": 0.08208036422729492,
      "learning_rate": 0.002999796211859668,
      "loss": 2.244,
      "step": 3820
    },
    {
      "epoch": 0.025282452397882595,
      "grad_norm": 0.08271029591560364,
      "learning_rate": 0.0029997856411786194,
      "loss": 2.2451,
      "step": 3840
    },
    {
      "epoch": 0.0254141318374549,
      "grad_norm": 0.08439147472381592,
      "learning_rate": 0.002999774803249496,
      "loss": 2.2411,
      "step": 3860
    },
    {
      "epoch": 0.025545811277027205,
      "grad_norm": 0.08189237862825394,
      "learning_rate": 0.0029997636980742294,
      "loss": 2.2461,
      "step": 3880
    },
    {
      "epoch": 0.02567749071659951,
      "grad_norm": 0.08077215403318405,
      "learning_rate": 0.002999752325654798,
      "loss": 2.2334,
      "step": 3900
    },
    {
      "epoch": 0.025809170156171814,
      "grad_norm": 0.07819052785634995,
      "learning_rate": 0.002999740685993228,
      "loss": 2.2288,
      "step": 3920
    },
    {
      "epoch": 0.02594084959574412,
      "grad_norm": 0.08109986782073975,
      "learning_rate": 0.002999728779091594,
      "loss": 2.2333,
      "step": 3940
    },
    {
      "epoch": 0.026072529035316427,
      "grad_norm": 0.0791846290230751,
      "learning_rate": 0.0029997166049520172,
      "loss": 2.2197,
      "step": 3960
    },
    {
      "epoch": 0.026204208474888732,
      "grad_norm": 0.08189737796783447,
      "learning_rate": 0.0029997041635766676,
      "loss": 2.2204,
      "step": 3980
    },
    {
      "epoch": 0.026335887914461037,
      "grad_norm": 0.08211075514554977,
      "learning_rate": 0.0029996914549677615,
      "loss": 2.2199,
      "step": 4000
    },
    {
      "epoch": 0.026335887914461037,
      "eval_loss": 2.2871885299682617,
      "eval_runtime": 67.7873,
      "eval_samples_per_second": 14.752,
      "eval_steps_per_second": 14.752,
      "step": 4000
    },
    {
      "epoch": 0.02646756735403334,
      "grad_norm": 0.08011506497859955,
      "learning_rate": 0.0029996784791275636,
      "loss": 2.211,
      "step": 4020
    },
    {
      "epoch": 0.026599246793605646,
      "grad_norm": 0.08271001279354095,
      "learning_rate": 0.0029996652360583863,
      "loss": 2.2097,
      "step": 4040
    },
    {
      "epoch": 0.02673092623317795,
      "grad_norm": 0.08106506615877151,
      "learning_rate": 0.0029996517257625892,
      "loss": 2.2117,
      "step": 4060
    },
    {
      "epoch": 0.026862605672750256,
      "grad_norm": 0.08120162785053253,
      "learning_rate": 0.0029996379482425803,
      "loss": 2.2011,
      "step": 4080
    },
    {
      "epoch": 0.02699428511232256,
      "grad_norm": 0.07838305830955505,
      "learning_rate": 0.002999623903500814,
      "loss": 2.209,
      "step": 4100
    },
    {
      "epoch": 0.027125964551894866,
      "grad_norm": 0.07733200490474701,
      "learning_rate": 0.0029996095915397927,
      "loss": 2.2049,
      "step": 4120
    },
    {
      "epoch": 0.027257643991467174,
      "grad_norm": 0.08035876601934433,
      "learning_rate": 0.002999595012362067,
      "loss": 2.2351,
      "step": 4140
    },
    {
      "epoch": 0.02738932343103948,
      "grad_norm": 0.07693365216255188,
      "learning_rate": 0.0029995801659702353,
      "loss": 2.2437,
      "step": 4160
    },
    {
      "epoch": 0.027521002870611783,
      "grad_norm": 0.07832279056310654,
      "learning_rate": 0.002999565052366942,
      "loss": 2.2516,
      "step": 4180
    },
    {
      "epoch": 0.027652682310184088,
      "grad_norm": 0.07737457007169724,
      "learning_rate": 0.0029995496715548814,
      "loss": 2.2528,
      "step": 4200
    },
    {
      "epoch": 0.027784361749756393,
      "grad_norm": 0.07880548387765884,
      "learning_rate": 0.0029995340235367935,
      "loss": 2.2481,
      "step": 4220
    },
    {
      "epoch": 0.027916041189328698,
      "grad_norm": 0.07748763263225555,
      "learning_rate": 0.0029995181083154665,
      "loss": 2.2366,
      "step": 4240
    },
    {
      "epoch": 0.028047720628901002,
      "grad_norm": 0.07669378817081451,
      "learning_rate": 0.0029995019258937366,
      "loss": 2.2354,
      "step": 4260
    },
    {
      "epoch": 0.028179400068473307,
      "grad_norm": 0.07612152397632599,
      "learning_rate": 0.0029994854762744874,
      "loss": 2.2306,
      "step": 4280
    },
    {
      "epoch": 0.028311079508045615,
      "grad_norm": 0.07380878925323486,
      "learning_rate": 0.0029994687594606505,
      "loss": 2.2222,
      "step": 4300
    },
    {
      "epoch": 0.02844275894761792,
      "grad_norm": 0.08108644932508469,
      "learning_rate": 0.0029994517754552037,
      "loss": 2.2188,
      "step": 4320
    },
    {
      "epoch": 0.028574438387190225,
      "grad_norm": 0.07742933928966522,
      "learning_rate": 0.002999434524261174,
      "loss": 2.2244,
      "step": 4340
    },
    {
      "epoch": 0.02870611782676253,
      "grad_norm": 0.07378648966550827,
      "learning_rate": 0.0029994170058816358,
      "loss": 2.2179,
      "step": 4360
    },
    {
      "epoch": 0.028837797266334834,
      "grad_norm": 0.07612361758947372,
      "learning_rate": 0.0029993992203197098,
      "loss": 2.215,
      "step": 4380
    },
    {
      "epoch": 0.02896947670590714,
      "grad_norm": 0.07602405548095703,
      "learning_rate": 0.0029993811675785654,
      "loss": 2.2174,
      "step": 4400
    },
    {
      "epoch": 0.029101156145479444,
      "grad_norm": 0.07372742891311646,
      "learning_rate": 0.00299936284766142,
      "loss": 2.2115,
      "step": 4420
    },
    {
      "epoch": 0.02923283558505175,
      "grad_norm": 0.07210436463356018,
      "learning_rate": 0.002999344260571538,
      "loss": 2.2031,
      "step": 4440
    },
    {
      "epoch": 0.029364515024624054,
      "grad_norm": 0.07500762492418289,
      "learning_rate": 0.002999325406312231,
      "loss": 2.2029,
      "step": 4460
    },
    {
      "epoch": 0.029496194464196362,
      "grad_norm": 0.0756341964006424,
      "learning_rate": 0.0029993062848868582,
      "loss": 2.2036,
      "step": 4480
    },
    {
      "epoch": 0.029627873903768667,
      "grad_norm": 0.07349260151386261,
      "learning_rate": 0.0029992868962988284,
      "loss": 2.199,
      "step": 4500
    },
    {
      "epoch": 0.02975955334334097,
      "grad_norm": 0.07115154713392258,
      "learning_rate": 0.002999267240551595,
      "loss": 2.1924,
      "step": 4520
    },
    {
      "epoch": 0.029891232782913276,
      "grad_norm": 0.07597554475069046,
      "learning_rate": 0.0029992473176486613,
      "loss": 2.2,
      "step": 4540
    },
    {
      "epoch": 0.03002291222248558,
      "grad_norm": 0.07525033503770828,
      "learning_rate": 0.0029992271275935773,
      "loss": 2.19,
      "step": 4560
    },
    {
      "epoch": 0.030154591662057886,
      "grad_norm": 0.07225248217582703,
      "learning_rate": 0.0029992066703899405,
      "loss": 2.1817,
      "step": 4580
    },
    {
      "epoch": 0.03028627110163019,
      "grad_norm": 0.07293014973402023,
      "learning_rate": 0.002999185946041396,
      "loss": 2.1886,
      "step": 4600
    },
    {
      "epoch": 0.030417950541202495,
      "grad_norm": 0.07140370458364487,
      "learning_rate": 0.002999164954551637,
      "loss": 2.1811,
      "step": 4620
    },
    {
      "epoch": 0.030549629980774803,
      "grad_norm": 0.0731213241815567,
      "learning_rate": 0.0029991436959244037,
      "loss": 2.1827,
      "step": 4640
    },
    {
      "epoch": 0.030681309420347108,
      "grad_norm": 0.06939541548490524,
      "learning_rate": 0.0029991221701634843,
      "loss": 2.1807,
      "step": 4660
    },
    {
      "epoch": 0.030812988859919413,
      "grad_norm": 0.0718071386218071,
      "learning_rate": 0.0029991003772727146,
      "loss": 2.1807,
      "step": 4680
    },
    {
      "epoch": 0.030944668299491718,
      "grad_norm": 0.0721597746014595,
      "learning_rate": 0.002999078317255978,
      "loss": 2.1679,
      "step": 4700
    },
    {
      "epoch": 0.031076347739064022,
      "grad_norm": 0.388390451669693,
      "learning_rate": 0.0029990559901172053,
      "loss": 2.4076,
      "step": 4720
    },
    {
      "epoch": 0.031208027178636327,
      "grad_norm": 0.12360863387584686,
      "learning_rate": 0.0029990333958603748,
      "loss": 2.7131,
      "step": 4740
    },
    {
      "epoch": 0.03133970661820863,
      "grad_norm": 0.088360495865345,
      "learning_rate": 0.002999010534489513,
      "loss": 2.425,
      "step": 4760
    },
    {
      "epoch": 0.03147138605778094,
      "grad_norm": 0.07668942958116531,
      "learning_rate": 0.002998987406008693,
      "loss": 2.3163,
      "step": 4780
    },
    {
      "epoch": 0.03160306549735324,
      "grad_norm": 0.07627064734697342,
      "learning_rate": 0.0029989640104220363,
      "loss": 2.2667,
      "step": 4800
    },
    {
      "epoch": 0.031734744936925546,
      "grad_norm": 0.07259105890989304,
      "learning_rate": 0.0029989403477337123,
      "loss": 2.2479,
      "step": 4820
    },
    {
      "epoch": 0.03186642437649785,
      "grad_norm": 0.06848350912332535,
      "learning_rate": 0.0029989164179479366,
      "loss": 2.2312,
      "step": 4840
    },
    {
      "epoch": 0.031998103816070156,
      "grad_norm": 0.0695989802479744,
      "learning_rate": 0.0029988922210689743,
      "loss": 2.2274,
      "step": 4860
    },
    {
      "epoch": 0.03212978325564246,
      "grad_norm": 0.07058060169219971,
      "learning_rate": 0.002998867757101136,
      "loss": 2.2065,
      "step": 4880
    },
    {
      "epoch": 0.03226146269521477,
      "grad_norm": 0.07192632555961609,
      "learning_rate": 0.0029988430260487822,
      "loss": 2.1976,
      "step": 4900
    },
    {
      "epoch": 0.03239314213478708,
      "grad_norm": 0.07093453407287598,
      "learning_rate": 0.002998818027916318,
      "loss": 2.1925,
      "step": 4920
    },
    {
      "epoch": 0.03252482157435938,
      "grad_norm": 0.0686832144856453,
      "learning_rate": 0.002998792762708199,
      "loss": 2.1805,
      "step": 4940
    },
    {
      "epoch": 0.03265650101393169,
      "grad_norm": 0.06718004494905472,
      "learning_rate": 0.0029987672304289275,
      "loss": 2.177,
      "step": 4960
    },
    {
      "epoch": 0.03278818045350399,
      "grad_norm": 0.07362579554319382,
      "learning_rate": 0.0029987414310830526,
      "loss": 2.173,
      "step": 4980
    },
    {
      "epoch": 0.032919859893076296,
      "grad_norm": 0.06656539440155029,
      "learning_rate": 0.0029987153646751715,
      "loss": 2.1698,
      "step": 5000
    },
    {
      "epoch": 0.032919859893076296,
      "eval_loss": 2.136354446411133,
      "eval_runtime": 67.5662,
      "eval_samples_per_second": 14.8,
      "eval_steps_per_second": 14.8,
      "step": 5000
    },
    {
      "epoch": 0.0330515393326486,
      "grad_norm": 0.06965313106775284,
      "learning_rate": 0.0029986890312099284,
      "loss": 2.1551,
      "step": 5020
    },
    {
      "epoch": 0.033183218772220906,
      "grad_norm": 0.06893343478441238,
      "learning_rate": 0.0029986624306920177,
      "loss": 2.1559,
      "step": 5040
    },
    {
      "epoch": 0.03331489821179321,
      "grad_norm": 0.06808198988437653,
      "learning_rate": 0.0029986355631261767,
      "loss": 2.1516,
      "step": 5060
    },
    {
      "epoch": 0.033446577651365515,
      "grad_norm": 0.06501699984073639,
      "learning_rate": 0.002998608428517195,
      "loss": 2.1496,
      "step": 5080
    },
    {
      "epoch": 0.03357825709093782,
      "grad_norm": 0.0687108114361763,
      "learning_rate": 0.002998581026869907,
      "loss": 2.1428,
      "step": 5100
    },
    {
      "epoch": 0.033709936530510125,
      "grad_norm": 0.06481294333934784,
      "learning_rate": 0.0029985533581891955,
      "loss": 2.1408,
      "step": 5120
    },
    {
      "epoch": 0.03384161597008243,
      "grad_norm": 0.06615699827671051,
      "learning_rate": 0.002998525422479991,
      "loss": 2.1389,
      "step": 5140
    },
    {
      "epoch": 0.033973295409654734,
      "grad_norm": 0.06763967871665955,
      "learning_rate": 0.00299849721974727,
      "loss": 2.1272,
      "step": 5160
    },
    {
      "epoch": 0.03410497484922704,
      "grad_norm": 0.0660347044467926,
      "learning_rate": 0.0029984687499960603,
      "loss": 2.1297,
      "step": 5180
    },
    {
      "epoch": 0.034236654288799344,
      "grad_norm": 0.06650801748037338,
      "learning_rate": 0.002998440013231433,
      "loss": 2.131,
      "step": 5200
    },
    {
      "epoch": 0.03436833372837165,
      "grad_norm": 0.06661241501569748,
      "learning_rate": 0.0029984110094585102,
      "loss": 2.1246,
      "step": 5220
    },
    {
      "epoch": 0.03450001316794396,
      "grad_norm": 0.06411544978618622,
      "learning_rate": 0.0029983817386824586,
      "loss": 2.1241,
      "step": 5240
    },
    {
      "epoch": 0.034631692607516265,
      "grad_norm": 0.06469565629959106,
      "learning_rate": 0.0029983522009084953,
      "loss": 2.1161,
      "step": 5260
    },
    {
      "epoch": 0.03476337204708857,
      "grad_norm": 0.06469576805830002,
      "learning_rate": 0.0029983223961418835,
      "loss": 2.121,
      "step": 5280
    },
    {
      "epoch": 0.034895051486660875,
      "grad_norm": 0.06342877447605133,
      "learning_rate": 0.002998292324387933,
      "loss": 2.1244,
      "step": 5300
    },
    {
      "epoch": 0.03502673092623318,
      "grad_norm": 0.06592860817909241,
      "learning_rate": 0.0029982619856520035,
      "loss": 2.1312,
      "step": 5320
    },
    {
      "epoch": 0.035158410365805484,
      "grad_norm": 0.06745469570159912,
      "learning_rate": 0.0029982313799395005,
      "loss": 2.1259,
      "step": 5340
    },
    {
      "epoch": 0.03529008980537779,
      "grad_norm": 0.06721679121255875,
      "learning_rate": 0.0029982005072558774,
      "loss": 2.1212,
      "step": 5360
    },
    {
      "epoch": 0.035421769244950094,
      "grad_norm": 0.0677921324968338,
      "learning_rate": 0.0029981693676066367,
      "loss": 2.1187,
      "step": 5380
    },
    {
      "epoch": 0.0355534486845224,
      "grad_norm": 0.06510468572378159,
      "learning_rate": 0.0029981379609973257,
      "loss": 2.1186,
      "step": 5400
    },
    {
      "epoch": 0.0356851281240947,
      "grad_norm": 0.06507667899131775,
      "learning_rate": 0.002998106287433542,
      "loss": 2.1085,
      "step": 5420
    },
    {
      "epoch": 0.03581680756366701,
      "grad_norm": 0.06307834386825562,
      "learning_rate": 0.0029980743469209285,
      "loss": 2.1086,
      "step": 5440
    },
    {
      "epoch": 0.03594848700323931,
      "grad_norm": 0.06396503001451492,
      "learning_rate": 0.0029980421394651775,
      "loss": 2.1012,
      "step": 5460
    },
    {
      "epoch": 0.03608016644281162,
      "grad_norm": 0.0639200434088707,
      "learning_rate": 0.0029980096650720277,
      "loss": 2.1048,
      "step": 5480
    },
    {
      "epoch": 0.03621184588238392,
      "grad_norm": 0.06461615860462189,
      "learning_rate": 0.0029979769237472656,
      "loss": 2.098,
      "step": 5500
    },
    {
      "epoch": 0.03634352532195623,
      "grad_norm": 0.06527513265609741,
      "learning_rate": 0.0029979439154967256,
      "loss": 2.0921,
      "step": 5520
    },
    {
      "epoch": 0.03647520476152853,
      "grad_norm": 0.06093462184071541,
      "learning_rate": 0.0029979106403262897,
      "loss": 2.0834,
      "step": 5540
    },
    {
      "epoch": 0.036606884201100844,
      "grad_norm": 0.06476835161447525,
      "learning_rate": 0.002997877098241887,
      "loss": 2.0851,
      "step": 5560
    },
    {
      "epoch": 0.03673856364067315,
      "grad_norm": 0.06311192363500595,
      "learning_rate": 0.002997843289249494,
      "loss": 2.0799,
      "step": 5580
    },
    {
      "epoch": 0.03687024308024545,
      "grad_norm": 0.06339168548583984,
      "learning_rate": 0.002997809213355136,
      "loss": 2.0858,
      "step": 5600
    },
    {
      "epoch": 0.03700192251981776,
      "grad_norm": 0.06146325170993805,
      "learning_rate": 0.002997774870564884,
      "loss": 2.0948,
      "step": 5620
    },
    {
      "epoch": 0.03713360195939006,
      "grad_norm": 0.06187060475349426,
      "learning_rate": 0.0029977402608848585,
      "loss": 2.0822,
      "step": 5640
    },
    {
      "epoch": 0.03726528139896237,
      "grad_norm": 0.06805236637592316,
      "learning_rate": 0.0029977053843212266,
      "loss": 2.0759,
      "step": 5660
    },
    {
      "epoch": 0.03739696083853467,
      "grad_norm": 0.06448160111904144,
      "learning_rate": 0.002997670240880202,
      "loss": 2.08,
      "step": 5680
    },
    {
      "epoch": 0.03752864027810698,
      "grad_norm": 0.06372448801994324,
      "learning_rate": 0.002997634830568048,
      "loss": 2.0823,
      "step": 5700
    },
    {
      "epoch": 0.03766031971767928,
      "grad_norm": 0.061139971017837524,
      "learning_rate": 0.0029975991533910734,
      "loss": 2.0742,
      "step": 5720
    },
    {
      "epoch": 0.037791999157251586,
      "grad_norm": 0.06403601914644241,
      "learning_rate": 0.0029975632093556365,
      "loss": 2.0664,
      "step": 5740
    },
    {
      "epoch": 0.03792367859682389,
      "grad_norm": 0.06160760670900345,
      "learning_rate": 0.002997526998468142,
      "loss": 2.0591,
      "step": 5760
    },
    {
      "epoch": 0.038055358036396196,
      "grad_norm": 0.06119397282600403,
      "learning_rate": 0.0029974905207350417,
      "loss": 2.0637,
      "step": 5780
    },
    {
      "epoch": 0.0381870374759685,
      "grad_norm": 0.062456753104925156,
      "learning_rate": 0.0029974537761628364,
      "loss": 2.0611,
      "step": 5800
    },
    {
      "epoch": 0.038318716915540806,
      "grad_norm": 0.06182549521327019,
      "learning_rate": 0.002997416764758073,
      "loss": 2.0619,
      "step": 5820
    },
    {
      "epoch": 0.03845039635511311,
      "grad_norm": 0.06361284852027893,
      "learning_rate": 0.002997379486527347,
      "loss": 2.054,
      "step": 5840
    },
    {
      "epoch": 0.038582075794685415,
      "grad_norm": 0.06169150397181511,
      "learning_rate": 0.0029973419414773017,
      "loss": 2.058,
      "step": 5860
    },
    {
      "epoch": 0.03871375523425772,
      "grad_norm": 0.06227455288171768,
      "learning_rate": 0.002997304129614625,
      "loss": 2.0513,
      "step": 5880
    },
    {
      "epoch": 0.03884543467383003,
      "grad_norm": 0.06137228012084961,
      "learning_rate": 0.0029972660509460574,
      "loss": 2.0706,
      "step": 5900
    },
    {
      "epoch": 0.038977114113402336,
      "grad_norm": 0.06067802384495735,
      "learning_rate": 0.0029972277054783826,
      "loss": 2.0988,
      "step": 5920
    },
    {
      "epoch": 0.03910879355297464,
      "grad_norm": 0.062081821262836456,
      "learning_rate": 0.0029971890932184338,
      "loss": 2.1072,
      "step": 5940
    },
    {
      "epoch": 0.039240472992546946,
      "grad_norm": 0.058524634689092636,
      "learning_rate": 0.0029971502141730912,
      "loss": 2.112,
      "step": 5960
    },
    {
      "epoch": 0.03937215243211925,
      "grad_norm": 0.06115134805440903,
      "learning_rate": 0.0029971110683492826,
      "loss": 2.1138,
      "step": 5980
    },
    {
      "epoch": 0.039503831871691555,
      "grad_norm": 0.05962108075618744,
      "learning_rate": 0.0029970716557539837,
      "loss": 2.1112,
      "step": 6000
    },
    {
      "epoch": 0.039503831871691555,
      "eval_loss": 2.0984599590301514,
      "eval_runtime": 67.374,
      "eval_samples_per_second": 14.843,
      "eval_steps_per_second": 14.843,
      "step": 6000
    },
    {
      "epoch": 0.03963551131126386,
      "grad_norm": 0.06020989269018173,
      "learning_rate": 0.0029970319763942175,
      "loss": 2.1038,
      "step": 6020
    },
    {
      "epoch": 0.039767190750836165,
      "grad_norm": 0.059578102082014084,
      "learning_rate": 0.0029969920302770543,
      "loss": 2.0974,
      "step": 6040
    },
    {
      "epoch": 0.03989887019040847,
      "grad_norm": 0.05818217247724533,
      "learning_rate": 0.002996951817409612,
      "loss": 2.1037,
      "step": 6060
    },
    {
      "epoch": 0.040030549629980774,
      "grad_norm": 0.05842572823166847,
      "learning_rate": 0.0029969113377990566,
      "loss": 2.0998,
      "step": 6080
    },
    {
      "epoch": 0.04016222906955308,
      "grad_norm": 0.057952892035245895,
      "learning_rate": 0.002996870591452601,
      "loss": 2.1024,
      "step": 6100
    },
    {
      "epoch": 0.040293908509125384,
      "grad_norm": 0.0585886612534523,
      "learning_rate": 0.0029968295783775055,
      "loss": 2.0989,
      "step": 6120
    },
    {
      "epoch": 0.04042558794869769,
      "grad_norm": 0.061875827610492706,
      "learning_rate": 0.0029967882985810785,
      "loss": 2.0848,
      "step": 6140
    },
    {
      "epoch": 0.040557267388269994,
      "grad_norm": 0.059371497482061386,
      "learning_rate": 0.002996746752070676,
      "loss": 2.0865,
      "step": 6160
    },
    {
      "epoch": 0.0406889468278423,
      "grad_norm": 0.06270122528076172,
      "learning_rate": 0.002996704938853701,
      "loss": 2.0847,
      "step": 6180
    },
    {
      "epoch": 0.0408206262674146,
      "grad_norm": 0.05946968123316765,
      "learning_rate": 0.002996662858937604,
      "loss": 2.0773,
      "step": 6200
    },
    {
      "epoch": 0.04095230570698691,
      "grad_norm": 0.05964416265487671,
      "learning_rate": 0.002996620512329883,
      "loss": 2.0827,
      "step": 6220
    },
    {
      "epoch": 0.04108398514655922,
      "grad_norm": 0.059607379138469696,
      "learning_rate": 0.0029965778990380847,
      "loss": 2.0802,
      "step": 6240
    },
    {
      "epoch": 0.041215664586131524,
      "grad_norm": 0.059771690517663956,
      "learning_rate": 0.0029965350190698016,
      "loss": 2.0753,
      "step": 6260
    },
    {
      "epoch": 0.04134734402570383,
      "grad_norm": 0.05873079225420952,
      "learning_rate": 0.002996491872432675,
      "loss": 2.0742,
      "step": 6280
    },
    {
      "epoch": 0.041479023465276134,
      "grad_norm": 0.05816845968365669,
      "learning_rate": 0.0029964484591343933,
      "loss": 2.0689,
      "step": 6300
    },
    {
      "epoch": 0.04161070290484844,
      "grad_norm": 0.057672835886478424,
      "learning_rate": 0.002996404779182692,
      "loss": 2.0648,
      "step": 6320
    },
    {
      "epoch": 0.04174238234442074,
      "grad_norm": 0.06043649837374687,
      "learning_rate": 0.0029963608325853544,
      "loss": 2.0719,
      "step": 6340
    },
    {
      "epoch": 0.04187406178399305,
      "grad_norm": 0.05902590975165367,
      "learning_rate": 0.0029963166193502115,
      "loss": 2.0662,
      "step": 6360
    },
    {
      "epoch": 0.04200574122356535,
      "grad_norm": 0.061324600130319595,
      "learning_rate": 0.002996272139485142,
      "loss": 2.0672,
      "step": 6380
    },
    {
      "epoch": 0.04213742066313766,
      "grad_norm": 0.05997491627931595,
      "learning_rate": 0.002996227392998071,
      "loss": 2.0733,
      "step": 6400
    },
    {
      "epoch": 0.04226910010270996,
      "grad_norm": 0.0597202442586422,
      "learning_rate": 0.0029961823798969733,
      "loss": 2.062,
      "step": 6420
    },
    {
      "epoch": 0.04240077954228227,
      "grad_norm": 0.10704315453767776,
      "learning_rate": 0.002996137100189868,
      "loss": 2.0669,
      "step": 6440
    },
    {
      "epoch": 0.04253245898185457,
      "grad_norm": 0.056581802666187286,
      "learning_rate": 0.0029960915538848254,
      "loss": 2.0585,
      "step": 6460
    },
    {
      "epoch": 0.04266413842142688,
      "grad_norm": 0.0565234012901783,
      "learning_rate": 0.00299604574098996,
      "loss": 2.0554,
      "step": 6480
    },
    {
      "epoch": 0.04279581786099918,
      "grad_norm": 0.056596871465444565,
      "learning_rate": 0.0029959996615134357,
      "loss": 2.077,
      "step": 6500
    },
    {
      "epoch": 0.042927497300571486,
      "grad_norm": 0.05787625536322594,
      "learning_rate": 0.002995953315463464,
      "loss": 2.0741,
      "step": 6520
    },
    {
      "epoch": 0.04305917674014379,
      "grad_norm": 0.059064172208309174,
      "learning_rate": 0.002995906702848302,
      "loss": 2.0792,
      "step": 6540
    },
    {
      "epoch": 0.043190856179716096,
      "grad_norm": 0.057197242975234985,
      "learning_rate": 0.0029958598236762574,
      "loss": 2.0702,
      "step": 6560
    },
    {
      "epoch": 0.04332253561928841,
      "grad_norm": 0.05579278990626335,
      "learning_rate": 0.002995812677955683,
      "loss": 2.0677,
      "step": 6580
    },
    {
      "epoch": 0.04345421505886071,
      "grad_norm": 0.05777699500322342,
      "learning_rate": 0.0029957652656949787,
      "loss": 2.0649,
      "step": 6600
    },
    {
      "epoch": 0.04358589449843302,
      "grad_norm": 0.05634402483701706,
      "learning_rate": 0.002995717586902594,
      "loss": 2.0681,
      "step": 6620
    },
    {
      "epoch": 0.04371757393800532,
      "grad_norm": 0.05572090670466423,
      "learning_rate": 0.0029956696415870242,
      "loss": 2.0587,
      "step": 6640
    },
    {
      "epoch": 0.04384925337757763,
      "grad_norm": 0.05806443840265274,
      "learning_rate": 0.0029956214297568134,
      "loss": 2.0555,
      "step": 6660
    },
    {
      "epoch": 0.04398093281714993,
      "grad_norm": 0.05726337805390358,
      "learning_rate": 0.0029955729514205523,
      "loss": 2.0474,
      "step": 6680
    },
    {
      "epoch": 0.044112612256722236,
      "grad_norm": 0.05729576200246811,
      "learning_rate": 0.0029955242065868794,
      "loss": 2.0489,
      "step": 6700
    },
    {
      "epoch": 0.04424429169629454,
      "grad_norm": 0.06136414781212807,
      "learning_rate": 0.0029954751952644795,
      "loss": 2.0509,
      "step": 6720
    },
    {
      "epoch": 0.044375971135866846,
      "grad_norm": 0.058006297796964645,
      "learning_rate": 0.0029954259174620875,
      "loss": 2.0515,
      "step": 6740
    },
    {
      "epoch": 0.04450765057543915,
      "grad_norm": 0.05813615396618843,
      "learning_rate": 0.0029953763731884833,
      "loss": 2.048,
      "step": 6760
    },
    {
      "epoch": 0.044639330015011455,
      "grad_norm": 0.05697380378842354,
      "learning_rate": 0.0029953265624524956,
      "loss": 2.0433,
      "step": 6780
    },
    {
      "epoch": 0.04477100945458376,
      "grad_norm": 0.05840064957737923,
      "learning_rate": 0.002995276485263,
      "loss": 2.0405,
      "step": 6800
    },
    {
      "epoch": 0.044902688894156065,
      "grad_norm": 0.05951813980937004,
      "learning_rate": 0.00299522614162892,
      "loss": 2.0418,
      "step": 6820
    },
    {
      "epoch": 0.04503436833372837,
      "grad_norm": 0.05788157880306244,
      "learning_rate": 0.0029951755315592264,
      "loss": 2.0378,
      "step": 6840
    },
    {
      "epoch": 0.045166047773300674,
      "grad_norm": 0.05446796864271164,
      "learning_rate": 0.002995124655062937,
      "loss": 2.0412,
      "step": 6860
    },
    {
      "epoch": 0.04529772721287298,
      "grad_norm": 0.05822540074586868,
      "learning_rate": 0.002995073512149118,
      "loss": 2.0357,
      "step": 6880
    },
    {
      "epoch": 0.045429406652445284,
      "grad_norm": 0.0586545467376709,
      "learning_rate": 0.002995022102826883,
      "loss": 2.0284,
      "step": 6900
    },
    {
      "epoch": 0.045561086092017596,
      "grad_norm": 0.056676048785448074,
      "learning_rate": 0.0029949704271053914,
      "loss": 2.0363,
      "step": 6920
    },
    {
      "epoch": 0.0456927655315899,
      "grad_norm": 0.0642879530787468,
      "learning_rate": 0.0029949184849938524,
      "loss": 2.0271,
      "step": 6940
    },
    {
      "epoch": 0.045824444971162205,
      "grad_norm": 0.054941147565841675,
      "learning_rate": 0.002994866276501521,
      "loss": 2.0297,
      "step": 6960
    },
    {
      "epoch": 0.04595612441073451,
      "grad_norm": 0.057689156383275986,
      "learning_rate": 0.0029948138016377005,
      "loss": 2.032,
      "step": 6980
    },
    {
      "epoch": 0.046087803850306815,
      "grad_norm": 0.05917928367853165,
      "learning_rate": 0.0029947610604117423,
      "loss": 2.0272,
      "step": 7000
    },
    {
      "epoch": 0.046087803850306815,
      "eval_loss": 1.9085378646850586,
      "eval_runtime": 68.9073,
      "eval_samples_per_second": 14.512,
      "eval_steps_per_second": 14.512,
      "step": 7000
    },
    {
      "epoch": 0.04621948328987912,
      "grad_norm": 0.05777193233370781,
      "learning_rate": 0.002994708052833043,
      "loss": 2.0243,
      "step": 7020
    },
    {
      "epoch": 0.046351162729451424,
      "grad_norm": 0.056196246296167374,
      "learning_rate": 0.0029946547789110487,
      "loss": 2.0237,
      "step": 7040
    },
    {
      "epoch": 0.04648284216902373,
      "grad_norm": 0.05449189990758896,
      "learning_rate": 0.002994601238655252,
      "loss": 2.0229,
      "step": 7060
    },
    {
      "epoch": 0.046614521608596034,
      "grad_norm": 0.05653191730380058,
      "learning_rate": 0.0029945474320751944,
      "loss": 2.0243,
      "step": 7080
    },
    {
      "epoch": 0.04674620104816834,
      "grad_norm": 0.05763334780931473,
      "learning_rate": 0.0029944933591804623,
      "loss": 2.0279,
      "step": 7100
    },
    {
      "epoch": 0.04687788048774064,
      "grad_norm": 0.05874243751168251,
      "learning_rate": 0.002994439019980692,
      "loss": 2.0081,
      "step": 7120
    },
    {
      "epoch": 0.04700955992731295,
      "grad_norm": 0.06031106784939766,
      "learning_rate": 0.0029943844144855658,
      "loss": 1.9983,
      "step": 7140
    },
    {
      "epoch": 0.04714123936688525,
      "grad_norm": 0.05547622591257095,
      "learning_rate": 0.002994329542704814,
      "loss": 1.9954,
      "step": 7160
    },
    {
      "epoch": 0.04727291880645756,
      "grad_norm": 0.05814756825566292,
      "learning_rate": 0.002994274404648214,
      "loss": 1.9944,
      "step": 7180
    },
    {
      "epoch": 0.04740459824602986,
      "grad_norm": 0.05669866129755974,
      "learning_rate": 0.002994219000325591,
      "loss": 1.9852,
      "step": 7200
    },
    {
      "epoch": 0.04753627768560217,
      "grad_norm": 0.05776144191622734,
      "learning_rate": 0.0029941633297468177,
      "loss": 1.9788,
      "step": 7220
    },
    {
      "epoch": 0.04766795712517447,
      "grad_norm": 0.05656075105071068,
      "learning_rate": 0.002994107392921814,
      "loss": 1.9872,
      "step": 7240
    },
    {
      "epoch": 0.047799636564746784,
      "grad_norm": 0.05627656355500221,
      "learning_rate": 0.0029940511898605476,
      "loss": 1.9787,
      "step": 7260
    },
    {
      "epoch": 0.04793131600431909,
      "grad_norm": 0.05843086168169975,
      "learning_rate": 0.002993994720573033,
      "loss": 1.9782,
      "step": 7280
    },
    {
      "epoch": 0.04806299544389139,
      "grad_norm": 0.05826539546251297,
      "learning_rate": 0.002993937985069333,
      "loss": 1.9819,
      "step": 7300
    },
    {
      "epoch": 0.0481946748834637,
      "grad_norm": 0.06292963773012161,
      "learning_rate": 0.0029938809833595565,
      "loss": 1.9696,
      "step": 7320
    },
    {
      "epoch": 0.048326354323036,
      "grad_norm": 0.057060081511735916,
      "learning_rate": 0.0029938237154538616,
      "loss": 1.9633,
      "step": 7340
    },
    {
      "epoch": 0.04845803376260831,
      "grad_norm": 0.06161024048924446,
      "learning_rate": 0.0029937661813624525,
      "loss": 1.9623,
      "step": 7360
    },
    {
      "epoch": 0.04858971320218061,
      "grad_norm": 0.05789177492260933,
      "learning_rate": 0.0029937083810955807,
      "loss": 1.9724,
      "step": 7380
    },
    {
      "epoch": 0.04872139264175292,
      "grad_norm": 0.055680371820926666,
      "learning_rate": 0.0029936503146635466,
      "loss": 1.9657,
      "step": 7400
    },
    {
      "epoch": 0.04885307208132522,
      "grad_norm": 0.05655812472105026,
      "learning_rate": 0.0029935919820766967,
      "loss": 1.9664,
      "step": 7420
    },
    {
      "epoch": 0.048984751520897526,
      "grad_norm": 0.05809199810028076,
      "learning_rate": 0.0029935333833454253,
      "loss": 1.9728,
      "step": 7440
    },
    {
      "epoch": 0.04911643096046983,
      "grad_norm": 0.057024531066417694,
      "learning_rate": 0.0029934745184801745,
      "loss": 1.9637,
      "step": 7460
    },
    {
      "epoch": 0.049248110400042136,
      "grad_norm": 0.05507376790046692,
      "learning_rate": 0.0029934153874914327,
      "loss": 1.9646,
      "step": 7480
    },
    {
      "epoch": 0.04937978983961444,
      "grad_norm": 0.0562782883644104,
      "learning_rate": 0.0029933559903897373,
      "loss": 1.9642,
      "step": 7500
    },
    {
      "epoch": 0.049511469279186746,
      "grad_norm": 0.05876750499010086,
      "learning_rate": 0.002993296327185672,
      "loss": 1.96,
      "step": 7520
    },
    {
      "epoch": 0.04964314871875905,
      "grad_norm": 0.0542948879301548,
      "learning_rate": 0.0029932363978898684,
      "loss": 1.957,
      "step": 7540
    },
    {
      "epoch": 0.049774828158331355,
      "grad_norm": 0.05514788627624512,
      "learning_rate": 0.0029931762025130052,
      "loss": 1.956,
      "step": 7560
    },
    {
      "epoch": 0.04990650759790366,
      "grad_norm": 0.055275484919548035,
      "learning_rate": 0.0029931157410658087,
      "loss": 1.9586,
      "step": 7580
    },
    {
      "epoch": 0.05003818703747597,
      "grad_norm": 0.06243159994482994,
      "learning_rate": 0.0029930550135590524,
      "loss": 1.9528,
      "step": 7600
    },
    {
      "epoch": 0.050169866477048276,
      "grad_norm": 0.055291496217250824,
      "learning_rate": 0.0029929940200035575,
      "loss": 1.9535,
      "step": 7620
    },
    {
      "epoch": 0.05030154591662058,
      "grad_norm": 0.0554860420525074,
      "learning_rate": 0.002992932760410193,
      "loss": 1.9573,
      "step": 7640
    },
    {
      "epoch": 0.050433225356192886,
      "grad_norm": 0.058307211846113205,
      "learning_rate": 0.0029928712347898736,
      "loss": 1.9489,
      "step": 7660
    },
    {
      "epoch": 0.05056490479576519,
      "grad_norm": 0.057058483362197876,
      "learning_rate": 0.0029928094431535637,
      "loss": 1.9564,
      "step": 7680
    },
    {
      "epoch": 0.050696584235337495,
      "grad_norm": 0.055521443486213684,
      "learning_rate": 0.0029927473855122736,
      "loss": 1.967,
      "step": 7700
    },
    {
      "epoch": 0.0508282636749098,
      "grad_norm": 0.05751781165599823,
      "learning_rate": 0.0029926850618770618,
      "loss": 1.9694,
      "step": 7720
    },
    {
      "epoch": 0.050959943114482105,
      "grad_norm": 0.05835530161857605,
      "learning_rate": 0.002992622472259033,
      "loss": 1.9706,
      "step": 7740
    },
    {
      "epoch": 0.05109162255405441,
      "grad_norm": 0.05460565164685249,
      "learning_rate": 0.0029925596166693403,
      "loss": 1.9652,
      "step": 7760
    },
    {
      "epoch": 0.051223301993626714,
      "grad_norm": 0.05999600142240524,
      "learning_rate": 0.0029924964951191847,
      "loss": 1.9628,
      "step": 7780
    },
    {
      "epoch": 0.05135498143319902,
      "grad_norm": 0.05896478518843651,
      "learning_rate": 0.0029924331076198133,
      "loss": 1.9625,
      "step": 7800
    },
    {
      "epoch": 0.051486660872771324,
      "grad_norm": 0.05324622988700867,
      "learning_rate": 0.0029923694541825214,
      "loss": 1.9593,
      "step": 7820
    },
    {
      "epoch": 0.05161834031234363,
      "grad_norm": 0.056877490133047104,
      "learning_rate": 0.002992305534818651,
      "loss": 1.9547,
      "step": 7840
    },
    {
      "epoch": 0.051750019751915934,
      "grad_norm": 0.05267615243792534,
      "learning_rate": 0.0029922413495395927,
      "loss": 1.9584,
      "step": 7860
    },
    {
      "epoch": 0.05188169919148824,
      "grad_norm": 0.05451074242591858,
      "learning_rate": 0.002992176898356783,
      "loss": 1.957,
      "step": 7880
    },
    {
      "epoch": 0.05201337863106054,
      "grad_norm": 0.0581444576382637,
      "learning_rate": 0.0029921121812817074,
      "loss": 1.948,
      "step": 7900
    },
    {
      "epoch": 0.052145058070632855,
      "grad_norm": 0.0567544661462307,
      "learning_rate": 0.0029920471983258964,
      "loss": 1.947,
      "step": 7920
    },
    {
      "epoch": 0.05227673751020516,
      "grad_norm": 0.05727114900946617,
      "learning_rate": 0.0029919819495009313,
      "loss": 1.9494,
      "step": 7940
    },
    {
      "epoch": 0.052408416949777464,
      "grad_norm": 0.05433478206396103,
      "learning_rate": 0.002991916434818437,
      "loss": 1.9443,
      "step": 7960
    },
    {
      "epoch": 0.05254009638934977,
      "grad_norm": 0.055305104702711105,
      "learning_rate": 0.002991850654290089,
      "loss": 1.9461,
      "step": 7980
    },
    {
      "epoch": 0.052671775828922074,
      "grad_norm": 0.056631047278642654,
      "learning_rate": 0.0029917846079276084,
      "loss": 1.9424,
      "step": 8000
    },
    {
      "epoch": 0.052671775828922074,
      "eval_loss": 1.780034065246582,
      "eval_runtime": 67.4451,
      "eval_samples_per_second": 14.827,
      "eval_steps_per_second": 14.827,
      "step": 8000
    },
    {
      "epoch": 0.05280345526849438,
      "grad_norm": 0.06008228659629822,
      "learning_rate": 0.002991718295742763,
      "loss": 1.9402,
      "step": 8020
    },
    {
      "epoch": 0.05293513470806668,
      "grad_norm": 0.05446495860815048,
      "learning_rate": 0.0029916517177473708,
      "loss": 1.942,
      "step": 8040
    },
    {
      "epoch": 0.05306681414763899,
      "grad_norm": 0.06055721640586853,
      "learning_rate": 0.002991584873953294,
      "loss": 1.9409,
      "step": 8060
    },
    {
      "epoch": 0.05319849358721129,
      "grad_norm": 0.056658077985048294,
      "learning_rate": 0.002991517764372444,
      "loss": 1.9384,
      "step": 8080
    },
    {
      "epoch": 0.0533301730267836,
      "grad_norm": 0.054917097091674805,
      "learning_rate": 0.00299145038901678,
      "loss": 1.9312,
      "step": 8100
    },
    {
      "epoch": 0.0534618524663559,
      "grad_norm": 0.05398482084274292,
      "learning_rate": 0.0029913827478983057,
      "loss": 1.9454,
      "step": 8120
    },
    {
      "epoch": 0.05359353190592821,
      "grad_norm": 0.05773143097758293,
      "learning_rate": 0.002991314841029076,
      "loss": 1.9394,
      "step": 8140
    },
    {
      "epoch": 0.05372521134550051,
      "grad_norm": 0.0562208853662014,
      "learning_rate": 0.0029912466684211907,
      "loss": 1.94,
      "step": 8160
    },
    {
      "epoch": 0.05385689078507282,
      "grad_norm": 0.05213721841573715,
      "learning_rate": 0.002991178230086797,
      "loss": 1.928,
      "step": 8180
    },
    {
      "epoch": 0.05398857022464512,
      "grad_norm": 0.05434253811836243,
      "learning_rate": 0.0029911095260380903,
      "loss": 1.9361,
      "step": 8200
    },
    {
      "epoch": 0.054120249664217426,
      "grad_norm": 0.0537894032895565,
      "learning_rate": 0.0029910405562873135,
      "loss": 1.9368,
      "step": 8220
    },
    {
      "epoch": 0.05425192910378973,
      "grad_norm": 0.055712759494781494,
      "learning_rate": 0.0029909713208467557,
      "loss": 1.9253,
      "step": 8240
    },
    {
      "epoch": 0.05438360854336204,
      "grad_norm": 0.0564916618168354,
      "learning_rate": 0.0029909018197287543,
      "loss": 1.936,
      "step": 8260
    },
    {
      "epoch": 0.05451528798293435,
      "grad_norm": 0.059347622096538544,
      "learning_rate": 0.002990832052945694,
      "loss": 1.9872,
      "step": 8280
    },
    {
      "epoch": 0.05464696742250665,
      "grad_norm": 0.05435614287853241,
      "learning_rate": 0.002990762020510006,
      "loss": 2.0069,
      "step": 8300
    },
    {
      "epoch": 0.05477864686207896,
      "grad_norm": 0.05343768745660782,
      "learning_rate": 0.0029906917224341704,
      "loss": 2.0144,
      "step": 8320
    },
    {
      "epoch": 0.05491032630165126,
      "grad_norm": 0.054454583674669266,
      "learning_rate": 0.0029906211587307128,
      "loss": 2.0294,
      "step": 8340
    },
    {
      "epoch": 0.05504200574122357,
      "grad_norm": 0.05322398245334625,
      "learning_rate": 0.002990550329412207,
      "loss": 2.0255,
      "step": 8360
    },
    {
      "epoch": 0.05517368518079587,
      "grad_norm": 0.0558038055896759,
      "learning_rate": 0.002990479234491275,
      "loss": 2.0239,
      "step": 8380
    },
    {
      "epoch": 0.055305364620368176,
      "grad_norm": 0.0529707595705986,
      "learning_rate": 0.002990407873980584,
      "loss": 2.0235,
      "step": 8400
    },
    {
      "epoch": 0.05543704405994048,
      "grad_norm": 0.049640074372291565,
      "learning_rate": 0.002990336247892851,
      "loss": 2.0226,
      "step": 8420
    },
    {
      "epoch": 0.055568723499512786,
      "grad_norm": 0.05250220373272896,
      "learning_rate": 0.0029902643562408388,
      "loss": 2.0082,
      "step": 8440
    },
    {
      "epoch": 0.05570040293908509,
      "grad_norm": 0.05764392018318176,
      "learning_rate": 0.0029901921990373574,
      "loss": 2.0135,
      "step": 8460
    },
    {
      "epoch": 0.055832082378657395,
      "grad_norm": 0.0569184236228466,
      "learning_rate": 0.0029901197762952645,
      "loss": 2.0064,
      "step": 8480
    },
    {
      "epoch": 0.0559637618182297,
      "grad_norm": 0.05316290259361267,
      "learning_rate": 0.0029900470880274655,
      "loss": 1.9979,
      "step": 8500
    },
    {
      "epoch": 0.056095441257802005,
      "grad_norm": 0.05095953494310379,
      "learning_rate": 0.002989974134246913,
      "loss": 2.0072,
      "step": 8520
    },
    {
      "epoch": 0.05622712069737431,
      "grad_norm": 0.06118274852633476,
      "learning_rate": 0.002989900914966606,
      "loss": 2.0055,
      "step": 8540
    },
    {
      "epoch": 0.056358800136946614,
      "grad_norm": 0.05806516483426094,
      "learning_rate": 0.0029898274301995926,
      "loss": 1.9939,
      "step": 8560
    },
    {
      "epoch": 0.05649047957651892,
      "grad_norm": 0.05719646438956261,
      "learning_rate": 0.0029897536799589662,
      "loss": 1.9945,
      "step": 8580
    },
    {
      "epoch": 0.05662215901609123,
      "grad_norm": 0.05527440831065178,
      "learning_rate": 0.0029896796642578686,
      "loss": 1.9894,
      "step": 8600
    },
    {
      "epoch": 0.056753838455663536,
      "grad_norm": 0.0631207823753357,
      "learning_rate": 0.002989605383109489,
      "loss": 1.994,
      "step": 8620
    },
    {
      "epoch": 0.05688551789523584,
      "grad_norm": 0.054478082805871964,
      "learning_rate": 0.002989530836527063,
      "loss": 1.9867,
      "step": 8640
    },
    {
      "epoch": 0.057017197334808145,
      "grad_norm": 0.05206460878252983,
      "learning_rate": 0.002989456024523875,
      "loss": 1.9807,
      "step": 8660
    },
    {
      "epoch": 0.05714887677438045,
      "grad_norm": 0.05099276080727577,
      "learning_rate": 0.002989380947113255,
      "loss": 1.983,
      "step": 8680
    },
    {
      "epoch": 0.057280556213952755,
      "grad_norm": 0.05928613618016243,
      "learning_rate": 0.0029893056043085813,
      "loss": 1.9792,
      "step": 8700
    },
    {
      "epoch": 0.05741223565352506,
      "grad_norm": 0.04971221461892128,
      "learning_rate": 0.0029892299961232797,
      "loss": 1.9756,
      "step": 8720
    },
    {
      "epoch": 0.057543915093097364,
      "grad_norm": 0.060248006135225296,
      "learning_rate": 0.002989154122570823,
      "loss": 1.9837,
      "step": 8740
    },
    {
      "epoch": 0.05767559453266967,
      "grad_norm": 0.058486610651016235,
      "learning_rate": 0.0029890779836647305,
      "loss": 1.9781,
      "step": 8760
    },
    {
      "epoch": 0.057807273972241974,
      "grad_norm": 0.05129622295498848,
      "learning_rate": 0.00298900157941857,
      "loss": 1.972,
      "step": 8780
    },
    {
      "epoch": 0.05793895341181428,
      "grad_norm": 0.052501507103443146,
      "learning_rate": 0.002988924909845955,
      "loss": 1.9731,
      "step": 8800
    },
    {
      "epoch": 0.05807063285138658,
      "grad_norm": 0.055490028113126755,
      "learning_rate": 0.0029888479749605487,
      "loss": 1.9697,
      "step": 8820
    },
    {
      "epoch": 0.05820231229095889,
      "grad_norm": 0.05250798165798187,
      "learning_rate": 0.0029887707747760597,
      "loss": 1.9722,
      "step": 8840
    },
    {
      "epoch": 0.05833399173053119,
      "grad_norm": 0.053195856511592865,
      "learning_rate": 0.0029886933093062444,
      "loss": 1.9829,
      "step": 8860
    },
    {
      "epoch": 0.0584656711701035,
      "grad_norm": 0.05129017308354378,
      "learning_rate": 0.002988615578564906,
      "loss": 1.9818,
      "step": 8880
    },
    {
      "epoch": 0.0585973506096758,
      "grad_norm": 0.05016680434346199,
      "learning_rate": 0.0029885375825658956,
      "loss": 1.9915,
      "step": 8900
    },
    {
      "epoch": 0.05872903004924811,
      "grad_norm": 0.054276566952466965,
      "learning_rate": 0.0029884593213231115,
      "loss": 1.986,
      "step": 8920
    },
    {
      "epoch": 0.05886070948882042,
      "grad_norm": 0.05341557413339615,
      "learning_rate": 0.0029883807948504998,
      "loss": 1.9816,
      "step": 8940
    },
    {
      "epoch": 0.058992388928392724,
      "grad_norm": 0.05010661855340004,
      "learning_rate": 0.002988302003162052,
      "loss": 1.9822,
      "step": 8960
    },
    {
      "epoch": 0.05912406836796503,
      "grad_norm": 0.06433503329753876,
      "learning_rate": 0.0029882229462718088,
      "loss": 1.9838,
      "step": 8980
    },
    {
      "epoch": 0.05925574780753733,
      "grad_norm": 0.0578884556889534,
      "learning_rate": 0.002988143624193857,
      "loss": 1.9786,
      "step": 9000
    },
    {
      "epoch": 0.05925574780753733,
      "eval_loss": 1.9106996059417725,
      "eval_runtime": 67.8585,
      "eval_samples_per_second": 14.737,
      "eval_steps_per_second": 14.737,
      "step": 9000
    },
    {
      "epoch": 0.05938742724710964,
      "grad_norm": 0.0511951707303524,
      "learning_rate": 0.0029880640369423315,
      "loss": 1.9788,
      "step": 9020
    },
    {
      "epoch": 0.05951910668668194,
      "grad_norm": 0.04770907759666443,
      "learning_rate": 0.0029879841845314135,
      "loss": 1.975,
      "step": 9040
    },
    {
      "epoch": 0.05965078612625425,
      "grad_norm": 0.052814580500125885,
      "learning_rate": 0.0029879040669753324,
      "loss": 1.9674,
      "step": 9060
    },
    {
      "epoch": 0.05978246556582655,
      "grad_norm": 0.05619759112596512,
      "learning_rate": 0.0029878236842883644,
      "loss": 1.9686,
      "step": 9080
    },
    {
      "epoch": 0.05991414500539886,
      "grad_norm": 0.052855897694826126,
      "learning_rate": 0.0029877430364848327,
      "loss": 1.9591,
      "step": 9100
    },
    {
      "epoch": 0.06004582444497116,
      "grad_norm": 0.06869558244943619,
      "learning_rate": 0.002987662123579108,
      "loss": 1.9603,
      "step": 9120
    },
    {
      "epoch": 0.060177503884543466,
      "grad_norm": 0.04893792048096657,
      "learning_rate": 0.002987580945585609,
      "loss": 1.9669,
      "step": 9140
    },
    {
      "epoch": 0.06030918332411577,
      "grad_norm": 0.06035724654793739,
      "learning_rate": 0.0029874995025188,
      "loss": 1.9693,
      "step": 9160
    },
    {
      "epoch": 0.060440862763688076,
      "grad_norm": 0.05003412812948227,
      "learning_rate": 0.002987417794393193,
      "loss": 1.961,
      "step": 9180
    },
    {
      "epoch": 0.06057254220326038,
      "grad_norm": 0.051062557846307755,
      "learning_rate": 0.002987335821223349,
      "loss": 1.9568,
      "step": 9200
    },
    {
      "epoch": 0.060704221642832686,
      "grad_norm": 0.05019828677177429,
      "learning_rate": 0.002987253583023874,
      "loss": 1.9529,
      "step": 9220
    },
    {
      "epoch": 0.06083590108240499,
      "grad_norm": 0.05402472987771034,
      "learning_rate": 0.0029871710798094224,
      "loss": 1.9579,
      "step": 9240
    },
    {
      "epoch": 0.060967580521977295,
      "grad_norm": 0.06498553603887558,
      "learning_rate": 0.002987088311594695,
      "loss": 1.9526,
      "step": 9260
    },
    {
      "epoch": 0.06109925996154961,
      "grad_norm": 0.06526941061019897,
      "learning_rate": 0.0029870052783944412,
      "loss": 1.9554,
      "step": 9280
    },
    {
      "epoch": 0.06123093940112191,
      "grad_norm": 0.051123786717653275,
      "learning_rate": 0.002986921980223456,
      "loss": 1.9449,
      "step": 9300
    },
    {
      "epoch": 0.061362618840694216,
      "grad_norm": 0.05042225867509842,
      "learning_rate": 0.002986838417096583,
      "loss": 1.952,
      "step": 9320
    },
    {
      "epoch": 0.06149429828026652,
      "grad_norm": 0.052252884954214096,
      "learning_rate": 0.002986754589028711,
      "loss": 1.9459,
      "step": 9340
    },
    {
      "epoch": 0.061625977719838826,
      "grad_norm": 0.049406830221414566,
      "learning_rate": 0.002986670496034779,
      "loss": 1.9478,
      "step": 9360
    },
    {
      "epoch": 0.06175765715941113,
      "grad_norm": 0.051424238830804825,
      "learning_rate": 0.0029865861381297714,
      "loss": 1.9434,
      "step": 9380
    },
    {
      "epoch": 0.061889336598983435,
      "grad_norm": 0.05527883395552635,
      "learning_rate": 0.0029865015153287193,
      "loss": 1.9428,
      "step": 9400
    },
    {
      "epoch": 0.06202101603855574,
      "grad_norm": 0.048182398080825806,
      "learning_rate": 0.0029864166276467024,
      "loss": 1.9404,
      "step": 9420
    },
    {
      "epoch": 0.062152695478128045,
      "grad_norm": 0.04890589416027069,
      "learning_rate": 0.002986331475098846,
      "loss": 1.9342,
      "step": 9440
    },
    {
      "epoch": 0.06228437491770035,
      "grad_norm": 0.052375294268131256,
      "learning_rate": 0.0029862460577003247,
      "loss": 1.9291,
      "step": 9460
    },
    {
      "epoch": 0.062416054357272654,
      "grad_norm": 0.0541619248688221,
      "learning_rate": 0.002986160375466358,
      "loss": 1.883,
      "step": 9480
    },
    {
      "epoch": 0.06254773379684496,
      "grad_norm": 0.05440934747457504,
      "learning_rate": 0.002986074428412214,
      "loss": 1.8321,
      "step": 9500
    },
    {
      "epoch": 0.06267941323641726,
      "grad_norm": 0.06579524278640747,
      "learning_rate": 0.0029859882165532087,
      "loss": 1.8184,
      "step": 9520
    },
    {
      "epoch": 0.06281109267598957,
      "grad_norm": 0.050777874886989594,
      "learning_rate": 0.0029859017399047026,
      "loss": 1.8232,
      "step": 9540
    },
    {
      "epoch": 0.06294277211556187,
      "grad_norm": 0.05892116203904152,
      "learning_rate": 0.002985814998482106,
      "loss": 1.8241,
      "step": 9560
    },
    {
      "epoch": 0.06307445155513418,
      "grad_norm": 0.054253749549388885,
      "learning_rate": 0.0029857279923008757,
      "loss": 1.812,
      "step": 9580
    },
    {
      "epoch": 0.06320613099470648,
      "grad_norm": 0.05559428408741951,
      "learning_rate": 0.0029856407213765144,
      "loss": 1.8166,
      "step": 9600
    },
    {
      "epoch": 0.06333781043427879,
      "grad_norm": 0.0527789331972599,
      "learning_rate": 0.0029855531857245735,
      "loss": 1.8244,
      "step": 9620
    },
    {
      "epoch": 0.06346948987385109,
      "grad_norm": 0.05062944442033768,
      "learning_rate": 0.0029854653853606515,
      "loss": 1.8276,
      "step": 9640
    },
    {
      "epoch": 0.0636011693134234,
      "grad_norm": 0.05144268646836281,
      "learning_rate": 0.002985377320300393,
      "loss": 1.8178,
      "step": 9660
    },
    {
      "epoch": 0.0637328487529957,
      "grad_norm": 0.051822349429130554,
      "learning_rate": 0.0029852889905594903,
      "loss": 1.8238,
      "step": 9680
    },
    {
      "epoch": 0.06386452819256801,
      "grad_norm": 0.05682462826371193,
      "learning_rate": 0.0029852003961536844,
      "loss": 1.8207,
      "step": 9700
    },
    {
      "epoch": 0.06399620763214031,
      "grad_norm": 0.06278064101934433,
      "learning_rate": 0.00298511153709876,
      "loss": 1.8311,
      "step": 9720
    },
    {
      "epoch": 0.06412788707171262,
      "grad_norm": 0.052284616976976395,
      "learning_rate": 0.0029850224134105514,
      "loss": 1.8243,
      "step": 9740
    },
    {
      "epoch": 0.06425956651128492,
      "grad_norm": 0.06791289150714874,
      "learning_rate": 0.0029849330251049406,
      "loss": 1.8252,
      "step": 9760
    },
    {
      "epoch": 0.06439124595085724,
      "grad_norm": 0.05337266996502876,
      "learning_rate": 0.0029848433721978555,
      "loss": 1.829,
      "step": 9780
    },
    {
      "epoch": 0.06452292539042954,
      "grad_norm": 0.05477340891957283,
      "learning_rate": 0.002984753454705271,
      "loss": 1.8226,
      "step": 9800
    },
    {
      "epoch": 0.06465460483000185,
      "grad_norm": 0.055485792458057404,
      "learning_rate": 0.00298466327264321,
      "loss": 1.8205,
      "step": 9820
    },
    {
      "epoch": 0.06478628426957415,
      "grad_norm": 0.05361747369170189,
      "learning_rate": 0.0029845728260277414,
      "loss": 1.8327,
      "step": 9840
    },
    {
      "epoch": 0.06491796370914646,
      "grad_norm": 0.06504777073860168,
      "learning_rate": 0.0029844821148749825,
      "loss": 1.8292,
      "step": 9860
    },
    {
      "epoch": 0.06504964314871876,
      "grad_norm": 0.05408887565135956,
      "learning_rate": 0.002984391139201097,
      "loss": 1.8295,
      "step": 9880
    },
    {
      "epoch": 0.06518132258829107,
      "grad_norm": 0.0615227073431015,
      "learning_rate": 0.0029842998990222964,
      "loss": 1.8233,
      "step": 9900
    },
    {
      "epoch": 0.06531300202786337,
      "grad_norm": 0.05818536877632141,
      "learning_rate": 0.002984208394354838,
      "loss": 1.8261,
      "step": 9920
    },
    {
      "epoch": 0.06544468146743568,
      "grad_norm": 0.057526398450136185,
      "learning_rate": 0.0029841166252150284,
      "loss": 1.8294,
      "step": 9940
    },
    {
      "epoch": 0.06557636090700798,
      "grad_norm": 0.05408645048737526,
      "learning_rate": 0.002984024591619219,
      "loss": 1.8248,
      "step": 9960
    },
    {
      "epoch": 0.06570804034658029,
      "grad_norm": 0.05534709617495537,
      "learning_rate": 0.0029839322935838095,
      "loss": 1.8263,
      "step": 9980
    },
    {
      "epoch": 0.06583971978615259,
      "grad_norm": 0.05544685944914818,
      "learning_rate": 0.002983839731125246,
      "loss": 1.8239,
      "step": 10000
    },
    {
      "epoch": 0.06583971978615259,
      "eval_loss": 1.968814492225647,
      "eval_runtime": 68.0355,
      "eval_samples_per_second": 14.698,
      "eval_steps_per_second": 14.698,
      "step": 10000
    },
    {
      "epoch": 0.0659713992257249,
      "grad_norm": 0.052884284406900406,
      "learning_rate": 0.002983746904260024,
      "loss": 1.8226,
      "step": 10020
    },
    {
      "epoch": 0.0661030786652972,
      "grad_norm": 0.06210578605532646,
      "learning_rate": 0.002983653813004683,
      "loss": 1.8562,
      "step": 10040
    },
    {
      "epoch": 0.0662347581048695,
      "grad_norm": 0.05301567539572716,
      "learning_rate": 0.002983560457375811,
      "loss": 1.9107,
      "step": 10060
    },
    {
      "epoch": 0.06636643754444181,
      "grad_norm": 0.056775398552417755,
      "learning_rate": 0.0029834668373900437,
      "loss": 1.9256,
      "step": 10080
    },
    {
      "epoch": 0.06649811698401412,
      "grad_norm": 0.05117955058813095,
      "learning_rate": 0.002983372953064063,
      "loss": 1.9391,
      "step": 10100
    },
    {
      "epoch": 0.06662979642358642,
      "grad_norm": 0.04930364713072777,
      "learning_rate": 0.0029832788044145985,
      "loss": 1.9415,
      "step": 10120
    },
    {
      "epoch": 0.06676147586315873,
      "grad_norm": 0.0544724240899086,
      "learning_rate": 0.002983184391458426,
      "loss": 1.9422,
      "step": 10140
    },
    {
      "epoch": 0.06689315530273103,
      "grad_norm": 0.05356570705771446,
      "learning_rate": 0.0029830897142123698,
      "loss": 1.9259,
      "step": 10160
    },
    {
      "epoch": 0.06702483474230334,
      "grad_norm": 0.05016016215085983,
      "learning_rate": 0.0029829947726933005,
      "loss": 1.9262,
      "step": 10180
    },
    {
      "epoch": 0.06715651418187564,
      "grad_norm": 0.05329560115933418,
      "learning_rate": 0.002982899566918135,
      "loss": 1.9342,
      "step": 10200
    },
    {
      "epoch": 0.06728819362144794,
      "grad_norm": 0.053507186472415924,
      "learning_rate": 0.0029828040969038386,
      "loss": 1.9217,
      "step": 10220
    },
    {
      "epoch": 0.06741987306102025,
      "grad_norm": 0.05592430755496025,
      "learning_rate": 0.002982708362667423,
      "loss": 1.9244,
      "step": 10240
    },
    {
      "epoch": 0.06755155250059255,
      "grad_norm": 0.05031822621822357,
      "learning_rate": 0.0029826123642259475,
      "loss": 1.9163,
      "step": 10260
    },
    {
      "epoch": 0.06768323194016486,
      "grad_norm": 0.05068674683570862,
      "learning_rate": 0.0029825161015965183,
      "loss": 1.9224,
      "step": 10280
    },
    {
      "epoch": 0.06781491137973716,
      "grad_norm": 0.05538354441523552,
      "learning_rate": 0.002982419574796288,
      "loss": 1.9166,
      "step": 10300
    },
    {
      "epoch": 0.06794659081930947,
      "grad_norm": 0.04972713813185692,
      "learning_rate": 0.002982322783842457,
      "loss": 1.9165,
      "step": 10320
    },
    {
      "epoch": 0.06807827025888177,
      "grad_norm": 0.05182690918445587,
      "learning_rate": 0.0029822257287522727,
      "loss": 1.9025,
      "step": 10340
    },
    {
      "epoch": 0.06820994969845408,
      "grad_norm": 0.0539088249206543,
      "learning_rate": 0.002982128409543029,
      "loss": 1.9039,
      "step": 10360
    },
    {
      "epoch": 0.06834162913802638,
      "grad_norm": 0.0529555045068264,
      "learning_rate": 0.0029820308262320677,
      "loss": 1.8987,
      "step": 10380
    },
    {
      "epoch": 0.06847330857759869,
      "grad_norm": 0.04998824745416641,
      "learning_rate": 0.0029819329788367773,
      "loss": 1.9015,
      "step": 10400
    },
    {
      "epoch": 0.06860498801717099,
      "grad_norm": 0.055027175694704056,
      "learning_rate": 0.0029818348673745928,
      "loss": 1.9022,
      "step": 10420
    },
    {
      "epoch": 0.0687366674567433,
      "grad_norm": 0.04936603456735611,
      "learning_rate": 0.002981736491862997,
      "loss": 1.8981,
      "step": 10440
    },
    {
      "epoch": 0.06886834689631562,
      "grad_norm": 0.05445968732237816,
      "learning_rate": 0.00298163785231952,
      "loss": 1.8922,
      "step": 10460
    },
    {
      "epoch": 0.06900002633588792,
      "grad_norm": 0.0514165423810482,
      "learning_rate": 0.0029815389487617377,
      "loss": 1.8941,
      "step": 10480
    },
    {
      "epoch": 0.06913170577546023,
      "grad_norm": 0.054617028683423996,
      "learning_rate": 0.002981439781207275,
      "loss": 1.8771,
      "step": 10500
    },
    {
      "epoch": 0.06926338521503253,
      "grad_norm": 0.051449310034513474,
      "learning_rate": 0.002981340349673801,
      "loss": 1.8979,
      "step": 10520
    },
    {
      "epoch": 0.06939506465460483,
      "grad_norm": 0.04865436255931854,
      "learning_rate": 0.0029812406541790347,
      "loss": 1.8929,
      "step": 10540
    },
    {
      "epoch": 0.06952674409417714,
      "grad_norm": 0.06338293850421906,
      "learning_rate": 0.0029811406947407404,
      "loss": 1.8849,
      "step": 10560
    },
    {
      "epoch": 0.06965842353374944,
      "grad_norm": 0.058406732976436615,
      "learning_rate": 0.00298104047137673,
      "loss": 1.8869,
      "step": 10580
    },
    {
      "epoch": 0.06979010297332175,
      "grad_norm": 0.05265038460493088,
      "learning_rate": 0.002980939984104863,
      "loss": 1.8807,
      "step": 10600
    },
    {
      "epoch": 0.06992178241289405,
      "grad_norm": 0.05860557034611702,
      "learning_rate": 0.0029808392329430445,
      "loss": 1.8926,
      "step": 10620
    },
    {
      "epoch": 0.07005346185246636,
      "grad_norm": 0.05650535598397255,
      "learning_rate": 0.0029807382179092277,
      "loss": 1.9136,
      "step": 10640
    },
    {
      "epoch": 0.07018514129203866,
      "grad_norm": 0.05268057808279991,
      "learning_rate": 0.0029806369390214123,
      "loss": 1.9313,
      "step": 10660
    },
    {
      "epoch": 0.07031682073161097,
      "grad_norm": 0.05681426078081131,
      "learning_rate": 0.002980535396297646,
      "loss": 1.9358,
      "step": 10680
    },
    {
      "epoch": 0.07044850017118327,
      "grad_norm": 0.06345421820878983,
      "learning_rate": 0.002980433589756022,
      "loss": 1.9411,
      "step": 10700
    },
    {
      "epoch": 0.07058017961075558,
      "grad_norm": 0.04853732883930206,
      "learning_rate": 0.002980331519414682,
      "loss": 1.9494,
      "step": 10720
    },
    {
      "epoch": 0.07071185905032788,
      "grad_norm": 0.047528378665447235,
      "learning_rate": 0.002980229185291814,
      "loss": 1.9364,
      "step": 10740
    },
    {
      "epoch": 0.07084353848990019,
      "grad_norm": 0.05809812992811203,
      "learning_rate": 0.002980126587405652,
      "loss": 1.9358,
      "step": 10760
    },
    {
      "epoch": 0.07097521792947249,
      "grad_norm": 0.05844232812523842,
      "learning_rate": 0.0029800237257744788,
      "loss": 1.9357,
      "step": 10780
    },
    {
      "epoch": 0.0711068973690448,
      "grad_norm": 0.04900583252310753,
      "learning_rate": 0.0029799206004166236,
      "loss": 1.9339,
      "step": 10800
    },
    {
      "epoch": 0.0712385768086171,
      "grad_norm": 0.056056343019008636,
      "learning_rate": 0.002979817211350462,
      "loss": 1.9213,
      "step": 10820
    },
    {
      "epoch": 0.0713702562481894,
      "grad_norm": 0.04929627478122711,
      "learning_rate": 0.002979713558594416,
      "loss": 1.9238,
      "step": 10840
    },
    {
      "epoch": 0.07150193568776171,
      "grad_norm": 0.051502469927072525,
      "learning_rate": 0.002979609642166958,
      "loss": 1.9297,
      "step": 10860
    },
    {
      "epoch": 0.07163361512733402,
      "grad_norm": 0.055544495582580566,
      "learning_rate": 0.002979505462086603,
      "loss": 1.9126,
      "step": 10880
    },
    {
      "epoch": 0.07176529456690632,
      "grad_norm": 0.05013841763138771,
      "learning_rate": 0.002979401018371915,
      "loss": 1.9167,
      "step": 10900
    },
    {
      "epoch": 0.07189697400647863,
      "grad_norm": 0.05101664736866951,
      "learning_rate": 0.002979296311041506,
      "loss": 1.9099,
      "step": 10920
    },
    {
      "epoch": 0.07202865344605093,
      "grad_norm": 0.06268607825040817,
      "learning_rate": 0.0029791913401140332,
      "loss": 1.9139,
      "step": 10940
    },
    {
      "epoch": 0.07216033288562324,
      "grad_norm": 0.05222015082836151,
      "learning_rate": 0.002979086105608202,
      "loss": 1.911,
      "step": 10960
    },
    {
      "epoch": 0.07229201232519554,
      "grad_norm": 0.054124996066093445,
      "learning_rate": 0.002978980607542763,
      "loss": 1.9097,
      "step": 10980
    },
    {
      "epoch": 0.07242369176476784,
      "grad_norm": 0.05831436067819595,
      "learning_rate": 0.0029788748459365164,
      "loss": 1.9166,
      "step": 11000
    },
    {
      "epoch": 0.07242369176476784,
      "eval_loss": 1.9284849166870117,
      "eval_runtime": 67.5794,
      "eval_samples_per_second": 14.797,
      "eval_steps_per_second": 14.797,
      "step": 11000
    },
    {
      "epoch": 0.07255537120434015,
      "grad_norm": 0.056505072861909866,
      "learning_rate": 0.0029787688208083073,
      "loss": 1.9091,
      "step": 11020
    },
    {
      "epoch": 0.07268705064391245,
      "grad_norm": 0.04782087355852127,
      "learning_rate": 0.002978662532177028,
      "loss": 1.9054,
      "step": 11040
    },
    {
      "epoch": 0.07281873008348476,
      "grad_norm": 0.05616345256567001,
      "learning_rate": 0.002978555980061619,
      "loss": 1.9027,
      "step": 11060
    },
    {
      "epoch": 0.07295040952305706,
      "grad_norm": 0.0646371841430664,
      "learning_rate": 0.002978449164481066,
      "loss": 1.8956,
      "step": 11080
    },
    {
      "epoch": 0.07308208896262937,
      "grad_norm": 0.0588691271841526,
      "learning_rate": 0.0029783420854544037,
      "loss": 1.8958,
      "step": 11100
    },
    {
      "epoch": 0.07321376840220169,
      "grad_norm": 0.05876161903142929,
      "learning_rate": 0.0029782347430007115,
      "loss": 1.898,
      "step": 11120
    },
    {
      "epoch": 0.07334544784177399,
      "grad_norm": 0.05978132411837578,
      "learning_rate": 0.0029781271371391175,
      "loss": 1.9033,
      "step": 11140
    },
    {
      "epoch": 0.0734771272813463,
      "grad_norm": 0.05216595530509949,
      "learning_rate": 0.0029780192678887964,
      "loss": 1.8997,
      "step": 11160
    },
    {
      "epoch": 0.0736088067209186,
      "grad_norm": 0.0518871508538723,
      "learning_rate": 0.002977911135268968,
      "loss": 1.9014,
      "step": 11180
    },
    {
      "epoch": 0.0737404861604909,
      "grad_norm": 0.05054188892245293,
      "learning_rate": 0.0029778027392989024,
      "loss": 1.8899,
      "step": 11200
    },
    {
      "epoch": 0.07387216560006321,
      "grad_norm": 0.059255532920360565,
      "learning_rate": 0.002977694079997913,
      "loss": 1.8928,
      "step": 11220
    },
    {
      "epoch": 0.07400384503963552,
      "grad_norm": 0.06282193213701248,
      "learning_rate": 0.002977585157385364,
      "loss": 1.9071,
      "step": 11240
    },
    {
      "epoch": 0.07413552447920782,
      "grad_norm": 0.048698440194129944,
      "learning_rate": 0.0029774759714806622,
      "loss": 1.9165,
      "step": 11260
    },
    {
      "epoch": 0.07426720391878013,
      "grad_norm": 0.05037577450275421,
      "learning_rate": 0.0029773665223032653,
      "loss": 1.9154,
      "step": 11280
    },
    {
      "epoch": 0.07439888335835243,
      "grad_norm": 0.049512796103954315,
      "learning_rate": 0.002977256809872675,
      "loss": 1.911,
      "step": 11300
    },
    {
      "epoch": 0.07453056279792473,
      "grad_norm": 0.05632052198052406,
      "learning_rate": 0.0029771468342084414,
      "loss": 1.9104,
      "step": 11320
    },
    {
      "epoch": 0.07466224223749704,
      "grad_norm": 0.050531480461359024,
      "learning_rate": 0.0029770365953301616,
      "loss": 1.9071,
      "step": 11340
    },
    {
      "epoch": 0.07479392167706934,
      "grad_norm": 0.057267602533102036,
      "learning_rate": 0.0029769260932574785,
      "loss": 1.9078,
      "step": 11360
    },
    {
      "epoch": 0.07492560111664165,
      "grad_norm": 0.05797697231173515,
      "learning_rate": 0.002976815328010083,
      "loss": 1.898,
      "step": 11380
    },
    {
      "epoch": 0.07505728055621395,
      "grad_norm": 0.04657141864299774,
      "learning_rate": 0.0029767042996077116,
      "loss": 1.8966,
      "step": 11400
    },
    {
      "epoch": 0.07518895999578626,
      "grad_norm": 0.05821000784635544,
      "learning_rate": 0.00297659300807015,
      "loss": 1.897,
      "step": 11420
    },
    {
      "epoch": 0.07532063943535856,
      "grad_norm": 0.050238363444805145,
      "learning_rate": 0.0029764814534172282,
      "loss": 1.8937,
      "step": 11440
    },
    {
      "epoch": 0.07545231887493087,
      "grad_norm": 0.04804101213812828,
      "learning_rate": 0.0029763696356688244,
      "loss": 1.8935,
      "step": 11460
    },
    {
      "epoch": 0.07558399831450317,
      "grad_norm": 0.057969000190496445,
      "learning_rate": 0.002976257554844864,
      "loss": 1.8896,
      "step": 11480
    },
    {
      "epoch": 0.07571567775407548,
      "grad_norm": 0.05278369039297104,
      "learning_rate": 0.002976145210965319,
      "loss": 1.8918,
      "step": 11500
    },
    {
      "epoch": 0.07584735719364778,
      "grad_norm": 0.05272908881306648,
      "learning_rate": 0.0029760326040502065,
      "loss": 1.8896,
      "step": 11520
    },
    {
      "epoch": 0.07597903663322009,
      "grad_norm": 0.0534030944108963,
      "learning_rate": 0.0029759197341195933,
      "loss": 1.8833,
      "step": 11540
    },
    {
      "epoch": 0.07611071607279239,
      "grad_norm": 0.05342519283294678,
      "learning_rate": 0.002975806601193592,
      "loss": 1.8828,
      "step": 11560
    },
    {
      "epoch": 0.0762423955123647,
      "grad_norm": 0.05289285629987717,
      "learning_rate": 0.002975693205292361,
      "loss": 1.8786,
      "step": 11580
    },
    {
      "epoch": 0.076374074951937,
      "grad_norm": 0.05132855102419853,
      "learning_rate": 0.0029755795464361063,
      "loss": 1.882,
      "step": 11600
    },
    {
      "epoch": 0.0765057543915093,
      "grad_norm": 0.057877812534570694,
      "learning_rate": 0.0029754656246450822,
      "loss": 1.8881,
      "step": 11620
    },
    {
      "epoch": 0.07663743383108161,
      "grad_norm": 0.054198529571294785,
      "learning_rate": 0.002975351439939587,
      "loss": 1.8849,
      "step": 11640
    },
    {
      "epoch": 0.07676911327065392,
      "grad_norm": 0.04742525890469551,
      "learning_rate": 0.0029752369923399685,
      "loss": 1.8819,
      "step": 11660
    },
    {
      "epoch": 0.07690079271022622,
      "grad_norm": 0.057487789541482925,
      "learning_rate": 0.002975122281866619,
      "loss": 1.8751,
      "step": 11680
    },
    {
      "epoch": 0.07703247214979853,
      "grad_norm": 0.04776060953736305,
      "learning_rate": 0.002975007308539981,
      "loss": 1.8726,
      "step": 11700
    },
    {
      "epoch": 0.07716415158937083,
      "grad_norm": 0.05077700316905975,
      "learning_rate": 0.002974892072380539,
      "loss": 1.876,
      "step": 11720
    },
    {
      "epoch": 0.07729583102894313,
      "grad_norm": 0.0542939268052578,
      "learning_rate": 0.0029747765734088284,
      "loss": 1.8788,
      "step": 11740
    },
    {
      "epoch": 0.07742751046851544,
      "grad_norm": 0.04847322776913643,
      "learning_rate": 0.0029746608116454306,
      "loss": 1.8705,
      "step": 11760
    },
    {
      "epoch": 0.07755918990808774,
      "grad_norm": 0.0494987815618515,
      "learning_rate": 0.0029745447871109727,
      "loss": 1.8776,
      "step": 11780
    },
    {
      "epoch": 0.07769086934766006,
      "grad_norm": 0.05288725346326828,
      "learning_rate": 0.0029744284998261287,
      "loss": 1.8733,
      "step": 11800
    },
    {
      "epoch": 0.07782254878723237,
      "grad_norm": 0.0571809746325016,
      "learning_rate": 0.0029743119498116203,
      "loss": 1.8761,
      "step": 11820
    },
    {
      "epoch": 0.07795422822680467,
      "grad_norm": 0.050889961421489716,
      "learning_rate": 0.0029741951370882164,
      "loss": 1.8692,
      "step": 11840
    },
    {
      "epoch": 0.07808590766637698,
      "grad_norm": 0.0550447441637516,
      "learning_rate": 0.002974078061676731,
      "loss": 1.8739,
      "step": 11860
    },
    {
      "epoch": 0.07821758710594928,
      "grad_norm": 0.05145235359668732,
      "learning_rate": 0.002973960723598026,
      "loss": 1.8653,
      "step": 11880
    },
    {
      "epoch": 0.07834926654552159,
      "grad_norm": 0.05060429126024246,
      "learning_rate": 0.0029738431228730107,
      "loss": 1.8529,
      "step": 11900
    },
    {
      "epoch": 0.07848094598509389,
      "grad_norm": 0.05473101884126663,
      "learning_rate": 0.0029737252595226395,
      "loss": 1.8513,
      "step": 11920
    },
    {
      "epoch": 0.0786126254246662,
      "grad_norm": 0.05369679257273674,
      "learning_rate": 0.002973607133567915,
      "loss": 1.8473,
      "step": 11940
    },
    {
      "epoch": 0.0787443048642385,
      "grad_norm": 0.06132115051150322,
      "learning_rate": 0.0029734887450298866,
      "loss": 1.8426,
      "step": 11960
    },
    {
      "epoch": 0.0788759843038108,
      "grad_norm": 0.049720194190740585,
      "learning_rate": 0.0029733700939296487,
      "loss": 1.8412,
      "step": 11980
    },
    {
      "epoch": 0.07900766374338311,
      "grad_norm": 0.048152726143598557,
      "learning_rate": 0.0029732511802883456,
      "loss": 1.8388,
      "step": 12000
    },
    {
      "epoch": 0.07900766374338311,
      "eval_loss": 1.7857414484024048,
      "eval_runtime": 68.3192,
      "eval_samples_per_second": 14.637,
      "eval_steps_per_second": 14.637,
      "step": 12000
    },
    {
      "epoch": 0.07913934318295542,
      "grad_norm": 0.05637095496058464,
      "learning_rate": 0.0029731320041271655,
      "loss": 1.8317,
      "step": 12020
    },
    {
      "epoch": 0.07927102262252772,
      "grad_norm": 0.05279828608036041,
      "learning_rate": 0.002973012565467345,
      "loss": 1.8265,
      "step": 12040
    },
    {
      "epoch": 0.07940270206210003,
      "grad_norm": 0.05788380280137062,
      "learning_rate": 0.0029728928643301664,
      "loss": 1.8285,
      "step": 12060
    },
    {
      "epoch": 0.07953438150167233,
      "grad_norm": 0.05116511508822441,
      "learning_rate": 0.0029727729007369597,
      "loss": 1.8305,
      "step": 12080
    },
    {
      "epoch": 0.07966606094124463,
      "grad_norm": 0.05228397622704506,
      "learning_rate": 0.0029726526747091016,
      "loss": 1.8299,
      "step": 12100
    },
    {
      "epoch": 0.07979774038081694,
      "grad_norm": 0.06771951913833618,
      "learning_rate": 0.0029725321862680144,
      "loss": 1.8221,
      "step": 12120
    },
    {
      "epoch": 0.07992941982038924,
      "grad_norm": 0.05031878501176834,
      "learning_rate": 0.002972411435435169,
      "loss": 1.8244,
      "step": 12140
    },
    {
      "epoch": 0.08006109925996155,
      "grad_norm": 0.057394348084926605,
      "learning_rate": 0.0029722904222320825,
      "loss": 1.8279,
      "step": 12160
    },
    {
      "epoch": 0.08019277869953385,
      "grad_norm": 0.05116160586476326,
      "learning_rate": 0.0029721691466803165,
      "loss": 1.8192,
      "step": 12180
    },
    {
      "epoch": 0.08032445813910616,
      "grad_norm": 0.05168084427714348,
      "learning_rate": 0.002972047608801483,
      "loss": 1.8265,
      "step": 12200
    },
    {
      "epoch": 0.08045613757867846,
      "grad_norm": 0.053319819271564484,
      "learning_rate": 0.0029719258086172377,
      "loss": 1.8237,
      "step": 12220
    },
    {
      "epoch": 0.08058781701825077,
      "grad_norm": 0.09060586988925934,
      "learning_rate": 0.002971803746149285,
      "loss": 1.8176,
      "step": 12240
    },
    {
      "epoch": 0.08071949645782307,
      "grad_norm": 0.05889497324824333,
      "learning_rate": 0.002971681421419375,
      "loss": 1.8166,
      "step": 12260
    },
    {
      "epoch": 0.08085117589739538,
      "grad_norm": 0.056241389364004135,
      "learning_rate": 0.002971558834449305,
      "loss": 1.8189,
      "step": 12280
    },
    {
      "epoch": 0.08098285533696768,
      "grad_norm": 0.07331902533769608,
      "learning_rate": 0.002971435985260919,
      "loss": 1.8173,
      "step": 12300
    },
    {
      "epoch": 0.08111453477653999,
      "grad_norm": 0.05276619270443916,
      "learning_rate": 0.002971312873876107,
      "loss": 1.8194,
      "step": 12320
    },
    {
      "epoch": 0.08124621421611229,
      "grad_norm": 0.052701130509376526,
      "learning_rate": 0.002971189500316807,
      "loss": 1.808,
      "step": 12340
    },
    {
      "epoch": 0.0813778936556846,
      "grad_norm": 0.05825614184141159,
      "learning_rate": 0.0029710658646050028,
      "loss": 1.8206,
      "step": 12360
    },
    {
      "epoch": 0.0815095730952569,
      "grad_norm": 0.05325302854180336,
      "learning_rate": 0.002970941966762725,
      "loss": 1.8072,
      "step": 12380
    },
    {
      "epoch": 0.0816412525348292,
      "grad_norm": 0.05424511805176735,
      "learning_rate": 0.0029708178068120514,
      "loss": 1.814,
      "step": 12400
    },
    {
      "epoch": 0.08177293197440151,
      "grad_norm": 0.05035754665732384,
      "learning_rate": 0.002970693384775106,
      "loss": 1.8281,
      "step": 12420
    },
    {
      "epoch": 0.08190461141397382,
      "grad_norm": 0.0715329572558403,
      "learning_rate": 0.0029705687006740594,
      "loss": 1.8342,
      "step": 12440
    },
    {
      "epoch": 0.08203629085354612,
      "grad_norm": 0.05069407820701599,
      "learning_rate": 0.002970443754531129,
      "loss": 1.8306,
      "step": 12460
    },
    {
      "epoch": 0.08216797029311844,
      "grad_norm": 0.05642094835639,
      "learning_rate": 0.00297031854636858,
      "loss": 1.8252,
      "step": 12480
    },
    {
      "epoch": 0.08229964973269074,
      "grad_norm": 0.05098734796047211,
      "learning_rate": 0.0029701930762087232,
      "loss": 1.8283,
      "step": 12500
    },
    {
      "epoch": 0.08243132917226305,
      "grad_norm": 0.05360301956534386,
      "learning_rate": 0.002970067344073915,
      "loss": 1.8121,
      "step": 12520
    },
    {
      "epoch": 0.08256300861183535,
      "grad_norm": 0.064446359872818,
      "learning_rate": 0.002969941349986561,
      "loss": 1.8167,
      "step": 12540
    },
    {
      "epoch": 0.08269468805140766,
      "grad_norm": 0.05934213101863861,
      "learning_rate": 0.002969815093969112,
      "loss": 1.8104,
      "step": 12560
    },
    {
      "epoch": 0.08282636749097996,
      "grad_norm": 0.05697745084762573,
      "learning_rate": 0.0029696885760440647,
      "loss": 1.8118,
      "step": 12580
    },
    {
      "epoch": 0.08295804693055227,
      "grad_norm": 0.05313008651137352,
      "learning_rate": 0.0029695617962339645,
      "loss": 1.8073,
      "step": 12600
    },
    {
      "epoch": 0.08308972637012457,
      "grad_norm": 0.05284099653363228,
      "learning_rate": 0.002969434754561402,
      "loss": 1.8142,
      "step": 12620
    },
    {
      "epoch": 0.08322140580969688,
      "grad_norm": 0.054485708475112915,
      "learning_rate": 0.0029693074510490154,
      "loss": 1.8067,
      "step": 12640
    },
    {
      "epoch": 0.08335308524926918,
      "grad_norm": 0.050835754722356796,
      "learning_rate": 0.0029691798857194884,
      "loss": 1.8094,
      "step": 12660
    },
    {
      "epoch": 0.08348476468884149,
      "grad_norm": 0.05048472061753273,
      "learning_rate": 0.0029690520585955514,
      "loss": 1.8019,
      "step": 12680
    },
    {
      "epoch": 0.08361644412841379,
      "grad_norm": 0.05250314995646477,
      "learning_rate": 0.0029689239696999836,
      "loss": 1.7949,
      "step": 12700
    },
    {
      "epoch": 0.0837481235679861,
      "grad_norm": 0.05029289051890373,
      "learning_rate": 0.0029687956190556078,
      "loss": 1.8036,
      "step": 12720
    },
    {
      "epoch": 0.0838798030075584,
      "grad_norm": 0.053878892213106155,
      "learning_rate": 0.002968667006685296,
      "loss": 1.8019,
      "step": 12740
    },
    {
      "epoch": 0.0840114824471307,
      "grad_norm": 0.05751722306013107,
      "learning_rate": 0.0029685381326119645,
      "loss": 1.8014,
      "step": 12760
    },
    {
      "epoch": 0.08414316188670301,
      "grad_norm": 0.05387134477496147,
      "learning_rate": 0.002968408996858579,
      "loss": 1.7998,
      "step": 12780
    },
    {
      "epoch": 0.08427484132627532,
      "grad_norm": 0.056205861270427704,
      "learning_rate": 0.002968279599448149,
      "loss": 1.8051,
      "step": 12800
    },
    {
      "epoch": 0.08440652076584762,
      "grad_norm": 0.052303776144981384,
      "learning_rate": 0.0029681499404037325,
      "loss": 1.7967,
      "step": 12820
    },
    {
      "epoch": 0.08453820020541992,
      "grad_norm": 0.06636688113212585,
      "learning_rate": 0.002968020019748434,
      "loss": 1.799,
      "step": 12840
    },
    {
      "epoch": 0.08466987964499223,
      "grad_norm": 0.05160198360681534,
      "learning_rate": 0.0029678898375054036,
      "loss": 1.7987,
      "step": 12860
    },
    {
      "epoch": 0.08480155908456453,
      "grad_norm": 0.05559029057621956,
      "learning_rate": 0.002967759393697839,
      "loss": 1.8034,
      "step": 12880
    },
    {
      "epoch": 0.08493323852413684,
      "grad_norm": 0.057246867567300797,
      "learning_rate": 0.002967628688348983,
      "loss": 1.7888,
      "step": 12900
    },
    {
      "epoch": 0.08506491796370914,
      "grad_norm": 0.06301821768283844,
      "learning_rate": 0.0029674977214821274,
      "loss": 1.801,
      "step": 12920
    },
    {
      "epoch": 0.08519659740328145,
      "grad_norm": 0.055293645709753036,
      "learning_rate": 0.0029673664931206085,
      "loss": 1.7955,
      "step": 12940
    },
    {
      "epoch": 0.08532827684285375,
      "grad_norm": 0.05522451177239418,
      "learning_rate": 0.0029672350032878105,
      "loss": 1.788,
      "step": 12960
    },
    {
      "epoch": 0.08545995628242606,
      "grad_norm": 0.057026926428079605,
      "learning_rate": 0.0029671032520071634,
      "loss": 1.7898,
      "step": 12980
    },
    {
      "epoch": 0.08559163572199836,
      "grad_norm": 0.06190698221325874,
      "learning_rate": 0.002966971239302144,
      "loss": 1.8188,
      "step": 13000
    },
    {
      "epoch": 0.08559163572199836,
      "eval_loss": 1.6987569332122803,
      "eval_runtime": 70.9806,
      "eval_samples_per_second": 14.088,
      "eval_steps_per_second": 14.088,
      "step": 13000
    },
    {
      "epoch": 0.08572331516157067,
      "grad_norm": 0.05302195996046066,
      "learning_rate": 0.002966838965196276,
      "loss": 1.8589,
      "step": 13020
    },
    {
      "epoch": 0.08585499460114297,
      "grad_norm": 0.051264405250549316,
      "learning_rate": 0.0029667064297131296,
      "loss": 1.8669,
      "step": 13040
    },
    {
      "epoch": 0.08598667404071528,
      "grad_norm": 0.05437634140253067,
      "learning_rate": 0.0029665736328763205,
      "loss": 1.8685,
      "step": 13060
    },
    {
      "epoch": 0.08611835348028758,
      "grad_norm": 0.05606294050812721,
      "learning_rate": 0.0029664405747095134,
      "loss": 1.8787,
      "step": 13080
    },
    {
      "epoch": 0.08625003291985989,
      "grad_norm": 0.053556881844997406,
      "learning_rate": 0.0029663072552364163,
      "loss": 1.8788,
      "step": 13100
    },
    {
      "epoch": 0.08638171235943219,
      "grad_norm": 0.06021646037697792,
      "learning_rate": 0.0029661736744807867,
      "loss": 1.8733,
      "step": 13120
    },
    {
      "epoch": 0.08651339179900451,
      "grad_norm": 0.06918266415596008,
      "learning_rate": 0.002966039832466427,
      "loss": 1.8827,
      "step": 13140
    },
    {
      "epoch": 0.08664507123857682,
      "grad_norm": 0.052268948405981064,
      "learning_rate": 0.002965905729217187,
      "loss": 1.8688,
      "step": 13160
    },
    {
      "epoch": 0.08677675067814912,
      "grad_norm": 0.0564788356423378,
      "learning_rate": 0.0029657713647569623,
      "loss": 1.8599,
      "step": 13180
    },
    {
      "epoch": 0.08690843011772142,
      "grad_norm": 0.04878311604261398,
      "learning_rate": 0.0029656367391096955,
      "loss": 1.8654,
      "step": 13200
    },
    {
      "epoch": 0.08704010955729373,
      "grad_norm": 0.051273491233587265,
      "learning_rate": 0.002965501852299376,
      "loss": 1.8634,
      "step": 13220
    },
    {
      "epoch": 0.08717178899686603,
      "grad_norm": 0.06505248695611954,
      "learning_rate": 0.002965366704350039,
      "loss": 1.859,
      "step": 13240
    },
    {
      "epoch": 0.08730346843643834,
      "grad_norm": 0.05197957530617714,
      "learning_rate": 0.002965231295285767,
      "loss": 1.8631,
      "step": 13260
    },
    {
      "epoch": 0.08743514787601064,
      "grad_norm": 0.05774524062871933,
      "learning_rate": 0.0029650956251306877,
      "loss": 1.8508,
      "step": 13280
    },
    {
      "epoch": 0.08756682731558295,
      "grad_norm": 0.04711959511041641,
      "learning_rate": 0.0029649596939089772,
      "loss": 1.8488,
      "step": 13300
    },
    {
      "epoch": 0.08769850675515525,
      "grad_norm": 0.04737945273518562,
      "learning_rate": 0.0029648235016448573,
      "loss": 1.8546,
      "step": 13320
    },
    {
      "epoch": 0.08783018619472756,
      "grad_norm": 0.051619525998830795,
      "learning_rate": 0.0029646870483625953,
      "loss": 1.8514,
      "step": 13340
    },
    {
      "epoch": 0.08796186563429986,
      "grad_norm": 0.05580944940447807,
      "learning_rate": 0.002964550334086507,
      "loss": 1.85,
      "step": 13360
    },
    {
      "epoch": 0.08809354507387217,
      "grad_norm": 0.05655309185385704,
      "learning_rate": 0.002964413358840953,
      "loss": 1.8456,
      "step": 13380
    },
    {
      "epoch": 0.08822522451344447,
      "grad_norm": 0.05536419898271561,
      "learning_rate": 0.0029642761226503413,
      "loss": 1.8375,
      "step": 13400
    },
    {
      "epoch": 0.08835690395301678,
      "grad_norm": 0.055286381393671036,
      "learning_rate": 0.002964138625539126,
      "loss": 1.8406,
      "step": 13420
    },
    {
      "epoch": 0.08848858339258908,
      "grad_norm": 0.07900959998369217,
      "learning_rate": 0.0029640008675318077,
      "loss": 1.847,
      "step": 13440
    },
    {
      "epoch": 0.08862026283216139,
      "grad_norm": 0.05409523472189903,
      "learning_rate": 0.002963862848652934,
      "loss": 1.8483,
      "step": 13460
    },
    {
      "epoch": 0.08875194227173369,
      "grad_norm": 0.04646017774939537,
      "learning_rate": 0.0029637245689270978,
      "loss": 1.8357,
      "step": 13480
    },
    {
      "epoch": 0.088883621711306,
      "grad_norm": 0.05184982717037201,
      "learning_rate": 0.0029635860283789405,
      "loss": 1.8407,
      "step": 13500
    },
    {
      "epoch": 0.0890153011508783,
      "grad_norm": 0.05196777358651161,
      "learning_rate": 0.0029634472270331476,
      "loss": 1.8404,
      "step": 13520
    },
    {
      "epoch": 0.0891469805904506,
      "grad_norm": 0.0526873879134655,
      "learning_rate": 0.0029633081649144528,
      "loss": 1.8403,
      "step": 13540
    },
    {
      "epoch": 0.08927866003002291,
      "grad_norm": 0.06644752621650696,
      "learning_rate": 0.0029631688420476356,
      "loss": 1.8312,
      "step": 13560
    },
    {
      "epoch": 0.08941033946959522,
      "grad_norm": 0.06348369270563126,
      "learning_rate": 0.002963029258457522,
      "loss": 1.8343,
      "step": 13580
    },
    {
      "epoch": 0.08954201890916752,
      "grad_norm": 0.06941412389278412,
      "learning_rate": 0.002962889414168985,
      "loss": 1.8488,
      "step": 13600
    },
    {
      "epoch": 0.08967369834873982,
      "grad_norm": 0.0526285357773304,
      "learning_rate": 0.0029627493092069432,
      "loss": 1.8645,
      "step": 13620
    },
    {
      "epoch": 0.08980537778831213,
      "grad_norm": 0.04776741564273834,
      "learning_rate": 0.002962608943596362,
      "loss": 1.8647,
      "step": 13640
    },
    {
      "epoch": 0.08993705722788443,
      "grad_norm": 0.05642177537083626,
      "learning_rate": 0.0029624683173622526,
      "loss": 1.8651,
      "step": 13660
    },
    {
      "epoch": 0.09006873666745674,
      "grad_norm": 0.04894222691655159,
      "learning_rate": 0.0029623274305296743,
      "loss": 1.8652,
      "step": 13680
    },
    {
      "epoch": 0.09020041610702904,
      "grad_norm": 0.05177977308630943,
      "learning_rate": 0.0029621862831237314,
      "loss": 1.8606,
      "step": 13700
    },
    {
      "epoch": 0.09033209554660135,
      "grad_norm": 0.05325636267662048,
      "learning_rate": 0.0029620448751695757,
      "loss": 1.8562,
      "step": 13720
    },
    {
      "epoch": 0.09046377498617365,
      "grad_norm": 0.053102049976587296,
      "learning_rate": 0.0029619032066924037,
      "loss": 1.8607,
      "step": 13740
    },
    {
      "epoch": 0.09059545442574596,
      "grad_norm": 0.06039892137050629,
      "learning_rate": 0.0029617612777174604,
      "loss": 1.8598,
      "step": 13760
    },
    {
      "epoch": 0.09072713386531826,
      "grad_norm": 0.05456007272005081,
      "learning_rate": 0.0029616190882700358,
      "loss": 1.855,
      "step": 13780
    },
    {
      "epoch": 0.09085881330489057,
      "grad_norm": 0.05931520834565163,
      "learning_rate": 0.002961476638375467,
      "loss": 1.8562,
      "step": 13800
    },
    {
      "epoch": 0.09099049274446289,
      "grad_norm": 0.05977236106991768,
      "learning_rate": 0.0029613339280591365,
      "loss": 1.8562,
      "step": 13820
    },
    {
      "epoch": 0.09112217218403519,
      "grad_norm": 0.06268831342458725,
      "learning_rate": 0.0029611909573464745,
      "loss": 1.8472,
      "step": 13840
    },
    {
      "epoch": 0.0912538516236075,
      "grad_norm": 0.05633818358182907,
      "learning_rate": 0.002961047726262958,
      "loss": 1.8403,
      "step": 13860
    },
    {
      "epoch": 0.0913855310631798,
      "grad_norm": 0.0659051239490509,
      "learning_rate": 0.0029609042348341078,
      "loss": 1.8382,
      "step": 13880
    },
    {
      "epoch": 0.0915172105027521,
      "grad_norm": 0.050592195242643356,
      "learning_rate": 0.0029607604830854937,
      "loss": 1.8452,
      "step": 13900
    },
    {
      "epoch": 0.09164888994232441,
      "grad_norm": 0.049383338540792465,
      "learning_rate": 0.0029606164710427314,
      "loss": 1.8443,
      "step": 13920
    },
    {
      "epoch": 0.09178056938189671,
      "grad_norm": 0.06528118252754211,
      "learning_rate": 0.002960472198731481,
      "loss": 1.8422,
      "step": 13940
    },
    {
      "epoch": 0.09191224882146902,
      "grad_norm": 0.05411006882786751,
      "learning_rate": 0.002960327666177452,
      "loss": 1.8465,
      "step": 13960
    },
    {
      "epoch": 0.09204392826104132,
      "grad_norm": 0.0483255498111248,
      "learning_rate": 0.0029601828734063984,
      "loss": 1.8427,
      "step": 13980
    },
    {
      "epoch": 0.09217560770061363,
      "grad_norm": 0.050317153334617615,
      "learning_rate": 0.0029600378204441206,
      "loss": 1.8339,
      "step": 14000
    },
    {
      "epoch": 0.09217560770061363,
      "eval_loss": 1.785549283027649,
      "eval_runtime": 68.197,
      "eval_samples_per_second": 14.663,
      "eval_steps_per_second": 14.663,
      "step": 14000
    },
    {
      "epoch": 0.09230728714018593,
      "grad_norm": 0.05255632847547531,
      "learning_rate": 0.0029598925073164657,
      "loss": 1.8319,
      "step": 14020
    },
    {
      "epoch": 0.09243896657975824,
      "grad_norm": 0.053322333842515945,
      "learning_rate": 0.0029597469340493275,
      "loss": 1.8343,
      "step": 14040
    },
    {
      "epoch": 0.09257064601933054,
      "grad_norm": 0.046458128839731216,
      "learning_rate": 0.0029596011006686454,
      "loss": 1.8335,
      "step": 14060
    },
    {
      "epoch": 0.09270232545890285,
      "grad_norm": 0.051327046006917953,
      "learning_rate": 0.0029594550072004062,
      "loss": 1.8295,
      "step": 14080
    },
    {
      "epoch": 0.09283400489847515,
      "grad_norm": 0.06848567724227905,
      "learning_rate": 0.002959308653670642,
      "loss": 1.8375,
      "step": 14100
    },
    {
      "epoch": 0.09296568433804746,
      "grad_norm": 0.05707765370607376,
      "learning_rate": 0.0029591620401054316,
      "loss": 1.8341,
      "step": 14120
    },
    {
      "epoch": 0.09309736377761976,
      "grad_norm": 0.050898268818855286,
      "learning_rate": 0.0029590151665309004,
      "loss": 1.8324,
      "step": 14140
    },
    {
      "epoch": 0.09322904321719207,
      "grad_norm": 0.06690908968448639,
      "learning_rate": 0.0029588680329732196,
      "loss": 1.8357,
      "step": 14160
    },
    {
      "epoch": 0.09336072265676437,
      "grad_norm": 0.05576496198773384,
      "learning_rate": 0.0029587206394586073,
      "loss": 1.827,
      "step": 14180
    },
    {
      "epoch": 0.09349240209633668,
      "grad_norm": 0.060783933848142624,
      "learning_rate": 0.002958572986013328,
      "loss": 1.827,
      "step": 14200
    },
    {
      "epoch": 0.09362408153590898,
      "grad_norm": 0.05840138718485832,
      "learning_rate": 0.0029584250726636917,
      "loss": 1.8005,
      "step": 14220
    },
    {
      "epoch": 0.09375576097548129,
      "grad_norm": 0.06067179888486862,
      "learning_rate": 0.0029582768994360557,
      "loss": 1.7913,
      "step": 14240
    },
    {
      "epoch": 0.09388744041505359,
      "grad_norm": 0.04877826198935509,
      "learning_rate": 0.002958128466356822,
      "loss": 1.7808,
      "step": 14260
    },
    {
      "epoch": 0.0940191198546259,
      "grad_norm": 0.07199737429618835,
      "learning_rate": 0.0029579797734524413,
      "loss": 1.7692,
      "step": 14280
    },
    {
      "epoch": 0.0941507992941982,
      "grad_norm": 0.05159222334623337,
      "learning_rate": 0.002957830820749409,
      "loss": 1.7735,
      "step": 14300
    },
    {
      "epoch": 0.0942824787337705,
      "grad_norm": 0.05802571773529053,
      "learning_rate": 0.0029576816082742662,
      "loss": 1.7648,
      "step": 14320
    },
    {
      "epoch": 0.09441415817334281,
      "grad_norm": 0.05139509588479996,
      "learning_rate": 0.0029575321360536025,
      "loss": 1.7667,
      "step": 14340
    },
    {
      "epoch": 0.09454583761291512,
      "grad_norm": 0.06891103833913803,
      "learning_rate": 0.002957382404114052,
      "loss": 1.766,
      "step": 14360
    },
    {
      "epoch": 0.09467751705248742,
      "grad_norm": 0.054211389273405075,
      "learning_rate": 0.0029572324124822955,
      "loss": 1.7576,
      "step": 14380
    },
    {
      "epoch": 0.09480919649205972,
      "grad_norm": 0.0536891408264637,
      "learning_rate": 0.0029570821611850594,
      "loss": 1.7597,
      "step": 14400
    },
    {
      "epoch": 0.09494087593163203,
      "grad_norm": 0.052394088357686996,
      "learning_rate": 0.0029569316502491183,
      "loss": 1.7645,
      "step": 14420
    },
    {
      "epoch": 0.09507255537120433,
      "grad_norm": 0.05916254222393036,
      "learning_rate": 0.002956780879701291,
      "loss": 1.7631,
      "step": 14440
    },
    {
      "epoch": 0.09520423481077664,
      "grad_norm": 0.060256022959947586,
      "learning_rate": 0.0029566298495684444,
      "loss": 1.7566,
      "step": 14460
    },
    {
      "epoch": 0.09533591425034894,
      "grad_norm": 0.055872827768325806,
      "learning_rate": 0.0029564785598774897,
      "loss": 1.7536,
      "step": 14480
    },
    {
      "epoch": 0.09546759368992126,
      "grad_norm": 0.053552061319351196,
      "learning_rate": 0.0029563270106553858,
      "loss": 1.7567,
      "step": 14500
    },
    {
      "epoch": 0.09559927312949357,
      "grad_norm": 0.07257280498743057,
      "learning_rate": 0.0029561752019291376,
      "loss": 1.7526,
      "step": 14520
    },
    {
      "epoch": 0.09573095256906587,
      "grad_norm": 0.061258714646101,
      "learning_rate": 0.002956023133725795,
      "loss": 1.7556,
      "step": 14540
    },
    {
      "epoch": 0.09586263200863818,
      "grad_norm": 0.05536385625600815,
      "learning_rate": 0.002955870806072457,
      "loss": 1.7482,
      "step": 14560
    },
    {
      "epoch": 0.09599431144821048,
      "grad_norm": 0.05093999579548836,
      "learning_rate": 0.002955718218996265,
      "loss": 1.7564,
      "step": 14580
    },
    {
      "epoch": 0.09612599088778279,
      "grad_norm": 0.05484277009963989,
      "learning_rate": 0.00295556537252441,
      "loss": 1.7491,
      "step": 14600
    },
    {
      "epoch": 0.09625767032735509,
      "grad_norm": 0.05289185419678688,
      "learning_rate": 0.002955412266684127,
      "loss": 1.7516,
      "step": 14620
    },
    {
      "epoch": 0.0963893497669274,
      "grad_norm": 0.05835564061999321,
      "learning_rate": 0.0029552589015026986,
      "loss": 1.7516,
      "step": 14640
    },
    {
      "epoch": 0.0965210292064997,
      "grad_norm": 0.05260492116212845,
      "learning_rate": 0.0029551052770074525,
      "loss": 1.7534,
      "step": 14660
    },
    {
      "epoch": 0.096652708646072,
      "grad_norm": 0.052535686641931534,
      "learning_rate": 0.002954951393225764,
      "loss": 1.7482,
      "step": 14680
    },
    {
      "epoch": 0.09678438808564431,
      "grad_norm": 0.062096912413835526,
      "learning_rate": 0.0029547972501850536,
      "loss": 1.7496,
      "step": 14700
    },
    {
      "epoch": 0.09691606752521661,
      "grad_norm": 0.07016194611787796,
      "learning_rate": 0.0029546428479127877,
      "loss": 1.7456,
      "step": 14720
    },
    {
      "epoch": 0.09704774696478892,
      "grad_norm": 0.06487017124891281,
      "learning_rate": 0.0029544881864364794,
      "loss": 1.7443,
      "step": 14740
    },
    {
      "epoch": 0.09717942640436122,
      "grad_norm": 0.060498468577861786,
      "learning_rate": 0.002954333265783688,
      "loss": 1.7516,
      "step": 14760
    },
    {
      "epoch": 0.09731110584393353,
      "grad_norm": 0.0497431606054306,
      "learning_rate": 0.0029541780859820197,
      "loss": 1.7635,
      "step": 14780
    },
    {
      "epoch": 0.09744278528350583,
      "grad_norm": 0.06207452341914177,
      "learning_rate": 0.002954022647059125,
      "loss": 1.7834,
      "step": 14800
    },
    {
      "epoch": 0.09757446472307814,
      "grad_norm": 0.06484688818454742,
      "learning_rate": 0.002953866949042703,
      "loss": 1.7779,
      "step": 14820
    },
    {
      "epoch": 0.09770614416265044,
      "grad_norm": 0.06074841320514679,
      "learning_rate": 0.002953710991960496,
      "loss": 1.7828,
      "step": 14840
    },
    {
      "epoch": 0.09783782360222275,
      "grad_norm": 0.05373964086174965,
      "learning_rate": 0.0029535547758402954,
      "loss": 1.7808,
      "step": 14860
    },
    {
      "epoch": 0.09796950304179505,
      "grad_norm": 0.06373169273138046,
      "learning_rate": 0.002953398300709937,
      "loss": 1.7843,
      "step": 14880
    },
    {
      "epoch": 0.09810118248136736,
      "grad_norm": 0.05869118869304657,
      "learning_rate": 0.0029532415665973035,
      "loss": 1.7826,
      "step": 14900
    },
    {
      "epoch": 0.09823286192093966,
      "grad_norm": 0.05736270919442177,
      "learning_rate": 0.002953084573530323,
      "loss": 1.774,
      "step": 14920
    },
    {
      "epoch": 0.09836454136051197,
      "grad_norm": 0.07032917439937592,
      "learning_rate": 0.0029529273215369704,
      "loss": 1.7702,
      "step": 14940
    },
    {
      "epoch": 0.09849622080008427,
      "grad_norm": 0.0598934143781662,
      "learning_rate": 0.002952769810645267,
      "loss": 1.7708,
      "step": 14960
    },
    {
      "epoch": 0.09862790023965658,
      "grad_norm": 0.06163979694247246,
      "learning_rate": 0.0029526120408832793,
      "loss": 1.7771,
      "step": 14980
    },
    {
      "epoch": 0.09875957967922888,
      "grad_norm": 0.055272020399570465,
      "learning_rate": 0.002952454012279121,
      "loss": 1.7675,
      "step": 15000
    },
    {
      "epoch": 0.09875957967922888,
      "eval_loss": 1.6859747171401978,
      "eval_runtime": 68.5865,
      "eval_samples_per_second": 14.58,
      "eval_steps_per_second": 14.58,
      "step": 15000
    },
    {
      "epoch": 0.09889125911880119,
      "grad_norm": 0.06086422875523567,
      "learning_rate": 0.0029522957248609505,
      "loss": 1.7656,
      "step": 15020
    },
    {
      "epoch": 0.09902293855837349,
      "grad_norm": 0.061683736741542816,
      "learning_rate": 0.0029521371786569743,
      "loss": 1.7571,
      "step": 15040
    },
    {
      "epoch": 0.0991546179979458,
      "grad_norm": 0.0584699846804142,
      "learning_rate": 0.0029519783736954427,
      "loss": 1.7654,
      "step": 15060
    },
    {
      "epoch": 0.0992862974375181,
      "grad_norm": 0.052420806139707565,
      "learning_rate": 0.002951819310004654,
      "loss": 1.764,
      "step": 15080
    },
    {
      "epoch": 0.0994179768770904,
      "grad_norm": 0.0528109036386013,
      "learning_rate": 0.0029516599876129516,
      "loss": 1.7611,
      "step": 15100
    },
    {
      "epoch": 0.09954965631666271,
      "grad_norm": 0.061885856091976166,
      "learning_rate": 0.0029515004065487254,
      "loss": 1.7628,
      "step": 15120
    },
    {
      "epoch": 0.09968133575623501,
      "grad_norm": 0.04771438241004944,
      "learning_rate": 0.0029513405668404116,
      "loss": 1.7652,
      "step": 15140
    },
    {
      "epoch": 0.09981301519580732,
      "grad_norm": 0.05954625830054283,
      "learning_rate": 0.0029511804685164915,
      "loss": 1.7554,
      "step": 15160
    },
    {
      "epoch": 0.09994469463537964,
      "grad_norm": 0.05653547868132591,
      "learning_rate": 0.002951020111605494,
      "loss": 1.7607,
      "step": 15180
    },
    {
      "epoch": 0.10007637407495194,
      "grad_norm": 0.05563562363386154,
      "learning_rate": 0.0029508594961359922,
      "loss": 1.759,
      "step": 15200
    },
    {
      "epoch": 0.10020805351452425,
      "grad_norm": 0.07335685193538666,
      "learning_rate": 0.002950698622136607,
      "loss": 1.7543,
      "step": 15220
    },
    {
      "epoch": 0.10033973295409655,
      "grad_norm": 0.06355411559343338,
      "learning_rate": 0.0029505374896360045,
      "loss": 1.7564,
      "step": 15240
    },
    {
      "epoch": 0.10047141239366886,
      "grad_norm": 0.05546756461262703,
      "learning_rate": 0.0029503760986628975,
      "loss": 1.7522,
      "step": 15260
    },
    {
      "epoch": 0.10060309183324116,
      "grad_norm": 0.058881357312202454,
      "learning_rate": 0.002950214449246043,
      "loss": 1.7544,
      "step": 15280
    },
    {
      "epoch": 0.10073477127281347,
      "grad_norm": 0.06903751194477081,
      "learning_rate": 0.0029500525414142475,
      "loss": 1.7531,
      "step": 15300
    },
    {
      "epoch": 0.10086645071238577,
      "grad_norm": 0.053250331431627274,
      "learning_rate": 0.00294989037519636,
      "loss": 1.753,
      "step": 15320
    },
    {
      "epoch": 0.10099813015195808,
      "grad_norm": 0.059820231050252914,
      "learning_rate": 0.002949727950621277,
      "loss": 1.7506,
      "step": 15340
    },
    {
      "epoch": 0.10112980959153038,
      "grad_norm": 0.06770730763673782,
      "learning_rate": 0.0029495652677179423,
      "loss": 1.76,
      "step": 15360
    },
    {
      "epoch": 0.10126148903110269,
      "grad_norm": 0.0625774934887886,
      "learning_rate": 0.0029494023265153432,
      "loss": 1.7919,
      "step": 15380
    },
    {
      "epoch": 0.10139316847067499,
      "grad_norm": 0.06113676354289055,
      "learning_rate": 0.002949239127042515,
      "loss": 1.814,
      "step": 15400
    },
    {
      "epoch": 0.1015248479102473,
      "grad_norm": 0.06288577616214752,
      "learning_rate": 0.002949075669328538,
      "loss": 1.8281,
      "step": 15420
    },
    {
      "epoch": 0.1016565273498196,
      "grad_norm": 0.05384008586406708,
      "learning_rate": 0.0029489119534025394,
      "loss": 1.8323,
      "step": 15440
    },
    {
      "epoch": 0.1017882067893919,
      "grad_norm": 0.06467036157846451,
      "learning_rate": 0.002948747979293692,
      "loss": 1.8343,
      "step": 15460
    },
    {
      "epoch": 0.10191988622896421,
      "grad_norm": 0.05992506816983223,
      "learning_rate": 0.0029485837470312128,
      "loss": 1.8315,
      "step": 15480
    },
    {
      "epoch": 0.10205156566853651,
      "grad_norm": 0.054823681712150574,
      "learning_rate": 0.002948419256644369,
      "loss": 1.8361,
      "step": 15500
    },
    {
      "epoch": 0.10218324510810882,
      "grad_norm": 0.07598422467708588,
      "learning_rate": 0.002948254508162469,
      "loss": 1.8395,
      "step": 15520
    },
    {
      "epoch": 0.10231492454768112,
      "grad_norm": 0.060086604207754135,
      "learning_rate": 0.002948089501614871,
      "loss": 1.8341,
      "step": 15540
    },
    {
      "epoch": 0.10244660398725343,
      "grad_norm": 0.05407283455133438,
      "learning_rate": 0.0029479242370309766,
      "loss": 1.8277,
      "step": 15560
    },
    {
      "epoch": 0.10257828342682573,
      "grad_norm": 0.057680752128362656,
      "learning_rate": 0.0029477587144402357,
      "loss": 1.8196,
      "step": 15580
    },
    {
      "epoch": 0.10270996286639804,
      "grad_norm": 0.0544477254152298,
      "learning_rate": 0.0029475929338721417,
      "loss": 1.8108,
      "step": 15600
    },
    {
      "epoch": 0.10284164230597034,
      "grad_norm": 0.054894376546144485,
      "learning_rate": 0.002947426895356236,
      "loss": 1.8139,
      "step": 15620
    },
    {
      "epoch": 0.10297332174554265,
      "grad_norm": 0.061300814151763916,
      "learning_rate": 0.0029472605989221043,
      "loss": 1.8116,
      "step": 15640
    },
    {
      "epoch": 0.10310500118511495,
      "grad_norm": 0.07517964392900467,
      "learning_rate": 0.00294709404459938,
      "loss": 1.806,
      "step": 15660
    },
    {
      "epoch": 0.10323668062468726,
      "grad_norm": 0.06433480978012085,
      "learning_rate": 0.0029469272324177415,
      "loss": 1.8092,
      "step": 15680
    },
    {
      "epoch": 0.10336836006425956,
      "grad_norm": 0.05521683767437935,
      "learning_rate": 0.0029467601624069122,
      "loss": 1.814,
      "step": 15700
    },
    {
      "epoch": 0.10350003950383187,
      "grad_norm": 0.0895058736205101,
      "learning_rate": 0.002946592834596663,
      "loss": 1.81,
      "step": 15720
    },
    {
      "epoch": 0.10363171894340417,
      "grad_norm": 0.0530746690928936,
      "learning_rate": 0.0029464252490168114,
      "loss": 1.8098,
      "step": 15740
    },
    {
      "epoch": 0.10376339838297648,
      "grad_norm": 0.05384887382388115,
      "learning_rate": 0.0029462574056972174,
      "loss": 1.8097,
      "step": 15760
    },
    {
      "epoch": 0.10389507782254878,
      "grad_norm": 0.05644648149609566,
      "learning_rate": 0.002946089304667791,
      "loss": 1.8032,
      "step": 15780
    },
    {
      "epoch": 0.10402675726212109,
      "grad_norm": 0.06337836384773254,
      "learning_rate": 0.002945920945958485,
      "loss": 1.7987,
      "step": 15800
    },
    {
      "epoch": 0.10415843670169339,
      "grad_norm": 0.05301804468035698,
      "learning_rate": 0.0029457523295993,
      "loss": 1.7936,
      "step": 15820
    },
    {
      "epoch": 0.10429011614126571,
      "grad_norm": 0.0584891214966774,
      "learning_rate": 0.002945583455620282,
      "loss": 1.7933,
      "step": 15840
    },
    {
      "epoch": 0.10442179558083801,
      "grad_norm": 0.058483339846134186,
      "learning_rate": 0.0029454143240515225,
      "loss": 1.7977,
      "step": 15860
    },
    {
      "epoch": 0.10455347502041032,
      "grad_norm": 0.057218778878450394,
      "learning_rate": 0.0029452449349231594,
      "loss": 1.7886,
      "step": 15880
    },
    {
      "epoch": 0.10468515445998262,
      "grad_norm": 0.0614253394305706,
      "learning_rate": 0.0029450752882653757,
      "loss": 1.7898,
      "step": 15900
    },
    {
      "epoch": 0.10481683389955493,
      "grad_norm": 0.05331398919224739,
      "learning_rate": 0.0029449053841084016,
      "loss": 1.7881,
      "step": 15920
    },
    {
      "epoch": 0.10494851333912723,
      "grad_norm": 0.05099445581436157,
      "learning_rate": 0.002944735222482512,
      "loss": 1.7937,
      "step": 15940
    },
    {
      "epoch": 0.10508019277869954,
      "grad_norm": 0.05356656014919281,
      "learning_rate": 0.0029445648034180294,
      "loss": 1.7943,
      "step": 15960
    },
    {
      "epoch": 0.10521187221827184,
      "grad_norm": 0.07416883111000061,
      "learning_rate": 0.0029443941269453188,
      "loss": 1.8186,
      "step": 15980
    },
    {
      "epoch": 0.10534355165784415,
      "grad_norm": 0.05611014366149902,
      "learning_rate": 0.002944223193094795,
      "loss": 1.8195,
      "step": 16000
    },
    {
      "epoch": 0.10534355165784415,
      "eval_loss": 1.753896713256836,
      "eval_runtime": 68.3324,
      "eval_samples_per_second": 14.634,
      "eval_steps_per_second": 14.634,
      "step": 16000
    },
    {
      "epoch": 0.00013167943957230517,
      "grad_norm": 0.06669767946004868,
      "learning_rate": 0.0029440520018969153,
      "loss": 1.9527,
      "step": 16020
    },
    {
      "epoch": 0.00026335887914461035,
      "grad_norm": 0.05156511813402176,
      "learning_rate": 0.0029438805533821863,
      "loss": 1.9455,
      "step": 16040
    },
    {
      "epoch": 0.00039503831871691555,
      "grad_norm": 0.05357939004898071,
      "learning_rate": 0.0029437088475811565,
      "loss": 1.9429,
      "step": 16060
    },
    {
      "epoch": 0.0005267177582892207,
      "grad_norm": 0.05948431417346001,
      "learning_rate": 0.002943536884524424,
      "loss": 1.9276,
      "step": 16080
    },
    {
      "epoch": 0.0006583971978615259,
      "grad_norm": 0.05270468071103096,
      "learning_rate": 0.00294336466424263,
      "loss": 1.925,
      "step": 16100
    },
    {
      "epoch": 0.0007900766374338311,
      "grad_norm": 0.05456703528761864,
      "learning_rate": 0.002943192186766463,
      "loss": 1.9325,
      "step": 16120
    },
    {
      "epoch": 0.0009217560770061363,
      "grad_norm": 0.056205980479717255,
      "learning_rate": 0.0029430194521266563,
      "loss": 1.923,
      "step": 16140
    },
    {
      "epoch": 0.0010534355165784414,
      "grad_norm": 0.052037931978702545,
      "learning_rate": 0.002942846460353991,
      "loss": 1.925,
      "step": 16160
    },
    {
      "epoch": 0.0011851149561507466,
      "grad_norm": 0.05397897586226463,
      "learning_rate": 0.0029426732114792914,
      "loss": 1.9119,
      "step": 16180
    },
    {
      "epoch": 0.0013167943957230518,
      "grad_norm": 0.06542910635471344,
      "learning_rate": 0.0029424997055334293,
      "loss": 1.9183,
      "step": 16200
    },
    {
      "epoch": 0.001448473835295357,
      "grad_norm": 0.07107926905155182,
      "learning_rate": 0.002942325942547322,
      "loss": 1.917,
      "step": 16220
    },
    {
      "epoch": 0.0015801532748676622,
      "grad_norm": 0.051881249994039536,
      "learning_rate": 0.0029421519225519316,
      "loss": 1.905,
      "step": 16240
    },
    {
      "epoch": 0.0017118327144399674,
      "grad_norm": 0.06571921706199646,
      "learning_rate": 0.0029419776455782685,
      "loss": 1.9101,
      "step": 16260
    },
    {
      "epoch": 0.0018435121540122726,
      "grad_norm": 0.08709180355072021,
      "learning_rate": 0.0029418031116573854,
      "loss": 1.9145,
      "step": 16280
    },
    {
      "epoch": 0.0019751915935845776,
      "grad_norm": 0.05184926837682724,
      "learning_rate": 0.0029416283208203845,
      "loss": 1.9066,
      "step": 16300
    },
    {
      "epoch": 0.002106871033156883,
      "grad_norm": 0.06952713429927826,
      "learning_rate": 0.0029414532730984103,
      "loss": 1.9042,
      "step": 16320
    },
    {
      "epoch": 0.002238550472729188,
      "grad_norm": 0.06200730428099632,
      "learning_rate": 0.002941277968522656,
      "loss": 1.9042,
      "step": 16340
    },
    {
      "epoch": 0.002370229912301493,
      "grad_norm": 0.06455907970666885,
      "learning_rate": 0.0029411024071243587,
      "loss": 1.8952,
      "step": 16360
    },
    {
      "epoch": 0.0025019093518737984,
      "grad_norm": 0.051165301352739334,
      "learning_rate": 0.0029409265889348015,
      "loss": 1.8987,
      "step": 16380
    },
    {
      "epoch": 0.0026335887914461036,
      "grad_norm": 0.0553780272603035,
      "learning_rate": 0.0029407505139853137,
      "loss": 1.9017,
      "step": 16400
    },
    {
      "epoch": 0.002765268231018409,
      "grad_norm": 0.06370244920253754,
      "learning_rate": 0.002940574182307271,
      "loss": 1.8952,
      "step": 16420
    },
    {
      "epoch": 0.002896947670590714,
      "grad_norm": 0.05009014904499054,
      "learning_rate": 0.0029403975939320936,
      "loss": 1.9018,
      "step": 16440
    },
    {
      "epoch": 0.003028627110163019,
      "grad_norm": 0.053404245525598526,
      "learning_rate": 0.0029402207488912475,
      "loss": 1.895,
      "step": 16460
    },
    {
      "epoch": 0.0031603065497353244,
      "grad_norm": 0.060919389128685,
      "learning_rate": 0.0029400436472162453,
      "loss": 1.8967,
      "step": 16480
    },
    {
      "epoch": 0.0032919859893076296,
      "grad_norm": 0.05715464800596237,
      "learning_rate": 0.002939866288938645,
      "loss": 1.9042,
      "step": 16500
    },
    {
      "epoch": 0.003423665428879935,
      "grad_norm": 0.07242151349782944,
      "learning_rate": 0.0029396886740900496,
      "loss": 1.889,
      "step": 16520
    },
    {
      "epoch": 0.00355534486845224,
      "grad_norm": 0.07230902463197708,
      "learning_rate": 0.0029395108027021094,
      "loss": 1.8924,
      "step": 16540
    },
    {
      "epoch": 0.0036870243080245452,
      "grad_norm": 0.05646587163209915,
      "learning_rate": 0.002939332674806519,
      "loss": 1.8873,
      "step": 16560
    },
    {
      "epoch": 0.0038187037475968504,
      "grad_norm": 0.06602409482002258,
      "learning_rate": 0.002939154290435019,
      "loss": 1.8926,
      "step": 16580
    },
    {
      "epoch": 0.003950383187169155,
      "grad_norm": 0.051282357424497604,
      "learning_rate": 0.002938975649619396,
      "loss": 1.9128,
      "step": 16600
    },
    {
      "epoch": 0.004082062626741461,
      "grad_norm": 0.04802711680531502,
      "learning_rate": 0.002938796752391482,
      "loss": 1.9163,
      "step": 16620
    },
    {
      "epoch": 0.004213742066313766,
      "grad_norm": 0.05548110231757164,
      "learning_rate": 0.002938617598783155,
      "loss": 1.9177,
      "step": 16640
    },
    {
      "epoch": 0.004345421505886071,
      "grad_norm": 0.06551088392734528,
      "learning_rate": 0.002938438188826339,
      "loss": 1.9084,
      "step": 16660
    },
    {
      "epoch": 0.004477100945458376,
      "grad_norm": 0.051025088876485825,
      "learning_rate": 0.002938258522553002,
      "loss": 1.9111,
      "step": 16680
    },
    {
      "epoch": 0.004608780385030682,
      "grad_norm": 0.06420248001813889,
      "learning_rate": 0.00293807859999516,
      "loss": 1.9021,
      "step": 16700
    },
    {
      "epoch": 0.004740459824602986,
      "grad_norm": 0.05957198888063431,
      "learning_rate": 0.0029378984211848734,
      "loss": 1.9114,
      "step": 16720
    },
    {
      "epoch": 0.004872139264175292,
      "grad_norm": 0.05393804982304573,
      "learning_rate": 0.002937717986154248,
      "loss": 1.9018,
      "step": 16740
    },
    {
      "epoch": 0.005003818703747597,
      "grad_norm": 0.052903495728969574,
      "learning_rate": 0.0029375372949354355,
      "loss": 1.9099,
      "step": 16760
    },
    {
      "epoch": 0.0051354981433199024,
      "grad_norm": 0.0786515399813652,
      "learning_rate": 0.0029373563475606343,
      "loss": 1.8988,
      "step": 16780
    },
    {
      "epoch": 0.005267177582892207,
      "grad_norm": 0.05307811498641968,
      "learning_rate": 0.002937175144062087,
      "loss": 1.901,
      "step": 16800
    },
    {
      "epoch": 0.005398857022464512,
      "grad_norm": 0.07106801867485046,
      "learning_rate": 0.0029369936844720825,
      "loss": 1.8994,
      "step": 16820
    },
    {
      "epoch": 0.005530536462036818,
      "grad_norm": 0.05467914044857025,
      "learning_rate": 0.0029368119688229547,
      "loss": 1.8971,
      "step": 16840
    },
    {
      "epoch": 0.005662215901609122,
      "grad_norm": 0.07325299084186554,
      "learning_rate": 0.0029366299971470846,
      "loss": 1.8895,
      "step": 16860
    },
    {
      "epoch": 0.005793895341181428,
      "grad_norm": 0.05642605945467949,
      "learning_rate": 0.002936447769476897,
      "loss": 1.8979,
      "step": 16880
    },
    {
      "epoch": 0.005925574780753733,
      "grad_norm": 0.05529412254691124,
      "learning_rate": 0.0029362652858448642,
      "loss": 1.8891,
      "step": 16900
    },
    {
      "epoch": 0.006057254220326038,
      "grad_norm": 0.06359060108661652,
      "learning_rate": 0.0029360825462835024,
      "loss": 1.8841,
      "step": 16920
    },
    {
      "epoch": 0.006188933659898343,
      "grad_norm": 0.06403638422489166,
      "learning_rate": 0.002935899550825374,
      "loss": 1.8846,
      "step": 16940
    },
    {
      "epoch": 0.006320613099470649,
      "grad_norm": 0.06129438802599907,
      "learning_rate": 0.0029357162995030882,
      "loss": 1.8845,
      "step": 16960
    },
    {
      "epoch": 0.006452292539042954,
      "grad_norm": 0.059139031916856766,
      "learning_rate": 0.002935532792349297,
      "loss": 1.8841,
      "step": 16980
    },
    {
      "epoch": 0.006583971978615259,
      "grad_norm": 0.056415166705846786,
      "learning_rate": 0.002935349029396701,
      "loss": 1.8768,
      "step": 17000
    },
    {
      "epoch": 0.006583971978615259,
      "eval_loss": 1.7990331649780273,
      "eval_runtime": 68.3369,
      "eval_samples_per_second": 14.633,
      "eval_steps_per_second": 14.633,
      "step": 17000
    },
    {
      "epoch": 0.006715651418187564,
      "grad_norm": 0.05324932560324669,
      "learning_rate": 0.002935165010678045,
      "loss": 1.877,
      "step": 17020
    },
    {
      "epoch": 0.00684733085775987,
      "grad_norm": 0.09261898696422577,
      "learning_rate": 0.0029349807362261185,
      "loss": 1.8822,
      "step": 17040
    },
    {
      "epoch": 0.006979010297332174,
      "grad_norm": 0.05834802985191345,
      "learning_rate": 0.0029347962060737583,
      "loss": 1.8869,
      "step": 17060
    },
    {
      "epoch": 0.00711068973690448,
      "grad_norm": 0.058308567851781845,
      "learning_rate": 0.0029346114202538458,
      "loss": 1.8808,
      "step": 17080
    },
    {
      "epoch": 0.007242369176476785,
      "grad_norm": 0.06321283429861069,
      "learning_rate": 0.0029344263787993087,
      "loss": 1.88,
      "step": 17100
    },
    {
      "epoch": 0.0073740486160490904,
      "grad_norm": 0.05077283829450607,
      "learning_rate": 0.0029342410817431185,
      "loss": 1.8774,
      "step": 17120
    },
    {
      "epoch": 0.007505728055621395,
      "grad_norm": 0.06091293692588806,
      "learning_rate": 0.0029340555291182944,
      "loss": 1.8779,
      "step": 17140
    },
    {
      "epoch": 0.007637407495193701,
      "grad_norm": 0.06410450488328934,
      "learning_rate": 0.0029338697209579,
      "loss": 1.8789,
      "step": 17160
    },
    {
      "epoch": 0.007769086934766006,
      "grad_norm": 0.08083081990480423,
      "learning_rate": 0.002933683657295044,
      "loss": 1.899,
      "step": 17180
    },
    {
      "epoch": 0.00790076637433831,
      "grad_norm": 0.06501522660255432,
      "learning_rate": 0.002933497338162883,
      "loss": 1.9053,
      "step": 17200
    },
    {
      "epoch": 0.008032445813910615,
      "grad_norm": 0.048091478645801544,
      "learning_rate": 0.0029333107635946152,
      "loss": 1.9089,
      "step": 17220
    },
    {
      "epoch": 0.008164125253482922,
      "grad_norm": 0.053573913872241974,
      "learning_rate": 0.0029331239336234873,
      "loss": 1.9014,
      "step": 17240
    },
    {
      "epoch": 0.008295804693055226,
      "grad_norm": 0.07217760384082794,
      "learning_rate": 0.0029329368482827905,
      "loss": 1.9048,
      "step": 17260
    },
    {
      "epoch": 0.008427484132627531,
      "grad_norm": 0.053729306906461716,
      "learning_rate": 0.0029327495076058624,
      "loss": 1.8983,
      "step": 17280
    },
    {
      "epoch": 0.008559163572199836,
      "grad_norm": 0.05567470192909241,
      "learning_rate": 0.002932561911626085,
      "loss": 1.8935,
      "step": 17300
    },
    {
      "epoch": 0.008690843011772142,
      "grad_norm": 0.06024245545268059,
      "learning_rate": 0.002932374060376886,
      "loss": 1.8916,
      "step": 17320
    },
    {
      "epoch": 0.008822522451344447,
      "grad_norm": 0.05578263849020004,
      "learning_rate": 0.0029321859538917394,
      "loss": 1.8964,
      "step": 17340
    },
    {
      "epoch": 0.008954201890916752,
      "grad_norm": 0.0536380261182785,
      "learning_rate": 0.0029319975922041633,
      "loss": 1.8874,
      "step": 17360
    },
    {
      "epoch": 0.009085881330489057,
      "grad_norm": 0.04501722753047943,
      "learning_rate": 0.002931808975347722,
      "loss": 1.8824,
      "step": 17380
    },
    {
      "epoch": 0.009217560770061363,
      "grad_norm": 0.06172627955675125,
      "learning_rate": 0.0029316201033560262,
      "loss": 1.8865,
      "step": 17400
    },
    {
      "epoch": 0.009349240209633668,
      "grad_norm": 0.05938473343849182,
      "learning_rate": 0.00293143097626273,
      "loss": 1.8845,
      "step": 17420
    },
    {
      "epoch": 0.009480919649205973,
      "grad_norm": 0.06107376888394356,
      "learning_rate": 0.0029312415941015347,
      "loss": 1.8892,
      "step": 17440
    },
    {
      "epoch": 0.009612599088778278,
      "grad_norm": 0.05630321800708771,
      "learning_rate": 0.0029310519569061867,
      "loss": 1.885,
      "step": 17460
    },
    {
      "epoch": 0.009744278528350584,
      "grad_norm": 0.05061402544379234,
      "learning_rate": 0.0029308620647104772,
      "loss": 1.8818,
      "step": 17480
    },
    {
      "epoch": 0.009875957967922889,
      "grad_norm": 0.059643711894750595,
      "learning_rate": 0.0029306719175482437,
      "loss": 1.877,
      "step": 17500
    },
    {
      "epoch": 0.010007637407495194,
      "grad_norm": 0.060052718967199326,
      "learning_rate": 0.0029304815154533683,
      "loss": 1.8752,
      "step": 17520
    },
    {
      "epoch": 0.010139316847067498,
      "grad_norm": 0.0526292622089386,
      "learning_rate": 0.002930290858459779,
      "loss": 1.8804,
      "step": 17540
    },
    {
      "epoch": 0.010270996286639805,
      "grad_norm": 0.08102289587259293,
      "learning_rate": 0.0029300999466014494,
      "loss": 1.8775,
      "step": 17560
    },
    {
      "epoch": 0.01040267572621211,
      "grad_norm": 0.07066734880208969,
      "learning_rate": 0.0029299087799123974,
      "loss": 1.8757,
      "step": 17580
    },
    {
      "epoch": 0.010534355165784414,
      "grad_norm": 0.08068786561489105,
      "learning_rate": 0.002929717358426688,
      "loss": 1.8798,
      "step": 17600
    },
    {
      "epoch": 0.01066603460535672,
      "grad_norm": 0.06165233626961708,
      "learning_rate": 0.0029295256821784306,
      "loss": 1.8741,
      "step": 17620
    },
    {
      "epoch": 0.010797714044929024,
      "grad_norm": 0.05444507673382759,
      "learning_rate": 0.0029293337512017797,
      "loss": 1.8754,
      "step": 17640
    },
    {
      "epoch": 0.01092939348450133,
      "grad_norm": 0.06841527670621872,
      "learning_rate": 0.002929141565530937,
      "loss": 1.8747,
      "step": 17660
    },
    {
      "epoch": 0.011061072924073635,
      "grad_norm": 0.06230770796537399,
      "learning_rate": 0.0029289491252001464,
      "loss": 1.8765,
      "step": 17680
    },
    {
      "epoch": 0.01119275236364594,
      "grad_norm": 0.07215554267168045,
      "learning_rate": 0.0029287564302437,
      "loss": 1.8755,
      "step": 17700
    },
    {
      "epoch": 0.011324431803218245,
      "grad_norm": 0.05525139719247818,
      "learning_rate": 0.002928563480695934,
      "loss": 1.8625,
      "step": 17720
    },
    {
      "epoch": 0.011456111242790551,
      "grad_norm": 0.08375394344329834,
      "learning_rate": 0.0029283702765912304,
      "loss": 1.8653,
      "step": 17740
    },
    {
      "epoch": 0.011587790682362856,
      "grad_norm": 0.07031811773777008,
      "learning_rate": 0.0029281768179640166,
      "loss": 1.8788,
      "step": 17760
    },
    {
      "epoch": 0.01171947012193516,
      "grad_norm": 0.07093445956707001,
      "learning_rate": 0.0029279831048487645,
      "loss": 1.8865,
      "step": 17780
    },
    {
      "epoch": 0.011851149561507466,
      "grad_norm": 0.07754283398389816,
      "learning_rate": 0.0029277891372799926,
      "loss": 1.8915,
      "step": 17800
    },
    {
      "epoch": 0.011982829001079772,
      "grad_norm": 0.06764809042215347,
      "learning_rate": 0.002927594915292264,
      "loss": 1.906,
      "step": 17820
    },
    {
      "epoch": 0.012114508440652077,
      "grad_norm": 0.05239294841885567,
      "learning_rate": 0.0029274004389201872,
      "loss": 1.8853,
      "step": 17840
    },
    {
      "epoch": 0.012246187880224382,
      "grad_norm": 0.060666222125291824,
      "learning_rate": 0.002927205708198416,
      "loss": 1.8856,
      "step": 17860
    },
    {
      "epoch": 0.012377867319796686,
      "grad_norm": 0.05473129451274872,
      "learning_rate": 0.00292701072316165,
      "loss": 1.8849,
      "step": 17880
    },
    {
      "epoch": 0.012509546759368993,
      "grad_norm": 0.054005883634090424,
      "learning_rate": 0.0029268154838446338,
      "loss": 1.8877,
      "step": 17900
    },
    {
      "epoch": 0.012641226198941298,
      "grad_norm": 0.07338313013315201,
      "learning_rate": 0.0029266199902821563,
      "loss": 1.8815,
      "step": 17920
    },
    {
      "epoch": 0.012772905638513602,
      "grad_norm": 0.06305193156003952,
      "learning_rate": 0.002926424242509054,
      "loss": 1.8775,
      "step": 17940
    },
    {
      "epoch": 0.012904585078085907,
      "grad_norm": 0.06329913437366486,
      "learning_rate": 0.0029262282405602066,
      "loss": 1.8759,
      "step": 17960
    },
    {
      "epoch": 0.013036264517658214,
      "grad_norm": 0.08256923407316208,
      "learning_rate": 0.0029260319844705398,
      "loss": 1.8779,
      "step": 17980
    },
    {
      "epoch": 0.013167943957230518,
      "grad_norm": 0.0697188749909401,
      "learning_rate": 0.002925835474275025,
      "loss": 1.8799,
      "step": 18000
    },
    {
      "epoch": 0.013167943957230518,
      "eval_loss": 1.8294225931167603,
      "eval_runtime": 65.8499,
      "eval_samples_per_second": 15.186,
      "eval_steps_per_second": 15.186,
      "step": 18000
    },
    {
      "epoch": 0.013299623396802823,
      "grad_norm": 0.05664549395442009,
      "learning_rate": 0.002925638710008678,
      "loss": 1.8665,
      "step": 18020
    },
    {
      "epoch": 0.013431302836375128,
      "grad_norm": 0.063201904296875,
      "learning_rate": 0.002925441691706561,
      "loss": 1.875,
      "step": 18040
    },
    {
      "epoch": 0.013562982275947433,
      "grad_norm": 0.05750538036227226,
      "learning_rate": 0.0029252444194037804,
      "loss": 1.86,
      "step": 18060
    },
    {
      "epoch": 0.01369466171551974,
      "grad_norm": 0.073476642370224,
      "learning_rate": 0.0029250468931354885,
      "loss": 1.8567,
      "step": 18080
    },
    {
      "epoch": 0.013826341155092044,
      "grad_norm": 0.05800202861428261,
      "learning_rate": 0.002924849112936883,
      "loss": 1.8669,
      "step": 18100
    },
    {
      "epoch": 0.013958020594664349,
      "grad_norm": 0.053366515785455704,
      "learning_rate": 0.002924651078843206,
      "loss": 1.8683,
      "step": 18120
    },
    {
      "epoch": 0.014089700034236654,
      "grad_norm": 0.06024957075715065,
      "learning_rate": 0.0029244527908897453,
      "loss": 1.8654,
      "step": 18140
    },
    {
      "epoch": 0.01422137947380896,
      "grad_norm": 0.06702018529176712,
      "learning_rate": 0.0029242542491118343,
      "loss": 1.8669,
      "step": 18160
    },
    {
      "epoch": 0.014353058913381265,
      "grad_norm": 0.050999585539102554,
      "learning_rate": 0.002924055453544852,
      "loss": 1.863,
      "step": 18180
    },
    {
      "epoch": 0.01448473835295357,
      "grad_norm": 0.05390724167227745,
      "learning_rate": 0.0029238564042242204,
      "loss": 1.8686,
      "step": 18200
    },
    {
      "epoch": 0.014616417792525874,
      "grad_norm": 0.08430622518062592,
      "learning_rate": 0.0029236571011854095,
      "loss": 1.8591,
      "step": 18220
    },
    {
      "epoch": 0.014748097232098181,
      "grad_norm": 0.07103120535612106,
      "learning_rate": 0.0029234575444639327,
      "loss": 1.8687,
      "step": 18240
    },
    {
      "epoch": 0.014879776671670486,
      "grad_norm": 0.0677858293056488,
      "learning_rate": 0.0029232577340953495,
      "loss": 1.8611,
      "step": 18260
    },
    {
      "epoch": 0.01501145611124279,
      "grad_norm": 0.07199349999427795,
      "learning_rate": 0.002923057670115264,
      "loss": 1.8613,
      "step": 18280
    },
    {
      "epoch": 0.015143135550815095,
      "grad_norm": 0.04802101105451584,
      "learning_rate": 0.002922857352559326,
      "loss": 1.8585,
      "step": 18300
    },
    {
      "epoch": 0.015274814990387402,
      "grad_norm": 0.06592894345521927,
      "learning_rate": 0.00292265678146323,
      "loss": 1.8627,
      "step": 18320
    },
    {
      "epoch": 0.015406494429959706,
      "grad_norm": 0.06502063572406769,
      "learning_rate": 0.0029224559568627162,
      "loss": 1.8542,
      "step": 18340
    },
    {
      "epoch": 0.015538173869532011,
      "grad_norm": 0.08361044526100159,
      "learning_rate": 0.0029222548787935696,
      "loss": 1.8558,
      "step": 18360
    },
    {
      "epoch": 0.015669853309104316,
      "grad_norm": 0.09134091436862946,
      "learning_rate": 0.002922053547291621,
      "loss": 1.8395,
      "step": 18380
    },
    {
      "epoch": 0.01580153274867662,
      "grad_norm": 0.06279317289590836,
      "learning_rate": 0.0029218519623927445,
      "loss": 1.7925,
      "step": 18400
    },
    {
      "epoch": 0.015933212188248926,
      "grad_norm": 0.05970417335629463,
      "learning_rate": 0.0029216501241328616,
      "loss": 1.7958,
      "step": 18420
    },
    {
      "epoch": 0.01606489162782123,
      "grad_norm": 0.05772954598069191,
      "learning_rate": 0.0029214480325479384,
      "loss": 1.7899,
      "step": 18440
    },
    {
      "epoch": 0.01619657106739354,
      "grad_norm": 0.05578155443072319,
      "learning_rate": 0.002921245687673985,
      "loss": 1.7772,
      "step": 18460
    },
    {
      "epoch": 0.016328250506965843,
      "grad_norm": 0.05890611186623573,
      "learning_rate": 0.0029210430895470576,
      "loss": 1.7791,
      "step": 18480
    },
    {
      "epoch": 0.016459929946538148,
      "grad_norm": 0.053837426006793976,
      "learning_rate": 0.0029208402382032575,
      "loss": 1.7787,
      "step": 18500
    },
    {
      "epoch": 0.016591609386110453,
      "grad_norm": 0.05714136362075806,
      "learning_rate": 0.002920637133678731,
      "loss": 1.7725,
      "step": 18520
    },
    {
      "epoch": 0.016723288825682758,
      "grad_norm": 0.06816162914037704,
      "learning_rate": 0.0029204337760096693,
      "loss": 1.7821,
      "step": 18540
    },
    {
      "epoch": 0.016854968265255062,
      "grad_norm": 0.05500810220837593,
      "learning_rate": 0.002920230165232309,
      "loss": 1.77,
      "step": 18560
    },
    {
      "epoch": 0.016986647704827367,
      "grad_norm": 0.05971973389387131,
      "learning_rate": 0.0029200263013829312,
      "loss": 1.7673,
      "step": 18580
    },
    {
      "epoch": 0.017118327144399672,
      "grad_norm": 0.05585896596312523,
      "learning_rate": 0.0029198221844978635,
      "loss": 1.7719,
      "step": 18600
    },
    {
      "epoch": 0.01725000658397198,
      "grad_norm": 0.05968007817864418,
      "learning_rate": 0.0029196178146134767,
      "loss": 1.7725,
      "step": 18620
    },
    {
      "epoch": 0.017381686023544285,
      "grad_norm": 0.06808626651763916,
      "learning_rate": 0.0029194131917661885,
      "loss": 1.7704,
      "step": 18640
    },
    {
      "epoch": 0.01751336546311659,
      "grad_norm": 0.057727713137865067,
      "learning_rate": 0.00291920831599246,
      "loss": 1.7752,
      "step": 18660
    },
    {
      "epoch": 0.017645044902688894,
      "grad_norm": 0.05363877862691879,
      "learning_rate": 0.0029190031873287986,
      "loss": 1.7682,
      "step": 18680
    },
    {
      "epoch": 0.0177767243422612,
      "grad_norm": 0.08140948414802551,
      "learning_rate": 0.0029187978058117566,
      "loss": 1.7621,
      "step": 18700
    },
    {
      "epoch": 0.017908403781833504,
      "grad_norm": 0.05746082961559296,
      "learning_rate": 0.0029185921714779306,
      "loss": 1.7743,
      "step": 18720
    },
    {
      "epoch": 0.01804008322140581,
      "grad_norm": 0.06933200359344482,
      "learning_rate": 0.0029183862843639636,
      "loss": 1.7783,
      "step": 18740
    },
    {
      "epoch": 0.018171762660978114,
      "grad_norm": 0.06859739869832993,
      "learning_rate": 0.0029181801445065415,
      "loss": 1.7723,
      "step": 18760
    },
    {
      "epoch": 0.018303442100550422,
      "grad_norm": 0.05922358110547066,
      "learning_rate": 0.002917973751942397,
      "loss": 1.7745,
      "step": 18780
    },
    {
      "epoch": 0.018435121540122727,
      "grad_norm": 0.0643932893872261,
      "learning_rate": 0.0029177671067083084,
      "loss": 1.7625,
      "step": 18800
    },
    {
      "epoch": 0.01856680097969503,
      "grad_norm": 0.060253337025642395,
      "learning_rate": 0.002917560208841097,
      "loss": 1.7739,
      "step": 18820
    },
    {
      "epoch": 0.018698480419267336,
      "grad_norm": 0.06834478676319122,
      "learning_rate": 0.00291735305837763,
      "loss": 1.7714,
      "step": 18840
    },
    {
      "epoch": 0.01883015985883964,
      "grad_norm": 0.06220484897494316,
      "learning_rate": 0.00291714565535482,
      "loss": 1.7794,
      "step": 18860
    },
    {
      "epoch": 0.018961839298411946,
      "grad_norm": 0.059682924300432205,
      "learning_rate": 0.0029169379998096245,
      "loss": 1.7751,
      "step": 18880
    },
    {
      "epoch": 0.01909351873798425,
      "grad_norm": 0.06003641337156296,
      "learning_rate": 0.002916730091779046,
      "loss": 1.7709,
      "step": 18900
    },
    {
      "epoch": 0.019225198177556555,
      "grad_norm": 0.05494799092411995,
      "learning_rate": 0.0029165219313001304,
      "loss": 1.7704,
      "step": 18920
    },
    {
      "epoch": 0.01935687761712886,
      "grad_norm": 0.05533396452665329,
      "learning_rate": 0.002916313518409972,
      "loss": 1.7788,
      "step": 18940
    },
    {
      "epoch": 0.019488557056701168,
      "grad_norm": 0.06435955315828323,
      "learning_rate": 0.0029161048531457065,
      "loss": 1.8023,
      "step": 18960
    },
    {
      "epoch": 0.019620236496273473,
      "grad_norm": 0.056616947054862976,
      "learning_rate": 0.0029158959355445177,
      "loss": 1.8187,
      "step": 18980
    },
    {
      "epoch": 0.019751915935845778,
      "grad_norm": 0.05237163230776787,
      "learning_rate": 0.002915686765643631,
      "loss": 1.8274,
      "step": 19000
    },
    {
      "epoch": 0.019751915935845778,
      "eval_loss": 1.8479809761047363,
      "eval_runtime": 65.9646,
      "eval_samples_per_second": 15.16,
      "eval_steps_per_second": 15.16,
      "step": 19000
    },
    {
      "epoch": 0.019883595375418082,
      "grad_norm": 0.06315074861049652,
      "learning_rate": 0.00291547734348032,
      "loss": 1.8249,
      "step": 19020
    },
    {
      "epoch": 0.020015274814990387,
      "grad_norm": 0.061322640627622604,
      "learning_rate": 0.002915267669091901,
      "loss": 1.8305,
      "step": 19040
    },
    {
      "epoch": 0.020146954254562692,
      "grad_norm": 0.06135581433773041,
      "learning_rate": 0.0029150577425157355,
      "loss": 1.8285,
      "step": 19060
    },
    {
      "epoch": 0.020278633694134997,
      "grad_norm": 0.06242895498871803,
      "learning_rate": 0.0029148475637892322,
      "loss": 1.8291,
      "step": 19080
    },
    {
      "epoch": 0.0204103131337073,
      "grad_norm": 0.07442411780357361,
      "learning_rate": 0.002914637132949842,
      "loss": 1.8167,
      "step": 19100
    },
    {
      "epoch": 0.02054199257327961,
      "grad_norm": 0.07312595844268799,
      "learning_rate": 0.0029144264500350616,
      "loss": 1.8139,
      "step": 19120
    },
    {
      "epoch": 0.020673672012851915,
      "grad_norm": 0.07727395743131638,
      "learning_rate": 0.0029142155150824333,
      "loss": 1.8121,
      "step": 19140
    },
    {
      "epoch": 0.02080535145242422,
      "grad_norm": 0.06504666805267334,
      "learning_rate": 0.0029140043281295435,
      "loss": 1.8162,
      "step": 19160
    },
    {
      "epoch": 0.020937030891996524,
      "grad_norm": 0.06606578081846237,
      "learning_rate": 0.0029137928892140237,
      "loss": 1.8141,
      "step": 19180
    },
    {
      "epoch": 0.02106871033156883,
      "grad_norm": 0.07460979372262955,
      "learning_rate": 0.0029135811983735504,
      "loss": 1.811,
      "step": 19200
    },
    {
      "epoch": 0.021200389771141134,
      "grad_norm": 0.05553029850125313,
      "learning_rate": 0.002913369255645845,
      "loss": 1.8122,
      "step": 19220
    },
    {
      "epoch": 0.02133206921071344,
      "grad_norm": 0.1024724543094635,
      "learning_rate": 0.002913157061068674,
      "loss": 1.8186,
      "step": 19240
    },
    {
      "epoch": 0.021463748650285743,
      "grad_norm": 0.07562136650085449,
      "learning_rate": 0.002912944614679848,
      "loss": 1.8238,
      "step": 19260
    },
    {
      "epoch": 0.021595428089858048,
      "grad_norm": 0.054396726191043854,
      "learning_rate": 0.002912731916517224,
      "loss": 1.8072,
      "step": 19280
    },
    {
      "epoch": 0.021727107529430356,
      "grad_norm": 0.07781907916069031,
      "learning_rate": 0.0029125189666187015,
      "loss": 1.8066,
      "step": 19300
    },
    {
      "epoch": 0.02185878696900266,
      "grad_norm": 0.05102219805121422,
      "learning_rate": 0.0029123057650222274,
      "loss": 1.803,
      "step": 19320
    },
    {
      "epoch": 0.021990466408574966,
      "grad_norm": 0.05836215242743492,
      "learning_rate": 0.002912092311765792,
      "loss": 1.8013,
      "step": 19340
    },
    {
      "epoch": 0.02212214584814727,
      "grad_norm": 0.0890396237373352,
      "learning_rate": 0.0029118786068874298,
      "loss": 1.804,
      "step": 19360
    },
    {
      "epoch": 0.022253825287719575,
      "grad_norm": 0.06296063959598541,
      "learning_rate": 0.002911664650425222,
      "loss": 1.7988,
      "step": 19380
    },
    {
      "epoch": 0.02238550472729188,
      "grad_norm": 0.07317844778299332,
      "learning_rate": 0.0029114504424172937,
      "loss": 1.7993,
      "step": 19400
    },
    {
      "epoch": 0.022517184166864185,
      "grad_norm": 0.061669182032346725,
      "learning_rate": 0.002911235982901814,
      "loss": 1.7971,
      "step": 19420
    },
    {
      "epoch": 0.02264886360643649,
      "grad_norm": 0.07136490941047668,
      "learning_rate": 0.002911021271916998,
      "loss": 1.8064,
      "step": 19440
    },
    {
      "epoch": 0.022780543046008798,
      "grad_norm": 0.05423993617296219,
      "learning_rate": 0.0029108063095011063,
      "loss": 1.8017,
      "step": 19460
    },
    {
      "epoch": 0.022912222485581103,
      "grad_norm": 0.0703122541308403,
      "learning_rate": 0.002910591095692442,
      "loss": 1.7979,
      "step": 19480
    },
    {
      "epoch": 0.023043901925153407,
      "grad_norm": 0.06629758328199387,
      "learning_rate": 0.0029103756305293546,
      "loss": 1.7959,
      "step": 19500
    },
    {
      "epoch": 0.023175581364725712,
      "grad_norm": 0.06768941134214401,
      "learning_rate": 0.002910159914050238,
      "loss": 1.796,
      "step": 19520
    },
    {
      "epoch": 0.023307260804298017,
      "grad_norm": 0.060242656618356705,
      "learning_rate": 0.002909943946293531,
      "loss": 1.802,
      "step": 19540
    },
    {
      "epoch": 0.02343894024387032,
      "grad_norm": 0.06179652363061905,
      "learning_rate": 0.0029097277272977164,
      "loss": 1.8097,
      "step": 19560
    },
    {
      "epoch": 0.023570619683442626,
      "grad_norm": 0.10733772069215775,
      "learning_rate": 0.0029095112571013238,
      "loss": 1.8088,
      "step": 19580
    },
    {
      "epoch": 0.02370229912301493,
      "grad_norm": 0.07508611679077148,
      "learning_rate": 0.002909294535742925,
      "loss": 1.8135,
      "step": 19600
    },
    {
      "epoch": 0.023833978562587236,
      "grad_norm": 0.057078372687101364,
      "learning_rate": 0.0029090775632611377,
      "loss": 1.811,
      "step": 19620
    },
    {
      "epoch": 0.023965658002159544,
      "grad_norm": 0.06910204887390137,
      "learning_rate": 0.002908860339694626,
      "loss": 1.8024,
      "step": 19640
    },
    {
      "epoch": 0.02409733744173185,
      "grad_norm": 0.05256013572216034,
      "learning_rate": 0.0029086428650820953,
      "loss": 1.8001,
      "step": 19660
    },
    {
      "epoch": 0.024229016881304154,
      "grad_norm": 0.05280625820159912,
      "learning_rate": 0.0029084251394622986,
      "loss": 1.7961,
      "step": 19680
    },
    {
      "epoch": 0.02436069632087646,
      "grad_norm": 0.07600513845682144,
      "learning_rate": 0.0029082071628740323,
      "loss": 1.7916,
      "step": 19700
    },
    {
      "epoch": 0.024492375760448763,
      "grad_norm": 0.060011059045791626,
      "learning_rate": 0.002907988935356138,
      "loss": 1.7981,
      "step": 19720
    },
    {
      "epoch": 0.024624055200021068,
      "grad_norm": 0.06662587076425552,
      "learning_rate": 0.002907770456947501,
      "loss": 1.7979,
      "step": 19740
    },
    {
      "epoch": 0.024755734639593373,
      "grad_norm": 0.06288140267133713,
      "learning_rate": 0.002907551727687054,
      "loss": 1.7918,
      "step": 19760
    },
    {
      "epoch": 0.024887414079165678,
      "grad_norm": 0.07184966653585434,
      "learning_rate": 0.002907332747613771,
      "loss": 1.7932,
      "step": 19780
    },
    {
      "epoch": 0.025019093518737986,
      "grad_norm": 0.062474627047777176,
      "learning_rate": 0.0029071135167666722,
      "loss": 1.7897,
      "step": 19800
    },
    {
      "epoch": 0.02515077295831029,
      "grad_norm": 0.055238522589206696,
      "learning_rate": 0.0029068940351848235,
      "loss": 1.7866,
      "step": 19820
    },
    {
      "epoch": 0.025282452397882595,
      "grad_norm": 0.060174524784088135,
      "learning_rate": 0.002906674302907334,
      "loss": 1.7854,
      "step": 19840
    },
    {
      "epoch": 0.0254141318374549,
      "grad_norm": 0.05686888098716736,
      "learning_rate": 0.002906454319973358,
      "loss": 1.7794,
      "step": 19860
    },
    {
      "epoch": 0.025545811277027205,
      "grad_norm": 0.05989696830511093,
      "learning_rate": 0.002906234086422094,
      "loss": 1.7802,
      "step": 19880
    },
    {
      "epoch": 0.02567749071659951,
      "grad_norm": 0.05739467218518257,
      "learning_rate": 0.0029060136022927867,
      "loss": 1.7859,
      "step": 19900
    },
    {
      "epoch": 0.025809170156171814,
      "grad_norm": 0.06893962621688843,
      "learning_rate": 0.0029057928676247233,
      "loss": 1.7903,
      "step": 19920
    },
    {
      "epoch": 0.02594084959574412,
      "grad_norm": 0.05944626033306122,
      "learning_rate": 0.002905571882457237,
      "loss": 1.7905,
      "step": 19940
    },
    {
      "epoch": 0.026072529035316427,
      "grad_norm": 0.10011964291334152,
      "learning_rate": 0.002905350646829706,
      "loss": 1.7824,
      "step": 19960
    },
    {
      "epoch": 0.026204208474888732,
      "grad_norm": 0.055610090494155884,
      "learning_rate": 0.0029051291607815515,
      "loss": 1.7803,
      "step": 19980
    },
    {
      "epoch": 0.026335887914461037,
      "grad_norm": 0.061704959720373154,
      "learning_rate": 0.0029049074243522404,
      "loss": 1.7793,
      "step": 20000
    },
    {
      "epoch": 0.026335887914461037,
      "eval_loss": 1.8498175144195557,
      "eval_runtime": 65.8184,
      "eval_samples_per_second": 15.193,
      "eval_steps_per_second": 15.193,
      "step": 20000
    },
    {
      "epoch": 0.02646756735403334,
      "grad_norm": 0.06252337247133255,
      "learning_rate": 0.002904685437581285,
      "loss": 1.7794,
      "step": 20020
    },
    {
      "epoch": 0.026599246793605646,
      "grad_norm": 0.05822984129190445,
      "learning_rate": 0.00290446320050824,
      "loss": 1.7819,
      "step": 20040
    },
    {
      "epoch": 0.02673092623317795,
      "grad_norm": 0.05482339486479759,
      "learning_rate": 0.0029042407131727073,
      "loss": 1.7867,
      "step": 20060
    },
    {
      "epoch": 0.026862605672750256,
      "grad_norm": 0.05103064700961113,
      "learning_rate": 0.0029040179756143316,
      "loss": 1.7764,
      "step": 20080
    },
    {
      "epoch": 0.02699428511232256,
      "grad_norm": 0.06983567029237747,
      "learning_rate": 0.002903794987872802,
      "loss": 1.779,
      "step": 20100
    },
    {
      "epoch": 0.027125964551894866,
      "grad_norm": 0.06782662123441696,
      "learning_rate": 0.0029035717499878537,
      "loss": 1.7814,
      "step": 20120
    },
    {
      "epoch": 0.027257643991467174,
      "grad_norm": 0.07617141306400299,
      "learning_rate": 0.0029033482619992656,
      "loss": 1.8055,
      "step": 20140
    },
    {
      "epoch": 0.02738932343103948,
      "grad_norm": 0.07056764513254166,
      "learning_rate": 0.002903124523946861,
      "loss": 1.8267,
      "step": 20160
    },
    {
      "epoch": 0.027521002870611783,
      "grad_norm": 0.05244629457592964,
      "learning_rate": 0.0029029005358705085,
      "loss": 1.8346,
      "step": 20180
    },
    {
      "epoch": 0.027652682310184088,
      "grad_norm": 0.05634482949972153,
      "learning_rate": 0.0029026762978101197,
      "loss": 1.8347,
      "step": 20200
    },
    {
      "epoch": 0.027784361749756393,
      "grad_norm": 0.060985077172517776,
      "learning_rate": 0.0029024518098056526,
      "loss": 1.8397,
      "step": 20220
    },
    {
      "epoch": 0.027916041189328698,
      "grad_norm": 0.060473646968603134,
      "learning_rate": 0.0029022270718971083,
      "loss": 1.8363,
      "step": 20240
    },
    {
      "epoch": 0.028047720628901002,
      "grad_norm": 0.05223367363214493,
      "learning_rate": 0.002902002084124533,
      "loss": 1.8378,
      "step": 20260
    },
    {
      "epoch": 0.028179400068473307,
      "grad_norm": 0.06747066229581833,
      "learning_rate": 0.0029017768465280187,
      "loss": 1.8394,
      "step": 20280
    },
    {
      "epoch": 0.028311079508045615,
      "grad_norm": 0.06481572240591049,
      "learning_rate": 0.0029015513591476995,
      "loss": 1.829,
      "step": 20300
    },
    {
      "epoch": 0.02844275894761792,
      "grad_norm": 0.07189793139696121,
      "learning_rate": 0.002901325622023755,
      "loss": 1.8283,
      "step": 20320
    },
    {
      "epoch": 0.028574438387190225,
      "grad_norm": 0.0644344687461853,
      "learning_rate": 0.0029010996351964105,
      "loss": 1.8265,
      "step": 20340
    },
    {
      "epoch": 0.02870611782676253,
      "grad_norm": 0.06990751624107361,
      "learning_rate": 0.002900873398705934,
      "loss": 1.824,
      "step": 20360
    },
    {
      "epoch": 0.028837797266334834,
      "grad_norm": 0.07374297082424164,
      "learning_rate": 0.0029006469125926397,
      "loss": 1.8223,
      "step": 20380
    },
    {
      "epoch": 0.02896947670590714,
      "grad_norm": 0.08984444290399551,
      "learning_rate": 0.0029004201768968838,
      "loss": 1.8232,
      "step": 20400
    },
    {
      "epoch": 0.029101156145479444,
      "grad_norm": 0.06562740355730057,
      "learning_rate": 0.00290019319165907,
      "loss": 1.8235,
      "step": 20420
    },
    {
      "epoch": 0.02923283558505175,
      "grad_norm": 0.07431894540786743,
      "learning_rate": 0.0028999659569196442,
      "loss": 1.8141,
      "step": 20440
    },
    {
      "epoch": 0.029364515024624054,
      "grad_norm": 0.057788606733083725,
      "learning_rate": 0.0028997384727190976,
      "loss": 1.8159,
      "step": 20460
    },
    {
      "epoch": 0.029496194464196362,
      "grad_norm": 0.07726121693849564,
      "learning_rate": 0.002899510739097966,
      "loss": 1.8156,
      "step": 20480
    },
    {
      "epoch": 0.029627873903768667,
      "grad_norm": 0.07005012780427933,
      "learning_rate": 0.0028992827560968297,
      "loss": 1.8172,
      "step": 20500
    },
    {
      "epoch": 0.02975955334334097,
      "grad_norm": 0.06550168991088867,
      "learning_rate": 0.002899054523756313,
      "loss": 1.8067,
      "step": 20520
    },
    {
      "epoch": 0.029891232782913276,
      "grad_norm": 0.07354078441858292,
      "learning_rate": 0.002898826042117084,
      "loss": 1.8109,
      "step": 20540
    },
    {
      "epoch": 0.03002291222248558,
      "grad_norm": 0.06678362935781479,
      "learning_rate": 0.0028985973112198573,
      "loss": 1.8158,
      "step": 20560
    },
    {
      "epoch": 0.030154591662057886,
      "grad_norm": 0.08662049472332001,
      "learning_rate": 0.0028983683311053898,
      "loss": 1.8077,
      "step": 20580
    },
    {
      "epoch": 0.03028627110163019,
      "grad_norm": 0.06835264712572098,
      "learning_rate": 0.0028981391018144845,
      "loss": 1.8074,
      "step": 20600
    },
    {
      "epoch": 0.030417950541202495,
      "grad_norm": 0.06030866131186485,
      "learning_rate": 0.0028979096233879872,
      "loss": 1.8096,
      "step": 20620
    },
    {
      "epoch": 0.030549629980774803,
      "grad_norm": 0.07386446744203568,
      "learning_rate": 0.0028976798958667885,
      "loss": 1.8134,
      "step": 20640
    },
    {
      "epoch": 0.030681309420347108,
      "grad_norm": 0.07323277741670609,
      "learning_rate": 0.002897449919291825,
      "loss": 1.8073,
      "step": 20660
    },
    {
      "epoch": 0.030812988859919413,
      "grad_norm": 0.08217355608940125,
      "learning_rate": 0.0028972196937040754,
      "loss": 1.8044,
      "step": 20680
    },
    {
      "epoch": 0.030944668299491718,
      "grad_norm": 0.06611933559179306,
      "learning_rate": 0.002896989219144564,
      "loss": 1.8016,
      "step": 20700
    },
    {
      "epoch": 0.031076347739064022,
      "grad_norm": 0.06082133203744888,
      "learning_rate": 0.0028967584956543597,
      "loss": 1.7986,
      "step": 20720
    },
    {
      "epoch": 0.031208027178636327,
      "grad_norm": 0.06548244506120682,
      "learning_rate": 0.002896527523274575,
      "loss": 1.8025,
      "step": 20740
    },
    {
      "epoch": 0.03133970661820863,
      "grad_norm": 0.08356646448373795,
      "learning_rate": 0.0028962963020463667,
      "loss": 1.783,
      "step": 20760
    },
    {
      "epoch": 0.03147138605778094,
      "grad_norm": 0.06322002410888672,
      "learning_rate": 0.002896064832010937,
      "loss": 1.774,
      "step": 20780
    },
    {
      "epoch": 0.03160306549735324,
      "grad_norm": 0.0772123858332634,
      "learning_rate": 0.002895833113209531,
      "loss": 1.7628,
      "step": 20800
    },
    {
      "epoch": 0.031734744936925546,
      "grad_norm": 0.0836576297879219,
      "learning_rate": 0.0028956011456834395,
      "loss": 1.7589,
      "step": 20820
    },
    {
      "epoch": 0.03186642437649785,
      "grad_norm": 0.06800664216279984,
      "learning_rate": 0.002895368929473997,
      "loss": 1.7455,
      "step": 20840
    },
    {
      "epoch": 0.031998103816070156,
      "grad_norm": 0.058029696345329285,
      "learning_rate": 0.0028951364646225826,
      "loss": 1.745,
      "step": 20860
    },
    {
      "epoch": 0.03212978325564246,
      "grad_norm": 0.05537112057209015,
      "learning_rate": 0.0028949037511706178,
      "loss": 1.748,
      "step": 20880
    },
    {
      "epoch": 0.03226146269521477,
      "grad_norm": 0.06035890802741051,
      "learning_rate": 0.0028946707891595717,
      "loss": 1.7463,
      "step": 20900
    },
    {
      "epoch": 0.03239314213478708,
      "grad_norm": 0.06673149764537811,
      "learning_rate": 0.002894437578630955,
      "loss": 1.7429,
      "step": 20920
    },
    {
      "epoch": 0.03252482157435938,
      "grad_norm": 0.08091327548027039,
      "learning_rate": 0.0028942041196263245,
      "loss": 1.7388,
      "step": 20940
    },
    {
      "epoch": 0.03265650101393169,
      "grad_norm": 0.05750538781285286,
      "learning_rate": 0.00289397041218728,
      "loss": 1.7322,
      "step": 20960
    },
    {
      "epoch": 0.03278818045350399,
      "grad_norm": 0.059635717421770096,
      "learning_rate": 0.0028937364563554656,
      "loss": 1.7425,
      "step": 20980
    },
    {
      "epoch": 0.032919859893076296,
      "grad_norm": 0.07420338690280914,
      "learning_rate": 0.002893502252172571,
      "loss": 1.7369,
      "step": 21000
    },
    {
      "epoch": 0.032919859893076296,
      "eval_loss": 1.7132776975631714,
      "eval_runtime": 66.0009,
      "eval_samples_per_second": 15.151,
      "eval_steps_per_second": 15.151,
      "step": 21000
    },
    {
      "epoch": 0.0330515393326486,
      "grad_norm": 0.06639964133501053,
      "learning_rate": 0.0028932677996803286,
      "loss": 1.7301,
      "step": 21020
    },
    {
      "epoch": 0.033183218772220906,
      "grad_norm": 0.06292149424552917,
      "learning_rate": 0.0028930330989205156,
      "loss": 1.7371,
      "step": 21040
    },
    {
      "epoch": 0.03331489821179321,
      "grad_norm": 0.07750008255243301,
      "learning_rate": 0.002892798149934954,
      "loss": 1.7395,
      "step": 21060
    },
    {
      "epoch": 0.033446577651365515,
      "grad_norm": 0.056288789957761765,
      "learning_rate": 0.0028925629527655097,
      "loss": 1.7346,
      "step": 21080
    },
    {
      "epoch": 0.03357825709093782,
      "grad_norm": 0.07288241386413574,
      "learning_rate": 0.002892327507454092,
      "loss": 1.7306,
      "step": 21100
    },
    {
      "epoch": 0.033709936530510125,
      "grad_norm": 0.06977301836013794,
      "learning_rate": 0.002892091814042655,
      "loss": 1.7293,
      "step": 21120
    },
    {
      "epoch": 0.03384161597008243,
      "grad_norm": 0.07677018642425537,
      "learning_rate": 0.0028918558725731986,
      "loss": 1.7334,
      "step": 21140
    },
    {
      "epoch": 0.033973295409654734,
      "grad_norm": 0.06132907420396805,
      "learning_rate": 0.0028916196830877637,
      "loss": 1.7352,
      "step": 21160
    },
    {
      "epoch": 0.03410497484922704,
      "grad_norm": 0.08819183707237244,
      "learning_rate": 0.0028913832456284377,
      "loss": 1.7383,
      "step": 21180
    },
    {
      "epoch": 0.034236654288799344,
      "grad_norm": 0.0902981087565422,
      "learning_rate": 0.0028911465602373524,
      "loss": 1.7296,
      "step": 21200
    },
    {
      "epoch": 0.03436833372837165,
      "grad_norm": 0.07254979014396667,
      "learning_rate": 0.0028909096269566815,
      "loss": 1.7314,
      "step": 21220
    },
    {
      "epoch": 0.03450001316794396,
      "grad_norm": 0.08108898252248764,
      "learning_rate": 0.0028906724458286456,
      "loss": 1.7365,
      "step": 21240
    },
    {
      "epoch": 0.034631692607516265,
      "grad_norm": 0.061406515538692474,
      "learning_rate": 0.002890435016895507,
      "loss": 1.7338,
      "step": 21260
    },
    {
      "epoch": 0.03476337204708857,
      "grad_norm": 0.07603522390127182,
      "learning_rate": 0.0028901973401995744,
      "loss": 1.727,
      "step": 21280
    },
    {
      "epoch": 0.034895051486660875,
      "grad_norm": 0.08022353053092957,
      "learning_rate": 0.002889959415783199,
      "loss": 1.7258,
      "step": 21300
    },
    {
      "epoch": 0.03502673092623318,
      "grad_norm": 0.061437543481588364,
      "learning_rate": 0.0028897212436887775,
      "loss": 1.7416,
      "step": 21320
    },
    {
      "epoch": 0.035158410365805484,
      "grad_norm": 0.07316586375236511,
      "learning_rate": 0.0028894828239587494,
      "loss": 1.7415,
      "step": 21340
    },
    {
      "epoch": 0.03529008980537779,
      "grad_norm": 0.09495033323764801,
      "learning_rate": 0.002889244156635599,
      "loss": 1.7403,
      "step": 21360
    },
    {
      "epoch": 0.035421769244950094,
      "grad_norm": 0.06723074615001678,
      "learning_rate": 0.002889005241761854,
      "loss": 1.7497,
      "step": 21380
    },
    {
      "epoch": 0.0355534486845224,
      "grad_norm": 0.05688747763633728,
      "learning_rate": 0.002888766079380088,
      "loss": 1.7408,
      "step": 21400
    },
    {
      "epoch": 0.0356851281240947,
      "grad_norm": 0.06509803235530853,
      "learning_rate": 0.002888526669532917,
      "loss": 1.7461,
      "step": 21420
    },
    {
      "epoch": 0.03581680756366701,
      "grad_norm": 0.07571250200271606,
      "learning_rate": 0.002888287012263002,
      "loss": 1.7385,
      "step": 21440
    },
    {
      "epoch": 0.03594848700323931,
      "grad_norm": 0.0633770301938057,
      "learning_rate": 0.002888047107613047,
      "loss": 1.743,
      "step": 21460
    },
    {
      "epoch": 0.03608016644281162,
      "grad_norm": 0.07172537595033646,
      "learning_rate": 0.0028878069556258013,
      "loss": 1.7289,
      "step": 21480
    },
    {
      "epoch": 0.03621184588238392,
      "grad_norm": 0.09092337638139725,
      "learning_rate": 0.002887566556344058,
      "loss": 1.7392,
      "step": 21500
    },
    {
      "epoch": 0.03634352532195623,
      "grad_norm": 0.06401342898607254,
      "learning_rate": 0.002887325909810653,
      "loss": 1.7372,
      "step": 21520
    },
    {
      "epoch": 0.03647520476152853,
      "grad_norm": 0.06551776826381683,
      "learning_rate": 0.0028870850160684688,
      "loss": 1.7221,
      "step": 21540
    },
    {
      "epoch": 0.036606884201100844,
      "grad_norm": 0.08069788664579391,
      "learning_rate": 0.0028868438751604294,
      "loss": 1.7201,
      "step": 21560
    },
    {
      "epoch": 0.03673856364067315,
      "grad_norm": 0.05819667875766754,
      "learning_rate": 0.002886602487129504,
      "loss": 1.7256,
      "step": 21580
    },
    {
      "epoch": 0.03687024308024545,
      "grad_norm": 0.06299937516450882,
      "learning_rate": 0.0028863608520187066,
      "loss": 1.7216,
      "step": 21600
    },
    {
      "epoch": 0.03700192251981776,
      "grad_norm": 0.08698736876249313,
      "learning_rate": 0.002886118969871093,
      "loss": 1.7208,
      "step": 21620
    },
    {
      "epoch": 0.03713360195939006,
      "grad_norm": 0.07046283036470413,
      "learning_rate": 0.0028858768407297656,
      "loss": 1.7141,
      "step": 21640
    },
    {
      "epoch": 0.03726528139896237,
      "grad_norm": 0.06260740756988525,
      "learning_rate": 0.0028856344646378687,
      "loss": 1.7243,
      "step": 21660
    },
    {
      "epoch": 0.03739696083853467,
      "grad_norm": 0.0720597505569458,
      "learning_rate": 0.0028853918416385928,
      "loss": 1.7195,
      "step": 21680
    },
    {
      "epoch": 0.03752864027810698,
      "grad_norm": 0.07073231786489487,
      "learning_rate": 0.0028851489717751696,
      "loss": 1.7241,
      "step": 21700
    },
    {
      "epoch": 0.03766031971767928,
      "grad_norm": 0.05944892391562462,
      "learning_rate": 0.0028849058550908767,
      "loss": 1.7201,
      "step": 21720
    },
    {
      "epoch": 0.037791999157251586,
      "grad_norm": 0.06408489495515823,
      "learning_rate": 0.0028846624916290357,
      "loss": 1.7228,
      "step": 21740
    },
    {
      "epoch": 0.03792367859682389,
      "grad_norm": 0.07245621085166931,
      "learning_rate": 0.002884418881433012,
      "loss": 1.7191,
      "step": 21760
    },
    {
      "epoch": 0.038055358036396196,
      "grad_norm": 0.06913050264120102,
      "learning_rate": 0.0028841750245462137,
      "loss": 1.7228,
      "step": 21780
    },
    {
      "epoch": 0.0381870374759685,
      "grad_norm": 0.057451095432043076,
      "learning_rate": 0.002883930921012094,
      "loss": 1.7163,
      "step": 21800
    },
    {
      "epoch": 0.038318716915540806,
      "grad_norm": 0.07231597602367401,
      "learning_rate": 0.002883686570874151,
      "loss": 1.7143,
      "step": 21820
    },
    {
      "epoch": 0.03845039635511311,
      "grad_norm": 0.08483259379863739,
      "learning_rate": 0.0028834419741759244,
      "loss": 1.7216,
      "step": 21840
    },
    {
      "epoch": 0.038582075794685415,
      "grad_norm": 0.0528041310608387,
      "learning_rate": 0.0028831971309610004,
      "loss": 1.7174,
      "step": 21860
    },
    {
      "epoch": 0.03871375523425772,
      "grad_norm": 0.06268829107284546,
      "learning_rate": 0.0028829520412730065,
      "loss": 1.721,
      "step": 21880
    },
    {
      "epoch": 0.03884543467383003,
      "grad_norm": 0.08661678433418274,
      "learning_rate": 0.0028827067051556163,
      "loss": 1.7196,
      "step": 21900
    },
    {
      "epoch": 0.038977114113402336,
      "grad_norm": 0.062315549701452255,
      "learning_rate": 0.0028824611226525455,
      "loss": 1.7481,
      "step": 21920
    },
    {
      "epoch": 0.03910879355297464,
      "grad_norm": 0.0838395208120346,
      "learning_rate": 0.002882215293807556,
      "loss": 1.7647,
      "step": 21940
    },
    {
      "epoch": 0.039240472992546946,
      "grad_norm": 0.06503219902515411,
      "learning_rate": 0.0028819692186644514,
      "loss": 1.7775,
      "step": 21960
    },
    {
      "epoch": 0.03937215243211925,
      "grad_norm": 0.06856662034988403,
      "learning_rate": 0.0028817228972670803,
      "loss": 1.7717,
      "step": 21980
    },
    {
      "epoch": 0.039503831871691555,
      "grad_norm": 0.055163074284791946,
      "learning_rate": 0.002881476329659335,
      "loss": 1.7772,
      "step": 22000
    },
    {
      "epoch": 0.039503831871691555,
      "eval_loss": 1.7622946500778198,
      "eval_runtime": 67.7091,
      "eval_samples_per_second": 14.769,
      "eval_steps_per_second": 14.769,
      "step": 22000
    },
    {
      "epoch": 0.03963551131126386,
      "grad_norm": 0.09752371162176132,
      "learning_rate": 0.002881229515885151,
      "loss": 1.7653,
      "step": 22020
    },
    {
      "epoch": 0.039767190750836165,
      "grad_norm": 0.09371519088745117,
      "learning_rate": 0.0028809824559885085,
      "loss": 1.7757,
      "step": 22040
    },
    {
      "epoch": 0.03989887019040847,
      "grad_norm": 0.06047109514474869,
      "learning_rate": 0.002880735150013432,
      "loss": 1.7666,
      "step": 22060
    },
    {
      "epoch": 0.040030549629980774,
      "grad_norm": 0.07913815230131149,
      "learning_rate": 0.0028804875980039885,
      "loss": 1.7664,
      "step": 22080
    },
    {
      "epoch": 0.04016222906955308,
      "grad_norm": 0.06452839821577072,
      "learning_rate": 0.0028802398000042895,
      "loss": 1.7689,
      "step": 22100
    },
    {
      "epoch": 0.040293908509125384,
      "grad_norm": 0.07609876990318298,
      "learning_rate": 0.0028799917560584907,
      "loss": 1.767,
      "step": 22120
    },
    {
      "epoch": 0.04042558794869769,
      "grad_norm": 0.07160958647727966,
      "learning_rate": 0.0028797434662107906,
      "loss": 1.7677,
      "step": 22140
    },
    {
      "epoch": 0.040557267388269994,
      "grad_norm": 0.06023573875427246,
      "learning_rate": 0.002879494930505433,
      "loss": 1.766,
      "step": 22160
    },
    {
      "epoch": 0.0406889468278423,
      "grad_norm": 0.06579054147005081,
      "learning_rate": 0.0028792461489867043,
      "loss": 1.7568,
      "step": 22180
    },
    {
      "epoch": 0.0408206262674146,
      "grad_norm": 0.065089151263237,
      "learning_rate": 0.0028789971216989347,
      "loss": 1.7607,
      "step": 22200
    },
    {
      "epoch": 0.04095230570698691,
      "grad_norm": 0.06882410496473312,
      "learning_rate": 0.002878747848686499,
      "loss": 1.7594,
      "step": 22220
    },
    {
      "epoch": 0.04108398514655922,
      "grad_norm": 0.08353780955076218,
      "learning_rate": 0.0028784983299938163,
      "loss": 1.7538,
      "step": 22240
    },
    {
      "epoch": 0.041215664586131524,
      "grad_norm": 0.06264527142047882,
      "learning_rate": 0.002878248565665347,
      "loss": 1.7511,
      "step": 22260
    },
    {
      "epoch": 0.04134734402570383,
      "grad_norm": 0.06513512134552002,
      "learning_rate": 0.0028779985557455968,
      "loss": 1.7527,
      "step": 22280
    },
    {
      "epoch": 0.041479023465276134,
      "grad_norm": 0.06116553023457527,
      "learning_rate": 0.0028777483002791165,
      "loss": 1.7505,
      "step": 22300
    },
    {
      "epoch": 0.04161070290484844,
      "grad_norm": 0.07746737450361252,
      "learning_rate": 0.002877497799310498,
      "loss": 1.7506,
      "step": 22320
    },
    {
      "epoch": 0.04174238234442074,
      "grad_norm": 0.05574808269739151,
      "learning_rate": 0.00287724705288438,
      "loss": 1.7452,
      "step": 22340
    },
    {
      "epoch": 0.04187406178399305,
      "grad_norm": 0.0626768171787262,
      "learning_rate": 0.002876996061045441,
      "loss": 1.7506,
      "step": 22360
    },
    {
      "epoch": 0.04200574122356535,
      "grad_norm": 0.07668133080005646,
      "learning_rate": 0.0028767448238384073,
      "loss": 1.7489,
      "step": 22380
    },
    {
      "epoch": 0.04213742066313766,
      "grad_norm": 0.08054382354021072,
      "learning_rate": 0.002876493341308046,
      "loss": 1.7492,
      "step": 22400
    },
    {
      "epoch": 0.04226910010270996,
      "grad_norm": 0.06886732578277588,
      "learning_rate": 0.0028762416134991697,
      "loss": 1.7482,
      "step": 22420
    },
    {
      "epoch": 0.04240077954228227,
      "grad_norm": 0.11965472251176834,
      "learning_rate": 0.0028759896404566333,
      "loss": 1.7443,
      "step": 22440
    },
    {
      "epoch": 0.04253245898185457,
      "grad_norm": 0.07691692560911179,
      "learning_rate": 0.0028757374222253365,
      "loss": 1.7421,
      "step": 22460
    },
    {
      "epoch": 0.04266413842142688,
      "grad_norm": 0.06630564481019974,
      "learning_rate": 0.002875484958850222,
      "loss": 1.7439,
      "step": 22480
    },
    {
      "epoch": 0.04279581786099918,
      "grad_norm": 0.06624884158372879,
      "learning_rate": 0.002875232250376277,
      "loss": 1.7532,
      "step": 22500
    },
    {
      "epoch": 0.042927497300571486,
      "grad_norm": 0.06218545883893967,
      "learning_rate": 0.002874979296848531,
      "loss": 1.7689,
      "step": 22520
    },
    {
      "epoch": 0.04305917674014379,
      "grad_norm": 0.08492391556501389,
      "learning_rate": 0.002874726098312059,
      "loss": 1.7707,
      "step": 22540
    },
    {
      "epoch": 0.043190856179716096,
      "grad_norm": 0.07384103536605835,
      "learning_rate": 0.0028744726548119784,
      "loss": 1.7758,
      "step": 22560
    },
    {
      "epoch": 0.04332253561928841,
      "grad_norm": 0.07037446647882462,
      "learning_rate": 0.0028742189663934496,
      "loss": 1.7694,
      "step": 22580
    },
    {
      "epoch": 0.04345421505886071,
      "grad_norm": 0.06753331422805786,
      "learning_rate": 0.0028739650331016785,
      "loss": 1.7684,
      "step": 22600
    },
    {
      "epoch": 0.04358589449843302,
      "grad_norm": 0.0832001268863678,
      "learning_rate": 0.0028737108549819136,
      "loss": 1.7703,
      "step": 22620
    },
    {
      "epoch": 0.04371757393800532,
      "grad_norm": 0.07064741849899292,
      "learning_rate": 0.002873456432079447,
      "loss": 1.7782,
      "step": 22640
    },
    {
      "epoch": 0.04384925337757763,
      "grad_norm": 0.06530604511499405,
      "learning_rate": 0.0028732017644396137,
      "loss": 1.7668,
      "step": 22660
    },
    {
      "epoch": 0.04398093281714993,
      "grad_norm": 0.09668663144111633,
      "learning_rate": 0.002872946852107795,
      "loss": 1.7674,
      "step": 22680
    },
    {
      "epoch": 0.044112612256722236,
      "grad_norm": 0.07591857761144638,
      "learning_rate": 0.002872691695129412,
      "loss": 1.7575,
      "step": 22700
    },
    {
      "epoch": 0.04424429169629454,
      "grad_norm": 0.08346087485551834,
      "learning_rate": 0.002872436293549932,
      "loss": 1.7595,
      "step": 22720
    },
    {
      "epoch": 0.044375971135866846,
      "grad_norm": 0.06290127336978912,
      "learning_rate": 0.0028721806474148663,
      "loss": 1.7576,
      "step": 22740
    },
    {
      "epoch": 0.04450765057543915,
      "grad_norm": 0.06168945133686066,
      "learning_rate": 0.002871924756769767,
      "loss": 1.7596,
      "step": 22760
    },
    {
      "epoch": 0.044639330015011455,
      "grad_norm": 0.07243330031633377,
      "learning_rate": 0.002871668621660233,
      "loss": 1.7604,
      "step": 22780
    },
    {
      "epoch": 0.04477100945458376,
      "grad_norm": 0.06622735410928726,
      "learning_rate": 0.002871412242131904,
      "loss": 1.7537,
      "step": 22800
    },
    {
      "epoch": 0.044902688894156065,
      "grad_norm": 0.057980917394161224,
      "learning_rate": 0.0028711556182304653,
      "loss": 1.7485,
      "step": 22820
    },
    {
      "epoch": 0.04503436833372837,
      "grad_norm": 0.07098107784986496,
      "learning_rate": 0.0028708987500016443,
      "loss": 1.748,
      "step": 22840
    },
    {
      "epoch": 0.045166047773300674,
      "grad_norm": 0.08247384428977966,
      "learning_rate": 0.002870641637491213,
      "loss": 1.7512,
      "step": 22860
    },
    {
      "epoch": 0.04529772721287298,
      "grad_norm": 0.07915489375591278,
      "learning_rate": 0.002870384280744987,
      "loss": 1.7551,
      "step": 22880
    },
    {
      "epoch": 0.045429406652445284,
      "grad_norm": 0.08467677235603333,
      "learning_rate": 0.0028701266798088236,
      "loss": 1.7454,
      "step": 22900
    },
    {
      "epoch": 0.045561086092017596,
      "grad_norm": 0.0650927871465683,
      "learning_rate": 0.002869868834728626,
      "loss": 1.7409,
      "step": 22920
    },
    {
      "epoch": 0.0456927655315899,
      "grad_norm": 0.08258968591690063,
      "learning_rate": 0.002869610745550339,
      "loss": 1.747,
      "step": 22940
    },
    {
      "epoch": 0.045824444971162205,
      "grad_norm": 0.06526441127061844,
      "learning_rate": 0.0028693524123199524,
      "loss": 1.7494,
      "step": 22960
    },
    {
      "epoch": 0.04595612441073451,
      "grad_norm": 0.08214429020881653,
      "learning_rate": 0.0028690938350834994,
      "loss": 1.7537,
      "step": 22980
    },
    {
      "epoch": 0.046087803850306815,
      "grad_norm": 0.06021787226200104,
      "learning_rate": 0.0028688350138870544,
      "loss": 1.7492,
      "step": 23000
    },
    {
      "epoch": 0.046087803850306815,
      "eval_loss": 1.614139437675476,
      "eval_runtime": 64.9061,
      "eval_samples_per_second": 15.407,
      "eval_steps_per_second": 15.407,
      "step": 23000
    },
    {
      "epoch": 0.04621948328987912,
      "grad_norm": 0.09553001075983047,
      "learning_rate": 0.002868575948776738,
      "loss": 1.751,
      "step": 23020
    },
    {
      "epoch": 0.046351162729451424,
      "grad_norm": 0.06959230452775955,
      "learning_rate": 0.0028683166397987137,
      "loss": 1.7443,
      "step": 23040
    },
    {
      "epoch": 0.04648284216902373,
      "grad_norm": 0.07799062877893448,
      "learning_rate": 0.0028680570869991863,
      "loss": 1.7412,
      "step": 23060
    },
    {
      "epoch": 0.046614521608596034,
      "grad_norm": 0.06044070050120354,
      "learning_rate": 0.002867797290424408,
      "loss": 1.7404,
      "step": 23080
    },
    {
      "epoch": 0.04674620104816834,
      "grad_norm": 0.06427774578332901,
      "learning_rate": 0.00286753725012067,
      "loss": 1.7424,
      "step": 23100
    },
    {
      "epoch": 0.04687788048774064,
      "grad_norm": 0.0699395090341568,
      "learning_rate": 0.002867276966134311,
      "loss": 1.7364,
      "step": 23120
    },
    {
      "epoch": 0.04700955992731295,
      "grad_norm": 0.05850491672754288,
      "learning_rate": 0.00286701643851171,
      "loss": 1.6901,
      "step": 23140
    },
    {
      "epoch": 0.04714123936688525,
      "grad_norm": 0.07173515111207962,
      "learning_rate": 0.002866755667299291,
      "loss": 1.6853,
      "step": 23160
    },
    {
      "epoch": 0.04727291880645756,
      "grad_norm": 0.10769501328468323,
      "learning_rate": 0.002866494652543521,
      "loss": 1.6832,
      "step": 23180
    },
    {
      "epoch": 0.04740459824602986,
      "grad_norm": 0.08263743668794632,
      "learning_rate": 0.0028662333942909104,
      "loss": 1.6789,
      "step": 23200
    },
    {
      "epoch": 0.04753627768560217,
      "grad_norm": 0.06656038016080856,
      "learning_rate": 0.0028659718925880128,
      "loss": 1.6759,
      "step": 23220
    },
    {
      "epoch": 0.04766795712517447,
      "grad_norm": 0.0891193225979805,
      "learning_rate": 0.0028657101474814264,
      "loss": 1.6761,
      "step": 23240
    },
    {
      "epoch": 0.047799636564746784,
      "grad_norm": 0.08038073033094406,
      "learning_rate": 0.0028654481590177907,
      "loss": 1.6701,
      "step": 23260
    },
    {
      "epoch": 0.04793131600431909,
      "grad_norm": 0.08320681005716324,
      "learning_rate": 0.0028651859272437897,
      "loss": 1.6733,
      "step": 23280
    },
    {
      "epoch": 0.04806299544389139,
      "grad_norm": 0.0696556568145752,
      "learning_rate": 0.0028649234522061516,
      "loss": 1.6736,
      "step": 23300
    },
    {
      "epoch": 0.0481946748834637,
      "grad_norm": 0.08962351828813553,
      "learning_rate": 0.002864660733951646,
      "loss": 1.6655,
      "step": 23320
    },
    {
      "epoch": 0.048326354323036,
      "grad_norm": 0.07071159034967422,
      "learning_rate": 0.0028643977725270877,
      "loss": 1.6765,
      "step": 23340
    },
    {
      "epoch": 0.04845803376260831,
      "grad_norm": 0.07777654379606247,
      "learning_rate": 0.0028641345679793337,
      "loss": 1.6659,
      "step": 23360
    },
    {
      "epoch": 0.04858971320218061,
      "grad_norm": 0.0689220279455185,
      "learning_rate": 0.002863871120355285,
      "loss": 1.6657,
      "step": 23380
    },
    {
      "epoch": 0.04872139264175292,
      "grad_norm": 0.08802857995033264,
      "learning_rate": 0.002863607429701884,
      "loss": 1.67,
      "step": 23400
    },
    {
      "epoch": 0.04885307208132522,
      "grad_norm": 0.06455916166305542,
      "learning_rate": 0.00286334349606612,
      "loss": 1.6648,
      "step": 23420
    },
    {
      "epoch": 0.048984751520897526,
      "grad_norm": 0.06590025126934052,
      "learning_rate": 0.0028630793194950226,
      "loss": 1.6652,
      "step": 23440
    },
    {
      "epoch": 0.04911643096046983,
      "grad_norm": 0.07274986058473587,
      "learning_rate": 0.0028628149000356658,
      "loss": 1.6589,
      "step": 23460
    },
    {
      "epoch": 0.049248110400042136,
      "grad_norm": 0.079200379550457,
      "learning_rate": 0.0028625502377351664,
      "loss": 1.6592,
      "step": 23480
    },
    {
      "epoch": 0.04937978983961444,
      "grad_norm": 0.08473269641399384,
      "learning_rate": 0.0028622853326406854,
      "loss": 1.6639,
      "step": 23500
    },
    {
      "epoch": 0.049511469279186746,
      "grad_norm": 0.0783248320221901,
      "learning_rate": 0.002862020184799426,
      "loss": 1.6609,
      "step": 23520
    },
    {
      "epoch": 0.04964314871875905,
      "grad_norm": 0.08220840990543365,
      "learning_rate": 0.002861754794258635,
      "loss": 1.6648,
      "step": 23540
    },
    {
      "epoch": 0.049774828158331355,
      "grad_norm": 0.06549611687660217,
      "learning_rate": 0.002861489161065603,
      "loss": 1.6619,
      "step": 23560
    },
    {
      "epoch": 0.04990650759790366,
      "grad_norm": 0.07199726998806,
      "learning_rate": 0.0028612232852676636,
      "loss": 1.6671,
      "step": 23580
    },
    {
      "epoch": 0.05003818703747597,
      "grad_norm": 0.08660496026277542,
      "learning_rate": 0.0028609571669121927,
      "loss": 1.6667,
      "step": 23600
    },
    {
      "epoch": 0.050169866477048276,
      "grad_norm": 0.10165523737668991,
      "learning_rate": 0.002860690806046611,
      "loss": 1.6579,
      "step": 23620
    },
    {
      "epoch": 0.05030154591662058,
      "grad_norm": 0.07491469383239746,
      "learning_rate": 0.00286042420271838,
      "loss": 1.662,
      "step": 23640
    },
    {
      "epoch": 0.050433225356192886,
      "grad_norm": 0.08252348750829697,
      "learning_rate": 0.0028601573569750085,
      "loss": 1.6665,
      "step": 23660
    },
    {
      "epoch": 0.05056490479576519,
      "grad_norm": 0.06357965618371964,
      "learning_rate": 0.0028598902688640434,
      "loss": 1.6753,
      "step": 23680
    },
    {
      "epoch": 0.050696584235337495,
      "grad_norm": 0.06429125368595123,
      "learning_rate": 0.0028596229384330787,
      "loss": 1.6727,
      "step": 23700
    },
    {
      "epoch": 0.0508282636749098,
      "grad_norm": 0.0685480460524559,
      "learning_rate": 0.0028593553657297504,
      "loss": 1.6862,
      "step": 23720
    },
    {
      "epoch": 0.050959943114482105,
      "grad_norm": 0.06911630928516388,
      "learning_rate": 0.002859087550801737,
      "loss": 1.6896,
      "step": 23740
    },
    {
      "epoch": 0.05109162255405441,
      "grad_norm": 0.06673968583345413,
      "learning_rate": 0.0028588194936967604,
      "loss": 1.6872,
      "step": 23760
    },
    {
      "epoch": 0.051223301993626714,
      "grad_norm": 0.08397223055362701,
      "learning_rate": 0.0028585511944625866,
      "loss": 1.6822,
      "step": 23780
    },
    {
      "epoch": 0.05135498143319902,
      "grad_norm": 0.0685010775923729,
      "learning_rate": 0.002858282653147024,
      "loss": 1.6926,
      "step": 23800
    },
    {
      "epoch": 0.051486660872771324,
      "grad_norm": 0.06479799002408981,
      "learning_rate": 0.0028580138697979237,
      "loss": 1.6875,
      "step": 23820
    },
    {
      "epoch": 0.05161834031234363,
      "grad_norm": 0.06732255220413208,
      "learning_rate": 0.002857744844463181,
      "loss": 1.6848,
      "step": 23840
    },
    {
      "epoch": 0.051750019751915934,
      "grad_norm": 0.06983063369989395,
      "learning_rate": 0.0028574755771907335,
      "loss": 1.6812,
      "step": 23860
    },
    {
      "epoch": 0.05188169919148824,
      "grad_norm": 0.07041189819574356,
      "learning_rate": 0.0028572060680285617,
      "loss": 1.6756,
      "step": 23880
    },
    {
      "epoch": 0.05201337863106054,
      "grad_norm": 0.06055330112576485,
      "learning_rate": 0.0028569363170246904,
      "loss": 1.6781,
      "step": 23900
    },
    {
      "epoch": 0.052145058070632855,
      "grad_norm": 0.07243553549051285,
      "learning_rate": 0.0028566663242271862,
      "loss": 1.6781,
      "step": 23920
    },
    {
      "epoch": 0.05227673751020516,
      "grad_norm": 0.08006641268730164,
      "learning_rate": 0.00285639608968416,
      "loss": 1.6846,
      "step": 23940
    },
    {
      "epoch": 0.052408416949777464,
      "grad_norm": 0.10588499903678894,
      "learning_rate": 0.0028561256134437647,
      "loss": 1.6757,
      "step": 23960
    },
    {
      "epoch": 0.05254009638934977,
      "grad_norm": 0.07262659072875977,
      "learning_rate": 0.0028558548955541965,
      "loss": 1.6781,
      "step": 23980
    },
    {
      "epoch": 0.052671775828922074,
      "grad_norm": 0.06767702847719193,
      "learning_rate": 0.002855583936063695,
      "loss": 1.672,
      "step": 24000
    },
    {
      "epoch": 0.052671775828922074,
      "eval_loss": 1.5004775524139404,
      "eval_runtime": 67.8841,
      "eval_samples_per_second": 14.731,
      "eval_steps_per_second": 14.731,
      "step": 24000
    },
    {
      "epoch": 0.05280345526849438,
      "grad_norm": 0.08169445395469666,
      "learning_rate": 0.002855312735020543,
      "loss": 1.675,
      "step": 24020
    },
    {
      "epoch": 0.05293513470806668,
      "grad_norm": 0.06959936767816544,
      "learning_rate": 0.0028550412924730657,
      "loss": 1.6778,
      "step": 24040
    },
    {
      "epoch": 0.05306681414763899,
      "grad_norm": 0.08186500519514084,
      "learning_rate": 0.002854769608469632,
      "loss": 1.6797,
      "step": 24060
    },
    {
      "epoch": 0.05319849358721129,
      "grad_norm": 0.07614962756633759,
      "learning_rate": 0.002854497683058653,
      "loss": 1.6682,
      "step": 24080
    },
    {
      "epoch": 0.0533301730267836,
      "grad_norm": 0.06458086520433426,
      "learning_rate": 0.002854225516288584,
      "loss": 1.6777,
      "step": 24100
    },
    {
      "epoch": 0.0534618524663559,
      "grad_norm": 0.07567057013511658,
      "learning_rate": 0.002853953108207922,
      "loss": 1.6709,
      "step": 24120
    },
    {
      "epoch": 0.05359353190592821,
      "grad_norm": 0.06437589973211288,
      "learning_rate": 0.002853680458865208,
      "loss": 1.672,
      "step": 24140
    },
    {
      "epoch": 0.05372521134550051,
      "grad_norm": 0.09382762014865875,
      "learning_rate": 0.0028534075683090254,
      "loss": 1.6754,
      "step": 24160
    },
    {
      "epoch": 0.05385689078507282,
      "grad_norm": 0.06308145076036453,
      "learning_rate": 0.002853134436588001,
      "loss": 1.6796,
      "step": 24180
    },
    {
      "epoch": 0.05398857022464512,
      "grad_norm": 0.07414955645799637,
      "learning_rate": 0.002852861063750804,
      "loss": 1.6753,
      "step": 24200
    },
    {
      "epoch": 0.054120249664217426,
      "grad_norm": 0.08207368105649948,
      "learning_rate": 0.0028525874498461475,
      "loss": 1.6737,
      "step": 24220
    },
    {
      "epoch": 0.05425192910378973,
      "grad_norm": 0.1388959437608719,
      "learning_rate": 0.0028523135949227864,
      "loss": 1.6711,
      "step": 24240
    },
    {
      "epoch": 0.05438360854336204,
      "grad_norm": 0.07871943712234497,
      "learning_rate": 0.0028520394990295193,
      "loss": 1.6799,
      "step": 24260
    },
    {
      "epoch": 0.05451528798293435,
      "grad_norm": 0.0691281110048294,
      "learning_rate": 0.0028517651622151884,
      "loss": 1.6845,
      "step": 24280
    },
    {
      "epoch": 0.05464696742250665,
      "grad_norm": 0.07609487324953079,
      "learning_rate": 0.002851490584528677,
      "loss": 1.7203,
      "step": 24300
    },
    {
      "epoch": 0.05477864686207896,
      "grad_norm": 0.07939011603593826,
      "learning_rate": 0.002851215766018913,
      "loss": 1.7355,
      "step": 24320
    },
    {
      "epoch": 0.05491032630165126,
      "grad_norm": 0.0882917046546936,
      "learning_rate": 0.0028509407067348652,
      "loss": 1.7458,
      "step": 24340
    },
    {
      "epoch": 0.05504200574122357,
      "grad_norm": 0.08259212225675583,
      "learning_rate": 0.0028506654067255487,
      "loss": 1.7469,
      "step": 24360
    },
    {
      "epoch": 0.05517368518079587,
      "grad_norm": 0.0653611272573471,
      "learning_rate": 0.0028503898660400175,
      "loss": 1.7565,
      "step": 24380
    },
    {
      "epoch": 0.055305364620368176,
      "grad_norm": 0.07420220971107483,
      "learning_rate": 0.002850114084727372,
      "loss": 1.7467,
      "step": 24400
    },
    {
      "epoch": 0.05543704405994048,
      "grad_norm": 0.12061482667922974,
      "learning_rate": 0.002849838062836753,
      "loss": 1.7476,
      "step": 24420
    },
    {
      "epoch": 0.055568723499512786,
      "grad_norm": 0.08582179993391037,
      "learning_rate": 0.0028495618004173453,
      "loss": 1.743,
      "step": 24440
    },
    {
      "epoch": 0.05570040293908509,
      "grad_norm": 0.11333516985177994,
      "learning_rate": 0.0028492852975183767,
      "loss": 1.7446,
      "step": 24460
    },
    {
      "epoch": 0.055832082378657395,
      "grad_norm": 0.06701093912124634,
      "learning_rate": 0.0028490085541891166,
      "loss": 1.7329,
      "step": 24480
    },
    {
      "epoch": 0.0559637618182297,
      "grad_norm": 0.07070864737033844,
      "learning_rate": 0.0028487315704788787,
      "loss": 1.741,
      "step": 24500
    },
    {
      "epoch": 0.056095441257802005,
      "grad_norm": 0.08799562603235245,
      "learning_rate": 0.0028484543464370187,
      "loss": 1.7349,
      "step": 24520
    },
    {
      "epoch": 0.05622712069737431,
      "grad_norm": 0.08984426409006119,
      "learning_rate": 0.002848176882112936,
      "loss": 1.7277,
      "step": 24540
    },
    {
      "epoch": 0.056358800136946614,
      "grad_norm": 0.08583886176347733,
      "learning_rate": 0.002847899177556072,
      "loss": 1.7304,
      "step": 24560
    },
    {
      "epoch": 0.05649047957651892,
      "grad_norm": 0.07309599220752716,
      "learning_rate": 0.0028476212328159105,
      "loss": 1.736,
      "step": 24580
    },
    {
      "epoch": 0.05662215901609123,
      "grad_norm": 0.10758428275585175,
      "learning_rate": 0.0028473430479419794,
      "loss": 1.7309,
      "step": 24600
    },
    {
      "epoch": 0.056753838455663536,
      "grad_norm": 0.07512976229190826,
      "learning_rate": 0.0028470646229838475,
      "loss": 1.7258,
      "step": 24620
    },
    {
      "epoch": 0.05688551789523584,
      "grad_norm": 0.09403645247220993,
      "learning_rate": 0.00284678595799113,
      "loss": 1.7245,
      "step": 24640
    },
    {
      "epoch": 0.057017197334808145,
      "grad_norm": 0.06885070353746414,
      "learning_rate": 0.0028465070530134797,
      "loss": 1.7257,
      "step": 24660
    },
    {
      "epoch": 0.05714887677438045,
      "grad_norm": 0.07384008169174194,
      "learning_rate": 0.0028462279081005967,
      "loss": 1.7237,
      "step": 24680
    },
    {
      "epoch": 0.057280556213952755,
      "grad_norm": 0.08138833940029144,
      "learning_rate": 0.0028459485233022215,
      "loss": 1.7216,
      "step": 24700
    },
    {
      "epoch": 0.05741223565352506,
      "grad_norm": 0.057137709110975266,
      "learning_rate": 0.002845668898668138,
      "loss": 1.7229,
      "step": 24720
    },
    {
      "epoch": 0.057543915093097364,
      "grad_norm": 0.06821935623884201,
      "learning_rate": 0.0028453890342481727,
      "loss": 1.7139,
      "step": 24740
    },
    {
      "epoch": 0.05767559453266967,
      "grad_norm": 0.07039092481136322,
      "learning_rate": 0.002845108930092195,
      "loss": 1.7257,
      "step": 24760
    },
    {
      "epoch": 0.057807273972241974,
      "grad_norm": 0.06676448881626129,
      "learning_rate": 0.0028448285862501164,
      "loss": 1.7204,
      "step": 24780
    },
    {
      "epoch": 0.05793895341181428,
      "grad_norm": 0.0648011863231659,
      "learning_rate": 0.0028445480027718922,
      "loss": 1.7128,
      "step": 24800
    },
    {
      "epoch": 0.05807063285138658,
      "grad_norm": 0.07681864500045776,
      "learning_rate": 0.0028442671797075194,
      "loss": 1.7201,
      "step": 24820
    },
    {
      "epoch": 0.05820231229095889,
      "grad_norm": 0.06368867307901382,
      "learning_rate": 0.0028439861171070386,
      "loss": 1.7169,
      "step": 24840
    },
    {
      "epoch": 0.05833399173053119,
      "grad_norm": 0.08015841245651245,
      "learning_rate": 0.0028437048150205322,
      "loss": 1.7143,
      "step": 24860
    },
    {
      "epoch": 0.0584656711701035,
      "grad_norm": 0.06920666247606277,
      "learning_rate": 0.0028434232734981255,
      "loss": 1.7262,
      "step": 24880
    },
    {
      "epoch": 0.0585973506096758,
      "grad_norm": 0.1089114099740982,
      "learning_rate": 0.002843141492589987,
      "loss": 1.7356,
      "step": 24900
    },
    {
      "epoch": 0.05872903004924811,
      "grad_norm": 0.06988651305437088,
      "learning_rate": 0.0028428594723463275,
      "loss": 1.7347,
      "step": 24920
    },
    {
      "epoch": 0.05886070948882042,
      "grad_norm": 0.076754629611969,
      "learning_rate": 0.0028425772128173998,
      "loss": 1.7332,
      "step": 24940
    },
    {
      "epoch": 0.058992388928392724,
      "grad_norm": 0.07548579573631287,
      "learning_rate": 0.0028422947140535008,
      "loss": 1.7258,
      "step": 24960
    },
    {
      "epoch": 0.05912406836796503,
      "grad_norm": 0.0791606530547142,
      "learning_rate": 0.0028420119761049687,
      "loss": 1.7298,
      "step": 24980
    },
    {
      "epoch": 0.05925574780753733,
      "grad_norm": 0.06826753914356232,
      "learning_rate": 0.0028417289990221853,
      "loss": 1.7286,
      "step": 25000
    },
    {
      "epoch": 0.05925574780753733,
      "eval_loss": 1.641815423965454,
      "eval_runtime": 66.1639,
      "eval_samples_per_second": 15.114,
      "eval_steps_per_second": 15.114,
      "step": 25000
    },
    {
      "epoch": 0.05938742724710964,
      "grad_norm": 0.06096088886260986,
      "learning_rate": 0.0028414457828555735,
      "loss": 1.7278,
      "step": 25020
    },
    {
      "epoch": 0.05951910668668194,
      "grad_norm": 0.0753740519285202,
      "learning_rate": 0.0028411623276556005,
      "loss": 1.7184,
      "step": 25040
    },
    {
      "epoch": 0.05965078612625425,
      "grad_norm": 0.06860582530498505,
      "learning_rate": 0.0028408786334727758,
      "loss": 1.7238,
      "step": 25060
    },
    {
      "epoch": 0.05978246556582655,
      "grad_norm": 0.08930736780166626,
      "learning_rate": 0.0028405947003576505,
      "loss": 1.717,
      "step": 25080
    },
    {
      "epoch": 0.05991414500539886,
      "grad_norm": 0.06507642567157745,
      "learning_rate": 0.002840310528360819,
      "loss": 1.7187,
      "step": 25100
    },
    {
      "epoch": 0.06004582444497116,
      "grad_norm": 0.07465964555740356,
      "learning_rate": 0.0028400261175329186,
      "loss": 1.7214,
      "step": 25120
    },
    {
      "epoch": 0.060177503884543466,
      "grad_norm": 0.08822104334831238,
      "learning_rate": 0.002839741467924628,
      "loss": 1.7129,
      "step": 25140
    },
    {
      "epoch": 0.06030918332411577,
      "grad_norm": 0.12070819735527039,
      "learning_rate": 0.00283945657958667,
      "loss": 1.7093,
      "step": 25160
    },
    {
      "epoch": 0.060440862763688076,
      "grad_norm": 0.06325086951255798,
      "learning_rate": 0.002839171452569808,
      "loss": 1.7043,
      "step": 25180
    },
    {
      "epoch": 0.06057254220326038,
      "grad_norm": 0.07239518314599991,
      "learning_rate": 0.00283888608692485,
      "loss": 1.7006,
      "step": 25200
    },
    {
      "epoch": 0.060704221642832686,
      "grad_norm": 0.08891430497169495,
      "learning_rate": 0.002838600482702645,
      "loss": 1.7102,
      "step": 25220
    },
    {
      "epoch": 0.06083590108240499,
      "grad_norm": 0.06925225257873535,
      "learning_rate": 0.0028383146399540856,
      "loss": 1.7128,
      "step": 25240
    },
    {
      "epoch": 0.060967580521977295,
      "grad_norm": 0.10547026246786118,
      "learning_rate": 0.0028380285587301054,
      "loss": 1.7051,
      "step": 25260
    },
    {
      "epoch": 0.06109925996154961,
      "grad_norm": 0.07027033716440201,
      "learning_rate": 0.0028377422390816823,
      "loss": 1.7204,
      "step": 25280
    },
    {
      "epoch": 0.06123093940112191,
      "grad_norm": 0.06962334364652634,
      "learning_rate": 0.0028374556810598357,
      "loss": 1.714,
      "step": 25300
    },
    {
      "epoch": 0.061362618840694216,
      "grad_norm": 0.09216281771659851,
      "learning_rate": 0.0028371688847156277,
      "loss": 1.7058,
      "step": 25320
    },
    {
      "epoch": 0.06149429828026652,
      "grad_norm": 0.07846725732088089,
      "learning_rate": 0.0028368818501001624,
      "loss": 1.712,
      "step": 25340
    },
    {
      "epoch": 0.061625977719838826,
      "grad_norm": 0.11076796799898148,
      "learning_rate": 0.002836594577264587,
      "loss": 1.7075,
      "step": 25360
    },
    {
      "epoch": 0.06175765715941113,
      "grad_norm": 0.07543148845434189,
      "learning_rate": 0.00283630706626009,
      "loss": 1.7033,
      "step": 25380
    },
    {
      "epoch": 0.061889336598983435,
      "grad_norm": 0.08785652369260788,
      "learning_rate": 0.002836019317137905,
      "loss": 1.7076,
      "step": 25400
    },
    {
      "epoch": 0.06202101603855574,
      "grad_norm": 0.0767190009355545,
      "learning_rate": 0.002835731329949305,
      "loss": 1.6997,
      "step": 25420
    },
    {
      "epoch": 0.062152695478128045,
      "grad_norm": 0.06853576004505157,
      "learning_rate": 0.002835443104745607,
      "loss": 1.7111,
      "step": 25440
    },
    {
      "epoch": 0.06228437491770035,
      "grad_norm": 0.07039535045623779,
      "learning_rate": 0.0028351546415781703,
      "loss": 1.6986,
      "step": 25460
    },
    {
      "epoch": 0.062416054357272654,
      "grad_norm": 0.06429757922887802,
      "learning_rate": 0.002834865940498396,
      "loss": 1.7016,
      "step": 25480
    },
    {
      "epoch": 0.06254773379684496,
      "grad_norm": 0.06879903376102448,
      "learning_rate": 0.0028345770015577275,
      "loss": 1.6425,
      "step": 25500
    },
    {
      "epoch": 0.06267941323641726,
      "grad_norm": 0.0694214403629303,
      "learning_rate": 0.0028342878248076527,
      "loss": 1.5967,
      "step": 25520
    },
    {
      "epoch": 0.06281109267598957,
      "grad_norm": 0.07605341076850891,
      "learning_rate": 0.002833998410299699,
      "loss": 1.5833,
      "step": 25540
    },
    {
      "epoch": 0.06294277211556187,
      "grad_norm": 0.06291548907756805,
      "learning_rate": 0.002833708758085438,
      "loss": 1.5772,
      "step": 25560
    },
    {
      "epoch": 0.06307445155513418,
      "grad_norm": 0.06334402412176132,
      "learning_rate": 0.0028334188682164825,
      "loss": 1.5762,
      "step": 25580
    },
    {
      "epoch": 0.06320613099470648,
      "grad_norm": 0.0940147116780281,
      "learning_rate": 0.002833128740744488,
      "loss": 1.5723,
      "step": 25600
    },
    {
      "epoch": 0.06333781043427879,
      "grad_norm": 0.0760306864976883,
      "learning_rate": 0.002832838375721154,
      "loss": 1.5759,
      "step": 25620
    },
    {
      "epoch": 0.06346948987385109,
      "grad_norm": 0.08564443141222,
      "learning_rate": 0.0028325477731982194,
      "loss": 1.5828,
      "step": 25640
    },
    {
      "epoch": 0.0636011693134234,
      "grad_norm": 0.07691756635904312,
      "learning_rate": 0.0028322569332274675,
      "loss": 1.5797,
      "step": 25660
    },
    {
      "epoch": 0.0637328487529957,
      "grad_norm": 0.07960242033004761,
      "learning_rate": 0.002831965855860723,
      "loss": 1.575,
      "step": 25680
    },
    {
      "epoch": 0.06386452819256801,
      "grad_norm": 0.08058685809373856,
      "learning_rate": 0.0028316745411498543,
      "loss": 1.5784,
      "step": 25700
    },
    {
      "epoch": 0.06399620763214031,
      "grad_norm": 0.07303532212972641,
      "learning_rate": 0.00283138298914677,
      "loss": 1.5821,
      "step": 25720
    },
    {
      "epoch": 0.06412788707171262,
      "grad_norm": 0.09177567064762115,
      "learning_rate": 0.002831091199903422,
      "loss": 1.5731,
      "step": 25740
    },
    {
      "epoch": 0.06425956651128492,
      "grad_norm": 0.07598966360092163,
      "learning_rate": 0.0028307991734718044,
      "loss": 1.5735,
      "step": 25760
    },
    {
      "epoch": 0.06439124595085724,
      "grad_norm": 0.07235264033079147,
      "learning_rate": 0.0028305069099039543,
      "loss": 1.5749,
      "step": 25780
    },
    {
      "epoch": 0.06452292539042954,
      "grad_norm": 0.08868271857500076,
      "learning_rate": 0.00283021440925195,
      "loss": 1.5741,
      "step": 25800
    },
    {
      "epoch": 0.06465460483000185,
      "grad_norm": 0.08162138611078262,
      "learning_rate": 0.0028299216715679124,
      "loss": 1.578,
      "step": 25820
    },
    {
      "epoch": 0.06478628426957415,
      "grad_norm": 0.0805659294128418,
      "learning_rate": 0.002829628696904005,
      "loss": 1.5809,
      "step": 25840
    },
    {
      "epoch": 0.06491796370914646,
      "grad_norm": 0.07114304602146149,
      "learning_rate": 0.0028293354853124327,
      "loss": 1.5869,
      "step": 25860
    },
    {
      "epoch": 0.06504964314871876,
      "grad_norm": 0.06668581813573837,
      "learning_rate": 0.0028290420368454433,
      "loss": 1.5876,
      "step": 25880
    },
    {
      "epoch": 0.06518132258829107,
      "grad_norm": 0.07414104044437408,
      "learning_rate": 0.0028287483515553272,
      "loss": 1.5846,
      "step": 25900
    },
    {
      "epoch": 0.06531300202786337,
      "grad_norm": 0.07118961215019226,
      "learning_rate": 0.002828454429494415,
      "loss": 1.5878,
      "step": 25920
    },
    {
      "epoch": 0.06544468146743568,
      "grad_norm": 0.08453836292028427,
      "learning_rate": 0.0028281602707150824,
      "loss": 1.5897,
      "step": 25940
    },
    {
      "epoch": 0.06557636090700798,
      "grad_norm": 0.07505936175584793,
      "learning_rate": 0.0028278658752697453,
      "loss": 1.5864,
      "step": 25960
    },
    {
      "epoch": 0.06570804034658029,
      "grad_norm": 0.07345987111330032,
      "learning_rate": 0.0028275712432108626,
      "loss": 1.5844,
      "step": 25980
    },
    {
      "epoch": 0.06583971978615259,
      "grad_norm": 0.09382303059101105,
      "learning_rate": 0.002827276374590934,
      "loss": 1.5951,
      "step": 26000
    },
    {
      "epoch": 0.06583971978615259,
      "eval_loss": 1.7229478359222412,
      "eval_runtime": 67.2917,
      "eval_samples_per_second": 14.861,
      "eval_steps_per_second": 14.861,
      "step": 26000
    },
    {
      "epoch": 0.0659713992257249,
      "grad_norm": 0.07171901315450668,
      "learning_rate": 0.0028269812694625036,
      "loss": 1.5965,
      "step": 26020
    },
    {
      "epoch": 0.0661030786652972,
      "grad_norm": 0.11737576127052307,
      "learning_rate": 0.002826685927878155,
      "loss": 1.5861,
      "step": 26040
    },
    {
      "epoch": 0.0662347581048695,
      "grad_norm": 0.1007668673992157,
      "learning_rate": 0.002826390349890517,
      "loss": 1.6206,
      "step": 26060
    },
    {
      "epoch": 0.06636643754444181,
      "grad_norm": 0.07489984482526779,
      "learning_rate": 0.0028260945355522583,
      "loss": 1.6584,
      "step": 26080
    },
    {
      "epoch": 0.06649811698401412,
      "grad_norm": 0.07443971931934357,
      "learning_rate": 0.0028257984849160895,
      "loss": 1.6751,
      "step": 26100
    },
    {
      "epoch": 0.06662979642358642,
      "grad_norm": 0.07076618820428848,
      "learning_rate": 0.0028255021980347654,
      "loss": 1.6871,
      "step": 26120
    },
    {
      "epoch": 0.06676147586315873,
      "grad_norm": 0.08177798986434937,
      "learning_rate": 0.0028252056749610804,
      "loss": 1.6991,
      "step": 26140
    },
    {
      "epoch": 0.06689315530273103,
      "grad_norm": 0.08380310237407684,
      "learning_rate": 0.0028249089157478735,
      "loss": 1.6959,
      "step": 26160
    },
    {
      "epoch": 0.06702483474230334,
      "grad_norm": 0.08319336175918579,
      "learning_rate": 0.0028246119204480232,
      "loss": 1.6876,
      "step": 26180
    },
    {
      "epoch": 0.06715651418187564,
      "grad_norm": 0.07932873070240021,
      "learning_rate": 0.0028243146891144518,
      "loss": 1.6902,
      "step": 26200
    },
    {
      "epoch": 0.06728819362144794,
      "grad_norm": 0.06260685622692108,
      "learning_rate": 0.0028240172218001233,
      "loss": 1.6899,
      "step": 26220
    },
    {
      "epoch": 0.06741987306102025,
      "grad_norm": 0.09259559959173203,
      "learning_rate": 0.002823719518558044,
      "loss": 1.6746,
      "step": 26240
    },
    {
      "epoch": 0.06755155250059255,
      "grad_norm": 0.08179523050785065,
      "learning_rate": 0.0028234215794412615,
      "loss": 1.6782,
      "step": 26260
    },
    {
      "epoch": 0.06768323194016486,
      "grad_norm": 0.08897533267736435,
      "learning_rate": 0.002823123404502866,
      "loss": 1.6725,
      "step": 26280
    },
    {
      "epoch": 0.06781491137973716,
      "grad_norm": 0.06979696452617645,
      "learning_rate": 0.0028228249937959893,
      "loss": 1.6782,
      "step": 26300
    },
    {
      "epoch": 0.06794659081930947,
      "grad_norm": 0.08062341064214706,
      "learning_rate": 0.002822526347373805,
      "loss": 1.6731,
      "step": 26320
    },
    {
      "epoch": 0.06807827025888177,
      "grad_norm": 0.08686768263578415,
      "learning_rate": 0.0028222274652895305,
      "loss": 1.6711,
      "step": 26340
    },
    {
      "epoch": 0.06820994969845408,
      "grad_norm": 0.07862361520528793,
      "learning_rate": 0.0028219283475964226,
      "loss": 1.6678,
      "step": 26360
    },
    {
      "epoch": 0.06834162913802638,
      "grad_norm": 0.08037963509559631,
      "learning_rate": 0.0028216289943477817,
      "loss": 1.6723,
      "step": 26380
    },
    {
      "epoch": 0.06847330857759869,
      "grad_norm": 0.06775148212909698,
      "learning_rate": 0.0028213294055969506,
      "loss": 1.6675,
      "step": 26400
    },
    {
      "epoch": 0.06860498801717099,
      "grad_norm": 0.07389701157808304,
      "learning_rate": 0.0028210295813973123,
      "loss": 1.6645,
      "step": 26420
    },
    {
      "epoch": 0.0687366674567433,
      "grad_norm": 0.0821567252278328,
      "learning_rate": 0.0028207295218022925,
      "loss": 1.664,
      "step": 26440
    },
    {
      "epoch": 0.06886834689631562,
      "grad_norm": 0.0829467922449112,
      "learning_rate": 0.0028204292268653604,
      "loss": 1.66,
      "step": 26460
    },
    {
      "epoch": 0.06900002633588792,
      "grad_norm": 0.07580474019050598,
      "learning_rate": 0.0028201286966400246,
      "loss": 1.6631,
      "step": 26480
    },
    {
      "epoch": 0.06913170577546023,
      "grad_norm": 0.07539396733045578,
      "learning_rate": 0.0028198279311798372,
      "loss": 1.6594,
      "step": 26500
    },
    {
      "epoch": 0.06926338521503253,
      "grad_norm": 0.08187814056873322,
      "learning_rate": 0.0028195269305383917,
      "loss": 1.6656,
      "step": 26520
    },
    {
      "epoch": 0.06939506465460483,
      "grad_norm": 0.08854663372039795,
      "learning_rate": 0.0028192256947693244,
      "loss": 1.6616,
      "step": 26540
    },
    {
      "epoch": 0.06952674409417714,
      "grad_norm": 0.08540527522563934,
      "learning_rate": 0.0028189242239263124,
      "loss": 1.6572,
      "step": 26560
    },
    {
      "epoch": 0.06965842353374944,
      "grad_norm": 0.07513197511434555,
      "learning_rate": 0.0028186225180630744,
      "loss": 1.658,
      "step": 26580
    },
    {
      "epoch": 0.06979010297332175,
      "grad_norm": 0.09598047286272049,
      "learning_rate": 0.002818320577233372,
      "loss": 1.667,
      "step": 26600
    },
    {
      "epoch": 0.06992178241289405,
      "grad_norm": 0.07224708050489426,
      "learning_rate": 0.0028180184014910082,
      "loss": 1.6529,
      "step": 26620
    },
    {
      "epoch": 0.07005346185246636,
      "grad_norm": 0.08062656968832016,
      "learning_rate": 0.0028177159908898282,
      "loss": 1.6542,
      "step": 26640
    },
    {
      "epoch": 0.07018514129203866,
      "grad_norm": 0.07291059195995331,
      "learning_rate": 0.0028174133454837193,
      "loss": 1.6783,
      "step": 26660
    },
    {
      "epoch": 0.07031682073161097,
      "grad_norm": 0.07700487226247787,
      "learning_rate": 0.002817110465326609,
      "loss": 1.7044,
      "step": 26680
    },
    {
      "epoch": 0.07044850017118327,
      "grad_norm": 0.23015104234218597,
      "learning_rate": 0.002816807350472469,
      "loss": 1.7165,
      "step": 26700
    },
    {
      "epoch": 0.07058017961075558,
      "grad_norm": 0.10937873274087906,
      "learning_rate": 0.00281650400097531,
      "loss": 1.7157,
      "step": 26720
    },
    {
      "epoch": 0.07071185905032788,
      "grad_norm": 0.0746932178735733,
      "learning_rate": 0.0028162004168891877,
      "loss": 1.7176,
      "step": 26740
    },
    {
      "epoch": 0.07084353848990019,
      "grad_norm": 0.07187461853027344,
      "learning_rate": 0.0028158965982681976,
      "loss": 1.7088,
      "step": 26760
    },
    {
      "epoch": 0.07097521792947249,
      "grad_norm": 0.0845644474029541,
      "learning_rate": 0.0028155925451664766,
      "loss": 1.7109,
      "step": 26780
    },
    {
      "epoch": 0.0711068973690448,
      "grad_norm": 0.0642613023519516,
      "learning_rate": 0.002815288257638205,
      "loss": 1.7109,
      "step": 26800
    },
    {
      "epoch": 0.0712385768086171,
      "grad_norm": 0.06674417108297348,
      "learning_rate": 0.0028149837357376033,
      "loss": 1.7072,
      "step": 26820
    },
    {
      "epoch": 0.0713702562481894,
      "grad_norm": 0.0797337219119072,
      "learning_rate": 0.0028146789795189353,
      "loss": 1.7036,
      "step": 26840
    },
    {
      "epoch": 0.07150193568776171,
      "grad_norm": 0.06999875605106354,
      "learning_rate": 0.0028143739890365053,
      "loss": 1.6998,
      "step": 26860
    },
    {
      "epoch": 0.07163361512733402,
      "grad_norm": 0.10101525485515594,
      "learning_rate": 0.0028140687643446603,
      "loss": 1.6987,
      "step": 26880
    },
    {
      "epoch": 0.07176529456690632,
      "grad_norm": 0.06726180762052536,
      "learning_rate": 0.002813763305497788,
      "loss": 1.7011,
      "step": 26900
    },
    {
      "epoch": 0.07189697400647863,
      "grad_norm": 0.07989557832479477,
      "learning_rate": 0.002813457612550319,
      "loss": 1.6982,
      "step": 26920
    },
    {
      "epoch": 0.07202865344605093,
      "grad_norm": 0.08156982809305191,
      "learning_rate": 0.0028131516855567236,
      "loss": 1.6946,
      "step": 26940
    },
    {
      "epoch": 0.07216033288562324,
      "grad_norm": 0.09064619243144989,
      "learning_rate": 0.0028128455245715173,
      "loss": 1.6942,
      "step": 26960
    },
    {
      "epoch": 0.07229201232519554,
      "grad_norm": 0.08123116940259933,
      "learning_rate": 0.0028125391296492533,
      "loss": 1.6931,
      "step": 26980
    },
    {
      "epoch": 0.07242369176476784,
      "grad_norm": 0.07538691908121109,
      "learning_rate": 0.00281223250084453,
      "loss": 1.6849,
      "step": 27000
    },
    {
      "epoch": 0.07242369176476784,
      "eval_loss": 1.7134445905685425,
      "eval_runtime": 66.2906,
      "eval_samples_per_second": 15.085,
      "eval_steps_per_second": 15.085,
      "step": 27000
    },
    {
      "epoch": 0.07255537120434015,
      "grad_norm": 0.07675933837890625,
      "learning_rate": 0.0028119256382119844,
      "loss": 1.6956,
      "step": 27020
    },
    {
      "epoch": 0.07268705064391245,
      "grad_norm": 0.07131894677877426,
      "learning_rate": 0.0028116185418062965,
      "loss": 1.6849,
      "step": 27040
    },
    {
      "epoch": 0.07281873008348476,
      "grad_norm": 0.07346808910369873,
      "learning_rate": 0.0028113112116821895,
      "loss": 1.6826,
      "step": 27060
    },
    {
      "epoch": 0.07295040952305706,
      "grad_norm": 0.10725018382072449,
      "learning_rate": 0.002811003647894426,
      "loss": 1.6802,
      "step": 27080
    },
    {
      "epoch": 0.07308208896262937,
      "grad_norm": 0.07160640507936478,
      "learning_rate": 0.0028106958504978113,
      "loss": 1.6832,
      "step": 27100
    },
    {
      "epoch": 0.07321376840220169,
      "grad_norm": 0.09838573634624481,
      "learning_rate": 0.0028103878195471915,
      "loss": 1.6867,
      "step": 27120
    },
    {
      "epoch": 0.07334544784177399,
      "grad_norm": 0.08039765805006027,
      "learning_rate": 0.0028100795550974555,
      "loss": 1.6859,
      "step": 27140
    },
    {
      "epoch": 0.0734771272813463,
      "grad_norm": 0.096266008913517,
      "learning_rate": 0.0028097710572035327,
      "loss": 1.6887,
      "step": 27160
    },
    {
      "epoch": 0.0736088067209186,
      "grad_norm": 0.08196264505386353,
      "learning_rate": 0.0028094623259203953,
      "loss": 1.681,
      "step": 27180
    },
    {
      "epoch": 0.0737404861604909,
      "grad_norm": 0.0661395713686943,
      "learning_rate": 0.002809153361303056,
      "loss": 1.6812,
      "step": 27200
    },
    {
      "epoch": 0.07387216560006321,
      "grad_norm": 0.07243622839450836,
      "learning_rate": 0.002808844163406569,
      "loss": 1.6778,
      "step": 27220
    },
    {
      "epoch": 0.07400384503963552,
      "grad_norm": 0.06151533126831055,
      "learning_rate": 0.002808534732286032,
      "loss": 1.684,
      "step": 27240
    },
    {
      "epoch": 0.07413552447920782,
      "grad_norm": 0.06907381862401962,
      "learning_rate": 0.0028082250679965803,
      "loss": 1.6891,
      "step": 27260
    },
    {
      "epoch": 0.07426720391878013,
      "grad_norm": 0.07179991155862808,
      "learning_rate": 0.0028079151705933955,
      "loss": 1.7013,
      "step": 27280
    },
    {
      "epoch": 0.07439888335835243,
      "grad_norm": 0.0859934464097023,
      "learning_rate": 0.002807605040131698,
      "loss": 1.6998,
      "step": 27300
    },
    {
      "epoch": 0.07453056279792473,
      "grad_norm": 0.05939393863081932,
      "learning_rate": 0.0028072946766667497,
      "loss": 1.7046,
      "step": 27320
    },
    {
      "epoch": 0.07466224223749704,
      "grad_norm": 0.08368372917175293,
      "learning_rate": 0.0028069840802538546,
      "loss": 1.7019,
      "step": 27340
    },
    {
      "epoch": 0.07479392167706934,
      "grad_norm": 0.06464136391878128,
      "learning_rate": 0.0028066732509483584,
      "loss": 1.7044,
      "step": 27360
    },
    {
      "epoch": 0.07492560111664165,
      "grad_norm": 0.08051209896802902,
      "learning_rate": 0.002806362188805648,
      "loss": 1.7057,
      "step": 27380
    },
    {
      "epoch": 0.07505728055621395,
      "grad_norm": 0.08102471381425858,
      "learning_rate": 0.0028060508938811514,
      "loss": 1.6942,
      "step": 27400
    },
    {
      "epoch": 0.07518895999578626,
      "grad_norm": 0.0877891480922699,
      "learning_rate": 0.002805739366230339,
      "loss": 1.6992,
      "step": 27420
    },
    {
      "epoch": 0.07532063943535856,
      "grad_norm": 0.0812918171286583,
      "learning_rate": 0.002805427605908722,
      "loss": 1.6922,
      "step": 27440
    },
    {
      "epoch": 0.07545231887493087,
      "grad_norm": 0.07641803473234177,
      "learning_rate": 0.0028051156129718538,
      "loss": 1.6876,
      "step": 27460
    },
    {
      "epoch": 0.07558399831450317,
      "grad_norm": 0.12771321833133698,
      "learning_rate": 0.0028048033874753277,
      "loss": 1.6847,
      "step": 27480
    },
    {
      "epoch": 0.07571567775407548,
      "grad_norm": 0.06846945732831955,
      "learning_rate": 0.00280449092947478,
      "loss": 1.6961,
      "step": 27500
    },
    {
      "epoch": 0.07584735719364778,
      "grad_norm": 0.0757148414850235,
      "learning_rate": 0.0028041782390258872,
      "loss": 1.6922,
      "step": 27520
    },
    {
      "epoch": 0.07597903663322009,
      "grad_norm": 0.07266736030578613,
      "learning_rate": 0.0028038653161843686,
      "loss": 1.6867,
      "step": 27540
    },
    {
      "epoch": 0.07611071607279239,
      "grad_norm": 0.10380854457616806,
      "learning_rate": 0.0028035521610059845,
      "loss": 1.686,
      "step": 27560
    },
    {
      "epoch": 0.0762423955123647,
      "grad_norm": 0.063069187104702,
      "learning_rate": 0.0028032387735465354,
      "loss": 1.6815,
      "step": 27580
    },
    {
      "epoch": 0.076374074951937,
      "grad_norm": 0.08514589071273804,
      "learning_rate": 0.0028029251538618646,
      "loss": 1.6811,
      "step": 27600
    },
    {
      "epoch": 0.0765057543915093,
      "grad_norm": 0.07171927392482758,
      "learning_rate": 0.002802611302007856,
      "loss": 1.6866,
      "step": 27620
    },
    {
      "epoch": 0.07663743383108161,
      "grad_norm": 0.07096509635448456,
      "learning_rate": 0.002802297218040436,
      "loss": 1.6754,
      "step": 27640
    },
    {
      "epoch": 0.07676911327065392,
      "grad_norm": 0.07615986466407776,
      "learning_rate": 0.0028019829020155696,
      "loss": 1.679,
      "step": 27660
    },
    {
      "epoch": 0.07690079271022622,
      "grad_norm": 0.09297886490821838,
      "learning_rate": 0.0028016683539892665,
      "loss": 1.6825,
      "step": 27680
    },
    {
      "epoch": 0.07703247214979853,
      "grad_norm": 0.07821375876665115,
      "learning_rate": 0.002801353574017576,
      "loss": 1.6857,
      "step": 27700
    },
    {
      "epoch": 0.07716415158937083,
      "grad_norm": 0.08595246821641922,
      "learning_rate": 0.002801038562156589,
      "loss": 1.6732,
      "step": 27720
    },
    {
      "epoch": 0.07729583102894313,
      "grad_norm": 0.08263617008924484,
      "learning_rate": 0.0028007233184624385,
      "loss": 1.6812,
      "step": 27740
    },
    {
      "epoch": 0.07742751046851544,
      "grad_norm": 0.06740175932645798,
      "learning_rate": 0.002800407842991296,
      "loss": 1.6786,
      "step": 27760
    },
    {
      "epoch": 0.07755918990808774,
      "grad_norm": 0.07347702234983444,
      "learning_rate": 0.002800092135799378,
      "loss": 1.6849,
      "step": 27780
    },
    {
      "epoch": 0.07769086934766006,
      "grad_norm": 0.06432320922613144,
      "learning_rate": 0.0027997761969429406,
      "loss": 1.6729,
      "step": 27800
    },
    {
      "epoch": 0.07782254878723237,
      "grad_norm": 0.06912487000226974,
      "learning_rate": 0.0027994600264782806,
      "loss": 1.6729,
      "step": 27820
    },
    {
      "epoch": 0.07795422822680467,
      "grad_norm": 0.07025229185819626,
      "learning_rate": 0.0027991436244617367,
      "loss": 1.6717,
      "step": 27840
    },
    {
      "epoch": 0.07808590766637698,
      "grad_norm": 0.06698182970285416,
      "learning_rate": 0.002798826990949689,
      "loss": 1.6741,
      "step": 27860
    },
    {
      "epoch": 0.07821758710594928,
      "grad_norm": 0.07972477376461029,
      "learning_rate": 0.0027985101259985594,
      "loss": 1.6626,
      "step": 27880
    },
    {
      "epoch": 0.07834926654552159,
      "grad_norm": 0.08026134967803955,
      "learning_rate": 0.0027981930296648084,
      "loss": 1.6533,
      "step": 27900
    },
    {
      "epoch": 0.07848094598509389,
      "grad_norm": 0.07009740173816681,
      "learning_rate": 0.002797875702004942,
      "loss": 1.6447,
      "step": 27920
    },
    {
      "epoch": 0.0786126254246662,
      "grad_norm": 0.0895591527223587,
      "learning_rate": 0.002797558143075504,
      "loss": 1.6316,
      "step": 27940
    },
    {
      "epoch": 0.0787443048642385,
      "grad_norm": 0.07302293926477432,
      "learning_rate": 0.00279724035293308,
      "loss": 1.6278,
      "step": 27960
    },
    {
      "epoch": 0.0788759843038108,
      "grad_norm": 0.06689465790987015,
      "learning_rate": 0.002796922331634298,
      "loss": 1.6277,
      "step": 27980
    },
    {
      "epoch": 0.07900766374338311,
      "grad_norm": 0.12900836765766144,
      "learning_rate": 0.002796604079235826,
      "loss": 1.6216,
      "step": 28000
    },
    {
      "epoch": 0.07900766374338311,
      "eval_loss": 1.57399582862854,
      "eval_runtime": 67.3732,
      "eval_samples_per_second": 14.843,
      "eval_steps_per_second": 14.843,
      "step": 28000
    },
    {
      "epoch": 0.07913934318295542,
      "grad_norm": 0.09559524804353714,
      "learning_rate": 0.002796285595794375,
      "loss": 1.6188,
      "step": 28020
    },
    {
      "epoch": 0.07927102262252772,
      "grad_norm": 0.09647485613822937,
      "learning_rate": 0.002795966881366694,
      "loss": 1.6111,
      "step": 28040
    },
    {
      "epoch": 0.07940270206210003,
      "grad_norm": 0.0804126113653183,
      "learning_rate": 0.002795647936009576,
      "loss": 1.624,
      "step": 28060
    },
    {
      "epoch": 0.07953438150167233,
      "grad_norm": 0.1041126698255539,
      "learning_rate": 0.0027953287597798538,
      "loss": 1.6142,
      "step": 28080
    },
    {
      "epoch": 0.07966606094124463,
      "grad_norm": 0.070301353931427,
      "learning_rate": 0.002795009352734402,
      "loss": 1.6135,
      "step": 28100
    },
    {
      "epoch": 0.07979774038081694,
      "grad_norm": 0.07575052231550217,
      "learning_rate": 0.0027946897149301354,
      "loss": 1.6113,
      "step": 28120
    },
    {
      "epoch": 0.07992941982038924,
      "grad_norm": 0.0882832482457161,
      "learning_rate": 0.0027943698464240117,
      "loss": 1.6073,
      "step": 28140
    },
    {
      "epoch": 0.08006109925996155,
      "grad_norm": 0.07412759959697723,
      "learning_rate": 0.002794049747273027,
      "loss": 1.6025,
      "step": 28160
    },
    {
      "epoch": 0.08019277869953385,
      "grad_norm": 0.10568110644817352,
      "learning_rate": 0.002793729417534221,
      "loss": 1.6053,
      "step": 28180
    },
    {
      "epoch": 0.08032445813910616,
      "grad_norm": 0.07552295923233032,
      "learning_rate": 0.002793408857264674,
      "loss": 1.6057,
      "step": 28200
    },
    {
      "epoch": 0.08045613757867846,
      "grad_norm": 0.07230095565319061,
      "learning_rate": 0.0027930880665215057,
      "loss": 1.6072,
      "step": 28220
    },
    {
      "epoch": 0.08058781701825077,
      "grad_norm": 0.07529731094837189,
      "learning_rate": 0.002792767045361879,
      "loss": 1.6136,
      "step": 28240
    },
    {
      "epoch": 0.08071949645782307,
      "grad_norm": 0.08432866632938385,
      "learning_rate": 0.0027924457938429967,
      "loss": 1.6122,
      "step": 28260
    },
    {
      "epoch": 0.08085117589739538,
      "grad_norm": 0.08389344066381454,
      "learning_rate": 0.002792124312022102,
      "loss": 1.6049,
      "step": 28280
    },
    {
      "epoch": 0.08098285533696768,
      "grad_norm": 0.08148514479398727,
      "learning_rate": 0.0027918025999564815,
      "loss": 1.6022,
      "step": 28300
    },
    {
      "epoch": 0.08111453477653999,
      "grad_norm": 0.07762300223112106,
      "learning_rate": 0.00279148065770346,
      "loss": 1.6088,
      "step": 28320
    },
    {
      "epoch": 0.08124621421611229,
      "grad_norm": 0.08115459978580475,
      "learning_rate": 0.0027911584853204056,
      "loss": 1.6068,
      "step": 28340
    },
    {
      "epoch": 0.0813778936556846,
      "grad_norm": 0.08367536962032318,
      "learning_rate": 0.0027908360828647265,
      "loss": 1.5971,
      "step": 28360
    },
    {
      "epoch": 0.0815095730952569,
      "grad_norm": 0.08000832796096802,
      "learning_rate": 0.002790513450393871,
      "loss": 1.6084,
      "step": 28380
    },
    {
      "epoch": 0.0816412525348292,
      "grad_norm": 0.08210103958845139,
      "learning_rate": 0.00279019058796533,
      "loss": 1.6077,
      "step": 28400
    },
    {
      "epoch": 0.08177293197440151,
      "grad_norm": 0.07329817861318588,
      "learning_rate": 0.002789867495636635,
      "loss": 1.61,
      "step": 28420
    },
    {
      "epoch": 0.08190461141397382,
      "grad_norm": 0.06839757412672043,
      "learning_rate": 0.002789544173465357,
      "loss": 1.6128,
      "step": 28440
    },
    {
      "epoch": 0.08203629085354612,
      "grad_norm": 0.06943529099225998,
      "learning_rate": 0.002789220621509109,
      "loss": 1.6285,
      "step": 28460
    },
    {
      "epoch": 0.08216797029311844,
      "grad_norm": 0.10486973077058792,
      "learning_rate": 0.0027888968398255467,
      "loss": 1.6335,
      "step": 28480
    },
    {
      "epoch": 0.08229964973269074,
      "grad_norm": 0.07937227934598923,
      "learning_rate": 0.0027885728284723628,
      "loss": 1.6372,
      "step": 28500
    },
    {
      "epoch": 0.08243132917226305,
      "grad_norm": 0.07697343081235886,
      "learning_rate": 0.002788248587507295,
      "loss": 1.6282,
      "step": 28520
    },
    {
      "epoch": 0.08256300861183535,
      "grad_norm": 0.08481860160827637,
      "learning_rate": 0.0027879241169881186,
      "loss": 1.6265,
      "step": 28540
    },
    {
      "epoch": 0.08269468805140766,
      "grad_norm": 0.08266434073448181,
      "learning_rate": 0.0027875994169726526,
      "loss": 1.628,
      "step": 28560
    },
    {
      "epoch": 0.08282636749097996,
      "grad_norm": 0.07758313417434692,
      "learning_rate": 0.0027872744875187547,
      "loss": 1.6168,
      "step": 28580
    },
    {
      "epoch": 0.08295804693055227,
      "grad_norm": 0.07969856262207031,
      "learning_rate": 0.002786949328684325,
      "loss": 1.6218,
      "step": 28600
    },
    {
      "epoch": 0.08308972637012457,
      "grad_norm": 0.06928009539842606,
      "learning_rate": 0.002786623940527303,
      "loss": 1.614,
      "step": 28620
    },
    {
      "epoch": 0.08322140580969688,
      "grad_norm": 0.08013467490673065,
      "learning_rate": 0.0027862983231056707,
      "loss": 1.6131,
      "step": 28640
    },
    {
      "epoch": 0.08335308524926918,
      "grad_norm": 0.07733161002397537,
      "learning_rate": 0.0027859724764774494,
      "loss": 1.6108,
      "step": 28660
    },
    {
      "epoch": 0.08348476468884149,
      "grad_norm": 0.10586979240179062,
      "learning_rate": 0.0027856464007007028,
      "loss": 1.6141,
      "step": 28680
    },
    {
      "epoch": 0.08361644412841379,
      "grad_norm": 0.07554934173822403,
      "learning_rate": 0.002785320095833534,
      "loss": 1.6104,
      "step": 28700
    },
    {
      "epoch": 0.0837481235679861,
      "grad_norm": 0.08737286925315857,
      "learning_rate": 0.0027849935619340873,
      "loss": 1.6126,
      "step": 28720
    },
    {
      "epoch": 0.0838798030075584,
      "grad_norm": 0.07767580449581146,
      "learning_rate": 0.002784666799060549,
      "loss": 1.6056,
      "step": 28740
    },
    {
      "epoch": 0.0840114824471307,
      "grad_norm": 0.09215500950813293,
      "learning_rate": 0.0027843398072711447,
      "loss": 1.605,
      "step": 28760
    },
    {
      "epoch": 0.08414316188670301,
      "grad_norm": 0.09147578477859497,
      "learning_rate": 0.002784012586624141,
      "loss": 1.6081,
      "step": 28780
    },
    {
      "epoch": 0.08427484132627532,
      "grad_norm": 0.09132102876901627,
      "learning_rate": 0.002783685137177846,
      "loss": 1.6054,
      "step": 28800
    },
    {
      "epoch": 0.08440652076584762,
      "grad_norm": 0.09715555608272552,
      "learning_rate": 0.0027833574589906087,
      "loss": 1.6011,
      "step": 28820
    },
    {
      "epoch": 0.08453820020541992,
      "grad_norm": 0.09655196219682693,
      "learning_rate": 0.0027830295521208176,
      "loss": 1.602,
      "step": 28840
    },
    {
      "epoch": 0.08466987964499223,
      "grad_norm": 0.09575804322957993,
      "learning_rate": 0.002782701416626903,
      "loss": 1.606,
      "step": 28860
    },
    {
      "epoch": 0.08480155908456453,
      "grad_norm": 0.08992674201726913,
      "learning_rate": 0.002782373052567336,
      "loss": 1.6016,
      "step": 28880
    },
    {
      "epoch": 0.08493323852413684,
      "grad_norm": 0.13212674856185913,
      "learning_rate": 0.002782044460000627,
      "loss": 1.6017,
      "step": 28900
    },
    {
      "epoch": 0.08506491796370914,
      "grad_norm": 0.08106255531311035,
      "learning_rate": 0.002781715638985329,
      "loss": 1.6053,
      "step": 28920
    },
    {
      "epoch": 0.08519659740328145,
      "grad_norm": 0.09823165833950043,
      "learning_rate": 0.002781386589580035,
      "loss": 1.6067,
      "step": 28940
    },
    {
      "epoch": 0.08532827684285375,
      "grad_norm": 0.07196325063705444,
      "learning_rate": 0.002781057311843379,
      "loss": 1.6017,
      "step": 28960
    },
    {
      "epoch": 0.08545995628242606,
      "grad_norm": 0.0797024518251419,
      "learning_rate": 0.0027807278058340334,
      "loss": 1.6051,
      "step": 28980
    },
    {
      "epoch": 0.08559163572199836,
      "grad_norm": 0.08300334960222244,
      "learning_rate": 0.0027803980716107147,
      "loss": 1.6088,
      "step": 29000
    },
    {
      "epoch": 0.08559163572199836,
      "eval_loss": 1.4939472675323486,
      "eval_runtime": 71.0831,
      "eval_samples_per_second": 14.068,
      "eval_steps_per_second": 14.068,
      "step": 29000
    },
    {
      "epoch": 0.08572331516157067,
      "grad_norm": 0.07488060742616653,
      "learning_rate": 0.0027800681092321785,
      "loss": 1.6085,
      "step": 29020
    },
    {
      "epoch": 0.08585499460114297,
      "grad_norm": 0.07563740015029907,
      "learning_rate": 0.0027797379187572205,
      "loss": 1.6419,
      "step": 29040
    },
    {
      "epoch": 0.08598667404071528,
      "grad_norm": 0.07274390012025833,
      "learning_rate": 0.0027794075002446775,
      "loss": 1.6611,
      "step": 29060
    },
    {
      "epoch": 0.08611835348028758,
      "grad_norm": 0.07751382887363434,
      "learning_rate": 0.0027790768537534283,
      "loss": 1.6802,
      "step": 29080
    },
    {
      "epoch": 0.08625003291985989,
      "grad_norm": 0.08200586587190628,
      "learning_rate": 0.0027787459793423897,
      "loss": 1.6833,
      "step": 29100
    },
    {
      "epoch": 0.08638171235943219,
      "grad_norm": 0.07423494756221771,
      "learning_rate": 0.002778414877070521,
      "loss": 1.6816,
      "step": 29120
    },
    {
      "epoch": 0.08651339179900451,
      "grad_norm": 0.09629736840724945,
      "learning_rate": 0.002778083546996822,
      "loss": 1.6779,
      "step": 29140
    },
    {
      "epoch": 0.08664507123857682,
      "grad_norm": 0.09053540974855423,
      "learning_rate": 0.002777751989180333,
      "loss": 1.6756,
      "step": 29160
    },
    {
      "epoch": 0.08677675067814912,
      "grad_norm": 0.08008238673210144,
      "learning_rate": 0.0027774202036801323,
      "loss": 1.6745,
      "step": 29180
    },
    {
      "epoch": 0.08690843011772142,
      "grad_norm": 0.07499855756759644,
      "learning_rate": 0.0027770881905553432,
      "loss": 1.669,
      "step": 29200
    },
    {
      "epoch": 0.08704010955729373,
      "grad_norm": 0.08784040063619614,
      "learning_rate": 0.002776755949865127,
      "loss": 1.6681,
      "step": 29220
    },
    {
      "epoch": 0.08717178899686603,
      "grad_norm": 0.10526277124881744,
      "learning_rate": 0.0027764234816686864,
      "loss": 1.6645,
      "step": 29240
    },
    {
      "epoch": 0.08730346843643834,
      "grad_norm": 0.08847960829734802,
      "learning_rate": 0.0027760907860252624,
      "loss": 1.659,
      "step": 29260
    },
    {
      "epoch": 0.08743514787601064,
      "grad_norm": 0.09215822070837021,
      "learning_rate": 0.00277575786299414,
      "loss": 1.6612,
      "step": 29280
    },
    {
      "epoch": 0.08756682731558295,
      "grad_norm": 0.08477826416492462,
      "learning_rate": 0.002775424712634642,
      "loss": 1.6656,
      "step": 29300
    },
    {
      "epoch": 0.08769850675515525,
      "grad_norm": 0.07661165297031403,
      "learning_rate": 0.002775091335006134,
      "loss": 1.6627,
      "step": 29320
    },
    {
      "epoch": 0.08783018619472756,
      "grad_norm": 0.06484279781579971,
      "learning_rate": 0.0027747577301680195,
      "loss": 1.6595,
      "step": 29340
    },
    {
      "epoch": 0.08796186563429986,
      "grad_norm": 0.0724259540438652,
      "learning_rate": 0.0027744238981797452,
      "loss": 1.6561,
      "step": 29360
    },
    {
      "epoch": 0.08809354507387217,
      "grad_norm": 0.07707709074020386,
      "learning_rate": 0.0027740898391007958,
      "loss": 1.6538,
      "step": 29380
    },
    {
      "epoch": 0.08822522451344447,
      "grad_norm": 0.08440696448087692,
      "learning_rate": 0.002773755552990697,
      "loss": 1.6546,
      "step": 29400
    },
    {
      "epoch": 0.08835690395301678,
      "grad_norm": 0.07948070019483566,
      "learning_rate": 0.0027734210399090174,
      "loss": 1.6462,
      "step": 29420
    },
    {
      "epoch": 0.08848858339258908,
      "grad_norm": 0.08475562930107117,
      "learning_rate": 0.002773086299915363,
      "loss": 1.6494,
      "step": 29440
    },
    {
      "epoch": 0.08862026283216139,
      "grad_norm": 0.0809844508767128,
      "learning_rate": 0.0027727513330693815,
      "loss": 1.6581,
      "step": 29460
    },
    {
      "epoch": 0.08875194227173369,
      "grad_norm": 0.07043874263763428,
      "learning_rate": 0.002772416139430761,
      "loss": 1.6499,
      "step": 29480
    },
    {
      "epoch": 0.088883621711306,
      "grad_norm": 0.079310841858387,
      "learning_rate": 0.0027720807190592302,
      "loss": 1.6508,
      "step": 29500
    },
    {
      "epoch": 0.0890153011508783,
      "grad_norm": 0.07824884355068207,
      "learning_rate": 0.002771745072014558,
      "loss": 1.6486,
      "step": 29520
    },
    {
      "epoch": 0.0891469805904506,
      "grad_norm": 0.06876828521490097,
      "learning_rate": 0.0027714091983565524,
      "loss": 1.6493,
      "step": 29540
    },
    {
      "epoch": 0.08927866003002291,
      "grad_norm": 0.09782326221466064,
      "learning_rate": 0.0027710730981450643,
      "loss": 1.654,
      "step": 29560
    },
    {
      "epoch": 0.08941033946959522,
      "grad_norm": 0.08521437644958496,
      "learning_rate": 0.0027707367714399837,
      "loss": 1.6469,
      "step": 29580
    },
    {
      "epoch": 0.08954201890916752,
      "grad_norm": 0.07844710350036621,
      "learning_rate": 0.0027704002183012405,
      "loss": 1.6494,
      "step": 29600
    },
    {
      "epoch": 0.08967369834873982,
      "grad_norm": 0.09010956436395645,
      "learning_rate": 0.002770063438788805,
      "loss": 1.6534,
      "step": 29620
    },
    {
      "epoch": 0.08980537778831213,
      "grad_norm": 0.07454434037208557,
      "learning_rate": 0.002769726432962689,
      "loss": 1.668,
      "step": 29640
    },
    {
      "epoch": 0.08993705722788443,
      "grad_norm": 0.09403835237026215,
      "learning_rate": 0.0027693892008829437,
      "loss": 1.6764,
      "step": 29660
    },
    {
      "epoch": 0.09006873666745674,
      "grad_norm": 0.09171565622091293,
      "learning_rate": 0.0027690517426096606,
      "loss": 1.6795,
      "step": 29680
    },
    {
      "epoch": 0.09020041610702904,
      "grad_norm": 0.08323973417282104,
      "learning_rate": 0.0027687140582029713,
      "loss": 1.6813,
      "step": 29700
    },
    {
      "epoch": 0.09033209554660135,
      "grad_norm": 0.06903329491615295,
      "learning_rate": 0.0027683761477230485,
      "loss": 1.6718,
      "step": 29720
    },
    {
      "epoch": 0.09046377498617365,
      "grad_norm": 0.06989486515522003,
      "learning_rate": 0.002768038011230105,
      "loss": 1.6684,
      "step": 29740
    },
    {
      "epoch": 0.09059545442574596,
      "grad_norm": 0.08322389423847198,
      "learning_rate": 0.002767699648784393,
      "loss": 1.6708,
      "step": 29760
    },
    {
      "epoch": 0.09072713386531826,
      "grad_norm": 0.09357618540525436,
      "learning_rate": 0.002767361060446206,
      "loss": 1.6716,
      "step": 29780
    },
    {
      "epoch": 0.09085881330489057,
      "grad_norm": 0.08602355420589447,
      "learning_rate": 0.0027670222462758774,
      "loss": 1.6658,
      "step": 29800
    },
    {
      "epoch": 0.09099049274446289,
      "grad_norm": 0.07419178634881973,
      "learning_rate": 0.0027666832063337806,
      "loss": 1.6682,
      "step": 29820
    },
    {
      "epoch": 0.09112217218403519,
      "grad_norm": 0.09919148683547974,
      "learning_rate": 0.0027663439406803293,
      "loss": 1.6616,
      "step": 29840
    },
    {
      "epoch": 0.0912538516236075,
      "grad_norm": 0.07616308331489563,
      "learning_rate": 0.002766004449375978,
      "loss": 1.661,
      "step": 29860
    },
    {
      "epoch": 0.0913855310631798,
      "grad_norm": 0.0842098742723465,
      "learning_rate": 0.002765664732481221,
      "loss": 1.6573,
      "step": 29880
    },
    {
      "epoch": 0.0915172105027521,
      "grad_norm": 0.09518759697675705,
      "learning_rate": 0.002765324790056592,
      "loss": 1.6638,
      "step": 29900
    },
    {
      "epoch": 0.09164888994232441,
      "grad_norm": 0.07474862039089203,
      "learning_rate": 0.002764984622162666,
      "loss": 1.66,
      "step": 29920
    },
    {
      "epoch": 0.09178056938189671,
      "grad_norm": 0.17661775648593903,
      "learning_rate": 0.0027646442288600578,
      "loss": 1.6541,
      "step": 29940
    },
    {
      "epoch": 0.09191224882146902,
      "grad_norm": 0.1082674190402031,
      "learning_rate": 0.002764303610209422,
      "loss": 1.6583,
      "step": 29960
    },
    {
      "epoch": 0.09204392826104132,
      "grad_norm": 0.0656692162156105,
      "learning_rate": 0.0027639627662714547,
      "loss": 1.6548,
      "step": 29980
    },
    {
      "epoch": 0.09217560770061363,
      "grad_norm": 0.06664387881755829,
      "learning_rate": 0.0027636216971068904,
      "loss": 1.6536,
      "step": 30000
    },
    {
      "epoch": 0.09217560770061363,
      "eval_loss": 1.593057632446289,
      "eval_runtime": 66.8776,
      "eval_samples_per_second": 14.953,
      "eval_steps_per_second": 14.953,
      "step": 30000
    },
    {
      "epoch": 0.09230728714018593,
      "grad_norm": 0.11122573167085648,
      "learning_rate": 0.0027632804027765048,
      "loss": 1.6526,
      "step": 30020
    },
    {
      "epoch": 0.09243896657975824,
      "grad_norm": 0.08378654718399048,
      "learning_rate": 0.0027629388833411133,
      "loss": 1.6516,
      "step": 30040
    },
    {
      "epoch": 0.09257064601933054,
      "grad_norm": 0.0903952419757843,
      "learning_rate": 0.0027625971388615714,
      "loss": 1.6495,
      "step": 30060
    },
    {
      "epoch": 0.09270232545890285,
      "grad_norm": 0.10157206654548645,
      "learning_rate": 0.002762255169398775,
      "loss": 1.6479,
      "step": 30080
    },
    {
      "epoch": 0.09283400489847515,
      "grad_norm": 0.06507019698619843,
      "learning_rate": 0.00276191297501366,
      "loss": 1.651,
      "step": 30100
    },
    {
      "epoch": 0.09296568433804746,
      "grad_norm": 0.07820893824100494,
      "learning_rate": 0.0027615705557672023,
      "loss": 1.6566,
      "step": 30120
    },
    {
      "epoch": 0.09309736377761976,
      "grad_norm": 0.07816014438867569,
      "learning_rate": 0.002761227911720418,
      "loss": 1.6532,
      "step": 30140
    },
    {
      "epoch": 0.09322904321719207,
      "grad_norm": 0.10330802947282791,
      "learning_rate": 0.002760885042934363,
      "loss": 1.6467,
      "step": 30160
    },
    {
      "epoch": 0.09336072265676437,
      "grad_norm": 0.07946458458900452,
      "learning_rate": 0.0027605419494701328,
      "loss": 1.651,
      "step": 30180
    },
    {
      "epoch": 0.09349240209633668,
      "grad_norm": 0.08045420050621033,
      "learning_rate": 0.002760198631388864,
      "loss": 1.6483,
      "step": 30200
    },
    {
      "epoch": 0.09362408153590898,
      "grad_norm": 0.08808773756027222,
      "learning_rate": 0.0027598550887517333,
      "loss": 1.6419,
      "step": 30220
    },
    {
      "epoch": 0.09375576097548129,
      "grad_norm": 0.12588857114315033,
      "learning_rate": 0.002759511321619956,
      "loss": 1.6357,
      "step": 30240
    },
    {
      "epoch": 0.09388744041505359,
      "grad_norm": 0.07971849292516708,
      "learning_rate": 0.002759167330054789,
      "loss": 1.6015,
      "step": 30260
    },
    {
      "epoch": 0.0940191198546259,
      "grad_norm": 0.09091313183307648,
      "learning_rate": 0.002758823114117527,
      "loss": 1.6004,
      "step": 30280
    },
    {
      "epoch": 0.0941507992941982,
      "grad_norm": 0.0841759741306305,
      "learning_rate": 0.0027584786738695083,
      "loss": 1.5936,
      "step": 30300
    },
    {
      "epoch": 0.0942824787337705,
      "grad_norm": 0.09897656738758087,
      "learning_rate": 0.0027581340093721075,
      "loss": 1.5878,
      "step": 30320
    },
    {
      "epoch": 0.09441415817334281,
      "grad_norm": 0.09095700830221176,
      "learning_rate": 0.002757789120686741,
      "loss": 1.5916,
      "step": 30340
    },
    {
      "epoch": 0.09454583761291512,
      "grad_norm": 0.07832010090351105,
      "learning_rate": 0.0027574440078748648,
      "loss": 1.589,
      "step": 30360
    },
    {
      "epoch": 0.09467751705248742,
      "grad_norm": 0.07373721152544022,
      "learning_rate": 0.0027570986709979746,
      "loss": 1.5799,
      "step": 30380
    },
    {
      "epoch": 0.09480919649205972,
      "grad_norm": 0.07081932574510574,
      "learning_rate": 0.002756753110117607,
      "loss": 1.5791,
      "step": 30400
    },
    {
      "epoch": 0.09494087593163203,
      "grad_norm": 0.08223600685596466,
      "learning_rate": 0.0027564073252953373,
      "loss": 1.5766,
      "step": 30420
    },
    {
      "epoch": 0.09507255537120433,
      "grad_norm": 0.08098476380109787,
      "learning_rate": 0.002756061316592781,
      "loss": 1.5818,
      "step": 30440
    },
    {
      "epoch": 0.09520423481077664,
      "grad_norm": 0.07812890410423279,
      "learning_rate": 0.002755715084071594,
      "loss": 1.5761,
      "step": 30460
    },
    {
      "epoch": 0.09533591425034894,
      "grad_norm": 0.07554410398006439,
      "learning_rate": 0.002755368627793472,
      "loss": 1.5781,
      "step": 30480
    },
    {
      "epoch": 0.09546759368992126,
      "grad_norm": 0.08680091798305511,
      "learning_rate": 0.002755021947820149,
      "loss": 1.5784,
      "step": 30500
    },
    {
      "epoch": 0.09559927312949357,
      "grad_norm": 0.08017537742853165,
      "learning_rate": 0.002754675044213402,
      "loss": 1.578,
      "step": 30520
    },
    {
      "epoch": 0.09573095256906587,
      "grad_norm": 0.08503107726573944,
      "learning_rate": 0.002754327917035045,
      "loss": 1.5727,
      "step": 30540
    },
    {
      "epoch": 0.09586263200863818,
      "grad_norm": 0.09204825013875961,
      "learning_rate": 0.002753980566346934,
      "loss": 1.5775,
      "step": 30560
    },
    {
      "epoch": 0.09599431144821048,
      "grad_norm": 0.07686753571033478,
      "learning_rate": 0.0027536329922109616,
      "loss": 1.5747,
      "step": 30580
    },
    {
      "epoch": 0.09612599088778279,
      "grad_norm": 0.08408256620168686,
      "learning_rate": 0.002753285194689064,
      "loss": 1.5722,
      "step": 30600
    },
    {
      "epoch": 0.09625767032735509,
      "grad_norm": 0.09992122650146484,
      "learning_rate": 0.0027529371738432154,
      "loss": 1.5725,
      "step": 30620
    },
    {
      "epoch": 0.0963893497669274,
      "grad_norm": 0.07421014457941055,
      "learning_rate": 0.0027525889297354295,
      "loss": 1.5709,
      "step": 30640
    },
    {
      "epoch": 0.0965210292064997,
      "grad_norm": 0.08402437716722488,
      "learning_rate": 0.0027522404624277606,
      "loss": 1.5711,
      "step": 30660
    },
    {
      "epoch": 0.096652708646072,
      "grad_norm": 0.10464178025722504,
      "learning_rate": 0.0027518917719823017,
      "loss": 1.5724,
      "step": 30680
    },
    {
      "epoch": 0.09678438808564431,
      "grad_norm": 0.09930146485567093,
      "learning_rate": 0.002751542858461187,
      "loss": 1.5768,
      "step": 30700
    },
    {
      "epoch": 0.09691606752521661,
      "grad_norm": 0.08711190521717072,
      "learning_rate": 0.0027511937219265896,
      "loss": 1.5783,
      "step": 30720
    },
    {
      "epoch": 0.09704774696478892,
      "grad_norm": 0.07746979594230652,
      "learning_rate": 0.0027508443624407217,
      "loss": 1.5724,
      "step": 30740
    },
    {
      "epoch": 0.09717942640436122,
      "grad_norm": 0.09594020992517471,
      "learning_rate": 0.0027504947800658366,
      "loss": 1.5755,
      "step": 30760
    },
    {
      "epoch": 0.09731110584393353,
      "grad_norm": 0.0837220624089241,
      "learning_rate": 0.002750144974864227,
      "loss": 1.5692,
      "step": 30780
    },
    {
      "epoch": 0.09744278528350583,
      "grad_norm": 0.07753045111894608,
      "learning_rate": 0.0027497949468982237,
      "loss": 1.5723,
      "step": 30800
    },
    {
      "epoch": 0.09757446472307814,
      "grad_norm": 0.0966426357626915,
      "learning_rate": 0.0027494446962302004,
      "loss": 1.5948,
      "step": 30820
    },
    {
      "epoch": 0.09770614416265044,
      "grad_norm": 0.0724552646279335,
      "learning_rate": 0.002749094222922567,
      "loss": 1.6,
      "step": 30840
    },
    {
      "epoch": 0.09783782360222275,
      "grad_norm": 0.07575785368680954,
      "learning_rate": 0.002748743527037775,
      "loss": 1.6046,
      "step": 30860
    },
    {
      "epoch": 0.09796950304179505,
      "grad_norm": 0.10300685465335846,
      "learning_rate": 0.0027483926086383156,
      "loss": 1.6014,
      "step": 30880
    },
    {
      "epoch": 0.09810118248136736,
      "grad_norm": 0.07341894507408142,
      "learning_rate": 0.0027480414677867188,
      "loss": 1.6002,
      "step": 30900
    },
    {
      "epoch": 0.09823286192093966,
      "grad_norm": 0.09823710471391678,
      "learning_rate": 0.002747690104545555,
      "loss": 1.5955,
      "step": 30920
    },
    {
      "epoch": 0.09836454136051197,
      "grad_norm": 0.08126625418663025,
      "learning_rate": 0.0027473385189774337,
      "loss": 1.5951,
      "step": 30940
    },
    {
      "epoch": 0.09849622080008427,
      "grad_norm": 0.08654072135686874,
      "learning_rate": 0.0027469867111450043,
      "loss": 1.592,
      "step": 30960
    },
    {
      "epoch": 0.09862790023965658,
      "grad_norm": 0.08610561490058899,
      "learning_rate": 0.002746634681110956,
      "loss": 1.5866,
      "step": 30980
    },
    {
      "epoch": 0.09875957967922888,
      "grad_norm": 0.10982340574264526,
      "learning_rate": 0.0027462824289380167,
      "loss": 1.5916,
      "step": 31000
    },
    {
      "epoch": 0.09875957967922888,
      "eval_loss": 1.4958887100219727,
      "eval_runtime": 65.9246,
      "eval_samples_per_second": 15.169,
      "eval_steps_per_second": 15.169,
      "step": 31000
    },
    {
      "epoch": 0.09889125911880119,
      "grad_norm": 0.08077514916658401,
      "learning_rate": 0.002745929954688955,
      "loss": 1.5905,
      "step": 31020
    },
    {
      "epoch": 0.09902293855837349,
      "grad_norm": 0.08327169716358185,
      "learning_rate": 0.0027455772584265782,
      "loss": 1.5853,
      "step": 31040
    },
    {
      "epoch": 0.0991546179979458,
      "grad_norm": 0.07334230840206146,
      "learning_rate": 0.0027452243402137345,
      "loss": 1.5903,
      "step": 31060
    },
    {
      "epoch": 0.0992862974375181,
      "grad_norm": 0.0786149725317955,
      "learning_rate": 0.0027448712001133095,
      "loss": 1.5844,
      "step": 31080
    },
    {
      "epoch": 0.0994179768770904,
      "grad_norm": 0.08628056943416595,
      "learning_rate": 0.00274451783818823,
      "loss": 1.5821,
      "step": 31100
    },
    {
      "epoch": 0.09954965631666271,
      "grad_norm": 0.09965677559375763,
      "learning_rate": 0.0027441642545014623,
      "loss": 1.5788,
      "step": 31120
    },
    {
      "epoch": 0.09968133575623501,
      "grad_norm": 0.09084340184926987,
      "learning_rate": 0.002743810449116011,
      "loss": 1.5821,
      "step": 31140
    },
    {
      "epoch": 0.09981301519580732,
      "grad_norm": 0.08165693283081055,
      "learning_rate": 0.002743456422094922,
      "loss": 1.5779,
      "step": 31160
    },
    {
      "epoch": 0.09994469463537964,
      "grad_norm": 0.08033919334411621,
      "learning_rate": 0.002743102173501278,
      "loss": 1.5799,
      "step": 31180
    },
    {
      "epoch": 0.10007637407495194,
      "grad_norm": 0.09969887882471085,
      "learning_rate": 0.0027427477033982045,
      "loss": 1.577,
      "step": 31200
    },
    {
      "epoch": 0.10020805351452425,
      "grad_norm": 0.08198243379592896,
      "learning_rate": 0.0027423930118488644,
      "loss": 1.5756,
      "step": 31220
    },
    {
      "epoch": 0.10033973295409655,
      "grad_norm": 0.08875197172164917,
      "learning_rate": 0.00274203809891646,
      "loss": 1.5766,
      "step": 31240
    },
    {
      "epoch": 0.10047141239366886,
      "grad_norm": 0.07156415283679962,
      "learning_rate": 0.002741682964664234,
      "loss": 1.5795,
      "step": 31260
    },
    {
      "epoch": 0.10060309183324116,
      "grad_norm": 0.1065966859459877,
      "learning_rate": 0.0027413276091554683,
      "loss": 1.5865,
      "step": 31280
    },
    {
      "epoch": 0.10073477127281347,
      "grad_norm": 0.08343872427940369,
      "learning_rate": 0.0027409720324534834,
      "loss": 1.58,
      "step": 31300
    },
    {
      "epoch": 0.10086645071238577,
      "grad_norm": 0.08024212718009949,
      "learning_rate": 0.00274061623462164,
      "loss": 1.5828,
      "step": 31320
    },
    {
      "epoch": 0.10099813015195808,
      "grad_norm": 0.08566683530807495,
      "learning_rate": 0.0027402602157233376,
      "loss": 1.5695,
      "step": 31340
    },
    {
      "epoch": 0.10112980959153038,
      "grad_norm": 0.08198649436235428,
      "learning_rate": 0.002739903975822016,
      "loss": 1.5811,
      "step": 31360
    },
    {
      "epoch": 0.10126148903110269,
      "grad_norm": 0.10375852882862091,
      "learning_rate": 0.0027395475149811542,
      "loss": 1.5786,
      "step": 31380
    },
    {
      "epoch": 0.10139316847067499,
      "grad_norm": 0.11077110469341278,
      "learning_rate": 0.0027391908332642704,
      "loss": 1.5964,
      "step": 31400
    },
    {
      "epoch": 0.1015248479102473,
      "grad_norm": 0.08183398842811584,
      "learning_rate": 0.0027388339307349205,
      "loss": 1.6251,
      "step": 31420
    },
    {
      "epoch": 0.1016565273498196,
      "grad_norm": 0.08363300561904907,
      "learning_rate": 0.002738476807456703,
      "loss": 1.6606,
      "step": 31440
    },
    {
      "epoch": 0.1017882067893919,
      "grad_norm": 0.07447555661201477,
      "learning_rate": 0.0027381194634932536,
      "loss": 1.6706,
      "step": 31460
    },
    {
      "epoch": 0.10191988622896421,
      "grad_norm": 0.10050380229949951,
      "learning_rate": 0.0027377618989082465,
      "loss": 1.6575,
      "step": 31480
    },
    {
      "epoch": 0.10205156566853651,
      "grad_norm": 0.07040868699550629,
      "learning_rate": 0.002737404113765398,
      "loss": 1.6624,
      "step": 31500
    },
    {
      "epoch": 0.10218324510810882,
      "grad_norm": 0.08749683946371078,
      "learning_rate": 0.002737046108128461,
      "loss": 1.6537,
      "step": 31520
    },
    {
      "epoch": 0.10231492454768112,
      "grad_norm": 0.09317103773355484,
      "learning_rate": 0.00273668788206123,
      "loss": 1.6564,
      "step": 31540
    },
    {
      "epoch": 0.10244660398725343,
      "grad_norm": 0.07827817648649216,
      "learning_rate": 0.0027363294356275367,
      "loss": 1.6535,
      "step": 31560
    },
    {
      "epoch": 0.10257828342682573,
      "grad_norm": 0.08093781024217606,
      "learning_rate": 0.002735970768891253,
      "loss": 1.643,
      "step": 31580
    },
    {
      "epoch": 0.10270996286639804,
      "grad_norm": 0.08157644420862198,
      "learning_rate": 0.00273561188191629,
      "loss": 1.6449,
      "step": 31600
    },
    {
      "epoch": 0.10284164230597034,
      "grad_norm": 0.09422040730714798,
      "learning_rate": 0.0027352527747665986,
      "loss": 1.6457,
      "step": 31620
    },
    {
      "epoch": 0.10297332174554265,
      "grad_norm": 0.08076277375221252,
      "learning_rate": 0.0027348934475061684,
      "loss": 1.641,
      "step": 31640
    },
    {
      "epoch": 0.10310500118511495,
      "grad_norm": 0.09262409806251526,
      "learning_rate": 0.0027345339001990272,
      "loss": 1.6404,
      "step": 31660
    },
    {
      "epoch": 0.10323668062468726,
      "grad_norm": 0.08169291168451309,
      "learning_rate": 0.0027341741329092446,
      "loss": 1.6405,
      "step": 31680
    },
    {
      "epoch": 0.10336836006425956,
      "grad_norm": 0.07819892466068268,
      "learning_rate": 0.002733814145700927,
      "loss": 1.6315,
      "step": 31700
    },
    {
      "epoch": 0.10350003950383187,
      "grad_norm": 0.07450778037309647,
      "learning_rate": 0.00273345393863822,
      "loss": 1.6268,
      "step": 31720
    },
    {
      "epoch": 0.10363171894340417,
      "grad_norm": 0.08458829671144485,
      "learning_rate": 0.0027330935117853106,
      "loss": 1.6344,
      "step": 31740
    },
    {
      "epoch": 0.10376339838297648,
      "grad_norm": 0.07392025738954544,
      "learning_rate": 0.0027327328652064233,
      "loss": 1.6322,
      "step": 31760
    },
    {
      "epoch": 0.10389507782254878,
      "grad_norm": 0.09063839167356491,
      "learning_rate": 0.002732371998965821,
      "loss": 1.636,
      "step": 31780
    },
    {
      "epoch": 0.10402675726212109,
      "grad_norm": 0.09151460975408554,
      "learning_rate": 0.002732010913127808,
      "loss": 1.6305,
      "step": 31800
    },
    {
      "epoch": 0.10415843670169339,
      "grad_norm": 0.07469334453344345,
      "learning_rate": 0.002731649607756726,
      "loss": 1.6291,
      "step": 31820
    },
    {
      "epoch": 0.10429011614126571,
      "grad_norm": 0.07378929853439331,
      "learning_rate": 0.0027312880829169564,
      "loss": 1.6337,
      "step": 31840
    },
    {
      "epoch": 0.10442179558083801,
      "grad_norm": 0.08763482421636581,
      "learning_rate": 0.002730926338672919,
      "loss": 1.6221,
      "step": 31860
    },
    {
      "epoch": 0.10455347502041032,
      "grad_norm": 0.09730922430753708,
      "learning_rate": 0.0027305643750890744,
      "loss": 1.6251,
      "step": 31880
    },
    {
      "epoch": 0.10468515445998262,
      "grad_norm": 0.0719449371099472,
      "learning_rate": 0.0027302021922299205,
      "loss": 1.6316,
      "step": 31900
    },
    {
      "epoch": 0.10481683389955493,
      "grad_norm": 0.08451465517282486,
      "learning_rate": 0.0027298397901599947,
      "loss": 1.6271,
      "step": 31920
    },
    {
      "epoch": 0.10494851333912723,
      "grad_norm": 0.07725101709365845,
      "learning_rate": 0.0027294771689438746,
      "loss": 1.6333,
      "step": 31940
    },
    {
      "epoch": 0.10508019277869954,
      "grad_norm": 0.10085799545049667,
      "learning_rate": 0.0027291143286461753,
      "loss": 1.6231,
      "step": 31960
    },
    {
      "epoch": 0.10521187221827184,
      "grad_norm": 0.10356923937797546,
      "learning_rate": 0.002728751269331552,
      "loss": 1.6268,
      "step": 31980
    },
    {
      "epoch": 0.10534355165784415,
      "grad_norm": 0.0808652937412262,
      "learning_rate": 0.0027283879910646983,
      "loss": 1.6371,
      "step": 32000
    },
    {
      "epoch": 0.10534355165784415,
      "eval_loss": 1.5755228996276855,
      "eval_runtime": 66.1417,
      "eval_samples_per_second": 15.119,
      "eval_steps_per_second": 15.119,
      "step": 32000
    },
    {
      "epoch": 0.10547523109741645,
      "grad_norm": 0.10892391949892044,
      "learning_rate": 0.002728024493910347,
      "loss": 1.6455,
      "step": 32020
    },
    {
      "epoch": 0.10560691053698876,
      "grad_norm": 0.08325547724962234,
      "learning_rate": 0.00272766077793327,
      "loss": 1.6526,
      "step": 32040
    },
    {
      "epoch": 0.10573858997656106,
      "grad_norm": 0.10805673897266388,
      "learning_rate": 0.0027272968431982787,
      "loss": 1.6663,
      "step": 32060
    },
    {
      "epoch": 0.10587026941613337,
      "grad_norm": 0.09780491143465042,
      "learning_rate": 0.0027269326897702223,
      "loss": 1.6522,
      "step": 32080
    },
    {
      "epoch": 0.10600194885570567,
      "grad_norm": 0.09470649063587189,
      "learning_rate": 0.0027265683177139896,
      "loss": 1.6546,
      "step": 32100
    },
    {
      "epoch": 0.10613362829527798,
      "grad_norm": 0.07851683348417282,
      "learning_rate": 0.002726203727094509,
      "loss": 1.6458,
      "step": 32120
    },
    {
      "epoch": 0.10626530773485028,
      "grad_norm": 0.0761096403002739,
      "learning_rate": 0.0027258389179767467,
      "loss": 1.6525,
      "step": 32140
    },
    {
      "epoch": 0.10639698717442259,
      "grad_norm": 0.1286764144897461,
      "learning_rate": 0.0027254738904257084,
      "loss": 1.6486,
      "step": 32160
    },
    {
      "epoch": 0.10652866661399489,
      "grad_norm": 0.0972808375954628,
      "learning_rate": 0.002725108644506438,
      "loss": 1.6394,
      "step": 32180
    },
    {
      "epoch": 0.1066603460535672,
      "grad_norm": 0.07394243031740189,
      "learning_rate": 0.0027247431802840207,
      "loss": 1.6393,
      "step": 32200
    },
    {
      "epoch": 0.1067920254931395,
      "grad_norm": 0.07845206558704376,
      "learning_rate": 0.0027243774978235775,
      "loss": 1.6381,
      "step": 32220
    },
    {
      "epoch": 0.1069237049327118,
      "grad_norm": 0.07898611575365067,
      "learning_rate": 0.00272401159719027,
      "loss": 1.6394,
      "step": 32240
    },
    {
      "epoch": 0.10705538437228411,
      "grad_norm": 0.08906087279319763,
      "learning_rate": 0.002723645478449298,
      "loss": 1.6368,
      "step": 32260
    },
    {
      "epoch": 0.10718706381185641,
      "grad_norm": 0.0694982185959816,
      "learning_rate": 0.0027232791416659016,
      "loss": 1.6367,
      "step": 32280
    },
    {
      "epoch": 0.10731874325142872,
      "grad_norm": 0.07341042160987854,
      "learning_rate": 0.0027229125869053577,
      "loss": 1.6301,
      "step": 32300
    },
    {
      "epoch": 0.10745042269100102,
      "grad_norm": 0.1038580909371376,
      "learning_rate": 0.0027225458142329834,
      "loss": 1.6339,
      "step": 32320
    },
    {
      "epoch": 0.10758210213057333,
      "grad_norm": 0.09001409262418747,
      "learning_rate": 0.002722178823714134,
      "loss": 1.6368,
      "step": 32340
    },
    {
      "epoch": 0.10771378157014563,
      "grad_norm": 0.07618994265794754,
      "learning_rate": 0.0027218116154142037,
      "loss": 1.6209,
      "step": 32360
    },
    {
      "epoch": 0.10784546100971794,
      "grad_norm": 0.08030746132135391,
      "learning_rate": 0.0027214441893986256,
      "loss": 1.6342,
      "step": 32380
    },
    {
      "epoch": 0.10797714044929024,
      "grad_norm": 0.07523296028375626,
      "learning_rate": 0.0027210765457328723,
      "loss": 1.6318,
      "step": 32400
    },
    {
      "epoch": 0.10810881988886255,
      "grad_norm": 0.08324859291315079,
      "learning_rate": 0.0027207086844824543,
      "loss": 1.6366,
      "step": 32420
    },
    {
      "epoch": 0.10824049932843485,
      "grad_norm": 0.09114082902669907,
      "learning_rate": 0.002720340605712921,
      "loss": 1.6323,
      "step": 32440
    },
    {
      "epoch": 0.10837217876800716,
      "grad_norm": 0.14000263810157776,
      "learning_rate": 0.0027199723094898606,
      "loss": 1.6284,
      "step": 32460
    },
    {
      "epoch": 0.10850385820757946,
      "grad_norm": 0.09753448516130447,
      "learning_rate": 0.0027196037958789002,
      "loss": 1.633,
      "step": 32480
    },
    {
      "epoch": 0.10863553764715177,
      "grad_norm": 0.09878184646368027,
      "learning_rate": 0.0027192350649457055,
      "loss": 1.6316,
      "step": 32500
    },
    {
      "epoch": 0.10876721708672409,
      "grad_norm": 0.08344288915395737,
      "learning_rate": 0.002718866116755981,
      "loss": 1.6188,
      "step": 32520
    },
    {
      "epoch": 0.10889889652629639,
      "grad_norm": 0.09065096825361252,
      "learning_rate": 0.00271849695137547,
      "loss": 1.6279,
      "step": 32540
    },
    {
      "epoch": 0.1090305759658687,
      "grad_norm": 0.08262189477682114,
      "learning_rate": 0.002718127568869954,
      "loss": 1.6301,
      "step": 32560
    },
    {
      "epoch": 0.109162255405441,
      "grad_norm": 0.11573157459497452,
      "learning_rate": 0.002717757969305254,
      "loss": 1.6204,
      "step": 32580
    },
    {
      "epoch": 0.1092939348450133,
      "grad_norm": 0.09369558840990067,
      "learning_rate": 0.002717388152747229,
      "loss": 1.6258,
      "step": 32600
    },
    {
      "epoch": 0.10942561428458561,
      "grad_norm": 0.08298981189727783,
      "learning_rate": 0.0027170181192617774,
      "loss": 1.597,
      "step": 32620
    },
    {
      "epoch": 0.10955729372415791,
      "grad_norm": 0.0729466900229454,
      "learning_rate": 0.0027166478689148355,
      "loss": 1.5801,
      "step": 32640
    },
    {
      "epoch": 0.10968897316373022,
      "grad_norm": 0.08472312241792679,
      "learning_rate": 0.002716277401772378,
      "loss": 1.574,
      "step": 32660
    },
    {
      "epoch": 0.10982065260330252,
      "grad_norm": 0.08250225335359573,
      "learning_rate": 0.00271590671790042,
      "loss": 1.5731,
      "step": 32680
    },
    {
      "epoch": 0.10995233204287483,
      "grad_norm": 0.07867790013551712,
      "learning_rate": 0.0027155358173650123,
      "loss": 1.5637,
      "step": 32700
    },
    {
      "epoch": 0.11008401148244713,
      "grad_norm": 0.10871689021587372,
      "learning_rate": 0.002715164700232247,
      "loss": 1.5652,
      "step": 32720
    },
    {
      "epoch": 0.11021569092201944,
      "grad_norm": 0.09244067221879959,
      "learning_rate": 0.002714793366568254,
      "loss": 1.564,
      "step": 32740
    },
    {
      "epoch": 0.11034737036159174,
      "grad_norm": 0.10256188362836838,
      "learning_rate": 0.0027144218164392012,
      "loss": 1.5575,
      "step": 32760
    },
    {
      "epoch": 0.11047904980116405,
      "grad_norm": 0.10790042579174042,
      "learning_rate": 0.0027140500499112956,
      "loss": 1.5585,
      "step": 32780
    },
    {
      "epoch": 0.11061072924073635,
      "grad_norm": 0.07326998561620712,
      "learning_rate": 0.002713678067050782,
      "loss": 1.5602,
      "step": 32800
    },
    {
      "epoch": 0.11074240868030866,
      "grad_norm": 0.12128879874944687,
      "learning_rate": 0.0027133058679239446,
      "loss": 1.5558,
      "step": 32820
    },
    {
      "epoch": 0.11087408811988096,
      "grad_norm": 0.0924782082438469,
      "learning_rate": 0.0027129334525971068,
      "loss": 1.5546,
      "step": 32840
    },
    {
      "epoch": 0.11100576755945327,
      "grad_norm": 0.10736776888370514,
      "learning_rate": 0.002712560821136628,
      "loss": 1.5555,
      "step": 32860
    },
    {
      "epoch": 0.11113744699902557,
      "grad_norm": 0.11395737528800964,
      "learning_rate": 0.0027121879736089097,
      "loss": 1.5535,
      "step": 32880
    },
    {
      "epoch": 0.11126912643859788,
      "grad_norm": 0.08395570516586304,
      "learning_rate": 0.0027118149100803877,
      "loss": 1.5501,
      "step": 32900
    },
    {
      "epoch": 0.11140080587817018,
      "grad_norm": 0.08537106215953827,
      "learning_rate": 0.0027114416306175397,
      "loss": 1.5457,
      "step": 32920
    },
    {
      "epoch": 0.11153248531774249,
      "grad_norm": 0.06968770921230316,
      "learning_rate": 0.002711068135286881,
      "loss": 1.5526,
      "step": 32940
    },
    {
      "epoch": 0.11166416475731479,
      "grad_norm": 0.10123132914304733,
      "learning_rate": 0.0027106944241549647,
      "loss": 1.5524,
      "step": 32960
    },
    {
      "epoch": 0.1117958441968871,
      "grad_norm": 0.11128298938274384,
      "learning_rate": 0.0027103204972883824,
      "loss": 1.5412,
      "step": 32980
    },
    {
      "epoch": 0.1119275236364594,
      "grad_norm": 0.08465181291103363,
      "learning_rate": 0.002709946354753765,
      "loss": 1.5507,
      "step": 33000
    },
    {
      "epoch": 0.1119275236364594,
      "eval_loss": 1.6168590784072876,
      "eval_runtime": 66.027,
      "eval_samples_per_second": 15.145,
      "eval_steps_per_second": 15.145,
      "step": 33000
    },
    {
      "epoch": 0.1120592030760317,
      "grad_norm": 0.09486215561628342,
      "learning_rate": 0.0027095719966177805,
      "loss": 1.5482,
      "step": 33020
    },
    {
      "epoch": 0.11219088251560401,
      "grad_norm": 0.09311071038246155,
      "learning_rate": 0.0027091974229471365,
      "loss": 1.5502,
      "step": 33040
    },
    {
      "epoch": 0.11232256195517631,
      "grad_norm": 0.16168630123138428,
      "learning_rate": 0.002708822633808579,
      "loss": 1.5475,
      "step": 33060
    },
    {
      "epoch": 0.11245424139474862,
      "grad_norm": 0.09388844668865204,
      "learning_rate": 0.002708447629268891,
      "loss": 1.5533,
      "step": 33080
    },
    {
      "epoch": 0.11258592083432092,
      "grad_norm": 0.08351096510887146,
      "learning_rate": 0.002708072409394896,
      "loss": 1.5522,
      "step": 33100
    },
    {
      "epoch": 0.11271760027389323,
      "grad_norm": 0.07432214170694351,
      "learning_rate": 0.0027076969742534543,
      "loss": 1.5471,
      "step": 33120
    },
    {
      "epoch": 0.11284927971346553,
      "grad_norm": 0.09724659472703934,
      "learning_rate": 0.002707321323911465,
      "loss": 1.5502,
      "step": 33140
    },
    {
      "epoch": 0.11298095915303784,
      "grad_norm": 0.08879043906927109,
      "learning_rate": 0.0027069454584358658,
      "loss": 1.5441,
      "step": 33160
    },
    {
      "epoch": 0.11311263859261014,
      "grad_norm": 0.07506344467401505,
      "learning_rate": 0.0027065693778936316,
      "loss": 1.5691,
      "step": 33180
    },
    {
      "epoch": 0.11324431803218246,
      "grad_norm": 0.10346978157758713,
      "learning_rate": 0.0027061930823517777,
      "loss": 1.5975,
      "step": 33200
    },
    {
      "epoch": 0.11337599747175477,
      "grad_norm": 0.08535962551832199,
      "learning_rate": 0.002705816571877355,
      "loss": 1.6033,
      "step": 33220
    },
    {
      "epoch": 0.11350767691132707,
      "grad_norm": 0.10019640624523163,
      "learning_rate": 0.002705439846537456,
      "loss": 1.6107,
      "step": 33240
    },
    {
      "epoch": 0.11363935635089938,
      "grad_norm": 0.08656740933656693,
      "learning_rate": 0.002705062906399209,
      "loss": 1.6062,
      "step": 33260
    },
    {
      "epoch": 0.11377103579047168,
      "grad_norm": 0.10879825055599213,
      "learning_rate": 0.0027046857515297808,
      "loss": 1.6123,
      "step": 33280
    },
    {
      "epoch": 0.11390271523004399,
      "grad_norm": 0.11161535233259201,
      "learning_rate": 0.002704308381996377,
      "loss": 1.6132,
      "step": 33300
    },
    {
      "epoch": 0.11403439466961629,
      "grad_norm": 0.07457931339740753,
      "learning_rate": 0.002703930797866242,
      "loss": 1.6062,
      "step": 33320
    },
    {
      "epoch": 0.1141660741091886,
      "grad_norm": 0.08514375239610672,
      "learning_rate": 0.002703552999206657,
      "loss": 1.6045,
      "step": 33340
    },
    {
      "epoch": 0.1142977535487609,
      "grad_norm": 0.07797901332378387,
      "learning_rate": 0.0027031749860849438,
      "loss": 1.6044,
      "step": 33360
    },
    {
      "epoch": 0.1144294329883332,
      "grad_norm": 0.10032613575458527,
      "learning_rate": 0.0027027967585684597,
      "loss": 1.6016,
      "step": 33380
    },
    {
      "epoch": 0.11456111242790551,
      "grad_norm": 0.07706362754106522,
      "learning_rate": 0.0027024183167246012,
      "loss": 1.5949,
      "step": 33400
    },
    {
      "epoch": 0.11469279186747781,
      "grad_norm": 0.07698775827884674,
      "learning_rate": 0.0027020396606208036,
      "loss": 1.6026,
      "step": 33420
    },
    {
      "epoch": 0.11482447130705012,
      "grad_norm": 0.09712442755699158,
      "learning_rate": 0.00270166079032454,
      "loss": 1.5968,
      "step": 33440
    },
    {
      "epoch": 0.11495615074662242,
      "grad_norm": 0.0798928514122963,
      "learning_rate": 0.002701281705903321,
      "loss": 1.5939,
      "step": 33460
    },
    {
      "epoch": 0.11508783018619473,
      "grad_norm": 0.09378674626350403,
      "learning_rate": 0.002700902407424697,
      "loss": 1.5946,
      "step": 33480
    },
    {
      "epoch": 0.11521950962576703,
      "grad_norm": 0.1083533987402916,
      "learning_rate": 0.0027005228949562544,
      "loss": 1.5982,
      "step": 33500
    },
    {
      "epoch": 0.11535118906533934,
      "grad_norm": 0.09362529218196869,
      "learning_rate": 0.00270014316856562,
      "loss": 1.5938,
      "step": 33520
    },
    {
      "epoch": 0.11548286850491164,
      "grad_norm": 0.10311439633369446,
      "learning_rate": 0.002699763228320457,
      "loss": 1.5909,
      "step": 33540
    },
    {
      "epoch": 0.11561454794448395,
      "grad_norm": 0.07770277559757233,
      "learning_rate": 0.0026993830742884674,
      "loss": 1.5921,
      "step": 33560
    },
    {
      "epoch": 0.11574622738405625,
      "grad_norm": 0.08439300209283829,
      "learning_rate": 0.002699002706537391,
      "loss": 1.5847,
      "step": 33580
    },
    {
      "epoch": 0.11587790682362856,
      "grad_norm": 0.09440808743238449,
      "learning_rate": 0.0026986221251350056,
      "loss": 1.5852,
      "step": 33600
    },
    {
      "epoch": 0.11600958626320086,
      "grad_norm": 0.08001919090747833,
      "learning_rate": 0.0026982413301491277,
      "loss": 1.5938,
      "step": 33620
    },
    {
      "epoch": 0.11614126570277317,
      "grad_norm": 0.11483153700828552,
      "learning_rate": 0.002697860321647611,
      "loss": 1.5927,
      "step": 33640
    },
    {
      "epoch": 0.11627294514234547,
      "grad_norm": 0.10114043205976486,
      "learning_rate": 0.0026974790996983486,
      "loss": 1.5898,
      "step": 33660
    },
    {
      "epoch": 0.11640462458191778,
      "grad_norm": 0.08223793655633926,
      "learning_rate": 0.002697097664369271,
      "loss": 1.5898,
      "step": 33680
    },
    {
      "epoch": 0.11653630402149008,
      "grad_norm": 0.0857810378074646,
      "learning_rate": 0.002696716015728345,
      "loss": 1.5936,
      "step": 33700
    },
    {
      "epoch": 0.11666798346106239,
      "grad_norm": 0.12150619924068451,
      "learning_rate": 0.0026963341538435775,
      "loss": 1.587,
      "step": 33720
    },
    {
      "epoch": 0.11679966290063469,
      "grad_norm": 0.09355217963457108,
      "learning_rate": 0.0026959520787830133,
      "loss": 1.5856,
      "step": 33740
    },
    {
      "epoch": 0.116931342340207,
      "grad_norm": 0.08837936073541641,
      "learning_rate": 0.0026955697906147344,
      "loss": 1.5895,
      "step": 33760
    },
    {
      "epoch": 0.1170630217797793,
      "grad_norm": 0.08591131120920181,
      "learning_rate": 0.002695187289406861,
      "loss": 1.5967,
      "step": 33780
    },
    {
      "epoch": 0.1171947012193516,
      "grad_norm": 0.11289795488119125,
      "learning_rate": 0.002694804575227551,
      "loss": 1.5971,
      "step": 33800
    },
    {
      "epoch": 0.11732638065892391,
      "grad_norm": 0.09001953154802322,
      "learning_rate": 0.002694421648145001,
      "loss": 1.6129,
      "step": 33820
    },
    {
      "epoch": 0.11745806009849621,
      "grad_norm": 0.10741370916366577,
      "learning_rate": 0.002694038508227445,
      "loss": 1.6142,
      "step": 33840
    },
    {
      "epoch": 0.11758973953806853,
      "grad_norm": 0.08802719414234161,
      "learning_rate": 0.0026936551555431553,
      "loss": 1.6164,
      "step": 33860
    },
    {
      "epoch": 0.11772141897764084,
      "grad_norm": 0.10249514877796173,
      "learning_rate": 0.002693271590160441,
      "loss": 1.6068,
      "step": 33880
    },
    {
      "epoch": 0.11785309841721314,
      "grad_norm": 0.12778210639953613,
      "learning_rate": 0.002692887812147651,
      "loss": 1.604,
      "step": 33900
    },
    {
      "epoch": 0.11798477785678545,
      "grad_norm": 0.10118341445922852,
      "learning_rate": 0.0026925038215731703,
      "loss": 1.6052,
      "step": 33920
    },
    {
      "epoch": 0.11811645729635775,
      "grad_norm": 0.12567481398582458,
      "learning_rate": 0.002692119618505423,
      "loss": 1.6026,
      "step": 33940
    },
    {
      "epoch": 0.11824813673593006,
      "grad_norm": 0.0919373631477356,
      "learning_rate": 0.0026917352030128698,
      "loss": 1.5962,
      "step": 33960
    },
    {
      "epoch": 0.11837981617550236,
      "grad_norm": 0.10136484354734421,
      "learning_rate": 0.002691350575164011,
      "loss": 1.5898,
      "step": 33980
    },
    {
      "epoch": 0.11851149561507467,
      "grad_norm": 0.07460084557533264,
      "learning_rate": 0.0026909657350273827,
      "loss": 1.5924,
      "step": 34000
    },
    {
      "epoch": 0.11851149561507467,
      "eval_loss": 1.5006837844848633,
      "eval_runtime": 66.9994,
      "eval_samples_per_second": 14.926,
      "eval_steps_per_second": 14.926,
      "step": 34000
    },
    {
      "epoch": 0.11864317505464697,
      "grad_norm": 0.09306183457374573,
      "learning_rate": 0.0026905806826715605,
      "loss": 1.5871,
      "step": 34020
    },
    {
      "epoch": 0.11877485449421928,
      "grad_norm": 0.08243928104639053,
      "learning_rate": 0.002690195418165158,
      "loss": 1.5946,
      "step": 34040
    },
    {
      "epoch": 0.11890653393379158,
      "grad_norm": 0.08196361362934113,
      "learning_rate": 0.002689809941576824,
      "loss": 1.5878,
      "step": 34060
    },
    {
      "epoch": 0.11903821337336389,
      "grad_norm": 0.09074049443006516,
      "learning_rate": 0.0026894242529752485,
      "loss": 1.5903,
      "step": 34080
    },
    {
      "epoch": 0.11916989281293619,
      "grad_norm": 0.08904928714036942,
      "learning_rate": 0.002689038352429157,
      "loss": 1.5858,
      "step": 34100
    },
    {
      "epoch": 0.1193015722525085,
      "grad_norm": 0.08144096285104752,
      "learning_rate": 0.0026886522400073136,
      "loss": 1.5845,
      "step": 34120
    },
    {
      "epoch": 0.1194332516920808,
      "grad_norm": 0.1431560218334198,
      "learning_rate": 0.00268826591577852,
      "loss": 1.5868,
      "step": 34140
    },
    {
      "epoch": 0.1195649311316531,
      "grad_norm": 0.09951731562614441,
      "learning_rate": 0.002687879379811615,
      "loss": 1.5827,
      "step": 34160
    },
    {
      "epoch": 0.11969661057122541,
      "grad_norm": 0.12058053910732269,
      "learning_rate": 0.0026874926321754766,
      "loss": 1.5822,
      "step": 34180
    },
    {
      "epoch": 0.11982829001079771,
      "grad_norm": 0.10634467750787735,
      "learning_rate": 0.0026871056729390196,
      "loss": 1.5869,
      "step": 34200
    },
    {
      "epoch": 0.11995996945037002,
      "grad_norm": 0.09280095249414444,
      "learning_rate": 0.002686718502171196,
      "loss": 1.5891,
      "step": 34220
    },
    {
      "epoch": 0.12009164888994232,
      "grad_norm": 0.11709116399288177,
      "learning_rate": 0.002686331119940997,
      "loss": 1.5816,
      "step": 34240
    },
    {
      "epoch": 0.12022332832951463,
      "grad_norm": 0.07095829397439957,
      "learning_rate": 0.00268594352631745,
      "loss": 1.5885,
      "step": 34260
    },
    {
      "epoch": 0.12035500776908693,
      "grad_norm": 0.0902029350399971,
      "learning_rate": 0.00268555572136962,
      "loss": 1.5819,
      "step": 34280
    },
    {
      "epoch": 0.12048668720865924,
      "grad_norm": 0.08400021493434906,
      "learning_rate": 0.0026851677051666113,
      "loss": 1.5836,
      "step": 34300
    },
    {
      "epoch": 0.12061836664823154,
      "grad_norm": 0.07917217165231705,
      "learning_rate": 0.002684779477777565,
      "loss": 1.5784,
      "step": 34320
    },
    {
      "epoch": 0.12075004608780385,
      "grad_norm": 0.10317684710025787,
      "learning_rate": 0.002684391039271659,
      "loss": 1.5754,
      "step": 34340
    },
    {
      "epoch": 0.12088172552737615,
      "grad_norm": 0.08167439699172974,
      "learning_rate": 0.0026840023897181098,
      "loss": 1.586,
      "step": 34360
    },
    {
      "epoch": 0.12101340496694846,
      "grad_norm": 0.0829315185546875,
      "learning_rate": 0.00268361352918617,
      "loss": 1.6143,
      "step": 34380
    },
    {
      "epoch": 0.12114508440652076,
      "grad_norm": 0.08923465013504028,
      "learning_rate": 0.0026832244577451336,
      "loss": 1.6303,
      "step": 34400
    },
    {
      "epoch": 0.12127676384609307,
      "grad_norm": 0.0999564528465271,
      "learning_rate": 0.002682835175464327,
      "loss": 1.632,
      "step": 34420
    },
    {
      "epoch": 0.12140844328566537,
      "grad_norm": 0.08035658299922943,
      "learning_rate": 0.002682445682413118,
      "loss": 1.6289,
      "step": 34440
    },
    {
      "epoch": 0.12154012272523768,
      "grad_norm": 0.06893916428089142,
      "learning_rate": 0.0026820559786609106,
      "loss": 1.6304,
      "step": 34460
    },
    {
      "epoch": 0.12167180216480998,
      "grad_norm": 0.10471726953983307,
      "learning_rate": 0.0026816660642771466,
      "loss": 1.6257,
      "step": 34480
    },
    {
      "epoch": 0.12180348160438229,
      "grad_norm": 0.06977444887161255,
      "learning_rate": 0.0026812759393313045,
      "loss": 1.6235,
      "step": 34500
    },
    {
      "epoch": 0.12193516104395459,
      "grad_norm": 0.0887407585978508,
      "learning_rate": 0.0026808856038929014,
      "loss": 1.6233,
      "step": 34520
    },
    {
      "epoch": 0.12206684048352691,
      "grad_norm": 0.09635386615991592,
      "learning_rate": 0.0026804950580314917,
      "loss": 1.6219,
      "step": 34540
    },
    {
      "epoch": 0.12219851992309921,
      "grad_norm": 0.0804891362786293,
      "learning_rate": 0.0026801043018166667,
      "loss": 1.619,
      "step": 34560
    },
    {
      "epoch": 0.12233019936267152,
      "grad_norm": 0.08388678729534149,
      "learning_rate": 0.0026797133353180557,
      "loss": 1.6072,
      "step": 34580
    },
    {
      "epoch": 0.12246187880224382,
      "grad_norm": 0.1082354336977005,
      "learning_rate": 0.0026793221586053255,
      "loss": 1.6141,
      "step": 34600
    },
    {
      "epoch": 0.12259355824181613,
      "grad_norm": 0.08664815127849579,
      "learning_rate": 0.0026789307717481805,
      "loss": 1.6101,
      "step": 34620
    },
    {
      "epoch": 0.12272523768138843,
      "grad_norm": 0.09746722877025604,
      "learning_rate": 0.002678539174816361,
      "loss": 1.6086,
      "step": 34640
    },
    {
      "epoch": 0.12285691712096074,
      "grad_norm": 0.08757676929235458,
      "learning_rate": 0.0026781473678796474,
      "loss": 1.6021,
      "step": 34660
    },
    {
      "epoch": 0.12298859656053304,
      "grad_norm": 0.10005202144384384,
      "learning_rate": 0.0026777553510078556,
      "loss": 1.6083,
      "step": 34680
    },
    {
      "epoch": 0.12312027600010535,
      "grad_norm": 0.0922529324889183,
      "learning_rate": 0.0026773631242708396,
      "loss": 1.6052,
      "step": 34700
    },
    {
      "epoch": 0.12325195543967765,
      "grad_norm": 0.0967714935541153,
      "learning_rate": 0.00267697068773849,
      "loss": 1.6057,
      "step": 34720
    },
    {
      "epoch": 0.12338363487924996,
      "grad_norm": 0.07341698557138443,
      "learning_rate": 0.002676578041480736,
      "loss": 1.6049,
      "step": 34740
    },
    {
      "epoch": 0.12351531431882226,
      "grad_norm": 0.09357757121324539,
      "learning_rate": 0.0026761851855675433,
      "loss": 1.6074,
      "step": 34760
    },
    {
      "epoch": 0.12364699375839457,
      "grad_norm": 0.07980691641569138,
      "learning_rate": 0.0026757921200689155,
      "loss": 1.6043,
      "step": 34780
    },
    {
      "epoch": 0.12377867319796687,
      "grad_norm": 0.1163335070014,
      "learning_rate": 0.002675398845054893,
      "loss": 1.6085,
      "step": 34800
    },
    {
      "epoch": 0.12391035263753918,
      "grad_norm": 0.08081631362438202,
      "learning_rate": 0.0026750053605955537,
      "loss": 1.6018,
      "step": 34820
    },
    {
      "epoch": 0.12404203207711148,
      "grad_norm": 0.0792580246925354,
      "learning_rate": 0.002674611666761013,
      "loss": 1.6092,
      "step": 34840
    },
    {
      "epoch": 0.12417371151668379,
      "grad_norm": 0.0822744071483612,
      "learning_rate": 0.002674217763621424,
      "loss": 1.6052,
      "step": 34860
    },
    {
      "epoch": 0.12430539095625609,
      "grad_norm": 0.10850385576486588,
      "learning_rate": 0.002673823651246976,
      "loss": 1.6001,
      "step": 34880
    },
    {
      "epoch": 0.1244370703958284,
      "grad_norm": 0.10312029719352722,
      "learning_rate": 0.002673429329707897,
      "loss": 1.599,
      "step": 34900
    },
    {
      "epoch": 0.1245687498354007,
      "grad_norm": 0.09553300589323044,
      "learning_rate": 0.0026730347990744506,
      "loss": 1.6001,
      "step": 34920
    },
    {
      "epoch": 0.124700429274973,
      "grad_norm": 0.07871738821268082,
      "learning_rate": 0.0026726400594169392,
      "loss": 1.5985,
      "step": 34940
    },
    {
      "epoch": 0.12483210871454531,
      "grad_norm": 0.1027330532670021,
      "learning_rate": 0.0026722451108057016,
      "loss": 1.5977,
      "step": 34960
    },
    {
      "epoch": 0.12496378815411761,
      "grad_norm": 0.08380601555109024,
      "learning_rate": 0.0026718499533111137,
      "loss": 1.6039,
      "step": 34980
    },
    {
      "epoch": 0.12509546759368992,
      "grad_norm": 0.09733425080776215,
      "learning_rate": 0.002671454587003589,
      "loss": 1.6059,
      "step": 35000
    },
    {
      "epoch": 0.12509546759368992,
      "eval_loss": 1.5703957080841064,
      "eval_runtime": 66.1504,
      "eval_samples_per_second": 15.117,
      "eval_steps_per_second": 15.117,
      "step": 35000
    },
    {
      "epoch": 0.12522714703326224,
      "grad_norm": 0.08837222307920456,
      "learning_rate": 0.0026710590119535787,
      "loss": 1.5829,
      "step": 35020
    },
    {
      "epoch": 0.12535882647283453,
      "grad_norm": 0.09444919973611832,
      "learning_rate": 0.0026706632282315706,
      "loss": 1.5766,
      "step": 35040
    },
    {
      "epoch": 0.12549050591240685,
      "grad_norm": 0.09045890718698502,
      "learning_rate": 0.0026702672359080887,
      "loss": 1.5706,
      "step": 35060
    },
    {
      "epoch": 0.12562218535197914,
      "grad_norm": 0.10616089403629303,
      "learning_rate": 0.002669871035053697,
      "loss": 1.5704,
      "step": 35080
    },
    {
      "epoch": 0.12575386479155146,
      "grad_norm": 0.08494651317596436,
      "learning_rate": 0.002669474625738993,
      "loss": 1.5693,
      "step": 35100
    },
    {
      "epoch": 0.12588554423112375,
      "grad_norm": 0.09198006242513657,
      "learning_rate": 0.0026690780080346143,
      "loss": 1.5588,
      "step": 35120
    },
    {
      "epoch": 0.12601722367069607,
      "grad_norm": 0.07722193002700806,
      "learning_rate": 0.0026686811820112346,
      "loss": 1.5653,
      "step": 35140
    },
    {
      "epoch": 0.12614890311026836,
      "grad_norm": 0.08537338674068451,
      "learning_rate": 0.002668284147739564,
      "loss": 1.5594,
      "step": 35160
    },
    {
      "epoch": 0.12628058254984068,
      "grad_norm": 0.10831312835216522,
      "learning_rate": 0.002667886905290351,
      "loss": 1.5523,
      "step": 35180
    },
    {
      "epoch": 0.12641226198941297,
      "grad_norm": 0.0943257287144661,
      "learning_rate": 0.00266748945473438,
      "loss": 1.553,
      "step": 35200
    },
    {
      "epoch": 0.12654394142898528,
      "grad_norm": 0.1066376119852066,
      "learning_rate": 0.0026670917961424735,
      "loss": 1.5569,
      "step": 35220
    },
    {
      "epoch": 0.12667562086855758,
      "grad_norm": 0.09195302426815033,
      "learning_rate": 0.0026666939295854905,
      "loss": 1.5524,
      "step": 35240
    },
    {
      "epoch": 0.1268073003081299,
      "grad_norm": 0.09674491733312607,
      "learning_rate": 0.002666295855134327,
      "loss": 1.5522,
      "step": 35260
    },
    {
      "epoch": 0.12693897974770219,
      "grad_norm": 0.12088479846715927,
      "learning_rate": 0.0026658975728599166,
      "loss": 1.5529,
      "step": 35280
    },
    {
      "epoch": 0.1270706591872745,
      "grad_norm": 0.09849090874195099,
      "learning_rate": 0.0026654990828332294,
      "loss": 1.5502,
      "step": 35300
    },
    {
      "epoch": 0.1272023386268468,
      "grad_norm": 0.08557942509651184,
      "learning_rate": 0.0026651003851252725,
      "loss": 1.5469,
      "step": 35320
    },
    {
      "epoch": 0.1273340180664191,
      "grad_norm": 0.0989651307463646,
      "learning_rate": 0.00266470147980709,
      "loss": 1.5467,
      "step": 35340
    },
    {
      "epoch": 0.1274656975059914,
      "grad_norm": 0.08966153860092163,
      "learning_rate": 0.002664302366949764,
      "loss": 1.5445,
      "step": 35360
    },
    {
      "epoch": 0.12759737694556372,
      "grad_norm": 0.0903390496969223,
      "learning_rate": 0.002663903046624412,
      "loss": 1.5489,
      "step": 35380
    },
    {
      "epoch": 0.12772905638513601,
      "grad_norm": 0.0847504660487175,
      "learning_rate": 0.0026635035189021892,
      "loss": 1.5421,
      "step": 35400
    },
    {
      "epoch": 0.12786073582470833,
      "grad_norm": 0.11433856189250946,
      "learning_rate": 0.0026631037838542883,
      "loss": 1.5404,
      "step": 35420
    },
    {
      "epoch": 0.12799241526428062,
      "grad_norm": 0.09573832899332047,
      "learning_rate": 0.0026627038415519386,
      "loss": 1.5505,
      "step": 35440
    },
    {
      "epoch": 0.12812409470385294,
      "grad_norm": 0.10996285080909729,
      "learning_rate": 0.002662303692066405,
      "loss": 1.5422,
      "step": 35460
    },
    {
      "epoch": 0.12825577414342523,
      "grad_norm": 0.13538189232349396,
      "learning_rate": 0.0026619033354689917,
      "loss": 1.5431,
      "step": 35480
    },
    {
      "epoch": 0.12838745358299755,
      "grad_norm": 0.09150131791830063,
      "learning_rate": 0.0026615027718310375,
      "loss": 1.5489,
      "step": 35500
    },
    {
      "epoch": 0.12851913302256984,
      "grad_norm": 0.08535768836736679,
      "learning_rate": 0.00266110200122392,
      "loss": 1.5474,
      "step": 35520
    },
    {
      "epoch": 0.12865081246214216,
      "grad_norm": 0.0952446460723877,
      "learning_rate": 0.002660701023719053,
      "loss": 1.5451,
      "step": 35540
    },
    {
      "epoch": 0.12878249190171448,
      "grad_norm": 0.1442309468984604,
      "learning_rate": 0.0026602998393878865,
      "loss": 1.5548,
      "step": 35560
    },
    {
      "epoch": 0.12891417134128677,
      "grad_norm": 0.08960012346506119,
      "learning_rate": 0.0026598984483019077,
      "loss": 1.5595,
      "step": 35580
    },
    {
      "epoch": 0.1290458507808591,
      "grad_norm": 0.09016614407300949,
      "learning_rate": 0.002659496850532641,
      "loss": 1.5564,
      "step": 35600
    },
    {
      "epoch": 0.12917753022043138,
      "grad_norm": 0.11049813777208328,
      "learning_rate": 0.0026590950461516482,
      "loss": 1.5583,
      "step": 35620
    },
    {
      "epoch": 0.1293092096600037,
      "grad_norm": 0.11803802847862244,
      "learning_rate": 0.002658693035230526,
      "loss": 1.5528,
      "step": 35640
    },
    {
      "epoch": 0.129440889099576,
      "grad_norm": 0.12277698516845703,
      "learning_rate": 0.00265829081784091,
      "loss": 1.5452,
      "step": 35660
    },
    {
      "epoch": 0.1295725685391483,
      "grad_norm": 0.08969718217849731,
      "learning_rate": 0.002657888394054471,
      "loss": 1.5436,
      "step": 35680
    },
    {
      "epoch": 0.1297042479787206,
      "grad_norm": 0.07710202783346176,
      "learning_rate": 0.002657485763942918,
      "loss": 1.5429,
      "step": 35700
    },
    {
      "epoch": 0.12983592741829292,
      "grad_norm": 0.10966431349515915,
      "learning_rate": 0.0026570829275779947,
      "loss": 1.5381,
      "step": 35720
    },
    {
      "epoch": 0.1299676068578652,
      "grad_norm": 0.09629274159669876,
      "learning_rate": 0.002656679885031484,
      "loss": 1.5386,
      "step": 35740
    },
    {
      "epoch": 0.13009928629743753,
      "grad_norm": 0.11513884365558624,
      "learning_rate": 0.0026562766363752042,
      "loss": 1.5343,
      "step": 35760
    },
    {
      "epoch": 0.13023096573700982,
      "grad_norm": 0.10110338777303696,
      "learning_rate": 0.0026558731816810105,
      "loss": 1.5315,
      "step": 35780
    },
    {
      "epoch": 0.13036264517658214,
      "grad_norm": 0.08176104724407196,
      "learning_rate": 0.0026554695210207944,
      "loss": 1.5365,
      "step": 35800
    },
    {
      "epoch": 0.13049432461615443,
      "grad_norm": 0.09711804240942001,
      "learning_rate": 0.0026550656544664848,
      "loss": 1.5305,
      "step": 35820
    },
    {
      "epoch": 0.13062600405572675,
      "grad_norm": 0.09326780587434769,
      "learning_rate": 0.0026546615820900467,
      "loss": 1.5236,
      "step": 35840
    },
    {
      "epoch": 0.13075768349529904,
      "grad_norm": 0.10439273715019226,
      "learning_rate": 0.002654257303963483,
      "loss": 1.5217,
      "step": 35860
    },
    {
      "epoch": 0.13088936293487136,
      "grad_norm": 0.09977664798498154,
      "learning_rate": 0.002653852820158832,
      "loss": 1.5241,
      "step": 35880
    },
    {
      "epoch": 0.13102104237444365,
      "grad_norm": 0.08770251274108887,
      "learning_rate": 0.0026534481307481685,
      "loss": 1.5319,
      "step": 35900
    },
    {
      "epoch": 0.13115272181401597,
      "grad_norm": 0.10881366580724716,
      "learning_rate": 0.002653043235803605,
      "loss": 1.5383,
      "step": 35920
    },
    {
      "epoch": 0.13128440125358826,
      "grad_norm": 0.09104442596435547,
      "learning_rate": 0.0026526381353972894,
      "loss": 1.5326,
      "step": 35940
    },
    {
      "epoch": 0.13141608069316058,
      "grad_norm": 0.12487766891717911,
      "learning_rate": 0.0026522328296014076,
      "loss": 1.5339,
      "step": 35960
    },
    {
      "epoch": 0.13154776013273287,
      "grad_norm": 0.15558366477489471,
      "learning_rate": 0.0026518273184881814,
      "loss": 1.5229,
      "step": 35980
    },
    {
      "epoch": 0.13167943957230518,
      "grad_norm": 0.1283443719148636,
      "learning_rate": 0.0026514216021298687,
      "loss": 1.5299,
      "step": 36000
    },
    {
      "epoch": 0.13167943957230518,
      "eval_loss": 1.617323875427246,
      "eval_runtime": 67.0387,
      "eval_samples_per_second": 14.917,
      "eval_steps_per_second": 14.917,
      "step": 36000
    },
    {
      "epoch": 0.13181111901187748,
      "grad_norm": 0.14764146506786346,
      "learning_rate": 0.0026510156805987645,
      "loss": 1.5301,
      "step": 36020
    },
    {
      "epoch": 0.1319427984514498,
      "grad_norm": 0.09800808131694794,
      "learning_rate": 0.0026506095539672006,
      "loss": 1.5251,
      "step": 36040
    },
    {
      "epoch": 0.13207447789102209,
      "grad_norm": 0.09196116775274277,
      "learning_rate": 0.0026502032223075446,
      "loss": 1.5279,
      "step": 36060
    },
    {
      "epoch": 0.1322061573305944,
      "grad_norm": 0.10031158477067947,
      "learning_rate": 0.0026497966856922016,
      "loss": 1.5282,
      "step": 36080
    },
    {
      "epoch": 0.1323378367701667,
      "grad_norm": 0.12328962981700897,
      "learning_rate": 0.0026493899441936124,
      "loss": 1.5328,
      "step": 36100
    },
    {
      "epoch": 0.132469516209739,
      "grad_norm": 0.1079377755522728,
      "learning_rate": 0.0026489829978842546,
      "loss": 1.5307,
      "step": 36120
    },
    {
      "epoch": 0.1326011956493113,
      "grad_norm": 0.08541183918714523,
      "learning_rate": 0.0026485758468366427,
      "loss": 1.5356,
      "step": 36140
    },
    {
      "epoch": 0.13273287508888362,
      "grad_norm": 0.10611236095428467,
      "learning_rate": 0.0026481684911233267,
      "loss": 1.5834,
      "step": 36160
    },
    {
      "epoch": 0.1328645545284559,
      "grad_norm": 0.087572380900383,
      "learning_rate": 0.0026477609308168944,
      "loss": 1.6016,
      "step": 36180
    },
    {
      "epoch": 0.13299623396802823,
      "grad_norm": 0.1034546047449112,
      "learning_rate": 0.002647353165989968,
      "loss": 1.6119,
      "step": 36200
    },
    {
      "epoch": 0.13312791340760055,
      "grad_norm": 0.09647917747497559,
      "learning_rate": 0.002646945196715209,
      "loss": 1.6142,
      "step": 36220
    },
    {
      "epoch": 0.13325959284717284,
      "grad_norm": 0.08855011314153671,
      "learning_rate": 0.002646537023065313,
      "loss": 1.6148,
      "step": 36240
    },
    {
      "epoch": 0.13339127228674516,
      "grad_norm": 0.10243046283721924,
      "learning_rate": 0.0026461286451130127,
      "loss": 1.6155,
      "step": 36260
    },
    {
      "epoch": 0.13352295172631745,
      "grad_norm": 0.105488620698452,
      "learning_rate": 0.0026457200629310777,
      "loss": 1.614,
      "step": 36280
    },
    {
      "epoch": 0.13365463116588977,
      "grad_norm": 0.08305686712265015,
      "learning_rate": 0.0026453112765923133,
      "loss": 1.6059,
      "step": 36300
    },
    {
      "epoch": 0.13378631060546206,
      "grad_norm": 0.09603402763605118,
      "learning_rate": 0.0026449022861695618,
      "loss": 1.6018,
      "step": 36320
    },
    {
      "epoch": 0.13391799004503438,
      "grad_norm": 0.11666828393936157,
      "learning_rate": 0.0026444930917357014,
      "loss": 1.6017,
      "step": 36340
    },
    {
      "epoch": 0.13404966948460667,
      "grad_norm": 0.1235344409942627,
      "learning_rate": 0.0026440836933636466,
      "loss": 1.5965,
      "step": 36360
    },
    {
      "epoch": 0.134181348924179,
      "grad_norm": 0.11957265436649323,
      "learning_rate": 0.002643674091126349,
      "loss": 1.5957,
      "step": 36380
    },
    {
      "epoch": 0.13431302836375128,
      "grad_norm": 0.11756687611341476,
      "learning_rate": 0.0026432642850967953,
      "loss": 1.5999,
      "step": 36400
    },
    {
      "epoch": 0.1344447078033236,
      "grad_norm": 0.10342928767204285,
      "learning_rate": 0.002642854275348009,
      "loss": 1.592,
      "step": 36420
    },
    {
      "epoch": 0.1345763872428959,
      "grad_norm": 0.10075769573450089,
      "learning_rate": 0.002642444061953051,
      "loss": 1.5932,
      "step": 36440
    },
    {
      "epoch": 0.1347080666824682,
      "grad_norm": 0.10493767261505127,
      "learning_rate": 0.002642033644985017,
      "loss": 1.5928,
      "step": 36460
    },
    {
      "epoch": 0.1348397461220405,
      "grad_norm": 0.1051807627081871,
      "learning_rate": 0.00264162302451704,
      "loss": 1.5842,
      "step": 36480
    },
    {
      "epoch": 0.13497142556161282,
      "grad_norm": 0.10073640197515488,
      "learning_rate": 0.0026412122006222883,
      "loss": 1.5831,
      "step": 36500
    },
    {
      "epoch": 0.1351031050011851,
      "grad_norm": 0.10793869942426682,
      "learning_rate": 0.0026408011733739674,
      "loss": 1.5888,
      "step": 36520
    },
    {
      "epoch": 0.13523478444075743,
      "grad_norm": 0.08608701080083847,
      "learning_rate": 0.002640389942845318,
      "loss": 1.588,
      "step": 36540
    },
    {
      "epoch": 0.13536646388032972,
      "grad_norm": 0.08790282160043716,
      "learning_rate": 0.0026399785091096176,
      "loss": 1.5869,
      "step": 36560
    },
    {
      "epoch": 0.13549814331990204,
      "grad_norm": 0.08713233470916748,
      "learning_rate": 0.002639566872240181,
      "loss": 1.5829,
      "step": 36580
    },
    {
      "epoch": 0.13562982275947433,
      "grad_norm": 0.08939250558614731,
      "learning_rate": 0.0026391550323103573,
      "loss": 1.582,
      "step": 36600
    },
    {
      "epoch": 0.13576150219904665,
      "grad_norm": 0.09528082609176636,
      "learning_rate": 0.0026387429893935327,
      "loss": 1.5845,
      "step": 36620
    },
    {
      "epoch": 0.13589318163861894,
      "grad_norm": 0.12500585615634918,
      "learning_rate": 0.0026383307435631294,
      "loss": 1.5823,
      "step": 36640
    },
    {
      "epoch": 0.13602486107819126,
      "grad_norm": 0.0984441414475441,
      "learning_rate": 0.0026379182948926055,
      "loss": 1.5762,
      "step": 36660
    },
    {
      "epoch": 0.13615654051776355,
      "grad_norm": 0.11194117367267609,
      "learning_rate": 0.0026375056434554565,
      "loss": 1.5809,
      "step": 36680
    },
    {
      "epoch": 0.13628821995733587,
      "grad_norm": 0.11597609519958496,
      "learning_rate": 0.002637092789325212,
      "loss": 1.5774,
      "step": 36700
    },
    {
      "epoch": 0.13641989939690816,
      "grad_norm": 0.07991953939199448,
      "learning_rate": 0.0026366797325754403,
      "loss": 1.581,
      "step": 36720
    },
    {
      "epoch": 0.13655157883648048,
      "grad_norm": 0.10064183175563812,
      "learning_rate": 0.0026362664732797426,
      "loss": 1.5853,
      "step": 36740
    },
    {
      "epoch": 0.13668325827605277,
      "grad_norm": 0.1042194738984108,
      "learning_rate": 0.0026358530115117595,
      "loss": 1.5975,
      "step": 36760
    },
    {
      "epoch": 0.13681493771562508,
      "grad_norm": 0.10666348040103912,
      "learning_rate": 0.002635439347345165,
      "loss": 1.6029,
      "step": 36780
    },
    {
      "epoch": 0.13694661715519738,
      "grad_norm": 0.1007642075419426,
      "learning_rate": 0.0026350254808536707,
      "loss": 1.6028,
      "step": 36800
    },
    {
      "epoch": 0.1370782965947697,
      "grad_norm": 0.08334644138813019,
      "learning_rate": 0.002634611412111024,
      "loss": 1.6097,
      "step": 36820
    },
    {
      "epoch": 0.13720997603434198,
      "grad_norm": 0.09047345817089081,
      "learning_rate": 0.0026341971411910077,
      "loss": 1.6055,
      "step": 36840
    },
    {
      "epoch": 0.1373416554739143,
      "grad_norm": 0.16159166395664215,
      "learning_rate": 0.002633782668167442,
      "loss": 1.5987,
      "step": 36860
    },
    {
      "epoch": 0.1374733349134866,
      "grad_norm": 0.10918816179037094,
      "learning_rate": 0.002633367993114181,
      "loss": 1.604,
      "step": 36880
    },
    {
      "epoch": 0.1376050143530589,
      "grad_norm": 0.10157617926597595,
      "learning_rate": 0.002632953116105116,
      "loss": 1.5927,
      "step": 36900
    },
    {
      "epoch": 0.13773669379263123,
      "grad_norm": 0.09241142123937607,
      "learning_rate": 0.0026325380372141756,
      "loss": 1.5981,
      "step": 36920
    },
    {
      "epoch": 0.13786837323220352,
      "grad_norm": 0.12108739465475082,
      "learning_rate": 0.002632122756515322,
      "loss": 1.592,
      "step": 36940
    },
    {
      "epoch": 0.13800005267177584,
      "grad_norm": 0.09442280977964401,
      "learning_rate": 0.0026317072740825545,
      "loss": 1.5884,
      "step": 36960
    },
    {
      "epoch": 0.13813173211134813,
      "grad_norm": 0.09058595448732376,
      "learning_rate": 0.0026312915899899085,
      "loss": 1.5859,
      "step": 36980
    },
    {
      "epoch": 0.13826341155092045,
      "grad_norm": 0.08916100859642029,
      "learning_rate": 0.002630875704311455,
      "loss": 1.5841,
      "step": 37000
    },
    {
      "epoch": 0.13826341155092045,
      "eval_loss": 1.493425726890564,
      "eval_runtime": 67.4059,
      "eval_samples_per_second": 14.836,
      "eval_steps_per_second": 14.836,
      "step": 37000
    },
    {
      "epoch": 0.13839509099049274,
      "grad_norm": 0.08742368966341019,
      "learning_rate": 0.0026304596171213015,
      "loss": 1.588,
      "step": 37020
    },
    {
      "epoch": 0.13852677043006506,
      "grad_norm": 0.07880369573831558,
      "learning_rate": 0.0026300433284935907,
      "loss": 1.5861,
      "step": 37040
    },
    {
      "epoch": 0.13865844986963735,
      "grad_norm": 0.08849447965621948,
      "learning_rate": 0.0026296268385025004,
      "loss": 1.5832,
      "step": 37060
    },
    {
      "epoch": 0.13879012930920967,
      "grad_norm": 0.08267071843147278,
      "learning_rate": 0.002629210147222247,
      "loss": 1.585,
      "step": 37080
    },
    {
      "epoch": 0.13892180874878196,
      "grad_norm": 0.08915061503648758,
      "learning_rate": 0.0026287932547270796,
      "loss": 1.5886,
      "step": 37100
    },
    {
      "epoch": 0.13905348818835428,
      "grad_norm": 0.11180838197469711,
      "learning_rate": 0.0026283761610912853,
      "loss": 1.5786,
      "step": 37120
    },
    {
      "epoch": 0.13918516762792657,
      "grad_norm": 0.14976435899734497,
      "learning_rate": 0.002627958866389186,
      "loss": 1.5809,
      "step": 37140
    },
    {
      "epoch": 0.1393168470674989,
      "grad_norm": 0.09413915127515793,
      "learning_rate": 0.002627541370695141,
      "loss": 1.5791,
      "step": 37160
    },
    {
      "epoch": 0.13944852650707118,
      "grad_norm": 0.10926341265439987,
      "learning_rate": 0.0026271236740835428,
      "loss": 1.5841,
      "step": 37180
    },
    {
      "epoch": 0.1395802059466435,
      "grad_norm": 0.09298263490200043,
      "learning_rate": 0.002626705776628822,
      "loss": 1.587,
      "step": 37200
    },
    {
      "epoch": 0.1397118853862158,
      "grad_norm": 0.1033315360546112,
      "learning_rate": 0.002626287678405443,
      "loss": 1.5787,
      "step": 37220
    },
    {
      "epoch": 0.1398435648257881,
      "grad_norm": 0.10745333135128021,
      "learning_rate": 0.002625869379487908,
      "loss": 1.5761,
      "step": 37240
    },
    {
      "epoch": 0.1399752442653604,
      "grad_norm": 0.11225748062133789,
      "learning_rate": 0.0026254508799507543,
      "loss": 1.5859,
      "step": 37260
    },
    {
      "epoch": 0.14010692370493272,
      "grad_norm": 0.09762544184923172,
      "learning_rate": 0.002625032179868554,
      "loss": 1.5775,
      "step": 37280
    },
    {
      "epoch": 0.140238603144505,
      "grad_norm": 0.0946531891822815,
      "learning_rate": 0.002624613279315916,
      "loss": 1.5794,
      "step": 37300
    },
    {
      "epoch": 0.14037028258407733,
      "grad_norm": 0.07943738996982574,
      "learning_rate": 0.0026241941783674842,
      "loss": 1.5804,
      "step": 37320
    },
    {
      "epoch": 0.14050196202364962,
      "grad_norm": 0.09314402937889099,
      "learning_rate": 0.0026237748770979387,
      "loss": 1.5749,
      "step": 37340
    },
    {
      "epoch": 0.14063364146322194,
      "grad_norm": 0.104873888194561,
      "learning_rate": 0.0026233553755819956,
      "loss": 1.5279,
      "step": 37360
    },
    {
      "epoch": 0.14076532090279423,
      "grad_norm": 0.09734835475683212,
      "learning_rate": 0.0026229356738944054,
      "loss": 1.4679,
      "step": 37380
    },
    {
      "epoch": 0.14089700034236655,
      "grad_norm": 0.10877981036901474,
      "learning_rate": 0.0026225157721099556,
      "loss": 1.4734,
      "step": 37400
    },
    {
      "epoch": 0.14102867978193884,
      "grad_norm": 0.10789605230093002,
      "learning_rate": 0.002622095670303469,
      "loss": 1.4646,
      "step": 37420
    },
    {
      "epoch": 0.14116035922151116,
      "grad_norm": 0.08844413608312607,
      "learning_rate": 0.0026216753685498035,
      "loss": 1.4594,
      "step": 37440
    },
    {
      "epoch": 0.14129203866108345,
      "grad_norm": 0.07974802702665329,
      "learning_rate": 0.0026212548669238534,
      "loss": 1.4563,
      "step": 37460
    },
    {
      "epoch": 0.14142371810065577,
      "grad_norm": 0.1050298884510994,
      "learning_rate": 0.002620834165500548,
      "loss": 1.4593,
      "step": 37480
    },
    {
      "epoch": 0.14155539754022806,
      "grad_norm": 0.11050856858491898,
      "learning_rate": 0.0026204132643548525,
      "loss": 1.4532,
      "step": 37500
    },
    {
      "epoch": 0.14168707697980037,
      "grad_norm": 0.09087752550840378,
      "learning_rate": 0.0026199921635617672,
      "loss": 1.4568,
      "step": 37520
    },
    {
      "epoch": 0.14181875641937267,
      "grad_norm": 0.08688180893659592,
      "learning_rate": 0.0026195708631963294,
      "loss": 1.4584,
      "step": 37540
    },
    {
      "epoch": 0.14195043585894498,
      "grad_norm": 0.08796868473291397,
      "learning_rate": 0.00261914936333361,
      "loss": 1.4513,
      "step": 37560
    },
    {
      "epoch": 0.1420821152985173,
      "grad_norm": 0.1213894784450531,
      "learning_rate": 0.002618727664048717,
      "loss": 1.4539,
      "step": 37580
    },
    {
      "epoch": 0.1422137947380896,
      "grad_norm": 0.11895439028739929,
      "learning_rate": 0.0026183057654167933,
      "loss": 1.4568,
      "step": 37600
    },
    {
      "epoch": 0.1423454741776619,
      "grad_norm": 0.10927541553974152,
      "learning_rate": 0.002617883667513017,
      "loss": 1.4576,
      "step": 37620
    },
    {
      "epoch": 0.1424771536172342,
      "grad_norm": 0.11971597373485565,
      "learning_rate": 0.0026174613704126027,
      "loss": 1.4586,
      "step": 37640
    },
    {
      "epoch": 0.14260883305680652,
      "grad_norm": 0.11147040128707886,
      "learning_rate": 0.0026170388741907995,
      "loss": 1.4556,
      "step": 37660
    },
    {
      "epoch": 0.1427405124963788,
      "grad_norm": 0.09948437660932541,
      "learning_rate": 0.0026166161789228917,
      "loss": 1.4559,
      "step": 37680
    },
    {
      "epoch": 0.14287219193595113,
      "grad_norm": 0.11996204406023026,
      "learning_rate": 0.0026161932846842013,
      "loss": 1.458,
      "step": 37700
    },
    {
      "epoch": 0.14300387137552342,
      "grad_norm": 0.1091773509979248,
      "learning_rate": 0.0026157701915500826,
      "loss": 1.4576,
      "step": 37720
    },
    {
      "epoch": 0.14313555081509574,
      "grad_norm": 0.27840930223464966,
      "learning_rate": 0.002615346899595928,
      "loss": 1.464,
      "step": 37740
    },
    {
      "epoch": 0.14326723025466803,
      "grad_norm": 0.09474640339612961,
      "learning_rate": 0.002614923408897164,
      "loss": 1.4745,
      "step": 37760
    },
    {
      "epoch": 0.14339890969424035,
      "grad_norm": 0.13245952129364014,
      "learning_rate": 0.002614499719529252,
      "loss": 1.4685,
      "step": 37780
    },
    {
      "epoch": 0.14353058913381264,
      "grad_norm": 0.10501421242952347,
      "learning_rate": 0.0026140758315676906,
      "loss": 1.4674,
      "step": 37800
    },
    {
      "epoch": 0.14366226857338496,
      "grad_norm": 0.10588248074054718,
      "learning_rate": 0.0026136517450880124,
      "loss": 1.4648,
      "step": 37820
    },
    {
      "epoch": 0.14379394801295725,
      "grad_norm": 0.10599175095558167,
      "learning_rate": 0.0026132274601657853,
      "loss": 1.4685,
      "step": 37840
    },
    {
      "epoch": 0.14392562745252957,
      "grad_norm": 0.11727793514728546,
      "learning_rate": 0.002612802976876614,
      "loss": 1.4643,
      "step": 37860
    },
    {
      "epoch": 0.14405730689210186,
      "grad_norm": 0.09941857308149338,
      "learning_rate": 0.0026123782952961363,
      "loss": 1.4645,
      "step": 37880
    },
    {
      "epoch": 0.14418898633167418,
      "grad_norm": 0.09774893522262573,
      "learning_rate": 0.0026119534155000274,
      "loss": 1.4684,
      "step": 37900
    },
    {
      "epoch": 0.14432066577124647,
      "grad_norm": 0.10435054451227188,
      "learning_rate": 0.002611528337563997,
      "loss": 1.4912,
      "step": 37920
    },
    {
      "epoch": 0.1444523452108188,
      "grad_norm": 0.1452368199825287,
      "learning_rate": 0.0026111030615637893,
      "loss": 1.5458,
      "step": 37940
    },
    {
      "epoch": 0.14458402465039108,
      "grad_norm": 0.1285165399312973,
      "learning_rate": 0.0026106775875751856,
      "loss": 1.5752,
      "step": 37960
    },
    {
      "epoch": 0.1447157040899634,
      "grad_norm": 0.11321987956762314,
      "learning_rate": 0.0026102519156740007,
      "loss": 1.5846,
      "step": 37980
    },
    {
      "epoch": 0.1448473835295357,
      "grad_norm": 0.0958334282040596,
      "learning_rate": 0.0026098260459360865,
      "loss": 1.5965,
      "step": 38000
    },
    {
      "epoch": 0.1448473835295357,
      "eval_loss": 1.5834728479385376,
      "eval_runtime": 67.4429,
      "eval_samples_per_second": 14.827,
      "eval_steps_per_second": 14.827,
      "step": 38000
    },
    {
      "epoch": 0.144979062969108,
      "grad_norm": 0.12879179418087006,
      "learning_rate": 0.0026093999784373273,
      "loss": 1.595,
      "step": 38020
    },
    {
      "epoch": 0.1451107424086803,
      "grad_norm": 0.10820400714874268,
      "learning_rate": 0.0026089737132536464,
      "loss": 1.5911,
      "step": 38040
    },
    {
      "epoch": 0.14524242184825262,
      "grad_norm": 0.09675215184688568,
      "learning_rate": 0.002608547250460999,
      "loss": 1.589,
      "step": 38060
    },
    {
      "epoch": 0.1453741012878249,
      "grad_norm": 0.11472152918577194,
      "learning_rate": 0.0026081205901353774,
      "loss": 1.5921,
      "step": 38080
    },
    {
      "epoch": 0.14550578072739723,
      "grad_norm": 0.12123990058898926,
      "learning_rate": 0.0026076937323528085,
      "loss": 1.5846,
      "step": 38100
    },
    {
      "epoch": 0.14563746016696952,
      "grad_norm": 0.11360788345336914,
      "learning_rate": 0.0026072666771893545,
      "loss": 1.586,
      "step": 38120
    },
    {
      "epoch": 0.14576913960654184,
      "grad_norm": 0.10187938064336777,
      "learning_rate": 0.0026068394247211126,
      "loss": 1.5779,
      "step": 38140
    },
    {
      "epoch": 0.14590081904611413,
      "grad_norm": 0.1109524741768837,
      "learning_rate": 0.002606411975024215,
      "loss": 1.582,
      "step": 38160
    },
    {
      "epoch": 0.14603249848568645,
      "grad_norm": 0.13011963665485382,
      "learning_rate": 0.0026059843281748298,
      "loss": 1.5822,
      "step": 38180
    },
    {
      "epoch": 0.14616417792525874,
      "grad_norm": 0.14091023802757263,
      "learning_rate": 0.0026055564842491596,
      "loss": 1.5789,
      "step": 38200
    },
    {
      "epoch": 0.14629585736483106,
      "grad_norm": 0.11127016693353653,
      "learning_rate": 0.002605128443323442,
      "loss": 1.5769,
      "step": 38220
    },
    {
      "epoch": 0.14642753680440337,
      "grad_norm": 0.11978462338447571,
      "learning_rate": 0.002604700205473951,
      "loss": 1.5744,
      "step": 38240
    },
    {
      "epoch": 0.14655921624397567,
      "grad_norm": 0.09419921040534973,
      "learning_rate": 0.002604271770776993,
      "loss": 1.5677,
      "step": 38260
    },
    {
      "epoch": 0.14669089568354798,
      "grad_norm": 0.10313443094491959,
      "learning_rate": 0.0026038431393089127,
      "loss": 1.5681,
      "step": 38280
    },
    {
      "epoch": 0.14682257512312027,
      "grad_norm": 0.12435714900493622,
      "learning_rate": 0.0026034143111460877,
      "loss": 1.5635,
      "step": 38300
    },
    {
      "epoch": 0.1469542545626926,
      "grad_norm": 0.09440986067056656,
      "learning_rate": 0.002602985286364931,
      "loss": 1.5695,
      "step": 38320
    },
    {
      "epoch": 0.14708593400226488,
      "grad_norm": 0.12459173798561096,
      "learning_rate": 0.0026025560650418914,
      "loss": 1.5649,
      "step": 38340
    },
    {
      "epoch": 0.1472176134418372,
      "grad_norm": 0.1102808341383934,
      "learning_rate": 0.0026021266472534515,
      "loss": 1.5629,
      "step": 38360
    },
    {
      "epoch": 0.1473492928814095,
      "grad_norm": 0.10280980169773102,
      "learning_rate": 0.00260169703307613,
      "loss": 1.5676,
      "step": 38380
    },
    {
      "epoch": 0.1474809723209818,
      "grad_norm": 0.1017128974199295,
      "learning_rate": 0.0026012672225864805,
      "loss": 1.5665,
      "step": 38400
    },
    {
      "epoch": 0.1476126517605541,
      "grad_norm": 0.11201263964176178,
      "learning_rate": 0.002600837215861091,
      "loss": 1.5607,
      "step": 38420
    },
    {
      "epoch": 0.14774433120012642,
      "grad_norm": 0.09382157772779465,
      "learning_rate": 0.0026004070129765845,
      "loss": 1.5699,
      "step": 38440
    },
    {
      "epoch": 0.1478760106396987,
      "grad_norm": 0.09265664219856262,
      "learning_rate": 0.00259997661400962,
      "loss": 1.5608,
      "step": 38460
    },
    {
      "epoch": 0.14800769007927103,
      "grad_norm": 0.13978609442710876,
      "learning_rate": 0.00259954601903689,
      "loss": 1.5632,
      "step": 38480
    },
    {
      "epoch": 0.14813936951884332,
      "grad_norm": 0.09976094961166382,
      "learning_rate": 0.0025991152281351223,
      "loss": 1.5662,
      "step": 38500
    },
    {
      "epoch": 0.14827104895841564,
      "grad_norm": 0.13578425347805023,
      "learning_rate": 0.002598684241381081,
      "loss": 1.5691,
      "step": 38520
    },
    {
      "epoch": 0.14840272839798793,
      "grad_norm": 0.09976605325937271,
      "learning_rate": 0.002598253058851563,
      "loss": 1.5962,
      "step": 38540
    },
    {
      "epoch": 0.14853440783756025,
      "grad_norm": 0.10971450805664062,
      "learning_rate": 0.0025978216806234015,
      "loss": 1.592,
      "step": 38560
    },
    {
      "epoch": 0.14866608727713254,
      "grad_norm": 0.11708176881074905,
      "learning_rate": 0.0025973901067734644,
      "loss": 1.5908,
      "step": 38580
    },
    {
      "epoch": 0.14879776671670486,
      "grad_norm": 0.08862312138080597,
      "learning_rate": 0.0025969583373786536,
      "loss": 1.5871,
      "step": 38600
    },
    {
      "epoch": 0.14892944615627715,
      "grad_norm": 0.12013594061136246,
      "learning_rate": 0.0025965263725159066,
      "loss": 1.5808,
      "step": 38620
    },
    {
      "epoch": 0.14906112559584947,
      "grad_norm": 0.093787781894207,
      "learning_rate": 0.0025960942122621963,
      "loss": 1.5854,
      "step": 38640
    },
    {
      "epoch": 0.14919280503542176,
      "grad_norm": 0.08362217992544174,
      "learning_rate": 0.0025956618566945287,
      "loss": 1.5826,
      "step": 38660
    },
    {
      "epoch": 0.14932448447499408,
      "grad_norm": 0.1498386114835739,
      "learning_rate": 0.002595229305889946,
      "loss": 1.5732,
      "step": 38680
    },
    {
      "epoch": 0.14945616391456637,
      "grad_norm": 0.09603530913591385,
      "learning_rate": 0.0025947965599255247,
      "loss": 1.5698,
      "step": 38700
    },
    {
      "epoch": 0.1495878433541387,
      "grad_norm": 0.0774497240781784,
      "learning_rate": 0.0025943636188783766,
      "loss": 1.5715,
      "step": 38720
    },
    {
      "epoch": 0.14971952279371098,
      "grad_norm": 0.08751595765352249,
      "learning_rate": 0.002593930482825647,
      "loss": 1.576,
      "step": 38740
    },
    {
      "epoch": 0.1498512022332833,
      "grad_norm": 0.0924377366900444,
      "learning_rate": 0.002593497151844518,
      "loss": 1.5715,
      "step": 38760
    },
    {
      "epoch": 0.1499828816728556,
      "grad_norm": 0.10947736352682114,
      "learning_rate": 0.0025930636260122044,
      "loss": 1.5582,
      "step": 38780
    },
    {
      "epoch": 0.1501145611124279,
      "grad_norm": 0.09718424081802368,
      "learning_rate": 0.0025926299054059567,
      "loss": 1.5644,
      "step": 38800
    },
    {
      "epoch": 0.1502462405520002,
      "grad_norm": 0.1110563725233078,
      "learning_rate": 0.0025921959901030597,
      "loss": 1.5656,
      "step": 38820
    },
    {
      "epoch": 0.15037791999157252,
      "grad_norm": 0.09171340614557266,
      "learning_rate": 0.0025917618801808333,
      "loss": 1.5597,
      "step": 38840
    },
    {
      "epoch": 0.1505095994311448,
      "grad_norm": 0.12216000258922577,
      "learning_rate": 0.0025913275757166324,
      "loss": 1.5548,
      "step": 38860
    },
    {
      "epoch": 0.15064127887071713,
      "grad_norm": 0.08841335028409958,
      "learning_rate": 0.002590893076787845,
      "loss": 1.5601,
      "step": 38880
    },
    {
      "epoch": 0.15077295831028942,
      "grad_norm": 0.10563861578702927,
      "learning_rate": 0.002590458383471896,
      "loss": 1.5535,
      "step": 38900
    },
    {
      "epoch": 0.15090463774986174,
      "grad_norm": 0.08755604177713394,
      "learning_rate": 0.0025900234958462433,
      "loss": 1.5501,
      "step": 38920
    },
    {
      "epoch": 0.15103631718943406,
      "grad_norm": 0.13694095611572266,
      "learning_rate": 0.0025895884139883795,
      "loss": 1.5627,
      "step": 38940
    },
    {
      "epoch": 0.15116799662900635,
      "grad_norm": 0.10837943106889725,
      "learning_rate": 0.002589153137975833,
      "loss": 1.5613,
      "step": 38960
    },
    {
      "epoch": 0.15129967606857866,
      "grad_norm": 0.08162432163953781,
      "learning_rate": 0.0025887176678861647,
      "loss": 1.5607,
      "step": 38980
    },
    {
      "epoch": 0.15143135550815096,
      "grad_norm": 0.14715518057346344,
      "learning_rate": 0.002588282003796973,
      "loss": 1.5544,
      "step": 39000
    },
    {
      "epoch": 0.15143135550815096,
      "eval_loss": 1.427059292793274,
      "eval_runtime": 66.7992,
      "eval_samples_per_second": 14.97,
      "eval_steps_per_second": 14.97,
      "step": 39000
    },
    {
      "epoch": 0.15156303494772327,
      "grad_norm": 0.10396742075681686,
      "learning_rate": 0.002587846145785888,
      "loss": 1.5584,
      "step": 39020
    },
    {
      "epoch": 0.15169471438729557,
      "grad_norm": 0.12578904628753662,
      "learning_rate": 0.002587410093930576,
      "loss": 1.5543,
      "step": 39040
    },
    {
      "epoch": 0.15182639382686788,
      "grad_norm": 0.12245041877031326,
      "learning_rate": 0.0025869738483087386,
      "loss": 1.5484,
      "step": 39060
    },
    {
      "epoch": 0.15195807326644017,
      "grad_norm": 0.10764380544424057,
      "learning_rate": 0.0025865374089981087,
      "loss": 1.5526,
      "step": 39080
    },
    {
      "epoch": 0.1520897527060125,
      "grad_norm": 0.10245970636606216,
      "learning_rate": 0.002586100776076457,
      "loss": 1.5609,
      "step": 39100
    },
    {
      "epoch": 0.15222143214558478,
      "grad_norm": 0.11582383513450623,
      "learning_rate": 0.0025856639496215873,
      "loss": 1.5789,
      "step": 39120
    },
    {
      "epoch": 0.1523531115851571,
      "grad_norm": 0.09922783076763153,
      "learning_rate": 0.0025852269297113382,
      "loss": 1.6029,
      "step": 39140
    },
    {
      "epoch": 0.1524847910247294,
      "grad_norm": 0.08174051344394684,
      "learning_rate": 0.002584789716423583,
      "loss": 1.5928,
      "step": 39160
    },
    {
      "epoch": 0.1526164704643017,
      "grad_norm": 0.14259853959083557,
      "learning_rate": 0.0025843523098362277,
      "loss": 1.6008,
      "step": 39180
    },
    {
      "epoch": 0.152748149903874,
      "grad_norm": 0.10020045936107635,
      "learning_rate": 0.0025839147100272155,
      "loss": 1.6035,
      "step": 39200
    },
    {
      "epoch": 0.15287982934344632,
      "grad_norm": 0.1230427622795105,
      "learning_rate": 0.002583476917074522,
      "loss": 1.6032,
      "step": 39220
    },
    {
      "epoch": 0.1530115087830186,
      "grad_norm": 0.0876646339893341,
      "learning_rate": 0.002583038931056159,
      "loss": 1.5921,
      "step": 39240
    },
    {
      "epoch": 0.15314318822259093,
      "grad_norm": 0.14442962408065796,
      "learning_rate": 0.0025826007520501704,
      "loss": 1.5894,
      "step": 39260
    },
    {
      "epoch": 0.15327486766216322,
      "grad_norm": 0.130369171500206,
      "learning_rate": 0.0025821623801346354,
      "loss": 1.5833,
      "step": 39280
    },
    {
      "epoch": 0.15340654710173554,
      "grad_norm": 0.09191744774580002,
      "learning_rate": 0.002581723815387669,
      "loss": 1.5819,
      "step": 39300
    },
    {
      "epoch": 0.15353822654130783,
      "grad_norm": 0.12108132243156433,
      "learning_rate": 0.002581285057887419,
      "loss": 1.5859,
      "step": 39320
    },
    {
      "epoch": 0.15366990598088015,
      "grad_norm": 0.09722419083118439,
      "learning_rate": 0.002580846107712068,
      "loss": 1.5811,
      "step": 39340
    },
    {
      "epoch": 0.15380158542045244,
      "grad_norm": 0.0996871218085289,
      "learning_rate": 0.0025804069649398324,
      "loss": 1.5806,
      "step": 39360
    },
    {
      "epoch": 0.15393326486002476,
      "grad_norm": 0.1356852501630783,
      "learning_rate": 0.002579967629648964,
      "loss": 1.5764,
      "step": 39380
    },
    {
      "epoch": 0.15406494429959705,
      "grad_norm": 0.0935957059264183,
      "learning_rate": 0.0025795281019177483,
      "loss": 1.5694,
      "step": 39400
    },
    {
      "epoch": 0.15419662373916937,
      "grad_norm": 0.10903099179267883,
      "learning_rate": 0.002579088381824505,
      "loss": 1.5665,
      "step": 39420
    },
    {
      "epoch": 0.15432830317874166,
      "grad_norm": 0.10038184374570847,
      "learning_rate": 0.0025786484694475882,
      "loss": 1.5714,
      "step": 39440
    },
    {
      "epoch": 0.15445998261831398,
      "grad_norm": 0.08461392670869827,
      "learning_rate": 0.002578208364865386,
      "loss": 1.5699,
      "step": 39460
    },
    {
      "epoch": 0.15459166205788627,
      "grad_norm": 0.09134264290332794,
      "learning_rate": 0.0025777680681563215,
      "loss": 1.5695,
      "step": 39480
    },
    {
      "epoch": 0.1547233414974586,
      "grad_norm": 0.11467129737138748,
      "learning_rate": 0.002577327579398851,
      "loss": 1.5743,
      "step": 39500
    },
    {
      "epoch": 0.15485502093703088,
      "grad_norm": 0.09146096557378769,
      "learning_rate": 0.002576886898671466,
      "loss": 1.5711,
      "step": 39520
    },
    {
      "epoch": 0.1549867003766032,
      "grad_norm": 0.08838360756635666,
      "learning_rate": 0.0025764460260526917,
      "loss": 1.5699,
      "step": 39540
    },
    {
      "epoch": 0.1551183798161755,
      "grad_norm": 0.09347314387559891,
      "learning_rate": 0.0025760049616210872,
      "loss": 1.568,
      "step": 39560
    },
    {
      "epoch": 0.1552500592557478,
      "grad_norm": 0.11676613986492157,
      "learning_rate": 0.002575563705455247,
      "loss": 1.5667,
      "step": 39580
    },
    {
      "epoch": 0.15538173869532013,
      "grad_norm": 0.11522109806537628,
      "learning_rate": 0.002575122257633798,
      "loss": 1.5708,
      "step": 39600
    },
    {
      "epoch": 0.15551341813489242,
      "grad_norm": 0.09783374518156052,
      "learning_rate": 0.0025746806182354023,
      "loss": 1.5925,
      "step": 39620
    },
    {
      "epoch": 0.15564509757446474,
      "grad_norm": 0.0943576991558075,
      "learning_rate": 0.0025742387873387567,
      "loss": 1.5825,
      "step": 39640
    },
    {
      "epoch": 0.15577677701403703,
      "grad_norm": 0.09120535105466843,
      "learning_rate": 0.0025737967650225906,
      "loss": 1.5754,
      "step": 39660
    },
    {
      "epoch": 0.15590845645360935,
      "grad_norm": 0.11549575626850128,
      "learning_rate": 0.0025733545513656683,
      "loss": 1.5741,
      "step": 39680
    },
    {
      "epoch": 0.15604013589318164,
      "grad_norm": 0.08214199542999268,
      "learning_rate": 0.0025729121464467897,
      "loss": 1.5684,
      "step": 39700
    },
    {
      "epoch": 0.15617181533275395,
      "grad_norm": 0.09407494962215424,
      "learning_rate": 0.0025724695503447853,
      "loss": 1.5623,
      "step": 39720
    },
    {
      "epoch": 0.15630349477232625,
      "grad_norm": 0.10400024056434631,
      "learning_rate": 0.0025720267631385227,
      "loss": 1.5565,
      "step": 39740
    },
    {
      "epoch": 0.15643517421189856,
      "grad_norm": 0.09701120853424072,
      "learning_rate": 0.002571583784906902,
      "loss": 1.5475,
      "step": 39760
    },
    {
      "epoch": 0.15656685365147086,
      "grad_norm": 0.09145724773406982,
      "learning_rate": 0.0025711406157288594,
      "loss": 1.5366,
      "step": 39780
    },
    {
      "epoch": 0.15669853309104317,
      "grad_norm": 0.10197409987449646,
      "learning_rate": 0.0025706972556833615,
      "loss": 1.5368,
      "step": 39800
    },
    {
      "epoch": 0.15683021253061546,
      "grad_norm": 0.10949882119894028,
      "learning_rate": 0.0025702537048494126,
      "loss": 1.5324,
      "step": 39820
    },
    {
      "epoch": 0.15696189197018778,
      "grad_norm": 0.08680877834558487,
      "learning_rate": 0.0025698099633060484,
      "loss": 1.5246,
      "step": 39840
    },
    {
      "epoch": 0.15709357140976007,
      "grad_norm": 0.09298034757375717,
      "learning_rate": 0.0025693660311323398,
      "loss": 1.5244,
      "step": 39860
    },
    {
      "epoch": 0.1572252508493324,
      "grad_norm": 0.08509939163923264,
      "learning_rate": 0.002568921908407392,
      "loss": 1.5269,
      "step": 39880
    },
    {
      "epoch": 0.15735693028890468,
      "grad_norm": 0.13192257285118103,
      "learning_rate": 0.002568477595210343,
      "loss": 1.5189,
      "step": 39900
    },
    {
      "epoch": 0.157488609728477,
      "grad_norm": 0.09820321202278137,
      "learning_rate": 0.0025680330916203653,
      "loss": 1.5172,
      "step": 39920
    },
    {
      "epoch": 0.1576202891680493,
      "grad_norm": 0.12466151267290115,
      "learning_rate": 0.0025675883977166656,
      "loss": 1.5168,
      "step": 39940
    },
    {
      "epoch": 0.1577519686076216,
      "grad_norm": 0.1055043414235115,
      "learning_rate": 0.002567143513578485,
      "loss": 1.5127,
      "step": 39960
    },
    {
      "epoch": 0.1578836480471939,
      "grad_norm": 0.0885360836982727,
      "learning_rate": 0.0025666984392850966,
      "loss": 1.5102,
      "step": 39980
    },
    {
      "epoch": 0.15801532748676622,
      "grad_norm": 0.09092732518911362,
      "learning_rate": 0.0025662531749158085,
      "loss": 1.5128,
      "step": 40000
    },
    {
      "epoch": 0.15801532748676622,
      "eval_loss": 1.5180437564849854,
      "eval_runtime": 66.3341,
      "eval_samples_per_second": 15.075,
      "eval_steps_per_second": 15.075,
      "step": 40000
    },
    {
      "epoch": 0.1581470069263385,
      "grad_norm": 0.11039165407419205,
      "learning_rate": 0.002565807720549964,
      "loss": 1.5056,
      "step": 40020
    },
    {
      "epoch": 0.15827868636591083,
      "grad_norm": 0.09053885191679001,
      "learning_rate": 0.002565362076266938,
      "loss": 1.5117,
      "step": 40040
    },
    {
      "epoch": 0.15841036580548312,
      "grad_norm": 0.10280958563089371,
      "learning_rate": 0.002564916242146141,
      "loss": 1.5026,
      "step": 40060
    },
    {
      "epoch": 0.15854204524505544,
      "grad_norm": 0.10042385011911392,
      "learning_rate": 0.0025644702182670153,
      "loss": 1.5079,
      "step": 40080
    },
    {
      "epoch": 0.15867372468462773,
      "grad_norm": 0.15328729152679443,
      "learning_rate": 0.0025640240047090394,
      "loss": 1.5084,
      "step": 40100
    },
    {
      "epoch": 0.15880540412420005,
      "grad_norm": 0.10443607717752457,
      "learning_rate": 0.002563577601551724,
      "loss": 1.5017,
      "step": 40120
    },
    {
      "epoch": 0.15893708356377234,
      "grad_norm": 0.1254567801952362,
      "learning_rate": 0.0025631310088746143,
      "loss": 1.5015,
      "step": 40140
    },
    {
      "epoch": 0.15906876300334466,
      "grad_norm": 0.11930776387453079,
      "learning_rate": 0.002562684226757289,
      "loss": 1.508,
      "step": 40160
    },
    {
      "epoch": 0.15920044244291695,
      "grad_norm": 0.08289260417222977,
      "learning_rate": 0.0025622372552793603,
      "loss": 1.5091,
      "step": 40180
    },
    {
      "epoch": 0.15933212188248927,
      "grad_norm": 0.10695404559373856,
      "learning_rate": 0.0025617900945204743,
      "loss": 1.5045,
      "step": 40200
    },
    {
      "epoch": 0.15946380132206156,
      "grad_norm": 0.11500686407089233,
      "learning_rate": 0.0025613427445603117,
      "loss": 1.5029,
      "step": 40220
    },
    {
      "epoch": 0.15959548076163388,
      "grad_norm": 0.09452733397483826,
      "learning_rate": 0.0025608952054785856,
      "loss": 1.5029,
      "step": 40240
    },
    {
      "epoch": 0.1597271602012062,
      "grad_norm": 0.10225624591112137,
      "learning_rate": 0.002560447477355043,
      "loss": 1.5061,
      "step": 40260
    },
    {
      "epoch": 0.1598588396407785,
      "grad_norm": 0.10620761662721634,
      "learning_rate": 0.002559999560269466,
      "loss": 1.5015,
      "step": 40280
    },
    {
      "epoch": 0.1599905190803508,
      "grad_norm": 0.11584058403968811,
      "learning_rate": 0.002559551454301669,
      "loss": 1.5255,
      "step": 40300
    },
    {
      "epoch": 0.1601221985199231,
      "grad_norm": 0.12006339430809021,
      "learning_rate": 0.0025591031595314996,
      "loss": 1.5512,
      "step": 40320
    },
    {
      "epoch": 0.16025387795949542,
      "grad_norm": 0.10483147948980331,
      "learning_rate": 0.0025586546760388413,
      "loss": 1.5584,
      "step": 40340
    },
    {
      "epoch": 0.1603855573990677,
      "grad_norm": 0.12490592896938324,
      "learning_rate": 0.002558206003903608,
      "loss": 1.5647,
      "step": 40360
    },
    {
      "epoch": 0.16051723683864003,
      "grad_norm": 0.09851817041635513,
      "learning_rate": 0.0025577571432057497,
      "loss": 1.558,
      "step": 40380
    },
    {
      "epoch": 0.16064891627821232,
      "grad_norm": 0.10723355412483215,
      "learning_rate": 0.0025573080940252497,
      "loss": 1.5505,
      "step": 40400
    },
    {
      "epoch": 0.16078059571778464,
      "grad_norm": 0.08668304234743118,
      "learning_rate": 0.002556858856442124,
      "loss": 1.5473,
      "step": 40420
    },
    {
      "epoch": 0.16091227515735693,
      "grad_norm": 0.13923218846321106,
      "learning_rate": 0.002556409430536423,
      "loss": 1.5451,
      "step": 40440
    },
    {
      "epoch": 0.16104395459692925,
      "grad_norm": 0.11584249883890152,
      "learning_rate": 0.00255595981638823,
      "loss": 1.5435,
      "step": 40460
    },
    {
      "epoch": 0.16117563403650154,
      "grad_norm": 0.09080822020769119,
      "learning_rate": 0.0025555100140776618,
      "loss": 1.5435,
      "step": 40480
    },
    {
      "epoch": 0.16130731347607385,
      "grad_norm": 0.0972919836640358,
      "learning_rate": 0.0025550600236848697,
      "loss": 1.5367,
      "step": 40500
    },
    {
      "epoch": 0.16143899291564615,
      "grad_norm": 0.10431429743766785,
      "learning_rate": 0.0025546098452900374,
      "loss": 1.5362,
      "step": 40520
    },
    {
      "epoch": 0.16157067235521846,
      "grad_norm": 0.1297275573015213,
      "learning_rate": 0.002554159478973383,
      "loss": 1.5287,
      "step": 40540
    },
    {
      "epoch": 0.16170235179479076,
      "grad_norm": 0.08693351596593857,
      "learning_rate": 0.002553708924815157,
      "loss": 1.5322,
      "step": 40560
    },
    {
      "epoch": 0.16183403123436307,
      "grad_norm": 0.10223674029111862,
      "learning_rate": 0.002553258182895645,
      "loss": 1.5283,
      "step": 40580
    },
    {
      "epoch": 0.16196571067393536,
      "grad_norm": 0.13661502301692963,
      "learning_rate": 0.0025528072532951646,
      "loss": 1.5335,
      "step": 40600
    },
    {
      "epoch": 0.16209739011350768,
      "grad_norm": 0.0959714725613594,
      "learning_rate": 0.0025523561360940666,
      "loss": 1.5409,
      "step": 40620
    },
    {
      "epoch": 0.16222906955307997,
      "grad_norm": 0.11499602347612381,
      "learning_rate": 0.0025519048313727365,
      "loss": 1.534,
      "step": 40640
    },
    {
      "epoch": 0.1623607489926523,
      "grad_norm": 0.10685988515615463,
      "learning_rate": 0.0025514533392115932,
      "loss": 1.5268,
      "step": 40660
    },
    {
      "epoch": 0.16249242843222458,
      "grad_norm": 0.12656660377979279,
      "learning_rate": 0.002551001659691088,
      "loss": 1.5253,
      "step": 40680
    },
    {
      "epoch": 0.1626241078717969,
      "grad_norm": 0.11715365201234818,
      "learning_rate": 0.002550549792891706,
      "loss": 1.5259,
      "step": 40700
    },
    {
      "epoch": 0.1627557873113692,
      "grad_norm": 0.1329299360513687,
      "learning_rate": 0.0025500977388939667,
      "loss": 1.525,
      "step": 40720
    },
    {
      "epoch": 0.1628874667509415,
      "grad_norm": 0.0944843739271164,
      "learning_rate": 0.00254964549777842,
      "loss": 1.5223,
      "step": 40740
    },
    {
      "epoch": 0.1630191461905138,
      "grad_norm": 0.09008379280567169,
      "learning_rate": 0.002549193069625653,
      "loss": 1.5215,
      "step": 40760
    },
    {
      "epoch": 0.16315082563008612,
      "grad_norm": 0.1306624561548233,
      "learning_rate": 0.0025487404545162825,
      "loss": 1.5174,
      "step": 40780
    },
    {
      "epoch": 0.1632825050696584,
      "grad_norm": 0.08844735473394394,
      "learning_rate": 0.0025482876525309618,
      "loss": 1.5239,
      "step": 40800
    },
    {
      "epoch": 0.16341418450923073,
      "grad_norm": 0.09808750450611115,
      "learning_rate": 0.0025478346637503764,
      "loss": 1.5219,
      "step": 40820
    },
    {
      "epoch": 0.16354586394880302,
      "grad_norm": 0.10559525340795517,
      "learning_rate": 0.002547381488255243,
      "loss": 1.5201,
      "step": 40840
    },
    {
      "epoch": 0.16367754338837534,
      "grad_norm": 0.09977331012487411,
      "learning_rate": 0.0025469281261263155,
      "loss": 1.5204,
      "step": 40860
    },
    {
      "epoch": 0.16380922282794763,
      "grad_norm": 0.11581449955701828,
      "learning_rate": 0.0025464745774443766,
      "loss": 1.5199,
      "step": 40880
    },
    {
      "epoch": 0.16394090226751995,
      "grad_norm": 0.10969722270965576,
      "learning_rate": 0.0025460208422902458,
      "loss": 1.5318,
      "step": 40900
    },
    {
      "epoch": 0.16407258170709224,
      "grad_norm": 0.12273216247558594,
      "learning_rate": 0.0025455669207447747,
      "loss": 1.5515,
      "step": 40920
    },
    {
      "epoch": 0.16420426114666456,
      "grad_norm": 0.09553756564855576,
      "learning_rate": 0.0025451128128888474,
      "loss": 1.5507,
      "step": 40940
    },
    {
      "epoch": 0.16433594058623688,
      "grad_norm": 0.1060953214764595,
      "learning_rate": 0.002544658518803382,
      "loss": 1.5528,
      "step": 40960
    },
    {
      "epoch": 0.16446762002580917,
      "grad_norm": 0.12041070312261581,
      "learning_rate": 0.00254420403856933,
      "loss": 1.5417,
      "step": 40980
    },
    {
      "epoch": 0.1645992994653815,
      "grad_norm": 0.09940062463283539,
      "learning_rate": 0.0025437493722676747,
      "loss": 1.5393,
      "step": 41000
    },
    {
      "epoch": 0.1645992994653815,
      "eval_loss": 1.5645551681518555,
      "eval_runtime": 67.891,
      "eval_samples_per_second": 14.729,
      "eval_steps_per_second": 14.729,
      "step": 41000
    },
    {
      "epoch": 0.16473097890495378,
      "grad_norm": 0.09117773175239563,
      "learning_rate": 0.0025432945199794343,
      "loss": 1.5385,
      "step": 41020
    },
    {
      "epoch": 0.1648626583445261,
      "grad_norm": 0.10133585333824158,
      "learning_rate": 0.0025428394817856585,
      "loss": 1.5357,
      "step": 41040
    },
    {
      "epoch": 0.1649943377840984,
      "grad_norm": 0.09552034735679626,
      "learning_rate": 0.0025423842577674315,
      "loss": 1.5355,
      "step": 41060
    },
    {
      "epoch": 0.1651260172236707,
      "grad_norm": 0.11098198592662811,
      "learning_rate": 0.00254192884800587,
      "loss": 1.5294,
      "step": 41080
    },
    {
      "epoch": 0.165257696663243,
      "grad_norm": 0.11762163043022156,
      "learning_rate": 0.0025414732525821243,
      "loss": 1.5263,
      "step": 41100
    },
    {
      "epoch": 0.16538937610281532,
      "grad_norm": 0.10004954785108566,
      "learning_rate": 0.0025410174715773766,
      "loss": 1.5232,
      "step": 41120
    },
    {
      "epoch": 0.1655210555423876,
      "grad_norm": 0.10917594283819199,
      "learning_rate": 0.0025405615050728427,
      "loss": 1.5209,
      "step": 41140
    },
    {
      "epoch": 0.16565273498195993,
      "grad_norm": 0.09700624644756317,
      "learning_rate": 0.002540105353149773,
      "loss": 1.5246,
      "step": 41160
    },
    {
      "epoch": 0.16578441442153222,
      "grad_norm": 0.1017017513513565,
      "learning_rate": 0.002539649015889448,
      "loss": 1.5235,
      "step": 41180
    },
    {
      "epoch": 0.16591609386110454,
      "grad_norm": 0.09013497084379196,
      "learning_rate": 0.002539192493373184,
      "loss": 1.5123,
      "step": 41200
    },
    {
      "epoch": 0.16604777330067683,
      "grad_norm": 0.117029570043087,
      "learning_rate": 0.0025387357856823277,
      "loss": 1.5192,
      "step": 41220
    },
    {
      "epoch": 0.16617945274024915,
      "grad_norm": 0.11237141489982605,
      "learning_rate": 0.002538278892898262,
      "loss": 1.5113,
      "step": 41240
    },
    {
      "epoch": 0.16631113217982144,
      "grad_norm": 0.09884624928236008,
      "learning_rate": 0.0025378218151024,
      "loss": 1.5188,
      "step": 41260
    },
    {
      "epoch": 0.16644281161939375,
      "grad_norm": 0.11448849737644196,
      "learning_rate": 0.002537364552376189,
      "loss": 1.5255,
      "step": 41280
    },
    {
      "epoch": 0.16657449105896605,
      "grad_norm": 0.12878195941448212,
      "learning_rate": 0.0025369071048011093,
      "loss": 1.5255,
      "step": 41300
    },
    {
      "epoch": 0.16670617049853836,
      "grad_norm": 0.12284272909164429,
      "learning_rate": 0.0025364494724586734,
      "loss": 1.5184,
      "step": 41320
    },
    {
      "epoch": 0.16683784993811066,
      "grad_norm": 0.0931035578250885,
      "learning_rate": 0.0025359916554304274,
      "loss": 1.5149,
      "step": 41340
    },
    {
      "epoch": 0.16696952937768297,
      "grad_norm": 0.09691374748945236,
      "learning_rate": 0.0025355336537979498,
      "loss": 1.5157,
      "step": 41360
    },
    {
      "epoch": 0.16710120881725526,
      "grad_norm": 0.09920854866504669,
      "learning_rate": 0.0025350754676428527,
      "loss": 1.5215,
      "step": 41380
    },
    {
      "epoch": 0.16723288825682758,
      "grad_norm": 0.13028231263160706,
      "learning_rate": 0.0025346170970467806,
      "loss": 1.5156,
      "step": 41400
    },
    {
      "epoch": 0.16736456769639987,
      "grad_norm": 0.09974183887243271,
      "learning_rate": 0.002534158542091411,
      "loss": 1.5231,
      "step": 41420
    },
    {
      "epoch": 0.1674962471359722,
      "grad_norm": 0.09403842687606812,
      "learning_rate": 0.0025336998028584546,
      "loss": 1.5189,
      "step": 41440
    },
    {
      "epoch": 0.16762792657554448,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0025332408794296528,
      "loss": 1.5196,
      "step": 41460
    },
    {
      "epoch": 0.1677596060151168,
      "grad_norm": 0.11757822334766388,
      "learning_rate": 0.0025327817718867837,
      "loss": 1.5365,
      "step": 41480
    },
    {
      "epoch": 0.1678912854546891,
      "grad_norm": 0.11819227039813995,
      "learning_rate": 0.002532322480311655,
      "loss": 1.5616,
      "step": 41500
    },
    {
      "epoch": 0.1680229648942614,
      "grad_norm": 0.12360888719558716,
      "learning_rate": 0.0025318630047861082,
      "loss": 1.5725,
      "step": 41520
    },
    {
      "epoch": 0.1681546443338337,
      "grad_norm": 0.09366869181394577,
      "learning_rate": 0.0025314033453920186,
      "loss": 1.5816,
      "step": 41540
    },
    {
      "epoch": 0.16828632377340602,
      "grad_norm": 0.09530407935380936,
      "learning_rate": 0.0025309435022112916,
      "loss": 1.5805,
      "step": 41560
    },
    {
      "epoch": 0.1684180032129783,
      "grad_norm": 0.12878450751304626,
      "learning_rate": 0.0025304834753258685,
      "loss": 1.5807,
      "step": 41580
    },
    {
      "epoch": 0.16854968265255063,
      "grad_norm": 0.1358455866575241,
      "learning_rate": 0.0025300232648177213,
      "loss": 1.5794,
      "step": 41600
    },
    {
      "epoch": 0.16868136209212295,
      "grad_norm": 0.12388956546783447,
      "learning_rate": 0.002529562870768855,
      "loss": 1.5766,
      "step": 41620
    },
    {
      "epoch": 0.16881304153169524,
      "grad_norm": 0.10921106487512589,
      "learning_rate": 0.002529102293261308,
      "loss": 1.5705,
      "step": 41640
    },
    {
      "epoch": 0.16894472097126756,
      "grad_norm": 0.1032082810997963,
      "learning_rate": 0.0025286415323771514,
      "loss": 1.5692,
      "step": 41660
    },
    {
      "epoch": 0.16907640041083985,
      "grad_norm": 0.10300302505493164,
      "learning_rate": 0.0025281805881984887,
      "loss": 1.567,
      "step": 41680
    },
    {
      "epoch": 0.16920807985041217,
      "grad_norm": 0.10055007040500641,
      "learning_rate": 0.002527719460807454,
      "loss": 1.566,
      "step": 41700
    },
    {
      "epoch": 0.16933975928998446,
      "grad_norm": 0.12738655507564545,
      "learning_rate": 0.0025272581502862185,
      "loss": 1.561,
      "step": 41720
    },
    {
      "epoch": 0.16947143872955678,
      "grad_norm": 0.10861164331436157,
      "learning_rate": 0.0025267966567169816,
      "loss": 1.567,
      "step": 41740
    },
    {
      "epoch": 0.16960311816912907,
      "grad_norm": 0.12459450215101242,
      "learning_rate": 0.0025263349801819784,
      "loss": 1.5651,
      "step": 41760
    },
    {
      "epoch": 0.1697347976087014,
      "grad_norm": 0.09590265899896622,
      "learning_rate": 0.0025258731207634753,
      "loss": 1.5661,
      "step": 41780
    },
    {
      "epoch": 0.16986647704827368,
      "grad_norm": 0.11365272849798203,
      "learning_rate": 0.002525411078543771,
      "loss": 1.5619,
      "step": 41800
    },
    {
      "epoch": 0.169998156487846,
      "grad_norm": 0.08910445868968964,
      "learning_rate": 0.0025249488536051965,
      "loss": 1.5585,
      "step": 41820
    },
    {
      "epoch": 0.1701298359274183,
      "grad_norm": 0.18229752779006958,
      "learning_rate": 0.0025244864460301176,
      "loss": 1.5526,
      "step": 41840
    },
    {
      "epoch": 0.1702615153669906,
      "grad_norm": 0.09740415215492249,
      "learning_rate": 0.0025240238559009307,
      "loss": 1.5529,
      "step": 41860
    },
    {
      "epoch": 0.1703931948065629,
      "grad_norm": 0.12560056149959564,
      "learning_rate": 0.0025235610833000644,
      "loss": 1.5509,
      "step": 41880
    },
    {
      "epoch": 0.17052487424613522,
      "grad_norm": 0.15917159616947174,
      "learning_rate": 0.002523098128309981,
      "loss": 1.5542,
      "step": 41900
    },
    {
      "epoch": 0.1706565536857075,
      "grad_norm": 0.13030411303043365,
      "learning_rate": 0.002522634991013175,
      "loss": 1.5513,
      "step": 41920
    },
    {
      "epoch": 0.17078823312527983,
      "grad_norm": 0.09367945790290833,
      "learning_rate": 0.0025221716714921728,
      "loss": 1.5559,
      "step": 41940
    },
    {
      "epoch": 0.17091991256485212,
      "grad_norm": 0.10787492990493774,
      "learning_rate": 0.0025217081698295342,
      "loss": 1.5554,
      "step": 41960
    },
    {
      "epoch": 0.17105159200442444,
      "grad_norm": 0.10562490671873093,
      "learning_rate": 0.0025212444861078503,
      "loss": 1.5504,
      "step": 41980
    },
    {
      "epoch": 0.17118327144399673,
      "grad_norm": 0.1048375815153122,
      "learning_rate": 0.0025207806204097464,
      "loss": 1.5566,
      "step": 42000
    },
    {
      "epoch": 0.17118327144399673,
      "eval_loss": 1.424837350845337,
      "eval_runtime": 66.5617,
      "eval_samples_per_second": 15.024,
      "eval_steps_per_second": 15.024,
      "step": 42000
    },
    {
      "epoch": 0.00013167943957230517,
      "grad_norm": 0.10541951656341553,
      "learning_rate": 0.0025203165728178776,
      "loss": 1.4444,
      "step": 42020
    },
    {
      "epoch": 0.00026335887914461035,
      "grad_norm": 0.12595529854297638,
      "learning_rate": 0.0025198523434149345,
      "loss": 1.435,
      "step": 42040
    },
    {
      "epoch": 0.00039503831871691555,
      "grad_norm": 0.1330813616514206,
      "learning_rate": 0.002519387932283637,
      "loss": 1.4314,
      "step": 42060
    },
    {
      "epoch": 0.0005267177582892207,
      "grad_norm": 0.10303393006324768,
      "learning_rate": 0.00251892333950674,
      "loss": 1.4189,
      "step": 42080
    },
    {
      "epoch": 0.0006583971978615259,
      "grad_norm": 0.13066118955612183,
      "learning_rate": 0.00251845856516703,
      "loss": 1.4188,
      "step": 42100
    },
    {
      "epoch": 0.0007900766374338311,
      "grad_norm": 0.09912826120853424,
      "learning_rate": 0.002517993609347324,
      "loss": 1.4279,
      "step": 42120
    },
    {
      "epoch": 0.0009217560770061363,
      "grad_norm": 0.16035521030426025,
      "learning_rate": 0.0025175284721304747,
      "loss": 1.4217,
      "step": 42140
    },
    {
      "epoch": 0.0010534355165784414,
      "grad_norm": 0.09864924103021622,
      "learning_rate": 0.002517063153599363,
      "loss": 1.4261,
      "step": 42160
    },
    {
      "epoch": 0.0011851149561507466,
      "grad_norm": 0.09063073992729187,
      "learning_rate": 0.002516597653836907,
      "loss": 1.4148,
      "step": 42180
    },
    {
      "epoch": 0.0013167943957230518,
      "grad_norm": 0.10235504806041718,
      "learning_rate": 0.0025161319729260527,
      "loss": 1.4225,
      "step": 42200
    },
    {
      "epoch": 0.001448473835295357,
      "grad_norm": 0.10286398977041245,
      "learning_rate": 0.002515666110949781,
      "loss": 1.4209,
      "step": 42220
    },
    {
      "epoch": 0.0015801532748676622,
      "grad_norm": 0.11597388982772827,
      "learning_rate": 0.002515200067991104,
      "loss": 1.4143,
      "step": 42240
    },
    {
      "epoch": 0.0017118327144399674,
      "grad_norm": 0.1071164458990097,
      "learning_rate": 0.0025147338441330663,
      "loss": 1.4171,
      "step": 42260
    },
    {
      "epoch": 0.0018435121540122726,
      "grad_norm": 0.11942466348409653,
      "learning_rate": 0.0025142674394587447,
      "loss": 1.4249,
      "step": 42280
    },
    {
      "epoch": 0.0019751915935845776,
      "grad_norm": 0.11480249464511871,
      "learning_rate": 0.002513800854051248,
      "loss": 1.4203,
      "step": 42300
    },
    {
      "epoch": 0.002106871033156883,
      "grad_norm": 0.10616378486156464,
      "learning_rate": 0.002513334087993718,
      "loss": 1.423,
      "step": 42320
    },
    {
      "epoch": 0.002238550472729188,
      "grad_norm": 0.09758572280406952,
      "learning_rate": 0.0025128671413693275,
      "loss": 1.4221,
      "step": 42340
    },
    {
      "epoch": 0.002370229912301493,
      "grad_norm": 0.16627156734466553,
      "learning_rate": 0.002512400014261283,
      "loss": 1.4268,
      "step": 42360
    },
    {
      "epoch": 0.0025019093518737984,
      "grad_norm": 0.11762084066867828,
      "learning_rate": 0.0025119327067528213,
      "loss": 1.436,
      "step": 42380
    },
    {
      "epoch": 0.0026335887914461036,
      "grad_norm": 0.11083631217479706,
      "learning_rate": 0.0025114652189272128,
      "loss": 1.4393,
      "step": 42400
    },
    {
      "epoch": 0.002765268231018409,
      "grad_norm": 0.521308183670044,
      "learning_rate": 0.0025109975508677594,
      "loss": 1.4376,
      "step": 42420
    },
    {
      "epoch": 0.002896947670590714,
      "grad_norm": 0.11644919961690903,
      "learning_rate": 0.0025105297026577953,
      "loss": 1.4641,
      "step": 42440
    },
    {
      "epoch": 0.003028627110163019,
      "grad_norm": 0.09482905268669128,
      "learning_rate": 0.002510061674380687,
      "loss": 1.4611,
      "step": 42460
    },
    {
      "epoch": 0.0031603065497353244,
      "grad_norm": 0.14098559319972992,
      "learning_rate": 0.0025095934661198333,
      "loss": 1.4756,
      "step": 42480
    },
    {
      "epoch": 0.0032919859893076296,
      "grad_norm": 0.10550355911254883,
      "learning_rate": 0.002509125077958663,
      "loss": 1.5066,
      "step": 42500
    },
    {
      "epoch": 0.003423665428879935,
      "grad_norm": 0.10194244980812073,
      "learning_rate": 0.00250865650998064,
      "loss": 1.5265,
      "step": 42520
    },
    {
      "epoch": 0.00355534486845224,
      "grad_norm": 0.12069813162088394,
      "learning_rate": 0.0025081877622692584,
      "loss": 1.5486,
      "step": 42540
    },
    {
      "epoch": 0.0036870243080245452,
      "grad_norm": 0.09516333788633347,
      "learning_rate": 0.002507718834908045,
      "loss": 1.5594,
      "step": 42560
    },
    {
      "epoch": 0.0038187037475968504,
      "grad_norm": 0.11514731496572495,
      "learning_rate": 0.002507249727980558,
      "loss": 1.5632,
      "step": 42580
    },
    {
      "epoch": 0.003950383187169155,
      "grad_norm": 0.10155905038118362,
      "learning_rate": 0.002506780441570388,
      "loss": 1.5779,
      "step": 42600
    },
    {
      "epoch": 0.004082062626741461,
      "grad_norm": 0.09767820686101913,
      "learning_rate": 0.002506310975761158,
      "loss": 1.5785,
      "step": 42620
    },
    {
      "epoch": 0.004213742066313766,
      "grad_norm": 0.09538809210062027,
      "learning_rate": 0.0025058413306365216,
      "loss": 1.5823,
      "step": 42640
    },
    {
      "epoch": 0.004345421505886071,
      "grad_norm": 0.09154420346021652,
      "learning_rate": 0.0025053715062801664,
      "loss": 1.5665,
      "step": 42660
    },
    {
      "epoch": 0.004477100945458376,
      "grad_norm": 0.1306425929069519,
      "learning_rate": 0.0025049015027758096,
      "loss": 1.5653,
      "step": 42680
    },
    {
      "epoch": 0.004608780385030682,
      "grad_norm": 0.12897877395153046,
      "learning_rate": 0.002504431320207203,
      "loss": 1.5562,
      "step": 42700
    },
    {
      "epoch": 0.004740459824602986,
      "grad_norm": 0.10444645583629608,
      "learning_rate": 0.002503960958658127,
      "loss": 1.5588,
      "step": 42720
    },
    {
      "epoch": 0.004872139264175292,
      "grad_norm": 0.11443938314914703,
      "learning_rate": 0.0025034904182123976,
      "loss": 1.5489,
      "step": 42740
    },
    {
      "epoch": 0.005003818703747597,
      "grad_norm": 0.11464277654886246,
      "learning_rate": 0.002503019698953859,
      "loss": 1.5547,
      "step": 42760
    },
    {
      "epoch": 0.0051354981433199024,
      "grad_norm": 0.13487665355205536,
      "learning_rate": 0.0025025488009663907,
      "loss": 1.5458,
      "step": 42780
    },
    {
      "epoch": 0.005267177582892207,
      "grad_norm": 0.11822563409805298,
      "learning_rate": 0.002502077724333901,
      "loss": 1.5472,
      "step": 42800
    },
    {
      "epoch": 0.005398857022464512,
      "grad_norm": 0.08794966340065002,
      "learning_rate": 0.002501606469140333,
      "loss": 1.5441,
      "step": 42820
    },
    {
      "epoch": 0.005530536462036818,
      "grad_norm": 0.09283588081598282,
      "learning_rate": 0.0025011350354696583,
      "loss": 1.5405,
      "step": 42840
    },
    {
      "epoch": 0.005662215901609122,
      "grad_norm": 0.11356227099895477,
      "learning_rate": 0.0025006634234058835,
      "loss": 1.534,
      "step": 42860
    },
    {
      "epoch": 0.005793895341181428,
      "grad_norm": 0.10243146866559982,
      "learning_rate": 0.0025001916330330454,
      "loss": 1.5381,
      "step": 42880
    },
    {
      "epoch": 0.005925574780753733,
      "grad_norm": 0.13794657588005066,
      "learning_rate": 0.0024997196644352114,
      "loss": 1.5302,
      "step": 42900
    },
    {
      "epoch": 0.006057254220326038,
      "grad_norm": 0.11851724237203598,
      "learning_rate": 0.0024992475176964834,
      "loss": 1.5267,
      "step": 42920
    },
    {
      "epoch": 0.006188933659898343,
      "grad_norm": 0.1327144056558609,
      "learning_rate": 0.002498775192900993,
      "loss": 1.5292,
      "step": 42940
    },
    {
      "epoch": 0.006320613099470649,
      "grad_norm": 0.09289969503879547,
      "learning_rate": 0.0024983026901329044,
      "loss": 1.5283,
      "step": 42960
    },
    {
      "epoch": 0.006452292539042954,
      "grad_norm": 0.10237723588943481,
      "learning_rate": 0.002497830009476413,
      "loss": 1.5254,
      "step": 42980
    },
    {
      "epoch": 0.006583971978615259,
      "grad_norm": 0.10176314413547516,
      "learning_rate": 0.0024973571510157467,
      "loss": 1.5199,
      "step": 43000
    },
    {
      "epoch": 0.006583971978615259,
      "eval_loss": 1.4149765968322754,
      "eval_runtime": 65.3786,
      "eval_samples_per_second": 15.296,
      "eval_steps_per_second": 15.296,
      "step": 43000
    },
    {
      "epoch": 0.006715651418187564,
      "grad_norm": 0.09919629245996475,
      "learning_rate": 0.0024968841148351635,
      "loss": 1.5191,
      "step": 43020
    },
    {
      "epoch": 0.00684733085775987,
      "grad_norm": 0.10323656350374222,
      "learning_rate": 0.0024964109010189557,
      "loss": 1.5233,
      "step": 43040
    },
    {
      "epoch": 0.006979010297332174,
      "grad_norm": 0.1806810349225998,
      "learning_rate": 0.002495937509651444,
      "loss": 1.5246,
      "step": 43060
    },
    {
      "epoch": 0.00711068973690448,
      "grad_norm": 0.11572176963090897,
      "learning_rate": 0.0024954639408169836,
      "loss": 1.5231,
      "step": 43080
    },
    {
      "epoch": 0.007242369176476785,
      "grad_norm": 0.11550699919462204,
      "learning_rate": 0.0024949901945999593,
      "loss": 1.5218,
      "step": 43100
    },
    {
      "epoch": 0.0073740486160490904,
      "grad_norm": 0.10549616813659668,
      "learning_rate": 0.002494516271084789,
      "loss": 1.5182,
      "step": 43120
    },
    {
      "epoch": 0.007505728055621395,
      "grad_norm": 0.09112010896205902,
      "learning_rate": 0.0024940421703559218,
      "loss": 1.5215,
      "step": 43140
    },
    {
      "epoch": 0.007637407495193701,
      "grad_norm": 0.10110776126384735,
      "learning_rate": 0.002493567892497837,
      "loss": 1.5225,
      "step": 43160
    },
    {
      "epoch": 0.007769086934766006,
      "grad_norm": 0.12719328701496124,
      "learning_rate": 0.002493093437595047,
      "loss": 1.5472,
      "step": 43180
    },
    {
      "epoch": 0.00790076637433831,
      "grad_norm": 0.13183355331420898,
      "learning_rate": 0.002492618805732096,
      "loss": 1.5569,
      "step": 43200
    },
    {
      "epoch": 0.008032445813910615,
      "grad_norm": 0.1020386666059494,
      "learning_rate": 0.0024921439969935585,
      "loss": 1.561,
      "step": 43220
    },
    {
      "epoch": 0.008164125253482922,
      "grad_norm": 0.14664006233215332,
      "learning_rate": 0.002491669011464041,
      "loss": 1.5541,
      "step": 43240
    },
    {
      "epoch": 0.008295804693055226,
      "grad_norm": 0.09918078035116196,
      "learning_rate": 0.0024911938492281813,
      "loss": 1.5547,
      "step": 43260
    },
    {
      "epoch": 0.008427484132627531,
      "grad_norm": 0.10724213719367981,
      "learning_rate": 0.0024907185103706497,
      "loss": 1.5512,
      "step": 43280
    },
    {
      "epoch": 0.008559163572199836,
      "grad_norm": 0.0950016900897026,
      "learning_rate": 0.0024902429949761468,
      "loss": 1.5441,
      "step": 43300
    },
    {
      "epoch": 0.008690843011772142,
      "grad_norm": 0.10287702083587646,
      "learning_rate": 0.002489767303129405,
      "loss": 1.5428,
      "step": 43320
    },
    {
      "epoch": 0.008822522451344447,
      "grad_norm": 0.09938202053308487,
      "learning_rate": 0.0024892914349151887,
      "loss": 1.5466,
      "step": 43340
    },
    {
      "epoch": 0.008954201890916752,
      "grad_norm": 0.09565220773220062,
      "learning_rate": 0.002488815390418293,
      "loss": 1.5379,
      "step": 43360
    },
    {
      "epoch": 0.009085881330489057,
      "grad_norm": 0.13361500203609467,
      "learning_rate": 0.0024883391697235442,
      "loss": 1.5327,
      "step": 43380
    },
    {
      "epoch": 0.009217560770061363,
      "grad_norm": 0.09708066284656525,
      "learning_rate": 0.002487862772915802,
      "loss": 1.5391,
      "step": 43400
    },
    {
      "epoch": 0.009349240209633668,
      "grad_norm": 0.10566909611225128,
      "learning_rate": 0.0024873862000799538,
      "loss": 1.5424,
      "step": 43420
    },
    {
      "epoch": 0.009480919649205973,
      "grad_norm": 0.08183527737855911,
      "learning_rate": 0.002486909451300922,
      "loss": 1.5419,
      "step": 43440
    },
    {
      "epoch": 0.009612599088778278,
      "grad_norm": 0.09500976651906967,
      "learning_rate": 0.0024864325266636587,
      "loss": 1.5366,
      "step": 43460
    },
    {
      "epoch": 0.009744278528350584,
      "grad_norm": 0.11120256781578064,
      "learning_rate": 0.0024859554262531467,
      "loss": 1.5349,
      "step": 43480
    },
    {
      "epoch": 0.009875957967922889,
      "grad_norm": 0.13951987028121948,
      "learning_rate": 0.0024854781501544017,
      "loss": 1.5282,
      "step": 43500
    },
    {
      "epoch": 0.010007637407495194,
      "grad_norm": 0.10958447307348251,
      "learning_rate": 0.0024850006984524693,
      "loss": 1.5272,
      "step": 43520
    },
    {
      "epoch": 0.010139316847067498,
      "grad_norm": 0.09942590445280075,
      "learning_rate": 0.002484523071232428,
      "loss": 1.5326,
      "step": 43540
    },
    {
      "epoch": 0.010270996286639805,
      "grad_norm": 0.1157267838716507,
      "learning_rate": 0.002484045268579386,
      "loss": 1.5296,
      "step": 43560
    },
    {
      "epoch": 0.01040267572621211,
      "grad_norm": 0.1331472247838974,
      "learning_rate": 0.0024835672905784835,
      "loss": 1.5249,
      "step": 43580
    },
    {
      "epoch": 0.010534355165784414,
      "grad_norm": 0.13041837513446808,
      "learning_rate": 0.0024830891373148916,
      "loss": 1.5287,
      "step": 43600
    },
    {
      "epoch": 0.01066603460535672,
      "grad_norm": 0.13223400712013245,
      "learning_rate": 0.0024826108088738122,
      "loss": 1.5229,
      "step": 43620
    },
    {
      "epoch": 0.010797714044929024,
      "grad_norm": 0.11160468310117722,
      "learning_rate": 0.002482132305340481,
      "loss": 1.5257,
      "step": 43640
    },
    {
      "epoch": 0.01092939348450133,
      "grad_norm": 0.1368405967950821,
      "learning_rate": 0.002481653626800161,
      "loss": 1.5232,
      "step": 43660
    },
    {
      "epoch": 0.011061072924073635,
      "grad_norm": 0.11044129729270935,
      "learning_rate": 0.002481174773338149,
      "loss": 1.5258,
      "step": 43680
    },
    {
      "epoch": 0.01119275236364594,
      "grad_norm": 0.11385578662157059,
      "learning_rate": 0.002480695745039773,
      "loss": 1.5252,
      "step": 43700
    },
    {
      "epoch": 0.011324431803218245,
      "grad_norm": 0.13926897943019867,
      "learning_rate": 0.00248021654199039,
      "loss": 1.5145,
      "step": 43720
    },
    {
      "epoch": 0.011456111242790551,
      "grad_norm": 0.10542988032102585,
      "learning_rate": 0.002479737164275391,
      "loss": 1.518,
      "step": 43740
    },
    {
      "epoch": 0.011587790682362856,
      "grad_norm": 0.19769559800624847,
      "learning_rate": 0.002479257611980196,
      "loss": 1.5328,
      "step": 43760
    },
    {
      "epoch": 0.01171947012193516,
      "grad_norm": 0.14825715124607086,
      "learning_rate": 0.002478777885190257,
      "loss": 1.5437,
      "step": 43780
    },
    {
      "epoch": 0.011851149561507466,
      "grad_norm": 0.12765994668006897,
      "learning_rate": 0.002478297983991057,
      "loss": 1.5491,
      "step": 43800
    },
    {
      "epoch": 0.011982829001079772,
      "grad_norm": 0.12340036034584045,
      "learning_rate": 0.00247781790846811,
      "loss": 1.5623,
      "step": 43820
    },
    {
      "epoch": 0.012114508440652077,
      "grad_norm": 0.12714795768260956,
      "learning_rate": 0.002477337658706961,
      "loss": 1.5452,
      "step": 43840
    },
    {
      "epoch": 0.012246187880224382,
      "grad_norm": 0.10861831903457642,
      "learning_rate": 0.0024768572347931856,
      "loss": 1.5451,
      "step": 43860
    },
    {
      "epoch": 0.012377867319796686,
      "grad_norm": 0.09523173421621323,
      "learning_rate": 0.002476376636812392,
      "loss": 1.5425,
      "step": 43880
    },
    {
      "epoch": 0.012509546759368993,
      "grad_norm": 0.09822442382574081,
      "learning_rate": 0.002475895864850217,
      "loss": 1.545,
      "step": 43900
    },
    {
      "epoch": 0.012641226198941298,
      "grad_norm": 0.11365661770105362,
      "learning_rate": 0.0024754149189923315,
      "loss": 1.5376,
      "step": 43920
    },
    {
      "epoch": 0.012772905638513602,
      "grad_norm": 0.10581135004758835,
      "learning_rate": 0.002474933799324434,
      "loss": 1.5364,
      "step": 43940
    },
    {
      "epoch": 0.012904585078085907,
      "grad_norm": 0.11580084264278412,
      "learning_rate": 0.002474452505932257,
      "loss": 1.5321,
      "step": 43960
    },
    {
      "epoch": 0.013036264517658214,
      "grad_norm": 0.08592469990253448,
      "learning_rate": 0.0024739710389015614,
      "loss": 1.5331,
      "step": 43980
    },
    {
      "epoch": 0.013167943957230518,
      "grad_norm": 0.10360794514417648,
      "learning_rate": 0.0024734893983181408,
      "loss": 1.5342,
      "step": 44000
    },
    {
      "epoch": 0.013167943957230518,
      "eval_loss": 1.4869463443756104,
      "eval_runtime": 63.1265,
      "eval_samples_per_second": 15.841,
      "eval_steps_per_second": 15.841,
      "step": 44000
    },
    {
      "epoch": 0.013299623396802823,
      "grad_norm": 0.11372421681880951,
      "learning_rate": 0.002473007584267819,
      "loss": 1.5213,
      "step": 44020
    },
    {
      "epoch": 0.013431302836375128,
      "grad_norm": 0.1096460372209549,
      "learning_rate": 0.002472525596836451,
      "loss": 1.5294,
      "step": 44040
    },
    {
      "epoch": 0.013562982275947433,
      "grad_norm": 0.12670129537582397,
      "learning_rate": 0.0024720434361099226,
      "loss": 1.5169,
      "step": 44060
    },
    {
      "epoch": 0.01369466171551974,
      "grad_norm": 0.09985670447349548,
      "learning_rate": 0.00247156110217415,
      "loss": 1.513,
      "step": 44080
    },
    {
      "epoch": 0.013826341155092044,
      "grad_norm": 0.1046002060174942,
      "learning_rate": 0.002471078595115081,
      "loss": 1.5224,
      "step": 44100
    },
    {
      "epoch": 0.013958020594664349,
      "grad_norm": 0.15750165283679962,
      "learning_rate": 0.0024705959150186944,
      "loss": 1.5242,
      "step": 44120
    },
    {
      "epoch": 0.014089700034236654,
      "grad_norm": 0.1028977781534195,
      "learning_rate": 0.0024701130619709986,
      "loss": 1.5234,
      "step": 44140
    },
    {
      "epoch": 0.01422137947380896,
      "grad_norm": 0.16695408523082733,
      "learning_rate": 0.0024696300360580337,
      "loss": 1.522,
      "step": 44160
    },
    {
      "epoch": 0.014353058913381265,
      "grad_norm": 0.12498366087675095,
      "learning_rate": 0.0024691468373658713,
      "loss": 1.5188,
      "step": 44180
    },
    {
      "epoch": 0.01448473835295357,
      "grad_norm": 0.09633167088031769,
      "learning_rate": 0.002468663465980612,
      "loss": 1.5245,
      "step": 44200
    },
    {
      "epoch": 0.014616417792525874,
      "grad_norm": 0.1630498468875885,
      "learning_rate": 0.002468179921988389,
      "loss": 1.5192,
      "step": 44220
    },
    {
      "epoch": 0.014748097232098181,
      "grad_norm": 0.11872150003910065,
      "learning_rate": 0.0024676962054753643,
      "loss": 1.5278,
      "step": 44240
    },
    {
      "epoch": 0.014879776671670486,
      "grad_norm": 0.17083382606506348,
      "learning_rate": 0.0024672123165277323,
      "loss": 1.5212,
      "step": 44260
    },
    {
      "epoch": 0.01501145611124279,
      "grad_norm": 0.12562261521816254,
      "learning_rate": 0.002466728255231718,
      "loss": 1.5217,
      "step": 44280
    },
    {
      "epoch": 0.015143135550815095,
      "grad_norm": 0.11627357453107834,
      "learning_rate": 0.002466244021673577,
      "loss": 1.521,
      "step": 44300
    },
    {
      "epoch": 0.015274814990387402,
      "grad_norm": 0.10252583771944046,
      "learning_rate": 0.0024657596159395937,
      "loss": 1.5237,
      "step": 44320
    },
    {
      "epoch": 0.015406494429959706,
      "grad_norm": 0.1185808777809143,
      "learning_rate": 0.0024652750381160865,
      "loss": 1.5174,
      "step": 44340
    },
    {
      "epoch": 0.015538173869532011,
      "grad_norm": 0.0919899195432663,
      "learning_rate": 0.002464790288289401,
      "loss": 1.5184,
      "step": 44360
    },
    {
      "epoch": 0.015669853309104316,
      "grad_norm": 0.16310234367847443,
      "learning_rate": 0.0024643053665459173,
      "loss": 1.5144,
      "step": 44380
    },
    {
      "epoch": 0.01580153274867662,
      "grad_norm": 0.16470354795455933,
      "learning_rate": 0.0024638202729720424,
      "loss": 1.4866,
      "step": 44400
    },
    {
      "epoch": 0.015933212188248926,
      "grad_norm": 0.12554654479026794,
      "learning_rate": 0.0024633350076542155,
      "loss": 1.4843,
      "step": 44420
    },
    {
      "epoch": 0.01606489162782123,
      "grad_norm": 0.13595227897167206,
      "learning_rate": 0.0024628495706789074,
      "loss": 1.4784,
      "step": 44440
    },
    {
      "epoch": 0.01619657106739354,
      "grad_norm": 0.10951948165893555,
      "learning_rate": 0.002462363962132618,
      "loss": 1.4678,
      "step": 44460
    },
    {
      "epoch": 0.016328250506965843,
      "grad_norm": 0.14208072423934937,
      "learning_rate": 0.0024618781821018785,
      "loss": 1.4665,
      "step": 44480
    },
    {
      "epoch": 0.016459929946538148,
      "grad_norm": 0.10121913254261017,
      "learning_rate": 0.0024613922306732506,
      "loss": 1.4681,
      "step": 44500
    },
    {
      "epoch": 0.016591609386110453,
      "grad_norm": 0.11362706124782562,
      "learning_rate": 0.0024609061079333256,
      "loss": 1.4584,
      "step": 44520
    },
    {
      "epoch": 0.016723288825682758,
      "grad_norm": 0.14318877458572388,
      "learning_rate": 0.0024604198139687268,
      "loss": 1.4655,
      "step": 44540
    },
    {
      "epoch": 0.016854968265255062,
      "grad_norm": 0.12110181897878647,
      "learning_rate": 0.002459933348866107,
      "loss": 1.4528,
      "step": 44560
    },
    {
      "epoch": 0.016986647704827367,
      "grad_norm": 0.11001115292310715,
      "learning_rate": 0.0024594467127121507,
      "loss": 1.4501,
      "step": 44580
    },
    {
      "epoch": 0.017118327144399672,
      "grad_norm": 0.16570629179477692,
      "learning_rate": 0.0024589599055935708,
      "loss": 1.4536,
      "step": 44600
    },
    {
      "epoch": 0.01725000658397198,
      "grad_norm": 0.11245054006576538,
      "learning_rate": 0.002458472927597112,
      "loss": 1.4548,
      "step": 44620
    },
    {
      "epoch": 0.017381686023544285,
      "grad_norm": 0.13176459074020386,
      "learning_rate": 0.0024579857788095505,
      "loss": 1.4485,
      "step": 44640
    },
    {
      "epoch": 0.01751336546311659,
      "grad_norm": 0.09991902112960815,
      "learning_rate": 0.002457498459317691,
      "loss": 1.4534,
      "step": 44660
    },
    {
      "epoch": 0.017645044902688894,
      "grad_norm": 0.0981200635433197,
      "learning_rate": 0.002457010969208369,
      "loss": 1.4469,
      "step": 44680
    },
    {
      "epoch": 0.0177767243422612,
      "grad_norm": 0.11083146929740906,
      "learning_rate": 0.0024565233085684507,
      "loss": 1.442,
      "step": 44700
    },
    {
      "epoch": 0.017908403781833504,
      "grad_norm": 0.09444429725408554,
      "learning_rate": 0.002456035477484834,
      "loss": 1.4514,
      "step": 44720
    },
    {
      "epoch": 0.01804008322140581,
      "grad_norm": 0.11763214319944382,
      "learning_rate": 0.0024555474760444445,
      "loss": 1.4521,
      "step": 44740
    },
    {
      "epoch": 0.018171762660978114,
      "grad_norm": 0.12316908687353134,
      "learning_rate": 0.00245505930433424,
      "loss": 1.449,
      "step": 44760
    },
    {
      "epoch": 0.018303442100550422,
      "grad_norm": 0.1016848236322403,
      "learning_rate": 0.0024545709624412093,
      "loss": 1.4494,
      "step": 44780
    },
    {
      "epoch": 0.018435121540122727,
      "grad_norm": 0.10764171183109283,
      "learning_rate": 0.002454082450452369,
      "loss": 1.4399,
      "step": 44800
    },
    {
      "epoch": 0.01856680097969503,
      "grad_norm": 0.11485065519809723,
      "learning_rate": 0.0024535937684547686,
      "loss": 1.4497,
      "step": 44820
    },
    {
      "epoch": 0.018698480419267336,
      "grad_norm": 0.1278896927833557,
      "learning_rate": 0.0024531049165354853,
      "loss": 1.4439,
      "step": 44840
    },
    {
      "epoch": 0.01883015985883964,
      "grad_norm": 0.14847633242607117,
      "learning_rate": 0.0024526158947816294,
      "loss": 1.4523,
      "step": 44860
    },
    {
      "epoch": 0.018961839298411946,
      "grad_norm": 0.105763278901577,
      "learning_rate": 0.00245212670328034,
      "loss": 1.4496,
      "step": 44880
    },
    {
      "epoch": 0.01909351873798425,
      "grad_norm": 0.1055736318230629,
      "learning_rate": 0.0024516373421187856,
      "loss": 1.446,
      "step": 44900
    },
    {
      "epoch": 0.019225198177556555,
      "grad_norm": 0.0994647964835167,
      "learning_rate": 0.002451147811384167,
      "loss": 1.4465,
      "step": 44920
    },
    {
      "epoch": 0.01935687761712886,
      "grad_norm": 0.10782650858163834,
      "learning_rate": 0.0024506581111637137,
      "loss": 1.4587,
      "step": 44940
    },
    {
      "epoch": 0.019488557056701168,
      "grad_norm": 0.12616658210754395,
      "learning_rate": 0.0024501682415446853,
      "loss": 1.496,
      "step": 44960
    },
    {
      "epoch": 0.019620236496273473,
      "grad_norm": 0.12615470588207245,
      "learning_rate": 0.002449678202614372,
      "loss": 1.5177,
      "step": 44980
    },
    {
      "epoch": 0.019751915935845778,
      "grad_norm": 0.10382698476314545,
      "learning_rate": 0.0024491879944600957,
      "loss": 1.5299,
      "step": 45000
    },
    {
      "epoch": 0.019751915935845778,
      "eval_loss": 1.5385295152664185,
      "eval_runtime": 66.2896,
      "eval_samples_per_second": 15.085,
      "eval_steps_per_second": 15.085,
      "step": 45000
    },
    {
      "epoch": 0.019883595375418082,
      "grad_norm": 0.11794433742761612,
      "learning_rate": 0.0024486976171692056,
      "loss": 1.528,
      "step": 45020
    },
    {
      "epoch": 0.020015274814990387,
      "grad_norm": 0.1409338116645813,
      "learning_rate": 0.002448207070829083,
      "loss": 1.5341,
      "step": 45040
    },
    {
      "epoch": 0.020146954254562692,
      "grad_norm": 0.12568840384483337,
      "learning_rate": 0.002447716355527138,
      "loss": 1.5318,
      "step": 45060
    },
    {
      "epoch": 0.020278633694134997,
      "grad_norm": 0.15237917006015778,
      "learning_rate": 0.0024472254713508133,
      "loss": 1.5449,
      "step": 45080
    },
    {
      "epoch": 0.0204103131337073,
      "grad_norm": 0.11090464144945145,
      "learning_rate": 0.002446734418387578,
      "loss": 1.5253,
      "step": 45100
    },
    {
      "epoch": 0.02054199257327961,
      "grad_norm": 0.11170685291290283,
      "learning_rate": 0.0024462431967249344,
      "loss": 1.5203,
      "step": 45120
    },
    {
      "epoch": 0.020673672012851915,
      "grad_norm": 0.10791482031345367,
      "learning_rate": 0.0024457518064504133,
      "loss": 1.5176,
      "step": 45140
    },
    {
      "epoch": 0.02080535145242422,
      "grad_norm": 0.09597095102071762,
      "learning_rate": 0.002445260247651576,
      "loss": 1.5193,
      "step": 45160
    },
    {
      "epoch": 0.020937030891996524,
      "grad_norm": 0.34346190094947815,
      "learning_rate": 0.0024447685204160145,
      "loss": 1.5182,
      "step": 45180
    },
    {
      "epoch": 0.02106871033156883,
      "grad_norm": 0.15261825919151306,
      "learning_rate": 0.0024442766248313494,
      "loss": 1.5151,
      "step": 45200
    },
    {
      "epoch": 0.021200389771141134,
      "grad_norm": 0.1070566177368164,
      "learning_rate": 0.0024437845609852314,
      "loss": 1.516,
      "step": 45220
    },
    {
      "epoch": 0.02133206921071344,
      "grad_norm": 0.10885826498270035,
      "learning_rate": 0.0024432923289653425,
      "loss": 1.5128,
      "step": 45240
    },
    {
      "epoch": 0.021463748650285743,
      "grad_norm": 0.13303355872631073,
      "learning_rate": 0.002442799928859394,
      "loss": 1.5122,
      "step": 45260
    },
    {
      "epoch": 0.021595428089858048,
      "grad_norm": 0.10217521339654922,
      "learning_rate": 0.002442307360755127,
      "loss": 1.5032,
      "step": 45280
    },
    {
      "epoch": 0.021727107529430356,
      "grad_norm": 0.11426650732755661,
      "learning_rate": 0.0024418146247403125,
      "loss": 1.5029,
      "step": 45300
    },
    {
      "epoch": 0.02185878696900266,
      "grad_norm": 0.09248542040586472,
      "learning_rate": 0.0024413217209027518,
      "loss": 1.4998,
      "step": 45320
    },
    {
      "epoch": 0.021990466408574966,
      "grad_norm": 0.12633563578128815,
      "learning_rate": 0.0024408286493302757,
      "loss": 1.4982,
      "step": 45340
    },
    {
      "epoch": 0.02212214584814727,
      "grad_norm": 0.16023258864879608,
      "learning_rate": 0.0024403354101107445,
      "loss": 1.5003,
      "step": 45360
    },
    {
      "epoch": 0.022253825287719575,
      "grad_norm": 0.12979178130626678,
      "learning_rate": 0.00243984200333205,
      "loss": 1.4963,
      "step": 45380
    },
    {
      "epoch": 0.02238550472729188,
      "grad_norm": 0.127953439950943,
      "learning_rate": 0.002439348429082112,
      "loss": 1.4973,
      "step": 45400
    },
    {
      "epoch": 0.022517184166864185,
      "grad_norm": 0.10096823424100876,
      "learning_rate": 0.002438854687448881,
      "loss": 1.4964,
      "step": 45420
    },
    {
      "epoch": 0.02264886360643649,
      "grad_norm": 0.1225760206580162,
      "learning_rate": 0.0024383607785203377,
      "loss": 1.5038,
      "step": 45440
    },
    {
      "epoch": 0.022780543046008798,
      "grad_norm": 0.15798920392990112,
      "learning_rate": 0.002437866702384491,
      "loss": 1.5002,
      "step": 45460
    },
    {
      "epoch": 0.022912222485581103,
      "grad_norm": 0.11615593731403351,
      "learning_rate": 0.0024373724591293827,
      "loss": 1.4992,
      "step": 45480
    },
    {
      "epoch": 0.023043901925153407,
      "grad_norm": 0.11516784876585007,
      "learning_rate": 0.0024368780488430806,
      "loss": 1.4973,
      "step": 45500
    },
    {
      "epoch": 0.023175581364725712,
      "grad_norm": 0.13365384936332703,
      "learning_rate": 0.002436383471613685,
      "loss": 1.4992,
      "step": 45520
    },
    {
      "epoch": 0.023307260804298017,
      "grad_norm": 0.13123133778572083,
      "learning_rate": 0.0024358887275293243,
      "loss": 1.5064,
      "step": 45540
    },
    {
      "epoch": 0.02343894024387032,
      "grad_norm": 0.106318898499012,
      "learning_rate": 0.0024353938166781584,
      "loss": 1.5163,
      "step": 45560
    },
    {
      "epoch": 0.023570619683442626,
      "grad_norm": 0.10438498854637146,
      "learning_rate": 0.0024348987391483753,
      "loss": 1.5152,
      "step": 45580
    },
    {
      "epoch": 0.02370229912301493,
      "grad_norm": 0.1130092591047287,
      "learning_rate": 0.0024344034950281933,
      "loss": 1.519,
      "step": 45600
    },
    {
      "epoch": 0.023833978562587236,
      "grad_norm": 0.12775199115276337,
      "learning_rate": 0.0024339080844058606,
      "loss": 1.5139,
      "step": 45620
    },
    {
      "epoch": 0.023965658002159544,
      "grad_norm": 0.09360072016716003,
      "learning_rate": 0.002433412507369654,
      "loss": 1.5066,
      "step": 45640
    },
    {
      "epoch": 0.02409733744173185,
      "grad_norm": 0.1027098149061203,
      "learning_rate": 0.002432916764007882,
      "loss": 1.5034,
      "step": 45660
    },
    {
      "epoch": 0.024229016881304154,
      "grad_norm": 0.11048934608697891,
      "learning_rate": 0.0024324208544088803,
      "loss": 1.4977,
      "step": 45680
    },
    {
      "epoch": 0.02436069632087646,
      "grad_norm": 0.11964282393455505,
      "learning_rate": 0.002431924778661017,
      "loss": 1.4941,
      "step": 45700
    },
    {
      "epoch": 0.024492375760448763,
      "grad_norm": 0.12463732808828354,
      "learning_rate": 0.002431428536852687,
      "loss": 1.4987,
      "step": 45720
    },
    {
      "epoch": 0.024624055200021068,
      "grad_norm": 0.13735800981521606,
      "learning_rate": 0.0024309321290723164,
      "loss": 1.4951,
      "step": 45740
    },
    {
      "epoch": 0.024755734639593373,
      "grad_norm": 0.11579196900129318,
      "learning_rate": 0.0024304355554083607,
      "loss": 1.4909,
      "step": 45760
    },
    {
      "epoch": 0.024887414079165678,
      "grad_norm": 0.1210319772362709,
      "learning_rate": 0.002429938815949304,
      "loss": 1.49,
      "step": 45780
    },
    {
      "epoch": 0.025019093518737986,
      "grad_norm": 0.11229152232408524,
      "learning_rate": 0.0024294419107836625,
      "loss": 1.4849,
      "step": 45800
    },
    {
      "epoch": 0.02515077295831029,
      "grad_norm": 0.09796669334173203,
      "learning_rate": 0.002428944839999978,
      "loss": 1.4831,
      "step": 45820
    },
    {
      "epoch": 0.025282452397882595,
      "grad_norm": 0.12028352171182632,
      "learning_rate": 0.0024284476036868256,
      "loss": 1.4836,
      "step": 45840
    },
    {
      "epoch": 0.0254141318374549,
      "grad_norm": 0.11071626096963882,
      "learning_rate": 0.002427950201932807,
      "loss": 1.4798,
      "step": 45860
    },
    {
      "epoch": 0.025545811277027205,
      "grad_norm": 0.10366615653038025,
      "learning_rate": 0.002427452634826556,
      "loss": 1.4781,
      "step": 45880
    },
    {
      "epoch": 0.02567749071659951,
      "grad_norm": 0.10610003024339676,
      "learning_rate": 0.0024269549024567333,
      "loss": 1.4838,
      "step": 45900
    },
    {
      "epoch": 0.025809170156171814,
      "grad_norm": 0.0969921201467514,
      "learning_rate": 0.002426457004912031,
      "loss": 1.4874,
      "step": 45920
    },
    {
      "epoch": 0.02594084959574412,
      "grad_norm": 0.11021365970373154,
      "learning_rate": 0.00242595894228117,
      "loss": 1.4853,
      "step": 45940
    },
    {
      "epoch": 0.026072529035316427,
      "grad_norm": 0.11574053019285202,
      "learning_rate": 0.0024254607146528993,
      "loss": 1.4812,
      "step": 45960
    },
    {
      "epoch": 0.026204208474888732,
      "grad_norm": 0.09534565359354019,
      "learning_rate": 0.002424962322116,
      "loss": 1.478,
      "step": 45980
    },
    {
      "epoch": 0.026335887914461037,
      "grad_norm": 0.10188112407922745,
      "learning_rate": 0.0024244637647592797,
      "loss": 1.4782,
      "step": 46000
    },
    {
      "epoch": 0.026335887914461037,
      "eval_loss": 1.558879017829895,
      "eval_runtime": 77.2288,
      "eval_samples_per_second": 12.949,
      "eval_steps_per_second": 12.949,
      "step": 46000
    },
    {
      "epoch": 2e-06,
      "grad_norm": 0.1454385668039322,
      "learning_rate": 0.000690285,
      "loss": 2.076281929016113,
      "step": 46020
    },
    {
      "epoch": 4e-06,
      "grad_norm": 0.07851192355155945,
      "learning_rate": 0.0006905850000000001,
      "loss": 2.036176300048828,
      "step": 46040
    },
    {
      "epoch": 6e-06,
      "grad_norm": 0.057373788207769394,
      "learning_rate": 0.000690885,
      "loss": 2.018076515197754,
      "step": 46060
    },
    {
      "epoch": 8e-06,
      "grad_norm": 0.058911461383104324,
      "learning_rate": 0.000691185,
      "loss": 2.0062496185302736,
      "step": 46080
    },
    {
      "epoch": 1e-05,
      "grad_norm": 0.08635017275810242,
      "learning_rate": 0.000691485,
      "loss": 2.0066757202148438,
      "step": 46100
    },
    {
      "epoch": 1.2e-05,
      "grad_norm": 0.0559137687087059,
      "learning_rate": 0.000691785,
      "loss": 1.999479103088379,
      "step": 46120
    },
    {
      "epoch": 1.4e-05,
      "grad_norm": 0.07066166400909424,
      "learning_rate": 0.000692085,
      "loss": 1.9976621627807618,
      "step": 46140
    },
    {
      "epoch": 1.6e-05,
      "grad_norm": 0.06597048789262772,
      "learning_rate": 0.000692385,
      "loss": 1.9841796875,
      "step": 46160
    },
    {
      "epoch": 1.8e-05,
      "grad_norm": 0.07282619923353195,
      "learning_rate": 0.000692685,
      "loss": 1.9873577117919923,
      "step": 46180
    },
    {
      "epoch": 2e-05,
      "grad_norm": 0.06868898123502731,
      "learning_rate": 0.000692985,
      "loss": 1.9878196716308594,
      "step": 46200
    },
    {
      "epoch": 2.2e-05,
      "grad_norm": 0.056995589286088943,
      "learning_rate": 0.000693285,
      "loss": 1.9771448135375977,
      "step": 46220
    },
    {
      "epoch": 2.4e-05,
      "grad_norm": 0.0641326829791069,
      "learning_rate": 0.000693585,
      "loss": 1.9761754989624023,
      "step": 46240
    },
    {
      "epoch": 2.6e-05,
      "grad_norm": 0.07175535708665848,
      "learning_rate": 0.000693885,
      "loss": 1.970853042602539,
      "step": 46260
    },
    {
      "epoch": 2.8e-05,
      "grad_norm": 0.06924993544816971,
      "learning_rate": 0.0006941849999999999,
      "loss": 1.9718612670898437,
      "step": 46280
    },
    {
      "epoch": 3e-05,
      "grad_norm": 0.06896559149026871,
      "learning_rate": 0.000694485,
      "loss": 1.9643774032592773,
      "step": 46300
    },
    {
      "epoch": 3.2e-05,
      "grad_norm": 0.058981869369745255,
      "learning_rate": 0.000694785,
      "loss": 1.9701652526855469,
      "step": 46320
    },
    {
      "epoch": 3.4e-05,
      "grad_norm": 0.06701400130987167,
      "learning_rate": 0.000695085,
      "loss": 1.957753562927246,
      "step": 46340
    },
    {
      "epoch": 3.6e-05,
      "grad_norm": 0.07226961851119995,
      "learning_rate": 0.000695385,
      "loss": 1.9627220153808593,
      "step": 46360
    },
    {
      "epoch": 3.8e-05,
      "grad_norm": 0.07805868238210678,
      "learning_rate": 0.000695685,
      "loss": 1.9605262756347657,
      "step": 46380
    },
    {
      "epoch": 4e-05,
      "grad_norm": 0.07272197306156158,
      "learning_rate": 0.000695985,
      "loss": 1.9611610412597655,
      "step": 46400
    },
    {
      "epoch": 4.2e-05,
      "grad_norm": 0.06122796609997749,
      "learning_rate": 0.000696285,
      "loss": 1.947123146057129,
      "step": 46420
    },
    {
      "epoch": 4.4e-05,
      "grad_norm": 0.08275068551301956,
      "learning_rate": 0.000696585,
      "loss": 1.9539257049560548,
      "step": 46440
    },
    {
      "epoch": 4.6e-05,
      "grad_norm": 0.0847841426730156,
      "learning_rate": 0.0006968850000000001,
      "loss": 1.9514825820922852,
      "step": 46460
    },
    {
      "epoch": 4.8e-05,
      "grad_norm": 0.05611637979745865,
      "learning_rate": 0.000697185,
      "loss": 1.948338508605957,
      "step": 46480
    },
    {
      "epoch": 5e-05,
      "grad_norm": 0.06900504976511002,
      "learning_rate": 0.0006974850000000001,
      "loss": 1.9451566696166993,
      "step": 46500
    },
    {
      "epoch": 5.2e-05,
      "grad_norm": 0.07195140421390533,
      "learning_rate": 0.000697785,
      "loss": 1.9498029708862306,
      "step": 46520
    },
    {
      "epoch": 5.4e-05,
      "grad_norm": 0.07324763387441635,
      "learning_rate": 0.0006980850000000001,
      "loss": 1.9459251403808593,
      "step": 46540
    },
    {
      "epoch": 5.6e-05,
      "grad_norm": 0.056651510298252106,
      "learning_rate": 0.000698385,
      "loss": 1.9457489013671876,
      "step": 46560
    },
    {
      "epoch": 5.8e-05,
      "grad_norm": 0.08276001363992691,
      "learning_rate": 0.000698685,
      "loss": 1.9360223770141602,
      "step": 46580
    },
    {
      "epoch": 6e-05,
      "grad_norm": 0.08096142113208771,
      "learning_rate": 0.0006989850000000001,
      "loss": 1.9375598907470704,
      "step": 46600
    },
    {
      "epoch": 6.2e-05,
      "grad_norm": 0.0721917450428009,
      "learning_rate": 0.000699285,
      "loss": 1.9336822509765625,
      "step": 46620
    },
    {
      "epoch": 6.4e-05,
      "grad_norm": 0.0872625857591629,
      "learning_rate": 0.0006995850000000001,
      "loss": 1.9410781860351562,
      "step": 46640
    },
    {
      "epoch": 6.6e-05,
      "grad_norm": 0.05740909278392792,
      "learning_rate": 0.000699885,
      "loss": 1.9338249206542968,
      "step": 46660
    },
    {
      "epoch": 6.8e-05,
      "grad_norm": 0.06273666769266129,
      "learning_rate": 0.000700185,
      "loss": 1.9296646118164062,
      "step": 46680
    },
    {
      "epoch": 7e-05,
      "grad_norm": 0.07139860838651657,
      "learning_rate": 0.000700485,
      "loss": 1.941641616821289,
      "step": 46700
    },
    {
      "epoch": 7.2e-05,
      "grad_norm": 0.07698002457618713,
      "learning_rate": 0.000700785,
      "loss": 1.9317914962768554,
      "step": 46720
    },
    {
      "epoch": 7.4e-05,
      "grad_norm": 0.062162358313798904,
      "learning_rate": 0.0007010850000000001,
      "loss": 1.9283344268798828,
      "step": 46740
    },
    {
      "epoch": 7.6e-05,
      "grad_norm": 0.0760132372379303,
      "learning_rate": 0.000701385,
      "loss": 1.9273748397827148,
      "step": 46760
    },
    {
      "epoch": 7.8e-05,
      "grad_norm": 0.09244687855243683,
      "learning_rate": 0.000701685,
      "loss": 1.9315763473510743,
      "step": 46780
    },
    {
      "epoch": 8e-05,
      "grad_norm": 0.06585084646940231,
      "learning_rate": 0.000701985,
      "loss": 1.9171140670776368,
      "step": 46800
    },
    {
      "epoch": 8.2e-05,
      "grad_norm": 0.07399197667837143,
      "learning_rate": 0.000702285,
      "loss": 1.9259872436523438,
      "step": 46820
    },
    {
      "epoch": 8.4e-05,
      "grad_norm": 0.0975058525800705,
      "learning_rate": 0.0007025849999999999,
      "loss": 1.9262987136840821,
      "step": 46840
    },
    {
      "epoch": 8.6e-05,
      "grad_norm": 0.062050431966781616,
      "learning_rate": 0.000702885,
      "loss": 1.9207113265991211,
      "step": 46860
    },
    {
      "epoch": 8.8e-05,
      "grad_norm": 0.06593676656484604,
      "learning_rate": 0.0007031849999999999,
      "loss": 1.9189161300659179,
      "step": 46880
    },
    {
      "epoch": 9e-05,
      "grad_norm": 0.0780588909983635,
      "learning_rate": 0.000703485,
      "loss": 1.9223506927490235,
      "step": 46900
    },
    {
      "epoch": 9.2e-05,
      "grad_norm": 0.06823083758354187,
      "learning_rate": 0.000703785,
      "loss": 1.912630844116211,
      "step": 46920
    },
    {
      "epoch": 9.4e-05,
      "grad_norm": 0.08263996988534927,
      "learning_rate": 0.000704085,
      "loss": 1.9180915832519532,
      "step": 46940
    },
    {
      "epoch": 9.6e-05,
      "grad_norm": 0.06720611453056335,
      "learning_rate": 0.000704385,
      "loss": 1.9188003540039062,
      "step": 46960
    },
    {
      "epoch": 9.8e-05,
      "grad_norm": 0.09634207934141159,
      "learning_rate": 0.000704685,
      "loss": 1.9179636001586915,
      "step": 46980
    },
    {
      "epoch": 0.0001,
      "grad_norm": 0.1031593307852745,
      "learning_rate": 0.0007049850000000001,
      "loss": 1.9118635177612304,
      "step": 47000
    },
    {
      "epoch": 0.0001,
      "eval_loss": 1.8519573211669922,
      "eval_runtime": 103.0547,
      "eval_samples_per_second": 9.704,
      "eval_steps_per_second": 9.704,
      "step": 47000
    },
    {
      "epoch": 0.000102,
      "grad_norm": 0.06469423323869705,
      "learning_rate": 0.000705285,
      "loss": 1.9127388000488281,
      "step": 47020
    },
    {
      "epoch": 0.000104,
      "grad_norm": 0.0648200586438179,
      "learning_rate": 0.000705585,
      "loss": 1.9220613479614257,
      "step": 47040
    },
    {
      "epoch": 0.000106,
      "grad_norm": 0.08908385783433914,
      "learning_rate": 0.0007058850000000001,
      "loss": 1.9115385055541991,
      "step": 47060
    },
    {
      "epoch": 0.000108,
      "grad_norm": 0.0783170759677887,
      "learning_rate": 0.000706185,
      "loss": 1.9148534774780273,
      "step": 47080
    },
    {
      "epoch": 0.00011,
      "grad_norm": 0.08982904255390167,
      "learning_rate": 0.0007064850000000001,
      "loss": 1.9030448913574218,
      "step": 47100
    },
    {
      "epoch": 0.000112,
      "grad_norm": 0.0841226652264595,
      "learning_rate": 0.000706785,
      "loss": 1.912649154663086,
      "step": 47120
    },
    {
      "epoch": 0.000114,
      "grad_norm": 0.08642134070396423,
      "learning_rate": 0.000707085,
      "loss": 1.9114418029785156,
      "step": 47140
    },
    {
      "epoch": 0.000116,
      "grad_norm": 0.06373965740203857,
      "learning_rate": 0.000707385,
      "loss": 1.9125936508178711,
      "step": 47160
    },
    {
      "epoch": 0.000118,
      "grad_norm": 0.07079387456178665,
      "learning_rate": 0.000707685,
      "loss": 1.9213525772094726,
      "step": 47180
    },
    {
      "epoch": 0.00012,
      "grad_norm": 0.07068858295679092,
      "learning_rate": 0.0007079850000000001,
      "loss": 1.9196317672729493,
      "step": 47200
    },
    {
      "epoch": 0.000122,
      "grad_norm": 0.0658358484506607,
      "learning_rate": 0.000708285,
      "loss": 1.928008460998535,
      "step": 47220
    },
    {
      "epoch": 0.000124,
      "grad_norm": 0.07867607474327087,
      "learning_rate": 0.000708585,
      "loss": 1.9092967987060547,
      "step": 47240
    },
    {
      "epoch": 0.000126,
      "grad_norm": 0.0598660483956337,
      "learning_rate": 0.000708885,
      "loss": 1.914600944519043,
      "step": 47260
    },
    {
      "epoch": 0.000128,
      "grad_norm": 0.06065090373158455,
      "learning_rate": 0.000709185,
      "loss": 1.9233844757080079,
      "step": 47280
    },
    {
      "epoch": 0.00013,
      "grad_norm": 0.09242924302816391,
      "learning_rate": 0.000709485,
      "loss": 1.9209175109863281,
      "step": 47300
    },
    {
      "epoch": 0.000132,
      "grad_norm": 0.06336706876754761,
      "learning_rate": 0.000709785,
      "loss": 1.9172807693481446,
      "step": 47320
    },
    {
      "epoch": 0.000134,
      "grad_norm": 0.06539217382669449,
      "learning_rate": 0.000710085,
      "loss": 1.921341896057129,
      "step": 47340
    },
    {
      "epoch": 0.000136,
      "grad_norm": 0.06496850401163101,
      "learning_rate": 0.000710385,
      "loss": 1.9217124938964845,
      "step": 47360
    },
    {
      "epoch": 0.000138,
      "grad_norm": 0.06589950621128082,
      "learning_rate": 0.000710685,
      "loss": 1.917654037475586,
      "step": 47380
    },
    {
      "epoch": 0.00014,
      "grad_norm": 0.10992733389139175,
      "learning_rate": 0.000710985,
      "loss": 1.909129524230957,
      "step": 47400
    },
    {
      "epoch": 0.000142,
      "grad_norm": 0.07374437898397446,
      "learning_rate": 0.000711285,
      "loss": 1.9101346969604491,
      "step": 47420
    },
    {
      "epoch": 0.000144,
      "grad_norm": 0.062163710594177246,
      "learning_rate": 0.0007115849999999999,
      "loss": 1.9079486846923828,
      "step": 47440
    },
    {
      "epoch": 0.000146,
      "grad_norm": 0.08394552022218704,
      "learning_rate": 0.000711885,
      "loss": 1.9161951065063476,
      "step": 47460
    },
    {
      "epoch": 0.000148,
      "grad_norm": 0.11592655628919601,
      "learning_rate": 0.000712185,
      "loss": 1.9125499725341797,
      "step": 47480
    },
    {
      "epoch": 0.00015,
      "grad_norm": 0.07041492313146591,
      "learning_rate": 0.000712485,
      "loss": 1.9119735717773438,
      "step": 47500
    },
    {
      "epoch": 0.000152,
      "grad_norm": 0.05417707934975624,
      "learning_rate": 0.0007127850000000001,
      "loss": 1.9030326843261718,
      "step": 47520
    },
    {
      "epoch": 0.000154,
      "grad_norm": 0.0702233612537384,
      "learning_rate": 0.000713085,
      "loss": 1.9038074493408204,
      "step": 47540
    },
    {
      "epoch": 0.000156,
      "grad_norm": 0.05859832465648651,
      "learning_rate": 0.0007133850000000001,
      "loss": 1.900547981262207,
      "step": 47560
    },
    {
      "epoch": 0.000158,
      "grad_norm": 0.08888107538223267,
      "learning_rate": 0.000713685,
      "loss": 1.9098312377929687,
      "step": 47580
    },
    {
      "epoch": 0.00016,
      "grad_norm": 0.0922444686293602,
      "learning_rate": 0.0007139850000000001,
      "loss": 1.902810287475586,
      "step": 47600
    },
    {
      "epoch": 0.000162,
      "grad_norm": 0.06493257731199265,
      "learning_rate": 0.000714285,
      "loss": 1.9034292221069335,
      "step": 47620
    },
    {
      "epoch": 0.000164,
      "grad_norm": 0.07373186200857162,
      "learning_rate": 0.000714585,
      "loss": 1.9035234451293945,
      "step": 47640
    },
    {
      "epoch": 0.000166,
      "grad_norm": 0.06822580099105835,
      "learning_rate": 0.0007148850000000001,
      "loss": 1.902390480041504,
      "step": 47660
    },
    {
      "epoch": 0.000168,
      "grad_norm": 0.06901398301124573,
      "learning_rate": 0.000715185,
      "loss": 1.89398136138916,
      "step": 47680
    },
    {
      "epoch": 0.00017,
      "grad_norm": 0.09062466770410538,
      "learning_rate": 0.0007154850000000001,
      "loss": 1.9004844665527343,
      "step": 47700
    },
    {
      "epoch": 0.000172,
      "grad_norm": 0.14246119558811188,
      "learning_rate": 0.000715785,
      "loss": 1.8940235137939454,
      "step": 47720
    },
    {
      "epoch": 0.000174,
      "grad_norm": 0.09346599876880646,
      "learning_rate": 0.000716085,
      "loss": 1.8926639556884766,
      "step": 47740
    },
    {
      "epoch": 0.000176,
      "grad_norm": 0.07609276473522186,
      "learning_rate": 0.000716385,
      "loss": 1.8961585998535155,
      "step": 47760
    },
    {
      "epoch": 0.000178,
      "grad_norm": 0.06778562813997269,
      "learning_rate": 0.000716685,
      "loss": 1.8936416625976562,
      "step": 47780
    },
    {
      "epoch": 0.00018,
      "grad_norm": 0.12500469386577606,
      "learning_rate": 0.0007169850000000001,
      "loss": 1.8971038818359376,
      "step": 47800
    },
    {
      "epoch": 0.000182,
      "grad_norm": 0.0834629014134407,
      "learning_rate": 0.000717285,
      "loss": 1.9050609588623046,
      "step": 47820
    },
    {
      "epoch": 0.000184,
      "grad_norm": 0.07134594768285751,
      "learning_rate": 0.000717585,
      "loss": 1.894935417175293,
      "step": 47840
    },
    {
      "epoch": 0.000186,
      "grad_norm": 0.06053246557712555,
      "learning_rate": 0.000717885,
      "loss": 1.8934173583984375,
      "step": 47860
    },
    {
      "epoch": 0.000188,
      "grad_norm": 0.07200822979211807,
      "learning_rate": 0.000718185,
      "loss": 1.895335578918457,
      "step": 47880
    },
    {
      "epoch": 0.00019,
      "grad_norm": 0.07255461066961288,
      "learning_rate": 0.000718485,
      "loss": 1.8858306884765625,
      "step": 47900
    },
    {
      "epoch": 0.000192,
      "grad_norm": 0.1142456904053688,
      "learning_rate": 0.000718785,
      "loss": 1.8956653594970703,
      "step": 47920
    },
    {
      "epoch": 0.000194,
      "grad_norm": 0.11873633414506912,
      "learning_rate": 0.000719085,
      "loss": 1.8932693481445313,
      "step": 47940
    },
    {
      "epoch": 0.000196,
      "grad_norm": 0.06071966513991356,
      "learning_rate": 0.000719385,
      "loss": 1.8919456481933594,
      "step": 47960
    },
    {
      "epoch": 0.000198,
      "grad_norm": 0.10144706070423126,
      "learning_rate": 0.000719685,
      "loss": 1.886199951171875,
      "step": 47980
    },
    {
      "epoch": 0.0002,
      "grad_norm": 0.07306289672851562,
      "learning_rate": 0.000719985,
      "loss": 1.886566734313965,
      "step": 48000
    },
    {
      "epoch": 0.0002,
      "eval_loss": 1.8350651264190674,
      "eval_runtime": 99.0695,
      "eval_samples_per_second": 10.094,
      "eval_steps_per_second": 10.094,
      "step": 48000
    },
    {
      "epoch": 0.000202,
      "grad_norm": 0.07433490455150604,
      "learning_rate": 0.000720285,
      "loss": 1.8901538848876953,
      "step": 48020
    },
    {
      "epoch": 0.000204,
      "grad_norm": 0.06506136059761047,
      "learning_rate": 0.000720585,
      "loss": 1.8856683731079102,
      "step": 48040
    },
    {
      "epoch": 0.000206,
      "grad_norm": 0.07532741129398346,
      "learning_rate": 0.000720885,
      "loss": 1.885334587097168,
      "step": 48060
    },
    {
      "epoch": 0.000208,
      "grad_norm": 0.08687019348144531,
      "learning_rate": 0.0007211850000000001,
      "loss": 1.8871675491333009,
      "step": 48080
    },
    {
      "epoch": 0.00021,
      "grad_norm": 0.06796624511480331,
      "learning_rate": 0.000721485,
      "loss": 1.88447265625,
      "step": 48100
    },
    {
      "epoch": 0.000212,
      "grad_norm": 0.07211881875991821,
      "learning_rate": 0.0007217850000000001,
      "loss": 1.8851350784301757,
      "step": 48120
    },
    {
      "epoch": 0.000214,
      "grad_norm": 0.09619178622961044,
      "learning_rate": 0.000722085,
      "loss": 1.8860481262207032,
      "step": 48140
    },
    {
      "epoch": 0.000216,
      "grad_norm": 0.09473865479230881,
      "learning_rate": 0.0007223850000000001,
      "loss": 1.8759679794311523,
      "step": 48160
    },
    {
      "epoch": 0.000218,
      "grad_norm": 0.06743013858795166,
      "learning_rate": 0.000722685,
      "loss": 1.8819561004638672,
      "step": 48180
    },
    {
      "epoch": 0.00022,
      "grad_norm": 0.08684151619672775,
      "learning_rate": 0.000722985,
      "loss": 1.877330207824707,
      "step": 48200
    },
    {
      "epoch": 0.000222,
      "grad_norm": 0.06397704780101776,
      "learning_rate": 0.000723285,
      "loss": 1.8899486541748047,
      "step": 48220
    },
    {
      "epoch": 0.000224,
      "grad_norm": 0.0900849923491478,
      "learning_rate": 0.000723585,
      "loss": 1.8874429702758788,
      "step": 48240
    },
    {
      "epoch": 0.000226,
      "grad_norm": 0.07488174736499786,
      "learning_rate": 0.0007238850000000001,
      "loss": 1.8861740112304688,
      "step": 48260
    },
    {
      "epoch": 0.000228,
      "grad_norm": 0.09316900372505188,
      "learning_rate": 0.000724185,
      "loss": 1.8858060836791992,
      "step": 48280
    },
    {
      "epoch": 0.00023,
      "grad_norm": 0.0723276361823082,
      "learning_rate": 0.000724485,
      "loss": 1.8740653991699219,
      "step": 48300
    },
    {
      "epoch": 0.000232,
      "grad_norm": 0.0841292217373848,
      "learning_rate": 0.000724785,
      "loss": 1.8807933807373047,
      "step": 48320
    },
    {
      "epoch": 0.000234,
      "grad_norm": 0.09044871479272842,
      "learning_rate": 0.000725085,
      "loss": 1.8928699493408203,
      "step": 48340
    },
    {
      "epoch": 0.000236,
      "grad_norm": 0.08968909084796906,
      "learning_rate": 0.000725385,
      "loss": 1.8885452270507812,
      "step": 48360
    },
    {
      "epoch": 0.000238,
      "grad_norm": 0.07492511719465256,
      "learning_rate": 0.000725685,
      "loss": 1.881599235534668,
      "step": 48380
    },
    {
      "epoch": 0.00024,
      "grad_norm": 0.06670362502336502,
      "learning_rate": 0.000725985,
      "loss": 1.886728858947754,
      "step": 48400
    },
    {
      "epoch": 0.000242,
      "grad_norm": 0.06100883707404137,
      "learning_rate": 0.000726285,
      "loss": 1.898359489440918,
      "step": 48420
    },
    {
      "epoch": 0.000244,
      "grad_norm": 0.07540082186460495,
      "learning_rate": 0.000726585,
      "loss": 1.8863595962524413,
      "step": 48440
    },
    {
      "epoch": 0.000246,
      "grad_norm": 0.06365904211997986,
      "learning_rate": 0.000726885,
      "loss": 1.8902715682983398,
      "step": 48460
    },
    {
      "epoch": 0.000248,
      "grad_norm": 0.07440081238746643,
      "learning_rate": 0.000727185,
      "loss": 1.8887521743774414,
      "step": 48480
    },
    {
      "epoch": 0.00025,
      "grad_norm": 0.06741853058338165,
      "learning_rate": 0.0007274849999999999,
      "loss": 1.8897829055786133,
      "step": 48500
    },
    {
      "epoch": 0.000252,
      "grad_norm": 0.06400787085294724,
      "learning_rate": 0.000727785,
      "loss": 1.892916488647461,
      "step": 48520
    },
    {
      "epoch": 0.000254,
      "grad_norm": 0.084056057035923,
      "learning_rate": 0.000728085,
      "loss": 1.8810224533081055,
      "step": 48540
    },
    {
      "epoch": 0.000256,
      "grad_norm": 0.07808465510606766,
      "learning_rate": 0.000728385,
      "loss": 1.8811614990234375,
      "step": 48560
    },
    {
      "epoch": 0.000258,
      "grad_norm": 0.06912193447351456,
      "learning_rate": 0.000728685,
      "loss": 1.8807933807373047,
      "step": 48580
    },
    {
      "epoch": 0.00026,
      "grad_norm": 0.11728200316429138,
      "learning_rate": 0.000728985,
      "loss": 1.8793157577514648,
      "step": 48600
    },
    {
      "epoch": 0.000262,
      "grad_norm": 0.07116056978702545,
      "learning_rate": 0.000729285,
      "loss": 1.8806476593017578,
      "step": 48620
    },
    {
      "epoch": 0.000264,
      "grad_norm": 0.07659170776605606,
      "learning_rate": 0.000729585,
      "loss": 1.8804895401000976,
      "step": 48640
    },
    {
      "epoch": 0.000266,
      "grad_norm": 0.05822935327887535,
      "learning_rate": 0.000729885,
      "loss": 1.8786319732666015,
      "step": 48660
    },
    {
      "epoch": 0.000268,
      "grad_norm": 0.07382626831531525,
      "learning_rate": 0.0007301850000000001,
      "loss": 1.8767101287841796,
      "step": 48680
    },
    {
      "epoch": 0.00027,
      "grad_norm": 0.07334620505571365,
      "learning_rate": 0.000730485,
      "loss": 1.8743818283081055,
      "step": 48700
    },
    {
      "epoch": 0.000272,
      "grad_norm": 0.07567453384399414,
      "learning_rate": 0.0007307850000000001,
      "loss": 1.8763639450073242,
      "step": 48720
    },
    {
      "epoch": 0.000274,
      "grad_norm": 0.08777511864900589,
      "learning_rate": 0.000731085,
      "loss": 1.8786584854125976,
      "step": 48740
    },
    {
      "epoch": 0.000276,
      "grad_norm": 0.09745394438505173,
      "learning_rate": 0.0007313850000000001,
      "loss": 1.8746225357055664,
      "step": 48760
    },
    {
      "epoch": 0.000278,
      "grad_norm": 0.06679695099592209,
      "learning_rate": 0.000731685,
      "loss": 1.8859651565551758,
      "step": 48780
    },
    {
      "epoch": 0.00028,
      "grad_norm": 0.054582711309194565,
      "learning_rate": 0.000731985,
      "loss": 1.8781494140625,
      "step": 48800
    },
    {
      "epoch": 0.000282,
      "grad_norm": 0.06270311772823334,
      "learning_rate": 0.000732285,
      "loss": 1.8780647277832032,
      "step": 48820
    },
    {
      "epoch": 0.000284,
      "grad_norm": 0.07959431409835815,
      "learning_rate": 0.000732585,
      "loss": 1.872858428955078,
      "step": 48840
    },
    {
      "epoch": 0.000286,
      "grad_norm": 0.08098466694355011,
      "learning_rate": 0.0007328850000000001,
      "loss": 1.8745456695556642,
      "step": 48860
    },
    {
      "epoch": 0.000288,
      "grad_norm": 0.0944877415895462,
      "learning_rate": 0.000733185,
      "loss": 1.8757490158081054,
      "step": 48880
    },
    {
      "epoch": 0.00029,
      "grad_norm": 0.09513352811336517,
      "learning_rate": 0.000733485,
      "loss": 1.886286735534668,
      "step": 48900
    },
    {
      "epoch": 0.000292,
      "grad_norm": 0.10484618693590164,
      "learning_rate": 0.000733785,
      "loss": 1.8716920852661132,
      "step": 48920
    },
    {
      "epoch": 0.000294,
      "grad_norm": 0.06396955996751785,
      "learning_rate": 0.000734085,
      "loss": 1.876509666442871,
      "step": 48940
    },
    {
      "epoch": 0.000296,
      "grad_norm": 0.06441602110862732,
      "learning_rate": 0.000734385,
      "loss": 1.8659820556640625,
      "step": 48960
    },
    {
      "epoch": 0.000298,
      "grad_norm": 0.08893131464719772,
      "learning_rate": 0.000734685,
      "loss": 1.865779495239258,
      "step": 48980
    },
    {
      "epoch": 0.0003,
      "grad_norm": 0.07854273170232773,
      "learning_rate": 0.000734985,
      "loss": 1.8671316146850585,
      "step": 49000
    },
    {
      "epoch": 0.0003,
      "eval_loss": 1.8231958150863647,
      "eval_runtime": 99.9569,
      "eval_samples_per_second": 10.004,
      "eval_steps_per_second": 10.004,
      "step": 49000
    },
    {
      "epoch": 0.000302,
      "grad_norm": 0.08891486376523972,
      "learning_rate": 0.000735285,
      "loss": 1.8616506576538085,
      "step": 49020
    },
    {
      "epoch": 0.000304,
      "grad_norm": 0.08781826496124268,
      "learning_rate": 0.000735585,
      "loss": 1.865854835510254,
      "step": 49040
    },
    {
      "epoch": 0.000306,
      "grad_norm": 0.12502416968345642,
      "learning_rate": 0.000735885,
      "loss": 1.8626695632934571,
      "step": 49060
    },
    {
      "epoch": 0.000308,
      "grad_norm": 0.06342129409313202,
      "learning_rate": 0.000736185,
      "loss": 1.8653135299682617,
      "step": 49080
    },
    {
      "epoch": 0.00031,
      "grad_norm": 0.08215075731277466,
      "learning_rate": 0.0007364849999999999,
      "loss": 1.8684637069702148,
      "step": 49100
    },
    {
      "epoch": 0.000312,
      "grad_norm": 0.09643544256687164,
      "learning_rate": 0.000736785,
      "loss": 1.8758543014526368,
      "step": 49120
    },
    {
      "epoch": 0.000314,
      "grad_norm": 0.08002220094203949,
      "learning_rate": 0.000737085,
      "loss": 1.8670568466186523,
      "step": 49140
    },
    {
      "epoch": 0.000316,
      "grad_norm": 0.05769426375627518,
      "learning_rate": 0.000737385,
      "loss": 1.8668659210205079,
      "step": 49160
    },
    {
      "epoch": 0.000318,
      "grad_norm": 0.09401297569274902,
      "learning_rate": 0.0007376850000000001,
      "loss": 1.8710269927978516,
      "step": 49180
    },
    {
      "epoch": 0.00032,
      "grad_norm": 0.1117960587143898,
      "learning_rate": 0.000737985,
      "loss": 1.870610237121582,
      "step": 49200
    },
    {
      "epoch": 0.000322,
      "grad_norm": 0.08143491297960281,
      "learning_rate": 0.0007382850000000001,
      "loss": 1.866176223754883,
      "step": 49220
    },
    {
      "epoch": 0.000324,
      "grad_norm": 0.0771065503358841,
      "learning_rate": 0.000738585,
      "loss": 1.867238426208496,
      "step": 49240
    },
    {
      "epoch": 0.000326,
      "grad_norm": 0.04945933073759079,
      "learning_rate": 0.000738885,
      "loss": 1.8657207489013672,
      "step": 49260
    },
    {
      "epoch": 0.000328,
      "grad_norm": 0.08263809978961945,
      "learning_rate": 0.0007391850000000001,
      "loss": 1.8570655822753905,
      "step": 49280
    },
    {
      "epoch": 0.00033,
      "grad_norm": 0.0943840742111206,
      "learning_rate": 0.000739485,
      "loss": 1.8613241195678711,
      "step": 49300
    },
    {
      "epoch": 0.000332,
      "grad_norm": 0.06208841875195503,
      "learning_rate": 0.0007397850000000001,
      "loss": 1.8662214279174805,
      "step": 49320
    },
    {
      "epoch": 0.000334,
      "grad_norm": 0.06485865265130997,
      "learning_rate": 0.000740085,
      "loss": 1.8635101318359375,
      "step": 49340
    },
    {
      "epoch": 0.000336,
      "grad_norm": 0.08560191094875336,
      "learning_rate": 0.000740385,
      "loss": 1.8564205169677734,
      "step": 49360
    },
    {
      "epoch": 0.000338,
      "grad_norm": 0.12357484549283981,
      "learning_rate": 0.000740685,
      "loss": 1.8634933471679687,
      "step": 49380
    },
    {
      "epoch": 0.00034,
      "grad_norm": 0.08451298624277115,
      "learning_rate": 0.000740985,
      "loss": 1.8603437423706055,
      "step": 49400
    },
    {
      "epoch": 0.000342,
      "grad_norm": 0.09098716825246811,
      "learning_rate": 0.0007412850000000001,
      "loss": 1.8705177307128906,
      "step": 49420
    },
    {
      "epoch": 0.000344,
      "grad_norm": 0.0718744546175003,
      "learning_rate": 0.000741585,
      "loss": 1.8548259735107422,
      "step": 49440
    },
    {
      "epoch": 0.000346,
      "grad_norm": 0.06135766953229904,
      "learning_rate": 0.000741885,
      "loss": 1.8501501083374023,
      "step": 49460
    },
    {
      "epoch": 0.000348,
      "grad_norm": 0.09332036226987839,
      "learning_rate": 0.000742185,
      "loss": 1.8546777725219727,
      "step": 49480
    },
    {
      "epoch": 0.00035,
      "grad_norm": 0.1360894739627838,
      "learning_rate": 0.000742485,
      "loss": 1.864227294921875,
      "step": 49500
    },
    {
      "epoch": 0.000352,
      "grad_norm": 0.08547444641590118,
      "learning_rate": 0.000742785,
      "loss": 1.8685710906982422,
      "step": 49520
    },
    {
      "epoch": 0.000354,
      "grad_norm": 0.10424457490444183,
      "learning_rate": 0.000743085,
      "loss": 1.8618698120117188,
      "step": 49540
    },
    {
      "epoch": 0.000356,
      "grad_norm": 0.09299690276384354,
      "learning_rate": 0.0007433849999999999,
      "loss": 1.8716371536254883,
      "step": 49560
    },
    {
      "epoch": 0.000358,
      "grad_norm": 0.06361870467662811,
      "learning_rate": 0.000743685,
      "loss": 1.8801010131835938,
      "step": 49580
    },
    {
      "epoch": 0.00036,
      "grad_norm": 0.08042526245117188,
      "learning_rate": 0.000743985,
      "loss": 1.8659257888793945,
      "step": 49600
    },
    {
      "epoch": 0.000362,
      "grad_norm": 0.10201700031757355,
      "learning_rate": 0.000744285,
      "loss": 1.8701637268066407,
      "step": 49620
    },
    {
      "epoch": 0.000364,
      "grad_norm": 0.10048798471689224,
      "learning_rate": 0.000744585,
      "loss": 1.8687246322631836,
      "step": 49640
    },
    {
      "epoch": 0.000366,
      "grad_norm": 0.10112548619508743,
      "learning_rate": 0.0007448849999999999,
      "loss": 1.867312240600586,
      "step": 49660
    },
    {
      "epoch": 0.000368,
      "grad_norm": 0.06898175179958344,
      "learning_rate": 0.000745185,
      "loss": 1.8694904327392579,
      "step": 49680
    },
    {
      "epoch": 0.00037,
      "grad_norm": 0.08607565611600876,
      "learning_rate": 0.000745485,
      "loss": 1.8709667205810547,
      "step": 49700
    },
    {
      "epoch": 0.000372,
      "grad_norm": 0.08153693377971649,
      "learning_rate": 0.000745785,
      "loss": 1.8627836227416992,
      "step": 49720
    },
    {
      "epoch": 0.000374,
      "grad_norm": 0.07933573424816132,
      "learning_rate": 0.0007460850000000001,
      "loss": 1.8717823028564453,
      "step": 49740
    },
    {
      "epoch": 0.000376,
      "grad_norm": 0.0822339653968811,
      "learning_rate": 0.000746385,
      "loss": 1.8707391738891601,
      "step": 49760
    },
    {
      "epoch": 0.000378,
      "grad_norm": 0.08838674426078796,
      "learning_rate": 0.0007466850000000001,
      "loss": 1.8709238052368165,
      "step": 49780
    },
    {
      "epoch": 0.00038,
      "grad_norm": 0.09605666249990463,
      "learning_rate": 0.000746985,
      "loss": 1.8747797012329102,
      "step": 49800
    },
    {
      "epoch": 0.000382,
      "grad_norm": 0.07646477222442627,
      "learning_rate": 0.0007472850000000001,
      "loss": 1.859670639038086,
      "step": 49820
    },
    {
      "epoch": 0.000384,
      "grad_norm": 0.07252638041973114,
      "learning_rate": 0.000747585,
      "loss": 1.856839942932129,
      "step": 49840
    },
    {
      "epoch": 0.000386,
      "grad_norm": 0.08199745416641235,
      "learning_rate": 0.000747885,
      "loss": 1.867083740234375,
      "step": 49860
    },
    {
      "epoch": 0.000388,
      "grad_norm": 0.07510737329721451,
      "learning_rate": 0.0007481850000000001,
      "loss": 1.8708480834960937,
      "step": 49880
    },
    {
      "epoch": 0.00039,
      "grad_norm": 0.10873018205165863,
      "learning_rate": 0.000748485,
      "loss": 1.8570371627807618,
      "step": 49900
    },
    {
      "epoch": 0.000392,
      "grad_norm": 0.07855768501758575,
      "learning_rate": 0.0007487850000000001,
      "loss": 1.8611907958984375,
      "step": 49920
    },
    {
      "epoch": 0.000394,
      "grad_norm": 0.0734357237815857,
      "learning_rate": 0.000749085,
      "loss": 1.8603155136108398,
      "step": 49940
    },
    {
      "epoch": 0.000396,
      "grad_norm": 0.11408625543117523,
      "learning_rate": 0.000749385,
      "loss": 1.862259292602539,
      "step": 49960
    },
    {
      "epoch": 0.000398,
      "grad_norm": 0.10029411315917969,
      "learning_rate": 0.000749685,
      "loss": 1.86767578125,
      "step": 49980
    },
    {
      "epoch": 0.0004,
      "grad_norm": 0.07305110991001129,
      "learning_rate": 0.000749985,
      "loss": 1.8574264526367188,
      "step": 50000
    },
    {
      "epoch": 0.0004,
      "eval_loss": 1.8127195835113525,
      "eval_runtime": 104.6958,
      "eval_samples_per_second": 9.551,
      "eval_steps_per_second": 9.551,
      "step": 50000
    },
    {
      "epoch": 0.000402,
      "grad_norm": 0.07434222847223282,
      "learning_rate": 0.0007502850000000001,
      "loss": 1.8548992156982422,
      "step": 50020
    },
    {
      "epoch": 0.000404,
      "grad_norm": 0.10143296420574188,
      "learning_rate": 0.000750585,
      "loss": 1.853312110900879,
      "step": 50040
    },
    {
      "epoch": 0.000406,
      "grad_norm": 0.08113838732242584,
      "learning_rate": 0.000750885,
      "loss": 1.8585927963256836,
      "step": 50060
    },
    {
      "epoch": 0.000408,
      "grad_norm": 0.06600955873727798,
      "learning_rate": 0.0007511849999999999,
      "loss": 1.853336715698242,
      "step": 50080
    },
    {
      "epoch": 0.00041,
      "grad_norm": 0.09208826720714569,
      "learning_rate": 0.0007514850000000001,
      "loss": 1.8564899444580079,
      "step": 50100
    },
    {
      "epoch": 0.000412,
      "grad_norm": 0.05692492425441742,
      "learning_rate": 0.000751785,
      "loss": 1.8583003997802734,
      "step": 50120
    },
    {
      "epoch": 0.000414,
      "grad_norm": 0.09818264096975327,
      "learning_rate": 0.000752085,
      "loss": 1.851161575317383,
      "step": 50140
    },
    {
      "epoch": 0.000416,
      "grad_norm": 0.11614762246608734,
      "learning_rate": 0.000752385,
      "loss": 1.8539705276489258,
      "step": 50160
    },
    {
      "epoch": 0.000418,
      "grad_norm": 0.06572745740413666,
      "learning_rate": 0.0007526849999999999,
      "loss": 1.862236785888672,
      "step": 50180
    },
    {
      "epoch": 0.00042,
      "grad_norm": 0.08823873102664948,
      "learning_rate": 0.0007529850000000001,
      "loss": 1.8488019943237304,
      "step": 50200
    },
    {
      "epoch": 0.000422,
      "grad_norm": 0.06876957416534424,
      "learning_rate": 0.000753285,
      "loss": 1.854522132873535,
      "step": 50220
    },
    {
      "epoch": 0.000424,
      "grad_norm": 0.1036091297864914,
      "learning_rate": 0.000753585,
      "loss": 1.8490419387817383,
      "step": 50240
    },
    {
      "epoch": 0.000426,
      "grad_norm": 0.08654841780662537,
      "learning_rate": 0.000753885,
      "loss": 1.8570022583007812,
      "step": 50260
    },
    {
      "epoch": 0.000428,
      "grad_norm": 0.0797325074672699,
      "learning_rate": 0.000754185,
      "loss": 1.852284049987793,
      "step": 50280
    },
    {
      "epoch": 0.00043,
      "grad_norm": 0.09358066320419312,
      "learning_rate": 0.000754485,
      "loss": 1.845249557495117,
      "step": 50300
    },
    {
      "epoch": 0.000432,
      "grad_norm": 0.09743467718362808,
      "learning_rate": 0.000754785,
      "loss": 1.8510974884033202,
      "step": 50320
    },
    {
      "epoch": 0.000434,
      "grad_norm": 0.07994414120912552,
      "learning_rate": 0.0007550850000000001,
      "loss": 1.8470056533813477,
      "step": 50340
    },
    {
      "epoch": 0.000436,
      "grad_norm": 0.07373213022947311,
      "learning_rate": 0.000755385,
      "loss": 1.8472797393798828,
      "step": 50360
    },
    {
      "epoch": 0.000438,
      "grad_norm": 0.07436708360910416,
      "learning_rate": 0.000755685,
      "loss": 1.8461109161376954,
      "step": 50380
    },
    {
      "epoch": 0.00044,
      "grad_norm": 0.06742985546588898,
      "learning_rate": 0.0007559850000000001,
      "loss": 1.8556638717651368,
      "step": 50400
    },
    {
      "epoch": 0.000442,
      "grad_norm": 0.07899279147386551,
      "learning_rate": 0.0007562850000000001,
      "loss": 1.8491823196411132,
      "step": 50420
    },
    {
      "epoch": 0.000444,
      "grad_norm": 0.0626215934753418,
      "learning_rate": 0.000756585,
      "loss": 1.8468595504760743,
      "step": 50440
    },
    {
      "epoch": 0.000446,
      "grad_norm": 0.07054529339075089,
      "learning_rate": 0.000756885,
      "loss": 1.8442632675170898,
      "step": 50460
    },
    {
      "epoch": 0.000448,
      "grad_norm": 0.08473960310220718,
      "learning_rate": 0.000757185,
      "loss": 1.8389867782592773,
      "step": 50480
    },
    {
      "epoch": 0.00045,
      "grad_norm": 0.08779094368219376,
      "learning_rate": 0.0007574850000000001,
      "loss": 1.847212028503418,
      "step": 50500
    },
    {
      "epoch": 0.000452,
      "grad_norm": 0.06304117292165756,
      "learning_rate": 0.0007577850000000001,
      "loss": 1.838887596130371,
      "step": 50520
    },
    {
      "epoch": 0.000454,
      "grad_norm": 0.07276386022567749,
      "learning_rate": 0.000758085,
      "loss": 1.8409353256225587,
      "step": 50540
    },
    {
      "epoch": 0.000456,
      "grad_norm": 0.07081222534179688,
      "learning_rate": 0.000758385,
      "loss": 1.8434917449951171,
      "step": 50560
    },
    {
      "epoch": 0.000458,
      "grad_norm": 0.09153230488300323,
      "learning_rate": 0.0007586849999999999,
      "loss": 1.8405410766601562,
      "step": 50580
    },
    {
      "epoch": 0.00046,
      "grad_norm": 0.08038565516471863,
      "learning_rate": 0.0007589850000000001,
      "loss": 1.8444704055786132,
      "step": 50600
    },
    {
      "epoch": 0.000462,
      "grad_norm": 0.08140019327402115,
      "learning_rate": 0.0007592850000000001,
      "loss": 1.8456674575805665,
      "step": 50620
    },
    {
      "epoch": 0.000464,
      "grad_norm": 0.07837044447660446,
      "learning_rate": 0.000759585,
      "loss": 1.8389007568359375,
      "step": 50640
    },
    {
      "epoch": 0.000466,
      "grad_norm": 0.08453674614429474,
      "learning_rate": 0.000759885,
      "loss": 1.8354032516479493,
      "step": 50660
    },
    {
      "epoch": 0.000468,
      "grad_norm": 0.08782172948122025,
      "learning_rate": 0.0007601849999999999,
      "loss": 1.8379409790039063,
      "step": 50680
    },
    {
      "epoch": 0.00047,
      "grad_norm": 0.06480462849140167,
      "learning_rate": 0.0007604850000000001,
      "loss": 1.8450876235961915,
      "step": 50700
    },
    {
      "epoch": 0.000472,
      "grad_norm": 0.09490370750427246,
      "learning_rate": 0.000760785,
      "loss": 1.8424947738647461,
      "step": 50720
    },
    {
      "epoch": 0.000474,
      "grad_norm": 0.09134998172521591,
      "learning_rate": 0.000761085,
      "loss": 1.8489044189453125,
      "step": 50740
    },
    {
      "epoch": 0.000476,
      "grad_norm": 0.08516672253608704,
      "learning_rate": 0.000761385,
      "loss": 1.8510494232177734,
      "step": 50760
    },
    {
      "epoch": 0.000478,
      "grad_norm": 0.09990148991346359,
      "learning_rate": 0.000761685,
      "loss": 1.8433395385742188,
      "step": 50780
    },
    {
      "epoch": 0.00048,
      "grad_norm": 0.07432044297456741,
      "learning_rate": 0.0007619850000000001,
      "loss": 1.8616621017456054,
      "step": 50800
    },
    {
      "epoch": 0.000482,
      "grad_norm": 0.07730050384998322,
      "learning_rate": 0.000762285,
      "loss": 1.857870101928711,
      "step": 50820
    },
    {
      "epoch": 0.000484,
      "grad_norm": 0.09263300895690918,
      "learning_rate": 0.000762585,
      "loss": 1.852854347229004,
      "step": 50840
    },
    {
      "epoch": 0.000486,
      "grad_norm": 0.1193644255399704,
      "learning_rate": 0.000762885,
      "loss": 1.8558570861816406,
      "step": 50860
    },
    {
      "epoch": 0.000488,
      "grad_norm": 0.06789831817150116,
      "learning_rate": 0.000763185,
      "loss": 1.8561784744262695,
      "step": 50880
    },
    {
      "epoch": 0.00049,
      "grad_norm": 0.09891850501298904,
      "learning_rate": 0.000763485,
      "loss": 1.8547218322753907,
      "step": 50900
    },
    {
      "epoch": 0.000492,
      "grad_norm": 0.07384803146123886,
      "learning_rate": 0.0007637850000000001,
      "loss": 1.856735610961914,
      "step": 50920
    },
    {
      "epoch": 0.000494,
      "grad_norm": 0.07126709073781967,
      "learning_rate": 0.0007640850000000001,
      "loss": 1.855617141723633,
      "step": 50940
    },
    {
      "epoch": 0.000496,
      "grad_norm": 0.0794643834233284,
      "learning_rate": 0.000764385,
      "loss": 1.851529312133789,
      "step": 50960
    },
    {
      "epoch": 0.000498,
      "grad_norm": 0.07430235296487808,
      "learning_rate": 0.000764685,
      "loss": 1.84918270111084,
      "step": 50980
    },
    {
      "epoch": 0.0005,
      "grad_norm": 0.05850321426987648,
      "learning_rate": 0.0007649850000000001,
      "loss": 1.8446617126464844,
      "step": 51000
    },
    {
      "epoch": 0.0005,
      "eval_loss": 1.8034069538116455,
      "eval_runtime": 99.2653,
      "eval_samples_per_second": 10.074,
      "eval_steps_per_second": 10.074,
      "step": 51000
    },
    {
      "epoch": 0.000502,
      "grad_norm": 0.07530247420072556,
      "learning_rate": 0.0007652850000000001,
      "loss": 1.8506610870361329,
      "step": 51020
    },
    {
      "epoch": 0.000504,
      "grad_norm": 0.11097286641597748,
      "learning_rate": 0.000765585,
      "loss": 1.8453807830810547,
      "step": 51040
    },
    {
      "epoch": 0.000506,
      "grad_norm": 0.06734902411699295,
      "learning_rate": 0.000765885,
      "loss": 1.8480216979980468,
      "step": 51060
    },
    {
      "epoch": 0.000508,
      "grad_norm": 0.08474509418010712,
      "learning_rate": 0.000766185,
      "loss": 1.847792625427246,
      "step": 51080
    },
    {
      "epoch": 0.00051,
      "grad_norm": 0.07088479399681091,
      "learning_rate": 0.0007664850000000001,
      "loss": 1.8439413070678712,
      "step": 51100
    },
    {
      "epoch": 0.000512,
      "grad_norm": 0.08443014323711395,
      "learning_rate": 0.0007667850000000001,
      "loss": 1.8392679214477539,
      "step": 51120
    },
    {
      "epoch": 0.000514,
      "grad_norm": 0.07145456969738007,
      "learning_rate": 0.000767085,
      "loss": 1.840279769897461,
      "step": 51140
    },
    {
      "epoch": 0.000516,
      "grad_norm": 0.07933073490858078,
      "learning_rate": 0.000767385,
      "loss": 1.8390647888183593,
      "step": 51160
    },
    {
      "epoch": 0.000518,
      "grad_norm": 0.0909983366727829,
      "learning_rate": 0.0007676849999999999,
      "loss": 1.846038246154785,
      "step": 51180
    },
    {
      "epoch": 0.00052,
      "grad_norm": 0.06586463749408722,
      "learning_rate": 0.0007679849999999999,
      "loss": 1.841064453125,
      "step": 51200
    },
    {
      "epoch": 0.000522,
      "grad_norm": 0.10752559453248978,
      "learning_rate": 0.0007682850000000001,
      "loss": 1.8423982620239259,
      "step": 51220
    },
    {
      "epoch": 0.000524,
      "grad_norm": 0.09651283174753189,
      "learning_rate": 0.000768585,
      "loss": 1.8416040420532227,
      "step": 51240
    },
    {
      "epoch": 0.000526,
      "grad_norm": 0.08183816075325012,
      "learning_rate": 0.000768885,
      "loss": 1.832880401611328,
      "step": 51260
    },
    {
      "epoch": 0.000528,
      "grad_norm": 0.08817758411169052,
      "learning_rate": 0.0007691849999999999,
      "loss": 1.8434953689575195,
      "step": 51280
    },
    {
      "epoch": 0.00053,
      "grad_norm": 0.09007537364959717,
      "learning_rate": 0.000769485,
      "loss": 1.83581485748291,
      "step": 51300
    },
    {
      "epoch": 0.000532,
      "grad_norm": 0.06463437527418137,
      "learning_rate": 0.000769785,
      "loss": 1.8326082229614258,
      "step": 51320
    },
    {
      "epoch": 0.000534,
      "grad_norm": 0.09915205091238022,
      "learning_rate": 0.000770085,
      "loss": 1.832014274597168,
      "step": 51340
    },
    {
      "epoch": 0.000536,
      "grad_norm": 0.06702706217765808,
      "learning_rate": 0.000770385,
      "loss": 1.8369335174560546,
      "step": 51360
    },
    {
      "epoch": 0.000538,
      "grad_norm": 0.08649089932441711,
      "learning_rate": 0.000770685,
      "loss": 1.8364547729492187,
      "step": 51380
    },
    {
      "epoch": 0.00054,
      "grad_norm": 0.0951608344912529,
      "learning_rate": 0.000770985,
      "loss": 1.8306005477905274,
      "step": 51400
    },
    {
      "epoch": 0.000542,
      "grad_norm": 0.09944529086351395,
      "learning_rate": 0.000771285,
      "loss": 1.8428287506103516,
      "step": 51420
    },
    {
      "epoch": 0.000544,
      "grad_norm": 0.07423659414052963,
      "learning_rate": 0.0007715850000000001,
      "loss": 1.8415821075439454,
      "step": 51440
    },
    {
      "epoch": 0.000546,
      "grad_norm": 0.08337030559778214,
      "learning_rate": 0.000771885,
      "loss": 1.8288591384887696,
      "step": 51460
    },
    {
      "epoch": 0.000548,
      "grad_norm": 0.06490614265203476,
      "learning_rate": 0.000772185,
      "loss": 1.8359104156494142,
      "step": 51480
    },
    {
      "epoch": 0.00055,
      "grad_norm": 0.09034062176942825,
      "learning_rate": 0.000772485,
      "loss": 1.8376876831054687,
      "step": 51500
    },
    {
      "epoch": 0.000552,
      "grad_norm": 0.11499785631895065,
      "learning_rate": 0.0007727850000000001,
      "loss": 1.8324405670166015,
      "step": 51520
    },
    {
      "epoch": 0.000554,
      "grad_norm": 0.10064467787742615,
      "learning_rate": 0.0007730850000000001,
      "loss": 1.8379056930541993,
      "step": 51540
    },
    {
      "epoch": 0.000556,
      "grad_norm": 0.08637844771146774,
      "learning_rate": 0.000773385,
      "loss": 1.8284292221069336,
      "step": 51560
    },
    {
      "epoch": 0.000558,
      "grad_norm": 0.08579288423061371,
      "learning_rate": 0.000773685,
      "loss": 1.8326189041137695,
      "step": 51580
    },
    {
      "epoch": 0.00056,
      "grad_norm": 0.08832810819149017,
      "learning_rate": 0.0007739849999999999,
      "loss": 1.827412223815918,
      "step": 51600
    },
    {
      "epoch": 0.000562,
      "grad_norm": 0.06998966634273529,
      "learning_rate": 0.0007742850000000001,
      "loss": 1.8301448822021484,
      "step": 51620
    },
    {
      "epoch": 0.000564,
      "grad_norm": 0.06854008883237839,
      "learning_rate": 0.000774585,
      "loss": 1.8331510543823242,
      "step": 51640
    },
    {
      "epoch": 0.000566,
      "grad_norm": 0.09752868115901947,
      "learning_rate": 0.000774885,
      "loss": 1.8248409271240233,
      "step": 51660
    },
    {
      "epoch": 0.000568,
      "grad_norm": 0.07244161516427994,
      "learning_rate": 0.000775185,
      "loss": 1.8324460983276367,
      "step": 51680
    },
    {
      "epoch": 0.00057,
      "grad_norm": 0.15373754501342773,
      "learning_rate": 0.0007754849999999999,
      "loss": 1.8331266403198243,
      "step": 51700
    },
    {
      "epoch": 0.000572,
      "grad_norm": 0.08941349387168884,
      "learning_rate": 0.0007757850000000001,
      "loss": 1.8267711639404296,
      "step": 51720
    },
    {
      "epoch": 0.000574,
      "grad_norm": 0.07003956288099289,
      "learning_rate": 0.000776085,
      "loss": 1.8321565628051757,
      "step": 51740
    },
    {
      "epoch": 0.000576,
      "grad_norm": 0.11117007583379745,
      "learning_rate": 0.000776385,
      "loss": 1.8288171768188477,
      "step": 51760
    },
    {
      "epoch": 0.000578,
      "grad_norm": 0.09099876880645752,
      "learning_rate": 0.0007766849999999999,
      "loss": 1.8312715530395507,
      "step": 51780
    },
    {
      "epoch": 0.00058,
      "grad_norm": 0.07872996479272842,
      "learning_rate": 0.0007769849999999999,
      "loss": 1.8285728454589845,
      "step": 51800
    },
    {
      "epoch": 0.000582,
      "grad_norm": 0.06602588295936584,
      "learning_rate": 0.0007772850000000001,
      "loss": 1.8299373626708983,
      "step": 51820
    },
    {
      "epoch": 0.000584,
      "grad_norm": 0.09315641224384308,
      "learning_rate": 0.000777585,
      "loss": 1.8216743469238281,
      "step": 51840
    },
    {
      "epoch": 0.000586,
      "grad_norm": 0.07509292662143707,
      "learning_rate": 0.000777885,
      "loss": 1.8301273345947267,
      "step": 51860
    },
    {
      "epoch": 0.000588,
      "grad_norm": 0.06267867982387543,
      "learning_rate": 0.0007781849999999999,
      "loss": 1.8248620986938477,
      "step": 51880
    },
    {
      "epoch": 0.00059,
      "grad_norm": 0.07973038405179977,
      "learning_rate": 0.000778485,
      "loss": 1.8356369018554688,
      "step": 51900
    },
    {
      "epoch": 0.000592,
      "grad_norm": 0.08409136533737183,
      "learning_rate": 0.000778785,
      "loss": 1.843098258972168,
      "step": 51920
    },
    {
      "epoch": 0.000594,
      "grad_norm": 0.0873011127114296,
      "learning_rate": 0.000779085,
      "loss": 1.8354225158691406,
      "step": 51940
    },
    {
      "epoch": 0.000596,
      "grad_norm": 0.07883521914482117,
      "learning_rate": 0.0007793850000000001,
      "loss": 1.8480518341064454,
      "step": 51960
    },
    {
      "epoch": 0.000598,
      "grad_norm": 0.0805177241563797,
      "learning_rate": 0.000779685,
      "loss": 1.8400152206420899,
      "step": 51980
    },
    {
      "epoch": 0.0006,
      "grad_norm": 0.08319337666034698,
      "learning_rate": 0.000779985,
      "loss": 1.8391363143920898,
      "step": 52000
    },
    {
      "epoch": 0.0006,
      "eval_loss": 1.7966078519821167,
      "eval_runtime": 99.67,
      "eval_samples_per_second": 10.033,
      "eval_steps_per_second": 10.033,
      "step": 52000
    },
    {
      "epoch": 0.000602,
      "grad_norm": 0.07502605020999908,
      "learning_rate": 0.000780285,
      "loss": 1.846877098083496,
      "step": 52020
    },
    {
      "epoch": 0.000604,
      "grad_norm": 0.09582391381263733,
      "learning_rate": 0.0007805850000000001,
      "loss": 1.8414690017700195,
      "step": 52040
    },
    {
      "epoch": 0.000606,
      "grad_norm": 0.0701102539896965,
      "learning_rate": 0.000780885,
      "loss": 1.829909896850586,
      "step": 52060
    },
    {
      "epoch": 0.000608,
      "grad_norm": 0.08239813148975372,
      "learning_rate": 0.000781185,
      "loss": 1.8336013793945312,
      "step": 52080
    },
    {
      "epoch": 0.00061,
      "grad_norm": 0.07423824071884155,
      "learning_rate": 0.000781485,
      "loss": 1.8355764389038085,
      "step": 52100
    },
    {
      "epoch": 0.000612,
      "grad_norm": 0.0922013521194458,
      "learning_rate": 0.0007817850000000001,
      "loss": 1.8389268875122071,
      "step": 52120
    },
    {
      "epoch": 0.000614,
      "grad_norm": 0.09734614938497543,
      "learning_rate": 0.0007820850000000001,
      "loss": 1.8420366287231444,
      "step": 52140
    },
    {
      "epoch": 0.000616,
      "grad_norm": 0.0785527154803276,
      "learning_rate": 0.000782385,
      "loss": 1.8435489654541015,
      "step": 52160
    },
    {
      "epoch": 0.000618,
      "grad_norm": 0.07604758441448212,
      "learning_rate": 0.000782685,
      "loss": 1.8343082427978517,
      "step": 52180
    },
    {
      "epoch": 0.00062,
      "grad_norm": 0.09369660168886185,
      "learning_rate": 0.0007829849999999999,
      "loss": 1.832362747192383,
      "step": 52200
    },
    {
      "epoch": 0.000622,
      "grad_norm": 0.06830932945013046,
      "learning_rate": 0.0007832850000000001,
      "loss": 1.8346200942993165,
      "step": 52220
    },
    {
      "epoch": 0.000624,
      "grad_norm": 0.08439896255731583,
      "learning_rate": 0.000783585,
      "loss": 1.8268022537231445,
      "step": 52240
    },
    {
      "epoch": 0.000626,
      "grad_norm": 0.08515721559524536,
      "learning_rate": 0.000783885,
      "loss": 1.8351348876953124,
      "step": 52260
    },
    {
      "epoch": 0.000628,
      "grad_norm": 0.0754883736371994,
      "learning_rate": 0.000784185,
      "loss": 1.83795166015625,
      "step": 52280
    },
    {
      "epoch": 0.00063,
      "grad_norm": 0.132758229970932,
      "learning_rate": 0.0007844849999999999,
      "loss": 1.8354440689086915,
      "step": 52300
    },
    {
      "epoch": 0.000632,
      "grad_norm": 0.08209332078695297,
      "learning_rate": 0.0007847850000000001,
      "loss": 1.8279428482055664,
      "step": 52320
    },
    {
      "epoch": 0.000634,
      "grad_norm": 0.08820170909166336,
      "learning_rate": 0.000785085,
      "loss": 1.8352716445922852,
      "step": 52340
    },
    {
      "epoch": 0.000636,
      "grad_norm": 0.09391432255506516,
      "learning_rate": 0.000785385,
      "loss": 1.838319969177246,
      "step": 52360
    },
    {
      "epoch": 0.000638,
      "grad_norm": 0.0945584699511528,
      "learning_rate": 0.0007856849999999999,
      "loss": 1.8296791076660157,
      "step": 52380
    },
    {
      "epoch": 0.00064,
      "grad_norm": 0.06577519327402115,
      "learning_rate": 0.000785985,
      "loss": 1.8366313934326173,
      "step": 52400
    },
    {
      "epoch": 0.000642,
      "grad_norm": 0.10168013721704483,
      "learning_rate": 0.0007862850000000001,
      "loss": 1.829878044128418,
      "step": 52420
    },
    {
      "epoch": 0.000644,
      "grad_norm": 0.10518037527799606,
      "learning_rate": 0.000786585,
      "loss": 1.8307470321655273,
      "step": 52440
    },
    {
      "epoch": 0.000646,
      "grad_norm": 0.10130952298641205,
      "learning_rate": 0.000786885,
      "loss": 1.822878646850586,
      "step": 52460
    },
    {
      "epoch": 0.000648,
      "grad_norm": 0.10156406462192535,
      "learning_rate": 0.000787185,
      "loss": 1.8218299865722656,
      "step": 52480
    },
    {
      "epoch": 0.00065,
      "grad_norm": 0.1337849199771881,
      "learning_rate": 0.000787485,
      "loss": 1.8343978881835938,
      "step": 52500
    },
    {
      "epoch": 0.000652,
      "grad_norm": 0.11971254646778107,
      "learning_rate": 0.000787785,
      "loss": 1.8258487701416015,
      "step": 52520
    },
    {
      "epoch": 0.000654,
      "grad_norm": 0.08975455164909363,
      "learning_rate": 0.000788085,
      "loss": 1.8330595016479492,
      "step": 52540
    },
    {
      "epoch": 0.000656,
      "grad_norm": 0.10773259401321411,
      "learning_rate": 0.0007883850000000001,
      "loss": 1.8237396240234376,
      "step": 52560
    },
    {
      "epoch": 0.000658,
      "grad_norm": 0.08031214028596878,
      "learning_rate": 0.000788685,
      "loss": 1.8248126983642579,
      "step": 52580
    },
    {
      "epoch": 0.00066,
      "grad_norm": 0.10008715838193893,
      "learning_rate": 0.000788985,
      "loss": 1.8211429595947266,
      "step": 52600
    },
    {
      "epoch": 0.000662,
      "grad_norm": 0.0784764438867569,
      "learning_rate": 0.0007892850000000001,
      "loss": 1.8209182739257812,
      "step": 52620
    },
    {
      "epoch": 0.000664,
      "grad_norm": 0.07093469053506851,
      "learning_rate": 0.0007895850000000001,
      "loss": 1.827080535888672,
      "step": 52640
    },
    {
      "epoch": 0.000666,
      "grad_norm": 0.08998201787471771,
      "learning_rate": 0.000789885,
      "loss": 1.8244873046875,
      "step": 52660
    },
    {
      "epoch": 0.000668,
      "grad_norm": 0.06582771241664886,
      "learning_rate": 0.000790185,
      "loss": 1.8138311386108399,
      "step": 52680
    },
    {
      "epoch": 0.00067,
      "grad_norm": 0.0939745232462883,
      "learning_rate": 0.000790485,
      "loss": 1.8177841186523438,
      "step": 52700
    },
    {
      "epoch": 0.000672,
      "grad_norm": 0.09036336094141006,
      "learning_rate": 0.0007907850000000001,
      "loss": 1.8163984298706055,
      "step": 52720
    },
    {
      "epoch": 0.000674,
      "grad_norm": 0.10736818611621857,
      "learning_rate": 0.0007910850000000001,
      "loss": 1.817758560180664,
      "step": 52740
    },
    {
      "epoch": 0.000676,
      "grad_norm": 0.14584214985370636,
      "learning_rate": 0.000791385,
      "loss": 1.8186847686767578,
      "step": 52760
    },
    {
      "epoch": 0.000678,
      "grad_norm": 0.09771035611629486,
      "learning_rate": 0.000791685,
      "loss": 1.828989601135254,
      "step": 52780
    },
    {
      "epoch": 0.00068,
      "grad_norm": 0.0757264718413353,
      "learning_rate": 0.0007919849999999999,
      "loss": 1.8295337677001953,
      "step": 52800
    },
    {
      "epoch": 0.000682,
      "grad_norm": 0.0830211341381073,
      "learning_rate": 0.0007922850000000001,
      "loss": 1.816922950744629,
      "step": 52820
    },
    {
      "epoch": 0.000684,
      "grad_norm": 0.13753241300582886,
      "learning_rate": 0.0007925850000000001,
      "loss": 1.8184217453002929,
      "step": 52840
    },
    {
      "epoch": 0.000686,
      "grad_norm": 0.08165712654590607,
      "learning_rate": 0.000792885,
      "loss": 1.8111629486083984,
      "step": 52860
    },
    {
      "epoch": 0.000688,
      "grad_norm": 0.08611007779836655,
      "learning_rate": 0.000793185,
      "loss": 1.8186038970947265,
      "step": 52880
    },
    {
      "epoch": 0.00069,
      "grad_norm": 0.08151885122060776,
      "learning_rate": 0.0007934849999999999,
      "loss": 1.8108427047729492,
      "step": 52900
    },
    {
      "epoch": 0.000692,
      "grad_norm": 0.07097958028316498,
      "learning_rate": 0.0007937850000000001,
      "loss": 1.8154457092285157,
      "step": 52920
    },
    {
      "epoch": 0.000694,
      "grad_norm": 0.06560239940881729,
      "learning_rate": 0.000794085,
      "loss": 1.8204071044921875,
      "step": 52940
    },
    {
      "epoch": 0.000696,
      "grad_norm": 0.0753859281539917,
      "learning_rate": 0.000794385,
      "loss": 1.8198514938354493,
      "step": 52960
    },
    {
      "epoch": 0.000698,
      "grad_norm": 0.14132261276245117,
      "learning_rate": 0.0007946849999999999,
      "loss": 1.8124382019042968,
      "step": 52980
    },
    {
      "epoch": 0.0007,
      "grad_norm": 0.07146037369966507,
      "learning_rate": 0.000794985,
      "loss": 1.8026058197021484,
      "step": 53000
    },
    {
      "epoch": 0.0007,
      "eval_loss": 1.7909702062606812,
      "eval_runtime": 98.0311,
      "eval_samples_per_second": 10.201,
      "eval_steps_per_second": 10.201,
      "step": 53000
    },
    {
      "epoch": 0.000702,
      "grad_norm": 0.06795086711645126,
      "learning_rate": 0.0007952850000000001,
      "loss": 1.8154428482055665,
      "step": 53020
    },
    {
      "epoch": 0.000704,
      "grad_norm": 0.09198163449764252,
      "learning_rate": 0.000795585,
      "loss": 1.8090250015258789,
      "step": 53040
    },
    {
      "epoch": 0.000706,
      "grad_norm": 0.07970103621482849,
      "learning_rate": 0.000795885,
      "loss": 1.8174896240234375,
      "step": 53060
    },
    {
      "epoch": 0.000708,
      "grad_norm": 0.09615591913461685,
      "learning_rate": 0.000796185,
      "loss": 1.820881462097168,
      "step": 53080
    },
    {
      "epoch": 0.00071,
      "grad_norm": 0.09093371033668518,
      "learning_rate": 0.000796485,
      "loss": 1.8219066619873048,
      "step": 53100
    },
    {
      "epoch": 0.000712,
      "grad_norm": 0.10867125540971756,
      "learning_rate": 0.000796785,
      "loss": 1.8259807586669923,
      "step": 53120
    },
    {
      "epoch": 0.000714,
      "grad_norm": 0.1117551326751709,
      "learning_rate": 0.0007970850000000001,
      "loss": 1.8272289276123046,
      "step": 53140
    },
    {
      "epoch": 0.000716,
      "grad_norm": 0.09976839274168015,
      "learning_rate": 0.0007973850000000001,
      "loss": 1.8372200012207032,
      "step": 53160
    },
    {
      "epoch": 0.000718,
      "grad_norm": 0.05409502238035202,
      "learning_rate": 0.000797685,
      "loss": 1.8297842025756836,
      "step": 53180
    },
    {
      "epoch": 0.00072,
      "grad_norm": 0.09373551607131958,
      "learning_rate": 0.000797985,
      "loss": 1.8313852310180665,
      "step": 53200
    },
    {
      "epoch": 0.000722,
      "grad_norm": 0.06937111169099808,
      "learning_rate": 0.0007982850000000001,
      "loss": 1.8413938522338866,
      "step": 53220
    },
    {
      "epoch": 0.000724,
      "grad_norm": 0.10701680183410645,
      "learning_rate": 0.0007985850000000001,
      "loss": 1.830308723449707,
      "step": 53240
    },
    {
      "epoch": 0.000726,
      "grad_norm": 0.1108751967549324,
      "learning_rate": 0.000798885,
      "loss": 1.838094139099121,
      "step": 53260
    },
    {
      "epoch": 0.000728,
      "grad_norm": 0.07272954285144806,
      "learning_rate": 0.000799185,
      "loss": 1.8260072708129882,
      "step": 53280
    },
    {
      "epoch": 0.00073,
      "grad_norm": 0.08393100649118423,
      "learning_rate": 0.000799485,
      "loss": 1.8310443878173828,
      "step": 53300
    },
    {
      "epoch": 0.000732,
      "grad_norm": 0.09824348986148834,
      "learning_rate": 0.0007997850000000001,
      "loss": 1.8303199768066407,
      "step": 53320
    },
    {
      "epoch": 0.000734,
      "grad_norm": 0.08609996736049652,
      "learning_rate": 0.0008000850000000001,
      "loss": 1.836837387084961,
      "step": 53340
    },
    {
      "epoch": 0.000736,
      "grad_norm": 0.12561504542827606,
      "learning_rate": 0.000800385,
      "loss": 1.8288450241088867,
      "step": 53360
    },
    {
      "epoch": 0.000738,
      "grad_norm": 0.09802212566137314,
      "learning_rate": 0.000800685,
      "loss": 1.8354440689086915,
      "step": 53380
    },
    {
      "epoch": 0.00074,
      "grad_norm": 0.08278319239616394,
      "learning_rate": 0.0008009849999999999,
      "loss": 1.8270971298217773,
      "step": 53400
    },
    {
      "epoch": 0.000742,
      "grad_norm": 0.13082146644592285,
      "learning_rate": 0.0008012850000000001,
      "loss": 1.8322364807128906,
      "step": 53420
    },
    {
      "epoch": 0.000744,
      "grad_norm": 0.07086450606584549,
      "learning_rate": 0.0008015850000000001,
      "loss": 1.826413917541504,
      "step": 53440
    },
    {
      "epoch": 0.000746,
      "grad_norm": 0.10642746835947037,
      "learning_rate": 0.000801885,
      "loss": 1.8282964706420899,
      "step": 53460
    },
    {
      "epoch": 0.000748,
      "grad_norm": 0.08327937126159668,
      "learning_rate": 0.000802185,
      "loss": 1.826943588256836,
      "step": 53480
    },
    {
      "epoch": 0.00075,
      "grad_norm": 0.10795646160840988,
      "learning_rate": 0.0008024849999999999,
      "loss": 1.8284631729125977,
      "step": 53500
    },
    {
      "epoch": 0.000752,
      "grad_norm": 0.08492525666952133,
      "learning_rate": 0.0008027850000000001,
      "loss": 1.8233604431152344,
      "step": 53520
    },
    {
      "epoch": 0.000754,
      "grad_norm": 0.06594720482826233,
      "learning_rate": 0.000803085,
      "loss": 1.8216796875,
      "step": 53540
    },
    {
      "epoch": 0.000756,
      "grad_norm": 0.12677860260009766,
      "learning_rate": 0.000803385,
      "loss": 1.8157302856445312,
      "step": 53560
    },
    {
      "epoch": 0.000758,
      "grad_norm": 0.09722927212715149,
      "learning_rate": 0.000803685,
      "loss": 1.8294761657714844,
      "step": 53580
    },
    {
      "epoch": 0.00076,
      "grad_norm": 0.11191555112600327,
      "learning_rate": 0.000803985,
      "loss": 1.8252492904663087,
      "step": 53600
    },
    {
      "epoch": 0.000762,
      "grad_norm": 0.11013641208410263,
      "learning_rate": 0.0008042850000000001,
      "loss": 1.8220636367797851,
      "step": 53620
    },
    {
      "epoch": 0.000764,
      "grad_norm": 0.07095854729413986,
      "learning_rate": 0.000804585,
      "loss": 1.818947982788086,
      "step": 53640
    },
    {
      "epoch": 0.000766,
      "grad_norm": 0.07869666814804077,
      "learning_rate": 0.0008048850000000001,
      "loss": 1.8186241149902345,
      "step": 53660
    },
    {
      "epoch": 0.000768,
      "grad_norm": 0.13918404281139374,
      "learning_rate": 0.000805185,
      "loss": 1.825152587890625,
      "step": 53680
    },
    {
      "epoch": 0.00077,
      "grad_norm": 0.08962216973304749,
      "learning_rate": 0.000805485,
      "loss": 1.821095848083496,
      "step": 53700
    },
    {
      "epoch": 0.000772,
      "grad_norm": 0.06319589167833328,
      "learning_rate": 0.0008057849999999999,
      "loss": 1.8147672653198241,
      "step": 53720
    },
    {
      "epoch": 0.000774,
      "grad_norm": 0.06963087618350983,
      "learning_rate": 0.0008060850000000001,
      "loss": 1.8127172470092774,
      "step": 53740
    },
    {
      "epoch": 0.000776,
      "grad_norm": 0.099605992436409,
      "learning_rate": 0.0008063850000000001,
      "loss": 1.8179948806762696,
      "step": 53760
    },
    {
      "epoch": 0.000778,
      "grad_norm": 0.13146516680717468,
      "learning_rate": 0.000806685,
      "loss": 1.8190895080566407,
      "step": 53780
    },
    {
      "epoch": 0.00078,
      "grad_norm": 0.09383892267942429,
      "learning_rate": 0.000806985,
      "loss": 1.8101465225219726,
      "step": 53800
    },
    {
      "epoch": 0.000782,
      "grad_norm": 0.06841962784528732,
      "learning_rate": 0.0008072849999999999,
      "loss": 1.8121726989746094,
      "step": 53820
    },
    {
      "epoch": 0.000784,
      "grad_norm": 0.09442818909883499,
      "learning_rate": 0.0008075850000000001,
      "loss": 1.8150535583496095,
      "step": 53840
    },
    {
      "epoch": 0.000786,
      "grad_norm": 0.07694031298160553,
      "learning_rate": 0.000807885,
      "loss": 1.8190279006958008,
      "step": 53860
    },
    {
      "epoch": 0.000788,
      "grad_norm": 0.087990403175354,
      "learning_rate": 0.000808185,
      "loss": 1.8216196060180665,
      "step": 53880
    },
    {
      "epoch": 0.00079,
      "grad_norm": 0.08122462034225464,
      "learning_rate": 0.000808485,
      "loss": 1.81506404876709,
      "step": 53900
    },
    {
      "epoch": 0.000792,
      "grad_norm": 0.08414269983768463,
      "learning_rate": 0.0008087849999999999,
      "loss": 1.8170902252197265,
      "step": 53920
    },
    {
      "epoch": 0.000794,
      "grad_norm": 0.10983981192111969,
      "learning_rate": 0.0008090850000000001,
      "loss": 1.8090068817138671,
      "step": 53940
    },
    {
      "epoch": 0.000796,
      "grad_norm": 0.10096807777881622,
      "learning_rate": 0.000809385,
      "loss": 1.812697982788086,
      "step": 53960
    },
    {
      "epoch": 0.000798,
      "grad_norm": 0.10480429977178574,
      "learning_rate": 0.000809685,
      "loss": 1.8131752014160156,
      "step": 53980
    },
    {
      "epoch": 0.0008,
      "grad_norm": 0.0834827795624733,
      "learning_rate": 0.0008099849999999999,
      "loss": 1.8161527633666992,
      "step": 54000
    },
    {
      "epoch": 0.0008,
      "eval_loss": 1.7843924760818481,
      "eval_runtime": 101.0883,
      "eval_samples_per_second": 9.892,
      "eval_steps_per_second": 9.892,
      "step": 54000
    },
    {
      "epoch": 0.000802,
      "grad_norm": 0.09298814833164215,
      "learning_rate": 0.0008102849999999999,
      "loss": 1.8114311218261718,
      "step": 54020
    },
    {
      "epoch": 0.000804,
      "grad_norm": 0.0750308409333229,
      "learning_rate": 0.0008105850000000001,
      "loss": 1.813578224182129,
      "step": 54040
    },
    {
      "epoch": 0.000806,
      "grad_norm": 0.08869277685880661,
      "learning_rate": 0.000810885,
      "loss": 1.8096160888671875,
      "step": 54060
    },
    {
      "epoch": 0.000808,
      "grad_norm": 0.08057360351085663,
      "learning_rate": 0.000811185,
      "loss": 1.8082204818725587,
      "step": 54080
    },
    {
      "epoch": 0.00081,
      "grad_norm": 0.07752317935228348,
      "learning_rate": 0.0008114849999999999,
      "loss": 1.809225082397461,
      "step": 54100
    },
    {
      "epoch": 0.000812,
      "grad_norm": 0.10853946208953857,
      "learning_rate": 0.000811785,
      "loss": 1.8025445938110352,
      "step": 54120
    },
    {
      "epoch": 0.000814,
      "grad_norm": 0.08279047906398773,
      "learning_rate": 0.000812085,
      "loss": 1.8129579544067382,
      "step": 54140
    },
    {
      "epoch": 0.000816,
      "grad_norm": 0.08732631057500839,
      "learning_rate": 0.000812385,
      "loss": 1.821772003173828,
      "step": 54160
    },
    {
      "epoch": 0.000818,
      "grad_norm": 0.06940227001905441,
      "learning_rate": 0.0008126850000000001,
      "loss": 1.8139148712158204,
      "step": 54180
    },
    {
      "epoch": 0.00082,
      "grad_norm": 0.09871824830770493,
      "learning_rate": 0.000812985,
      "loss": 1.8094743728637694,
      "step": 54200
    },
    {
      "epoch": 0.000822,
      "grad_norm": 0.0697028711438179,
      "learning_rate": 0.000813285,
      "loss": 1.8084943771362305,
      "step": 54220
    },
    {
      "epoch": 0.000824,
      "grad_norm": 0.10332415997982025,
      "learning_rate": 0.000813585,
      "loss": 1.8151615142822266,
      "step": 54240
    },
    {
      "epoch": 0.000826,
      "grad_norm": 0.12199199944734573,
      "learning_rate": 0.0008138850000000001,
      "loss": 1.81812744140625,
      "step": 54260
    },
    {
      "epoch": 0.000828,
      "grad_norm": 0.07809198647737503,
      "learning_rate": 0.000814185,
      "loss": 1.8145551681518555,
      "step": 54280
    },
    {
      "epoch": 0.00083,
      "grad_norm": 0.07807798683643341,
      "learning_rate": 0.000814485,
      "loss": 1.8237224578857423,
      "step": 54300
    },
    {
      "epoch": 0.000832,
      "grad_norm": 0.09394682198762894,
      "learning_rate": 0.0008147849999999999,
      "loss": 1.823185920715332,
      "step": 54320
    },
    {
      "epoch": 0.000834,
      "grad_norm": 0.07117526233196259,
      "learning_rate": 0.0008150850000000001,
      "loss": 1.8289709091186523,
      "step": 54340
    },
    {
      "epoch": 0.000836,
      "grad_norm": 0.09141800552606583,
      "learning_rate": 0.0008153850000000001,
      "loss": 1.8260204315185546,
      "step": 54360
    },
    {
      "epoch": 0.000838,
      "grad_norm": 0.08138151466846466,
      "learning_rate": 0.000815685,
      "loss": 1.8253677368164063,
      "step": 54380
    },
    {
      "epoch": 0.00084,
      "grad_norm": 0.1094491258263588,
      "learning_rate": 0.000815985,
      "loss": 1.8203096389770508,
      "step": 54400
    },
    {
      "epoch": 0.000842,
      "grad_norm": 0.11064698547124863,
      "learning_rate": 0.0008162849999999999,
      "loss": 1.82598876953125,
      "step": 54420
    },
    {
      "epoch": 0.000844,
      "grad_norm": 0.0654069036245346,
      "learning_rate": 0.0008165850000000001,
      "loss": 1.8315662384033202,
      "step": 54440
    },
    {
      "epoch": 0.000846,
      "grad_norm": 0.08133042603731155,
      "learning_rate": 0.000816885,
      "loss": 1.8194246292114258,
      "step": 54460
    },
    {
      "epoch": 0.000848,
      "grad_norm": 0.10478977859020233,
      "learning_rate": 0.000817185,
      "loss": 1.829688835144043,
      "step": 54480
    },
    {
      "epoch": 0.00085,
      "grad_norm": 0.07572335749864578,
      "learning_rate": 0.000817485,
      "loss": 1.8275995254516602,
      "step": 54500
    },
    {
      "epoch": 0.000852,
      "grad_norm": 0.07680771499872208,
      "learning_rate": 0.0008177849999999999,
      "loss": 1.8319358825683594,
      "step": 54520
    },
    {
      "epoch": 0.000854,
      "grad_norm": 0.07684208452701569,
      "learning_rate": 0.0008180850000000001,
      "loss": 1.8229276657104492,
      "step": 54540
    },
    {
      "epoch": 0.000856,
      "grad_norm": 0.09422148764133453,
      "learning_rate": 0.000818385,
      "loss": 1.8243440628051757,
      "step": 54560
    },
    {
      "epoch": 0.000858,
      "grad_norm": 0.10780831426382065,
      "learning_rate": 0.000818685,
      "loss": 1.823923110961914,
      "step": 54580
    },
    {
      "epoch": 0.00086,
      "grad_norm": 0.07486619055271149,
      "learning_rate": 0.0008189849999999999,
      "loss": 1.8238357543945312,
      "step": 54600
    },
    {
      "epoch": 0.000862,
      "grad_norm": 0.07407692819833755,
      "learning_rate": 0.000819285,
      "loss": 1.8205194473266602,
      "step": 54620
    },
    {
      "epoch": 0.000864,
      "grad_norm": 0.09588978439569473,
      "learning_rate": 0.0008195850000000001,
      "loss": 1.8264301300048829,
      "step": 54640
    },
    {
      "epoch": 0.000866,
      "grad_norm": 0.06922370940446854,
      "learning_rate": 0.000819885,
      "loss": 1.8174089431762694,
      "step": 54660
    },
    {
      "epoch": 0.000868,
      "grad_norm": 0.06955870240926743,
      "learning_rate": 0.000820185,
      "loss": 1.8233623504638672,
      "step": 54680
    },
    {
      "epoch": 0.00087,
      "grad_norm": 0.09113681316375732,
      "learning_rate": 0.000820485,
      "loss": 1.8206260681152344,
      "step": 54700
    },
    {
      "epoch": 0.000872,
      "grad_norm": 0.11233562231063843,
      "learning_rate": 0.000820785,
      "loss": 1.8202404022216796,
      "step": 54720
    },
    {
      "epoch": 0.000874,
      "grad_norm": 0.0829581469297409,
      "learning_rate": 0.000821085,
      "loss": 1.8094083786010742,
      "step": 54740
    },
    {
      "epoch": 0.000876,
      "grad_norm": 0.11381429433822632,
      "learning_rate": 0.000821385,
      "loss": 1.8112443923950194,
      "step": 54760
    },
    {
      "epoch": 0.000878,
      "grad_norm": 0.0887814611196518,
      "learning_rate": 0.0008216850000000001,
      "loss": 1.8204311370849608,
      "step": 54780
    },
    {
      "epoch": 0.00088,
      "grad_norm": 0.06194445863366127,
      "learning_rate": 0.000821985,
      "loss": 1.8063531875610352,
      "step": 54800
    },
    {
      "epoch": 0.000882,
      "grad_norm": 0.14742892980575562,
      "learning_rate": 0.000822285,
      "loss": 1.8119380950927735,
      "step": 54820
    },
    {
      "epoch": 0.000884,
      "grad_norm": 0.11888439953327179,
      "learning_rate": 0.0008225850000000001,
      "loss": 1.8191036224365233,
      "step": 54840
    },
    {
      "epoch": 0.000886,
      "grad_norm": 0.10441844910383224,
      "learning_rate": 0.0008228850000000001,
      "loss": 1.8138225555419922,
      "step": 54860
    },
    {
      "epoch": 0.000888,
      "grad_norm": 0.07273080199956894,
      "learning_rate": 0.000823185,
      "loss": 1.8145854949951172,
      "step": 54880
    },
    {
      "epoch": 0.00089,
      "grad_norm": 0.09075014293193817,
      "learning_rate": 0.000823485,
      "loss": 1.8046476364135742,
      "step": 54900
    },
    {
      "epoch": 0.000892,
      "grad_norm": 0.07432058453559875,
      "learning_rate": 0.000823785,
      "loss": 1.8108104705810546,
      "step": 54920
    },
    {
      "epoch": 0.000894,
      "grad_norm": 0.06345789134502411,
      "learning_rate": 0.0008240850000000001,
      "loss": 1.8109086990356444,
      "step": 54940
    },
    {
      "epoch": 0.000896,
      "grad_norm": 0.08759179711341858,
      "learning_rate": 0.0008243850000000001,
      "loss": 1.8053667068481445,
      "step": 54960
    },
    {
      "epoch": 0.000898,
      "grad_norm": 0.07735589891672134,
      "learning_rate": 0.000824685,
      "loss": 1.8095775604248048,
      "step": 54980
    },
    {
      "epoch": 0.0009,
      "grad_norm": 0.06968027353286743,
      "learning_rate": 0.000824985,
      "loss": 1.8140180587768555,
      "step": 55000
    },
    {
      "epoch": 0.0009,
      "eval_loss": 1.7782225608825684,
      "eval_runtime": 105.1123,
      "eval_samples_per_second": 9.514,
      "eval_steps_per_second": 9.514,
      "step": 55000
    },
    {
      "epoch": 0.000902,
      "grad_norm": 0.09342938661575317,
      "learning_rate": 0.0008252849999999999,
      "loss": 1.806143569946289,
      "step": 55020
    },
    {
      "epoch": 0.000904,
      "grad_norm": 0.12704946100711823,
      "learning_rate": 0.0008255850000000001,
      "loss": 1.8222143173217773,
      "step": 55040
    },
    {
      "epoch": 0.000906,
      "grad_norm": 0.08495233952999115,
      "learning_rate": 0.000825885,
      "loss": 1.8138565063476562,
      "step": 55060
    },
    {
      "epoch": 0.000908,
      "grad_norm": 0.0882241502404213,
      "learning_rate": 0.000826185,
      "loss": 1.816044807434082,
      "step": 55080
    },
    {
      "epoch": 0.00091,
      "grad_norm": 0.07508769631385803,
      "learning_rate": 0.000826485,
      "loss": 1.8081441879272462,
      "step": 55100
    },
    {
      "epoch": 0.000912,
      "grad_norm": 0.07477625459432602,
      "learning_rate": 0.0008267849999999999,
      "loss": 1.813369369506836,
      "step": 55120
    },
    {
      "epoch": 0.000914,
      "grad_norm": 0.10097239911556244,
      "learning_rate": 0.0008270850000000001,
      "loss": 1.807389259338379,
      "step": 55140
    },
    {
      "epoch": 0.000916,
      "grad_norm": 0.0737549215555191,
      "learning_rate": 0.000827385,
      "loss": 1.8057111740112304,
      "step": 55160
    },
    {
      "epoch": 0.000918,
      "grad_norm": 0.09195511788129807,
      "learning_rate": 0.000827685,
      "loss": 1.8020875930786133,
      "step": 55180
    },
    {
      "epoch": 0.00092,
      "grad_norm": 0.06454228609800339,
      "learning_rate": 0.0008279849999999999,
      "loss": 1.8037670135498047,
      "step": 55200
    },
    {
      "epoch": 0.000922,
      "grad_norm": 0.08498887717723846,
      "learning_rate": 0.000828285,
      "loss": 1.8082290649414063,
      "step": 55220
    },
    {
      "epoch": 0.000924,
      "grad_norm": 0.07089615613222122,
      "learning_rate": 0.0008285850000000001,
      "loss": 1.8005500793457032,
      "step": 55240
    },
    {
      "epoch": 0.000926,
      "grad_norm": 0.08960192650556564,
      "learning_rate": 0.000828885,
      "loss": 1.7943479537963867,
      "step": 55260
    },
    {
      "epoch": 0.000928,
      "grad_norm": 0.08665843307971954,
      "learning_rate": 0.000829185,
      "loss": 1.794640350341797,
      "step": 55280
    },
    {
      "epoch": 0.00093,
      "grad_norm": 0.08016286045312881,
      "learning_rate": 0.000829485,
      "loss": 1.8026283264160157,
      "step": 55300
    },
    {
      "epoch": 0.000932,
      "grad_norm": 0.06712556630373001,
      "learning_rate": 0.000829785,
      "loss": 1.7970748901367188,
      "step": 55320
    },
    {
      "epoch": 0.000934,
      "grad_norm": 0.09190410375595093,
      "learning_rate": 0.000830085,
      "loss": 1.7996068954467774,
      "step": 55340
    },
    {
      "epoch": 0.000936,
      "grad_norm": 0.06959818303585052,
      "learning_rate": 0.0008303850000000001,
      "loss": 1.8049400329589844,
      "step": 55360
    },
    {
      "epoch": 0.000938,
      "grad_norm": 0.09189862757921219,
      "learning_rate": 0.0008306850000000001,
      "loss": 1.8001893997192382,
      "step": 55380
    },
    {
      "epoch": 0.00094,
      "grad_norm": 0.0734458640217781,
      "learning_rate": 0.000830985,
      "loss": 1.8056737899780273,
      "step": 55400
    },
    {
      "epoch": 0.000942,
      "grad_norm": 0.08112155646085739,
      "learning_rate": 0.000831285,
      "loss": 1.8051637649536132,
      "step": 55420
    },
    {
      "epoch": 0.000944,
      "grad_norm": 0.10185881704092026,
      "learning_rate": 0.0008315850000000001,
      "loss": 1.8078388214111327,
      "step": 55440
    },
    {
      "epoch": 0.000946,
      "grad_norm": 0.07119335979223251,
      "learning_rate": 0.0008318850000000001,
      "loss": 1.8174514770507812,
      "step": 55460
    },
    {
      "epoch": 0.000948,
      "grad_norm": 0.06879117339849472,
      "learning_rate": 0.000832185,
      "loss": 1.8166082382202149,
      "step": 55480
    },
    {
      "epoch": 0.00095,
      "grad_norm": 0.07796020805835724,
      "learning_rate": 0.000832485,
      "loss": 1.811993408203125,
      "step": 55500
    },
    {
      "epoch": 0.000952,
      "grad_norm": 0.08357212692499161,
      "learning_rate": 0.000832785,
      "loss": 1.8183017730712892,
      "step": 55520
    },
    {
      "epoch": 0.000954,
      "grad_norm": 0.08831249177455902,
      "learning_rate": 0.0008330850000000001,
      "loss": 1.8182289123535156,
      "step": 55540
    },
    {
      "epoch": 0.000956,
      "grad_norm": 0.06869123876094818,
      "learning_rate": 0.0008333850000000001,
      "loss": 1.8175786972045898,
      "step": 55560
    },
    {
      "epoch": 0.000958,
      "grad_norm": 0.10531241446733475,
      "learning_rate": 0.000833685,
      "loss": 1.8232967376708984,
      "step": 55580
    },
    {
      "epoch": 0.00096,
      "grad_norm": 0.0827803835272789,
      "learning_rate": 0.000833985,
      "loss": 1.8102272033691407,
      "step": 55600
    },
    {
      "epoch": 0.000962,
      "grad_norm": 0.09131187945604324,
      "learning_rate": 0.0008342849999999999,
      "loss": 1.8116649627685546,
      "step": 55620
    },
    {
      "epoch": 0.000964,
      "grad_norm": 0.09667201340198517,
      "learning_rate": 0.0008345850000000001,
      "loss": 1.814039421081543,
      "step": 55640
    },
    {
      "epoch": 0.000966,
      "grad_norm": 0.08257482945919037,
      "learning_rate": 0.000834885,
      "loss": 1.8080272674560547,
      "step": 55660
    },
    {
      "epoch": 0.000968,
      "grad_norm": 0.1165599524974823,
      "learning_rate": 0.000835185,
      "loss": 1.8163698196411133,
      "step": 55680
    },
    {
      "epoch": 0.00097,
      "grad_norm": 0.08054927736520767,
      "learning_rate": 0.000835485,
      "loss": 1.820367431640625,
      "step": 55700
    },
    {
      "epoch": 0.000972,
      "grad_norm": 0.11378245800733566,
      "learning_rate": 0.0008357849999999999,
      "loss": 1.8148984909057617,
      "step": 55720
    },
    {
      "epoch": 0.000974,
      "grad_norm": 0.08413427323102951,
      "learning_rate": 0.0008360850000000001,
      "loss": 1.8135534286499024,
      "step": 55740
    },
    {
      "epoch": 0.000976,
      "grad_norm": 0.09522600471973419,
      "learning_rate": 0.000836385,
      "loss": 1.806572151184082,
      "step": 55760
    },
    {
      "epoch": 0.000978,
      "grad_norm": 0.12314817309379578,
      "learning_rate": 0.000836685,
      "loss": 1.8052108764648438,
      "step": 55780
    },
    {
      "epoch": 0.00098,
      "grad_norm": 0.12068308889865875,
      "learning_rate": 0.0008369849999999999,
      "loss": 1.8041009902954102,
      "step": 55800
    },
    {
      "epoch": 0.000982,
      "grad_norm": 0.06996431946754456,
      "learning_rate": 0.000837285,
      "loss": 1.8152761459350586,
      "step": 55820
    },
    {
      "epoch": 0.000984,
      "grad_norm": 0.09179938584566116,
      "learning_rate": 0.0008375850000000001,
      "loss": 1.8125650405883789,
      "step": 55840
    },
    {
      "epoch": 0.000986,
      "grad_norm": 0.09448322653770447,
      "learning_rate": 0.000837885,
      "loss": 1.8097614288330077,
      "step": 55860
    },
    {
      "epoch": 0.000988,
      "grad_norm": 0.09040167927742004,
      "learning_rate": 0.0008381850000000001,
      "loss": 1.8013286590576172,
      "step": 55880
    },
    {
      "epoch": 0.00099,
      "grad_norm": 0.08906266838312149,
      "learning_rate": 0.000838485,
      "loss": 1.8066373825073243,
      "step": 55900
    },
    {
      "epoch": 0.000992,
      "grad_norm": 0.1082487404346466,
      "learning_rate": 0.000838785,
      "loss": 1.8038511276245117,
      "step": 55920
    },
    {
      "epoch": 0.000994,
      "grad_norm": 0.09656911343336105,
      "learning_rate": 0.000839085,
      "loss": 1.8036479949951172,
      "step": 55940
    },
    {
      "epoch": 0.000996,
      "grad_norm": 0.10782207548618317,
      "learning_rate": 0.0008393850000000001,
      "loss": 1.8062810897827148,
      "step": 55960
    },
    {
      "epoch": 0.000998,
      "grad_norm": 0.0980667844414711,
      "learning_rate": 0.0008396850000000001,
      "loss": 1.806480598449707,
      "step": 55980
    },
    {
      "epoch": 0.001,
      "grad_norm": 0.07666339725255966,
      "learning_rate": 0.000839985,
      "loss": 1.807319450378418,
      "step": 56000
    },
    {
      "epoch": 0.001,
      "eval_loss": 1.7735828161239624,
      "eval_runtime": 100.1932,
      "eval_samples_per_second": 9.981,
      "eval_steps_per_second": 9.981,
      "step": 56000
    },
    {
      "epoch": 0.001002,
      "grad_norm": 0.06917919218540192,
      "learning_rate": 0.000840285,
      "loss": 1.8006320953369142,
      "step": 56020
    },
    {
      "epoch": 0.001004,
      "grad_norm": 0.10188503563404083,
      "learning_rate": 0.0008405850000000001,
      "loss": 1.7991653442382813,
      "step": 56040
    },
    {
      "epoch": 0.001006,
      "grad_norm": 0.10053511708974838,
      "learning_rate": 0.0008408850000000001,
      "loss": 1.8023815155029297,
      "step": 56060
    },
    {
      "epoch": 0.001008,
      "grad_norm": 0.0830378606915474,
      "learning_rate": 0.000841185,
      "loss": 1.7994457244873048,
      "step": 56080
    },
    {
      "epoch": 0.00101,
      "grad_norm": 0.10750024765729904,
      "learning_rate": 0.000841485,
      "loss": 1.8020074844360352,
      "step": 56100
    },
    {
      "epoch": 0.001012,
      "grad_norm": 0.09629049152135849,
      "learning_rate": 0.000841785,
      "loss": 1.8058887481689454,
      "step": 56120
    },
    {
      "epoch": 0.001014,
      "grad_norm": 0.09682925790548325,
      "learning_rate": 0.0008420849999999999,
      "loss": 1.8038902282714844,
      "step": 56140
    },
    {
      "epoch": 0.001016,
      "grad_norm": 0.0911613404750824,
      "learning_rate": 0.0008423850000000001,
      "loss": 1.804712677001953,
      "step": 56160
    },
    {
      "epoch": 0.001018,
      "grad_norm": 0.08601225167512894,
      "learning_rate": 0.000842685,
      "loss": 1.7994209289550782,
      "step": 56180
    },
    {
      "epoch": 0.00102,
      "grad_norm": 0.10935106873512268,
      "learning_rate": 0.000842985,
      "loss": 1.7846817016601562,
      "step": 56200
    },
    {
      "epoch": 0.001022,
      "grad_norm": 0.06160512566566467,
      "learning_rate": 0.0008432849999999999,
      "loss": 1.7995597839355468,
      "step": 56220
    },
    {
      "epoch": 0.001024,
      "grad_norm": 0.09608394652605057,
      "learning_rate": 0.0008435849999999999,
      "loss": 1.8008371353149415,
      "step": 56240
    },
    {
      "epoch": 0.001026,
      "grad_norm": 0.07297331094741821,
      "learning_rate": 0.0008438850000000001,
      "loss": 1.7997770309448242,
      "step": 56260
    },
    {
      "epoch": 0.001028,
      "grad_norm": 0.06868540495634079,
      "learning_rate": 0.000844185,
      "loss": 1.7942707061767578,
      "step": 56280
    },
    {
      "epoch": 0.00103,
      "grad_norm": 0.09523054957389832,
      "learning_rate": 0.000844485,
      "loss": 1.8014406204223632,
      "step": 56300
    },
    {
      "epoch": 0.001032,
      "grad_norm": 0.07361695170402527,
      "learning_rate": 0.000844785,
      "loss": 1.7993066787719727,
      "step": 56320
    },
    {
      "epoch": 0.001034,
      "grad_norm": 0.07631966471672058,
      "learning_rate": 0.000845085,
      "loss": 1.7900287628173828,
      "step": 56340
    },
    {
      "epoch": 0.001036,
      "grad_norm": 0.07280699908733368,
      "learning_rate": 0.000845385,
      "loss": 1.7911439895629884,
      "step": 56360
    },
    {
      "epoch": 0.001038,
      "grad_norm": 0.07538430392742157,
      "learning_rate": 0.000845685,
      "loss": 1.802274513244629,
      "step": 56380
    },
    {
      "epoch": 0.00104,
      "grad_norm": 0.1117163896560669,
      "learning_rate": 0.000845985,
      "loss": 1.8023576736450195,
      "step": 56400
    },
    {
      "epoch": 0.001042,
      "grad_norm": 0.08323267102241516,
      "learning_rate": 0.000846285,
      "loss": 1.796982192993164,
      "step": 56420
    },
    {
      "epoch": 0.001044,
      "grad_norm": 0.08631003648042679,
      "learning_rate": 0.000846585,
      "loss": 1.7945629119873048,
      "step": 56440
    },
    {
      "epoch": 0.001046,
      "grad_norm": 0.09242662787437439,
      "learning_rate": 0.000846885,
      "loss": 1.7980558395385742,
      "step": 56460
    },
    {
      "epoch": 0.001048,
      "grad_norm": 0.09422673285007477,
      "learning_rate": 0.0008471850000000001,
      "loss": 1.790669822692871,
      "step": 56480
    },
    {
      "epoch": 0.00105,
      "grad_norm": 0.08711153268814087,
      "learning_rate": 0.000847485,
      "loss": 1.7953336715698243,
      "step": 56500
    },
    {
      "epoch": 0.001052,
      "grad_norm": 0.0852414071559906,
      "learning_rate": 0.000847785,
      "loss": 1.7831697463989258,
      "step": 56520
    },
    {
      "epoch": 0.001054,
      "grad_norm": 0.1263018548488617,
      "learning_rate": 0.0008480849999999999,
      "loss": 1.7890201568603517,
      "step": 56540
    },
    {
      "epoch": 0.001056,
      "grad_norm": 0.08097770065069199,
      "learning_rate": 0.0008483850000000001,
      "loss": 1.7894300460815429,
      "step": 56560
    },
    {
      "epoch": 0.001058,
      "grad_norm": 0.08591044694185257,
      "learning_rate": 0.0008486850000000001,
      "loss": 1.7861698150634766,
      "step": 56580
    },
    {
      "epoch": 0.00106,
      "grad_norm": 0.1244858130812645,
      "learning_rate": 0.000848985,
      "loss": 1.7880922317504884,
      "step": 56600
    },
    {
      "epoch": 0.001062,
      "grad_norm": 0.07644981890916824,
      "learning_rate": 0.000849285,
      "loss": 1.7950071334838866,
      "step": 56620
    },
    {
      "epoch": 0.001064,
      "grad_norm": 0.06376186013221741,
      "learning_rate": 0.0008495849999999999,
      "loss": 1.7975915908813476,
      "step": 56640
    },
    {
      "epoch": 0.001066,
      "grad_norm": 0.08314666152000427,
      "learning_rate": 0.0008498850000000001,
      "loss": 1.8020807266235352,
      "step": 56660
    },
    {
      "epoch": 0.001068,
      "grad_norm": 0.0650612935423851,
      "learning_rate": 0.000850185,
      "loss": 1.8081113815307617,
      "step": 56680
    },
    {
      "epoch": 0.00107,
      "grad_norm": 0.09447851032018661,
      "learning_rate": 0.000850485,
      "loss": 1.8041149139404298,
      "step": 56700
    },
    {
      "epoch": 0.001072,
      "grad_norm": 0.12133060395717621,
      "learning_rate": 0.000850785,
      "loss": 1.8069007873535157,
      "step": 56720
    },
    {
      "epoch": 0.001074,
      "grad_norm": 0.10875920951366425,
      "learning_rate": 0.0008510849999999999,
      "loss": 1.8077239990234375,
      "step": 56740
    },
    {
      "epoch": 0.001076,
      "grad_norm": 0.09694947302341461,
      "learning_rate": 0.0008513850000000001,
      "loss": 1.8075384140014648,
      "step": 56760
    },
    {
      "epoch": 0.001078,
      "grad_norm": 0.06666617840528488,
      "learning_rate": 0.000851685,
      "loss": 1.8122877120971679,
      "step": 56780
    },
    {
      "epoch": 0.00108,
      "grad_norm": 0.08390645682811737,
      "learning_rate": 0.000851985,
      "loss": 1.803468894958496,
      "step": 56800
    },
    {
      "epoch": 0.001082,
      "grad_norm": 0.10730978846549988,
      "learning_rate": 0.0008522849999999999,
      "loss": 1.8112421035766602,
      "step": 56820
    },
    {
      "epoch": 0.001084,
      "grad_norm": 0.08279474824666977,
      "learning_rate": 0.000852585,
      "loss": 1.8050979614257812,
      "step": 56840
    },
    {
      "epoch": 0.001086,
      "grad_norm": 0.07948900759220123,
      "learning_rate": 0.0008528850000000001,
      "loss": 1.7999177932739259,
      "step": 56860
    },
    {
      "epoch": 0.001088,
      "grad_norm": 0.06034329906105995,
      "learning_rate": 0.000853185,
      "loss": 1.8109682083129883,
      "step": 56880
    },
    {
      "epoch": 0.00109,
      "grad_norm": 0.1001698300242424,
      "learning_rate": 0.000853485,
      "loss": 1.808699607849121,
      "step": 56900
    },
    {
      "epoch": 0.001092,
      "grad_norm": 0.09527815878391266,
      "learning_rate": 0.000853785,
      "loss": 1.7977937698364257,
      "step": 56920
    },
    {
      "epoch": 0.001094,
      "grad_norm": 0.1068359762430191,
      "learning_rate": 0.000854085,
      "loss": 1.797940444946289,
      "step": 56940
    },
    {
      "epoch": 0.001096,
      "grad_norm": 0.09304238855838776,
      "learning_rate": 0.000854385,
      "loss": 1.8067825317382813,
      "step": 56960
    },
    {
      "epoch": 0.001098,
      "grad_norm": 0.07920155674219131,
      "learning_rate": 0.000854685,
      "loss": 1.8016019821166993,
      "step": 56980
    },
    {
      "epoch": 0.0011,
      "grad_norm": 0.08701524138450623,
      "learning_rate": 0.0008549850000000001,
      "loss": 1.8006498336791992,
      "step": 57000
    },
    {
      "epoch": 0.0011,
      "eval_loss": 1.7683178186416626,
      "eval_runtime": 99.6685,
      "eval_samples_per_second": 10.033,
      "eval_steps_per_second": 10.033,
      "step": 57000
    },
    {
      "epoch": 0.001102,
      "grad_norm": 0.0937202125787735,
      "learning_rate": 0.000855285,
      "loss": 1.801673698425293,
      "step": 57020
    },
    {
      "epoch": 0.001104,
      "grad_norm": 0.10670318454504013,
      "learning_rate": 0.000855585,
      "loss": 1.8076778411865235,
      "step": 57040
    },
    {
      "epoch": 0.001106,
      "grad_norm": 0.10937001556158066,
      "learning_rate": 0.0008558850000000001,
      "loss": 1.7999671936035155,
      "step": 57060
    },
    {
      "epoch": 0.001108,
      "grad_norm": 0.0757187157869339,
      "learning_rate": 0.0008561850000000001,
      "loss": 1.8009063720703125,
      "step": 57080
    },
    {
      "epoch": 0.00111,
      "grad_norm": 0.0944443792104721,
      "learning_rate": 0.000856485,
      "loss": 1.8032840728759765,
      "step": 57100
    },
    {
      "epoch": 0.001112,
      "grad_norm": 0.09052728861570358,
      "learning_rate": 0.000856785,
      "loss": 1.8027692794799806,
      "step": 57120
    },
    {
      "epoch": 0.001114,
      "grad_norm": 0.0733652412891388,
      "learning_rate": 0.0008570849999999999,
      "loss": 1.8043195724487304,
      "step": 57140
    },
    {
      "epoch": 0.001116,
      "grad_norm": 0.07075349986553192,
      "learning_rate": 0.0008573850000000001,
      "loss": 1.791477584838867,
      "step": 57160
    },
    {
      "epoch": 0.001118,
      "grad_norm": 0.07513801753520966,
      "learning_rate": 0.0008576850000000001,
      "loss": 1.8048818588256836,
      "step": 57180
    },
    {
      "epoch": 0.00112,
      "grad_norm": 0.06912592053413391,
      "learning_rate": 0.000857985,
      "loss": 1.7968517303466798,
      "step": 57200
    },
    {
      "epoch": 0.001122,
      "grad_norm": 0.09329573810100555,
      "learning_rate": 0.000858285,
      "loss": 1.7901058197021484,
      "step": 57220
    },
    {
      "epoch": 0.001124,
      "grad_norm": 0.06965938210487366,
      "learning_rate": 0.0008585849999999999,
      "loss": 1.8001869201660157,
      "step": 57240
    },
    {
      "epoch": 0.001126,
      "grad_norm": 0.09302552044391632,
      "learning_rate": 0.0008588850000000001,
      "loss": 1.7976009368896484,
      "step": 57260
    },
    {
      "epoch": 0.001128,
      "grad_norm": 0.09167779982089996,
      "learning_rate": 0.000859185,
      "loss": 1.7888769149780273,
      "step": 57280
    },
    {
      "epoch": 0.00113,
      "grad_norm": 0.09916319698095322,
      "learning_rate": 0.000859485,
      "loss": 1.7904916763305665,
      "step": 57300
    },
    {
      "epoch": 0.001132,
      "grad_norm": 0.09171207249164581,
      "learning_rate": 0.000859785,
      "loss": 1.7911870956420899,
      "step": 57320
    },
    {
      "epoch": 0.001134,
      "grad_norm": 0.069407157599926,
      "learning_rate": 0.0008600849999999999,
      "loss": 1.7973270416259766,
      "step": 57340
    },
    {
      "epoch": 0.001136,
      "grad_norm": 0.078703872859478,
      "learning_rate": 0.0008603850000000001,
      "loss": 1.7949394226074218,
      "step": 57360
    },
    {
      "epoch": 0.001138,
      "grad_norm": 0.10920059680938721,
      "learning_rate": 0.000860685,
      "loss": 1.7930070877075195,
      "step": 57380
    },
    {
      "epoch": 0.00114,
      "grad_norm": 0.11591575294733047,
      "learning_rate": 0.000860985,
      "loss": 1.7912139892578125,
      "step": 57400
    },
    {
      "epoch": 0.001142,
      "grad_norm": 0.12172117084264755,
      "learning_rate": 0.0008612849999999999,
      "loss": 1.7863794326782227,
      "step": 57420
    },
    {
      "epoch": 0.001144,
      "grad_norm": 0.06724731624126434,
      "learning_rate": 0.000861585,
      "loss": 1.7922708511352539,
      "step": 57440
    },
    {
      "epoch": 0.001146,
      "grad_norm": 0.11076223105192184,
      "learning_rate": 0.0008618850000000001,
      "loss": 1.7924116134643555,
      "step": 57460
    },
    {
      "epoch": 0.001148,
      "grad_norm": 0.10175531357526779,
      "learning_rate": 0.000862185,
      "loss": 1.791329574584961,
      "step": 57480
    },
    {
      "epoch": 0.00115,
      "grad_norm": 0.08731532096862793,
      "learning_rate": 0.000862485,
      "loss": 1.7879472732543946,
      "step": 57500
    },
    {
      "epoch": 0.001152,
      "grad_norm": 0.09933855384588242,
      "learning_rate": 0.000862785,
      "loss": 1.789200210571289,
      "step": 57520
    },
    {
      "epoch": 0.001154,
      "grad_norm": 0.09416564553976059,
      "learning_rate": 0.000863085,
      "loss": 1.7867406845092773,
      "step": 57540
    },
    {
      "epoch": 0.001156,
      "grad_norm": 0.11345028132200241,
      "learning_rate": 0.000863385,
      "loss": 1.7869626998901367,
      "step": 57560
    },
    {
      "epoch": 0.001158,
      "grad_norm": 0.10422199219465256,
      "learning_rate": 0.0008636850000000001,
      "loss": 1.792588233947754,
      "step": 57580
    },
    {
      "epoch": 0.00116,
      "grad_norm": 0.11995210498571396,
      "learning_rate": 0.0008639850000000001,
      "loss": 1.785790252685547,
      "step": 57600
    },
    {
      "epoch": 0.001162,
      "grad_norm": 0.08032912760972977,
      "learning_rate": 0.000864285,
      "loss": 1.7840200424194337,
      "step": 57620
    },
    {
      "epoch": 0.001164,
      "grad_norm": 0.10984625667333603,
      "learning_rate": 0.000864585,
      "loss": 1.7997098922729493,
      "step": 57640
    },
    {
      "epoch": 0.001166,
      "grad_norm": 0.08549176901578903,
      "learning_rate": 0.0008648850000000001,
      "loss": 1.7887275695800782,
      "step": 57660
    },
    {
      "epoch": 0.001168,
      "grad_norm": 0.0953085646033287,
      "learning_rate": 0.0008651850000000001,
      "loss": 1.7921417236328125,
      "step": 57680
    },
    {
      "epoch": 0.00117,
      "grad_norm": 0.08001036196947098,
      "learning_rate": 0.000865485,
      "loss": 1.7886322021484375,
      "step": 57700
    },
    {
      "epoch": 0.001172,
      "grad_norm": 0.1295686960220337,
      "learning_rate": 0.000865785,
      "loss": 1.7780420303344726,
      "step": 57720
    },
    {
      "epoch": 0.001174,
      "grad_norm": 0.09134498238563538,
      "learning_rate": 0.0008660849999999999,
      "loss": 1.7858009338378906,
      "step": 57740
    },
    {
      "epoch": 0.001176,
      "grad_norm": 0.06826510280370712,
      "learning_rate": 0.0008663850000000001,
      "loss": 1.794895553588867,
      "step": 57760
    },
    {
      "epoch": 0.001178,
      "grad_norm": 0.09383084625005722,
      "learning_rate": 0.0008666850000000001,
      "loss": 1.7879268646240234,
      "step": 57780
    },
    {
      "epoch": 0.00118,
      "grad_norm": 0.09734877943992615,
      "learning_rate": 0.000866985,
      "loss": 1.7864860534667968,
      "step": 57800
    },
    {
      "epoch": 0.001182,
      "grad_norm": 0.07767576724290848,
      "learning_rate": 0.000867285,
      "loss": 1.799258041381836,
      "step": 57820
    },
    {
      "epoch": 0.001184,
      "grad_norm": 0.06870760768651962,
      "learning_rate": 0.0008675849999999999,
      "loss": 1.7924396514892578,
      "step": 57840
    },
    {
      "epoch": 0.001186,
      "grad_norm": 0.07649760693311691,
      "learning_rate": 0.0008678850000000001,
      "loss": 1.8006465911865235,
      "step": 57860
    },
    {
      "epoch": 0.001188,
      "grad_norm": 0.0633130595088005,
      "learning_rate": 0.000868185,
      "loss": 1.8013498306274414,
      "step": 57880
    },
    {
      "epoch": 0.00119,
      "grad_norm": 0.06914418190717697,
      "learning_rate": 0.000868485,
      "loss": 1.8069189071655274,
      "step": 57900
    },
    {
      "epoch": 0.001192,
      "grad_norm": 0.0731736272573471,
      "learning_rate": 0.000868785,
      "loss": 1.7978586196899413,
      "step": 57920
    },
    {
      "epoch": 0.001194,
      "grad_norm": 0.09937088936567307,
      "learning_rate": 0.0008690849999999999,
      "loss": 1.7953372955322267,
      "step": 57940
    },
    {
      "epoch": 0.001196,
      "grad_norm": 0.07816176116466522,
      "learning_rate": 0.0008693850000000001,
      "loss": 1.8023591995239259,
      "step": 57960
    },
    {
      "epoch": 0.001198,
      "grad_norm": 0.07683317363262177,
      "learning_rate": 0.000869685,
      "loss": 1.7934598922729492,
      "step": 57980
    },
    {
      "epoch": 0.0012,
      "grad_norm": 0.09890981763601303,
      "learning_rate": 0.000869985,
      "loss": 1.8024948120117188,
      "step": 58000
    },
    {
      "epoch": 0.0012,
      "eval_loss": 1.7639732360839844,
      "eval_runtime": 101.1214,
      "eval_samples_per_second": 9.889,
      "eval_steps_per_second": 9.889,
      "step": 58000
    },
    {
      "epoch": 0.001202,
      "grad_norm": 0.08846702426671982,
      "learning_rate": 0.000870285,
      "loss": 1.7955366134643556,
      "step": 58020
    },
    {
      "epoch": 0.001204,
      "grad_norm": 0.07499402016401291,
      "learning_rate": 0.000870585,
      "loss": 1.7973455429077148,
      "step": 58040
    },
    {
      "epoch": 0.001206,
      "grad_norm": 0.10481803119182587,
      "learning_rate": 0.0008708850000000001,
      "loss": 1.793146514892578,
      "step": 58060
    },
    {
      "epoch": 0.001208,
      "grad_norm": 0.064764603972435,
      "learning_rate": 0.000871185,
      "loss": 1.8076301574707032,
      "step": 58080
    },
    {
      "epoch": 0.00121,
      "grad_norm": 0.08116547018289566,
      "learning_rate": 0.0008714850000000001,
      "loss": 1.798056411743164,
      "step": 58100
    },
    {
      "epoch": 0.001212,
      "grad_norm": 0.10666067898273468,
      "learning_rate": 0.000871785,
      "loss": 1.7961101531982422,
      "step": 58120
    },
    {
      "epoch": 0.001214,
      "grad_norm": 0.15765641629695892,
      "learning_rate": 0.000872085,
      "loss": 1.7949405670166017,
      "step": 58140
    },
    {
      "epoch": 0.001216,
      "grad_norm": 0.11962568014860153,
      "learning_rate": 0.000872385,
      "loss": 1.802354621887207,
      "step": 58160
    },
    {
      "epoch": 0.001218,
      "grad_norm": 0.0923636183142662,
      "learning_rate": 0.0008726850000000001,
      "loss": 1.8007795333862304,
      "step": 58180
    },
    {
      "epoch": 0.00122,
      "grad_norm": 0.10511373728513718,
      "learning_rate": 0.0008729850000000001,
      "loss": 1.7984931945800782,
      "step": 58200
    },
    {
      "epoch": 0.001222,
      "grad_norm": 0.10952931642532349,
      "learning_rate": 0.000873285,
      "loss": 1.7882881164550781,
      "step": 58220
    },
    {
      "epoch": 0.001224,
      "grad_norm": 0.07120373845100403,
      "learning_rate": 0.000873585,
      "loss": 1.7912660598754884,
      "step": 58240
    },
    {
      "epoch": 0.001226,
      "grad_norm": 0.13143955171108246,
      "learning_rate": 0.0008738850000000001,
      "loss": 1.7980297088623047,
      "step": 58260
    },
    {
      "epoch": 0.001228,
      "grad_norm": 0.07388322055339813,
      "learning_rate": 0.0008741850000000001,
      "loss": 1.7870193481445313,
      "step": 58280
    },
    {
      "epoch": 0.00123,
      "grad_norm": 0.08142983913421631,
      "learning_rate": 0.000874485,
      "loss": 1.7930824279785156,
      "step": 58300
    },
    {
      "epoch": 0.001232,
      "grad_norm": 0.08841890096664429,
      "learning_rate": 0.000874785,
      "loss": 1.7895841598510742,
      "step": 58320
    },
    {
      "epoch": 0.001234,
      "grad_norm": 0.0918579250574112,
      "learning_rate": 0.000875085,
      "loss": 1.7884759902954102,
      "step": 58340
    },
    {
      "epoch": 0.001236,
      "grad_norm": 0.09915296733379364,
      "learning_rate": 0.0008753850000000001,
      "loss": 1.7873065948486329,
      "step": 58360
    },
    {
      "epoch": 0.001238,
      "grad_norm": 0.06550557166337967,
      "learning_rate": 0.0008756850000000001,
      "loss": 1.788595962524414,
      "step": 58380
    },
    {
      "epoch": 0.00124,
      "grad_norm": 0.07651044428348541,
      "learning_rate": 0.000875985,
      "loss": 1.791623306274414,
      "step": 58400
    },
    {
      "epoch": 0.001242,
      "grad_norm": 0.1300506591796875,
      "learning_rate": 0.000876285,
      "loss": 1.7822643280029298,
      "step": 58420
    },
    {
      "epoch": 0.001244,
      "grad_norm": 0.0868898332118988,
      "learning_rate": 0.0008765849999999999,
      "loss": 1.7846914291381837,
      "step": 58440
    },
    {
      "epoch": 0.001246,
      "grad_norm": 0.08382240682840347,
      "learning_rate": 0.0008768850000000001,
      "loss": 1.791547966003418,
      "step": 58460
    },
    {
      "epoch": 0.001248,
      "grad_norm": 0.08530152589082718,
      "learning_rate": 0.000877185,
      "loss": 1.7961118698120118,
      "step": 58480
    },
    {
      "epoch": 0.00125,
      "grad_norm": 0.09241317957639694,
      "learning_rate": 0.000877485,
      "loss": 1.7916952133178712,
      "step": 58500
    },
    {
      "epoch": 0.001252,
      "grad_norm": 0.08250889182090759,
      "learning_rate": 0.000877785,
      "loss": 1.786859130859375,
      "step": 58520
    },
    {
      "epoch": 0.001254,
      "grad_norm": 0.06770889461040497,
      "learning_rate": 0.000878085,
      "loss": 1.7918638229370116,
      "step": 58540
    },
    {
      "epoch": 0.001256,
      "grad_norm": 0.06325265765190125,
      "learning_rate": 0.0008783850000000001,
      "loss": 1.7936885833740235,
      "step": 58560
    },
    {
      "epoch": 0.001258,
      "grad_norm": 0.06696333736181259,
      "learning_rate": 0.000878685,
      "loss": 1.7841094970703124,
      "step": 58580
    },
    {
      "epoch": 0.00126,
      "grad_norm": 0.09796358644962311,
      "learning_rate": 0.000878985,
      "loss": 1.7848304748535155,
      "step": 58600
    },
    {
      "epoch": 0.001262,
      "grad_norm": 0.09959303587675095,
      "learning_rate": 0.000879285,
      "loss": 1.7836418151855469,
      "step": 58620
    },
    {
      "epoch": 0.001264,
      "grad_norm": 0.09237895905971527,
      "learning_rate": 0.000879585,
      "loss": 1.783563232421875,
      "step": 58640
    },
    {
      "epoch": 0.001266,
      "grad_norm": 0.07377872616052628,
      "learning_rate": 0.000879885,
      "loss": 1.7867752075195313,
      "step": 58660
    },
    {
      "epoch": 0.001268,
      "grad_norm": 0.08524210751056671,
      "learning_rate": 0.000880185,
      "loss": 1.7875303268432616,
      "step": 58680
    },
    {
      "epoch": 0.00127,
      "grad_norm": 0.08337651193141937,
      "learning_rate": 0.0008804850000000001,
      "loss": 1.787185287475586,
      "step": 58700
    },
    {
      "epoch": 0.001272,
      "grad_norm": 0.09430639445781708,
      "learning_rate": 0.000880785,
      "loss": 1.7876216888427734,
      "step": 58720
    },
    {
      "epoch": 0.001274,
      "grad_norm": 0.07649505138397217,
      "learning_rate": 0.000881085,
      "loss": 1.790460205078125,
      "step": 58740
    },
    {
      "epoch": 0.001276,
      "grad_norm": 0.09520697593688965,
      "learning_rate": 0.0008813849999999999,
      "loss": 1.7819580078125,
      "step": 58760
    },
    {
      "epoch": 0.001278,
      "grad_norm": 0.10098861902952194,
      "learning_rate": 0.0008816850000000001,
      "loss": 1.7848100662231445,
      "step": 58780
    },
    {
      "epoch": 0.00128,
      "grad_norm": 0.13069051504135132,
      "learning_rate": 0.0008819850000000001,
      "loss": 1.781412696838379,
      "step": 58800
    },
    {
      "epoch": 0.001282,
      "grad_norm": 0.0984887182712555,
      "learning_rate": 0.000882285,
      "loss": 1.7869014739990234,
      "step": 58820
    },
    {
      "epoch": 0.001284,
      "grad_norm": 0.098292276263237,
      "learning_rate": 0.000882585,
      "loss": 1.7787504196166992,
      "step": 58840
    },
    {
      "epoch": 0.001286,
      "grad_norm": 0.14762945473194122,
      "learning_rate": 0.0008828849999999999,
      "loss": 1.785466766357422,
      "step": 58860
    },
    {
      "epoch": 0.001288,
      "grad_norm": 0.07387299090623856,
      "learning_rate": 0.0008831850000000001,
      "loss": 1.7865690231323241,
      "step": 58880
    },
    {
      "epoch": 0.00129,
      "grad_norm": 0.09571004658937454,
      "learning_rate": 0.000883485,
      "loss": 1.7807199478149414,
      "step": 58900
    },
    {
      "epoch": 0.001292,
      "grad_norm": 0.09966664761304855,
      "learning_rate": 0.000883785,
      "loss": 1.7810815811157226,
      "step": 58920
    },
    {
      "epoch": 0.001294,
      "grad_norm": 0.061778098344802856,
      "learning_rate": 0.000884085,
      "loss": 1.7799869537353517,
      "step": 58940
    },
    {
      "epoch": 0.001296,
      "grad_norm": 0.09198242425918579,
      "learning_rate": 0.0008843849999999999,
      "loss": 1.7813076019287108,
      "step": 58960
    },
    {
      "epoch": 0.001298,
      "grad_norm": 0.0708242729306221,
      "learning_rate": 0.0008846850000000001,
      "loss": 1.7849723815917968,
      "step": 58980
    },
    {
      "epoch": 0.0013,
      "grad_norm": 0.0876058042049408,
      "learning_rate": 0.000884985,
      "loss": 1.7796497344970703,
      "step": 59000
    },
    {
      "epoch": 0.0013,
      "eval_loss": 1.7605947256088257,
      "eval_runtime": 102.4761,
      "eval_samples_per_second": 9.758,
      "eval_steps_per_second": 9.758,
      "step": 59000
    },
    {
      "epoch": 0.001302,
      "grad_norm": 0.11163485050201416,
      "learning_rate": 0.000885285,
      "loss": 1.7904579162597656,
      "step": 59020
    },
    {
      "epoch": 0.001304,
      "grad_norm": 0.07118188589811325,
      "learning_rate": 0.0008855849999999999,
      "loss": 1.7947458267211913,
      "step": 59040
    },
    {
      "epoch": 0.001306,
      "grad_norm": 0.11685152351856232,
      "learning_rate": 0.000885885,
      "loss": 1.800405502319336,
      "step": 59060
    },
    {
      "epoch": 0.001308,
      "grad_norm": 0.08772759139537811,
      "learning_rate": 0.0008861850000000001,
      "loss": 1.8008148193359375,
      "step": 59080
    },
    {
      "epoch": 0.00131,
      "grad_norm": 0.08764131367206573,
      "learning_rate": 0.000886485,
      "loss": 1.7950908660888671,
      "step": 59100
    },
    {
      "epoch": 0.001312,
      "grad_norm": 0.09448276460170746,
      "learning_rate": 0.000886785,
      "loss": 1.7981008529663085,
      "step": 59120
    },
    {
      "epoch": 0.001314,
      "grad_norm": 0.08192802965641022,
      "learning_rate": 0.000887085,
      "loss": 1.7959772109985352,
      "step": 59140
    },
    {
      "epoch": 0.001316,
      "grad_norm": 0.13744975626468658,
      "learning_rate": 0.000887385,
      "loss": 1.802775764465332,
      "step": 59160
    },
    {
      "epoch": 0.001318,
      "grad_norm": 0.07230217009782791,
      "learning_rate": 0.000887685,
      "loss": 1.7934690475463868,
      "step": 59180
    },
    {
      "epoch": 0.00132,
      "grad_norm": 0.08101034164428711,
      "learning_rate": 0.000887985,
      "loss": 1.7986469268798828,
      "step": 59200
    },
    {
      "epoch": 0.001322,
      "grad_norm": 0.07884607464075089,
      "learning_rate": 0.000888285,
      "loss": 1.7923391342163086,
      "step": 59220
    },
    {
      "epoch": 0.001324,
      "grad_norm": 0.09071575105190277,
      "learning_rate": 0.000888585,
      "loss": 1.7922843933105468,
      "step": 59240
    },
    {
      "epoch": 0.001326,
      "grad_norm": 0.10291945189237595,
      "learning_rate": 0.000888885,
      "loss": 1.7996133804321288,
      "step": 59260
    },
    {
      "epoch": 0.001328,
      "grad_norm": 0.10892489552497864,
      "learning_rate": 0.0008891850000000001,
      "loss": 1.7933570861816406,
      "step": 59280
    },
    {
      "epoch": 0.00133,
      "grad_norm": 0.07337154448032379,
      "learning_rate": 0.0008894850000000001,
      "loss": 1.7932397842407226,
      "step": 59300
    },
    {
      "epoch": 0.001332,
      "grad_norm": 0.09053796529769897,
      "learning_rate": 0.000889785,
      "loss": 1.7875089645385742,
      "step": 59320
    },
    {
      "epoch": 0.001334,
      "grad_norm": 0.11005237698554993,
      "learning_rate": 0.000890085,
      "loss": 1.7968561172485351,
      "step": 59340
    },
    {
      "epoch": 0.001336,
      "grad_norm": 0.08401455730199814,
      "learning_rate": 0.0008903849999999999,
      "loss": 1.7960100173950195,
      "step": 59360
    },
    {
      "epoch": 0.001338,
      "grad_norm": 0.08846209198236465,
      "learning_rate": 0.0008906850000000001,
      "loss": 1.7931608200073241,
      "step": 59380
    },
    {
      "epoch": 0.00134,
      "grad_norm": 0.06116236373782158,
      "learning_rate": 0.0008909850000000001,
      "loss": 1.790717887878418,
      "step": 59400
    },
    {
      "epoch": 0.001342,
      "grad_norm": 0.07480533421039581,
      "learning_rate": 0.000891285,
      "loss": 1.7880945205688477,
      "step": 59420
    },
    {
      "epoch": 0.001344,
      "grad_norm": 0.08431543409824371,
      "learning_rate": 0.000891585,
      "loss": 1.7937543869018555,
      "step": 59440
    },
    {
      "epoch": 0.001346,
      "grad_norm": 0.09316226094961166,
      "learning_rate": 0.0008918849999999999,
      "loss": 1.7945869445800782,
      "step": 59460
    },
    {
      "epoch": 0.001348,
      "grad_norm": 0.12234386801719666,
      "learning_rate": 0.0008921850000000001,
      "loss": 1.7928190231323242,
      "step": 59480
    },
    {
      "epoch": 0.00135,
      "grad_norm": 0.1400688886642456,
      "learning_rate": 0.000892485,
      "loss": 1.793172264099121,
      "step": 59500
    },
    {
      "epoch": 0.001352,
      "grad_norm": 0.12014756351709366,
      "learning_rate": 0.000892785,
      "loss": 1.7820804595947266,
      "step": 59520
    },
    {
      "epoch": 0.001354,
      "grad_norm": 0.07136882096529007,
      "learning_rate": 0.000893085,
      "loss": 1.7874395370483398,
      "step": 59540
    },
    {
      "epoch": 0.001356,
      "grad_norm": 0.10591576993465424,
      "learning_rate": 0.0008933849999999999,
      "loss": 1.7914464950561524,
      "step": 59560
    },
    {
      "epoch": 0.001358,
      "grad_norm": 0.13441281020641327,
      "learning_rate": 0.0008936850000000001,
      "loss": 1.787251853942871,
      "step": 59580
    },
    {
      "epoch": 0.00136,
      "grad_norm": 0.08462520688772202,
      "learning_rate": 0.000893985,
      "loss": 1.7891172409057616,
      "step": 59600
    },
    {
      "epoch": 0.001362,
      "grad_norm": 0.1170402392745018,
      "learning_rate": 0.000894285,
      "loss": 1.7866937637329101,
      "step": 59620
    },
    {
      "epoch": 0.001364,
      "grad_norm": 0.13150587677955627,
      "learning_rate": 0.0008945849999999999,
      "loss": 1.786465835571289,
      "step": 59640
    },
    {
      "epoch": 0.001366,
      "grad_norm": 0.10751062631607056,
      "learning_rate": 0.000894885,
      "loss": 1.7842748641967774,
      "step": 59660
    },
    {
      "epoch": 0.001368,
      "grad_norm": 0.15118949115276337,
      "learning_rate": 0.0008951850000000001,
      "loss": 1.792129135131836,
      "step": 59680
    },
    {
      "epoch": 0.00137,
      "grad_norm": 0.11457185447216034,
      "learning_rate": 0.000895485,
      "loss": 1.789076042175293,
      "step": 59700
    },
    {
      "epoch": 0.001372,
      "grad_norm": 0.09175355732440948,
      "learning_rate": 0.000895785,
      "loss": 1.7840227127075194,
      "step": 59720
    },
    {
      "epoch": 0.001374,
      "grad_norm": 0.0791926383972168,
      "learning_rate": 0.000896085,
      "loss": 1.7862699508666993,
      "step": 59740
    },
    {
      "epoch": 0.001376,
      "grad_norm": 0.07817155122756958,
      "learning_rate": 0.000896385,
      "loss": 1.7845630645751953,
      "step": 59760
    },
    {
      "epoch": 0.001378,
      "grad_norm": 0.07499396055936813,
      "learning_rate": 0.000896685,
      "loss": 1.7851142883300781,
      "step": 59780
    },
    {
      "epoch": 0.00138,
      "grad_norm": 0.09369134157896042,
      "learning_rate": 0.0008969850000000001,
      "loss": 1.777730369567871,
      "step": 59800
    },
    {
      "epoch": 0.001382,
      "grad_norm": 0.07674891501665115,
      "learning_rate": 0.000897285,
      "loss": 1.7841289520263672,
      "step": 59820
    },
    {
      "epoch": 0.001384,
      "grad_norm": 0.10798847675323486,
      "learning_rate": 0.000897585,
      "loss": 1.7842435836791992,
      "step": 59840
    },
    {
      "epoch": 0.001386,
      "grad_norm": 0.12474353611469269,
      "learning_rate": 0.000897885,
      "loss": 1.7782463073730468,
      "step": 59860
    },
    {
      "epoch": 0.001388,
      "grad_norm": 0.07487408816814423,
      "learning_rate": 0.0008981850000000001,
      "loss": 1.788216781616211,
      "step": 59880
    },
    {
      "epoch": 0.00139,
      "grad_norm": 0.08795047551393509,
      "learning_rate": 0.0008984850000000001,
      "loss": 1.788589096069336,
      "step": 59900
    },
    {
      "epoch": 0.001392,
      "grad_norm": 0.06374309211969376,
      "learning_rate": 0.000898785,
      "loss": 1.7882522583007812,
      "step": 59920
    },
    {
      "epoch": 0.001394,
      "grad_norm": 0.08231166005134583,
      "learning_rate": 0.000899085,
      "loss": 1.7771085739135741,
      "step": 59940
    },
    {
      "epoch": 0.001396,
      "grad_norm": 0.09305830299854279,
      "learning_rate": 0.0008993849999999999,
      "loss": 1.7856315612792968,
      "step": 59960
    },
    {
      "epoch": 0.001398,
      "grad_norm": 0.07728376984596252,
      "learning_rate": 0.0008996850000000001,
      "loss": 1.7800445556640625,
      "step": 59980
    },
    {
      "epoch": 0.0014,
      "grad_norm": 0.14031851291656494,
      "learning_rate": 0.0008999850000000001,
      "loss": 1.7799678802490235,
      "step": 60000
    },
    {
      "epoch": 0.0014,
      "eval_loss": 1.757127285003662,
      "eval_runtime": 101.7258,
      "eval_samples_per_second": 9.83,
      "eval_steps_per_second": 9.83,
      "step": 60000
    },
    {
      "epoch": 0.001402,
      "grad_norm": 0.07864392548799515,
      "learning_rate": 0.000900285,
      "loss": 1.780336570739746,
      "step": 60020
    },
    {
      "epoch": 0.001404,
      "grad_norm": 0.10762085765600204,
      "learning_rate": 0.000900585,
      "loss": 1.774503517150879,
      "step": 60040
    },
    {
      "epoch": 0.001406,
      "grad_norm": 0.08554880321025848,
      "learning_rate": 0.0009008849999999999,
      "loss": 1.7854381561279298,
      "step": 60060
    },
    {
      "epoch": 0.001408,
      "grad_norm": 0.12507252395153046,
      "learning_rate": 0.0009011850000000001,
      "loss": 1.7746183395385742,
      "step": 60080
    },
    {
      "epoch": 0.00141,
      "grad_norm": 0.0879814401268959,
      "learning_rate": 0.000901485,
      "loss": 1.7791704177856444,
      "step": 60100
    },
    {
      "epoch": 0.001412,
      "grad_norm": 0.09369395673274994,
      "learning_rate": 0.000901785,
      "loss": 1.7811712265014648,
      "step": 60120
    },
    {
      "epoch": 0.001414,
      "grad_norm": 0.12184352427721024,
      "learning_rate": 0.000902085,
      "loss": 1.7823884963989258,
      "step": 60140
    },
    {
      "epoch": 0.001416,
      "grad_norm": 0.0922425165772438,
      "learning_rate": 0.0009023849999999999,
      "loss": 1.784942054748535,
      "step": 60160
    },
    {
      "epoch": 0.001418,
      "grad_norm": 0.11703264713287354,
      "learning_rate": 0.0009026850000000001,
      "loss": 1.7911079406738282,
      "step": 60180
    },
    {
      "epoch": 0.00142,
      "grad_norm": 0.06786137819290161,
      "learning_rate": 0.000902985,
      "loss": 1.7841667175292968,
      "step": 60200
    },
    {
      "epoch": 0.001422,
      "grad_norm": 0.12831048667430878,
      "learning_rate": 0.000903285,
      "loss": 1.7822101593017579,
      "step": 60220
    },
    {
      "epoch": 0.001424,
      "grad_norm": 0.06657948344945908,
      "learning_rate": 0.000903585,
      "loss": 1.7872278213500976,
      "step": 60240
    },
    {
      "epoch": 0.001426,
      "grad_norm": 0.11561829596757889,
      "learning_rate": 0.000903885,
      "loss": 1.7912164688110352,
      "step": 60260
    },
    {
      "epoch": 0.001428,
      "grad_norm": 0.06207296624779701,
      "learning_rate": 0.0009041850000000001,
      "loss": 1.7937608718872071,
      "step": 60280
    },
    {
      "epoch": 0.00143,
      "grad_norm": 0.12583515048027039,
      "learning_rate": 0.000904485,
      "loss": 1.7964292526245118,
      "step": 60300
    },
    {
      "epoch": 0.001432,
      "grad_norm": 0.08955791592597961,
      "learning_rate": 0.0009047850000000001,
      "loss": 1.793429183959961,
      "step": 60320
    },
    {
      "epoch": 0.001434,
      "grad_norm": 0.15952014923095703,
      "learning_rate": 0.000905085,
      "loss": 1.7835594177246095,
      "step": 60340
    },
    {
      "epoch": 0.001436,
      "grad_norm": 0.13086777925491333,
      "learning_rate": 0.000905385,
      "loss": 1.789198112487793,
      "step": 60360
    },
    {
      "epoch": 0.001438,
      "grad_norm": 0.11018704622983932,
      "learning_rate": 0.000905685,
      "loss": 1.7987808227539062,
      "step": 60380
    },
    {
      "epoch": 0.00144,
      "grad_norm": 0.12111669778823853,
      "learning_rate": 0.0009059850000000001,
      "loss": 1.785379409790039,
      "step": 60400
    },
    {
      "epoch": 0.001442,
      "grad_norm": 0.10074923187494278,
      "learning_rate": 0.0009062850000000001,
      "loss": 1.7914039611816406,
      "step": 60420
    },
    {
      "epoch": 0.001444,
      "grad_norm": 0.12976816296577454,
      "learning_rate": 0.000906585,
      "loss": 1.782525634765625,
      "step": 60440
    },
    {
      "epoch": 0.001446,
      "grad_norm": 0.11517872661352158,
      "learning_rate": 0.000906885,
      "loss": 1.7910505294799806,
      "step": 60460
    },
    {
      "epoch": 0.001448,
      "grad_norm": 0.07611226290464401,
      "learning_rate": 0.0009071850000000001,
      "loss": 1.7883560180664062,
      "step": 60480
    },
    {
      "epoch": 0.00145,
      "grad_norm": 0.09393489360809326,
      "learning_rate": 0.0009074850000000001,
      "loss": 1.7844127655029296,
      "step": 60500
    },
    {
      "epoch": 0.001452,
      "grad_norm": 0.06728903949260712,
      "learning_rate": 0.000907785,
      "loss": 1.7861913681030273,
      "step": 60520
    },
    {
      "epoch": 0.001454,
      "grad_norm": 0.08470039069652557,
      "learning_rate": 0.000908085,
      "loss": 1.7802160263061524,
      "step": 60540
    },
    {
      "epoch": 0.001456,
      "grad_norm": 0.09300052374601364,
      "learning_rate": 0.0009083849999999999,
      "loss": 1.7824180603027344,
      "step": 60560
    },
    {
      "epoch": 0.001458,
      "grad_norm": 0.07797509431838989,
      "learning_rate": 0.0009086850000000001,
      "loss": 1.7897750854492187,
      "step": 60580
    },
    {
      "epoch": 0.00146,
      "grad_norm": 0.07057961076498032,
      "learning_rate": 0.0009089850000000001,
      "loss": 1.7705690383911132,
      "step": 60600
    },
    {
      "epoch": 0.001462,
      "grad_norm": 0.09517934918403625,
      "learning_rate": 0.000909285,
      "loss": 1.7854423522949219,
      "step": 60620
    },
    {
      "epoch": 0.001464,
      "grad_norm": 0.08621353656053543,
      "learning_rate": 0.000909585,
      "loss": 1.782230758666992,
      "step": 60640
    },
    {
      "epoch": 0.001466,
      "grad_norm": 0.1416175216436386,
      "learning_rate": 0.0009098849999999999,
      "loss": 1.7808948516845704,
      "step": 60660
    },
    {
      "epoch": 0.001468,
      "grad_norm": 0.08531250059604645,
      "learning_rate": 0.0009101850000000001,
      "loss": 1.7843456268310547,
      "step": 60680
    },
    {
      "epoch": 0.00147,
      "grad_norm": 0.0890691876411438,
      "learning_rate": 0.000910485,
      "loss": 1.775352668762207,
      "step": 60700
    },
    {
      "epoch": 0.001472,
      "grad_norm": 0.08158216625452042,
      "learning_rate": 0.000910785,
      "loss": 1.7796384811401367,
      "step": 60720
    },
    {
      "epoch": 0.001474,
      "grad_norm": 0.07983843237161636,
      "learning_rate": 0.000911085,
      "loss": 1.778581428527832,
      "step": 60740
    },
    {
      "epoch": 0.001476,
      "grad_norm": 0.07468003034591675,
      "learning_rate": 0.000911385,
      "loss": 1.7876134872436524,
      "step": 60760
    },
    {
      "epoch": 0.001478,
      "grad_norm": 0.0966465175151825,
      "learning_rate": 0.0009116850000000001,
      "loss": 1.7833978652954101,
      "step": 60780
    },
    {
      "epoch": 0.00148,
      "grad_norm": 0.102341428399086,
      "learning_rate": 0.000911985,
      "loss": 1.7740676879882813,
      "step": 60800
    },
    {
      "epoch": 0.001482,
      "grad_norm": 0.07008518278598785,
      "learning_rate": 0.000912285,
      "loss": 1.7881011962890625,
      "step": 60820
    },
    {
      "epoch": 0.001484,
      "grad_norm": 0.07029026001691818,
      "learning_rate": 0.000912585,
      "loss": 1.7808820724487304,
      "step": 60840
    },
    {
      "epoch": 0.001486,
      "grad_norm": 0.06320752203464508,
      "learning_rate": 0.000912885,
      "loss": 1.7813623428344727,
      "step": 60860
    },
    {
      "epoch": 0.001488,
      "grad_norm": 0.08225560933351517,
      "learning_rate": 0.0009131850000000001,
      "loss": 1.7788610458374023,
      "step": 60880
    },
    {
      "epoch": 0.00149,
      "grad_norm": 0.14620341360569,
      "learning_rate": 0.000913485,
      "loss": 1.7807167053222657,
      "step": 60900
    },
    {
      "epoch": 0.001492,
      "grad_norm": 0.091734878718853,
      "learning_rate": 0.0009137850000000001,
      "loss": 1.7744346618652345,
      "step": 60920
    },
    {
      "epoch": 0.001494,
      "grad_norm": 0.11475755274295807,
      "learning_rate": 0.000914085,
      "loss": 1.7763957977294922,
      "step": 60940
    },
    {
      "epoch": 0.001496,
      "grad_norm": 0.08693142980337143,
      "learning_rate": 0.000914385,
      "loss": 1.7710250854492187,
      "step": 60960
    },
    {
      "epoch": 0.001498,
      "grad_norm": 0.08143222332000732,
      "learning_rate": 0.0009146850000000002,
      "loss": 1.776578140258789,
      "step": 60980
    },
    {
      "epoch": 0.0015,
      "grad_norm": 0.08184456825256348,
      "learning_rate": 0.0009149850000000001,
      "loss": 1.7720607757568358,
      "step": 61000
    },
    {
      "epoch": 0.0015,
      "eval_loss": 1.7527116537094116,
      "eval_runtime": 96.9294,
      "eval_samples_per_second": 10.317,
      "eval_steps_per_second": 10.317,
      "step": 61000
    },
    {
      "epoch": 0.00013167943957230517,
      "grad_norm": 0.15917663276195526,
      "learning_rate": 0.0019556872017675816,
      "loss": 1.7689592361450195,
      "step": 61020,
      "train_bpb": 0.624479159802406,
      "train_bytes": 57285.4875,
      "train_loss_nats_per_token": 1.769419271408239
    },
    {
      "epoch": 0.00026335887914461035,
      "grad_norm": 0.19207613170146942,
      "learning_rate": 0.0019550959649518944,
      "loss": 1.7878398895263672,
      "step": 61040,
      "train_bpb": 0.6332869627067964,
      "train_bytes": 56225.5125,
      "train_loss_nats_per_token": 1.7878394894029352
    },
    {
      "epoch": 0.00039503831871691555,
      "grad_norm": 0.13957324624061584,
      "learning_rate": 0.0019545046502539264,
      "loss": 1.779135513305664,
      "step": 61060,
      "train_bpb": 0.6284148166299693,
      "train_bytes": 56917.51875,
      "train_loss_nats_per_token": 1.7795955909258072
    },
    {
      "epoch": 0.0005267177582892207,
      "grad_norm": 0.19360390305519104,
      "learning_rate": 0.001953913257774871,
      "loss": 1.7846166610717773,
      "step": 61080,
      "train_bpb": 0.6320205535939892,
      "train_bytes": 56260.525,
      "train_loss_nats_per_token": 1.784514928130163
    },
    {
      "epoch": 0.0006583971978615259,
      "grad_norm": 0.1390513926744461,
      "learning_rate": 0.0019533217876159356,
      "loss": 1.791531753540039,
      "step": 61100,
      "train_bpb": 0.6323405783127461,
      "train_bytes": 56795.56875,
      "train_loss_nats_per_token": 1.7918770866024658
    },
    {
      "epoch": 0.0007900766374338311,
      "grad_norm": 0.11623669415712357,
      "learning_rate": 0.0019527302398783402,
      "loss": 1.786199951171875,
      "step": 61120,
      "train_bpb": 0.629222283258163,
      "train_bytes": 56929.99375,
      "train_loss_nats_per_token": 1.7862105658113954
    },
    {
      "epoch": 0.0009217560770061363,
      "grad_norm": 0.12392310798168182,
      "learning_rate": 0.0019521386146633202,
      "loss": 1.779733657836914,
      "step": 61140,
      "train_bpb": 0.6278814477540333,
      "train_bytes": 56964.9875,
      "train_loss_nats_per_token": 1.7801071181111832
    },
    {
      "epoch": 0.0010534355165784414,
      "grad_norm": 0.14329653978347778,
      "learning_rate": 0.0019515469120721205,
      "loss": 1.7768293380737306,
      "step": 61160,
      "train_bpb": 0.6293027708173399,
      "train_bytes": 56336.2625,
      "train_loss_nats_per_token": 1.7768973502036727
    },
    {
      "epoch": 0.0011851149561507466,
      "grad_norm": 0.14658844470977783,
      "learning_rate": 0.0019509551322060036,
      "loss": 1.781470489501953,
      "step": 61180,
      "train_bpb": 0.6308210419779954,
      "train_bytes": 56659.75,
      "train_loss_nats_per_token": 1.781694116903357
    },
    {
      "epoch": 0.0013167943957230518,
      "grad_norm": 0.1357489973306656,
      "learning_rate": 0.0019503632751662406,
      "loss": 1.784101676940918,
      "step": 61200,
      "train_bpb": 0.6286416572116313,
      "train_bytes": 56591.2375,
      "train_loss_nats_per_token": 1.783926969335869
    },
    {
      "epoch": 0.001448473835295357,
      "grad_norm": 0.14259649813175201,
      "learning_rate": 0.0019497713410541204,
      "loss": 1.7761362075805665,
      "step": 61220,
      "train_bpb": 0.6287786742181933,
      "train_bytes": 56428.475,
      "train_loss_nats_per_token": 1.776385602011495
    },
    {
      "epoch": 0.0015801532748676622,
      "grad_norm": 0.17029036581516266,
      "learning_rate": 0.0019491793299709415,
      "loss": 1.7787519454956056,
      "step": 61240,
      "train_bpb": 0.6290644850335212,
      "train_bytes": 56637.1375,
      "train_loss_nats_per_token": 1.7787108882641405
    },
    {
      "epoch": 0.0017118327144399674,
      "grad_norm": 0.11382069438695908,
      "learning_rate": 0.0019485872420180175,
      "loss": 1.7769920349121093,
      "step": 61260,
      "train_bpb": 0.6269067905517248,
      "train_bytes": 56402.74375,
      "train_loss_nats_per_token": 1.777183683296432
    },
    {
      "epoch": 0.0018435121540122726,
      "grad_norm": 0.12319232523441315,
      "learning_rate": 0.0019479950772966746,
      "loss": 1.7725486755371094,
      "step": 61280,
      "train_bpb": 0.6284571601177934,
      "train_bytes": 56227.9125,
      "train_loss_nats_per_token": 1.772665407526244
    },
    {
      "epoch": 0.0019751915935845776,
      "grad_norm": 0.17766621708869934,
      "learning_rate": 0.0019474028359082523,
      "loss": 1.7788896560668945,
      "step": 61300,
      "train_bpb": 0.6282757738097056,
      "train_bytes": 56507.2,
      "train_loss_nats_per_token": 1.7795000779698864
    },
    {
      "epoch": 0.002106871033156883,
      "grad_norm": 0.22157756984233856,
      "learning_rate": 0.0019468105179541028,
      "loss": 1.7859193801879882,
      "step": 61320,
      "train_bpb": 0.6335497836523014,
      "train_bytes": 56519.20625,
      "train_loss_nats_per_token": 1.7861596322588515
    },
    {
      "epoch": 0.002238550472729188,
      "grad_norm": 0.12757186591625214,
      "learning_rate": 0.001946218123535592,
      "loss": 1.7696149826049805,
      "step": 61340,
      "train_bpb": 0.6261421234611524,
      "train_bytes": 56722.08125,
      "train_loss_nats_per_token": 1.769633024052579
    },
    {
      "epoch": 0.002370229912301493,
      "grad_norm": 0.18443775177001953,
      "learning_rate": 0.0019456256527540983,
      "loss": 1.7801097869873046,
      "step": 61360,
      "train_bpb": 0.6293789668013736,
      "train_bytes": 56245.75625,
      "train_loss_nats_per_token": 1.7799704810381
    },
    {
      "epoch": 0.0025019093518737984,
      "grad_norm": 0.12420649081468582,
      "learning_rate": 0.0019450331057110136,
      "loss": 1.7689334869384765,
      "step": 61380,
      "train_bpb": 0.6236885630055263,
      "train_bytes": 56647.2625,
      "train_loss_nats_per_token": 1.769181311484778
    },
    {
      "epoch": 0.0026335887914461036,
      "grad_norm": 0.10011699050664902,
      "learning_rate": 0.0019444404825077427,
      "loss": 1.7718803405761718,
      "step": 61400,
      "train_bpb": 0.625751695792086,
      "train_bytes": 56474.55,
      "train_loss_nats_per_token": 1.7719611993729882
    },
    {
      "epoch": 0.002765268231018409,
      "grad_norm": 0.18567967414855957,
      "learning_rate": 0.0019438477832457034,
      "loss": 1.768364143371582,
      "step": 61420,
      "train_bpb": 0.6258864428153574,
      "train_bytes": 56596.68125,
      "train_loss_nats_per_token": 1.7687656445781466
    },
    {
      "epoch": 0.002896947670590714,
      "grad_norm": 0.12001682817935944,
      "learning_rate": 0.0019432550080263264,
      "loss": 1.7719749450683593,
      "step": 61440,
      "train_bpb": 0.6255797476519676,
      "train_bytes": 56797.4375,
      "train_loss_nats_per_token": 1.7721568359050004
    },
    {
      "epoch": 0.003028627110163019,
      "grad_norm": 0.185967355966568,
      "learning_rate": 0.0019426621569510558,
      "loss": 1.7698919296264648,
      "step": 61460,
      "train_bpb": 0.6261617893357267,
      "train_bytes": 56435.13125,
      "train_loss_nats_per_token": 1.7706792069937654
    },
    {
      "epoch": 0.0031603065497353244,
      "grad_norm": 0.10721507668495178,
      "learning_rate": 0.0019420692301213487,
      "loss": 1.7653749465942383,
      "step": 61480,
      "train_bpb": 0.6244445340312212,
      "train_bytes": 56318.2375,
      "train_loss_nats_per_token": 1.7654478981667845
    },
    {
      "epoch": 0.0032919859893076296,
      "grad_norm": 0.12992048263549805,
      "learning_rate": 0.0019414762276386743,
      "loss": 1.7670888900756836,
      "step": 61500,
      "train_bpb": 0.624206059839131,
      "train_bytes": 56514.575,
      "train_loss_nats_per_token": 1.7675337122113983
    },
    {
      "epoch": 0.003423665428879935,
      "grad_norm": 0.1378113329410553,
      "learning_rate": 0.0019408831496045154,
      "loss": 1.773330307006836,
      "step": 61520,
      "train_bpb": 0.6270799829395985,
      "train_bytes": 56427.11875,
      "train_loss_nats_per_token": 1.7733300012285775
    },
    {
      "epoch": 0.00355534486845224,
      "grad_norm": 0.12402597069740295,
      "learning_rate": 0.0019402899961203683,
      "loss": 1.7699850082397461,
      "step": 61540,
      "train_bpb": 0.6219710775982273,
      "train_bytes": 57043.95625,
      "train_loss_nats_per_token": 1.7697111884815084
    },
    {
      "epoch": 0.0036870243080245452,
      "grad_norm": 0.14671672880649567,
      "learning_rate": 0.001939696767287741,
      "loss": 1.7713808059692382,
      "step": 61560,
      "train_bpb": 0.6249592885172314,
      "train_bytes": 56782.99375,
      "train_loss_nats_per_token": 1.7713262999885007
    },
    {
      "epoch": 0.0038187037475968504,
      "grad_norm": 0.12456189841032028,
      "learning_rate": 0.0019391034632081545,
      "loss": 1.764181137084961,
      "step": 61580,
      "train_bpb": 0.6241034138153788,
      "train_bytes": 56805.15625,
      "train_loss_nats_per_token": 1.7643190451817365
    },
    {
      "epoch": 0.003950383187169155,
      "grad_norm": 0.15086525678634644,
      "learning_rate": 0.0019385100839831443,
      "loss": 1.7586599349975587,
      "step": 61600,
      "train_bpb": 0.6227038171288081,
      "train_bytes": 56589.7875,
      "train_loss_nats_per_token": 1.759477132181976
    },
    {
      "epoch": 0.004082062626741461,
      "grad_norm": 0.16914771497249603,
      "learning_rate": 0.0019379166297142567,
      "loss": 1.7645421981811524,
      "step": 61620,
      "train_bpb": 0.62015483966336,
      "train_bytes": 56854.9125,
      "train_loss_nats_per_token": 1.7646593074728114
    },
    {
      "epoch": 0.004213742066313766,
      "grad_norm": 0.11670731008052826,
      "learning_rate": 0.0019373231005030528,
      "loss": 1.7573347091674805,
      "step": 61640,
      "train_bpb": 0.6225102203640901,
      "train_bytes": 56336.725,
      "train_loss_nats_per_token": 1.757779225033308
    },
    {
      "epoch": 0.004345421505886071,
      "grad_norm": 0.10884126275777817,
      "learning_rate": 0.0019367294964511043,
      "loss": 1.7507289886474608,
      "step": 61660,
      "train_bpb": 0.6195339281154585,
      "train_bytes": 56340.48125,
      "train_loss_nats_per_token": 1.7508099629251495
    },
    {
      "epoch": 0.004477100945458376,
      "grad_norm": 0.11791188269853592,
      "learning_rate": 0.0019361358176599975,
      "loss": 1.7582103729248046,
      "step": 61680,
      "train_bpb": 0.6205540484489418,
      "train_bytes": 56740.71875,
      "train_loss_nats_per_token": 1.7584190099908747
    },
    {
      "epoch": 0.004608780385030682,
      "grad_norm": 0.12153310328722,
      "learning_rate": 0.001935542064231331,
      "loss": 1.7683916091918945,
      "step": 61700,
      "train_bpb": 0.6241386862791074,
      "train_bytes": 56807.6375,
      "train_loss_nats_per_token": 1.7682926791779185
    },
    {
      "epoch": 0.004740459824602986,
      "grad_norm": 0.13308590650558472,
      "learning_rate": 0.0019349482362667157,
      "loss": 1.7664407730102538,
      "step": 61720,
      "train_bpb": 0.6248957458670582,
      "train_bytes": 56490.375,
      "train_loss_nats_per_token": 1.7662560395613687
    },
    {
      "epoch": 0.004872139264175292,
      "grad_norm": 0.10582863539457321,
      "learning_rate": 0.0019343543338677755,
      "loss": 1.759979248046875,
      "step": 61740,
      "train_bpb": 0.6206605576975344,
      "train_bytes": 57022.98125,
      "train_loss_nats_per_token": 1.7601138167693002
    },
    {
      "epoch": 0.005003818703747597,
      "grad_norm": 0.10867858678102493,
      "learning_rate": 0.0019337603571361476,
      "loss": 1.7591859817504882,
      "step": 61760,
      "train_bpb": 0.618893283201909,
      "train_bytes": 56616.2625,
      "train_loss_nats_per_token": 1.759045233716199
    },
    {
      "epoch": 0.0051354981433199024,
      "grad_norm": 0.14677183330059052,
      "learning_rate": 0.0019331663061734812,
      "loss": 1.7576532363891602,
      "step": 61780,
      "train_bpb": 0.6208415237457225,
      "train_bytes": 56404.51875,
      "train_loss_nats_per_token": 1.7577388332129984
    },
    {
      "epoch": 0.005267177582892207,
      "grad_norm": 0.16370470821857452,
      "learning_rate": 0.001932572181081439,
      "loss": 1.756208610534668,
      "step": 61800,
      "train_bpb": 0.6233595229684551,
      "train_bytes": 56361.63125,
      "train_loss_nats_per_token": 1.756642322315988
    },
    {
      "epoch": 0.005398857022464512,
      "grad_norm": 0.09954705089330673,
      "learning_rate": 0.0019319779819616948,
      "loss": 1.7551448822021485,
      "step": 61820,
      "train_bpb": 0.6218136553711042,
      "train_bytes": 56126.94375,
      "train_loss_nats_per_token": 1.7554528966019356
    },
    {
      "epoch": 0.005530536462036818,
      "grad_norm": 0.11029256135225296,
      "learning_rate": 0.001931383708915937,
      "loss": 1.7559398651123046,
      "step": 61840,
      "train_bpb": 0.6218979482827408,
      "train_bytes": 56335.15625,
      "train_loss_nats_per_token": 1.7561260614847012
    },
    {
      "epoch": 0.005662215901609122,
      "grad_norm": 0.14724649488925934,
      "learning_rate": 0.0019307893620458653,
      "loss": 1.7482669830322266,
      "step": 61860,
      "train_bpb": 0.617477358553122,
      "train_bytes": 56943.975,
      "train_loss_nats_per_token": 1.7486346369206247
    },
    {
      "epoch": 0.005793895341181428,
      "grad_norm": 0.1846814900636673,
      "learning_rate": 0.0019301949414531933,
      "loss": 1.7496850967407227,
      "step": 61880,
      "train_bpb": 0.6193719113692033,
      "train_bytes": 56260.18125,
      "train_loss_nats_per_token": 1.7503986545670658
    },
    {
      "epoch": 0.005925574780753733,
      "grad_norm": 0.10130450874567032,
      "learning_rate": 0.0019296004472396453,
      "loss": 1.747451400756836,
      "step": 61900,
      "train_bpb": 0.6197803302719931,
      "train_bytes": 56512.725,
      "train_loss_nats_per_token": 1.7479678488224406
    },
    {
      "epoch": 0.006057254220326038,
      "grad_norm": 0.17764247953891754,
      "learning_rate": 0.0019290058795069598,
      "loss": 1.7549179077148438,
      "step": 61920,
      "train_bpb": 0.6192572646787224,
      "train_bytes": 57069.98125,
      "train_loss_nats_per_token": 1.7551316356885147
    },
    {
      "epoch": 0.006188933659898343,
      "grad_norm": 0.10288644582033157,
      "learning_rate": 0.0019284112383568873,
      "loss": 1.7472471237182616,
      "step": 61940,
      "train_bpb": 0.6196760913387858,
      "train_bytes": 56464.46875,
      "train_loss_nats_per_token": 1.7474695581962063
    },
    {
      "epoch": 0.006320613099470649,
      "grad_norm": 0.11297981441020966,
      "learning_rate": 0.0019278165238911919,
      "loss": 1.7463382720947265,
      "step": 61960,
      "train_bpb": 0.6188157375860189,
      "train_bytes": 56099.6125,
      "train_loss_nats_per_token": 1.7464515904967113
    },
    {
      "epoch": 0.006452292539042954,
      "grad_norm": 0.16920043528079987,
      "learning_rate": 0.0019272217362116479,
      "loss": 1.7490440368652345,
      "step": 61980,
      "train_bpb": 0.619358771840242,
      "train_bytes": 56585.25625,
      "train_loss_nats_per_token": 1.749905126971789
    },
    {
      "epoch": 0.006583971978615259,
      "grad_norm": 0.16473084688186646,
      "learning_rate": 0.0019266268754200445,
      "loss": 1.7610960006713867,
      "step": 62000,
      "train_bpb": 0.6228261719429556,
      "train_bytes": 56809.4875,
      "train_loss_nats_per_token": 1.761069376979445
    },
    {
      "epoch": 0.006583971978615259,
      "eval_loss": 1.6318442821502686,
      "eval_runtime": 76.4764,
      "eval_samples_per_second": 13.076,
      "eval_steps_per_second": 13.076,
      "step": 62000,
      "train_bpb": 0.5753292046172002,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.6383147618233447
    },
    {
      "epoch": 0.006715651418187564,
      "grad_norm": 0.1256142556667328,
      "learning_rate": 0.0019260319416181818,
      "loss": 1.749222183227539,
      "step": 62020,
      "train_bpb": 0.6194383199517238,
      "train_bytes": 56232.99375,
      "train_loss_nats_per_token": 1.7494376755114711
    },
    {
      "epoch": 0.00684733085775987,
      "grad_norm": 0.10894767194986343,
      "learning_rate": 0.001925436934907874,
      "loss": 1.7566429138183595,
      "step": 62040,
      "train_bpb": 0.6197786026097003,
      "train_bytes": 56887.6125,
      "train_loss_nats_per_token": 1.756597370621486
    },
    {
      "epoch": 0.006979010297332174,
      "grad_norm": 0.14489591121673584,
      "learning_rate": 0.001924841855390946,
      "loss": 1.7518857955932616,
      "step": 62060,
      "train_bpb": 0.6210999759219875,
      "train_bytes": 56366.05,
      "train_loss_nats_per_token": 1.7519024670890846
    },
    {
      "epoch": 0.00711068973690448,
      "grad_norm": 0.13368278741836548,
      "learning_rate": 0.0019242467031692363,
      "loss": 1.7452600479125977,
      "step": 62080,
      "train_bpb": 0.6198154727116922,
      "train_bytes": 56245.95,
      "train_loss_nats_per_token": 1.7455805433087832
    },
    {
      "epoch": 0.007242369176476785,
      "grad_norm": 0.14757318794727325,
      "learning_rate": 0.001923651478344595,
      "loss": 1.7458837509155274,
      "step": 62100,
      "train_bpb": 0.6158564324662211,
      "train_bytes": 57026.4375,
      "train_loss_nats_per_token": 1.7462066170855792
    },
    {
      "epoch": 0.0073740486160490904,
      "grad_norm": 0.15032793581485748,
      "learning_rate": 0.001923056181018886,
      "loss": 1.7416406631469727,
      "step": 62120,
      "train_bpb": 0.6162609450539549,
      "train_bytes": 56746.2,
      "train_loss_nats_per_token": 1.7418991365423
    },
    {
      "epoch": 0.007505728055621395,
      "grad_norm": 0.14605239033699036,
      "learning_rate": 0.0019224608112939849,
      "loss": 1.7678943634033204,
      "step": 62140,
      "train_bpb": 0.6240668820964694,
      "train_bytes": 56957.3375,
      "train_loss_nats_per_token": 1.7683509430533095
    },
    {
      "epoch": 0.007637407495193701,
      "grad_norm": 0.1225280612707138,
      "learning_rate": 0.0019218653692717786,
      "loss": 1.7646080017089845,
      "step": 62160,
      "train_bpb": 0.6230380722542049,
      "train_bytes": 56879.6875,
      "train_loss_nats_per_token": 1.7649099922848515
    },
    {
      "epoch": 0.007769086934766006,
      "grad_norm": 0.10790197551250458,
      "learning_rate": 0.001921269855054168,
      "loss": 1.7782413482666015,
      "step": 62180,
      "train_bpb": 0.6289258402187764,
      "train_bytes": 56811.31875,
      "train_loss_nats_per_token": 1.777947698688714
    },
    {
      "epoch": 0.00790076637433831,
      "grad_norm": 0.18440155684947968,
      "learning_rate": 0.0019206742687430652,
      "loss": 1.7857942581176758,
      "step": 62200,
      "train_bpb": 0.6324715817001405,
      "train_bytes": 56628.66875,
      "train_loss_nats_per_token": 1.7857893620119016
    },
    {
      "epoch": 0.008032445813910615,
      "grad_norm": 0.10779275000095367,
      "learning_rate": 0.001920078610440396,
      "loss": 1.780332374572754,
      "step": 62220,
      "train_bpb": 0.6298762410393011,
      "train_bytes": 56580.625,
      "train_loss_nats_per_token": 1.7802636853303637
    },
    {
      "epoch": 0.008164125253482922,
      "grad_norm": 0.13158485293388367,
      "learning_rate": 0.0019194828802480963,
      "loss": 1.7823537826538085,
      "step": 62240,
      "train_bpb": 0.6339872730882118,
      "train_bytes": 56310.4375,
      "train_loss_nats_per_token": 1.7825659222181989
    },
    {
      "epoch": 0.008295804693055226,
      "grad_norm": 0.16808313131332397,
      "learning_rate": 0.0019188870782681168,
      "loss": 1.7894079208374023,
      "step": 62260,
      "train_bpb": 0.6345407967084842,
      "train_bytes": 56110.89375,
      "train_loss_nats_per_token": 1.7894951838512705
    },
    {
      "epoch": 0.008427484132627531,
      "grad_norm": 0.12873749434947968,
      "learning_rate": 0.0019182912046024187,
      "loss": 1.7840421676635743,
      "step": 62280,
      "train_bpb": 0.6279641668705678,
      "train_bytes": 56729.65625,
      "train_loss_nats_per_token": 1.7844275292619907
    },
    {
      "epoch": 0.008559163572199836,
      "grad_norm": 0.09539773315191269,
      "learning_rate": 0.0019176952593529765,
      "loss": 1.7863594055175782,
      "step": 62300,
      "train_bpb": 0.6326230796725983,
      "train_bytes": 56848.3875,
      "train_loss_nats_per_token": 1.7864485314199083
    },
    {
      "epoch": 0.008690843011772142,
      "grad_norm": 0.13712987303733826,
      "learning_rate": 0.0019170992426217762,
      "loss": 1.7881772994995118,
      "step": 62320,
      "train_bpb": 0.6323951089147853,
      "train_bytes": 56451.55,
      "train_loss_nats_per_token": 1.7879271602087115
    },
    {
      "epoch": 0.008822522451344447,
      "grad_norm": 0.1305018663406372,
      "learning_rate": 0.0019165031545108166,
      "loss": 1.7841411590576173,
      "step": 62340,
      "train_bpb": 0.631782897863029,
      "train_bytes": 56490.375,
      "train_loss_nats_per_token": 1.7841061871123294
    },
    {
      "epoch": 0.008954201890916752,
      "grad_norm": 0.1645195484161377,
      "learning_rate": 0.0019159069951221084,
      "loss": 1.7844476699829102,
      "step": 62360,
      "train_bpb": 0.630962180052814,
      "train_bytes": 56500.6875,
      "train_loss_nats_per_token": 1.7846340580462452
    },
    {
      "epoch": 0.009085881330489057,
      "grad_norm": 0.1745644360780716,
      "learning_rate": 0.001915310764557675,
      "loss": 1.7839271545410156,
      "step": 62380,
      "train_bpb": 0.6308853752474692,
      "train_bytes": 56790.86875,
      "train_loss_nats_per_token": 1.784064362990375
    },
    {
      "epoch": 0.009217560770061363,
      "grad_norm": 0.11974461376667023,
      "learning_rate": 0.0019147144629195507,
      "loss": 1.7823247909545898,
      "step": 62400,
      "train_bpb": 0.6296805458689111,
      "train_bytes": 56872.10625,
      "train_loss_nats_per_token": 1.7821394146094667
    },
    {
      "epoch": 0.009349240209633668,
      "grad_norm": 0.10348714888095856,
      "learning_rate": 0.0019141180903097835,
      "loss": 1.7801977157592774,
      "step": 62420,
      "train_bpb": 0.6313352425793528,
      "train_bytes": 56365.9875,
      "train_loss_nats_per_token": 1.780152054294454
    },
    {
      "epoch": 0.009480919649205973,
      "grad_norm": 0.13203227519989014,
      "learning_rate": 0.0019135216468304326,
      "loss": 1.7769199371337892,
      "step": 62440,
      "train_bpb": 0.6267474640872642,
      "train_bytes": 56837.74375,
      "train_loss_nats_per_token": 1.7767191338797004
    },
    {
      "epoch": 0.009612599088778278,
      "grad_norm": 0.1405482441186905,
      "learning_rate": 0.0019129251325835702,
      "loss": 1.7821220397949218,
      "step": 62460,
      "train_bpb": 0.6302602500125095,
      "train_bytes": 56581.3625,
      "train_loss_nats_per_token": 1.7823484634280262
    },
    {
      "epoch": 0.009744278528350584,
      "grad_norm": 0.1541583240032196,
      "learning_rate": 0.0019123285476712792,
      "loss": 1.7782817840576173,
      "step": 62480,
      "train_bpb": 0.6305136153492834,
      "train_bytes": 56223.50625,
      "train_loss_nats_per_token": 1.7782694956744676
    },
    {
      "epoch": 0.009875957967922889,
      "grad_norm": 0.1275201290845871,
      "learning_rate": 0.0019117318921956557,
      "loss": 1.7782909393310546,
      "step": 62500,
      "train_bpb": 0.6289028247919872,
      "train_bytes": 56647.58125,
      "train_loss_nats_per_token": 1.7786777735987562
    },
    {
      "epoch": 0.010007637407495194,
      "grad_norm": 0.11479001492261887,
      "learning_rate": 0.0019111351662588084,
      "loss": 1.7755208969116212,
      "step": 62520,
      "train_bpb": 0.6285422568344066,
      "train_bytes": 56485.15,
      "train_loss_nats_per_token": 1.776023522359223
    },
    {
      "epoch": 0.010139316847067498,
      "grad_norm": 0.10506820678710938,
      "learning_rate": 0.0019105383699628557,
      "loss": 1.7753276824951172,
      "step": 62540,
      "train_bpb": 0.6274289420284491,
      "train_bytes": 56672.8875,
      "train_loss_nats_per_token": 1.7754618883462472
    },
    {
      "epoch": 0.010270996286639805,
      "grad_norm": 0.12774519622325897,
      "learning_rate": 0.001909941503409931,
      "loss": 1.7802400588989258,
      "step": 62560,
      "train_bpb": 0.6279683699626438,
      "train_bytes": 56896.3,
      "train_loss_nats_per_token": 1.7803536156554023
    },
    {
      "epoch": 0.01040267572621211,
      "grad_norm": 0.22679783403873444,
      "learning_rate": 0.0019093445667021773,
      "loss": 1.782571792602539,
      "step": 62580,
      "train_bpb": 0.6291236840313551,
      "train_bytes": 56624.2875,
      "train_loss_nats_per_token": 1.7826451114434476
    },
    {
      "epoch": 0.010534355165784414,
      "grad_norm": 0.16639713943004608,
      "learning_rate": 0.0019087475599417513,
      "loss": 1.764303970336914,
      "step": 62600,
      "train_bpb": 0.6238385290748507,
      "train_bytes": 56356.5125,
      "train_loss_nats_per_token": 1.764926861250679
    },
    {
      "epoch": 0.01066603460535672,
      "grad_norm": 0.11111053824424744,
      "learning_rate": 0.0019081504832308208,
      "loss": 1.7819591522216798,
      "step": 62620,
      "train_bpb": 0.6309396241865605,
      "train_bytes": 56803.475,
      "train_loss_nats_per_token": 1.7824450168003851
    },
    {
      "epoch": 0.010797714044929024,
      "grad_norm": 0.11637653410434723,
      "learning_rate": 0.0019075533366715652,
      "loss": 1.7749759674072265,
      "step": 62640,
      "train_bpb": 0.6292787672802677,
      "train_bytes": 56288.28125,
      "train_loss_nats_per_token": 1.7755232804437795
    },
    {
      "epoch": 0.01092939348450133,
      "grad_norm": 0.13422615826129913,
      "learning_rate": 0.0019069561203661773,
      "loss": 1.764759635925293,
      "step": 62660,
      "train_bpb": 0.6242687437618313,
      "train_bytes": 56677.775,
      "train_loss_nats_per_token": 1.7651633215061844
    },
    {
      "epoch": 0.011061072924073635,
      "grad_norm": 0.10001391172409058,
      "learning_rate": 0.0019063588344168606,
      "loss": 1.766859245300293,
      "step": 62680,
      "train_bpb": 0.6254109753392681,
      "train_bytes": 56835.45625,
      "train_loss_nats_per_token": 1.76704123926647
    },
    {
      "epoch": 0.01119275236364594,
      "grad_norm": 0.1593712717294693,
      "learning_rate": 0.0019057614789258315,
      "loss": 1.7658124923706056,
      "step": 62700,
      "train_bpb": 0.6251774189747066,
      "train_bytes": 56348.6125,
      "train_loss_nats_per_token": 1.7658522646313781
    },
    {
      "epoch": 0.011324431803218245,
      "grad_norm": 0.12488920241594315,
      "learning_rate": 0.0019051640539953165,
      "loss": 1.765009307861328,
      "step": 62720,
      "train_bpb": 0.6241584859053644,
      "train_bytes": 56671.65625,
      "train_loss_nats_per_token": 1.7651890734167304
    },
    {
      "epoch": 0.011456111242790551,
      "grad_norm": 0.12321136146783829,
      "learning_rate": 0.0019045665597275557,
      "loss": 1.76617488861084,
      "step": 62740,
      "train_bpb": 0.625018202988811,
      "train_bytes": 56977.73125,
      "train_loss_nats_per_token": 1.7662115398192562
    },
    {
      "epoch": 0.011587790682362856,
      "grad_norm": 0.17460301518440247,
      "learning_rate": 0.0019039689962248006,
      "loss": 1.7706453323364257,
      "step": 62760,
      "train_bpb": 0.625686146306743,
      "train_bytes": 56931.0625,
      "train_loss_nats_per_token": 1.770448040197434
    },
    {
      "epoch": 0.01171947012193516,
      "grad_norm": 0.13196802139282227,
      "learning_rate": 0.0019033713635893147,
      "loss": 1.7548862457275392,
      "step": 62780,
      "train_bpb": 0.620246503945001,
      "train_bytes": 56546.1,
      "train_loss_nats_per_token": 1.7550993743355772
    },
    {
      "epoch": 0.011851149561507466,
      "grad_norm": 0.12338767945766449,
      "learning_rate": 0.0019027736619233733,
      "loss": 1.771772575378418,
      "step": 62800,
      "train_bpb": 0.6262262794625532,
      "train_bytes": 56334.225,
      "train_loss_nats_per_token": 1.7719592272943585
    },
    {
      "epoch": 0.011982829001079772,
      "grad_norm": 0.1355729103088379,
      "learning_rate": 0.0019021758913292627,
      "loss": 1.7570314407348633,
      "step": 62820,
      "train_bpb": 0.6241798579179059,
      "train_bytes": 56423.4375,
      "train_loss_nats_per_token": 1.7573517604230546
    },
    {
      "epoch": 0.012114508440652077,
      "grad_norm": 0.10685624182224274,
      "learning_rate": 0.001901578051909282,
      "loss": 1.7627410888671875,
      "step": 62840,
      "train_bpb": 0.6245839674159255,
      "train_bytes": 56612.975,
      "train_loss_nats_per_token": 1.7631987235371274
    },
    {
      "epoch": 0.012246187880224382,
      "grad_norm": 0.10736310482025146,
      "learning_rate": 0.001900980143765741,
      "loss": 1.763224983215332,
      "step": 62860,
      "train_bpb": 0.6232148387915107,
      "train_bytes": 56977.6875,
      "train_loss_nats_per_token": 1.7631862984563045
    },
    {
      "epoch": 0.012377867319796686,
      "grad_norm": 0.1110725849866867,
      "learning_rate": 0.0019003821670009631,
      "loss": 1.7600982666015625,
      "step": 62880,
      "train_bpb": 0.6221150330970139,
      "train_bytes": 56202.7875,
      "train_loss_nats_per_token": 1.7603128476681416
    },
    {
      "epoch": 0.012509546759368993,
      "grad_norm": 0.142109677195549,
      "learning_rate": 0.0018997841217172817,
      "loss": 1.765638542175293,
      "step": 62900,
      "train_bpb": 0.6263326242417033,
      "train_bytes": 56763.74375,
      "train_loss_nats_per_token": 1.7657771681581687
    },
    {
      "epoch": 0.012641226198941298,
      "grad_norm": 0.11788687855005264,
      "learning_rate": 0.0018991860080170425,
      "loss": 1.7701026916503906,
      "step": 62920,
      "train_bpb": 0.6249529635580896,
      "train_bytes": 56617.75,
      "train_loss_nats_per_token": 1.769762746749469
    },
    {
      "epoch": 0.012772905638513602,
      "grad_norm": 0.13505342602729797,
      "learning_rate": 0.0018985878260026029,
      "loss": 1.7570180892944336,
      "step": 62940,
      "train_bpb": 0.6229104624324007,
      "train_bytes": 56246.03125,
      "train_loss_nats_per_token": 1.7575544249140818
    },
    {
      "epoch": 0.012904585078085907,
      "grad_norm": 0.12131771445274353,
      "learning_rate": 0.0018979895757763325,
      "loss": 1.7501750946044923,
      "step": 62960,
      "train_bpb": 0.6207484221809032,
      "train_bytes": 56215.50625,
      "train_loss_nats_per_token": 1.7507055911000742
    },
    {
      "epoch": 0.013036264517658214,
      "grad_norm": 0.10991626977920532,
      "learning_rate": 0.0018973912574406114,
      "loss": 1.7573970794677733,
      "step": 62980,
      "train_bpb": 0.6213573178950063,
      "train_bytes": 56764.69375,
      "train_loss_nats_per_token": 1.757403989004647
    },
    {
      "epoch": 0.013167943957230518,
      "grad_norm": 0.1373276561498642,
      "learning_rate": 0.0018967928710978323,
      "loss": 1.7590560913085938,
      "step": 63000,
      "train_bpb": 0.620224440561614,
      "train_bytes": 56735.33125,
      "train_loss_nats_per_token": 1.759084572150992
    },
    {
      "epoch": 0.013167943957230518,
      "eval_loss": 1.6689921617507935,
      "eval_runtime": 75.5423,
      "eval_samples_per_second": 13.238,
      "eval_steps_per_second": 13.238,
      "step": 63000,
      "train_bpb": 0.5884653517859451,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.6757214216058254
    },
    {
      "epoch": 0.013299623396802823,
      "grad_norm": 0.10157153755426407,
      "learning_rate": 0.0018961944168503998,
      "loss": 1.7591888427734375,
      "step": 63020,
      "train_bpb": 0.6232433336735923,
      "train_bytes": 56429.2,
      "train_loss_nats_per_token": 1.7587621873640193
    },
    {
      "epoch": 0.013431302836375128,
      "grad_norm": 0.15738347172737122,
      "learning_rate": 0.0018955958948007295,
      "loss": 1.7578702926635743,
      "step": 63040,
      "train_bpb": 0.6188185276412344,
      "train_bytes": 57377.70625,
      "train_loss_nats_per_token": 1.7579316182949518
    },
    {
      "epoch": 0.013562982275947433,
      "grad_norm": 0.14567504823207855,
      "learning_rate": 0.0018949973050512484,
      "loss": 1.7547531127929688,
      "step": 63060,
      "train_bpb": 0.6214245267423555,
      "train_bytes": 56121.5375,
      "train_loss_nats_per_token": 1.7547137869808678
    },
    {
      "epoch": 0.01369466171551974,
      "grad_norm": 0.14020739495754242,
      "learning_rate": 0.0018943986477043952,
      "loss": 1.7541475296020508,
      "step": 63080,
      "train_bpb": 0.6212041545249202,
      "train_bytes": 56662.075,
      "train_loss_nats_per_token": 1.7544456229688312
    },
    {
      "epoch": 0.013826341155092044,
      "grad_norm": 0.1258753389120102,
      "learning_rate": 0.0018937999228626206,
      "loss": 1.759600830078125,
      "step": 63100,
      "train_bpb": 0.6223455303275037,
      "train_bytes": 56453.775,
      "train_loss_nats_per_token": 1.7598081350182893
    },
    {
      "epoch": 0.013958020594664349,
      "grad_norm": 0.12510743737220764,
      "learning_rate": 0.001893201130628387,
      "loss": 1.7535943984985352,
      "step": 63120,
      "train_bpb": 0.6219306821200631,
      "train_bytes": 56533.9875,
      "train_loss_nats_per_token": 1.7542756251746112
    },
    {
      "epoch": 0.014089700034236654,
      "grad_norm": 0.10973630845546722,
      "learning_rate": 0.0018926022711041675,
      "loss": 1.752177619934082,
      "step": 63140,
      "train_bpb": 0.6214176740896235,
      "train_bytes": 56140.94375,
      "train_loss_nats_per_token": 1.7523628737428252
    },
    {
      "epoch": 0.01422137947380896,
      "grad_norm": 0.09221392124891281,
      "learning_rate": 0.0018920033443924475,
      "loss": 1.7596546173095704,
      "step": 63160,
      "train_bpb": 0.623375675996879,
      "train_bytes": 56623.3625,
      "train_loss_nats_per_token": 1.7595694685738439
    },
    {
      "epoch": 0.014353058913381265,
      "grad_norm": 0.14081476628780365,
      "learning_rate": 0.0018914043505957236,
      "loss": 1.7529911041259765,
      "step": 63180,
      "train_bpb": 0.6203876619404499,
      "train_bytes": 56282.28125,
      "train_loss_nats_per_token": 1.7530027847952816
    },
    {
      "epoch": 0.01448473835295357,
      "grad_norm": 0.1228165253996849,
      "learning_rate": 0.0018908052898165032,
      "loss": 1.7495676040649415,
      "step": 63200,
      "train_bpb": 0.6178718371252654,
      "train_bytes": 56853.2125,
      "train_loss_nats_per_token": 1.749477170548349
    },
    {
      "epoch": 0.014616417792525874,
      "grad_norm": 0.1173931360244751,
      "learning_rate": 0.0018902061621573069,
      "loss": 1.7528549194335938,
      "step": 63220,
      "train_bpb": 0.6206915196457662,
      "train_bytes": 56691.075,
      "train_loss_nats_per_token": 1.7530598398560895
    },
    {
      "epoch": 0.014748097232098181,
      "grad_norm": 0.11473619192838669,
      "learning_rate": 0.0018896069677206645,
      "loss": 1.7458625793457032,
      "step": 63240,
      "train_bpb": 0.616160866441797,
      "train_bytes": 57091.675,
      "train_loss_nats_per_token": 1.7462601941385114
    },
    {
      "epoch": 0.014879776671670486,
      "grad_norm": 0.16624103486537933,
      "learning_rate": 0.0018890077066091194,
      "loss": 1.7553485870361327,
      "step": 63260,
      "train_bpb": 0.6193389547497082,
      "train_bytes": 56810.96875,
      "train_loss_nats_per_token": 1.7554144233724656
    },
    {
      "epoch": 0.01501145611124279,
      "grad_norm": 0.1392402946949005,
      "learning_rate": 0.0018884083789252248,
      "loss": 1.7427165985107422,
      "step": 63280,
      "train_bpb": 0.6185197663124349,
      "train_bytes": 56270.75,
      "train_loss_nats_per_token": 1.7426766941230183
    },
    {
      "epoch": 0.015143135550815095,
      "grad_norm": 0.16094620525836945,
      "learning_rate": 0.0018878089847715461,
      "loss": 1.7556787490844727,
      "step": 63300,
      "train_bpb": 0.6213197895562181,
      "train_bytes": 56747.20625,
      "train_loss_nats_per_token": 1.7552802296135943
    },
    {
      "epoch": 0.015274814990387402,
      "grad_norm": 0.09358065575361252,
      "learning_rate": 0.0018872095242506595,
      "loss": 1.7495384216308594,
      "step": 63320,
      "train_bpb": 0.6203683631858393,
      "train_bytes": 56604.51875,
      "train_loss_nats_per_token": 1.7499090098655772
    },
    {
      "epoch": 0.015406494429959706,
      "grad_norm": 0.10762621462345123,
      "learning_rate": 0.0018866099974651535,
      "loss": 1.7723360061645508,
      "step": 63340,
      "train_bpb": 0.6260602485243288,
      "train_bytes": 56681.525,
      "train_loss_nats_per_token": 1.7728629577938502
    },
    {
      "epoch": 0.015538173869532011,
      "grad_norm": 0.11712884157896042,
      "learning_rate": 0.0018860104045176273,
      "loss": 1.7746639251708984,
      "step": 63360,
      "train_bpb": 0.630135113212128,
      "train_bytes": 56290.19375,
      "train_loss_nats_per_token": 1.7751620145902838
    },
    {
      "epoch": 0.015669853309104316,
      "grad_norm": 0.11327071487903595,
      "learning_rate": 0.001885410745510691,
      "loss": 1.7801397323608399,
      "step": 63380,
      "train_bpb": 0.6312287238150456,
      "train_bytes": 56405.6625,
      "train_loss_nats_per_token": 1.780721380447015
    },
    {
      "epoch": 0.01580153274867662,
      "grad_norm": 0.14787380397319794,
      "learning_rate": 0.0018848110205469672,
      "loss": 1.7747804641723632,
      "step": 63400,
      "train_bpb": 0.6296944305337155,
      "train_bytes": 56437.2,
      "train_loss_nats_per_token": 1.7748890159302952
    },
    {
      "epoch": 0.015933212188248926,
      "grad_norm": 0.1872226595878601,
      "learning_rate": 0.0018842112297290884,
      "loss": 1.788611602783203,
      "step": 63420,
      "train_bpb": 0.6314911638538352,
      "train_bytes": 56651.86875,
      "train_loss_nats_per_token": 1.78906582465322
    },
    {
      "epoch": 0.01606489162782123,
      "grad_norm": 0.10592175275087357,
      "learning_rate": 0.0018836113731596999,
      "loss": 1.7855497360229493,
      "step": 63440,
      "train_bpb": 0.6316710042841067,
      "train_bytes": 56515.64375,
      "train_loss_nats_per_token": 1.7864240307371566
    },
    {
      "epoch": 0.01619657106739354,
      "grad_norm": 0.143657848238945,
      "learning_rate": 0.0018830114509414562,
      "loss": 1.783848190307617,
      "step": 63460,
      "train_bpb": 0.6295041546850556,
      "train_bytes": 56882.34375,
      "train_loss_nats_per_token": 1.784866591620197
    },
    {
      "epoch": 0.016328250506965843,
      "grad_norm": 0.12536552548408508,
      "learning_rate": 0.0018824114631770252,
      "loss": 1.7773574829101562,
      "step": 63480,
      "train_bpb": 0.6308511814468614,
      "train_bytes": 56164.46875,
      "train_loss_nats_per_token": 1.77772637637979
    },
    {
      "epoch": 0.016459929946538148,
      "grad_norm": 0.12971702218055725,
      "learning_rate": 0.0018818114099690844,
      "loss": 1.771673583984375,
      "step": 63500,
      "train_bpb": 0.6277891357960419,
      "train_bytes": 56333.8625,
      "train_loss_nats_per_token": 1.7717072206803417
    },
    {
      "epoch": 0.016591609386110453,
      "grad_norm": 0.17239081859588623,
      "learning_rate": 0.0018812112914203245,
      "loss": 1.7763727188110352,
      "step": 63520,
      "train_bpb": 0.6301139832593678,
      "train_bytes": 56231.45,
      "train_loss_nats_per_token": 1.7770076761464229
    },
    {
      "epoch": 0.016723288825682758,
      "grad_norm": 0.17456717789173126,
      "learning_rate": 0.0018806111076334444,
      "loss": 1.774092674255371,
      "step": 63540,
      "train_bpb": 0.628126843147184,
      "train_bytes": 56338.0625,
      "train_loss_nats_per_token": 1.7747051960381766
    },
    {
      "epoch": 0.016854968265255062,
      "grad_norm": 0.1358521580696106,
      "learning_rate": 0.0018800108587111568,
      "loss": 1.7809818267822266,
      "step": 63560,
      "train_bpb": 0.6299592887645886,
      "train_bytes": 56605.78125,
      "train_loss_nats_per_token": 1.7815781158494008
    },
    {
      "epoch": 0.016986647704827367,
      "grad_norm": 0.14188097417354584,
      "learning_rate": 0.0018794105447561837,
      "loss": 1.782798957824707,
      "step": 63580,
      "train_bpb": 0.6305928052907248,
      "train_bytes": 56887.49375,
      "train_loss_nats_per_token": 1.783156637572643
    },
    {
      "epoch": 0.017118327144399672,
      "grad_norm": 0.13208572566509247,
      "learning_rate": 0.0018788101658712597,
      "loss": 1.7681951522827148,
      "step": 63600,
      "train_bpb": 0.6269960904772092,
      "train_bytes": 56529.63125,
      "train_loss_nats_per_token": 1.7688435007406276
    },
    {
      "epoch": 0.01725000658397198,
      "grad_norm": 0.09853200614452362,
      "learning_rate": 0.0018782097221591305,
      "loss": 1.7687847137451171,
      "step": 63620,
      "train_bpb": 0.6267829371470116,
      "train_bytes": 57108.78125,
      "train_loss_nats_per_token": 1.7689105743609053
    },
    {
      "epoch": 0.017381686023544285,
      "grad_norm": 0.09453556686639786,
      "learning_rate": 0.001877609213722551,
      "loss": 1.7657785415649414,
      "step": 63640,
      "train_bpb": 0.6259778412717085,
      "train_bytes": 56899.96875,
      "train_loss_nats_per_token": 1.7660810703050602
    },
    {
      "epoch": 0.01751336546311659,
      "grad_norm": 0.0944974347949028,
      "learning_rate": 0.0018770086406642889,
      "loss": 1.778507614135742,
      "step": 63660,
      "train_bpb": 0.6296160006790142,
      "train_bytes": 56525.2125,
      "train_loss_nats_per_token": 1.7787596021929135
    },
    {
      "epoch": 0.017645044902688894,
      "grad_norm": 0.1238592192530632,
      "learning_rate": 0.001876408003087123,
      "loss": 1.758046531677246,
      "step": 63680,
      "train_bpb": 0.6220465383516208,
      "train_bytes": 56769.68125,
      "train_loss_nats_per_token": 1.7583063961273755
    },
    {
      "epoch": 0.0177767243422612,
      "grad_norm": 0.09906717389822006,
      "learning_rate": 0.0018758073010938425,
      "loss": 1.7715818405151367,
      "step": 63700,
      "train_bpb": 0.6275071733299665,
      "train_bytes": 56469.2,
      "train_loss_nats_per_token": 1.7720170063840168
    },
    {
      "epoch": 0.017908403781833504,
      "grad_norm": 0.1368124634027481,
      "learning_rate": 0.0018752065347872474,
      "loss": 1.7624378204345703,
      "step": 63720,
      "train_bpb": 0.6234465488804729,
      "train_bytes": 56413.8875,
      "train_loss_nats_per_token": 1.7625794176792802
    },
    {
      "epoch": 0.01804008322140581,
      "grad_norm": 0.14230164885520935,
      "learning_rate": 0.001874605704270149,
      "loss": 1.7500644683837892,
      "step": 63740,
      "train_bpb": 0.619467373070138,
      "train_bytes": 56640.575,
      "train_loss_nats_per_token": 1.750377630429683
    },
    {
      "epoch": 0.018171762660978114,
      "grad_norm": 0.1629158854484558,
      "learning_rate": 0.0018740048096453701,
      "loss": 1.764028549194336,
      "step": 63760,
      "train_bpb": 0.6253701724171671,
      "train_bytes": 56652.8375,
      "train_loss_nats_per_token": 1.7637753956522066
    },
    {
      "epoch": 0.018303442100550422,
      "grad_norm": 0.10744798183441162,
      "learning_rate": 0.0018734038510157445,
      "loss": 1.772380256652832,
      "step": 63780,
      "train_bpb": 0.6260907083496189,
      "train_bytes": 56680.79375,
      "train_loss_nats_per_token": 1.7726676117854572
    },
    {
      "epoch": 0.018435121540122727,
      "grad_norm": 0.19269350171089172,
      "learning_rate": 0.0018728028284841155,
      "loss": 1.7666484832763671,
      "step": 63800,
      "train_bpb": 0.6246088181514404,
      "train_bytes": 56467.54375,
      "train_loss_nats_per_token": 1.7665502466919936
    },
    {
      "epoch": 0.01856680097969503,
      "grad_norm": 0.10624369233846664,
      "learning_rate": 0.001872201742153339,
      "loss": 1.755765151977539,
      "step": 63820,
      "train_bpb": 0.6206369808456745,
      "train_bytes": 56724.9,
      "train_loss_nats_per_token": 1.7561542983859995
    },
    {
      "epoch": 0.018698480419267336,
      "grad_norm": 0.13163509964942932,
      "learning_rate": 0.0018716005921262811,
      "loss": 1.760056495666504,
      "step": 63840,
      "train_bpb": 0.6241102758674527,
      "train_bytes": 56323.54375,
      "train_loss_nats_per_token": 1.7605991377660606
    },
    {
      "epoch": 0.01883015985883964,
      "grad_norm": 0.1799865961074829,
      "learning_rate": 0.0018709993785058184,
      "loss": 1.767264175415039,
      "step": 63860,
      "train_bpb": 0.6249747321895248,
      "train_bytes": 56390.625,
      "train_loss_nats_per_token": 1.7678863978477115
    },
    {
      "epoch": 0.018961839298411946,
      "grad_norm": 0.11212833970785141,
      "learning_rate": 0.0018703981013948395,
      "loss": 1.761361312866211,
      "step": 63880,
      "train_bpb": 0.6256681121569294,
      "train_bytes": 56332.7,
      "train_loss_nats_per_token": 1.7617290562776422
    },
    {
      "epoch": 0.01909351873798425,
      "grad_norm": 0.10379034280776978,
      "learning_rate": 0.0018697967608962425,
      "loss": 1.750840950012207,
      "step": 63900,
      "train_bpb": 0.6187418073742522,
      "train_bytes": 56721.78125,
      "train_loss_nats_per_token": 1.7513697635160976
    },
    {
      "epoch": 0.019225198177556555,
      "grad_norm": 0.11276817321777344,
      "learning_rate": 0.0018691953571129375,
      "loss": 1.752840232849121,
      "step": 63920,
      "train_bpb": 0.6194070158918125,
      "train_bytes": 56750.60625,
      "train_loss_nats_per_token": 1.7535155784388747
    },
    {
      "epoch": 0.01935687761712886,
      "grad_norm": 0.10171769559383392,
      "learning_rate": 0.0018685938901478452,
      "loss": 1.74127197265625,
      "step": 63940,
      "train_bpb": 0.6177999374095071,
      "train_bytes": 56328.3875,
      "train_loss_nats_per_token": 1.741209883178063
    },
    {
      "epoch": 0.019488557056701168,
      "grad_norm": 0.1282004714012146,
      "learning_rate": 0.0018679923601038963,
      "loss": 1.7464399337768555,
      "step": 63960,
      "train_bpb": 0.6178984827703066,
      "train_bytes": 56676.35625,
      "train_loss_nats_per_token": 1.7466153875417443
    },
    {
      "epoch": 0.019620236496273473,
      "grad_norm": 0.11716566979885101,
      "learning_rate": 0.0018673907670840335,
      "loss": 1.7629877090454102,
      "step": 63980,
      "train_bpb": 0.6234754589938167,
      "train_bytes": 56293.83125,
      "train_loss_nats_per_token": 1.7628107748577853
    },
    {
      "epoch": 0.019751915935845778,
      "grad_norm": 0.12950582802295685,
      "learning_rate": 0.001866789111191209,
      "loss": 1.7497800827026366,
      "step": 64000,
      "train_bpb": 0.6188692730032028,
      "train_bytes": 56513.58125,
      "train_loss_nats_per_token": 1.7500143344791845
    },
    {
      "epoch": 0.019751915935845778,
      "eval_loss": 1.6860284805297852,
      "eval_runtime": 76.3352,
      "eval_samples_per_second": 13.1,
      "eval_steps_per_second": 13.1,
      "step": 64000,
      "train_bpb": 0.5944245585033697,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.6926909344613512
    },
    {
      "epoch": 0.019883595375418082,
      "grad_norm": 0.11891575157642365,
      "learning_rate": 0.0018661873925283872,
      "loss": 1.7486011505126953,
      "step": 64020,
      "train_bpb": 0.6182613610399195,
      "train_bytes": 56614.95,
      "train_loss_nats_per_token": 1.748804598987754
    },
    {
      "epoch": 0.020015274814990387,
      "grad_norm": 0.11116427183151245,
      "learning_rate": 0.001865585611198542,
      "loss": 1.754350471496582,
      "step": 64040,
      "train_bpb": 0.6181302529823262,
      "train_bytes": 56913.95,
      "train_loss_nats_per_token": 1.7542939214731685
    },
    {
      "epoch": 0.020146954254562692,
      "grad_norm": 0.11663824319839478,
      "learning_rate": 0.0018649837673046586,
      "loss": 1.754552459716797,
      "step": 64060,
      "train_bpb": 0.6189975991535586,
      "train_bytes": 56580.24375,
      "train_loss_nats_per_token": 1.754912253416025
    },
    {
      "epoch": 0.020278633694134997,
      "grad_norm": 0.11642865091562271,
      "learning_rate": 0.0018643818609497328,
      "loss": 1.752797508239746,
      "step": 64080,
      "train_bpb": 0.6185939519742618,
      "train_bytes": 57080.625,
      "train_loss_nats_per_token": 1.7528659056755225
    },
    {
      "epoch": 0.0204103131337073,
      "grad_norm": 0.10608592629432678,
      "learning_rate": 0.001863779892236771,
      "loss": 1.7468135833740235,
      "step": 64100,
      "train_bpb": 0.6204764745195106,
      "train_bytes": 56546.48125,
      "train_loss_nats_per_token": 1.7471790382029446
    },
    {
      "epoch": 0.02054199257327961,
      "grad_norm": 0.12963713705539703,
      "learning_rate": 0.0018631778612687904,
      "loss": 1.741278839111328,
      "step": 64120,
      "train_bpb": 0.6159586242575156,
      "train_bytes": 56619.5625,
      "train_loss_nats_per_token": 1.7415755186054052
    },
    {
      "epoch": 0.020673672012851915,
      "grad_norm": 0.10895021259784698,
      "learning_rate": 0.001862575768148819,
      "loss": 1.7461044311523437,
      "step": 64140,
      "train_bpb": 0.619333920433068,
      "train_bytes": 56487.0625,
      "train_loss_nats_per_token": 1.746052368668804
    },
    {
      "epoch": 0.02080535145242422,
      "grad_norm": 0.10060863941907883,
      "learning_rate": 0.0018619736129798952,
      "loss": 1.7542665481567383,
      "step": 64160,
      "train_bpb": 0.6198888580072623,
      "train_bytes": 56803.1,
      "train_loss_nats_per_token": 1.7547309784486051
    },
    {
      "epoch": 0.020937030891996524,
      "grad_norm": 0.096989706158638,
      "learning_rate": 0.0018613713958650683,
      "loss": 1.7430753707885742,
      "step": 64180,
      "train_bpb": 0.6182423639599679,
      "train_bytes": 56895.6125,
      "train_loss_nats_per_token": 1.743417892941417
    },
    {
      "epoch": 0.02106871033156883,
      "grad_norm": 0.15476413071155548,
      "learning_rate": 0.001860769116907397,
      "loss": 1.7512142181396484,
      "step": 64200,
      "train_bpb": 0.6178615870212195,
      "train_bytes": 56691.45,
      "train_loss_nats_per_token": 1.7509542722807538
    },
    {
      "epoch": 0.021200389771141134,
      "grad_norm": 0.10801466554403305,
      "learning_rate": 0.001860166776209953,
      "loss": 1.7460174560546875,
      "step": 64220,
      "train_bpb": 0.6179986001324165,
      "train_bytes": 56546.03125,
      "train_loss_nats_per_token": 1.746380268602827
    },
    {
      "epoch": 0.02133206921071344,
      "grad_norm": 0.13565224409103394,
      "learning_rate": 0.0018595643738758157,
      "loss": 1.7497320175170898,
      "step": 64240,
      "train_bpb": 0.6213920502587109,
      "train_bytes": 56025.73125,
      "train_loss_nats_per_token": 1.7499968006866977
    },
    {
      "epoch": 0.021463748650285743,
      "grad_norm": 0.09991683810949326,
      "learning_rate": 0.0018589619100080773,
      "loss": 1.7514360427856446,
      "step": 64260,
      "train_bpb": 0.6185526415721219,
      "train_bytes": 57186.2125,
      "train_loss_nats_per_token": 1.7517059818730671
    },
    {
      "epoch": 0.021595428089858048,
      "grad_norm": 0.2032524198293686,
      "learning_rate": 0.0018583593847098397,
      "loss": 1.7404657363891602,
      "step": 64280,
      "train_bpb": 0.6172124215377303,
      "train_bytes": 56325.1875,
      "train_loss_nats_per_token": 1.7405247197265576
    },
    {
      "epoch": 0.021727107529430356,
      "grad_norm": 0.12631218135356903,
      "learning_rate": 0.0018577567980842152,
      "loss": 1.740738296508789,
      "step": 64300,
      "train_bpb": 0.6166500691323835,
      "train_bytes": 56234.4125,
      "train_loss_nats_per_token": 1.7408280562102425
    },
    {
      "epoch": 0.02185878696900266,
      "grad_norm": 0.15879036486148834,
      "learning_rate": 0.0018571541502343265,
      "loss": 1.7445768356323241,
      "step": 64320,
      "train_bpb": 0.6187617837413681,
      "train_bytes": 56477.63125,
      "train_loss_nats_per_token": 1.7447102984746279
    },
    {
      "epoch": 0.021990466408574966,
      "grad_norm": 0.1627405881881714,
      "learning_rate": 0.001856551441263307,
      "loss": 1.756363868713379,
      "step": 64340,
      "train_bpb": 0.6249487041543555,
      "train_bytes": 55814.64375,
      "train_loss_nats_per_token": 1.756313111258582
    },
    {
      "epoch": 0.02212214584814727,
      "grad_norm": 0.14653439819812775,
      "learning_rate": 0.0018559486712743014,
      "loss": 1.7364974975585938,
      "step": 64360,
      "train_bpb": 0.6140769481480213,
      "train_bytes": 56349.31875,
      "train_loss_nats_per_token": 1.7361852283069892
    },
    {
      "epoch": 0.022253825287719575,
      "grad_norm": 0.15920495986938477,
      "learning_rate": 0.0018553458403704627,
      "loss": 1.73450927734375,
      "step": 64380,
      "train_bpb": 0.6111078891715178,
      "train_bytes": 57159.1,
      "train_loss_nats_per_token": 1.734520678914937
    },
    {
      "epoch": 0.02238550472729188,
      "grad_norm": 0.11347683519124985,
      "learning_rate": 0.0018547429486549564,
      "loss": 1.7384281158447266,
      "step": 64400,
      "train_bpb": 0.6120650426387823,
      "train_bytes": 56840.625,
      "train_loss_nats_per_token": 1.7387185858763752
    },
    {
      "epoch": 0.022517184166864185,
      "grad_norm": 0.1647258698940277,
      "learning_rate": 0.0018541399962309572,
      "loss": 1.738259506225586,
      "step": 64420,
      "train_bpb": 0.6164002361385981,
      "train_bytes": 56125.6875,
      "train_loss_nats_per_token": 1.7387567187900779
    },
    {
      "epoch": 0.02264886360643649,
      "grad_norm": 0.10896439105272293,
      "learning_rate": 0.0018535369832016516,
      "loss": 1.7492368698120118,
      "step": 64440,
      "train_bpb": 0.6203626568558797,
      "train_bytes": 56191.7,
      "train_loss_nats_per_token": 1.7490111849574443
    },
    {
      "epoch": 0.022780543046008798,
      "grad_norm": 0.1352044939994812,
      "learning_rate": 0.0018529339096702339,
      "loss": 1.7411384582519531,
      "step": 64460,
      "train_bpb": 0.6168982557176247,
      "train_bytes": 56629.9875,
      "train_loss_nats_per_token": 1.7412377135532555
    },
    {
      "epoch": 0.022912222485581103,
      "grad_norm": 0.14860899746418,
      "learning_rate": 0.0018523307757399114,
      "loss": 1.746278953552246,
      "step": 64480,
      "train_bpb": 0.6198202410283647,
      "train_bytes": 56680.0875,
      "train_loss_nats_per_token": 1.7467835496995183
    },
    {
      "epoch": 0.023043901925153407,
      "grad_norm": 0.10923822224140167,
      "learning_rate": 0.0018517275815139005,
      "loss": 1.7400382995605468,
      "step": 64500,
      "train_bpb": 0.6161985715397423,
      "train_bytes": 56377.84375,
      "train_loss_nats_per_token": 1.7399971126612117
    },
    {
      "epoch": 0.023175581364725712,
      "grad_norm": 0.12360353022813797,
      "learning_rate": 0.001851124327095428,
      "loss": 1.7517929077148438,
      "step": 64520,
      "train_bpb": 0.6202454017697927,
      "train_bytes": 56585.6625,
      "train_loss_nats_per_token": 1.7518942951235095
    },
    {
      "epoch": 0.023307260804298017,
      "grad_norm": 0.12313119322061539,
      "learning_rate": 0.0018505210125877316,
      "loss": 1.7614748001098632,
      "step": 64540,
      "train_bpb": 0.6233375177308366,
      "train_bytes": 56617.4125,
      "train_loss_nats_per_token": 1.7616333573553387
    },
    {
      "epoch": 0.02343894024387032,
      "grad_norm": 0.11500637978315353,
      "learning_rate": 0.001849917638094058,
      "loss": 1.7622629165649415,
      "step": 64560,
      "train_bpb": 0.6250420942604813,
      "train_bytes": 56750.525,
      "train_loss_nats_per_token": 1.762733424811165
    },
    {
      "epoch": 0.023570619683442626,
      "grad_norm": 0.09867610037326813,
      "learning_rate": 0.0018493142037176653,
      "loss": 1.7768150329589845,
      "step": 64580,
      "train_bpb": 0.6297290605516807,
      "train_bytes": 56673.575,
      "train_loss_nats_per_token": 1.7767399319535184
    },
    {
      "epoch": 0.02370229912301493,
      "grad_norm": 0.13972672820091248,
      "learning_rate": 0.0018487107095618213,
      "loss": 1.770728874206543,
      "step": 64600,
      "train_bpb": 0.6266389612114293,
      "train_bytes": 56233.2875,
      "train_loss_nats_per_token": 1.770305110369278
    },
    {
      "epoch": 0.023833978562587236,
      "grad_norm": 0.10179337859153748,
      "learning_rate": 0.0018481071557298044,
      "loss": 1.7638864517211914,
      "step": 64620,
      "train_bpb": 0.6253729867226778,
      "train_bytes": 56393.3,
      "train_loss_nats_per_token": 1.763949885914535
    },
    {
      "epoch": 0.023965658002159544,
      "grad_norm": 0.13786236941814423,
      "learning_rate": 0.0018475035423249028,
      "loss": 1.7660232543945313,
      "step": 64640,
      "train_bpb": 0.6243561986641518,
      "train_bytes": 56695.4875,
      "train_loss_nats_per_token": 1.7660139559860533
    },
    {
      "epoch": 0.02409733744173185,
      "grad_norm": 0.1399010866880417,
      "learning_rate": 0.0018468998694504157,
      "loss": 1.7764543533325194,
      "step": 64660,
      "train_bpb": 0.6305284641897241,
      "train_bytes": 56538.6125,
      "train_loss_nats_per_token": 1.7765292885305772
    },
    {
      "epoch": 0.024229016881304154,
      "grad_norm": 0.13358908891677856,
      "learning_rate": 0.0018462961372096517,
      "loss": 1.7741411209106446,
      "step": 64680,
      "train_bpb": 0.6303620359897038,
      "train_bytes": 55695.14375,
      "train_loss_nats_per_token": 1.7748535402419856
    },
    {
      "epoch": 0.02436069632087646,
      "grad_norm": 0.1331121027469635,
      "learning_rate": 0.001845692345705929,
      "loss": 1.775316047668457,
      "step": 64700,
      "train_bpb": 0.6300734363180864,
      "train_bytes": 56582.7625,
      "train_loss_nats_per_token": 1.7757031462974853
    },
    {
      "epoch": 0.024492375760448763,
      "grad_norm": 0.12405195087194443,
      "learning_rate": 0.0018450884950425776,
      "loss": 1.779593276977539,
      "step": 64720,
      "train_bpb": 0.6295954818215438,
      "train_bytes": 57312.25625,
      "train_loss_nats_per_token": 1.7798796339537883
    },
    {
      "epoch": 0.024624055200021068,
      "grad_norm": 0.09632065147161484,
      "learning_rate": 0.0018444845853229364,
      "loss": 1.7696037292480469,
      "step": 64740,
      "train_bpb": 0.6249528142232548,
      "train_bytes": 57664.53125,
      "train_loss_nats_per_token": 1.7700729132273978
    },
    {
      "epoch": 0.024755734639593373,
      "grad_norm": 0.09016328305006027,
      "learning_rate": 0.0018438806166503547,
      "loss": 1.7682765960693358,
      "step": 64760,
      "train_bpb": 0.6242668477807847,
      "train_bytes": 56850.3,
      "train_loss_nats_per_token": 1.7689991943439367
    },
    {
      "epoch": 0.024887414079165678,
      "grad_norm": 0.19884686172008514,
      "learning_rate": 0.001843276589128192,
      "loss": 1.773239517211914,
      "step": 64780,
      "train_bpb": 0.627552859319553,
      "train_bytes": 56341.95625,
      "train_loss_nats_per_token": 1.7733367215224085
    },
    {
      "epoch": 0.025019093518737986,
      "grad_norm": 0.1374020129442215,
      "learning_rate": 0.0018426725028598179,
      "loss": 1.773383331298828,
      "step": 64800,
      "train_bpb": 0.6264045842882233,
      "train_bytes": 56400.7125,
      "train_loss_nats_per_token": 1.773662410248799
    },
    {
      "epoch": 0.02515077295831029,
      "grad_norm": 0.09594950079917908,
      "learning_rate": 0.0018420683579486116,
      "loss": 1.7746393203735351,
      "step": 64820,
      "train_bpb": 0.6282746280711515,
      "train_bytes": 56189.0125,
      "train_loss_nats_per_token": 1.7745663126184987
    },
    {
      "epoch": 0.025282452397882595,
      "grad_norm": 0.13217893242835999,
      "learning_rate": 0.0018414641544979634,
      "loss": 1.7749906539916993,
      "step": 64840,
      "train_bpb": 0.627326922284183,
      "train_bytes": 56625.1125,
      "train_loss_nats_per_token": 1.7755441052309835
    },
    {
      "epoch": 0.0254141318374549,
      "grad_norm": 0.10296190530061722,
      "learning_rate": 0.0018408598926112723,
      "loss": 1.7739843368530273,
      "step": 64860,
      "train_bpb": 0.6265105961400614,
      "train_bytes": 56678.225,
      "train_loss_nats_per_token": 1.7742787289705604
    },
    {
      "epoch": 0.025545811277027205,
      "grad_norm": 0.1072382852435112,
      "learning_rate": 0.0018402555723919481,
      "loss": 1.7648448944091797,
      "step": 64880,
      "train_bpb": 0.6228117766591573,
      "train_bytes": 56907.21875,
      "train_loss_nats_per_token": 1.7657096603425722
    },
    {
      "epoch": 0.02567749071659951,
      "grad_norm": 0.1840733140707016,
      "learning_rate": 0.00183965119394341,
      "loss": 1.768271255493164,
      "step": 64900,
      "train_bpb": 0.6259557658224123,
      "train_bytes": 56491.5875,
      "train_loss_nats_per_token": 1.7686581726309158
    },
    {
      "epoch": 0.025809170156171814,
      "grad_norm": 0.1546754390001297,
      "learning_rate": 0.0018390467573690882,
      "loss": 1.7768306732177734,
      "step": 64920,
      "train_bpb": 0.6278240144087043,
      "train_bytes": 56663.7,
      "train_loss_nats_per_token": 1.7772119911764408
    },
    {
      "epoch": 0.02594084959574412,
      "grad_norm": 0.13413704931735992,
      "learning_rate": 0.0018384422627724221,
      "loss": 1.751815414428711,
      "step": 64940,
      "train_bpb": 0.6198328666684237,
      "train_bytes": 56544.1625,
      "train_loss_nats_per_token": 1.751609169015438
    },
    {
      "epoch": 0.026072529035316427,
      "grad_norm": 0.15267334878444672,
      "learning_rate": 0.0018378377102568602,
      "loss": 1.7583040237426757,
      "step": 64960,
      "train_bpb": 0.6227295948418429,
      "train_bytes": 56307.76875,
      "train_loss_nats_per_token": 1.7584230922792181
    },
    {
      "epoch": 0.026204208474888732,
      "grad_norm": 0.10096501559019089,
      "learning_rate": 0.0018372330999258631,
      "loss": 1.7627973556518555,
      "step": 64980,
      "train_bpb": 0.625957652171307,
      "train_bytes": 56465.81875,
      "train_loss_nats_per_token": 1.7628369077537722
    },
    {
      "epoch": 0.026335887914461037,
      "grad_norm": 0.16672393679618835,
      "learning_rate": 0.0018366284318828993,
      "loss": 1.7659345626831056,
      "step": 65000,
      "train_bpb": 0.6270829389790146,
      "train_bytes": 56406.4875,
      "train_loss_nats_per_token": 1.7657671096486274
    },
    {
      "epoch": 0.026335887914461037,
      "eval_loss": 1.6979228258132935,
      "eval_runtime": 76.4887,
      "eval_samples_per_second": 13.074,
      "eval_steps_per_second": 13.074,
      "step": 65000,
      "train_bpb": 0.598578813343806,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.7045206433911992
    },
    {
      "epoch": 0.02646756735403334,
      "grad_norm": 0.1449296921491623,
      "learning_rate": 0.0018360237062314483,
      "loss": 1.764761734008789,
      "step": 65020,
      "train_bpb": 0.6247281023463409,
      "train_bytes": 56860.3625,
      "train_loss_nats_per_token": 1.7651574192646935
    },
    {
      "epoch": 0.026599246793605646,
      "grad_norm": 0.1469276249408722,
      "learning_rate": 0.0018354189230749987,
      "loss": 1.7567232131958008,
      "step": 65040,
      "train_bpb": 0.6232288213494028,
      "train_bytes": 56096.725,
      "train_loss_nats_per_token": 1.7571499905538637
    },
    {
      "epoch": 0.02673092623317795,
      "grad_norm": 0.13042078912258148,
      "learning_rate": 0.0018348140825170493,
      "loss": 1.7605409622192383,
      "step": 65060,
      "train_bpb": 0.6219806622964095,
      "train_bytes": 56453.69375,
      "train_loss_nats_per_token": 1.760820871034311
    },
    {
      "epoch": 0.026862605672750256,
      "grad_norm": 0.12299855798482895,
      "learning_rate": 0.0018342091846611092,
      "loss": 1.7583702087402344,
      "step": 65080,
      "train_bpb": 0.6214079298422176,
      "train_bytes": 56648.925,
      "train_loss_nats_per_token": 1.758430203163433
    },
    {
      "epoch": 0.02699428511232256,
      "grad_norm": 0.13591144979000092,
      "learning_rate": 0.0018336042296106965,
      "loss": 1.756865882873535,
      "step": 65100,
      "train_bpb": 0.6240011894907073,
      "train_bytes": 56710.53125,
      "train_loss_nats_per_token": 1.7572049822968405
    },
    {
      "epoch": 0.027125964551894866,
      "grad_norm": 0.13546569645404816,
      "learning_rate": 0.0018329992174693397,
      "loss": 1.7548334121704101,
      "step": 65120,
      "train_bpb": 0.6217033584565105,
      "train_bytes": 56541.9875,
      "train_loss_nats_per_token": 1.7548741675530646
    },
    {
      "epoch": 0.027257643991467174,
      "grad_norm": 0.11460920423269272,
      "learning_rate": 0.0018323941483405764,
      "loss": 1.7566987991333007,
      "step": 65140,
      "train_bpb": 0.623240116396315,
      "train_bytes": 56223.91875,
      "train_loss_nats_per_token": 1.7571375211013367
    },
    {
      "epoch": 0.02738932343103948,
      "grad_norm": 0.1674625128507614,
      "learning_rate": 0.0018317890223279545,
      "loss": 1.7653774261474608,
      "step": 65160,
      "train_bpb": 0.6224772056010723,
      "train_bytes": 56978.26875,
      "train_loss_nats_per_token": 1.7652682995408318
    },
    {
      "epoch": 0.027521002870611783,
      "grad_norm": 0.12123852968215942,
      "learning_rate": 0.0018311838395350317,
      "loss": 1.759901237487793,
      "step": 65180,
      "train_bpb": 0.6229555358726437,
      "train_bytes": 56553.6125,
      "train_loss_nats_per_token": 1.7602923540342208
    },
    {
      "epoch": 0.027652682310184088,
      "grad_norm": 0.13497740030288696,
      "learning_rate": 0.001830578600065375,
      "loss": 1.7630739212036133,
      "step": 65200,
      "train_bpb": 0.6244375744003833,
      "train_bytes": 56943.40625,
      "train_loss_nats_per_token": 1.7631608077515841
    },
    {
      "epoch": 0.027784361749756393,
      "grad_norm": 0.22097095847129822,
      "learning_rate": 0.0018299733040225614,
      "loss": 1.7624841690063477,
      "step": 65220,
      "train_bpb": 0.6237964124753153,
      "train_bytes": 56626.06875,
      "train_loss_nats_per_token": 1.7628974813268279
    },
    {
      "epoch": 0.027916041189328698,
      "grad_norm": 0.09252142906188965,
      "learning_rate": 0.001829367951510177,
      "loss": 1.7568052291870118,
      "step": 65240,
      "train_bpb": 0.6207583997114864,
      "train_bytes": 56470.63125,
      "train_loss_nats_per_token": 1.7570648183980384
    },
    {
      "epoch": 0.028047720628901002,
      "grad_norm": 0.11635258793830872,
      "learning_rate": 0.0018287625426318193,
      "loss": 1.7659467697143554,
      "step": 65260,
      "train_bpb": 0.6220320114104386,
      "train_bytes": 57183.0625,
      "train_loss_nats_per_token": 1.7658361013661732
    },
    {
      "epoch": 0.028179400068473307,
      "grad_norm": 0.1255776286125183,
      "learning_rate": 0.0018281570774910927,
      "loss": 1.761505699157715,
      "step": 65280,
      "train_bpb": 0.6234980107049951,
      "train_bytes": 56402.23125,
      "train_loss_nats_per_token": 1.7619832968461497
    },
    {
      "epoch": 0.028311079508045615,
      "grad_norm": 0.1569579839706421,
      "learning_rate": 0.0018275515561916131,
      "loss": 1.7519027709960937,
      "step": 65300,
      "train_bpb": 0.6201181769598428,
      "train_bytes": 56713.60625,
      "train_loss_nats_per_token": 1.7524664574916329
    },
    {
      "epoch": 0.02844275894761792,
      "grad_norm": 0.16546931862831116,
      "learning_rate": 0.0018269459788370064,
      "loss": 1.752970314025879,
      "step": 65320,
      "train_bpb": 0.6216933645845436,
      "train_bytes": 56975.85,
      "train_loss_nats_per_token": 1.7533753893797908
    },
    {
      "epoch": 0.028574438387190225,
      "grad_norm": 0.15470819175243378,
      "learning_rate": 0.0018263403455309071,
      "loss": 1.7648845672607423,
      "step": 65340,
      "train_bpb": 0.6238857778038319,
      "train_bytes": 56582.70625,
      "train_loss_nats_per_token": 1.7648973335509013
    },
    {
      "epoch": 0.02870611782676253,
      "grad_norm": 0.11253272742033005,
      "learning_rate": 0.001825734656376959,
      "loss": 1.7625997543334961,
      "step": 65360,
      "train_bpb": 0.6206929961367598,
      "train_bytes": 56626.6875,
      "train_loss_nats_per_token": 1.7628366917120752
    },
    {
      "epoch": 0.028837797266334834,
      "grad_norm": 0.14017771184444427,
      "learning_rate": 0.001825128911478816,
      "loss": 1.756396484375,
      "step": 65380,
      "train_bpb": 0.6210338866284082,
      "train_bytes": 56881.60625,
      "train_loss_nats_per_token": 1.7565847754406243
    },
    {
      "epoch": 0.02896947670590714,
      "grad_norm": 0.15542899072170258,
      "learning_rate": 0.0018245231109401418,
      "loss": 1.7564102172851563,
      "step": 65400,
      "train_bpb": 0.6220924004438635,
      "train_bytes": 56891.90625,
      "train_loss_nats_per_token": 1.7571399155030485
    },
    {
      "epoch": 0.029101156145479444,
      "grad_norm": 0.12630757689476013,
      "learning_rate": 0.0018239172548646094,
      "loss": 1.7543754577636719,
      "step": 65420,
      "train_bpb": 0.6205444275655627,
      "train_bytes": 56575.025,
      "train_loss_nats_per_token": 1.7547273857271557
    },
    {
      "epoch": 0.02923283558505175,
      "grad_norm": 0.14113236963748932,
      "learning_rate": 0.001823311343355901,
      "loss": 1.7495370864868165,
      "step": 65440,
      "train_bpb": 0.6197702208789982,
      "train_bytes": 56393.30625,
      "train_loss_nats_per_token": 1.749519624237272
    },
    {
      "epoch": 0.029364515024624054,
      "grad_norm": 0.14338169991970062,
      "learning_rate": 0.0018227053765177085,
      "loss": 1.7608238220214845,
      "step": 65460,
      "train_bpb": 0.6218726164086682,
      "train_bytes": 56771.95625,
      "train_loss_nats_per_token": 1.76111671116096
    },
    {
      "epoch": 0.029496194464196362,
      "grad_norm": 0.15473738312721252,
      "learning_rate": 0.001822099354453733,
      "loss": 1.753028106689453,
      "step": 65480,
      "train_bpb": 0.6177273831103326,
      "train_bytes": 56933.35,
      "train_loss_nats_per_token": 1.7530954683494138
    },
    {
      "epoch": 0.029627873903768667,
      "grad_norm": 0.09051491320133209,
      "learning_rate": 0.001821493277267686,
      "loss": 1.7411779403686523,
      "step": 65500,
      "train_bpb": 0.6167592523814037,
      "train_bytes": 56788.9625,
      "train_loss_nats_per_token": 1.741119565437549
    },
    {
      "epoch": 0.02975955334334097,
      "grad_norm": 0.1601225584745407,
      "learning_rate": 0.001820887145063288,
      "loss": 1.7593490600585937,
      "step": 65520,
      "train_bpb": 0.6216045638327056,
      "train_bytes": 56706.16875,
      "train_loss_nats_per_token": 1.7596901283170714
    },
    {
      "epoch": 0.029891232782913276,
      "grad_norm": 0.11241374164819717,
      "learning_rate": 0.0018202809579442672,
      "loss": 1.753786849975586,
      "step": 65540,
      "train_bpb": 0.6157106729157668,
      "train_bytes": 57101.59375,
      "train_loss_nats_per_token": 1.7541465125073612
    },
    {
      "epoch": 0.03002291222248558,
      "grad_norm": 0.09816886484622955,
      "learning_rate": 0.0018196747160143642,
      "loss": 1.7503389358520507,
      "step": 65560,
      "train_bpb": 0.6207686319091456,
      "train_bytes": 56139.6125,
      "train_loss_nats_per_token": 1.7505196435903085
    },
    {
      "epoch": 0.030154591662057886,
      "grad_norm": 0.12562932074069977,
      "learning_rate": 0.0018190684193773266,
      "loss": 1.7505247116088867,
      "step": 65580,
      "train_bpb": 0.6198007644397356,
      "train_bytes": 56376.78125,
      "train_loss_nats_per_token": 1.7505287743506492
    },
    {
      "epoch": 0.03028627110163019,
      "grad_norm": 0.1265789270401001,
      "learning_rate": 0.001818462068136913,
      "loss": 1.7560331344604492,
      "step": 65600,
      "train_bpb": 0.6198455169552516,
      "train_bytes": 56652.51875,
      "train_loss_nats_per_token": 1.7561776411340144
    },
    {
      "epoch": 0.030417950541202495,
      "grad_norm": 0.1257970631122589,
      "learning_rate": 0.0018178556623968895,
      "loss": 1.7483354568481446,
      "step": 65620,
      "train_bpb": 0.6203455526110292,
      "train_bytes": 56062.5625,
      "train_loss_nats_per_token": 1.748523833272303
    },
    {
      "epoch": 0.030549629980774803,
      "grad_norm": 0.11794394999742508,
      "learning_rate": 0.0018172492022610332,
      "loss": 1.7488901138305664,
      "step": 65640,
      "train_bpb": 0.6185016746549021,
      "train_bytes": 56726.2875,
      "train_loss_nats_per_token": 1.7492824802816638
    },
    {
      "epoch": 0.030681309420347108,
      "grad_norm": 0.11422976106405258,
      "learning_rate": 0.0018166426878331295,
      "loss": 1.7469724655151366,
      "step": 65660,
      "train_bpb": 0.617713472483144,
      "train_bytes": 56921.075,
      "train_loss_nats_per_token": 1.7468044787614034
    },
    {
      "epoch": 0.030812988859919413,
      "grad_norm": 0.14921388030052185,
      "learning_rate": 0.0018160361192169744,
      "loss": 1.7458086013793945,
      "step": 65680,
      "train_bpb": 0.6179938197094433,
      "train_bytes": 56531.49375,
      "train_loss_nats_per_token": 1.7457549452421053
    },
    {
      "epoch": 0.030944668299491718,
      "grad_norm": 0.11006993055343628,
      "learning_rate": 0.0018154294965163713,
      "loss": 1.7656284332275392,
      "step": 65700,
      "train_bpb": 0.6263072003851187,
      "train_bytes": 56209.525,
      "train_loss_nats_per_token": 1.7654404403499142
    },
    {
      "epoch": 0.031076347739064022,
      "grad_norm": 0.1039053276181221,
      "learning_rate": 0.0018148228198351345,
      "loss": 1.7642967224121093,
      "step": 65720,
      "train_bpb": 0.6230630903543636,
      "train_bytes": 56712.09375,
      "train_loss_nats_per_token": 1.7644624194393979
    },
    {
      "epoch": 0.031208027178636327,
      "grad_norm": 0.09604157507419586,
      "learning_rate": 0.001814216089277086,
      "loss": 1.7636005401611328,
      "step": 65740,
      "train_bpb": 0.6265372805966687,
      "train_bytes": 56490.51875,
      "train_loss_nats_per_token": 1.7637435847300076
    },
    {
      "epoch": 0.03133970661820863,
      "grad_norm": 0.10771388560533524,
      "learning_rate": 0.0018136093049460592,
      "loss": 1.7675922393798829,
      "step": 65760,
      "train_bpb": 0.6237371774289967,
      "train_bytes": 56715.9875,
      "train_loss_nats_per_token": 1.7682601492006724
    },
    {
      "epoch": 0.03147138605778094,
      "grad_norm": 0.11332212388515472,
      "learning_rate": 0.0018130024669458945,
      "loss": 1.7697988510131837,
      "step": 65780,
      "train_bpb": 0.6265499739289517,
      "train_bytes": 56458.05625,
      "train_loss_nats_per_token": 1.7705935118330733
    },
    {
      "epoch": 0.03160306549735324,
      "grad_norm": 0.12479856610298157,
      "learning_rate": 0.001812395575380442,
      "loss": 1.7784015655517578,
      "step": 65800,
      "train_bpb": 0.6278925316064734,
      "train_bytes": 57141.40625,
      "train_loss_nats_per_token": 1.7788287704230288
    },
    {
      "epoch": 0.031734744936925546,
      "grad_norm": 0.12110721319913864,
      "learning_rate": 0.001811788630353562,
      "loss": 1.766665267944336,
      "step": 65820,
      "train_bpb": 0.6230609577966217,
      "train_bytes": 56760.875,
      "train_loss_nats_per_token": 1.7665642790693865
    },
    {
      "epoch": 0.03186642437649785,
      "grad_norm": 0.19775579869747162,
      "learning_rate": 0.0018111816319691235,
      "loss": 1.7672605514526367,
      "step": 65840,
      "train_bpb": 0.6236691055741445,
      "train_bytes": 56890.26875,
      "train_loss_nats_per_token": 1.7675389585016865
    },
    {
      "epoch": 0.031998103816070156,
      "grad_norm": 0.14023685455322266,
      "learning_rate": 0.0018105745803310034,
      "loss": 1.7728460311889649,
      "step": 65860,
      "train_bpb": 0.625798046455789,
      "train_bytes": 56530.7375,
      "train_loss_nats_per_token": 1.7732510383569562
    },
    {
      "epoch": 0.03212978325564246,
      "grad_norm": 0.15388800203800201,
      "learning_rate": 0.0018099674755430901,
      "loss": 1.7743217468261718,
      "step": 65880,
      "train_bpb": 0.6279809614975778,
      "train_bytes": 56400.76875,
      "train_loss_nats_per_token": 1.7744856895916747
    },
    {
      "epoch": 0.03226146269521477,
      "grad_norm": 0.12709969282150269,
      "learning_rate": 0.0018093603177092785,
      "loss": 1.7691682815551757,
      "step": 65900,
      "train_bpb": 0.625270396575003,
      "train_bytes": 56467.83125,
      "train_loss_nats_per_token": 1.7691966060706008
    },
    {
      "epoch": 0.03239314213478708,
      "grad_norm": 0.15010589361190796,
      "learning_rate": 0.0018087531069334744,
      "loss": 1.7673063278198242,
      "step": 65920,
      "train_bpb": 0.6275227008246463,
      "train_bytes": 55934.75625,
      "train_loss_nats_per_token": 1.7674214690697925
    },
    {
      "epoch": 0.03252482157435938,
      "grad_norm": 0.1336129754781723,
      "learning_rate": 0.0018081458433195918,
      "loss": 1.7614648818969727,
      "step": 65940,
      "train_bpb": 0.6230878764601391,
      "train_bytes": 56558.85,
      "train_loss_nats_per_token": 1.7619812037638864
    },
    {
      "epoch": 0.03265650101393169,
      "grad_norm": 0.13220910727977753,
      "learning_rate": 0.0018075385269715539,
      "loss": 1.7579992294311524,
      "step": 65960,
      "train_bpb": 0.6211978255997455,
      "train_bytes": 56919.875,
      "train_loss_nats_per_token": 1.7582092947413452
    },
    {
      "epoch": 0.03278818045350399,
      "grad_norm": 0.1266823709011078,
      "learning_rate": 0.0018069311579932933,
      "loss": 1.7622156143188477,
      "step": 65980,
      "train_bpb": 0.6183411872327591,
      "train_bytes": 57334.23125,
      "train_loss_nats_per_token": 1.7620647453225977
    },
    {
      "epoch": 0.032919859893076296,
      "grad_norm": 0.12702882289886475,
      "learning_rate": 0.0018063237364887518,
      "loss": 1.7586977005004882,
      "step": 66000,
      "train_bpb": 0.6207428124755741,
      "train_bytes": 56569.9375,
      "train_loss_nats_per_token": 1.7588783488443647
    },
    {
      "epoch": 0.032919859893076296,
      "eval_loss": 1.7008097171783447,
      "eval_runtime": 81.7538,
      "eval_samples_per_second": 12.232,
      "eval_steps_per_second": 12.232,
      "step": 66000,
      "train_bpb": 0.5996274187923551,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.7075066656057853
    },
    {
      "epoch": 0.0330515393326486,
      "grad_norm": 0.15249976515769958,
      "learning_rate": 0.0018057162625618789,
      "loss": 1.7588695526123046,
      "step": 66020,
      "train_bpb": 0.624592722349535,
      "train_bytes": 56125.48125,
      "train_loss_nats_per_token": 1.7591930810030552
    },
    {
      "epoch": 0.033183218772220906,
      "grad_norm": 0.1460212767124176,
      "learning_rate": 0.001805108736316634,
      "loss": 1.766263198852539,
      "step": 66040,
      "train_bpb": 0.6235381914002261,
      "train_bytes": 56595.71875,
      "train_loss_nats_per_token": 1.7665981634587216
    },
    {
      "epoch": 0.03331489821179321,
      "grad_norm": 0.12443271279335022,
      "learning_rate": 0.0018045011578569858,
      "loss": 1.7620964050292969,
      "step": 66060,
      "train_bpb": 0.6233628803926721,
      "train_bytes": 56528.84375,
      "train_loss_nats_per_token": 1.76215176147603
    },
    {
      "epoch": 0.033446577651365515,
      "grad_norm": 0.163056880235672,
      "learning_rate": 0.0018038935272869114,
      "loss": 1.7523185729980468,
      "step": 66080,
      "train_bpb": 0.6186325096948139,
      "train_bytes": 56296.86875,
      "train_loss_nats_per_token": 1.7524383691705372
    },
    {
      "epoch": 0.03357825709093782,
      "grad_norm": 0.12493759393692017,
      "learning_rate": 0.0018032858447103962,
      "loss": 1.7620996475219726,
      "step": 66100,
      "train_bpb": 0.6225525785862202,
      "train_bytes": 56928.84375,
      "train_loss_nats_per_token": 1.7628248432181859
    },
    {
      "epoch": 0.033709936530510125,
      "grad_norm": 0.14240586757659912,
      "learning_rate": 0.001802678110231436,
      "loss": 1.761892318725586,
      "step": 66120,
      "train_bpb": 0.6252467818820139,
      "train_bytes": 56321.4375,
      "train_loss_nats_per_token": 1.7624442429414948
    },
    {
      "epoch": 0.03384161597008243,
      "grad_norm": 0.09770048409700394,
      "learning_rate": 0.0018020703239540342,
      "loss": 1.7538791656494142,
      "step": 66140,
      "train_bpb": 0.6191778857642194,
      "train_bytes": 56488.375,
      "train_loss_nats_per_token": 1.7544684337652827
    },
    {
      "epoch": 0.033973295409654734,
      "grad_norm": 0.1390150487422943,
      "learning_rate": 0.001801462485982204,
      "loss": 1.7467054367065429,
      "step": 66160,
      "train_bpb": 0.6174334814935363,
      "train_bytes": 56470.59375,
      "train_loss_nats_per_token": 1.7471446982719476
    },
    {
      "epoch": 0.03410497484922704,
      "grad_norm": 0.1429586410522461,
      "learning_rate": 0.0018008545964199668,
      "loss": 1.752676010131836,
      "step": 66180,
      "train_bpb": 0.6216720224222492,
      "train_bytes": 56674.525,
      "train_loss_nats_per_token": 1.7529384301993816
    },
    {
      "epoch": 0.034236654288799344,
      "grad_norm": 0.18586716055870056,
      "learning_rate": 0.0018002466553713528,
      "loss": 1.7621088027954102,
      "step": 66200,
      "train_bpb": 0.6217908048591073,
      "train_bytes": 56513.40625,
      "train_loss_nats_per_token": 1.7625979740861133
    },
    {
      "epoch": 0.03436833372837165,
      "grad_norm": 0.12195756286382675,
      "learning_rate": 0.001799638662940402,
      "loss": 1.7651048660278321,
      "step": 66220,
      "train_bpb": 0.6232921951197159,
      "train_bytes": 56688.8625,
      "train_loss_nats_per_token": 1.7652163542048624
    },
    {
      "epoch": 0.03450001316794396,
      "grad_norm": 0.10573576390743256,
      "learning_rate": 0.0017990306192311612,
      "loss": 1.752359390258789,
      "step": 66240,
      "train_bpb": 0.6188294960333599,
      "train_bytes": 56660.86875,
      "train_loss_nats_per_token": 1.7528776344995158
    },
    {
      "epoch": 0.034631692607516265,
      "grad_norm": 0.13356566429138184,
      "learning_rate": 0.0017984225243476886,
      "loss": 1.7536745071411133,
      "step": 66260,
      "train_bpb": 0.6195871621221298,
      "train_bytes": 57073.275,
      "train_loss_nats_per_token": 1.7535562739748414
    },
    {
      "epoch": 0.03476337204708857,
      "grad_norm": 0.09456772357225418,
      "learning_rate": 0.0017978143783940483,
      "loss": 1.7552976608276367,
      "step": 66280,
      "train_bpb": 0.6188659969874808,
      "train_bytes": 56922.2625,
      "train_loss_nats_per_token": 1.755385950676788
    },
    {
      "epoch": 0.034895051486660875,
      "grad_norm": 0.0951802209019661,
      "learning_rate": 0.001797206181474315,
      "loss": 1.7546072006225586,
      "step": 66300,
      "train_bpb": 0.6192131836576023,
      "train_bytes": 56918.5375,
      "train_loss_nats_per_token": 1.7552833623320838
    },
    {
      "epoch": 0.03502673092623318,
      "grad_norm": 0.13678394258022308,
      "learning_rate": 0.0017965979336925725,
      "loss": 1.7563016891479493,
      "step": 66320,
      "train_bpb": 0.6196448861316511,
      "train_bytes": 56496.45,
      "train_loss_nats_per_token": 1.7565654665860242
    },
    {
      "epoch": 0.035158410365805484,
      "grad_norm": 0.16776540875434875,
      "learning_rate": 0.0017959896351529125,
      "loss": 1.7483545303344727,
      "step": 66340,
      "train_bpb": 0.6202139063161629,
      "train_bytes": 56304.725,
      "train_loss_nats_per_token": 1.748271362823276
    },
    {
      "epoch": 0.03529008980537779,
      "grad_norm": 0.1053045317530632,
      "learning_rate": 0.0017953812859594343,
      "loss": 1.7525400161743163,
      "step": 66360,
      "train_bpb": 0.6209972310402241,
      "train_bytes": 56168.525,
      "train_loss_nats_per_token": 1.752245503875135
    },
    {
      "epoch": 0.035421769244950094,
      "grad_norm": 0.1314372569322586,
      "learning_rate": 0.0017947728862162478,
      "loss": 1.7510343551635743,
      "step": 66380,
      "train_bpb": 0.6150260942727382,
      "train_bytes": 56974.29375,
      "train_loss_nats_per_token": 1.7511385719703114
    },
    {
      "epoch": 0.0355534486845224,
      "grad_norm": 0.17896795272827148,
      "learning_rate": 0.0017941644360274704,
      "loss": 1.7495370864868165,
      "step": 66400,
      "train_bpb": 0.6215426660082521,
      "train_bytes": 55688.65625,
      "train_loss_nats_per_token": 1.7493696345689318
    },
    {
      "epoch": 0.0356851281240947,
      "grad_norm": 0.11911572515964508,
      "learning_rate": 0.001793555935497229,
      "loss": 1.752629852294922,
      "step": 66420,
      "train_bpb": 0.6228956421775866,
      "train_bytes": 55803.6,
      "train_loss_nats_per_token": 1.7527154969445022
    },
    {
      "epoch": 0.03581680756366701,
      "grad_norm": 0.12834225594997406,
      "learning_rate": 0.001792947384729658,
      "loss": 1.7501964569091797,
      "step": 66440,
      "train_bpb": 0.6172468001757329,
      "train_bytes": 56913.025,
      "train_loss_nats_per_token": 1.75032659237265
    },
    {
      "epoch": 0.03594848700323931,
      "grad_norm": 0.15553921461105347,
      "learning_rate": 0.0017923387838289012,
      "loss": 1.749886703491211,
      "step": 66460,
      "train_bpb": 0.618462011805435,
      "train_bytes": 56846.425,
      "train_loss_nats_per_token": 1.7499113029521385
    },
    {
      "epoch": 0.03608016644281162,
      "grad_norm": 0.09298575669527054,
      "learning_rate": 0.001791730132899111,
      "loss": 1.7404754638671875,
      "step": 66480,
      "train_bpb": 0.614989341188935,
      "train_bytes": 56388.86875,
      "train_loss_nats_per_token": 1.7407633360958215
    },
    {
      "epoch": 0.03621184588238392,
      "grad_norm": 0.11655056476593018,
      "learning_rate": 0.0017911214320444479,
      "loss": 1.7440576553344727,
      "step": 66500,
      "train_bpb": 0.6155722123024193,
      "train_bytes": 56415.15,
      "train_loss_nats_per_token": 1.74406598547166
    },
    {
      "epoch": 0.03634352532195623,
      "grad_norm": 0.14331884682178497,
      "learning_rate": 0.001790512681369081,
      "loss": 1.7484418869018554,
      "step": 66520,
      "train_bpb": 0.6180610198459969,
      "train_bytes": 56891.76875,
      "train_loss_nats_per_token": 1.7488530196279009
    },
    {
      "epoch": 0.03647520476152853,
      "grad_norm": 0.12300813943147659,
      "learning_rate": 0.0017899038809771881,
      "loss": 1.743328857421875,
      "step": 66540,
      "train_bpb": 0.6170337864228076,
      "train_bytes": 56102.16875,
      "train_loss_nats_per_token": 1.7437316255454356
    },
    {
      "epoch": 0.036606884201100844,
      "grad_norm": 0.13815417885780334,
      "learning_rate": 0.001789295030972956,
      "loss": 1.7601776123046875,
      "step": 66560,
      "train_bpb": 0.6194021385093481,
      "train_bytes": 57203.05625,
      "train_loss_nats_per_token": 1.7606276156888827
    },
    {
      "epoch": 0.03673856364067315,
      "grad_norm": 0.14032725989818573,
      "learning_rate": 0.0017886861314605786,
      "loss": 1.7567913055419921,
      "step": 66580,
      "train_bpb": 0.6199468199055402,
      "train_bytes": 56865.30625,
      "train_loss_nats_per_token": 1.757169024104401
    },
    {
      "epoch": 0.03687024308024545,
      "grad_norm": 0.1347580850124359,
      "learning_rate": 0.0017880771825442604,
      "loss": 1.7457803726196288,
      "step": 66600,
      "train_bpb": 0.6163708444652597,
      "train_bytes": 56751.0625,
      "train_loss_nats_per_token": 1.745700699066743
    },
    {
      "epoch": 0.03700192251981776,
      "grad_norm": 0.19887541234493256,
      "learning_rate": 0.001787468184328212,
      "loss": 1.7487092971801759,
      "step": 66620,
      "train_bpb": 0.6171008289517163,
      "train_bytes": 57321.1625,
      "train_loss_nats_per_token": 1.748885370347079
    },
    {
      "epoch": 0.03713360195939006,
      "grad_norm": 0.13883525133132935,
      "learning_rate": 0.0017868591369166534,
      "loss": 1.7465459823608398,
      "step": 66640,
      "train_bpb": 0.6149444312621742,
      "train_bytes": 56662.21875,
      "train_loss_nats_per_token": 1.7466786428687333
    },
    {
      "epoch": 0.03726528139896237,
      "grad_norm": 0.13278844952583313,
      "learning_rate": 0.0017862500404138145,
      "loss": 1.7519451141357423,
      "step": 66660,
      "train_bpb": 0.6178016851313282,
      "train_bytes": 56644.18125,
      "train_loss_nats_per_token": 1.7520183793194921
    },
    {
      "epoch": 0.03739696083853467,
      "grad_norm": 0.10344759374856949,
      "learning_rate": 0.0017856408949239312,
      "loss": 1.7436071395874024,
      "step": 66680,
      "train_bpb": 0.6184387369687167,
      "train_bytes": 56764.76875,
      "train_loss_nats_per_token": 1.7437897883646776
    },
    {
      "epoch": 0.03752864027810698,
      "grad_norm": 0.13819292187690735,
      "learning_rate": 0.001785031700551249,
      "loss": 1.7461376190185547,
      "step": 66700,
      "train_bpb": 0.6169420069801362,
      "train_bytes": 56521.2375,
      "train_loss_nats_per_token": 1.7462954495282543
    },
    {
      "epoch": 0.03766031971767928,
      "grad_norm": 0.1457120180130005,
      "learning_rate": 0.0017844224574000216,
      "loss": 1.7445207595825196,
      "step": 66720,
      "train_bpb": 0.6184810084453779,
      "train_bytes": 56164.75,
      "train_loss_nats_per_token": 1.744687667282129
    },
    {
      "epoch": 0.037791999157251586,
      "grad_norm": 0.16211169958114624,
      "learning_rate": 0.0017838131655745118,
      "loss": 1.7469778060913086,
      "step": 66740,
      "train_bpb": 0.6153671493928917,
      "train_bytes": 56843.1375,
      "train_loss_nats_per_token": 1.747055975662981
    },
    {
      "epoch": 0.03792367859682389,
      "grad_norm": 0.11522988229990005,
      "learning_rate": 0.001783203825178989,
      "loss": 1.7457515716552734,
      "step": 66760,
      "train_bpb": 0.6179757612927224,
      "train_bytes": 56477.0125,
      "train_loss_nats_per_token": 1.7462300899232337
    },
    {
      "epoch": 0.038055358036396196,
      "grad_norm": 0.11231429129838943,
      "learning_rate": 0.0017825944363177324,
      "loss": 1.7401166915893556,
      "step": 66780,
      "train_bpb": 0.6163746126463402,
      "train_bytes": 56618.8625,
      "train_loss_nats_per_token": 1.7400333568828559
    },
    {
      "epoch": 0.0381870374759685,
      "grad_norm": 0.12767936289310455,
      "learning_rate": 0.0017819849990950288,
      "loss": 1.7431921005249023,
      "step": 66800,
      "train_bpb": 0.6163695552473107,
      "train_bytes": 56765.98125,
      "train_loss_nats_per_token": 1.7433584176589085
    },
    {
      "epoch": 0.038318716915540806,
      "grad_norm": 0.1538705676794052,
      "learning_rate": 0.0017813755136151736,
      "loss": 1.7457315444946289,
      "step": 66820,
      "train_bpb": 0.6149550231911648,
      "train_bytes": 56774.54375,
      "train_loss_nats_per_token": 1.7459854995840276
    },
    {
      "epoch": 0.03845039635511311,
      "grad_norm": 0.14725592732429504,
      "learning_rate": 0.001780765979982471,
      "loss": 1.7449726104736327,
      "step": 66840,
      "train_bpb": 0.6165792861893301,
      "train_bytes": 56858.1125,
      "train_loss_nats_per_token": 1.7456903246413549
    },
    {
      "epoch": 0.038582075794685415,
      "grad_norm": 0.12791959941387177,
      "learning_rate": 0.0017801563983012314,
      "loss": 1.7406732559204101,
      "step": 66860,
      "train_bpb": 0.6161965954725481,
      "train_bytes": 56441.95,
      "train_loss_nats_per_token": 1.7407915660092583
    },
    {
      "epoch": 0.03871375523425772,
      "grad_norm": 0.11042622476816177,
      "learning_rate": 0.001779546768675776,
      "loss": 1.7495645523071288,
      "step": 66880,
      "train_bpb": 0.6205376471774464,
      "train_bytes": 56415.53125,
      "train_loss_nats_per_token": 1.749932557153842
    },
    {
      "epoch": 0.03884543467383003,
      "grad_norm": 0.1126914694905281,
      "learning_rate": 0.0017789370912104324,
      "loss": 1.7535797119140626,
      "step": 66900,
      "train_bpb": 0.6234788506550774,
      "train_bytes": 56045.0875,
      "train_loss_nats_per_token": 1.7542937700299657
    },
    {
      "epoch": 0.038977114113402336,
      "grad_norm": 0.11664934456348419,
      "learning_rate": 0.0017783273660095373,
      "loss": 1.7551143646240235,
      "step": 66920,
      "train_bpb": 0.6224607271076271,
      "train_bytes": 56493.70625,
      "train_loss_nats_per_token": 1.7551458711581926
    },
    {
      "epoch": 0.03910879355297464,
      "grad_norm": 0.1340067982673645,
      "learning_rate": 0.0017777175931774347,
      "loss": 1.7696012496948241,
      "step": 66940,
      "train_bpb": 0.6226550311718331,
      "train_bytes": 57032.6375,
      "train_loss_nats_per_token": 1.7699110833822245
    },
    {
      "epoch": 0.039240472992546946,
      "grad_norm": 0.097479909658432,
      "learning_rate": 0.0017771077728184776,
      "loss": 1.760063934326172,
      "step": 66960,
      "train_bpb": 0.6227016687373202,
      "train_bytes": 56887.34375,
      "train_loss_nats_per_token": 1.7605088174679737
    },
    {
      "epoch": 0.03937215243211925,
      "grad_norm": 0.13523787260055542,
      "learning_rate": 0.0017764979050370275,
      "loss": 1.7656450271606445,
      "step": 66980,
      "train_bpb": 0.6244439069290668,
      "train_bytes": 56544.89375,
      "train_loss_nats_per_token": 1.7657375092479677
    },
    {
      "epoch": 0.039503831871691555,
      "grad_norm": 0.13579709827899933,
      "learning_rate": 0.0017758879899374528,
      "loss": 1.7569351196289062,
      "step": 67000,
      "train_bpb": 0.6229136680140755,
      "train_bytes": 56638.6,
      "train_loss_nats_per_token": 1.7569829496245493
    },
    {
      "epoch": 0.039503831871691555,
      "eval_loss": 1.7040061950683594,
      "eval_runtime": 77.6286,
      "eval_samples_per_second": 12.882,
      "eval_steps_per_second": 12.882,
      "step": 67000,
      "train_bpb": 0.6008410639518487,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.710962657000724
    },
    {
      "epoch": 0.03963551131126386,
      "grad_norm": 0.09360861778259277,
      "learning_rate": 0.0017752780276241303,
      "loss": 1.7670032501220703,
      "step": 67020,
      "train_bpb": 0.6236583269995207,
      "train_bytes": 56976.25,
      "train_loss_nats_per_token": 1.767141132471578
    },
    {
      "epoch": 0.039767190750836165,
      "grad_norm": 0.11249420046806335,
      "learning_rate": 0.0017746680182014452,
      "loss": 1.7691740036010741,
      "step": 67040,
      "train_bpb": 0.6238294211724529,
      "train_bytes": 56993.83125,
      "train_loss_nats_per_token": 1.7694869478251816
    },
    {
      "epoch": 0.03989887019040847,
      "grad_norm": 0.09180234372615814,
      "learning_rate": 0.0017740579617737911,
      "loss": 1.7611234664916993,
      "step": 67060,
      "train_bpb": 0.6238162656175731,
      "train_bytes": 56519.2,
      "train_loss_nats_per_token": 1.7612488273112437
    },
    {
      "epoch": 0.040030549629980774,
      "grad_norm": 0.10045190155506134,
      "learning_rate": 0.0017734478584455694,
      "loss": 1.7568624496459961,
      "step": 67080,
      "train_bpb": 0.617996706069556,
      "train_bytes": 56856.35625,
      "train_loss_nats_per_token": 1.7568037119747608
    },
    {
      "epoch": 0.04016222906955308,
      "grad_norm": 0.10826700180768967,
      "learning_rate": 0.0017728377083211888,
      "loss": 1.7566032409667969,
      "step": 67100,
      "train_bpb": 0.6232963562197034,
      "train_bytes": 56566.99375,
      "train_loss_nats_per_token": 1.7564162333608198
    },
    {
      "epoch": 0.040293908509125384,
      "grad_norm": 0.10119092464447021,
      "learning_rate": 0.0017722275115050667,
      "loss": 1.757288360595703,
      "step": 67120,
      "train_bpb": 0.6215623137792535,
      "train_bytes": 56722.6375,
      "train_loss_nats_per_token": 1.7573603106623414
    },
    {
      "epoch": 0.04042558794869769,
      "grad_norm": 0.12433522194623947,
      "learning_rate": 0.0017716172681016286,
      "loss": 1.7549156188964843,
      "step": 67140,
      "train_bpb": 0.6213598291513881,
      "train_bytes": 56418.4375,
      "train_loss_nats_per_token": 1.7554326672327378
    },
    {
      "epoch": 0.040557267388269994,
      "grad_norm": 0.09853285551071167,
      "learning_rate": 0.0017710069782153081,
      "loss": 1.7516807556152343,
      "step": 67160,
      "train_bpb": 0.6189448389318585,
      "train_bytes": 56897.15,
      "train_loss_nats_per_token": 1.7522361747963928
    },
    {
      "epoch": 0.0406889468278423,
      "grad_norm": 0.13604235649108887,
      "learning_rate": 0.0017703966419505455,
      "loss": 1.7534589767456055,
      "step": 67180,
      "train_bpb": 0.6203768075883531,
      "train_bytes": 56617.00625,
      "train_loss_nats_per_token": 1.7537672887711941
    },
    {
      "epoch": 0.0408206262674146,
      "grad_norm": 0.14241044223308563,
      "learning_rate": 0.00176978625941179,
      "loss": 1.7579910278320312,
      "step": 67200,
      "train_bpb": 0.6229967646108036,
      "train_bytes": 56278.03125,
      "train_loss_nats_per_token": 1.7583335180996678
    },
    {
      "epoch": 0.04095230570698691,
      "grad_norm": 0.1128389984369278,
      "learning_rate": 0.0017691758307035,
      "loss": 1.7534194946289063,
      "step": 67220,
      "train_bpb": 0.6214707995517762,
      "train_bytes": 56374.51875,
      "train_loss_nats_per_token": 1.7535298729957822
    },
    {
      "epoch": 0.04108398514655922,
      "grad_norm": 0.13558615744113922,
      "learning_rate": 0.0017685653559301386,
      "loss": 1.7532293319702148,
      "step": 67240,
      "train_bpb": 0.6217942717983311,
      "train_bytes": 56389.73125,
      "train_loss_nats_per_token": 1.7533453112727206
    },
    {
      "epoch": 0.041215664586131524,
      "grad_norm": 0.13748236000537872,
      "learning_rate": 0.0017679548351961802,
      "loss": 1.7481494903564454,
      "step": 67260,
      "train_bpb": 0.6196174261900729,
      "train_bytes": 56398.7625,
      "train_loss_nats_per_token": 1.7480007195649503
    },
    {
      "epoch": 0.04134734402570383,
      "grad_norm": 0.1321163773536682,
      "learning_rate": 0.0017673442686061045,
      "loss": 1.746068000793457,
      "step": 67280,
      "train_bpb": 0.6181315137673794,
      "train_bytes": 56461.6625,
      "train_loss_nats_per_token": 1.746235650512362
    },
    {
      "epoch": 0.041479023465276134,
      "grad_norm": 0.13446910679340363,
      "learning_rate": 0.0017667336562644004,
      "loss": 1.760188674926758,
      "step": 67300,
      "train_bpb": 0.6209666745987066,
      "train_bytes": 56890.24375,
      "train_loss_nats_per_token": 1.7609668727931076
    },
    {
      "epoch": 0.04161070290484844,
      "grad_norm": 0.1080336719751358,
      "learning_rate": 0.0017661229982755644,
      "loss": 1.7537275314331056,
      "step": 67320,
      "train_bpb": 0.6195863320355377,
      "train_bytes": 56655.2375,
      "train_loss_nats_per_token": 1.754067396179894
    },
    {
      "epoch": 0.04174238234442074,
      "grad_norm": 0.09537260234355927,
      "learning_rate": 0.0017655122947441003,
      "loss": 1.7485750198364258,
      "step": 67340,
      "train_bpb": 0.6177257371963593,
      "train_bytes": 56513.55,
      "train_loss_nats_per_token": 1.748717448957636
    },
    {
      "epoch": 0.04187406178399305,
      "grad_norm": 0.10151863843202591,
      "learning_rate": 0.0017649015457745206,
      "loss": 1.7487253189086913,
      "step": 67360,
      "train_bpb": 0.6204957581888882,
      "train_bytes": 56609.9125,
      "train_loss_nats_per_token": 1.7484851477018268
    },
    {
      "epoch": 0.04200574122356535,
      "grad_norm": 0.09507983177900314,
      "learning_rate": 0.0017642907514713447,
      "loss": 1.7460172653198243,
      "step": 67380,
      "train_bpb": 0.6169298622778249,
      "train_bytes": 56561.5875,
      "train_loss_nats_per_token": 1.7465896762049213
    },
    {
      "epoch": 0.04213742066313766,
      "grad_norm": 0.13690172135829926,
      "learning_rate": 0.0017636799119391002,
      "loss": 1.756070899963379,
      "step": 67400,
      "train_bpb": 0.6213987968903883,
      "train_bytes": 56845.0,
      "train_loss_nats_per_token": 1.7562377563249345
    },
    {
      "epoch": 0.04226910010270996,
      "grad_norm": 0.15823432803153992,
      "learning_rate": 0.0017630690272823224,
      "loss": 1.740869140625,
      "step": 67420,
      "train_bpb": 0.6151957104794983,
      "train_bytes": 56517.7375,
      "train_loss_nats_per_token": 1.7411604824234483
    },
    {
      "epoch": 0.04240077954228227,
      "grad_norm": 0.12306880205869675,
      "learning_rate": 0.0017624580976055544,
      "loss": 1.741773796081543,
      "step": 67440,
      "train_bpb": 0.6158144340485345,
      "train_bytes": 56512.61875,
      "train_loss_nats_per_token": 1.7420120230476879
    },
    {
      "epoch": 0.04253245898185457,
      "grad_norm": 0.11631748080253601,
      "learning_rate": 0.0017618471230133467,
      "loss": 1.7460275650024415,
      "step": 67460,
      "train_bpb": 0.6177208805966827,
      "train_bytes": 56481.4,
      "train_loss_nats_per_token": 1.7464735595677119
    },
    {
      "epoch": 0.04266413842142688,
      "grad_norm": 0.09470163285732269,
      "learning_rate": 0.001761236103610258,
      "loss": 1.7367265701293946,
      "step": 67480,
      "train_bpb": 0.6134675144148757,
      "train_bytes": 56726.90625,
      "train_loss_nats_per_token": 1.7370055763316743
    },
    {
      "epoch": 0.04279581786099918,
      "grad_norm": 0.12938092648983002,
      "learning_rate": 0.0017606250395008538,
      "loss": 1.741774559020996,
      "step": 67500,
      "train_bpb": 0.6171220087229113,
      "train_bytes": 56150.19375,
      "train_loss_nats_per_token": 1.7420209388436092
    },
    {
      "epoch": 0.042927497300571486,
      "grad_norm": 0.11507653445005417,
      "learning_rate": 0.0017600139307897083,
      "loss": 1.7440778732299804,
      "step": 67520,
      "train_bpb": 0.6183366422768851,
      "train_bytes": 56630.9875,
      "train_loss_nats_per_token": 1.7443673502022963
    },
    {
      "epoch": 0.04305917674014379,
      "grad_norm": 0.17593999207019806,
      "learning_rate": 0.0017594027775814022,
      "loss": 1.7396682739257812,
      "step": 67540,
      "train_bpb": 0.6140619992632864,
      "train_bytes": 56857.375,
      "train_loss_nats_per_token": 1.7397194195175951
    },
    {
      "epoch": 0.043190856179716096,
      "grad_norm": 0.1112256571650505,
      "learning_rate": 0.0017587915799805256,
      "loss": 1.735092544555664,
      "step": 67560,
      "train_bpb": 0.6165102860668541,
      "train_bytes": 56652.1,
      "train_loss_nats_per_token": 1.7354498519908024
    },
    {
      "epoch": 0.04332253561928841,
      "grad_norm": 0.14560790359973907,
      "learning_rate": 0.0017581803380916738,
      "loss": 1.7365510940551758,
      "step": 67580,
      "train_bpb": 0.6128233785403452,
      "train_bytes": 56718.6,
      "train_loss_nats_per_token": 1.7368788037345353
    },
    {
      "epoch": 0.04345421505886071,
      "grad_norm": 0.10159141570329666,
      "learning_rate": 0.0017575690520194517,
      "loss": 1.7316078186035155,
      "step": 67600,
      "train_bpb": 0.6124350196610958,
      "train_bytes": 56614.93125,
      "train_loss_nats_per_token": 1.7315756374548232
    },
    {
      "epoch": 0.04358589449843302,
      "grad_norm": 0.11562483012676239,
      "learning_rate": 0.0017569577218684703,
      "loss": 1.7388174057006835,
      "step": 67620,
      "train_bpb": 0.6151768976166229,
      "train_bytes": 56390.55625,
      "train_loss_nats_per_token": 1.7388074148680368
    },
    {
      "epoch": 0.04371757393800532,
      "grad_norm": 0.11210903525352478,
      "learning_rate": 0.0017563463477433493,
      "loss": 1.7354507446289062,
      "step": 67640,
      "train_bpb": 0.6152106429109979,
      "train_bytes": 56700.79375,
      "train_loss_nats_per_token": 1.735841718937821
    },
    {
      "epoch": 0.04384925337757763,
      "grad_norm": 0.1369801014661789,
      "learning_rate": 0.001755734929748716,
      "loss": 1.7322216033935547,
      "step": 67660,
      "train_bpb": 0.6116273896536198,
      "train_bytes": 57202.11875,
      "train_loss_nats_per_token": 1.7322703058011828
    },
    {
      "epoch": 0.04398093281714993,
      "grad_norm": 0.12356609106063843,
      "learning_rate": 0.0017551234679892035,
      "loss": 1.7473564147949219,
      "step": 67680,
      "train_bpb": 0.6199849624523359,
      "train_bytes": 55882.45,
      "train_loss_nats_per_token": 1.7473876058139115
    },
    {
      "epoch": 0.044112612256722236,
      "grad_norm": 0.11192473024129868,
      "learning_rate": 0.001754511962569454,
      "loss": 1.742447853088379,
      "step": 67700,
      "train_bpb": 0.6169144961250653,
      "train_bytes": 56453.80625,
      "train_loss_nats_per_token": 1.742646248635758
    },
    {
      "epoch": 0.04424429169629454,
      "grad_norm": 0.10719604790210724,
      "learning_rate": 0.001753900413594117,
      "loss": 1.7382843017578125,
      "step": 67720,
      "train_bpb": 0.6136890019807035,
      "train_bytes": 56552.4875,
      "train_loss_nats_per_token": 1.7390366883250192
    },
    {
      "epoch": 0.044375971135866846,
      "grad_norm": 0.14761324226856232,
      "learning_rate": 0.0017532888211678488,
      "loss": 1.7392040252685548,
      "step": 67740,
      "train_bpb": 0.6134338593834235,
      "train_bytes": 56694.51875,
      "train_loss_nats_per_token": 1.7393057943361554
    },
    {
      "epoch": 0.04450765057543915,
      "grad_norm": 0.161391943693161,
      "learning_rate": 0.0017526771853953141,
      "loss": 1.7323968887329102,
      "step": 67760,
      "train_bpb": 0.614033526876579,
      "train_bytes": 56603.98125,
      "train_loss_nats_per_token": 1.7328535424486953
    },
    {
      "epoch": 0.044639330015011455,
      "grad_norm": 0.10537908971309662,
      "learning_rate": 0.0017520655063811834,
      "loss": 1.7267471313476563,
      "step": 67780,
      "train_bpb": 0.6128659674269545,
      "train_bytes": 56500.8875,
      "train_loss_nats_per_token": 1.726705043029541
    },
    {
      "epoch": 0.04477100945458376,
      "grad_norm": 0.10909292846918106,
      "learning_rate": 0.0017514537842301366,
      "loss": 1.736473274230957,
      "step": 67800,
      "train_bpb": 0.6135659765923749,
      "train_bytes": 57046.35625,
      "train_loss_nats_per_token": 1.7367529796582186
    },
    {
      "epoch": 0.044902688894156065,
      "grad_norm": 0.1095421090722084,
      "learning_rate": 0.0017508420190468596,
      "loss": 1.7249736785888672,
      "step": 67820,
      "train_bpb": 0.6086770957100722,
      "train_bytes": 56954.78125,
      "train_loss_nats_per_token": 1.7248702032081378
    },
    {
      "epoch": 0.04503436833372837,
      "grad_norm": 0.14420661330223083,
      "learning_rate": 0.0017502302109360464,
      "loss": 1.7368558883666991,
      "step": 67840,
      "train_bpb": 0.6140440529376019,
      "train_bytes": 56440.0125,
      "train_loss_nats_per_token": 1.7367351233042918
    },
    {
      "epoch": 0.045166047773300674,
      "grad_norm": 0.09497535973787308,
      "learning_rate": 0.001749618360002397,
      "loss": 1.7399438858032226,
      "step": 67860,
      "train_bpb": 0.6151114929104627,
      "train_bytes": 56447.8625,
      "train_loss_nats_per_token": 1.740213225077458
    },
    {
      "epoch": 0.04529772721287298,
      "grad_norm": 0.12421896308660507,
      "learning_rate": 0.0017490064663506213,
      "loss": 1.7372198104858398,
      "step": 67880,
      "train_bpb": 0.615126999210811,
      "train_bytes": 56581.3375,
      "train_loss_nats_per_token": 1.7373263721863839
    },
    {
      "epoch": 0.045429406652445284,
      "grad_norm": 0.11862685531377792,
      "learning_rate": 0.0017483945300854338,
      "loss": 1.7286647796630858,
      "step": 67900,
      "train_bpb": 0.6118446641386429,
      "train_bytes": 56999.275,
      "train_loss_nats_per_token": 1.72905574229308
    },
    {
      "epoch": 0.045561086092017596,
      "grad_norm": 0.09324538707733154,
      "learning_rate": 0.0017477825513115578,
      "loss": 1.7306053161621093,
      "step": 67920,
      "train_bpb": 0.6133278384191984,
      "train_bytes": 56360.91875,
      "train_loss_nats_per_token": 1.7308761092940137
    },
    {
      "epoch": 0.0456927655315899,
      "grad_norm": 0.10412263870239258,
      "learning_rate": 0.0017471705301337235,
      "loss": 1.7286174774169922,
      "step": 67940,
      "train_bpb": 0.6128565196237855,
      "train_bytes": 56467.16875,
      "train_loss_nats_per_token": 1.7283842384542525
    },
    {
      "epoch": 0.045824444971162205,
      "grad_norm": 0.15043815970420837,
      "learning_rate": 0.0017465584666566682,
      "loss": 1.723621940612793,
      "step": 67960,
      "train_bpb": 0.6094278776616899,
      "train_bytes": 56924.7875,
      "train_loss_nats_per_token": 1.7240131154435057
    },
    {
      "epoch": 0.04595612441073451,
      "grad_norm": 0.14227575063705444,
      "learning_rate": 0.0017459463609851371,
      "loss": 1.7290281295776366,
      "step": 67980,
      "train_bpb": 0.6120607583849192,
      "train_bytes": 56308.90625,
      "train_loss_nats_per_token": 1.7286040539834282
    },
    {
      "epoch": 0.046087803850306815,
      "grad_norm": 0.13114085793495178,
      "learning_rate": 0.0017453342132238816,
      "loss": 1.7311513900756836,
      "step": 68000,
      "train_bpb": 0.6129886794106906,
      "train_bytes": 56468.29375,
      "train_loss_nats_per_token": 1.7314214196483904
    },
    {
      "epoch": 0.046087803850306815,
      "eval_loss": 1.7090251445770264,
      "eval_runtime": 77.0267,
      "eval_samples_per_second": 12.983,
      "eval_steps_per_second": 12.983,
      "step": 68000,
      "train_bpb": 0.6025960351215127,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.7159601352284226
    },
    {
      "epoch": 0.04621948328987912,
      "grad_norm": 0.10921578109264374,
      "learning_rate": 0.0017447220234776612,
      "loss": 1.7251373291015626,
      "step": 68020,
      "train_bpb": 0.6115475274116592,
      "train_bytes": 56375.675,
      "train_loss_nats_per_token": 1.7255044342613135
    },
    {
      "epoch": 0.046351162729451424,
      "grad_norm": 0.13529512286186218,
      "learning_rate": 0.0017441097918512423,
      "loss": 1.7334465026855468,
      "step": 68040,
      "train_bpb": 0.6110133269032632,
      "train_bytes": 56734.775,
      "train_loss_nats_per_token": 1.7334008243931935
    },
    {
      "epoch": 0.04648284216902373,
      "grad_norm": 0.11603549867868423,
      "learning_rate": 0.001743497518449398,
      "loss": 1.7373937606811523,
      "step": 68060,
      "train_bpb": 0.6171029394317212,
      "train_bytes": 56126.11875,
      "train_loss_nats_per_token": 1.737871382844481
    },
    {
      "epoch": 0.046614521608596034,
      "grad_norm": 0.09091587364673615,
      "learning_rate": 0.0017428852033769095,
      "loss": 1.7389154434204102,
      "step": 68080,
      "train_bpb": 0.6147109546742388,
      "train_bytes": 56878.40625,
      "train_loss_nats_per_token": 1.7386096613134066
    },
    {
      "epoch": 0.04674620104816834,
      "grad_norm": 0.1033649668097496,
      "learning_rate": 0.001742272846738564,
      "loss": 1.7550987243652343,
      "step": 68100,
      "train_bpb": 0.6185158068779705,
      "train_bytes": 57026.08125,
      "train_loss_nats_per_token": 1.755273755547577
    },
    {
      "epoch": 0.04687788048774064,
      "grad_norm": 0.14353741705417633,
      "learning_rate": 0.0017416604486391563,
      "loss": 1.7560901641845703,
      "step": 68120,
      "train_bpb": 0.6235896493857019,
      "train_bytes": 56309.36875,
      "train_loss_nats_per_token": 1.7562710140016489
    },
    {
      "epoch": 0.04700955992731295,
      "grad_norm": 0.0961536392569542,
      "learning_rate": 0.0017410480091834895,
      "loss": 1.7523826599121093,
      "step": 68140,
      "train_bpb": 0.6194768366821667,
      "train_bytes": 56679.04375,
      "train_loss_nats_per_token": 1.7529202459610875
    },
    {
      "epoch": 0.04714123936688525,
      "grad_norm": 0.15177172422409058,
      "learning_rate": 0.0017404355284763714,
      "loss": 1.7572574615478516,
      "step": 68160,
      "train_bpb": 0.621021079488049,
      "train_bytes": 56584.80625,
      "train_loss_nats_per_token": 1.757046372831132
    },
    {
      "epoch": 0.04727291880645756,
      "grad_norm": 0.12996254861354828,
      "learning_rate": 0.0017398230066226186,
      "loss": 1.7646583557128905,
      "step": 68180,
      "train_bpb": 0.6225724409338139,
      "train_bytes": 56778.675,
      "train_loss_nats_per_token": 1.7651770035685148
    },
    {
      "epoch": 0.04740459824602986,
      "grad_norm": 0.11878655105829239,
      "learning_rate": 0.0017392104437270542,
      "loss": 1.7624086380004882,
      "step": 68200,
      "train_bpb": 0.6230825685856506,
      "train_bytes": 56409.85625,
      "train_loss_nats_per_token": 1.762604092799269
    },
    {
      "epoch": 0.04753627768560217,
      "grad_norm": 0.13612288236618042,
      "learning_rate": 0.001738597839894509,
      "loss": 1.7658885955810546,
      "step": 68220,
      "train_bpb": 0.6232623748955158,
      "train_bytes": 56701.68125,
      "train_loss_nats_per_token": 1.765788759889759
    },
    {
      "epoch": 0.04766795712517447,
      "grad_norm": 0.10974232107400894,
      "learning_rate": 0.0017379851952298193,
      "loss": 1.7685802459716797,
      "step": 68240,
      "train_bpb": 0.6222600549721502,
      "train_bytes": 56651.49375,
      "train_loss_nats_per_token": 1.7686877999693724
    },
    {
      "epoch": 0.047799636564746784,
      "grad_norm": 0.11873923242092133,
      "learning_rate": 0.00173737250983783,
      "loss": 1.7656290054321289,
      "step": 68260,
      "train_bpb": 0.6250789491792549,
      "train_bytes": 56180.08125,
      "train_loss_nats_per_token": 1.7658805718257007
    },
    {
      "epoch": 0.04793131600431909,
      "grad_norm": 0.1710149198770523,
      "learning_rate": 0.0017367597838233914,
      "loss": 1.7614166259765625,
      "step": 68280,
      "train_bpb": 0.6208008721918539,
      "train_bytes": 57233.575,
      "train_loss_nats_per_token": 1.7611134717273722
    },
    {
      "epoch": 0.04806299544389139,
      "grad_norm": 0.11897904425859451,
      "learning_rate": 0.0017361470172913623,
      "loss": 1.765839195251465,
      "step": 68300,
      "train_bpb": 0.623545314892095,
      "train_bytes": 56690.95,
      "train_loss_nats_per_token": 1.7659982392578804
    },
    {
      "epoch": 0.0481946748834637,
      "grad_norm": 0.18269097805023193,
      "learning_rate": 0.0017355342103466077,
      "loss": 1.7584772109985352,
      "step": 68320,
      "train_bpb": 0.6217732863748514,
      "train_bytes": 56256.25,
      "train_loss_nats_per_token": 1.759060600845679
    },
    {
      "epoch": 0.048326354323036,
      "grad_norm": 0.15134495496749878,
      "learning_rate": 0.001734921363093999,
      "loss": 1.7594259262084961,
      "step": 68340,
      "train_bpb": 0.6232673870762894,
      "train_bytes": 56461.33125,
      "train_loss_nats_per_token": 1.7595840061438315
    },
    {
      "epoch": 0.04845803376260831,
      "grad_norm": 0.09823567420244217,
      "learning_rate": 0.0017343084756384157,
      "loss": 1.7585054397583009,
      "step": 68360,
      "train_bpb": 0.6209600565617838,
      "train_bytes": 56973.90625,
      "train_loss_nats_per_token": 1.7586542818122133
    },
    {
      "epoch": 0.04858971320218061,
      "grad_norm": 0.1364130675792694,
      "learning_rate": 0.0017336955480847431,
      "loss": 1.7574592590332032,
      "step": 68380,
      "train_bpb": 0.6192021826273473,
      "train_bytes": 57058.00625,
      "train_loss_nats_per_token": 1.757493230371645
    },
    {
      "epoch": 0.04872139264175292,
      "grad_norm": 0.1746111512184143,
      "learning_rate": 0.0017330825805378742,
      "loss": 1.7584991455078125,
      "step": 68400,
      "train_bpb": 0.622764713164824,
      "train_bytes": 56883.325,
      "train_loss_nats_per_token": 1.7587696458421187
    },
    {
      "epoch": 0.04885307208132522,
      "grad_norm": 0.11517908424139023,
      "learning_rate": 0.001732469573102708,
      "loss": 1.7667497634887694,
      "step": 68420,
      "train_bpb": 0.6242446506450197,
      "train_bytes": 56232.9625,
      "train_loss_nats_per_token": 1.7668788070493633
    },
    {
      "epoch": 0.048984751520897526,
      "grad_norm": 0.13675005733966827,
      "learning_rate": 0.0017318565258841508,
      "loss": 1.7575536727905274,
      "step": 68440,
      "train_bpb": 0.6209206468630436,
      "train_bytes": 57093.8625,
      "train_loss_nats_per_token": 1.7577409997273379
    },
    {
      "epoch": 0.04911643096046983,
      "grad_norm": 0.12679851055145264,
      "learning_rate": 0.0017312434389871161,
      "loss": 1.746677780151367,
      "step": 68460,
      "train_bpb": 0.6177383631562708,
      "train_bytes": 56563.96875,
      "train_loss_nats_per_token": 1.7467818845764715
    },
    {
      "epoch": 0.049248110400042136,
      "grad_norm": 0.11655280739068985,
      "learning_rate": 0.0017306303125165239,
      "loss": 1.752352523803711,
      "step": 68480,
      "train_bpb": 0.620831822368447,
      "train_bytes": 56478.725,
      "train_loss_nats_per_token": 1.7523662484318059
    },
    {
      "epoch": 0.04937978983961444,
      "grad_norm": 0.10246056318283081,
      "learning_rate": 0.0017300171465772997,
      "loss": 1.7455577850341797,
      "step": 68500,
      "train_bpb": 0.6153996709588182,
      "train_bytes": 57018.55,
      "train_loss_nats_per_token": 1.74554273515667
    },
    {
      "epoch": 0.049511469279186746,
      "grad_norm": 0.09432516992092133,
      "learning_rate": 0.001729403941274378,
      "loss": 1.7478919982910157,
      "step": 68520,
      "train_bpb": 0.6205918223422805,
      "train_bytes": 56523.1375,
      "train_loss_nats_per_token": 1.7485859010465665
    },
    {
      "epoch": 0.04964314871875905,
      "grad_norm": 0.11579181253910065,
      "learning_rate": 0.0017287906967126982,
      "loss": 1.7549814224243163,
      "step": 68540,
      "train_bpb": 0.6202859084648602,
      "train_bytes": 56396.2,
      "train_loss_nats_per_token": 1.7549782185664973
    },
    {
      "epoch": 0.049774828158331355,
      "grad_norm": 0.1416843831539154,
      "learning_rate": 0.001728177412997208,
      "loss": 1.7442008972167968,
      "step": 68560,
      "train_bpb": 0.6157051851909401,
      "train_bytes": 57301.9125,
      "train_loss_nats_per_token": 1.744458392355261
    },
    {
      "epoch": 0.04990650759790366,
      "grad_norm": 0.14826159179210663,
      "learning_rate": 0.0017275640902328605,
      "loss": 1.7515270233154296,
      "step": 68580,
      "train_bpb": 0.6209420832957769,
      "train_bytes": 56354.1625,
      "train_loss_nats_per_token": 1.7517042397500462
    },
    {
      "epoch": 0.05003818703747597,
      "grad_norm": 0.11685791611671448,
      "learning_rate": 0.0017269507285246157,
      "loss": 1.7559333801269532,
      "step": 68600,
      "train_bpb": 0.6175162165306306,
      "train_bytes": 57085.975,
      "train_loss_nats_per_token": 1.7564189752528365
    },
    {
      "epoch": 0.050169866477048276,
      "grad_norm": 0.1643613874912262,
      "learning_rate": 0.0017263373279774404,
      "loss": 1.7419418334960937,
      "step": 68620,
      "train_bpb": 0.6172206912005452,
      "train_bytes": 56387.73125,
      "train_loss_nats_per_token": 1.7421362808101473
    },
    {
      "epoch": 0.05030154591662058,
      "grad_norm": 0.1384725421667099,
      "learning_rate": 0.001725723888696309,
      "loss": 1.7476787567138672,
      "step": 68640,
      "train_bpb": 0.6180279411541335,
      "train_bytes": 56427.66875,
      "train_loss_nats_per_token": 1.7474382829167567
    },
    {
      "epoch": 0.050433225356192886,
      "grad_norm": 0.12071803212165833,
      "learning_rate": 0.0017251104107862004,
      "loss": 1.7565793991088867,
      "step": 68660,
      "train_bpb": 0.6210563364255954,
      "train_bytes": 56850.3,
      "train_loss_nats_per_token": 1.7568386093380413
    },
    {
      "epoch": 0.05056490479576519,
      "grad_norm": 0.10008357465267181,
      "learning_rate": 0.0017244968943521022,
      "loss": 1.7440914154052733,
      "step": 68680,
      "train_bpb": 0.6162997056095562,
      "train_bytes": 56886.16875,
      "train_loss_nats_per_token": 1.7444556594645395
    },
    {
      "epoch": 0.050696584235337495,
      "grad_norm": 0.1577955037355423,
      "learning_rate": 0.0017238833394990075,
      "loss": 1.7482124328613282,
      "step": 68700,
      "train_bpb": 0.6173269428811814,
      "train_bytes": 56585.49375,
      "train_loss_nats_per_token": 1.7484099260329644
    },
    {
      "epoch": 0.0508282636749098,
      "grad_norm": 0.10657801479101181,
      "learning_rate": 0.001723269746331917,
      "loss": 1.7553913116455078,
      "step": 68720,
      "train_bpb": 0.6222485739480147,
      "train_bytes": 56274.2125,
      "train_loss_nats_per_token": 1.7561646140981897
    },
    {
      "epoch": 0.050959943114482105,
      "grad_norm": 0.10856018960475922,
      "learning_rate": 0.0017226561149558358,
      "loss": 1.7502891540527343,
      "step": 68740,
      "train_bpb": 0.6170116687959734,
      "train_bytes": 57069.4,
      "train_loss_nats_per_token": 1.7500548441059178
    },
    {
      "epoch": 0.05109162255405441,
      "grad_norm": 0.1475059539079666,
      "learning_rate": 0.001722042445475778,
      "loss": 1.747053337097168,
      "step": 68760,
      "train_bpb": 0.6156250378610987,
      "train_bytes": 57221.7125,
      "train_loss_nats_per_token": 1.747127300390797
    },
    {
      "epoch": 0.051223301993626714,
      "grad_norm": 0.12851455807685852,
      "learning_rate": 0.001721428737996763,
      "loss": 1.7425302505493163,
      "step": 68780,
      "train_bpb": 0.6164560004783338,
      "train_bytes": 56138.06875,
      "train_loss_nats_per_token": 1.742838368126903
    },
    {
      "epoch": 0.05135498143319902,
      "grad_norm": 0.16194961965084076,
      "learning_rate": 0.0017208149926238169,
      "loss": 1.7439794540405273,
      "step": 68800,
      "train_bpb": 0.6173377218020162,
      "train_bytes": 56668.3,
      "train_loss_nats_per_token": 1.744229235499699
    },
    {
      "epoch": 0.051486660872771324,
      "grad_norm": 0.095216304063797,
      "learning_rate": 0.0017202012094619714,
      "loss": 1.7384283065795898,
      "step": 68820,
      "train_bpb": 0.6145405936657983,
      "train_bytes": 56696.5625,
      "train_loss_nats_per_token": 1.7382362064658126
    },
    {
      "epoch": 0.05161834031234363,
      "grad_norm": 0.21598264575004578,
      "learning_rate": 0.0017195873886162664,
      "loss": 1.7452983856201172,
      "step": 68840,
      "train_bpb": 0.6160506063810075,
      "train_bytes": 56775.7625,
      "train_loss_nats_per_token": 1.745511476624749
    },
    {
      "epoch": 0.051750019751915934,
      "grad_norm": 0.11842844635248184,
      "learning_rate": 0.0017189735301917472,
      "loss": 1.730061912536621,
      "step": 68860,
      "train_bpb": 0.6087630219260115,
      "train_bytes": 56862.01875,
      "train_loss_nats_per_token": 1.7300446424346565
    },
    {
      "epoch": 0.05188169919148824,
      "grad_norm": 0.10631615668535233,
      "learning_rate": 0.0017183596342934663,
      "loss": 1.7397266387939454,
      "step": 68880,
      "train_bpb": 0.6164614294438222,
      "train_bytes": 56315.61875,
      "train_loss_nats_per_token": 1.7397570064581627
    },
    {
      "epoch": 0.05201337863106054,
      "grad_norm": 0.12083332240581512,
      "learning_rate": 0.001717745701026481,
      "loss": 1.740973663330078,
      "step": 68900,
      "train_bpb": 0.616411443293268,
      "train_bytes": 56695.79375,
      "train_loss_nats_per_token": 1.7413172859225161
    },
    {
      "epoch": 0.052145058070632855,
      "grad_norm": 0.13787798583507538,
      "learning_rate": 0.0017171317304958563,
      "loss": 1.7352323532104492,
      "step": 68920,
      "train_bpb": 0.6162682569082368,
      "train_bytes": 56357.09375,
      "train_loss_nats_per_token": 1.7355473556781398
    },
    {
      "epoch": 0.05227673751020516,
      "grad_norm": 0.10850312560796738,
      "learning_rate": 0.0017165177228066631,
      "loss": 1.7462282180786133,
      "step": 68940,
      "train_bpb": 0.6182496216850144,
      "train_bytes": 56452.36875,
      "train_loss_nats_per_token": 1.7461896455725718
    },
    {
      "epoch": 0.052408416949777464,
      "grad_norm": 0.13137702643871307,
      "learning_rate": 0.0017159036780639793,
      "loss": 1.7392492294311523,
      "step": 68960,
      "train_bpb": 0.6141070916476461,
      "train_bytes": 56893.28125,
      "train_loss_nats_per_token": 1.739371976185883
    },
    {
      "epoch": 0.05254009638934977,
      "grad_norm": 0.1069563552737236,
      "learning_rate": 0.0017152895963728888,
      "loss": 1.7405105590820313,
      "step": 68980,
      "train_bpb": 0.6125047844276247,
      "train_bytes": 57031.45625,
      "train_loss_nats_per_token": 1.7408326086229375
    },
    {
      "epoch": 0.052671775828922074,
      "grad_norm": 0.13045178353786469,
      "learning_rate": 0.001714675477838481,
      "loss": 1.7295398712158203,
      "step": 69000,
      "train_bpb": 0.6123255195523242,
      "train_bytes": 56595.43125,
      "train_loss_nats_per_token": 1.7298144601926728
    },
    {
      "epoch": 0.052671775828922074,
      "eval_loss": 1.7077986001968384,
      "eval_runtime": 76.078,
      "eval_samples_per_second": 13.144,
      "eval_steps_per_second": 13.144,
      "step": 69000,
      "train_bpb": 0.6021220914123413,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.7146105270931549
    },
    {
      "epoch": 0.05280345526849438,
      "grad_norm": 0.1225469782948494,
      "learning_rate": 0.0017140613225658527,
      "loss": 1.7407466888427734,
      "step": 69020,
      "train_bpb": 0.6173213258428758,
      "train_bytes": 56299.6,
      "train_loss_nats_per_token": 1.7408929193114084
    },
    {
      "epoch": 0.05293513470806668,
      "grad_norm": 0.10395469516515732,
      "learning_rate": 0.0017134471306601065,
      "loss": 1.7372333526611328,
      "step": 69040,
      "train_bpb": 0.6125325162995318,
      "train_bytes": 56808.5,
      "train_loss_nats_per_token": 1.7374533543451227
    },
    {
      "epoch": 0.05306681414763899,
      "grad_norm": 0.11525048315525055,
      "learning_rate": 0.0017128329022263516,
      "loss": 1.7353818893432618,
      "step": 69060,
      "train_bpb": 0.6135367113849752,
      "train_bytes": 56846.1625,
      "train_loss_nats_per_token": 1.7361660583913963
    },
    {
      "epoch": 0.05319849358721129,
      "grad_norm": 0.11179927736520767,
      "learning_rate": 0.0017122186373697032,
      "loss": 1.739413070678711,
      "step": 69080,
      "train_bpb": 0.6155279274541076,
      "train_bytes": 56347.83125,
      "train_loss_nats_per_token": 1.7393776121895528
    },
    {
      "epoch": 0.0533301730267836,
      "grad_norm": 0.11327791959047318,
      "learning_rate": 0.0017116043361952826,
      "loss": 1.729496955871582,
      "step": 69100,
      "train_bpb": 0.6113152432995336,
      "train_bytes": 56882.04375,
      "train_loss_nats_per_token": 1.729431618670918
    },
    {
      "epoch": 0.0534618524663559,
      "grad_norm": 0.1354941576719284,
      "learning_rate": 0.0017109899988082173,
      "loss": 1.7394012451171874,
      "step": 69120,
      "train_bpb": 0.6132968255217918,
      "train_bytes": 56558.925,
      "train_loss_nats_per_token": 1.7397471695862141
    },
    {
      "epoch": 0.05359353190592821,
      "grad_norm": 0.09174686670303345,
      "learning_rate": 0.0017103756253136415,
      "loss": 1.7353641510009765,
      "step": 69140,
      "train_bpb": 0.6130863372569343,
      "train_bytes": 56862.71875,
      "train_loss_nats_per_token": 1.7356672193047846
    },
    {
      "epoch": 0.05372521134550051,
      "grad_norm": 0.11450312286615372,
      "learning_rate": 0.001709761215816695,
      "loss": 1.7372150421142578,
      "step": 69160,
      "train_bpb": 0.6145474279745945,
      "train_bytes": 56459.9875,
      "train_loss_nats_per_token": 1.7374880196398537
    },
    {
      "epoch": 0.05385689078507282,
      "grad_norm": 0.1526605784893036,
      "learning_rate": 0.001709146770422524,
      "loss": 1.7390140533447265,
      "step": 69180,
      "train_bpb": 0.6154099068346034,
      "train_bytes": 56327.4875,
      "train_loss_nats_per_token": 1.739540305426811
    },
    {
      "epoch": 0.05398857022464512,
      "grad_norm": 0.10622841864824295,
      "learning_rate": 0.0017085322892362812,
      "loss": 1.7387792587280273,
      "step": 69200,
      "train_bpb": 0.61279876717947,
      "train_bytes": 56989.6,
      "train_loss_nats_per_token": 1.7391239176506326
    },
    {
      "epoch": 0.054120249664217426,
      "grad_norm": 0.12610821425914764,
      "learning_rate": 0.001707917772363125,
      "loss": 1.732680320739746,
      "step": 69220,
      "train_bpb": 0.6141240193476185,
      "train_bytes": 56600.40625,
      "train_loss_nats_per_token": 1.7331256781664515
    },
    {
      "epoch": 0.05425192910378973,
      "grad_norm": 0.14352662861347198,
      "learning_rate": 0.0017073032199082195,
      "loss": 1.7382108688354492,
      "step": 69240,
      "train_bpb": 0.6139356629472409,
      "train_bytes": 56423.9875,
      "train_loss_nats_per_token": 1.7391779951213489
    },
    {
      "epoch": 0.05438360854336204,
      "grad_norm": 0.15801072120666504,
      "learning_rate": 0.0017066886319767362,
      "loss": 1.7301963806152343,
      "step": 69260,
      "train_bpb": 0.6140750612880286,
      "train_bytes": 56379.975,
      "train_loss_nats_per_token": 1.7307841963896125
    },
    {
      "epoch": 0.05451528798293435,
      "grad_norm": 0.10234980285167694,
      "learning_rate": 0.0017060740086738506,
      "loss": 1.746319580078125,
      "step": 69280,
      "train_bpb": 0.6164914236372138,
      "train_bytes": 56445.425,
      "train_loss_nats_per_token": 1.7462973415398992
    },
    {
      "epoch": 0.05464696742250665,
      "grad_norm": 0.12496054917573929,
      "learning_rate": 0.001705459350104747,
      "loss": 1.7547456741333007,
      "step": 69300,
      "train_bpb": 0.6205446350942906,
      "train_bytes": 56881.825,
      "train_loss_nats_per_token": 1.7549401517671939
    },
    {
      "epoch": 0.05477864686207896,
      "grad_norm": 0.14522233605384827,
      "learning_rate": 0.0017048446563746127,
      "loss": 1.7491983413696288,
      "step": 69320,
      "train_bpb": 0.6174673684406435,
      "train_bytes": 56558.43125,
      "train_loss_nats_per_token": 1.7492241130022306
    },
    {
      "epoch": 0.05491032630165126,
      "grad_norm": 0.12734948098659515,
      "learning_rate": 0.001704229927588643,
      "loss": 1.7614456176757813,
      "step": 69340,
      "train_bpb": 0.623121029349792,
      "train_bytes": 56584.1375,
      "train_loss_nats_per_token": 1.7618746780887151
    },
    {
      "epoch": 0.05504200574122357,
      "grad_norm": 0.0992262065410614,
      "learning_rate": 0.0017036151638520397,
      "loss": 1.758976364135742,
      "step": 69360,
      "train_bpb": 0.6201070320938017,
      "train_bytes": 56540.275,
      "train_loss_nats_per_token": 1.7592686112815306
    },
    {
      "epoch": 0.05517368518079587,
      "grad_norm": 0.1127210259437561,
      "learning_rate": 0.001703000365270009,
      "loss": 1.750307846069336,
      "step": 69380,
      "train_bpb": 0.622080619797096,
      "train_bytes": 56129.94375,
      "train_loss_nats_per_token": 1.750815083751808
    },
    {
      "epoch": 0.055305364620368176,
      "grad_norm": 0.09238316863775253,
      "learning_rate": 0.0017023855319477635,
      "loss": 1.7503311157226562,
      "step": 69400,
      "train_bpb": 0.6192341868648507,
      "train_bytes": 56736.53125,
      "train_loss_nats_per_token": 1.750419272156487
    },
    {
      "epoch": 0.05543704405994048,
      "grad_norm": 0.12219781428575516,
      "learning_rate": 0.0017017706639905219,
      "loss": 1.7535955429077148,
      "step": 69420,
      "train_bpb": 0.6207696140583399,
      "train_bytes": 56559.375,
      "train_loss_nats_per_token": 1.753436395092286
    },
    {
      "epoch": 0.055568723499512786,
      "grad_norm": 0.1062895879149437,
      "learning_rate": 0.0017011557615035089,
      "loss": 1.753286361694336,
      "step": 69440,
      "train_bpb": 0.6193588906059371,
      "train_bytes": 56408.63125,
      "train_loss_nats_per_token": 1.753504742411463
    },
    {
      "epoch": 0.05570040293908509,
      "grad_norm": 0.1761479526758194,
      "learning_rate": 0.0017005408245919556,
      "loss": 1.7424940109252929,
      "step": 69460,
      "train_bpb": 0.6174386403188719,
      "train_bytes": 56477.56875,
      "train_loss_nats_per_token": 1.7430160670085064
    },
    {
      "epoch": 0.055832082378657395,
      "grad_norm": 0.13355162739753723,
      "learning_rate": 0.0016999258533610978,
      "loss": 1.7458745956420898,
      "step": 69480,
      "train_bpb": 0.6203716513401007,
      "train_bytes": 56572.13125,
      "train_loss_nats_per_token": 1.7463808541555619
    },
    {
      "epoch": 0.0559637618182297,
      "grad_norm": 0.126582071185112,
      "learning_rate": 0.0016993108479161781,
      "loss": 1.7483613967895508,
      "step": 69500,
      "train_bpb": 0.6190523892407079,
      "train_bytes": 56541.94375,
      "train_loss_nats_per_token": 1.7481453897856212
    },
    {
      "epoch": 0.056095441257802005,
      "grad_norm": 0.08464136719703674,
      "learning_rate": 0.0016986958083624448,
      "loss": 1.7501449584960938,
      "step": 69520,
      "train_bpb": 0.6175621802040345,
      "train_bytes": 56928.3875,
      "train_loss_nats_per_token": 1.7503245677669448
    },
    {
      "epoch": 0.05622712069737431,
      "grad_norm": 0.11908835172653198,
      "learning_rate": 0.0016980807348051519,
      "loss": 1.7422473907470704,
      "step": 69540,
      "train_bpb": 0.6148987873775886,
      "train_bytes": 56819.85625,
      "train_loss_nats_per_token": 1.742675676499577
    },
    {
      "epoch": 0.056358800136946614,
      "grad_norm": 0.11111199110746384,
      "learning_rate": 0.0016974656273495592,
      "loss": 1.742528533935547,
      "step": 69560,
      "train_bpb": 0.6169755639461678,
      "train_bytes": 56647.8375,
      "train_loss_nats_per_token": 1.743079626083265
    },
    {
      "epoch": 0.05649047957651892,
      "grad_norm": 0.09571356326341629,
      "learning_rate": 0.0016968504861009323,
      "loss": 1.744684600830078,
      "step": 69580,
      "train_bpb": 0.616837827646692,
      "train_bytes": 57212.075,
      "train_loss_nats_per_token": 1.74527530549901
    },
    {
      "epoch": 0.05662215901609123,
      "grad_norm": 0.12213265895843506,
      "learning_rate": 0.0016962353111645428,
      "loss": 1.7419164657592774,
      "step": 69600,
      "train_bpb": 0.6154208974511606,
      "train_bytes": 56664.5625,
      "train_loss_nats_per_token": 1.741979311937543
    },
    {
      "epoch": 0.056753838455663536,
      "grad_norm": 0.12492791563272476,
      "learning_rate": 0.0016956201026456675,
      "loss": 1.7392982482910155,
      "step": 69620,
      "train_bpb": 0.6130576836738212,
      "train_bytes": 57048.7125,
      "train_loss_nats_per_token": 1.7396449114742225
    },
    {
      "epoch": 0.05688551789523584,
      "grad_norm": 0.1087220087647438,
      "learning_rate": 0.00169500486064959,
      "loss": 1.742051887512207,
      "step": 69640,
      "train_bpb": 0.6162406794571267,
      "train_bytes": 56449.46875,
      "train_loss_nats_per_token": 1.7424652388772763
    },
    {
      "epoch": 0.057017197334808145,
      "grad_norm": 0.1329224854707718,
      "learning_rate": 0.0016943895852815978,
      "loss": 1.7451723098754883,
      "step": 69660,
      "train_bpb": 0.6159571353748223,
      "train_bytes": 56936.73125,
      "train_loss_nats_per_token": 1.7450497703284857
    },
    {
      "epoch": 0.05714887677438045,
      "grad_norm": 0.12450263649225235,
      "learning_rate": 0.0016937742766469867,
      "loss": 1.7396595001220703,
      "step": 69680,
      "train_bpb": 0.6166808304368355,
      "train_bytes": 56815.225,
      "train_loss_nats_per_token": 1.7399126535227893
    },
    {
      "epoch": 0.057280556213952755,
      "grad_norm": 0.11714454740285873,
      "learning_rate": 0.0016931589348510557,
      "loss": 1.7390140533447265,
      "step": 69700,
      "train_bpb": 0.6150732177978109,
      "train_bytes": 56409.1,
      "train_loss_nats_per_token": 1.7390569824773097
    },
    {
      "epoch": 0.05741223565352506,
      "grad_norm": 0.16534774005413055,
      "learning_rate": 0.0016925435599991117,
      "loss": 1.7341142654418946,
      "step": 69720,
      "train_bpb": 0.6140955378791789,
      "train_bytes": 56142.9,
      "train_loss_nats_per_token": 1.7339835929050185
    },
    {
      "epoch": 0.057543915093097364,
      "grad_norm": 0.11600124090909958,
      "learning_rate": 0.0016919281521964645,
      "loss": 1.7265274047851562,
      "step": 69740,
      "train_bpb": 0.6107997371248053,
      "train_bytes": 56830.3875,
      "train_loss_nats_per_token": 1.7267834583836086
    },
    {
      "epoch": 0.05767559453266967,
      "grad_norm": 0.11774647235870361,
      "learning_rate": 0.0016913127115484325,
      "loss": 1.7297332763671875,
      "step": 69760,
      "train_bpb": 0.6132577050049317,
      "train_bytes": 56366.1875,
      "train_loss_nats_per_token": 1.729661553023296
    },
    {
      "epoch": 0.057807273972241974,
      "grad_norm": 0.10113774985074997,
      "learning_rate": 0.0016906972381603373,
      "loss": 1.7259651184082032,
      "step": 69780,
      "train_bpb": 0.6119835080602994,
      "train_bytes": 56741.73125,
      "train_loss_nats_per_token": 1.7262548670144278
    },
    {
      "epoch": 0.05793895341181428,
      "grad_norm": 0.10899896919727325,
      "learning_rate": 0.0016900817321375088,
      "loss": 1.7294694900512695,
      "step": 69800,
      "train_bpb": 0.6118528756406227,
      "train_bytes": 56609.8875,
      "train_loss_nats_per_token": 1.7295466150898573
    },
    {
      "epoch": 0.05807063285138658,
      "grad_norm": 0.11490485072135925,
      "learning_rate": 0.0016894661935852792,
      "loss": 1.7334844589233398,
      "step": 69820,
      "train_bpb": 0.6132031127826746,
      "train_bytes": 56777.7,
      "train_loss_nats_per_token": 1.733611574917725
    },
    {
      "epoch": 0.05820231229095889,
      "grad_norm": 0.10855840146541595,
      "learning_rate": 0.0016888506226089887,
      "loss": 1.7322589874267578,
      "step": 69840,
      "train_bpb": 0.6135158164201966,
      "train_bytes": 56374.85625,
      "train_loss_nats_per_token": 1.7324041798790237
    },
    {
      "epoch": 0.05833399173053119,
      "grad_norm": 0.12649796903133392,
      "learning_rate": 0.001688235019313982,
      "loss": 1.726606559753418,
      "step": 69860,
      "train_bpb": 0.6126670872670965,
      "train_bytes": 56248.41875,
      "train_loss_nats_per_token": 1.7267943418023541
    },
    {
      "epoch": 0.0584656711701035,
      "grad_norm": 0.13009482622146606,
      "learning_rate": 0.00168761938380561,
      "loss": 1.7286685943603515,
      "step": 69880,
      "train_bpb": 0.6128696995384272,
      "train_bytes": 56397.7875,
      "train_loss_nats_per_token": 1.729129031306117
    },
    {
      "epoch": 0.0585973506096758,
      "grad_norm": 0.17691923677921295,
      "learning_rate": 0.001687003716189228,
      "loss": 1.7346275329589844,
      "step": 69900,
      "train_bpb": 0.6134444170112492,
      "train_bytes": 56472.4625,
      "train_loss_nats_per_token": 1.7349626749286056
    },
    {
      "epoch": 0.05872903004924811,
      "grad_norm": 0.14012116193771362,
      "learning_rate": 0.0016863880165701986,
      "loss": 1.7250659942626954,
      "step": 69920,
      "train_bpb": 0.6108768691874755,
      "train_bytes": 56664.36875,
      "train_loss_nats_per_token": 1.7255679415405765
    },
    {
      "epoch": 0.05886070948882042,
      "grad_norm": 0.1260337233543396,
      "learning_rate": 0.0016857722850538877,
      "loss": 1.7241352081298829,
      "step": 69940,
      "train_bpb": 0.6139030559591308,
      "train_bytes": 56211.06875,
      "train_loss_nats_per_token": 1.7244456177682976
    },
    {
      "epoch": 0.058992388928392724,
      "grad_norm": 0.11018558591604233,
      "learning_rate": 0.0016851565217456685,
      "loss": 1.7331155776977538,
      "step": 69960,
      "train_bpb": 0.6136229190075132,
      "train_bytes": 56421.60625,
      "train_loss_nats_per_token": 1.7327153121611258
    },
    {
      "epoch": 0.05912406836796503,
      "grad_norm": 0.1260758489370346,
      "learning_rate": 0.0016845407267509183,
      "loss": 1.7309453964233399,
      "step": 69980,
      "train_bpb": 0.6120464660805914,
      "train_bytes": 56793.09375,
      "train_loss_nats_per_token": 1.7313789499659564
    },
    {
      "epoch": 0.05925574780753733,
      "grad_norm": 0.1335260570049286,
      "learning_rate": 0.0016839249001750207,
      "loss": 1.7260385513305665,
      "step": 70000,
      "train_bpb": 0.6106149463565073,
      "train_bytes": 56727.95,
      "train_loss_nats_per_token": 1.7264429535564472
    },
    {
      "epoch": 0.05925574780753733,
      "eval_loss": 1.708622932434082,
      "eval_runtime": 76.2224,
      "eval_samples_per_second": 13.12,
      "eval_steps_per_second": 13.12,
      "step": 70000,
      "train_bpb": 0.6024171539537923,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.7154507509396795
    },
    {
      "epoch": 0.05938742724710964,
      "grad_norm": 0.11723925173282623,
      "learning_rate": 0.0016833090421233643,
      "loss": 1.728260612487793,
      "step": 70020,
      "train_bpb": 0.614048495469996,
      "train_bytes": 56222.0375,
      "train_loss_nats_per_token": 1.7279937607344358
    },
    {
      "epoch": 0.05951910668668194,
      "grad_norm": 0.12407031655311584,
      "learning_rate": 0.0016826931527013438,
      "loss": 1.735132598876953,
      "step": 70040,
      "train_bpb": 0.6153942704154657,
      "train_bytes": 56226.86875,
      "train_loss_nats_per_token": 1.734984353114614
    },
    {
      "epoch": 0.05965078612625425,
      "grad_norm": 0.10960811376571655,
      "learning_rate": 0.0016820772320143577,
      "loss": 1.734769058227539,
      "step": 70060,
      "train_bpb": 0.6112482532321764,
      "train_bytes": 56999.6625,
      "train_loss_nats_per_token": 1.7350719229571512
    },
    {
      "epoch": 0.05978246556582655,
      "grad_norm": 0.1727869063615799,
      "learning_rate": 0.0016814612801678112,
      "loss": 1.7230836868286132,
      "step": 70080,
      "train_bpb": 0.6129491667514666,
      "train_bytes": 55904.26875,
      "train_loss_nats_per_token": 1.7233316311462958
    },
    {
      "epoch": 0.05991414500539886,
      "grad_norm": 0.13737550377845764,
      "learning_rate": 0.0016808452972671146,
      "loss": 1.722420883178711,
      "step": 70100,
      "train_bpb": 0.6092397554413067,
      "train_bytes": 56483.29375,
      "train_loss_nats_per_token": 1.7228855741308597
    },
    {
      "epoch": 0.06004582444497116,
      "grad_norm": 0.10050071775913239,
      "learning_rate": 0.001680229283417684,
      "loss": 1.7264396667480468,
      "step": 70120,
      "train_bpb": 0.6091772931503967,
      "train_bytes": 56932.86875,
      "train_loss_nats_per_token": 1.7267777945271359
    },
    {
      "epoch": 0.060177503884543466,
      "grad_norm": 0.11976324021816254,
      "learning_rate": 0.0016796132387249388,
      "loss": 1.7167184829711915,
      "step": 70140,
      "train_bpb": 0.6070952017883269,
      "train_bytes": 56494.1,
      "train_loss_nats_per_token": 1.7168409189112321
    },
    {
      "epoch": 0.06030918332411577,
      "grad_norm": 0.16113649308681488,
      "learning_rate": 0.0016789971632943056,
      "loss": 1.7265333175659179,
      "step": 70160,
      "train_bpb": 0.6135580918386941,
      "train_bytes": 56034.38125,
      "train_loss_nats_per_token": 1.7263787768047185
    },
    {
      "epoch": 0.060440862763688076,
      "grad_norm": 0.12247829139232635,
      "learning_rate": 0.0016783810572312161,
      "loss": 1.7122283935546876,
      "step": 70180,
      "train_bpb": 0.6081704962068927,
      "train_bytes": 55877.75625,
      "train_loss_nats_per_token": 1.7121256995309875
    },
    {
      "epoch": 0.06057254220326038,
      "grad_norm": 0.12242767214775085,
      "learning_rate": 0.0016777649206411063,
      "loss": 1.7215398788452148,
      "step": 70200,
      "train_bpb": 0.6097714338909879,
      "train_bytes": 56231.65625,
      "train_loss_nats_per_token": 1.7218059009603544
    },
    {
      "epoch": 0.060704221642832686,
      "grad_norm": 0.1449415683746338,
      "learning_rate": 0.0016771487536294187,
      "loss": 1.7272953033447265,
      "step": 70220,
      "train_bpb": 0.6129781387290184,
      "train_bytes": 56352.31875,
      "train_loss_nats_per_token": 1.7275676655302235
    },
    {
      "epoch": 0.06083590108240499,
      "grad_norm": 0.10652636736631393,
      "learning_rate": 0.0016765325563015995,
      "loss": 1.7169567108154298,
      "step": 70240,
      "train_bpb": 0.6098153517664727,
      "train_bytes": 56196.2375,
      "train_loss_nats_per_token": 1.717069187983726
    },
    {
      "epoch": 0.060967580521977295,
      "grad_norm": 0.12485978752374649,
      "learning_rate": 0.0016759163287631015,
      "loss": 1.7214468002319336,
      "step": 70260,
      "train_bpb": 0.6082817973616951,
      "train_bytes": 56830.1125,
      "train_loss_nats_per_token": 1.7215578818912602
    },
    {
      "epoch": 0.06109925996154961,
      "grad_norm": 0.11328538507223129,
      "learning_rate": 0.0016753000711193814,
      "loss": 1.7303829193115234,
      "step": 70280,
      "train_bpb": 0.6125591514820903,
      "train_bytes": 56794.1,
      "train_loss_nats_per_token": 1.7310523647411433
    },
    {
      "epoch": 0.06123093940112191,
      "grad_norm": 0.12800779938697815,
      "learning_rate": 0.0016746837834759022,
      "loss": 1.7200902938842773,
      "step": 70300,
      "train_bpb": 0.60926083366337,
      "train_bytes": 56475.4875,
      "train_loss_nats_per_token": 1.7199397792297708
    },
    {
      "epoch": 0.061362618840694216,
      "grad_norm": 0.1265990138053894,
      "learning_rate": 0.001674067465938131,
      "loss": 1.7249839782714844,
      "step": 70320,
      "train_bpb": 0.6135400678742169,
      "train_bytes": 55990.325,
      "train_loss_nats_per_token": 1.7251777562486754
    },
    {
      "epoch": 0.06149429828026652,
      "grad_norm": 0.160077303647995,
      "learning_rate": 0.0016734511186115406,
      "loss": 1.7144487380981446,
      "step": 70340,
      "train_bpb": 0.606027768744676,
      "train_bytes": 56429.2625,
      "train_loss_nats_per_token": 1.7144616318188752
    },
    {
      "epoch": 0.061625977719838826,
      "grad_norm": 0.10297481715679169,
      "learning_rate": 0.0016728347416016095,
      "loss": 1.7148799896240234,
      "step": 70360,
      "train_bpb": 0.6078560697334697,
      "train_bytes": 56579.375,
      "train_loss_nats_per_token": 1.71552904384059
    },
    {
      "epoch": 0.06175765715941113,
      "grad_norm": 0.0975503996014595,
      "learning_rate": 0.0016722183350138203,
      "loss": 1.7149600982666016,
      "step": 70380,
      "train_bpb": 0.6050268012199196,
      "train_bytes": 56538.95625,
      "train_loss_nats_per_token": 1.7149804021778317
    },
    {
      "epoch": 0.061889336598983435,
      "grad_norm": 0.09516748040914536,
      "learning_rate": 0.0016716018989536604,
      "loss": 1.7078044891357422,
      "step": 70400,
      "train_bpb": 0.6041377834553053,
      "train_bytes": 56488.6375,
      "train_loss_nats_per_token": 1.7080146831087335
    },
    {
      "epoch": 0.06202101603855574,
      "grad_norm": 0.11138808727264404,
      "learning_rate": 0.0016709854335266237,
      "loss": 1.7088891983032226,
      "step": 70420,
      "train_bpb": 0.6075465243476268,
      "train_bytes": 56089.08125,
      "train_loss_nats_per_token": 1.7091005941338204
    },
    {
      "epoch": 0.062152695478128045,
      "grad_norm": 0.12611979246139526,
      "learning_rate": 0.0016703689388382072,
      "loss": 1.7307512283325195,
      "step": 70440,
      "train_bpb": 0.6113745667241797,
      "train_bytes": 56751.8625,
      "train_loss_nats_per_token": 1.7311320541348638
    },
    {
      "epoch": 0.06228437491770035,
      "grad_norm": 0.12514522671699524,
      "learning_rate": 0.0016697524149939151,
      "loss": 1.736818504333496,
      "step": 70460,
      "train_bpb": 0.6150994773628278,
      "train_bytes": 56765.0375,
      "train_loss_nats_per_token": 1.7367502629630909
    },
    {
      "epoch": 0.062416054357272654,
      "grad_norm": 0.10736590623855591,
      "learning_rate": 0.001669135862099255,
      "loss": 1.7401582717895507,
      "step": 70480,
      "train_bpb": 0.6181157918050021,
      "train_bytes": 56506.35,
      "train_loss_nats_per_token": 1.7404170418593434
    },
    {
      "epoch": 0.06254773379684496,
      "grad_norm": 0.14283962547779083,
      "learning_rate": 0.0016685192802597393,
      "loss": 1.7577373504638671,
      "step": 70500,
      "train_bpb": 0.6231688943493459,
      "train_bytes": 56250.04375,
      "train_loss_nats_per_token": 1.7577777368480905
    },
    {
      "epoch": 0.06267941323641726,
      "grad_norm": 0.12723104655742645,
      "learning_rate": 0.0016679026695808866,
      "loss": 1.7486257553100586,
      "step": 70520,
      "train_bpb": 0.618228737184438,
      "train_bytes": 56622.96875,
      "train_loss_nats_per_token": 1.749164879999252
    },
    {
      "epoch": 0.06281109267598957,
      "grad_norm": 0.13018546998500824,
      "learning_rate": 0.0016672860301682202,
      "loss": 1.7432897567749024,
      "step": 70540,
      "train_bpb": 0.6176969933107593,
      "train_bytes": 56626.5125,
      "train_loss_nats_per_token": 1.7437919568444689
    },
    {
      "epoch": 0.06294277211556187,
      "grad_norm": 0.12599371373653412,
      "learning_rate": 0.001666669362127267,
      "loss": 1.753157615661621,
      "step": 70560,
      "train_bpb": 0.6211562371040046,
      "train_bytes": 56259.3125,
      "train_loss_nats_per_token": 1.7537857759291664
    },
    {
      "epoch": 0.06307445155513418,
      "grad_norm": 0.1453937292098999,
      "learning_rate": 0.00166605266556356,
      "loss": 1.7625577926635743,
      "step": 70580,
      "train_bpb": 0.6253466411302103,
      "train_bytes": 56367.29375,
      "train_loss_nats_per_token": 1.7632490252769193
    },
    {
      "epoch": 0.06320613099470648,
      "grad_norm": 0.10890887677669525,
      "learning_rate": 0.0016654359405826372,
      "loss": 1.7604036331176758,
      "step": 70600,
      "train_bpb": 0.623631509564101,
      "train_bytes": 56561.1625,
      "train_loss_nats_per_token": 1.7606165132253153
    },
    {
      "epoch": 0.06333781043427879,
      "grad_norm": 0.11065804213285446,
      "learning_rate": 0.0016648191872900406,
      "loss": 1.744016456604004,
      "step": 70620,
      "train_bpb": 0.6200061717587104,
      "train_bytes": 56169.30625,
      "train_loss_nats_per_token": 1.7439665910391142
    },
    {
      "epoch": 0.06346948987385109,
      "grad_norm": 0.09765302389860153,
      "learning_rate": 0.0016642024057913177,
      "loss": 1.7491018295288085,
      "step": 70640,
      "train_bpb": 0.6176018977724999,
      "train_bytes": 57193.025,
      "train_loss_nats_per_token": 1.7493124958833655
    },
    {
      "epoch": 0.0636011693134234,
      "grad_norm": 0.11611080914735794,
      "learning_rate": 0.0016635855961920206,
      "loss": 1.7449129104614258,
      "step": 70660,
      "train_bpb": 0.6192326264690533,
      "train_bytes": 56440.03125,
      "train_loss_nats_per_token": 1.7455725097672743
    },
    {
      "epoch": 0.0637328487529957,
      "grad_norm": 0.1460847556591034,
      "learning_rate": 0.0016629687585977063,
      "loss": 1.739097785949707,
      "step": 70680,
      "train_bpb": 0.6159001536931346,
      "train_bytes": 56792.0125,
      "train_loss_nats_per_token": 1.739522967115571
    },
    {
      "epoch": 0.06386452819256801,
      "grad_norm": 0.1376514732837677,
      "learning_rate": 0.0016623518931139368,
      "loss": 1.7437063217163087,
      "step": 70700,
      "train_bpb": 0.6152075654557415,
      "train_bytes": 56502.11875,
      "train_loss_nats_per_token": 1.744027426332224
    },
    {
      "epoch": 0.06399620763214031,
      "grad_norm": 0.12957990169525146,
      "learning_rate": 0.0016617349998462778,
      "loss": 1.7469324111938476,
      "step": 70720,
      "train_bpb": 0.6190832597592519,
      "train_bytes": 56658.86875,
      "train_loss_nats_per_token": 1.7472180849594559
    },
    {
      "epoch": 0.06412788707171262,
      "grad_norm": 0.13546597957611084,
      "learning_rate": 0.001661118078900301,
      "loss": 1.7548362731933593,
      "step": 70740,
      "train_bpb": 0.6182153888493394,
      "train_bytes": 56655.26875,
      "train_loss_nats_per_token": 1.7546254715137541
    },
    {
      "epoch": 0.06425956651128492,
      "grad_norm": 0.13813762366771698,
      "learning_rate": 0.0016605011303815827,
      "loss": 1.7449028015136718,
      "step": 70760,
      "train_bpb": 0.6200660618366913,
      "train_bytes": 56137.34375,
      "train_loss_nats_per_token": 1.745447097549602
    },
    {
      "epoch": 0.06439124595085724,
      "grad_norm": 0.13575415313243866,
      "learning_rate": 0.0016598841543957034,
      "loss": 1.7423234939575196,
      "step": 70780,
      "train_bpb": 0.615964293644564,
      "train_bytes": 56967.61875,
      "train_loss_nats_per_token": 1.742311429084816
    },
    {
      "epoch": 0.06452292539042954,
      "grad_norm": 0.1525917649269104,
      "learning_rate": 0.0016592671510482484,
      "loss": 1.7458593368530273,
      "step": 70800,
      "train_bpb": 0.6181027617641297,
      "train_bytes": 56794.2125,
      "train_loss_nats_per_token": 1.7461349301458584
    },
    {
      "epoch": 0.06465460483000185,
      "grad_norm": 0.11188030242919922,
      "learning_rate": 0.0016586501204448076,
      "loss": 1.7294391632080077,
      "step": 70820,
      "train_bpb": 0.6119886426607365,
      "train_bytes": 56476.65,
      "train_loss_nats_per_token": 1.7297255453313998
    },
    {
      "epoch": 0.06478628426957415,
      "grad_norm": 0.10109905153512955,
      "learning_rate": 0.0016580330626909763,
      "loss": 1.7318866729736329,
      "step": 70840,
      "train_bpb": 0.6155574822844228,
      "train_bytes": 55971.71875,
      "train_loss_nats_per_token": 1.7323092691759738
    },
    {
      "epoch": 0.06491796370914646,
      "grad_norm": 0.13708457350730896,
      "learning_rate": 0.0016574159778923534,
      "loss": 1.7374021530151367,
      "step": 70860,
      "train_bpb": 0.6147100444195666,
      "train_bytes": 56535.1125,
      "train_loss_nats_per_token": 1.7375175564745915
    },
    {
      "epoch": 0.06504964314871876,
      "grad_norm": 0.1093808263540268,
      "learning_rate": 0.0016567988661545438,
      "loss": 1.7440547943115234,
      "step": 70880,
      "train_bpb": 0.6171141453417007,
      "train_bytes": 57230.61875,
      "train_loss_nats_per_token": 1.7441008603680765
    },
    {
      "epoch": 0.06518132258829107,
      "grad_norm": 0.131907656788826,
      "learning_rate": 0.0016561817275831552,
      "loss": 1.732374382019043,
      "step": 70900,
      "train_bpb": 0.615146737721091,
      "train_bytes": 56142.31875,
      "train_loss_nats_per_token": 1.7326922829253888
    },
    {
      "epoch": 0.06531300202786337,
      "grad_norm": 0.13402847945690155,
      "learning_rate": 0.0016555645622838018,
      "loss": 1.7278533935546876,
      "step": 70920,
      "train_bpb": 0.6107299593486855,
      "train_bytes": 56364.76875,
      "train_loss_nats_per_token": 1.7282653293037038
    },
    {
      "epoch": 0.06544468146743568,
      "grad_norm": 0.1502784937620163,
      "learning_rate": 0.0016549473703620999,
      "loss": 1.7302972793579101,
      "step": 70940,
      "train_bpb": 0.6146126903053569,
      "train_bytes": 56537.76875,
      "train_loss_nats_per_token": 1.730223615416695
    },
    {
      "epoch": 0.06557636090700798,
      "grad_norm": 0.18078316748142242,
      "learning_rate": 0.0016543301519236736,
      "loss": 1.734229850769043,
      "step": 70960,
      "train_bpb": 0.6141017771060868,
      "train_bytes": 56502.43125,
      "train_loss_nats_per_token": 1.7346525125915466
    },
    {
      "epoch": 0.06570804034658029,
      "grad_norm": 0.12838701903820038,
      "learning_rate": 0.0016537129070741483,
      "loss": 1.7317693710327149,
      "step": 70980,
      "train_bpb": 0.6138193708580245,
      "train_bytes": 56220.04375,
      "train_loss_nats_per_token": 1.7319386251444162
    },
    {
      "epoch": 0.06583971978615259,
      "grad_norm": 0.1093895435333252,
      "learning_rate": 0.0016530956359191561,
      "loss": 1.728123664855957,
      "step": 71000,
      "train_bpb": 0.6105586230902067,
      "train_bytes": 57043.35,
      "train_loss_nats_per_token": 1.7282908279299771
    },
    {
      "epoch": 0.06583971978615259,
      "eval_loss": 1.7092033624649048,
      "eval_runtime": 75.7277,
      "eval_samples_per_second": 13.205,
      "eval_steps_per_second": 13.205,
      "step": 71000,
      "train_bpb": 0.6025922458621559,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.715949344885276
    },
    {
      "epoch": 0.0659713992257249,
      "grad_norm": 0.11919566243886948,
      "learning_rate": 0.001652478338564333,
      "loss": 1.724302101135254,
      "step": 71020,
      "train_bpb": 0.6117374809768809,
      "train_bytes": 56258.7375,
      "train_loss_nats_per_token": 1.724505361433923
    },
    {
      "epoch": 0.0661030786652972,
      "grad_norm": 0.15619410574436188,
      "learning_rate": 0.00165186101511532,
      "loss": 1.726461410522461,
      "step": 71040,
      "train_bpb": 0.6093908417187494,
      "train_bytes": 56966.9,
      "train_loss_nats_per_token": 1.7266057121214669
    },
    {
      "epoch": 0.0662347581048695,
      "grad_norm": 0.1135774478316307,
      "learning_rate": 0.0016512436656777607,
      "loss": 1.7218477249145507,
      "step": 71060,
      "train_bpb": 0.6092447383856506,
      "train_bytes": 56899.9,
      "train_loss_nats_per_token": 1.7220879233533668
    },
    {
      "epoch": 0.06636643754444181,
      "grad_norm": 0.14298568665981293,
      "learning_rate": 0.001650626290357305,
      "loss": 1.7199974060058594,
      "step": 71080,
      "train_bpb": 0.6101781735925247,
      "train_bytes": 56398.225,
      "train_loss_nats_per_token": 1.7200728717688805
    },
    {
      "epoch": 0.06649811698401412,
      "grad_norm": 0.09896348416805267,
      "learning_rate": 0.0016500088892596066,
      "loss": 1.7274444580078125,
      "step": 71100,
      "train_bpb": 0.6118988684837444,
      "train_bytes": 56332.4375,
      "train_loss_nats_per_token": 1.7274979112740982
    },
    {
      "epoch": 0.06662979642358642,
      "grad_norm": 0.15846510231494904,
      "learning_rate": 0.0016493914624903243,
      "loss": 1.7200828552246095,
      "step": 71120,
      "train_bpb": 0.6085765182013464,
      "train_bytes": 56063.675,
      "train_loss_nats_per_token": 1.7209894034203321
    },
    {
      "epoch": 0.06676147586315873,
      "grad_norm": 0.18469329178333282,
      "learning_rate": 0.00164877401015512,
      "loss": 1.7234525680541992,
      "step": 71140,
      "train_bpb": 0.6112608960506302,
      "train_bytes": 56582.425,
      "train_loss_nats_per_token": 1.72394963773795
    },
    {
      "epoch": 0.06689315530273103,
      "grad_norm": 0.13529664278030396,
      "learning_rate": 0.0016481565323596605,
      "loss": 1.732782745361328,
      "step": 71160,
      "train_bpb": 0.6113875785829038,
      "train_bytes": 56997.05,
      "train_loss_nats_per_token": 1.7330524729816492
    },
    {
      "epoch": 0.06702483474230334,
      "grad_norm": 0.10092106461524963,
      "learning_rate": 0.0016475390292096175,
      "loss": 1.7226762771606445,
      "step": 71180,
      "train_bpb": 0.6074358284404875,
      "train_bytes": 56832.825,
      "train_loss_nats_per_token": 1.7230423656369795
    },
    {
      "epoch": 0.06715651418187564,
      "grad_norm": 0.14729367196559906,
      "learning_rate": 0.001646921500810667,
      "loss": 1.72320556640625,
      "step": 71200,
      "train_bpb": 0.6081825530066466,
      "train_bytes": 56785.31875,
      "train_loss_nats_per_token": 1.7234105453259798
    },
    {
      "epoch": 0.06728819362144794,
      "grad_norm": 0.1045590415596962,
      "learning_rate": 0.001646303947268488,
      "loss": 1.7256038665771485,
      "step": 71220,
      "train_bpb": 0.608943317838007,
      "train_bytes": 56936.70625,
      "train_loss_nats_per_token": 1.7257288810909603
    },
    {
      "epoch": 0.06741987306102025,
      "grad_norm": 0.11669188737869263,
      "learning_rate": 0.0016456863686887652,
      "loss": 1.7214010238647461,
      "step": 71240,
      "train_bpb": 0.6103645246998539,
      "train_bytes": 56337.40625,
      "train_loss_nats_per_token": 1.7220287525938902
    },
    {
      "epoch": 0.06755155250059255,
      "grad_norm": 0.11126547306776047,
      "learning_rate": 0.0016450687651771872,
      "loss": 1.7150197982788087,
      "step": 71260,
      "train_bpb": 0.6071630393895509,
      "train_bytes": 56460.14375,
      "train_loss_nats_per_token": 1.7155794793591064
    },
    {
      "epoch": 0.06768323194016486,
      "grad_norm": 0.1291969120502472,
      "learning_rate": 0.0016444511368394468,
      "loss": 1.715569305419922,
      "step": 71280,
      "train_bpb": 0.6085523321541806,
      "train_bytes": 56352.50625,
      "train_loss_nats_per_token": 1.7157522906986251
    },
    {
      "epoch": 0.06781491137973716,
      "grad_norm": 0.12901835143566132,
      "learning_rate": 0.0016438334837812411,
      "loss": 1.7156465530395508,
      "step": 71300,
      "train_bpb": 0.6084001484620973,
      "train_bytes": 56433.86875,
      "train_loss_nats_per_token": 1.7161198458896971
    },
    {
      "epoch": 0.06794659081930947,
      "grad_norm": 0.1267179399728775,
      "learning_rate": 0.0016432158061082715,
      "loss": 1.7197763442993164,
      "step": 71320,
      "train_bpb": 0.6077062293708,
      "train_bytes": 57072.13125,
      "train_loss_nats_per_token": 1.7200563498560202
    },
    {
      "epoch": 0.06807827025888177,
      "grad_norm": 0.11829151213169098,
      "learning_rate": 0.001642598103926243,
      "loss": 1.7226211547851562,
      "step": 71340,
      "train_bpb": 0.6099430547258506,
      "train_bytes": 56532.575,
      "train_loss_nats_per_token": 1.7228017594491085
    },
    {
      "epoch": 0.06820994969845408,
      "grad_norm": 0.11463771015405655,
      "learning_rate": 0.0016419803773408657,
      "loss": 1.7218378067016602,
      "step": 71360,
      "train_bpb": 0.6120618539150146,
      "train_bytes": 56184.2375,
      "train_loss_nats_per_token": 1.7221156205536312
    },
    {
      "epoch": 0.06834162913802638,
      "grad_norm": 0.09724868834018707,
      "learning_rate": 0.0016413626264578534,
      "loss": 1.7229814529418945,
      "step": 71380,
      "train_bpb": 0.6103464550334596,
      "train_bytes": 56799.8875,
      "train_loss_nats_per_token": 1.7227684744309246
    },
    {
      "epoch": 0.06847330857759869,
      "grad_norm": 0.09809409826993942,
      "learning_rate": 0.0016407448513829238,
      "loss": 1.7184297561645507,
      "step": 71400,
      "train_bpb": 0.6108586800933717,
      "train_bytes": 56239.04375,
      "train_loss_nats_per_token": 1.718839100102499
    },
    {
      "epoch": 0.06860498801717099,
      "grad_norm": 0.1500450223684311,
      "learning_rate": 0.0016401270522217995,
      "loss": 1.7198804855346679,
      "step": 71420,
      "train_bpb": 0.6092755659131849,
      "train_bytes": 56865.58125,
      "train_loss_nats_per_token": 1.7201146806015526
    },
    {
      "epoch": 0.0687366674567433,
      "grad_norm": 0.11599515378475189,
      "learning_rate": 0.001639509229080207,
      "loss": 1.7190250396728515,
      "step": 71440,
      "train_bpb": 0.6059822561369491,
      "train_bytes": 56989.7875,
      "train_loss_nats_per_token": 1.719485505872185
    },
    {
      "epoch": 0.06886834689631562,
      "grad_norm": 0.1430763453245163,
      "learning_rate": 0.001638891382063876,
      "loss": 1.7030101776123048,
      "step": 71460,
      "train_bpb": 0.604173371473607,
      "train_bytes": 56659.33125,
      "train_loss_nats_per_token": 1.7030915959266086
    },
    {
      "epoch": 0.06900002633588792,
      "grad_norm": 0.0973486453294754,
      "learning_rate": 0.0016382735112785412,
      "loss": 1.7186796188354492,
      "step": 71480,
      "train_bpb": 0.6081657472407901,
      "train_bytes": 56952.71875,
      "train_loss_nats_per_token": 1.7189721911431586
    },
    {
      "epoch": 0.06913170577546023,
      "grad_norm": 0.1076570525765419,
      "learning_rate": 0.0016376556168299413,
      "loss": 1.7165727615356445,
      "step": 71500,
      "train_bpb": 0.6045361803205116,
      "train_bytes": 57100.9875,
      "train_loss_nats_per_token": 1.7167424022519884
    },
    {
      "epoch": 0.06926338521503253,
      "grad_norm": 0.12373335659503937,
      "learning_rate": 0.0016370376988238196,
      "loss": 1.7227867126464844,
      "step": 71520,
      "train_bpb": 0.6085833517510988,
      "train_bytes": 56597.475,
      "train_loss_nats_per_token": 1.7229993801969161
    },
    {
      "epoch": 0.06939506465460483,
      "grad_norm": 0.12103434652090073,
      "learning_rate": 0.0016364197573659213,
      "loss": 1.7057281494140626,
      "step": 71540,
      "train_bpb": 0.6053079145973422,
      "train_bytes": 56078.70625,
      "train_loss_nats_per_token": 1.7058833424083462
    },
    {
      "epoch": 0.06952674409417714,
      "grad_norm": 0.10595158487558365,
      "learning_rate": 0.001635801792561998,
      "loss": 1.710403823852539,
      "step": 71560,
      "train_bpb": 0.6058205621963992,
      "train_bytes": 56812.1375,
      "train_loss_nats_per_token": 1.7105977606692377
    },
    {
      "epoch": 0.06965842353374944,
      "grad_norm": 0.11148687452077866,
      "learning_rate": 0.0016351838045178038,
      "loss": 1.7094232559204101,
      "step": 71580,
      "train_bpb": 0.6055061822212908,
      "train_bytes": 56703.91875,
      "train_loss_nats_per_token": 1.7097280856574515
    },
    {
      "epoch": 0.06979010297332175,
      "grad_norm": 0.12053634971380234,
      "learning_rate": 0.0016345657933390981,
      "loss": 1.713827133178711,
      "step": 71600,
      "train_bpb": 0.603892384137233,
      "train_bytes": 56905.95625,
      "train_loss_nats_per_token": 1.714241784882135
    },
    {
      "epoch": 0.06992178241289405,
      "grad_norm": 0.1122678741812706,
      "learning_rate": 0.0016339477591316425,
      "loss": 1.717440414428711,
      "step": 71620,
      "train_bpb": 0.6102322694953081,
      "train_bytes": 56292.8125,
      "train_loss_nats_per_token": 1.717616268539819
    },
    {
      "epoch": 0.07005346185246636,
      "grad_norm": 0.10120486468076706,
      "learning_rate": 0.0016333297020012043,
      "loss": 1.7225513458251953,
      "step": 71640,
      "train_bpb": 0.61121048317341,
      "train_bytes": 56836.48125,
      "train_loss_nats_per_token": 1.7229750840203661
    },
    {
      "epoch": 0.07018514129203866,
      "grad_norm": 0.17207732796669006,
      "learning_rate": 0.0016327116220535536,
      "loss": 1.7349845886230468,
      "step": 71660,
      "train_bpb": 0.6145903199986357,
      "train_bytes": 56086.7625,
      "train_loss_nats_per_token": 1.7347633651226644
    },
    {
      "epoch": 0.07031682073161097,
      "grad_norm": 0.11795572191476822,
      "learning_rate": 0.0016320935193944643,
      "loss": 1.7330524444580078,
      "step": 71680,
      "train_bpb": 0.614798482300221,
      "train_bytes": 56718.08125,
      "train_loss_nats_per_token": 1.733380718681982
    },
    {
      "epoch": 0.07044850017118327,
      "grad_norm": 0.10767831653356552,
      "learning_rate": 0.001631475394129716,
      "loss": 1.7416828155517579,
      "step": 71700,
      "train_bpb": 0.6188487496778015,
      "train_bytes": 56419.2625,
      "train_loss_nats_per_token": 1.7416232737082304
    },
    {
      "epoch": 0.07058017961075558,
      "grad_norm": 0.11837542802095413,
      "learning_rate": 0.001630857246365089,
      "loss": 1.74798526763916,
      "step": 71720,
      "train_bpb": 0.6162879874937426,
      "train_bytes": 56814.8,
      "train_loss_nats_per_token": 1.7479281994154792
    },
    {
      "epoch": 0.07071185905032788,
      "grad_norm": 0.12259206920862198,
      "learning_rate": 0.0016302390762063703,
      "loss": 1.7506153106689453,
      "step": 71740,
      "train_bpb": 0.6186883874564812,
      "train_bytes": 56948.1,
      "train_loss_nats_per_token": 1.7509363924609516
    },
    {
      "epoch": 0.07084353848990019,
      "grad_norm": 0.11306197941303253,
      "learning_rate": 0.0016296208837593496,
      "loss": 1.7488338470458984,
      "step": 71760,
      "train_bpb": 0.6162607311573529,
      "train_bytes": 56986.775,
      "train_loss_nats_per_token": 1.749336721684199
    },
    {
      "epoch": 0.07097521792947249,
      "grad_norm": 0.11263972520828247,
      "learning_rate": 0.001629002669129821,
      "loss": 1.7487022399902343,
      "step": 71780,
      "train_bpb": 0.6215779662693873,
      "train_bytes": 56289.68125,
      "train_loss_nats_per_token": 1.7488157402855165
    },
    {
      "epoch": 0.0711068973690448,
      "grad_norm": 0.1300438642501831,
      "learning_rate": 0.0016283844324235805,
      "loss": 1.7524848937988282,
      "step": 71800,
      "train_bpb": 0.6155699254534809,
      "train_bytes": 56935.6875,
      "train_loss_nats_per_token": 1.7525125538398105
    },
    {
      "epoch": 0.0712385768086171,
      "grad_norm": 0.15386012196540833,
      "learning_rate": 0.0016277661737464303,
      "loss": 1.7368692398071288,
      "step": 71820,
      "train_bpb": 0.6153376896648146,
      "train_bytes": 56513.8,
      "train_loss_nats_per_token": 1.7369547754161698
    },
    {
      "epoch": 0.0713702562481894,
      "grad_norm": 0.1374281346797943,
      "learning_rate": 0.0016271478932041751,
      "loss": 1.744148063659668,
      "step": 71840,
      "train_bpb": 0.6145342341015109,
      "train_bytes": 56839.9375,
      "train_loss_nats_per_token": 1.7439506266814335
    },
    {
      "epoch": 0.07150193568776171,
      "grad_norm": 0.1284988820552826,
      "learning_rate": 0.001626529590902624,
      "loss": 1.7425407409667968,
      "step": 71860,
      "train_bpb": 0.6150197115604336,
      "train_bytes": 56646.05,
      "train_loss_nats_per_token": 1.7427704207597707
    },
    {
      "epoch": 0.07163361512733402,
      "grad_norm": 0.10182293504476547,
      "learning_rate": 0.0016259112669475884,
      "loss": 1.7423788070678712,
      "step": 71880,
      "train_bpb": 0.6153247670561129,
      "train_bytes": 56537.175,
      "train_loss_nats_per_token": 1.7424803210877673
    },
    {
      "epoch": 0.07176529456690632,
      "grad_norm": 0.13725511729717255,
      "learning_rate": 0.0016252929214448848,
      "loss": 1.7372566223144532,
      "step": 71900,
      "train_bpb": 0.6132936695496747,
      "train_bytes": 57149.7875,
      "train_loss_nats_per_token": 1.737975780747699
    },
    {
      "epoch": 0.07189697400647863,
      "grad_norm": 0.08208240568637848,
      "learning_rate": 0.0016246745545003339,
      "loss": 1.7482172012329102,
      "step": 71920,
      "train_bpb": 0.6172820785385734,
      "train_bytes": 56559.3625,
      "train_loss_nats_per_token": 1.7478683407324802
    },
    {
      "epoch": 0.07202865344605093,
      "grad_norm": 0.11876673996448517,
      "learning_rate": 0.0016240561662197578,
      "loss": 1.7321107864379883,
      "step": 71940,
      "train_bpb": 0.6128042754341014,
      "train_bytes": 56405.1875,
      "train_loss_nats_per_token": 1.7323290029958267
    },
    {
      "epoch": 0.07216033288562324,
      "grad_norm": 0.1616792231798172,
      "learning_rate": 0.0016234377567089847,
      "loss": 1.7484973907470702,
      "step": 71960,
      "train_bpb": 0.6172946291742573,
      "train_bytes": 57264.21875,
      "train_loss_nats_per_token": 1.7489317075865471
    },
    {
      "epoch": 0.07229201232519554,
      "grad_norm": 0.1094522476196289,
      "learning_rate": 0.0016228193260738447,
      "loss": 1.7411638259887696,
      "step": 71980,
      "train_bpb": 0.6155997397109138,
      "train_bytes": 56884.71875,
      "train_loss_nats_per_token": 1.7412159669421714
    },
    {
      "epoch": 0.07242369176476784,
      "grad_norm": 0.09437184780836105,
      "learning_rate": 0.0016222008744201722,
      "loss": 1.7375732421875,
      "step": 72000,
      "train_bpb": 0.6143334385262982,
      "train_bytes": 56584.13125,
      "train_loss_nats_per_token": 1.73765470490181
    },
    {
      "epoch": 0.07242369176476784,
      "eval_loss": 1.7067722082138062,
      "eval_runtime": 79.2499,
      "eval_samples_per_second": 12.618,
      "eval_steps_per_second": 12.618,
      "step": 72000,
      "train_bpb": 0.60171914566472,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.7134630936563626
    },
    {
      "epoch": 0.00013167943957230517,
      "grad_norm": 0.15449965000152588,
      "learning_rate": 0.0016215824018538053,
      "loss": 1.725496482849121,
      "step": 72020,
      "train_bpb": 0.6091751735509844,
      "train_bytes": 57285.4875,
      "train_loss_nats_per_token": 1.726056466136722
    },
    {
      "epoch": 0.00026335887914461035,
      "grad_norm": 0.12627114355564117,
      "learning_rate": 0.0016209639084805857,
      "loss": 1.723215103149414,
      "step": 72040,
      "train_bpb": 0.6104391500380569,
      "train_bytes": 56225.5125,
      "train_loss_nats_per_token": 1.7233375745663202
    },
    {
      "epoch": 0.00039503831871691555,
      "grad_norm": 0.10122272372245789,
      "learning_rate": 0.0016203453944063576,
      "loss": 1.7076606750488281,
      "step": 72060,
      "train_bpb": 0.6031983406548199,
      "train_bytes": 56917.51875,
      "train_loss_nats_per_token": 1.7081855473104817
    },
    {
      "epoch": 0.0005267177582892207,
      "grad_norm": 0.13552120327949524,
      "learning_rate": 0.0016197268597369708,
      "loss": 1.7114341735839844,
      "step": 72080,
      "train_bpb": 0.6061487056094189,
      "train_bytes": 56260.525,
      "train_loss_nats_per_token": 1.7114655649658777
    },
    {
      "epoch": 0.0006583971978615259,
      "grad_norm": 0.13187319040298462,
      "learning_rate": 0.0016191083045782773,
      "loss": 1.7146881103515625,
      "step": 72100,
      "train_bpb": 0.6052567469578705,
      "train_bytes": 56795.56875,
      "train_loss_nats_per_token": 1.7151290516246989
    },
    {
      "epoch": 0.0007900766374338311,
      "grad_norm": 0.12753134965896606,
      "learning_rate": 0.001618489729036132,
      "loss": 1.709431266784668,
      "step": 72120,
      "train_bpb": 0.6022253949259472,
      "train_bytes": 56929.99375,
      "train_loss_nats_per_token": 1.7095728998130832
    },
    {
      "epoch": 0.0009217560770061363,
      "grad_norm": 0.0886593610048294,
      "learning_rate": 0.0016178711332163946,
      "loss": 1.7021554946899413,
      "step": 72140,
      "train_bpb": 0.6005572001518301,
      "train_bytes": 56964.9875,
      "train_loss_nats_per_token": 1.7026401252135541
    },
    {
      "epoch": 0.0010534355165784414,
      "grad_norm": 0.31866204738616943,
      "learning_rate": 0.0016172525172249276,
      "loss": 1.7134624481201173,
      "step": 72160,
      "train_bpb": 0.6068890613093276,
      "train_bytes": 56336.2625,
      "train_loss_nats_per_token": 1.7136100696132897
    },
    {
      "epoch": 0.0011851149561507466,
      "grad_norm": 0.22569464147090912,
      "learning_rate": 0.0016166338811675974,
      "loss": 1.7874439239501954,
      "step": 72180,
      "train_bpb": 0.6330766184147425,
      "train_bytes": 56659.75,
      "train_loss_nats_per_token": 1.788064778311506
    },
    {
      "epoch": 0.0013167943957230518,
      "grad_norm": 0.15891608595848083,
      "learning_rate": 0.0016160152251502727,
      "loss": 1.7477151870727539,
      "step": 72200,
      "train_bpb": 0.6158914091005846,
      "train_bytes": 56591.2375,
      "train_loss_nats_per_token": 1.747744970879215
    },
    {
      "epoch": 0.001448473835295357,
      "grad_norm": 0.1306278556585312,
      "learning_rate": 0.0016153965492788276,
      "loss": 1.721139144897461,
      "step": 72220,
      "train_bpb": 0.6093591217463528,
      "train_bytes": 56428.475,
      "train_loss_nats_per_token": 1.7215227149210937
    },
    {
      "epoch": 0.0015801532748676622,
      "grad_norm": 0.10164865106344223,
      "learning_rate": 0.0016147778536591373,
      "loss": 1.7148557662963868,
      "step": 72240,
      "train_bpb": 0.6065040706320316,
      "train_bytes": 56637.1375,
      "train_loss_nats_per_token": 1.7149202027392019
    },
    {
      "epoch": 0.0017118327144399674,
      "grad_norm": 0.10884436964988708,
      "learning_rate": 0.0016141591383970822,
      "loss": 1.7097047805786132,
      "step": 72260,
      "train_bpb": 0.603235428907351,
      "train_bytes": 56402.74375,
      "train_loss_nats_per_token": 1.7100790382203002
    },
    {
      "epoch": 0.0018435121540122726,
      "grad_norm": 0.13649609684944153,
      "learning_rate": 0.001613540403598545,
      "loss": 1.703007698059082,
      "step": 72280,
      "train_bpb": 0.6038516014115232,
      "train_bytes": 56227.9125,
      "train_loss_nats_per_token": 1.7032614361508744
    },
    {
      "epoch": 0.0019751915935845776,
      "grad_norm": 0.11490056663751602,
      "learning_rate": 0.0016129216493694124,
      "loss": 1.7086153030395508,
      "step": 72300,
      "train_bpb": 0.6035022995157558,
      "train_bytes": 56507.2,
      "train_loss_nats_per_token": 1.7093328022680854
    },
    {
      "epoch": 0.002106871033156883,
      "grad_norm": 0.13819620013237,
      "learning_rate": 0.001612302875815574,
      "loss": 1.7154592514038085,
      "step": 72320,
      "train_bpb": 0.6086047029872668,
      "train_bytes": 56519.20625,
      "train_loss_nats_per_token": 1.7158322526952936
    },
    {
      "epoch": 0.002238550472729188,
      "grad_norm": 0.12772710621356964,
      "learning_rate": 0.0016116840830429227,
      "loss": 1.6987539291381837,
      "step": 72340,
      "train_bpb": 0.601109093888735,
      "train_bytes": 56722.08125,
      "train_loss_nats_per_token": 1.6988834702954234
    },
    {
      "epoch": 0.002370229912301493,
      "grad_norm": 0.14496876299381256,
      "learning_rate": 0.001611065271157355,
      "loss": 1.7073875427246095,
      "step": 72360,
      "train_bpb": 0.6037231040715965,
      "train_bytes": 56245.75625,
      "train_loss_nats_per_token": 1.707412164453968
    },
    {
      "epoch": 0.0025019093518737984,
      "grad_norm": 0.10883032530546188,
      "learning_rate": 0.0016104464402647706,
      "loss": 1.6967897415161133,
      "step": 72380,
      "train_bpb": 0.5983027296788535,
      "train_bytes": 56647.2625,
      "train_loss_nats_per_token": 1.6971707848180926
    },
    {
      "epoch": 0.0026335887914461036,
      "grad_norm": 0.09636475890874863,
      "learning_rate": 0.0016098275904710716,
      "loss": 1.701212501525879,
      "step": 72400,
      "train_bpb": 0.6008284114300351,
      "train_bytes": 56474.55,
      "train_loss_nats_per_token": 1.7013851335828811
    },
    {
      "epoch": 0.002765268231018409,
      "grad_norm": 0.11111840605735779,
      "learning_rate": 0.0016092087218821656,
      "loss": 1.6979570388793945,
      "step": 72420,
      "train_bpb": 0.6010211417240093,
      "train_bytes": 56596.68125,
      "train_loss_nats_per_token": 1.6984958842768476
    },
    {
      "epoch": 0.002896947670590714,
      "grad_norm": 0.12723204493522644,
      "learning_rate": 0.0016085898346039605,
      "loss": 1.699591827392578,
      "step": 72440,
      "train_bpb": 0.6000812893431525,
      "train_bytes": 56797.4375,
      "train_loss_nats_per_token": 1.6999242111011923
    },
    {
      "epoch": 0.003028627110163019,
      "grad_norm": 0.16377851366996765,
      "learning_rate": 0.0016079709287423693,
      "loss": 1.6973066329956055,
      "step": 72460,
      "train_bpb": 0.6005314734778388,
      "train_bytes": 56435.13125,
      "train_loss_nats_per_token": 1.698201026224559
    },
    {
      "epoch": 0.0031603065497353244,
      "grad_norm": 0.13833080232143402,
      "learning_rate": 0.0016073520044033078,
      "loss": 1.6920793533325196,
      "step": 72480,
      "train_bpb": 0.5985669862924708,
      "train_bytes": 56318.2375,
      "train_loss_nats_per_token": 1.692286136352397
    },
    {
      "epoch": 0.0032919859893076296,
      "grad_norm": 0.10332290828227997,
      "learning_rate": 0.0016067330616926945,
      "loss": 1.6931743621826172,
      "step": 72500,
      "train_bpb": 0.5981398847692849,
      "train_bytes": 56514.575,
      "train_loss_nats_per_token": 1.6937234015645726
    },
    {
      "epoch": 0.003423665428879935,
      "grad_norm": 0.11617614328861237,
      "learning_rate": 0.0016061141007164522,
      "loss": 1.6989873886108398,
      "step": 72520,
      "train_bpb": 0.6008447463677097,
      "train_bytes": 56427.11875,
      "train_loss_nats_per_token": 1.699138935705853
    },
    {
      "epoch": 0.00355534486845224,
      "grad_norm": 0.09954865276813507,
      "learning_rate": 0.001605495121580505,
      "loss": 1.6973705291748047,
      "step": 72540,
      "train_bpb": 0.596505904687564,
      "train_bytes": 57043.95625,
      "train_loss_nats_per_token": 1.6972544408291232
    },
    {
      "epoch": 0.0036870243080245452,
      "grad_norm": 0.12463074922561646,
      "learning_rate": 0.0016048761243907818,
      "loss": 1.6974618911743165,
      "step": 72560,
      "train_bpb": 0.5989444007879423,
      "train_bytes": 56782.99375,
      "train_loss_nats_per_token": 1.6975921293428122
    },
    {
      "epoch": 0.0038187037475968504,
      "grad_norm": 0.10350430756807327,
      "learning_rate": 0.0016042571092532132,
      "loss": 1.690636444091797,
      "step": 72580,
      "train_bpb": 0.5981325630720336,
      "train_bytes": 56805.15625,
      "train_loss_nats_per_token": 1.6909003367245345
    },
    {
      "epoch": 0.003950383187169155,
      "grad_norm": 0.13256074488162994,
      "learning_rate": 0.0016036380762737348,
      "loss": 1.6859991073608398,
      "step": 72600,
      "train_bpb": 0.597013656231901,
      "train_bytes": 56589.7875,
      "train_loss_nats_per_token": 1.6868884481610575
    },
    {
      "epoch": 0.004082062626741461,
      "grad_norm": 0.10376463830471039,
      "learning_rate": 0.001603019025558283,
      "loss": 1.691042709350586,
      "step": 72620,
      "train_bpb": 0.5943725190608985,
      "train_bytes": 56854.9125,
      "train_loss_nats_per_token": 1.691295352038584
    },
    {
      "epoch": 0.004213742066313766,
      "grad_norm": 0.1583099216222763,
      "learning_rate": 0.0016023999572127986,
      "loss": 1.684712028503418,
      "step": 72640,
      "train_bpb": 0.5968347774124692,
      "train_bytes": 56336.725,
      "train_loss_nats_per_token": 1.6852795957300482
    },
    {
      "epoch": 0.004345421505886071,
      "grad_norm": 0.11464938521385193,
      "learning_rate": 0.001601780871343225,
      "loss": 1.6776395797729493,
      "step": 72660,
      "train_bpb": 0.593727140468275,
      "train_bytes": 56340.48125,
      "train_loss_nats_per_token": 1.6778796860293814
    },
    {
      "epoch": 0.004477100945458376,
      "grad_norm": 0.14877471327781677,
      "learning_rate": 0.001601161768055509,
      "loss": 1.6856382369995118,
      "step": 72680,
      "train_bpb": 0.594993200192308,
      "train_bytes": 56740.71875,
      "train_loss_nats_per_token": 1.6859890877330137
    },
    {
      "epoch": 0.004608780385030682,
      "grad_norm": 0.15782245993614197,
      "learning_rate": 0.0016005426474555997,
      "loss": 1.6960277557373047,
      "step": 72700,
      "train_bpb": 0.598645971946582,
      "train_bytes": 56807.6375,
      "train_loss_nats_per_token": 1.6960674172007766
    },
    {
      "epoch": 0.004740459824602986,
      "grad_norm": 0.10218894481658936,
      "learning_rate": 0.0015999235096494492,
      "loss": 1.6943191528320312,
      "step": 72720,
      "train_bpb": 0.5994310939101938,
      "train_bytes": 56490.375,
      "train_loss_nats_per_token": 1.6942806810930007
    },
    {
      "epoch": 0.004872139264175292,
      "grad_norm": 0.12775462865829468,
      "learning_rate": 0.0015993043547430136,
      "loss": 1.6895364761352538,
      "step": 72740,
      "train_bpb": 0.5958638177705871,
      "train_bytes": 57022.98125,
      "train_loss_nats_per_token": 1.6897934395277285
    },
    {
      "epoch": 0.005003818703747597,
      "grad_norm": 0.12436585128307343,
      "learning_rate": 0.0015986851828422514,
      "loss": 1.6877307891845703,
      "step": 72760,
      "train_bpb": 0.5938062580319832,
      "train_bytes": 56616.2625,
      "train_loss_nats_per_token": 1.6877418066940648
    },
    {
      "epoch": 0.0051354981433199024,
      "grad_norm": 0.1008092388510704,
      "learning_rate": 0.0015980659940531229,
      "loss": 1.685641098022461,
      "step": 72780,
      "train_bpb": 0.5954617317598055,
      "train_bytes": 56404.51875,
      "train_loss_nats_per_token": 1.6858830628654178
    },
    {
      "epoch": 0.005267177582892207,
      "grad_norm": 0.17052553594112396,
      "learning_rate": 0.0015974467884815927,
      "loss": 1.6841962814331055,
      "step": 72800,
      "train_bpb": 0.5978342619190784,
      "train_bytes": 56361.63125,
      "train_loss_nats_per_token": 1.6847115148198968
    },
    {
      "epoch": 0.005398857022464512,
      "grad_norm": 0.13889636099338531,
      "learning_rate": 0.0015968275662336274,
      "loss": 1.6835893630981444,
      "step": 72820,
      "train_bpb": 0.5964955160107622,
      "train_bytes": 56126.94375,
      "train_loss_nats_per_token": 1.6839768190137738
    },
    {
      "epoch": 0.005530536462036818,
      "grad_norm": 0.13935205340385437,
      "learning_rate": 0.0015962083274151978,
      "loss": 1.6845739364624024,
      "step": 72840,
      "train_bpb": 0.5966641695329269,
      "train_bytes": 56335.15625,
      "train_loss_nats_per_token": 1.6848704855262153
    },
    {
      "epoch": 0.005662215901609122,
      "grad_norm": 0.14388425648212433,
      "learning_rate": 0.0015955890721322758,
      "loss": 1.6780147552490234,
      "step": 72860,
      "train_bpb": 0.592713296119889,
      "train_bytes": 56943.975,
      "train_loss_nats_per_token": 1.6785052682534338
    },
    {
      "epoch": 0.005793895341181428,
      "grad_norm": 0.1418432742357254,
      "learning_rate": 0.001594969800490837,
      "loss": 1.677579116821289,
      "step": 72880,
      "train_bpb": 0.5939071092977515,
      "train_bytes": 56260.18125,
      "train_loss_nats_per_token": 1.678432918849168
    },
    {
      "epoch": 0.005925574780753733,
      "grad_norm": 0.126267671585083,
      "learning_rate": 0.0015943505125968593,
      "loss": 1.6756200790405273,
      "step": 72900,
      "train_bpb": 0.5943658836639266,
      "train_bytes": 56512.725,
      "train_loss_nats_per_token": 1.6762914283283932
    },
    {
      "epoch": 0.006057254220326038,
      "grad_norm": 0.11142408102750778,
      "learning_rate": 0.0015937312085563245,
      "loss": 1.6847494125366211,
      "step": 72920,
      "train_bpb": 0.5945321390649752,
      "train_bytes": 57069.98125,
      "train_loss_nats_per_token": 1.685054379213251
    },
    {
      "epoch": 0.006188933659898343,
      "grad_norm": 0.14940614998340607,
      "learning_rate": 0.0015931118884752156,
      "loss": 1.6771839141845704,
      "step": 72940,
      "train_bpb": 0.5948652890074361,
      "train_bytes": 56464.46875,
      "train_loss_nats_per_token": 1.6775037770494992
    },
    {
      "epoch": 0.006320613099470649,
      "grad_norm": 0.09686575084924698,
      "learning_rate": 0.0015924925524595204,
      "loss": 1.6759767532348633,
      "step": 72960,
      "train_bpb": 0.5939348747239063,
      "train_bytes": 56099.6125,
      "train_loss_nats_per_token": 1.6762316205134384
    },
    {
      "epoch": 0.006452292539042954,
      "grad_norm": 0.12070279568433762,
      "learning_rate": 0.0015918732006152273,
      "loss": 1.6790449142456054,
      "step": 72980,
      "train_bpb": 0.5946351798986301,
      "train_bytes": 56585.25625,
      "train_loss_nats_per_token": 1.6800523336267634
    },
    {
      "epoch": 0.006583971978615259,
      "grad_norm": 0.11836624890565872,
      "learning_rate": 0.001591253833048328,
      "loss": 1.6910829544067383,
      "step": 73000,
      "train_bpb": 0.5981143828668243,
      "train_bytes": 56809.4875,
      "train_loss_nats_per_token": 1.6911956675035107
    },
    {
      "epoch": 0.006583971978615259,
      "eval_loss": 1.5881518125534058,
      "eval_runtime": 77.8099,
      "eval_samples_per_second": 12.852,
      "eval_steps_per_second": 12.852,
      "step": 73000,
      "train_bpb": 0.5601338796282249,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5950443610852034
    },
    {
      "epoch": 0.00013167943957230517,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.0016029141173155728,
      "loss": 1.592186737060547,
      "step": 73020,
      "train_bpb": 0.5599881403174174,
      "train_bytes": 28694.4625,
      "train_loss_nats_per_token": 1.5932105636219485
    },
    {
      "epoch": 0.00026335887914461035,
      "grad_norm": 0.140625,
      "learning_rate": 0.001602291932649563,
      "loss": 1.6086257934570312,
      "step": 73040,
      "train_bpb": 0.5706939774614106,
      "train_bytes": 28591.025,
      "train_loss_nats_per_token": 1.6104061516763877
    },
    {
      "epoch": 0.00039503831871691555,
      "grad_norm": 0.201171875,
      "learning_rate": 0.0016016697303014378,
      "loss": 1.5989774703979491,
      "step": 73060,
      "train_bpb": 0.5676456127440689,
      "train_bytes": 27972.325,
      "train_loss_nats_per_token": 1.6003658039525157
    },
    {
      "epoch": 0.0005267177582892207,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.0016010475103787497,
      "loss": 1.595706844329834,
      "step": 73080,
      "train_bpb": 0.5646298466515326,
      "train_bytes": 28253.1875,
      "train_loss_nats_per_token": 1.59614655598072
    },
    {
      "epoch": 0.0006583971978615259,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0016004252729890556,
      "loss": 1.5810138702392578,
      "step": 73100,
      "train_bpb": 0.5591420001623094,
      "train_bytes": 28480.175,
      "train_loss_nats_per_token": 1.5822857819845237
    },
    {
      "epoch": 0.0007900766374338311,
      "grad_norm": 0.19140625,
      "learning_rate": 0.0015998030182399151,
      "loss": 1.587076187133789,
      "step": 73120,
      "train_bpb": 0.5604837484446706,
      "train_bytes": 28437.34375,
      "train_loss_nats_per_token": 1.5883665017345823
    },
    {
      "epoch": 0.0009217560770061363,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0015991807462388907,
      "loss": 1.5789700508117677,
      "step": 73140,
      "train_bpb": 0.5617871232698132,
      "train_bytes": 27932.15,
      "train_loss_nats_per_token": 1.5792691564310128
    },
    {
      "epoch": 0.0010534355165784414,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0015985584570935483,
      "loss": 1.5959596633911133,
      "step": 73160,
      "train_bpb": 0.563014167146227,
      "train_bytes": 28328.375,
      "train_loss_nats_per_token": 1.5965939944002183
    },
    {
      "epoch": 0.0011851149561507466,
      "grad_norm": 0.126953125,
      "learning_rate": 0.001597936150911456,
      "loss": 1.5944327354431151,
      "step": 73180,
      "train_bpb": 0.5644041647604972,
      "train_bytes": 28231.58125,
      "train_loss_nats_per_token": 1.596213823182813
    },
    {
      "epoch": 0.0013167943957230518,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.001597313827800185,
      "loss": 1.5846613883972167,
      "step": 73200,
      "train_bpb": 0.5582516107872993,
      "train_bytes": 28563.9875,
      "train_loss_nats_per_token": 1.5850214806463594
    },
    {
      "epoch": 0.001448473835295357,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00159669148786731,
      "loss": 1.5893366813659668,
      "step": 73220,
      "train_bpb": 0.5591593800172568,
      "train_bytes": 28503.2375,
      "train_loss_nats_per_token": 1.5889347792028603
    },
    {
      "epoch": 0.0015801532748676622,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0015960691312204083,
      "loss": 1.582364845275879,
      "step": 73240,
      "train_bpb": 0.5582639832260436,
      "train_bytes": 28426.75625,
      "train_loss_nats_per_token": 1.5831626480498469
    },
    {
      "epoch": 0.0017118327144399674,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0015954467579670601,
      "loss": 1.5725903511047363,
      "step": 73260,
      "train_bpb": 0.5559986728500919,
      "train_bytes": 28296.7625,
      "train_loss_nats_per_token": 1.5741684470592063
    },
    {
      "epoch": 0.0018435121540122726,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0015948243682148483,
      "loss": 1.5882252693176269,
      "step": 73280,
      "train_bpb": 0.5598052856772612,
      "train_bytes": 28668.225,
      "train_loss_nats_per_token": 1.5892405945602297
    },
    {
      "epoch": 0.0019751915935845776,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0015942019620713587,
      "loss": 1.5795911788940429,
      "step": 73300,
      "train_bpb": 0.5606502595530488,
      "train_bytes": 28055.3625,
      "train_loss_nats_per_token": 1.5802993426900829
    },
    {
      "epoch": 0.002106871033156883,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0015935795396441802,
      "loss": 1.5844932556152345,
      "step": 73320,
      "train_bpb": 0.5602806886470049,
      "train_bytes": 28280.9,
      "train_loss_nats_per_token": 1.5847435389526907
    },
    {
      "epoch": 0.002238550472729188,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.001592957101040904,
      "loss": 1.6027862548828125,
      "step": 73340,
      "train_bpb": 0.5653479681935242,
      "train_bytes": 28529.8,
      "train_loss_nats_per_token": 1.6037244698281736
    },
    {
      "epoch": 0.002370229912301493,
      "grad_norm": 0.171875,
      "learning_rate": 0.0015923346463691253,
      "loss": 1.5963197708129884,
      "step": 73360,
      "train_bpb": 0.5677904243266144,
      "train_bytes": 28129.95,
      "train_loss_nats_per_token": 1.59664957142697
    },
    {
      "epoch": 0.0025019093518737984,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0015917121757364396,
      "loss": 1.5965024948120117,
      "step": 73380,
      "train_bpb": 0.5597450877664376,
      "train_bytes": 28551.35625,
      "train_loss_nats_per_token": 1.5969119617574403
    },
    {
      "epoch": 0.0026335887914461036,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0015910896892504486,
      "loss": 1.5944787979125976,
      "step": 73400,
      "train_bpb": 0.5652599910109858,
      "train_bytes": 28039.88125,
      "train_loss_nats_per_token": 1.5954226918295644
    },
    {
      "epoch": 0.002765268231018409,
      "grad_norm": 0.142578125,
      "learning_rate": 0.001590467187018754,
      "loss": 1.5803529739379882,
      "step": 73420,
      "train_bpb": 0.5587901739811837,
      "train_bytes": 28373.16875,
      "train_loss_nats_per_token": 1.5819137839904636
    },
    {
      "epoch": 0.002896947670590714,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0015898446691489606,
      "loss": 1.5897476196289062,
      "step": 73440,
      "train_bpb": 0.5641648602802536,
      "train_bytes": 28055.30625,
      "train_loss_nats_per_token": 1.5905326868196887
    },
    {
      "epoch": 0.003028627110163019,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0015892221357486777,
      "loss": 1.5847281455993651,
      "step": 73460,
      "train_bpb": 0.5606552358575446,
      "train_bytes": 28229.7125,
      "train_loss_nats_per_token": 1.5851970873722565
    },
    {
      "epoch": 0.0031603065497353244,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0015885995869255147,
      "loss": 1.586777114868164,
      "step": 73480,
      "train_bpb": 0.5613727963765612,
      "train_bytes": 28407.425,
      "train_loss_nats_per_token": 1.5873921618015034
    },
    {
      "epoch": 0.0032919859893076296,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.001587977022787086,
      "loss": 1.5865592956542969,
      "step": 73500,
      "train_bpb": 0.561959891270781,
      "train_bytes": 28123.025,
      "train_loss_nats_per_token": 1.5873968212509781
    },
    {
      "epoch": 0.003423665428879935,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.0015873544434410068,
      "loss": 1.582630443572998,
      "step": 73520,
      "train_bpb": 0.5566792007314799,
      "train_bytes": 28279.71875,
      "train_loss_nats_per_token": 1.5837273283612001
    },
    {
      "epoch": 0.00355534486845224,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0015867318489948966,
      "loss": 1.584315586090088,
      "step": 73540,
      "train_bpb": 0.561889635884137,
      "train_bytes": 28275.45625,
      "train_loss_nats_per_token": 1.584448973231327
    },
    {
      "epoch": 0.0036870243080245452,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0015861092395563759,
      "loss": 1.5738457679748534,
      "step": 73560,
      "train_bpb": 0.558196107799709,
      "train_bytes": 27952.45625,
      "train_loss_nats_per_token": 1.5749371146101163
    },
    {
      "epoch": 0.0038187037475968504,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0015854866152330692,
      "loss": 1.589449405670166,
      "step": 73580,
      "train_bpb": 0.5619597211596302,
      "train_bytes": 28375.91875,
      "train_loss_nats_per_token": 1.5911625232173465
    },
    {
      "epoch": 0.003950383187169155,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0015848639761326022,
      "loss": 1.5813647270202638,
      "step": 73600,
      "train_bpb": 0.5585138753605037,
      "train_bytes": 28131.28125,
      "train_loss_nats_per_token": 1.5824181553548304
    },
    {
      "epoch": 0.004082062626741461,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0015842413223626048,
      "loss": 1.594572162628174,
      "step": 73620,
      "train_bpb": 0.5656825517102925,
      "train_bytes": 28159.8125,
      "train_loss_nats_per_token": 1.5955778038450814
    },
    {
      "epoch": 0.004213742066313766,
      "grad_norm": 0.216796875,
      "learning_rate": 0.0015836186540307076,
      "loss": 1.5894320487976075,
      "step": 73640,
      "train_bpb": 0.564431334714108,
      "train_bytes": 28359.39375,
      "train_loss_nats_per_token": 1.590546985071196
    },
    {
      "epoch": 0.004345421505886071,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0015829959712445453,
      "loss": 1.5876875877380372,
      "step": 73660,
      "train_bpb": 0.5608448736333357,
      "train_bytes": 28469.91875,
      "train_loss_nats_per_token": 1.5877379663501627
    },
    {
      "epoch": 0.004477100945458376,
      "grad_norm": 0.234375,
      "learning_rate": 0.0015823732741117541,
      "loss": 1.5697440147399901,
      "step": 73680,
      "train_bpb": 0.5565767771535242,
      "train_bytes": 28252.1625,
      "train_loss_nats_per_token": 1.5703816880095134
    },
    {
      "epoch": 0.004608780385030682,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0015817505627399734,
      "loss": 1.5881710052490234,
      "step": 73700,
      "train_bpb": 0.5620400833259139,
      "train_bytes": 28079.51875,
      "train_loss_nats_per_token": 1.589467205841437
    },
    {
      "epoch": 0.004740459824602986,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.001581127837236844,
      "loss": 1.580660629272461,
      "step": 73720,
      "train_bpb": 0.5590396599244891,
      "train_bytes": 28166.2375,
      "train_loss_nats_per_token": 1.581100357806571
    },
    {
      "epoch": 0.004872139264175292,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0015805050977100103,
      "loss": 1.577392864227295,
      "step": 73740,
      "train_bpb": 0.5554805053013988,
      "train_bytes": 28459.70625,
      "train_loss_nats_per_token": 1.5781767405761897
    },
    {
      "epoch": 0.005003818703747597,
      "grad_norm": 0.259765625,
      "learning_rate": 0.0015798823442671188,
      "loss": 1.574063205718994,
      "step": 73760,
      "train_bpb": 0.5560501011778881,
      "train_bytes": 28187.55625,
      "train_loss_nats_per_token": 1.5748193122840677
    },
    {
      "epoch": 0.0051354981433199024,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0015792595770158177,
      "loss": 1.5776287078857423,
      "step": 73780,
      "train_bpb": 0.5572755264243989,
      "train_bytes": 28293.2125,
      "train_loss_nats_per_token": 1.5783474085513425
    },
    {
      "epoch": 0.005267177582892207,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0015786367960637585,
      "loss": 1.5812874794006349,
      "step": 73800,
      "train_bpb": 0.5589773562947692,
      "train_bytes": 28181.3375,
      "train_loss_nats_per_token": 1.5825798271273435
    },
    {
      "epoch": 0.005398857022464512,
      "grad_norm": 0.15625,
      "learning_rate": 0.001578014001518595,
      "loss": 1.5758980751037597,
      "step": 73820,
      "train_bpb": 0.5566159121216685,
      "train_bytes": 28359.1,
      "train_loss_nats_per_token": 1.5771097580336526
    },
    {
      "epoch": 0.005530536462036818,
      "grad_norm": 0.1640625,
      "learning_rate": 0.001577391193487982,
      "loss": 1.5748394012451172,
      "step": 73840,
      "train_bpb": 0.5590531266898022,
      "train_bytes": 28237.58125,
      "train_loss_nats_per_token": 1.5757755903859092
    },
    {
      "epoch": 0.005662215901609122,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0015767683720795788,
      "loss": 1.578141212463379,
      "step": 73860,
      "train_bpb": 0.5576983832665993,
      "train_bytes": 28213.18125,
      "train_loss_nats_per_token": 1.5795581900185156
    },
    {
      "epoch": 0.005793895341181428,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0015761455374010453,
      "loss": 1.577780532836914,
      "step": 73880,
      "train_bpb": 0.5572845071833317,
      "train_bytes": 28584.25625,
      "train_loss_nats_per_token": 1.5789871952338266
    },
    {
      "epoch": 0.005925574780753733,
      "grad_norm": 0.12890625,
      "learning_rate": 0.001575522689560045,
      "loss": 1.575568389892578,
      "step": 73900,
      "train_bpb": 0.5577287429732204,
      "train_bytes": 28175.975,
      "train_loss_nats_per_token": 1.576483619401618
    },
    {
      "epoch": 0.006057254220326038,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0015748998286642414,
      "loss": 1.5794594764709473,
      "step": 73920,
      "train_bpb": 0.5589077219881732,
      "train_bytes": 28259.15625,
      "train_loss_nats_per_token": 1.5811746949313352
    },
    {
      "epoch": 0.006188933659898343,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.001574276954821304,
      "loss": 1.5750330924987792,
      "step": 73940,
      "train_bpb": 0.5582957920628506,
      "train_bytes": 27874.025,
      "train_loss_nats_per_token": 1.5757405908570157
    },
    {
      "epoch": 0.006320613099470649,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0015736540681389006,
      "loss": 1.5678739547729492,
      "step": 73960,
      "train_bpb": 0.5538328218213514,
      "train_bytes": 28444.2125,
      "train_loss_nats_per_token": 1.5684359043791458
    },
    {
      "epoch": 0.006452292539042954,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0015730311687247036,
      "loss": 1.5716561317443847,
      "step": 73980,
      "train_bpb": 0.5539903875349109,
      "train_bytes": 28337.29375,
      "train_loss_nats_per_token": 1.5727397192388193
    },
    {
      "epoch": 0.006583971978615259,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0015724082566863874,
      "loss": 1.5730464935302735,
      "step": 74000,
      "train_bpb": 0.5573164511853499,
      "train_bytes": 28177.28125,
      "train_loss_nats_per_token": 1.5740670786065045
    },
    {
      "epoch": 0.006583971978615259,
      "eval_loss": 1.477449655532837,
      "eval_runtime": 72.8312,
      "eval_samples_per_second": 13.73,
      "eval_steps_per_second": 13.73,
      "step": 74000,
      "train_bpb": 0.5216204844879215,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.4853731271552733
    },
    {
      "epoch": 0.006715651418187564,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.001571785332131627,
      "loss": 1.5744507789611817,
      "step": 74020,
      "train_bpb": 0.5562337363829174,
      "train_bytes": 28422.79375,
      "train_loss_nats_per_token": 1.5752765578112857
    },
    {
      "epoch": 0.00684733085775987,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.001571162395168102,
      "loss": 1.5810954093933105,
      "step": 74040,
      "train_bpb": 0.5601201036906998,
      "train_bytes": 28004.325,
      "train_loss_nats_per_token": 1.581635515245129
    },
    {
      "epoch": 0.006979010297332174,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0015705394459034924,
      "loss": 1.5800589561462401,
      "step": 74060,
      "train_bpb": 0.5545474606637948,
      "train_bytes": 28615.3625,
      "train_loss_nats_per_token": 1.5811100876353916
    },
    {
      "epoch": 0.00711068973690448,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0015699164844454807,
      "loss": 1.5750792503356934,
      "step": 74080,
      "train_bpb": 0.5548507584496235,
      "train_bytes": 28428.59375,
      "train_loss_nats_per_token": 1.5754810974875266
    },
    {
      "epoch": 0.007242369176476785,
      "grad_norm": 0.205078125,
      "learning_rate": 0.0015692935109017517,
      "loss": 1.5759275436401368,
      "step": 74100,
      "train_bpb": 0.5568817643952387,
      "train_bytes": 28326.43125,
      "train_loss_nats_per_token": 1.577001501838015
    },
    {
      "epoch": 0.0073740486160490904,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0015686705253799913,
      "loss": 1.5777870178222657,
      "step": 74120,
      "train_bpb": 0.5564597636998894,
      "train_bytes": 28456.5625,
      "train_loss_nats_per_token": 1.578544914712156
    },
    {
      "epoch": 0.007505728055621395,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00156804752798789,
      "loss": 1.5707834243774415,
      "step": 74140,
      "train_bpb": 0.558129414990807,
      "train_bytes": 28408.86875,
      "train_loss_nats_per_token": 1.5720996041278887
    },
    {
      "epoch": 0.007637407495193701,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0015674245188331383,
      "loss": 1.5707672119140625,
      "step": 74160,
      "train_bpb": 0.5536355399140668,
      "train_bytes": 28396.2875,
      "train_loss_nats_per_token": 1.5708198483404252
    },
    {
      "epoch": 0.007769086934766006,
      "grad_norm": 0.203125,
      "learning_rate": 0.0015668014980234279,
      "loss": 1.5606981277465821,
      "step": 74180,
      "train_bpb": 0.5538207382261222,
      "train_bytes": 28162.23125,
      "train_loss_nats_per_token": 1.5626381454580394
    },
    {
      "epoch": 0.00790076637433831,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0015661784656664554,
      "loss": 1.5741880416870118,
      "step": 74200,
      "train_bpb": 0.5568233925866455,
      "train_bytes": 28427.55625,
      "train_loss_nats_per_token": 1.575533333207125
    },
    {
      "epoch": 0.008032445813910615,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0015655554218699165,
      "loss": 1.5715331077575683,
      "step": 74220,
      "train_bpb": 0.5518584027902971,
      "train_bytes": 28702.625,
      "train_loss_nats_per_token": 1.5724491921795825
    },
    {
      "epoch": 0.008164125253482922,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0015649323667415109,
      "loss": 1.5791749000549316,
      "step": 74240,
      "train_bpb": 0.5561635112104989,
      "train_bytes": 28152.2875,
      "train_loss_nats_per_token": 1.5803901635186557
    },
    {
      "epoch": 0.008295804693055226,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0015643093003889393,
      "loss": 1.5692078590393066,
      "step": 74260,
      "train_bpb": 0.5550204172363222,
      "train_bytes": 28206.275,
      "train_loss_nats_per_token": 1.5705053779508573
    },
    {
      "epoch": 0.008427484132627531,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0015636862229199042,
      "loss": 1.5650456428527832,
      "step": 74280,
      "train_bpb": 0.5558786511490051,
      "train_bytes": 28130.45,
      "train_loss_nats_per_token": 1.566335065893594
    },
    {
      "epoch": 0.008559163572199836,
      "grad_norm": 0.19921875,
      "learning_rate": 0.001563063134442111,
      "loss": 1.5672988891601562,
      "step": 74300,
      "train_bpb": 0.5535600973571339,
      "train_bytes": 28267.69375,
      "train_loss_nats_per_token": 1.5681328532755165
    },
    {
      "epoch": 0.008690843011772142,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0015624400350632657,
      "loss": 1.550775146484375,
      "step": 74320,
      "train_bpb": 0.5504566075595797,
      "train_bytes": 28072.7875,
      "train_loss_nats_per_token": 1.5518442172369467
    },
    {
      "epoch": 0.008822522451344447,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.001561816924891077,
      "loss": 1.564197063446045,
      "step": 74340,
      "train_bpb": 0.5535591836597159,
      "train_bytes": 28460.01875,
      "train_loss_nats_per_token": 1.5648949188094827
    },
    {
      "epoch": 0.008954201890916752,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0015611938040332562,
      "loss": 1.5656596183776856,
      "step": 74360,
      "train_bpb": 0.5517018916543559,
      "train_bytes": 28280.7,
      "train_loss_nats_per_token": 1.5670316800905286
    },
    {
      "epoch": 0.009085881330489057,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0015605706725975142,
      "loss": 1.583462619781494,
      "step": 74380,
      "train_bpb": 0.5585583146972305,
      "train_bytes": 28369.1,
      "train_loss_nats_per_token": 1.5838620104965704
    },
    {
      "epoch": 0.009217560770061363,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0015599475306915661,
      "loss": 1.5689911842346191,
      "step": 74400,
      "train_bpb": 0.5544911468375859,
      "train_bytes": 28438.5375,
      "train_loss_nats_per_token": 1.5696149603656444
    },
    {
      "epoch": 0.009349240209633668,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0015593243784231268,
      "loss": 1.5685805320739745,
      "step": 74420,
      "train_bpb": 0.5558463096787437,
      "train_bytes": 28460.14375,
      "train_loss_nats_per_token": 1.5702001797375915
    },
    {
      "epoch": 0.009480919649205973,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0015587012158999154,
      "loss": 1.5755216598510742,
      "step": 74440,
      "train_bpb": 0.5573074686488436,
      "train_bytes": 28030.23125,
      "train_loss_nats_per_token": 1.5761250191474656
    },
    {
      "epoch": 0.009612599088778278,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0015580780432296497,
      "loss": 1.564333152770996,
      "step": 74460,
      "train_bpb": 0.5522971700002125,
      "train_bytes": 28535.60625,
      "train_loss_nats_per_token": 1.5654459611801201
    },
    {
      "epoch": 0.009744278528350584,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0015574548605200526,
      "loss": 1.5712315559387207,
      "step": 74480,
      "train_bpb": 0.5541527227911907,
      "train_bytes": 28487.375,
      "train_loss_nats_per_token": 1.5723090218140265
    },
    {
      "epoch": 0.009875957967922889,
      "grad_norm": 0.1640625,
      "learning_rate": 0.001556831667878846,
      "loss": 1.5666656494140625,
      "step": 74500,
      "train_bpb": 0.552333623789545,
      "train_bytes": 28114.975,
      "train_loss_nats_per_token": 1.5681377378127197
    },
    {
      "epoch": 0.010007637407495194,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0015562084654137545,
      "loss": 1.5616464614868164,
      "step": 74520,
      "train_bpb": 0.5491347139391304,
      "train_bytes": 28501.2875,
      "train_loss_nats_per_token": 1.5624735777659593
    },
    {
      "epoch": 0.010139316847067498,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.001555585253232505,
      "loss": 1.563005256652832,
      "step": 74540,
      "train_bpb": 0.5507192316876633,
      "train_bytes": 28419.2625,
      "train_loss_nats_per_token": 1.5641013887552873
    },
    {
      "epoch": 0.010270996286639805,
      "grad_norm": 0.15234375,
      "learning_rate": 0.001554962031442825,
      "loss": 1.5604401588439942,
      "step": 74560,
      "train_bpb": 0.553492905750133,
      "train_bytes": 27985.25625,
      "train_loss_nats_per_token": 1.5620963813636548
    },
    {
      "epoch": 0.01040267572621211,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.001554338800152445,
      "loss": 1.5647913932800293,
      "step": 74580,
      "train_bpb": 0.5552533402198039,
      "train_bytes": 28187.1875,
      "train_loss_nats_per_token": 1.5657178170419053
    },
    {
      "epoch": 0.010534355165784414,
      "grad_norm": 0.15625,
      "learning_rate": 0.0015537155594690958,
      "loss": 1.559356689453125,
      "step": 74600,
      "train_bpb": 0.5542230174088485,
      "train_bytes": 28174.44375,
      "train_loss_nats_per_token": 1.560816269017442
    },
    {
      "epoch": 0.01066603460535672,
      "grad_norm": 0.2001953125,
      "learning_rate": 0.0015530923095005103,
      "loss": 1.5635217666625976,
      "step": 74620,
      "train_bpb": 0.5513200419319951,
      "train_bytes": 28165.88125,
      "train_loss_nats_per_token": 1.564306806548203
    },
    {
      "epoch": 0.010797714044929024,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0015524690503544232,
      "loss": 1.556387996673584,
      "step": 74640,
      "train_bpb": 0.5541877481896906,
      "train_bytes": 27961.0625,
      "train_loss_nats_per_token": 1.5566525287539923
    },
    {
      "epoch": 0.01092939348450133,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0015518457821385705,
      "loss": 1.554613208770752,
      "step": 74660,
      "train_bpb": 0.5539171639574071,
      "train_bytes": 27954.55,
      "train_loss_nats_per_token": 1.5560995479062423
    },
    {
      "epoch": 0.011061072924073635,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0015512225049606908,
      "loss": 1.5648545265197753,
      "step": 74680,
      "train_bpb": 0.5516196598853382,
      "train_bytes": 28380.60625,
      "train_loss_nats_per_token": 1.565662018057668
    },
    {
      "epoch": 0.01119275236364594,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0015505992189285224,
      "loss": 1.5547642707824707,
      "step": 74700,
      "train_bpb": 0.547843172992642,
      "train_bytes": 28643.375,
      "train_loss_nats_per_token": 1.55568603761146
    },
    {
      "epoch": 0.011324431803218245,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.001549975924149806,
      "loss": 1.5564884185791015,
      "step": 74720,
      "train_bpb": 0.551440398182455,
      "train_bytes": 28300.6,
      "train_loss_nats_per_token": 1.5573200329976051
    },
    {
      "epoch": 0.011456111242790551,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.0015493526207322847,
      "loss": 1.5572917938232422,
      "step": 74740,
      "train_bpb": 0.5535949708036367,
      "train_bytes": 28045.7,
      "train_loss_nats_per_token": 1.5587709814865358
    },
    {
      "epoch": 0.011587790682362856,
      "grad_norm": 0.240234375,
      "learning_rate": 0.0015487293087837017,
      "loss": 1.548112392425537,
      "step": 74760,
      "train_bpb": 0.5460967292252629,
      "train_bytes": 28214.48125,
      "train_loss_nats_per_token": 1.5489828675298742
    },
    {
      "epoch": 0.01171947012193516,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0015481059884118028,
      "loss": 1.5530743598937988,
      "step": 74780,
      "train_bpb": 0.5500423190124716,
      "train_bytes": 28652.88125,
      "train_loss_nats_per_token": 1.5540916540493122
    },
    {
      "epoch": 0.011851149561507466,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0015474826597243348,
      "loss": 1.5470980644226073,
      "step": 74800,
      "train_bpb": 0.5504128433171094,
      "train_bytes": 27859.84375,
      "train_loss_nats_per_token": 1.5494532344387058
    },
    {
      "epoch": 0.011982829001079772,
      "grad_norm": 0.185546875,
      "learning_rate": 0.001546859322829045,
      "loss": 1.5589518547058105,
      "step": 74820,
      "train_bpb": 0.5495562192071525,
      "train_bytes": 28587.7375,
      "train_loss_nats_per_token": 1.5600657296533875
    },
    {
      "epoch": 0.012114508440652077,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0015462359778336838,
      "loss": 1.5604893684387207,
      "step": 74840,
      "train_bpb": 0.5519293226883446,
      "train_bytes": 28482.24375,
      "train_loss_nats_per_token": 1.561810759025593
    },
    {
      "epoch": 0.012246187880224382,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0015456126248460016,
      "loss": 1.5527351379394532,
      "step": 74860,
      "train_bpb": 0.5517418499045682,
      "train_bytes": 28330.6625,
      "train_loss_nats_per_token": 1.5544932656505142
    },
    {
      "epoch": 0.012377867319796686,
      "grad_norm": 0.166015625,
      "learning_rate": 0.001544989263973752,
      "loss": 1.5497429847717286,
      "step": 74880,
      "train_bpb": 0.5490881968079216,
      "train_bytes": 28133.80625,
      "train_loss_nats_per_token": 1.5498228063949484
    },
    {
      "epoch": 0.012509546759368993,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0015443658953246872,
      "loss": 1.5383386611938477,
      "step": 74900,
      "train_bpb": 0.5454707644723968,
      "train_bytes": 28137.425,
      "train_loss_nats_per_token": 1.5395291745952453
    },
    {
      "epoch": 0.012641226198941298,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0015437425190065635,
      "loss": 1.5592269897460938,
      "step": 74920,
      "train_bpb": 0.552839411903882,
      "train_bytes": 27962.1875,
      "train_loss_nats_per_token": 1.5601732361608407
    },
    {
      "epoch": 0.012772905638513602,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0015431191351271368,
      "loss": 1.5533102989196776,
      "step": 74940,
      "train_bpb": 0.5489894472640253,
      "train_bytes": 28338.075,
      "train_loss_nats_per_token": 1.5549823753708083
    },
    {
      "epoch": 0.012904585078085907,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0015424957437941646,
      "loss": 1.5557566642761231,
      "step": 74960,
      "train_bpb": 0.552691386546919,
      "train_bytes": 28247.18125,
      "train_loss_nats_per_token": 1.5576320572330036
    },
    {
      "epoch": 0.013036264517658214,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0015418723451154064,
      "loss": 1.582798385620117,
      "step": 74980,
      "train_bpb": 0.5601996444626259,
      "train_bytes": 28274.875,
      "train_loss_nats_per_token": 1.583609778763813
    },
    {
      "epoch": 0.013167943957230518,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0015412489391986221,
      "loss": 1.600499153137207,
      "step": 75000,
      "train_bpb": 0.566216897723917,
      "train_bytes": 28534.6125,
      "train_loss_nats_per_token": 1.6013850815541846
    },
    {
      "epoch": 0.013167943957230518,
      "eval_loss": 1.4872945547103882,
      "eval_runtime": 78.4431,
      "eval_samples_per_second": 12.748,
      "eval_steps_per_second": 12.748,
      "step": 75000,
      "train_bpb": 0.5250395812876856,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.4951093906965687
    },
    {
      "epoch": 0.013299623396802823,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0015406255261515734,
      "loss": 1.6694866180419923,
      "step": 75020,
      "train_bpb": 0.5936911148401095,
      "train_bytes": 28019.63125,
      "train_loss_nats_per_token": 1.6701546453360645
    },
    {
      "epoch": 0.013431302836375128,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0015400021060820233,
      "loss": 1.666590690612793,
      "step": 75040,
      "train_bpb": 0.5880051324875046,
      "train_bytes": 28213.3625,
      "train_loss_nats_per_token": 1.6671732623202768
    },
    {
      "epoch": 0.013562982275947433,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0015393786790977354,
      "loss": 1.6741701126098634,
      "step": 75060,
      "train_bpb": 0.5913699015546002,
      "train_bytes": 28467.725,
      "train_loss_nats_per_token": 1.674363846972515
    },
    {
      "epoch": 0.01369466171551974,
      "grad_norm": 0.12890625,
      "learning_rate": 0.001538755245306475,
      "loss": 1.6748661041259765,
      "step": 75080,
      "train_bpb": 0.5903863345791336,
      "train_bytes": 28419.8875,
      "train_loss_nats_per_token": 1.6750129988804967
    },
    {
      "epoch": 0.013826341155092044,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.0015381318048160084,
      "loss": 1.6750896453857422,
      "step": 75100,
      "train_bpb": 0.5921392993364749,
      "train_bytes": 28361.78125,
      "train_loss_nats_per_token": 1.675275814823876
    },
    {
      "epoch": 0.013958020594664349,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0015375083577341033,
      "loss": 1.6645305633544922,
      "step": 75120,
      "train_bpb": 0.592222150410819,
      "train_bytes": 28004.26875,
      "train_loss_nats_per_token": 1.665352992610043
    },
    {
      "epoch": 0.014089700034236654,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0015368849041685285,
      "loss": 1.6599254608154297,
      "step": 75140,
      "train_bpb": 0.5884418656117291,
      "train_bytes": 28285.0375,
      "train_loss_nats_per_token": 1.6600235836164843
    },
    {
      "epoch": 0.01422137947380896,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0015362614442270526,
      "loss": 1.6665157318115233,
      "step": 75160,
      "train_bpb": 0.5931283970519643,
      "train_bytes": 27960.9125,
      "train_loss_nats_per_token": 1.6675763940399584
    },
    {
      "epoch": 0.014353058913381265,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0015356379780174483,
      "loss": 1.6711013793945313,
      "step": 75180,
      "train_bpb": 0.58903316282213,
      "train_bytes": 28560.14375,
      "train_loss_nats_per_token": 1.6716448993158968
    },
    {
      "epoch": 0.01448473835295357,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0015350145056474862,
      "loss": 1.6568284988403321,
      "step": 75200,
      "train_bpb": 0.5850687071267588,
      "train_bytes": 28466.29375,
      "train_loss_nats_per_token": 1.6574252590793224
    },
    {
      "epoch": 0.014616417792525874,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00153439102722494,
      "loss": 1.6590381622314454,
      "step": 75220,
      "train_bpb": 0.5855794122071784,
      "train_bytes": 28553.6375,
      "train_loss_nats_per_token": 1.6591056390485943
    },
    {
      "epoch": 0.014748097232098181,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0015337675428575837,
      "loss": 1.6632837295532226,
      "step": 75240,
      "train_bpb": 0.5901135330217718,
      "train_bytes": 28192.5625,
      "train_loss_nats_per_token": 1.6640000851409558
    },
    {
      "epoch": 0.014879776671670486,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0015331440526531917,
      "loss": 1.6791948318481444,
      "step": 75260,
      "train_bpb": 0.5942014790362995,
      "train_bytes": 28330.4125,
      "train_loss_nats_per_token": 1.6799095762340643
    },
    {
      "epoch": 0.01501145611124279,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.001532520556719541,
      "loss": 1.6906164169311524,
      "step": 75280,
      "train_bpb": 0.5955115756148572,
      "train_bytes": 28626.925,
      "train_loss_nats_per_token": 1.69123779729944
    },
    {
      "epoch": 0.015143135550815095,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0015318970551644081,
      "loss": 1.6911014556884765,
      "step": 75300,
      "train_bpb": 0.5963748137841562,
      "train_bytes": 28595.59375,
      "train_loss_nats_per_token": 1.6909372035653074
    },
    {
      "epoch": 0.015274814990387402,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0015312735480955716,
      "loss": 1.6717477798461915,
      "step": 75320,
      "train_bpb": 0.5910263667468326,
      "train_bytes": 28284.09375,
      "train_loss_nats_per_token": 1.6726712377243502
    },
    {
      "epoch": 0.015406494429959706,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0015306500356208103,
      "loss": 1.687736701965332,
      "step": 75340,
      "train_bpb": 0.5977369187849892,
      "train_bytes": 28533.725,
      "train_loss_nats_per_token": 1.6886036822422326
    },
    {
      "epoch": 0.015538173869532011,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0015300265178479035,
      "loss": 1.7004064559936523,
      "step": 75360,
      "train_bpb": 0.6013878829997871,
      "train_bytes": 28277.59375,
      "train_loss_nats_per_token": 1.701292763668272
    },
    {
      "epoch": 0.015669853309104316,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0015294029948846335,
      "loss": 1.7031085968017579,
      "step": 75380,
      "train_bpb": 0.6021286389709439,
      "train_bytes": 28377.0875,
      "train_loss_nats_per_token": 1.7032068490667047
    },
    {
      "epoch": 0.01580153274867662,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0015287794668387802,
      "loss": 1.7029790878295898,
      "step": 75400,
      "train_bpb": 0.6044001604829812,
      "train_bytes": 28251.58125,
      "train_loss_nats_per_token": 1.7034242497426253
    },
    {
      "epoch": 0.015933212188248926,
      "grad_norm": 0.166015625,
      "learning_rate": 0.001528155933818128,
      "loss": 1.7043779373168946,
      "step": 75420,
      "train_bpb": 0.6022033207761586,
      "train_bytes": 28294.3125,
      "train_loss_nats_per_token": 1.7047206167027291
    },
    {
      "epoch": 0.01606489162782123,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0015275323959304593,
      "loss": 1.691830825805664,
      "step": 75440,
      "train_bpb": 0.5994582229829127,
      "train_bytes": 28286.3125,
      "train_loss_nats_per_token": 1.6916399785509781
    },
    {
      "epoch": 0.01619657106739354,
      "grad_norm": 0.13671875,
      "learning_rate": 0.001526908853283559,
      "loss": 1.6931020736694335,
      "step": 75460,
      "train_bpb": 0.6042914693703644,
      "train_bytes": 27965.14375,
      "train_loss_nats_per_token": 1.6933169015669216
    },
    {
      "epoch": 0.016328250506965843,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0015262853059852123,
      "loss": 1.708877944946289,
      "step": 75480,
      "train_bpb": 0.6057177533924603,
      "train_bytes": 28345.29375,
      "train_loss_nats_per_token": 1.7088100856803183
    },
    {
      "epoch": 0.016459929946538148,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0015256617541432047,
      "loss": 1.7118890762329102,
      "step": 75500,
      "train_bpb": 0.6064259208066819,
      "train_bytes": 28070.04375,
      "train_loss_nats_per_token": 1.7125856423455508
    },
    {
      "epoch": 0.016591609386110453,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0015250381978653231,
      "loss": 1.7042327880859376,
      "step": 75520,
      "train_bpb": 0.6052275814712949,
      "train_bytes": 28040.85,
      "train_loss_nats_per_token": 1.7044581734450361
    },
    {
      "epoch": 0.016723288825682758,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0015244146372593553,
      "loss": 1.7097885131835937,
      "step": 75540,
      "train_bpb": 0.6044691424627125,
      "train_bytes": 28057.16875,
      "train_loss_nats_per_token": 1.7105967815116407
    },
    {
      "epoch": 0.016854968265255062,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0015237910724330893,
      "loss": 1.6979747772216798,
      "step": 75560,
      "train_bpb": 0.5951063259464286,
      "train_bytes": 28672.4875,
      "train_loss_nats_per_token": 1.6979226664288585
    },
    {
      "epoch": 0.016986647704827367,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0015231675034943144,
      "loss": 1.7128398895263672,
      "step": 75580,
      "train_bpb": 0.6087631455216459,
      "train_bytes": 28486.975,
      "train_loss_nats_per_token": 1.7132155110925296
    },
    {
      "epoch": 0.017118327144399672,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00152254393055082,
      "loss": 1.7018428802490235,
      "step": 75600,
      "train_bpb": 0.60059664884993,
      "train_bytes": 28361.4125,
      "train_loss_nats_per_token": 1.7018523541248698
    },
    {
      "epoch": 0.01725000658397198,
      "grad_norm": 0.15625,
      "learning_rate": 0.0015219203537103966,
      "loss": 1.702927017211914,
      "step": 75620,
      "train_bpb": 0.6030319071485528,
      "train_bytes": 28105.75625,
      "train_loss_nats_per_token": 1.7021239630162288
    },
    {
      "epoch": 0.017381686023544285,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0015212967730808356,
      "loss": 1.7152244567871093,
      "step": 75640,
      "train_bpb": 0.6060655772898017,
      "train_bytes": 28345.79375,
      "train_loss_nats_per_token": 1.7162734969393318
    },
    {
      "epoch": 0.01751336546311659,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.0015206731887699282,
      "loss": 1.7024728775024414,
      "step": 75660,
      "train_bpb": 0.604469538498991,
      "train_bytes": 28318.4625,
      "train_loss_nats_per_token": 1.7033641905837682
    },
    {
      "epoch": 0.017645044902688894,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0015200496008854671,
      "loss": 1.7107509613037108,
      "step": 75680,
      "train_bpb": 0.6044947831573421,
      "train_bytes": 28171.9125,
      "train_loss_nats_per_token": 1.7106922329130037
    },
    {
      "epoch": 0.0177767243422612,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0015194260095352451,
      "loss": 1.7138227462768554,
      "step": 75700,
      "train_bpb": 0.6048259843771318,
      "train_bytes": 28319.45,
      "train_loss_nats_per_token": 1.7151234985579158
    },
    {
      "epoch": 0.017908403781833504,
      "grad_norm": 0.15625,
      "learning_rate": 0.0015188024148270563,
      "loss": 1.6984050750732422,
      "step": 75720,
      "train_bpb": 0.6020490670473531,
      "train_bytes": 28181.2375,
      "train_loss_nats_per_token": 1.6984627924919506
    },
    {
      "epoch": 0.01804008322140581,
      "grad_norm": 0.15625,
      "learning_rate": 0.001518178816868694,
      "loss": 1.6987266540527344,
      "step": 75740,
      "train_bpb": 0.6011490862655772,
      "train_bytes": 28430.31875,
      "train_loss_nats_per_token": 1.6991571998687154
    },
    {
      "epoch": 0.018171762660978114,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0015175552157679537,
      "loss": 1.7149166107177733,
      "step": 75760,
      "train_bpb": 0.6063636309813015,
      "train_bytes": 28360.55,
      "train_loss_nats_per_token": 1.7155465493116786
    },
    {
      "epoch": 0.018303442100550422,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0015169316116326306,
      "loss": 1.718169593811035,
      "step": 75780,
      "train_bpb": 0.607957001696512,
      "train_bytes": 28463.06875,
      "train_loss_nats_per_token": 1.7182029599618127
    },
    {
      "epoch": 0.018435121540122727,
      "grad_norm": 0.17578125,
      "learning_rate": 0.00151630800457052,
      "loss": 1.6933250427246094,
      "step": 75800,
      "train_bpb": 0.5975748036884775,
      "train_bytes": 28409.0375,
      "train_loss_nats_per_token": 1.6936963251073203
    },
    {
      "epoch": 0.01856680097969503,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0015156843946894185,
      "loss": 1.7072767257690429,
      "step": 75820,
      "train_bpb": 0.6066805984846251,
      "train_bytes": 28236.1,
      "train_loss_nats_per_token": 1.7077546119102716
    },
    {
      "epoch": 0.018698480419267336,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0015150607820971229,
      "loss": 1.692247200012207,
      "step": 75840,
      "train_bpb": 0.599306765283134,
      "train_bytes": 28129.8875,
      "train_loss_nats_per_token": 1.6927076522534195
    },
    {
      "epoch": 0.01883015985883964,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0015144371669014304,
      "loss": 1.695254135131836,
      "step": 75860,
      "train_bpb": 0.5997416398897005,
      "train_bytes": 28418.3625,
      "train_loss_nats_per_token": 1.6960154043287017
    },
    {
      "epoch": 0.018961839298411946,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0015138135492101387,
      "loss": 1.7058107376098632,
      "step": 75880,
      "train_bpb": 0.6005025190168088,
      "train_bytes": 28419.38125,
      "train_loss_nats_per_token": 1.706491738621123
    },
    {
      "epoch": 0.01909351873798425,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.001513189929131046,
      "loss": 1.7144611358642579,
      "step": 75900,
      "train_bpb": 0.6064927030863174,
      "train_bytes": 28340.45625,
      "train_loss_nats_per_token": 1.7159107102912807
    },
    {
      "epoch": 0.019225198177556555,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0015125663067719515,
      "loss": 1.6979066848754882,
      "step": 75920,
      "train_bpb": 0.6009853942990332,
      "train_bytes": 28240.90625,
      "train_loss_nats_per_token": 1.6987896518455297
    },
    {
      "epoch": 0.01935687761712886,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0015119426822406525,
      "loss": 1.704678726196289,
      "step": 75940,
      "train_bpb": 0.6049253089826331,
      "train_bytes": 27910.1375,
      "train_loss_nats_per_token": 1.705151348222937
    },
    {
      "epoch": 0.019488557056701168,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0015113190556449501,
      "loss": 1.7023496627807617,
      "step": 75960,
      "train_bpb": 0.6032852221612258,
      "train_bytes": 28313.36875,
      "train_loss_nats_per_token": 1.7024109922745034
    },
    {
      "epoch": 0.019620236496273473,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0015106954270926428,
      "loss": 1.6948934555053712,
      "step": 75980,
      "train_bpb": 0.601727889094302,
      "train_bytes": 28334.13125,
      "train_loss_nats_per_token": 1.6959967820560802
    },
    {
      "epoch": 0.019751915935845778,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0015100717966915313,
      "loss": 1.7132913589477539,
      "step": 76000,
      "train_bpb": 0.6038593236248951,
      "train_bytes": 28313.45,
      "train_loss_nats_per_token": 1.713738578344792
    },
    {
      "epoch": 0.019751915935845778,
      "eval_loss": 1.5080417394638062,
      "eval_runtime": 71.7507,
      "eval_samples_per_second": 13.937,
      "eval_steps_per_second": 13.937,
      "step": 76000,
      "train_bpb": 0.5321993268228425,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5154975731615112
    },
    {
      "epoch": 0.019883595375418082,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0015094481645494154,
      "loss": 1.6969793319702149,
      "step": 76020,
      "train_bpb": 0.603046786263758,
      "train_bytes": 27906.95625,
      "train_loss_nats_per_token": 1.6979816521980684
    },
    {
      "epoch": 0.020015274814990387,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.001508824530774096,
      "loss": 1.7078634262084962,
      "step": 76040,
      "train_bpb": 0.6024889352446463,
      "train_bytes": 28578.19375,
      "train_loss_nats_per_token": 1.7083025990454697
    },
    {
      "epoch": 0.020146954254562692,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0015082008954733746,
      "loss": 1.7014331817626953,
      "step": 76060,
      "train_bpb": 0.6011586835466952,
      "train_bytes": 28307.8125,
      "train_loss_nats_per_token": 1.7013138099816194
    },
    {
      "epoch": 0.020278633694134997,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.001507577258755051,
      "loss": 1.7017406463623046,
      "step": 76080,
      "train_bpb": 0.6018633562298512,
      "train_bytes": 28365.075,
      "train_loss_nats_per_token": 1.7029281163021974
    },
    {
      "epoch": 0.0204103131337073,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0015069536207269283,
      "loss": 1.7114662170410155,
      "step": 76100,
      "train_bpb": 0.6052986791601987,
      "train_bytes": 28283.0125,
      "train_loss_nats_per_token": 1.7123831095997148
    },
    {
      "epoch": 0.02054199257327961,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0015063299814968068,
      "loss": 1.70513916015625,
      "step": 76120,
      "train_bpb": 0.6003230204112259,
      "train_bytes": 28613.2875,
      "train_loss_nats_per_token": 1.7056187617092546
    },
    {
      "epoch": 0.020673672012851915,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0015057063411724888,
      "loss": 1.7076837539672851,
      "step": 76140,
      "train_bpb": 0.5996614326687789,
      "train_bytes": 28399.09375,
      "train_loss_nats_per_token": 1.7078480075584241
    },
    {
      "epoch": 0.02080535145242422,
      "grad_norm": 0.171875,
      "learning_rate": 0.0015050826998617769,
      "loss": 1.707564353942871,
      "step": 76160,
      "train_bpb": 0.6058346362073295,
      "train_bytes": 28225.19375,
      "train_loss_nats_per_token": 1.707915610483521
    },
    {
      "epoch": 0.020937030891996524,
      "grad_norm": 0.1875,
      "learning_rate": 0.0015044590576724724,
      "loss": 1.6868814468383788,
      "step": 76180,
      "train_bpb": 0.5944696423250446,
      "train_bytes": 28177.275,
      "train_loss_nats_per_token": 1.6873169072264074
    },
    {
      "epoch": 0.02106871033156883,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.001503835414712378,
      "loss": 1.6960824966430663,
      "step": 76200,
      "train_bpb": 0.6017489464908091,
      "train_bytes": 28179.2375,
      "train_loss_nats_per_token": 1.6969227678269305
    },
    {
      "epoch": 0.021200389771141134,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.001503211771089297,
      "loss": 1.7111825942993164,
      "step": 76220,
      "train_bpb": 0.6027817408981636,
      "train_bytes": 28621.54375,
      "train_loss_nats_per_token": 1.7120960535556735
    },
    {
      "epoch": 0.02133206921071344,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0015025881269110307,
      "loss": 1.703521156311035,
      "step": 76240,
      "train_bpb": 0.6064442054479366,
      "train_bytes": 28181.93125,
      "train_loss_nats_per_token": 1.7039459937563208
    },
    {
      "epoch": 0.021463748650285743,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0015019644822853828,
      "loss": 1.7058683395385743,
      "step": 76260,
      "train_bpb": 0.6033447203160621,
      "train_bytes": 28241.4625,
      "train_loss_nats_per_token": 1.7069218842867002
    },
    {
      "epoch": 0.021595428089858048,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0015013408373201558,
      "loss": 1.6992862701416016,
      "step": 76280,
      "train_bpb": 0.6041885052306615,
      "train_bytes": 28046.81875,
      "train_loss_nats_per_token": 1.700145063337374
    },
    {
      "epoch": 0.021727107529430356,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0015007171921231524,
      "loss": 1.6927122116088866,
      "step": 76300,
      "train_bpb": 0.6002775318447614,
      "train_bytes": 28167.90625,
      "train_loss_nats_per_token": 1.6934435266482388
    },
    {
      "epoch": 0.02185878696900266,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0015000935468021759,
      "loss": 1.6938488006591796,
      "step": 76320,
      "train_bpb": 0.5978432167611251,
      "train_bytes": 28509.86875,
      "train_loss_nats_per_token": 1.6942816022843437
    },
    {
      "epoch": 0.021990466408574966,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.001499469901465029,
      "loss": 1.694013786315918,
      "step": 76340,
      "train_bpb": 0.5989289934774129,
      "train_bytes": 28565.675,
      "train_loss_nats_per_token": 1.6948568669030661
    },
    {
      "epoch": 0.02212214584814727,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0014988462562195142,
      "loss": 1.695703887939453,
      "step": 76360,
      "train_bpb": 0.6013924768536293,
      "train_bytes": 28269.78125,
      "train_loss_nats_per_token": 1.6965109318778528
    },
    {
      "epoch": 0.022253825287719575,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0014982226111734348,
      "loss": 1.6897285461425782,
      "step": 76380,
      "train_bpb": 0.6006277051632831,
      "train_bytes": 28019.5375,
      "train_loss_nats_per_token": 1.6899565481967427
    },
    {
      "epoch": 0.02238550472729188,
      "grad_norm": 0.1953125,
      "learning_rate": 0.0014975989664345938,
      "loss": 1.701338005065918,
      "step": 76400,
      "train_bpb": 0.6000487688670104,
      "train_bytes": 28329.075,
      "train_loss_nats_per_token": 1.701400658759598
    },
    {
      "epoch": 0.022517184166864185,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0014969753221107942,
      "loss": 1.699007797241211,
      "step": 76420,
      "train_bpb": 0.600005988020775,
      "train_bytes": 28404.525,
      "train_loss_nats_per_token": 1.6995310297742416
    },
    {
      "epoch": 0.02264886360643649,
      "grad_norm": 0.212890625,
      "learning_rate": 0.0014963516783098381,
      "loss": 1.6897279739379882,
      "step": 76440,
      "train_bpb": 0.5986116690979536,
      "train_bytes": 28267.13125,
      "train_loss_nats_per_token": 1.6902935626798061
    },
    {
      "epoch": 0.022780543046008798,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0014957280351395282,
      "loss": 1.6985048294067382,
      "step": 76460,
      "train_bpb": 0.6025473653495584,
      "train_bytes": 28456.475,
      "train_loss_nats_per_token": 1.6983640879527147
    },
    {
      "epoch": 0.022912222485581103,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0014951043927076682,
      "loss": 1.6952274322509766,
      "step": 76480,
      "train_bpb": 0.5985667215462843,
      "train_bytes": 28521.25625,
      "train_loss_nats_per_token": 1.695795116854074
    },
    {
      "epoch": 0.023043901925153407,
      "grad_norm": 0.16796875,
      "learning_rate": 0.001494480751122059,
      "loss": 1.6972753524780273,
      "step": 76500,
      "train_bpb": 0.6027132231234856,
      "train_bytes": 28195.4125,
      "train_loss_nats_per_token": 1.697168834703184
    },
    {
      "epoch": 0.023175581364725712,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0014938571104905034,
      "loss": 1.7038139343261718,
      "step": 76520,
      "train_bpb": 0.5993823301019862,
      "train_bytes": 28735.65,
      "train_loss_nats_per_token": 1.7041712015236423
    },
    {
      "epoch": 0.023307260804298017,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0014932334709208041,
      "loss": 1.687881088256836,
      "step": 76540,
      "train_bpb": 0.5967475750835051,
      "train_bytes": 28234.64375,
      "train_loss_nats_per_token": 1.688416094886661
    },
    {
      "epoch": 0.02343894024387032,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0014926098325207622,
      "loss": 1.6829231262207032,
      "step": 76560,
      "train_bpb": 0.5953380132879753,
      "train_bytes": 28311.45625,
      "train_loss_nats_per_token": 1.6848042125123142
    },
    {
      "epoch": 0.023570619683442626,
      "grad_norm": 0.220703125,
      "learning_rate": 0.0014919861953981797,
      "loss": 1.7024517059326172,
      "step": 76580,
      "train_bpb": 0.6003695675074735,
      "train_bytes": 28166.75,
      "train_loss_nats_per_token": 1.7035286055730423
    },
    {
      "epoch": 0.02370229912301493,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.0014913625596608582,
      "loss": 1.7010129928588866,
      "step": 76600,
      "train_bpb": 0.6030245977850857,
      "train_bytes": 28167.475,
      "train_loss_nats_per_token": 1.7015805870194645
    },
    {
      "epoch": 0.023833978562587236,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0014907389254165994,
      "loss": 1.6888458251953125,
      "step": 76620,
      "train_bpb": 0.5974920294802359,
      "train_bytes": 28391.7,
      "train_loss_nats_per_token": 1.6890990401349033
    },
    {
      "epoch": 0.023965658002159544,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0014901152927732034,
      "loss": 1.6858325958251954,
      "step": 76640,
      "train_bpb": 0.6015469294089723,
      "train_bytes": 28031.7375,
      "train_loss_nats_per_token": 1.686665619826691
    },
    {
      "epoch": 0.02409733744173185,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0014894916618384713,
      "loss": 1.6937061309814454,
      "step": 76660,
      "train_bpb": 0.6009291010575714,
      "train_bytes": 28282.9875,
      "train_loss_nats_per_token": 1.6942106085208106
    },
    {
      "epoch": 0.024229016881304154,
      "grad_norm": 0.15625,
      "learning_rate": 0.0014888680327202041,
      "loss": 1.6931449890136718,
      "step": 76680,
      "train_bpb": 0.5994684965717954,
      "train_bytes": 28329.9875,
      "train_loss_nats_per_token": 1.6945050347245652
    },
    {
      "epoch": 0.02436069632087646,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0014882444055262008,
      "loss": 1.7016210556030273,
      "step": 76700,
      "train_bpb": 0.6010492343219654,
      "train_bytes": 28483.0625,
      "train_loss_nats_per_token": 1.7014394882406123
    },
    {
      "epoch": 0.024492375760448763,
      "grad_norm": 0.1875,
      "learning_rate": 0.001487620780364262,
      "loss": 1.6894468307495116,
      "step": 76720,
      "train_bpb": 0.597506149352703,
      "train_bytes": 28494.625,
      "train_loss_nats_per_token": 1.689495317619701
    },
    {
      "epoch": 0.024624055200021068,
      "grad_norm": 0.158203125,
      "learning_rate": 0.001486997157342187,
      "loss": 1.6953405380249023,
      "step": 76740,
      "train_bpb": 0.6010081086611093,
      "train_bytes": 28004.44375,
      "train_loss_nats_per_token": 1.6958803256788928
    },
    {
      "epoch": 0.024755734639593373,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0014863735365677753,
      "loss": 1.685969352722168,
      "step": 76760,
      "train_bpb": 0.5943258609033579,
      "train_bytes": 28198.34375,
      "train_loss_nats_per_token": 1.686332136067605
    },
    {
      "epoch": 0.024887414079165678,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0014857499181488243,
      "loss": 1.7040328979492188,
      "step": 76780,
      "train_bpb": 0.6045764965386173,
      "train_bytes": 28420.71875,
      "train_loss_nats_per_token": 1.7051891853574401
    },
    {
      "epoch": 0.025019093518737986,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0014851263021931332,
      "loss": 1.688633346557617,
      "step": 76800,
      "train_bpb": 0.599263508461098,
      "train_bytes": 28343.025,
      "train_loss_nats_per_token": 1.6887206379014172
    },
    {
      "epoch": 0.02515077295831029,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0014845026888085002,
      "loss": 1.6986400604248046,
      "step": 76820,
      "train_bpb": 0.602270836737079,
      "train_bytes": 28216.96875,
      "train_loss_nats_per_token": 1.698561151857933
    },
    {
      "epoch": 0.025282452397882595,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0014838790781027216,
      "loss": 1.7058876037597657,
      "step": 76840,
      "train_bpb": 0.599873220509381,
      "train_bytes": 28400.78125,
      "train_loss_nats_per_token": 1.705694650990934
    },
    {
      "epoch": 0.0254141318374549,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.001483255470183595,
      "loss": 1.6930063247680665,
      "step": 76860,
      "train_bpb": 0.5986707514746962,
      "train_bytes": 28069.69375,
      "train_loss_nats_per_token": 1.6935856245126923
    },
    {
      "epoch": 0.025545811277027205,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0014826318651589175,
      "loss": 1.6859165191650392,
      "step": 76880,
      "train_bpb": 0.5991865874925129,
      "train_bytes": 28176.3375,
      "train_loss_nats_per_token": 1.6862288900887301
    },
    {
      "epoch": 0.02567749071659951,
      "grad_norm": 0.171875,
      "learning_rate": 0.0014820082631364835,
      "loss": 1.6769113540649414,
      "step": 76900,
      "train_bpb": 0.5974146385792365,
      "train_bytes": 27760.1125,
      "train_loss_nats_per_token": 1.677211947510569
    },
    {
      "epoch": 0.025809170156171814,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0014813846642240892,
      "loss": 1.6901996612548829,
      "step": 76920,
      "train_bpb": 0.5969182299132318,
      "train_bytes": 28455.39375,
      "train_loss_nats_per_token": 1.6910561112724796
    },
    {
      "epoch": 0.02594084959574412,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0014807610685295296,
      "loss": 1.6899160385131835,
      "step": 76940,
      "train_bpb": 0.5998198678080428,
      "train_bytes": 28247.0,
      "train_loss_nats_per_token": 1.6906319844775828
    },
    {
      "epoch": 0.026072529035316427,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0014801374761605997,
      "loss": 1.6828527450561523,
      "step": 76960,
      "train_bpb": 0.59325101460721,
      "train_bytes": 28517.69375,
      "train_loss_nats_per_token": 1.6836876162492407
    },
    {
      "epoch": 0.026204208474888732,
      "grad_norm": 0.12890625,
      "learning_rate": 0.001479513887225092,
      "loss": 1.691169548034668,
      "step": 76980,
      "train_bpb": 0.5953969454844418,
      "train_bytes": 28591.95625,
      "train_loss_nats_per_token": 1.6914461597957424
    },
    {
      "epoch": 0.026335887914461037,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0014788903018308002,
      "loss": 1.6903409957885742,
      "step": 77000,
      "train_bpb": 0.59709138937149,
      "train_bytes": 28143.375,
      "train_loss_nats_per_token": 1.6906540234814413
    },
    {
      "epoch": 0.026335887914461037,
      "eval_loss": 1.51248037815094,
      "eval_runtime": 71.1393,
      "eval_samples_per_second": 14.057,
      "eval_steps_per_second": 14.057,
      "step": 77000,
      "train_bpb": 0.5337678942835314,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5199642458162959
    },
    {
      "epoch": 0.02646756735403334,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0014782667200855175,
      "loss": 1.6918773651123047,
      "step": 77020,
      "train_bpb": 0.5969269471336434,
      "train_bytes": 28403.1,
      "train_loss_nats_per_token": 1.6925799611735444
    },
    {
      "epoch": 0.026599246793605646,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0014776431420970343,
      "loss": 1.6920934677124024,
      "step": 77040,
      "train_bpb": 0.602227170446893,
      "train_bytes": 28026.1,
      "train_loss_nats_per_token": 1.691271895935188
    },
    {
      "epoch": 0.02673092623317795,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0014770195679731431,
      "loss": 1.6898454666137694,
      "step": 77060,
      "train_bpb": 0.594916740891826,
      "train_bytes": 28733.775,
      "train_loss_nats_per_token": 1.6912103140901307
    },
    {
      "epoch": 0.026862605672750256,
      "grad_norm": 0.193359375,
      "learning_rate": 0.001476395997821635,
      "loss": 1.692764663696289,
      "step": 77080,
      "train_bpb": 0.5964629028088805,
      "train_bytes": 28643.93125,
      "train_loss_nats_per_token": 1.6932403539554353
    },
    {
      "epoch": 0.02699428511232256,
      "grad_norm": 0.146484375,
      "learning_rate": 0.001475772431750298,
      "loss": 1.6947647094726563,
      "step": 77100,
      "train_bpb": 0.5969771780686601,
      "train_bytes": 28202.75,
      "train_loss_nats_per_token": 1.6951931822282718
    },
    {
      "epoch": 0.027125964551894866,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0014751488698669228,
      "loss": 1.6806833267211914,
      "step": 77120,
      "train_bpb": 0.5987335104686332,
      "train_bytes": 27918.7875,
      "train_loss_nats_per_token": 1.6811130555721177
    },
    {
      "epoch": 0.027257643991467174,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0014745253122792977,
      "loss": 1.6892797470092773,
      "step": 77140,
      "train_bpb": 0.5985712950018136,
      "train_bytes": 28268.19375,
      "train_loss_nats_per_token": 1.6895430585771885
    },
    {
      "epoch": 0.02738932343103948,
      "grad_norm": 0.2236328125,
      "learning_rate": 0.0014739017590952106,
      "loss": 1.685490608215332,
      "step": 77160,
      "train_bpb": 0.5967531324251619,
      "train_bytes": 28393.88125,
      "train_loss_nats_per_token": 1.6863646106619994
    },
    {
      "epoch": 0.027521002870611783,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0014732782104224477,
      "loss": 1.691900634765625,
      "step": 77180,
      "train_bpb": 0.5977716460492143,
      "train_bytes": 28279.18125,
      "train_loss_nats_per_token": 1.6918519679812967
    },
    {
      "epoch": 0.027652682310184088,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0014726546663687958,
      "loss": 1.69884033203125,
      "step": 77200,
      "train_bpb": 0.6016956155652149,
      "train_bytes": 28174.59375,
      "train_loss_nats_per_token": 1.6998718718976658
    },
    {
      "epoch": 0.027784361749756393,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0014720311270420407,
      "loss": 1.691525650024414,
      "step": 77220,
      "train_bpb": 0.5994898584360675,
      "train_bytes": 28352.7375,
      "train_loss_nats_per_token": 1.6922919850739226
    },
    {
      "epoch": 0.027916041189328698,
      "grad_norm": 0.12890625,
      "learning_rate": 0.001471407592549966,
      "loss": 1.6833175659179687,
      "step": 77240,
      "train_bpb": 0.5978083375540929,
      "train_bytes": 28181.25,
      "train_loss_nats_per_token": 1.6849166186445852
    },
    {
      "epoch": 0.028047720628901002,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0014707840630003557,
      "loss": 1.679113006591797,
      "step": 77260,
      "train_bpb": 0.596312566703751,
      "train_bytes": 27955.61875,
      "train_loss_nats_per_token": 1.679523216492946
    },
    {
      "epoch": 0.028179400068473307,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0014701605385009934,
      "loss": 1.6906530380249023,
      "step": 77280,
      "train_bpb": 0.5991973185333034,
      "train_bytes": 28185.325,
      "train_loss_nats_per_token": 1.691745482999527
    },
    {
      "epoch": 0.028311079508045615,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0014695370191596612,
      "loss": 1.6934459686279297,
      "step": 77300,
      "train_bpb": 0.6008071537344541,
      "train_bytes": 28278.7,
      "train_loss_nats_per_token": 1.6940196070853306
    },
    {
      "epoch": 0.02844275894761792,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0014689135050841389,
      "loss": 1.6931989669799805,
      "step": 77320,
      "train_bpb": 0.59961183843431,
      "train_bytes": 28344.6625,
      "train_loss_nats_per_token": 1.6943354801483617
    },
    {
      "epoch": 0.028574438387190225,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.0014682899963822077,
      "loss": 1.6785816192626952,
      "step": 77340,
      "train_bpb": 0.5924778042681349,
      "train_bytes": 28327.41875,
      "train_loss_nats_per_token": 1.6789808384063152
    },
    {
      "epoch": 0.02870611782676253,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0014676664931616477,
      "loss": 1.6945840835571289,
      "step": 77360,
      "train_bpb": 0.6015608732821868,
      "train_bytes": 27954.8625,
      "train_loss_nats_per_token": 1.6948520912494889
    },
    {
      "epoch": 0.028837797266334834,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0014670429955302355,
      "loss": 1.683724594116211,
      "step": 77380,
      "train_bpb": 0.5963643121994728,
      "train_bytes": 28370.25625,
      "train_loss_nats_per_token": 1.6835085919808104
    },
    {
      "epoch": 0.02896947670590714,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0014664195035957494,
      "loss": 1.6853879928588866,
      "step": 77400,
      "train_bpb": 0.593411469075158,
      "train_bytes": 28482.95625,
      "train_loss_nats_per_token": 1.6852748989143622
    },
    {
      "epoch": 0.029101156145479444,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.0014657960174659663,
      "loss": 1.691403579711914,
      "step": 77420,
      "train_bpb": 0.5987068159370338,
      "train_bytes": 28486.66875,
      "train_loss_nats_per_token": 1.6915533810848895
    },
    {
      "epoch": 0.02923283558505175,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0014651725372486603,
      "loss": 1.6831640243530273,
      "step": 77440,
      "train_bpb": 0.5962832966674679,
      "train_bytes": 28204.40625,
      "train_loss_nats_per_token": 1.6835326475216337
    },
    {
      "epoch": 0.029364515024624054,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0014645490630516068,
      "loss": 1.6792739868164062,
      "step": 77460,
      "train_bpb": 0.592545261477014,
      "train_bytes": 28596.05,
      "train_loss_nats_per_token": 1.6805971827450379
    },
    {
      "epoch": 0.029496194464196362,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0014639255949825786,
      "loss": 1.6840660095214843,
      "step": 77480,
      "train_bpb": 0.5949640898983676,
      "train_bytes": 28495.625,
      "train_loss_nats_per_token": 1.684912807275938
    },
    {
      "epoch": 0.029627873903768667,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.001463302133149349,
      "loss": 1.679654312133789,
      "step": 77500,
      "train_bpb": 0.5929298385163867,
      "train_bytes": 28394.25625,
      "train_loss_nats_per_token": 1.6802236083828608
    },
    {
      "epoch": 0.02975955334334097,
      "grad_norm": 0.2099609375,
      "learning_rate": 0.0014626786776596877,
      "loss": 1.7027454376220703,
      "step": 77520,
      "train_bpb": 0.6005715877279713,
      "train_bytes": 28416.7125,
      "train_loss_nats_per_token": 1.7025641553006783
    },
    {
      "epoch": 0.029891232782913276,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0014620552286213658,
      "loss": 1.6862194061279296,
      "step": 77540,
      "train_bpb": 0.5948630822944234,
      "train_bytes": 28368.18125,
      "train_loss_nats_per_token": 1.687123591694218
    },
    {
      "epoch": 0.03002291222248558,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0014614317861421525,
      "loss": 1.6683383941650392,
      "step": 77560,
      "train_bpb": 0.5962533482907953,
      "train_bytes": 27902.56875,
      "train_loss_nats_per_token": 1.6687791433679884
    },
    {
      "epoch": 0.030154591662057886,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0014608083503298147,
      "loss": 1.6918952941894532,
      "step": 77580,
      "train_bpb": 0.5988071859253181,
      "train_bytes": 28368.00625,
      "train_loss_nats_per_token": 1.6919374399187426
    },
    {
      "epoch": 0.03028627110163019,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0014601849212921192,
      "loss": 1.6885618209838866,
      "step": 77600,
      "train_bpb": 0.5978885111574731,
      "train_bytes": 28379.2,
      "train_loss_nats_per_token": 1.688828294292195
    },
    {
      "epoch": 0.030417950541202495,
      "grad_norm": 0.2294921875,
      "learning_rate": 0.0014595614991368322,
      "loss": 1.6782169342041016,
      "step": 77620,
      "train_bpb": 0.5959217850555917,
      "train_bytes": 28203.61875,
      "train_loss_nats_per_token": 1.6793893148442758
    },
    {
      "epoch": 0.030549629980774803,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0014589380839717184,
      "loss": 1.683704948425293,
      "step": 77640,
      "train_bpb": 0.5963821721393374,
      "train_bytes": 28400.9,
      "train_loss_nats_per_token": 1.683804813832797
    },
    {
      "epoch": 0.030681309420347108,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0014583146759045399,
      "loss": 1.7062732696533203,
      "step": 77660,
      "train_bpb": 0.6014909888968999,
      "train_bytes": 28510.99375,
      "train_loss_nats_per_token": 1.7067502824897496
    },
    {
      "epoch": 0.030812988859919413,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0014576912750430585,
      "loss": 1.6911861419677734,
      "step": 77680,
      "train_bpb": 0.599013845968416,
      "train_bytes": 28170.53125,
      "train_loss_nats_per_token": 1.6927983939774807
    },
    {
      "epoch": 0.030944668299491718,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0014570678814950363,
      "loss": 1.687225914001465,
      "step": 77700,
      "train_bpb": 0.6006623103689149,
      "train_bytes": 28122.7125,
      "train_loss_nats_per_token": 1.6883400830287894
    },
    {
      "epoch": 0.031076347739064022,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0014564444953682311,
      "loss": 1.7040239334106446,
      "step": 77720,
      "train_bpb": 0.6039691563589502,
      "train_bytes": 28167.48125,
      "train_loss_nats_per_token": 1.7052752188551266
    },
    {
      "epoch": 0.031208027178636327,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0014558211167704017,
      "loss": 1.7024145126342773,
      "step": 77740,
      "train_bpb": 0.6018275158866624,
      "train_bytes": 28251.36875,
      "train_loss_nats_per_token": 1.7034683672715791
    },
    {
      "epoch": 0.03133970661820863,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0014551977458093055,
      "loss": 1.6955022811889648,
      "step": 77760,
      "train_bpb": 0.603302269278734,
      "train_bytes": 28154.29375,
      "train_loss_nats_per_token": 1.6962553302875338
    },
    {
      "epoch": 0.03147138605778094,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0014545743825926968,
      "loss": 1.6970550537109375,
      "step": 77780,
      "train_bpb": 0.6017438206020793,
      "train_bytes": 28263.69375,
      "train_loss_nats_per_token": 1.6972357920636003
    },
    {
      "epoch": 0.03160306549735324,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0014539510272283303,
      "loss": 1.6896766662597655,
      "step": 77800,
      "train_bpb": 0.600171190726336,
      "train_bytes": 28173.50625,
      "train_loss_nats_per_token": 1.6905443512615685
    },
    {
      "epoch": 0.031734744936925546,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0014533276798239588,
      "loss": 1.7068660736083985,
      "step": 77820,
      "train_bpb": 0.6034759652236055,
      "train_bytes": 28311.60625,
      "train_loss_nats_per_token": 1.7073819642579178
    },
    {
      "epoch": 0.03186642437649785,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0014527043404873343,
      "loss": 1.7061237335205077,
      "step": 77840,
      "train_bpb": 0.601691573614774,
      "train_bytes": 28340.2625,
      "train_loss_nats_per_token": 1.7069527646979812
    },
    {
      "epoch": 0.031998103816070156,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0014520810093262059,
      "loss": 1.7061399459838866,
      "step": 77860,
      "train_bpb": 0.6054697855871355,
      "train_bytes": 27968.66875,
      "train_loss_nats_per_token": 1.7071916872715525
    },
    {
      "epoch": 0.03212978325564246,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0014514576864483224,
      "loss": 1.700141716003418,
      "step": 77880,
      "train_bpb": 0.5995806171725112,
      "train_bytes": 28546.975,
      "train_loss_nats_per_token": 1.7006791548335223
    },
    {
      "epoch": 0.03226146269521477,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.001450834371961432,
      "loss": 1.7026840209960938,
      "step": 77900,
      "train_bpb": 0.6033344477637109,
      "train_bytes": 28131.76875,
      "train_loss_nats_per_token": 1.7039646284567986
    },
    {
      "epoch": 0.03239314213478708,
      "grad_norm": 0.140625,
      "learning_rate": 0.0014502110659732786,
      "loss": 1.7021163940429687,
      "step": 77920,
      "train_bpb": 0.5984496442235387,
      "train_bytes": 28750.575,
      "train_loss_nats_per_token": 1.7033711652540575
    },
    {
      "epoch": 0.03252482157435938,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0014495877685916077,
      "loss": 1.692356491088867,
      "step": 77940,
      "train_bpb": 0.6011504301208217,
      "train_bytes": 27995.56875,
      "train_loss_nats_per_token": 1.693797662562753
    },
    {
      "epoch": 0.03265650101393169,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0014489644799241621,
      "loss": 1.6980403900146483,
      "step": 77960,
      "train_bpb": 0.6025329400949668,
      "train_bytes": 28168.9,
      "train_loss_nats_per_token": 1.6981578596447306
    },
    {
      "epoch": 0.03278818045350399,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0014483412000786831,
      "loss": 1.6935537338256836,
      "step": 77980,
      "train_bpb": 0.6005694716352866,
      "train_bytes": 28161.8875,
      "train_loss_nats_per_token": 1.6940244227912655
    },
    {
      "epoch": 0.032919859893076296,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0014477179291629094,
      "loss": 1.6868318557739257,
      "step": 78000,
      "train_bpb": 0.5977817426818478,
      "train_bytes": 28171.975,
      "train_loss_nats_per_token": 1.6878839533069174
    },
    {
      "epoch": 0.032919859893076296,
      "eval_loss": 1.5172574520111084,
      "eval_runtime": 72.0711,
      "eval_samples_per_second": 13.875,
      "eval_steps_per_second": 13.875,
      "step": 78000,
      "train_bpb": 0.535474876140263,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5248250690660194
    },
    {
      "epoch": 0.0330515393326486,
      "grad_norm": 0.2216796875,
      "learning_rate": 0.0014470946672845804,
      "loss": 1.6856176376342773,
      "step": 78020,
      "train_bpb": 0.5988610935527097,
      "train_bytes": 27965.175,
      "train_loss_nats_per_token": 1.686802723581412
    },
    {
      "epoch": 0.033183218772220906,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0014464714145514326,
      "loss": 1.7101030349731445,
      "step": 78040,
      "train_bpb": 0.6058144263479525,
      "train_bytes": 28266.275,
      "train_loss_nats_per_token": 1.7105539038504625
    },
    {
      "epoch": 0.03331489821179321,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0014458481710712002,
      "loss": 1.7071727752685546,
      "step": 78060,
      "train_bpb": 0.6050159864840899,
      "train_bytes": 28128.2125,
      "train_loss_nats_per_token": 1.7075656181239889
    },
    {
      "epoch": 0.033446577651365515,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0014452249369516171,
      "loss": 1.6910585403442382,
      "step": 78080,
      "train_bpb": 0.5981742809375922,
      "train_bytes": 28209.85,
      "train_loss_nats_per_token": 1.6918975736437993
    },
    {
      "epoch": 0.03357825709093782,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0014446017123004161,
      "loss": 1.7041206359863281,
      "step": 78100,
      "train_bpb": 0.6012439818354264,
      "train_bytes": 28403.2875,
      "train_loss_nats_per_token": 1.7048474672476013
    },
    {
      "epoch": 0.033709936530510125,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0014439784972253259,
      "loss": 1.7038389205932618,
      "step": 78120,
      "train_bpb": 0.6049490912061616,
      "train_bytes": 28202.49375,
      "train_loss_nats_per_token": 1.7063326803230812
    },
    {
      "epoch": 0.03384161597008243,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0014433552918340755,
      "loss": 1.693025779724121,
      "step": 78140,
      "train_bpb": 0.6014002334891742,
      "train_bytes": 27979.78125,
      "train_loss_nats_per_token": 1.6933369390119533
    },
    {
      "epoch": 0.033973295409654734,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0014427320962343922,
      "loss": 1.7185195922851562,
      "step": 78160,
      "train_bpb": 0.605387625644956,
      "train_bytes": 28907.7125,
      "train_loss_nats_per_token": 1.71902666050344
    },
    {
      "epoch": 0.03410497484922704,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0014421089105340014,
      "loss": 1.6958621978759765,
      "step": 78180,
      "train_bpb": 0.6002582429075636,
      "train_bytes": 28333.625,
      "train_loss_nats_per_token": 1.6958689330688002
    },
    {
      "epoch": 0.034236654288799344,
      "grad_norm": 0.1875,
      "learning_rate": 0.0014414857348406254,
      "loss": 1.6897453308105468,
      "step": 78200,
      "train_bpb": 0.600323375865349,
      "train_bytes": 28196.00625,
      "train_loss_nats_per_token": 1.6911343180357603
    },
    {
      "epoch": 0.03436833372837165,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0014408625692619867,
      "loss": 1.6891374588012695,
      "step": 78220,
      "train_bpb": 0.5985801659290828,
      "train_bytes": 28437.3625,
      "train_loss_nats_per_token": 1.689170184865208
    },
    {
      "epoch": 0.03450001316794396,
      "grad_norm": 0.130859375,
      "learning_rate": 0.001440239413905806,
      "loss": 1.6971651077270509,
      "step": 78240,
      "train_bpb": 0.6016299962354478,
      "train_bytes": 28671.41875,
      "train_loss_nats_per_token": 1.6980697222545529
    },
    {
      "epoch": 0.034631692607516265,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0014396162688798,
      "loss": 1.6942461013793946,
      "step": 78260,
      "train_bpb": 0.5985557052176707,
      "train_bytes": 28493.51875,
      "train_loss_nats_per_token": 1.6948327587595866
    },
    {
      "epoch": 0.03476337204708857,
      "grad_norm": 0.203125,
      "learning_rate": 0.001438993134291686,
      "loss": 1.6875732421875,
      "step": 78280,
      "train_bpb": 0.6008734786388091,
      "train_bytes": 28406.45,
      "train_loss_nats_per_token": 1.6891374196621216
    },
    {
      "epoch": 0.034895051486660875,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.001438370010249179,
      "loss": 1.7109525680541993,
      "step": 78300,
      "train_bpb": 0.6052625488535335,
      "train_bytes": 28298.01875,
      "train_loss_nats_per_token": 1.7121146532832467
    },
    {
      "epoch": 0.03502673092623318,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0014377468968599912,
      "loss": 1.6958467483520507,
      "step": 78320,
      "train_bpb": 0.6014291858248885,
      "train_bytes": 28227.19375,
      "train_loss_nats_per_token": 1.6969841193272464
    },
    {
      "epoch": 0.035158410365805484,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.001437123794231833,
      "loss": 1.6785184860229492,
      "step": 78340,
      "train_bpb": 0.5925629994352949,
      "train_bytes": 28502.6625,
      "train_loss_nats_per_token": 1.679704532842838
    },
    {
      "epoch": 0.03529008980537779,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0014365007024724147,
      "loss": 1.6898578643798827,
      "step": 78360,
      "train_bpb": 0.5996792046569441,
      "train_bytes": 28267.01875,
      "train_loss_nats_per_token": 1.6902745541524382
    },
    {
      "epoch": 0.035421769244950094,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0014358776216894435,
      "loss": 1.6955854415893554,
      "step": 78380,
      "train_bpb": 0.6007074275759285,
      "train_bytes": 28253.175,
      "train_loss_nats_per_token": 1.6969436082531852
    },
    {
      "epoch": 0.0355534486845224,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.001435254551990624,
      "loss": 1.69860782623291,
      "step": 78400,
      "train_bpb": 0.6021478192334722,
      "train_bytes": 28216.025,
      "train_loss_nats_per_token": 1.699796544854165
    },
    {
      "epoch": 0.0356851281240947,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0014346314934836595,
      "loss": 1.6828365325927734,
      "step": 78420,
      "train_bpb": 0.5950792401683539,
      "train_bytes": 28388.0,
      "train_loss_nats_per_token": 1.6832542923455405
    },
    {
      "epoch": 0.03581680756366701,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0014340084462762528,
      "loss": 1.6947235107421874,
      "step": 78440,
      "train_bpb": 0.6000495165906139,
      "train_bytes": 28025.8875,
      "train_loss_nats_per_token": 1.6955410310567858
    },
    {
      "epoch": 0.03594848700323931,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0014333854104761018,
      "loss": 1.6782625198364258,
      "step": 78460,
      "train_bpb": 0.5943095820941537,
      "train_bytes": 28320.71875,
      "train_loss_nats_per_token": 1.679294255000767
    },
    {
      "epoch": 0.03608016644281162,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0014327623861909045,
      "loss": 1.6753961563110351,
      "step": 78480,
      "train_bpb": 0.5931270507872257,
      "train_bytes": 28319.85625,
      "train_loss_nats_per_token": 1.6759470597849533
    },
    {
      "epoch": 0.03621184588238392,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0014321393735283574,
      "loss": 1.688319778442383,
      "step": 78500,
      "train_bpb": 0.5995629659704901,
      "train_bytes": 28268.66875,
      "train_loss_nats_per_token": 1.6882650267604775
    },
    {
      "epoch": 0.03634352532195623,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0014315163725961532,
      "loss": 1.6961109161376953,
      "step": 78520,
      "train_bpb": 0.6002832021543034,
      "train_bytes": 28384.16875,
      "train_loss_nats_per_token": 1.6957462829809045
    },
    {
      "epoch": 0.03647520476152853,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0014308933835019833,
      "loss": 1.6948938369750977,
      "step": 78540,
      "train_bpb": 0.5984438757815312,
      "train_bytes": 28397.75625,
      "train_loss_nats_per_token": 1.6953502408159915
    },
    {
      "epoch": 0.036606884201100844,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0014302704063535376,
      "loss": 1.703961181640625,
      "step": 78560,
      "train_bpb": 0.6026772704560748,
      "train_bytes": 28283.0375,
      "train_loss_nats_per_token": 1.7054070432307815
    },
    {
      "epoch": 0.03673856364067315,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0014296474412585042,
      "loss": 1.6986579895019531,
      "step": 78580,
      "train_bpb": 0.5999889219257692,
      "train_bytes": 28176.50625,
      "train_loss_nats_per_token": 1.6989717722798094
    },
    {
      "epoch": 0.03687024308024545,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0014290244883245669,
      "loss": 1.689097785949707,
      "step": 78600,
      "train_bpb": 0.5982379403071769,
      "train_bytes": 28291.0375,
      "train_loss_nats_per_token": 1.6899332219598138
    },
    {
      "epoch": 0.03700192251981776,
      "grad_norm": 0.140625,
      "learning_rate": 0.0014284015476594095,
      "loss": 1.6824327468872071,
      "step": 78620,
      "train_bpb": 0.5932796998954706,
      "train_bytes": 28488.99375,
      "train_loss_nats_per_token": 1.6847080102463428
    },
    {
      "epoch": 0.03713360195939006,
      "grad_norm": 0.140625,
      "learning_rate": 0.001427778619370714,
      "loss": 1.6862525939941406,
      "step": 78640,
      "train_bpb": 0.598078122277963,
      "train_bytes": 28235.90625,
      "train_loss_nats_per_token": 1.6862987486288787
    },
    {
      "epoch": 0.03726528139896237,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0014271557035661585,
      "loss": 1.6908929824829102,
      "step": 78660,
      "train_bpb": 0.5990516797825473,
      "train_bytes": 28219.9625,
      "train_loss_nats_per_token": 1.6916770362515654
    },
    {
      "epoch": 0.03739696083853467,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0014265328003534195,
      "loss": 1.6864519119262695,
      "step": 78680,
      "train_bpb": 0.598779172492953,
      "train_bytes": 28103.58125,
      "train_loss_nats_per_token": 1.6873703836559726
    },
    {
      "epoch": 0.03752864027810698,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0014259099098401723,
      "loss": 1.69726619720459,
      "step": 78700,
      "train_bpb": 0.6001386549163167,
      "train_bytes": 28327.1625,
      "train_loss_nats_per_token": 1.6983433561419947
    },
    {
      "epoch": 0.03766031971767928,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0014252870321340898,
      "loss": 1.6958938598632813,
      "step": 78720,
      "train_bpb": 0.6000656524414464,
      "train_bytes": 28063.4625,
      "train_loss_nats_per_token": 1.6967076976508757
    },
    {
      "epoch": 0.037791999157251586,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.001424664167342841,
      "loss": 1.6883604049682617,
      "step": 78740,
      "train_bpb": 0.5988855147615418,
      "train_bytes": 28258.11875,
      "train_loss_nats_per_token": 1.6889166554582342
    },
    {
      "epoch": 0.03792367859682389,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0014240413155740944,
      "loss": 1.693502426147461,
      "step": 78760,
      "train_bpb": 0.6027024938771277,
      "train_bytes": 28074.58125,
      "train_loss_nats_per_token": 1.694437035277016
    },
    {
      "epoch": 0.038055358036396196,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0014234184769355163,
      "loss": 1.675084686279297,
      "step": 78780,
      "train_bpb": 0.5921536451531454,
      "train_bytes": 28628.925,
      "train_loss_nats_per_token": 1.676072054057062
    },
    {
      "epoch": 0.0381870374759685,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0014227956515347698,
      "loss": 1.6889854431152345,
      "step": 78800,
      "train_bpb": 0.5971854423583867,
      "train_bytes": 28092.85625,
      "train_loss_nats_per_token": 1.6903938326645067
    },
    {
      "epoch": 0.038318716915540806,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0014221728394795156,
      "loss": 1.6822931289672851,
      "step": 78820,
      "train_bpb": 0.5940728775505026,
      "train_bytes": 28380.8625,
      "train_loss_nats_per_token": 1.683557923669422
    },
    {
      "epoch": 0.03845039635511311,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0014215500408774134,
      "loss": 1.6838510513305665,
      "step": 78840,
      "train_bpb": 0.5960664582500314,
      "train_bytes": 28369.74375,
      "train_loss_nats_per_token": 1.6856742803669826
    },
    {
      "epoch": 0.038582075794685415,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0014209272558361192,
      "loss": 1.670450782775879,
      "step": 78860,
      "train_bpb": 0.5936884667207082,
      "train_bytes": 28001.68125,
      "train_loss_nats_per_token": 1.6706047697813284
    },
    {
      "epoch": 0.03871375523425772,
      "grad_norm": 0.177734375,
      "learning_rate": 0.001420304484463287,
      "loss": 1.6749605178833007,
      "step": 78880,
      "train_bpb": 0.593547471472502,
      "train_bytes": 28326.70625,
      "train_loss_nats_per_token": 1.6754831282665164
    },
    {
      "epoch": 0.03884543467383003,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0014196817268665692,
      "loss": 1.6767765045166017,
      "step": 78900,
      "train_bpb": 0.5946737282239993,
      "train_bytes": 28256.1625,
      "train_loss_nats_per_token": 1.6778287892948318
    },
    {
      "epoch": 0.038977114113402336,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0014190589831536155,
      "loss": 1.6795551300048828,
      "step": 78920,
      "train_bpb": 0.5932495789366724,
      "train_bytes": 28420.19375,
      "train_loss_nats_per_token": 1.6800634281104443
    },
    {
      "epoch": 0.03910879355297464,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0014184362534320721,
      "loss": 1.69957275390625,
      "step": 78940,
      "train_bpb": 0.6016052998899448,
      "train_bytes": 28077.78125,
      "train_loss_nats_per_token": 1.6992799025272918
    },
    {
      "epoch": 0.039240472992546946,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.001417813537809584,
      "loss": 1.6916547775268556,
      "step": 78960,
      "train_bpb": 0.5978457343914115,
      "train_bytes": 28216.05,
      "train_loss_nats_per_token": 1.6920252508584193
    },
    {
      "epoch": 0.03937215243211925,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0014171908363937939,
      "loss": 1.6776002883911132,
      "step": 78980,
      "train_bpb": 0.5927188964849116,
      "train_bytes": 28337.95625,
      "train_loss_nats_per_token": 1.6784826171359355
    },
    {
      "epoch": 0.039503831871691555,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0014165681492923409,
      "loss": 1.6877754211425782,
      "step": 79000,
      "train_bpb": 0.597982654417132,
      "train_bytes": 28175.625,
      "train_loss_nats_per_token": 1.6885082508008513
    },
    {
      "epoch": 0.039503831871691555,
      "eval_loss": 1.5184521675109863,
      "eval_runtime": 71.6968,
      "eval_samples_per_second": 13.948,
      "eval_steps_per_second": 13.948,
      "step": 79000,
      "train_bpb": 0.5358855726439858,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5259945736544849
    },
    {
      "epoch": 0.03963551131126386,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0014159454766128625,
      "loss": 1.690185546875,
      "step": 79020,
      "train_bpb": 0.5967276655260991,
      "train_bytes": 28411.50625,
      "train_loss_nats_per_token": 1.6899873574390067
    },
    {
      "epoch": 0.039767190750836165,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0014153228184629937,
      "loss": 1.6728246688842774,
      "step": 79040,
      "train_bpb": 0.5924333050418202,
      "train_bytes": 28203.44375,
      "train_loss_nats_per_token": 1.6736600456865685
    },
    {
      "epoch": 0.03989887019040847,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0014147001749503665,
      "loss": 1.681999397277832,
      "step": 79060,
      "train_bpb": 0.5909319226331758,
      "train_bytes": 28637.44375,
      "train_loss_nats_per_token": 1.6818919575827886
    },
    {
      "epoch": 0.040030549629980774,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0014140775461826108,
      "loss": 1.6935585021972657,
      "step": 79080,
      "train_bpb": 0.5985582476906349,
      "train_bytes": 28276.50625,
      "train_loss_nats_per_token": 1.6938629821244626
    },
    {
      "epoch": 0.04016222906955308,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0014134549322673533,
      "loss": 1.6753795623779297,
      "step": 79100,
      "train_bpb": 0.5934050153063277,
      "train_bytes": 28056.8,
      "train_loss_nats_per_token": 1.6763994009199736
    },
    {
      "epoch": 0.040293908509125384,
      "grad_norm": 0.185546875,
      "learning_rate": 0.00141283233331222,
      "loss": 1.6987340927124024,
      "step": 79120,
      "train_bpb": 0.5975452908814409,
      "train_bytes": 28523.44375,
      "train_loss_nats_per_token": 1.7000338848928669
    },
    {
      "epoch": 0.04042558794869769,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0014122097494248314,
      "loss": 1.6816415786743164,
      "step": 79140,
      "train_bpb": 0.5944356092277779,
      "train_bytes": 28695.15625,
      "train_loss_nats_per_token": 1.6818655300925514
    },
    {
      "epoch": 0.040557267388269994,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0014115871807128077,
      "loss": 1.692144012451172,
      "step": 79160,
      "train_bpb": 0.5966286411316302,
      "train_bytes": 28385.46875,
      "train_loss_nats_per_token": 1.6932140669342586
    },
    {
      "epoch": 0.0406889468278423,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0014109646272837657,
      "loss": 1.6758153915405274,
      "step": 79180,
      "train_bpb": 0.5943127181942286,
      "train_bytes": 28194.91875,
      "train_loss_nats_per_token": 1.6763453442452272
    },
    {
      "epoch": 0.0408206262674146,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0014103420892453194,
      "loss": 1.6837926864624024,
      "step": 79200,
      "train_bpb": 0.5993501095962781,
      "train_bytes": 28351.5625,
      "train_loss_nats_per_token": 1.6848514149199854
    },
    {
      "epoch": 0.04095230570698691,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0014097195667050802,
      "loss": 1.6758466720581056,
      "step": 79220,
      "train_bpb": 0.5943123885685314,
      "train_bytes": 28334.5125,
      "train_loss_nats_per_token": 1.6769203373302262
    },
    {
      "epoch": 0.04108398514655922,
      "grad_norm": 0.15625,
      "learning_rate": 0.001409097059770657,
      "loss": 1.677947425842285,
      "step": 79240,
      "train_bpb": 0.59264406463803,
      "train_bytes": 28285.05,
      "train_loss_nats_per_token": 1.6791180866870166
    },
    {
      "epoch": 0.041215664586131524,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0014084745685496566,
      "loss": 1.689889907836914,
      "step": 79260,
      "train_bpb": 0.599480964175318,
      "train_bytes": 28141.0125,
      "train_loss_nats_per_token": 1.6904960818767496
    },
    {
      "epoch": 0.04134734402570383,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0014078520931496813,
      "loss": 1.6717391967773438,
      "step": 79280,
      "train_bpb": 0.5932972172800046,
      "train_bytes": 28346.05,
      "train_loss_nats_per_token": 1.672242020617127
    },
    {
      "epoch": 0.041479023465276134,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0014072296336783325,
      "loss": 1.6819957733154296,
      "step": 79300,
      "train_bpb": 0.5937900921855016,
      "train_bytes": 28583.7375,
      "train_loss_nats_per_token": 1.6827284947531986
    },
    {
      "epoch": 0.04161070290484844,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0014066071902432082,
      "loss": 1.6983301162719726,
      "step": 79320,
      "train_bpb": 0.6010051623684264,
      "train_bytes": 28219.3625,
      "train_loss_nats_per_token": 1.6993577706410823
    },
    {
      "epoch": 0.04174238234442074,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0014059847629519034,
      "loss": 1.6806570053100587,
      "step": 79340,
      "train_bpb": 0.5950408739369666,
      "train_bytes": 28480.20625,
      "train_loss_nats_per_token": 1.6817875380410419
    },
    {
      "epoch": 0.04187406178399305,
      "grad_norm": 0.15234375,
      "learning_rate": 0.00140536235191201,
      "loss": 1.6784297943115234,
      "step": 79360,
      "train_bpb": 0.5966719431243956,
      "train_bytes": 28415.40625,
      "train_loss_nats_per_token": 1.6787914321583952
    },
    {
      "epoch": 0.04200574122356535,
      "grad_norm": 0.2421875,
      "learning_rate": 0.0014047399572311183,
      "loss": 1.6877431869506836,
      "step": 79380,
      "train_bpb": 0.5952715379446712,
      "train_bytes": 28221.44375,
      "train_loss_nats_per_token": 1.688634053136842
    },
    {
      "epoch": 0.04213742066313766,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0014041175790168148,
      "loss": 1.685025978088379,
      "step": 79400,
      "train_bpb": 0.5953686590397086,
      "train_bytes": 28470.00625,
      "train_loss_nats_per_token": 1.6855320235269888
    },
    {
      "epoch": 0.04226910010270996,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0014034952173766829,
      "loss": 1.6749755859375,
      "step": 79420,
      "train_bpb": 0.593519256148443,
      "train_bytes": 28260.725,
      "train_loss_nats_per_token": 1.676105019521307
    },
    {
      "epoch": 0.04240077954228227,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.001402872872418304,
      "loss": 1.6891170501708985,
      "step": 79440,
      "train_bpb": 0.5975509446083594,
      "train_bytes": 28285.30625,
      "train_loss_nats_per_token": 1.689705869220961
    },
    {
      "epoch": 0.04253245898185457,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0014022505442492566,
      "loss": 1.6830257415771483,
      "step": 79460,
      "train_bpb": 0.5965925825152328,
      "train_bytes": 28132.69375,
      "train_loss_nats_per_token": 1.683796547507723
    },
    {
      "epoch": 0.04266413842142688,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0014016282329771157,
      "loss": 1.685413932800293,
      "step": 79480,
      "train_bpb": 0.5999038285077584,
      "train_bytes": 27893.0375,
      "train_loss_nats_per_token": 1.685803577516774
    },
    {
      "epoch": 0.04279581786099918,
      "grad_norm": 0.146484375,
      "learning_rate": 0.001401005938709453,
      "loss": 1.6839406967163086,
      "step": 79500,
      "train_bpb": 0.5951610860076998,
      "train_bytes": 28607.975,
      "train_loss_nats_per_token": 1.6845231091332986
    },
    {
      "epoch": 0.042927497300571486,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0014003836615538389,
      "loss": 1.690140151977539,
      "step": 79520,
      "train_bpb": 0.5968120425609973,
      "train_bytes": 28578.2375,
      "train_loss_nats_per_token": 1.691081792267875
    },
    {
      "epoch": 0.04305917674014379,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0013997614016178393,
      "loss": 1.6794157028198242,
      "step": 79540,
      "train_bpb": 0.5941933407792053,
      "train_bytes": 28085.25,
      "train_loss_nats_per_token": 1.6800860787021488
    },
    {
      "epoch": 0.043190856179716096,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0013991391590090173,
      "loss": 1.6722570419311524,
      "step": 79560,
      "train_bpb": 0.5948451365533113,
      "train_bytes": 28239.9375,
      "train_loss_nats_per_token": 1.6730180931816354
    },
    {
      "epoch": 0.04332253561928841,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0013985169338349335,
      "loss": 1.6765289306640625,
      "step": 79580,
      "train_bpb": 0.5944244697027372,
      "train_bytes": 28193.71875,
      "train_loss_nats_per_token": 1.6768886987863638
    },
    {
      "epoch": 0.04345421505886071,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0013978947262031464,
      "loss": 1.6751808166503905,
      "step": 79600,
      "train_bpb": 0.5932654131117767,
      "train_bytes": 28040.69375,
      "train_loss_nats_per_token": 1.676013717355352
    },
    {
      "epoch": 0.04358589449843302,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0013972725362212087,
      "loss": 1.6793363571166993,
      "step": 79620,
      "train_bpb": 0.5956372972697066,
      "train_bytes": 28218.20625,
      "train_loss_nats_per_token": 1.6795753366979191
    },
    {
      "epoch": 0.04371757393800532,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0013966503639966728,
      "loss": 1.6790416717529297,
      "step": 79640,
      "train_bpb": 0.5956924564448424,
      "train_bytes": 28259.425,
      "train_loss_nats_per_token": 1.6795935815125178
    },
    {
      "epoch": 0.04384925337757763,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0013960282096370869,
      "loss": 1.6938724517822266,
      "step": 79660,
      "train_bpb": 0.6022727108993942,
      "train_bytes": 27946.975,
      "train_loss_nats_per_token": 1.6948478996773066
    },
    {
      "epoch": 0.04398093281714993,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0013954060732499962,
      "loss": 1.6894308090209962,
      "step": 79680,
      "train_bpb": 0.6022473559581485,
      "train_bytes": 27867.66875,
      "train_loss_nats_per_token": 1.6902526580114179
    },
    {
      "epoch": 0.044112612256722236,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0013947839549429421,
      "loss": 1.6719879150390624,
      "step": 79700,
      "train_bpb": 0.5913966035613655,
      "train_bytes": 28140.63125,
      "train_loss_nats_per_token": 1.6727439005886071
    },
    {
      "epoch": 0.04424429169629454,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0013941618548234645,
      "loss": 1.6761489868164063,
      "step": 79720,
      "train_bpb": 0.592904093577309,
      "train_bytes": 28208.6875,
      "train_loss_nats_per_token": 1.6756404831771812
    },
    {
      "epoch": 0.044375971135866846,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.001393539772999099,
      "loss": 1.6787918090820313,
      "step": 79740,
      "train_bpb": 0.5907815882227805,
      "train_bytes": 28502.19375,
      "train_loss_nats_per_token": 1.6788864655311335
    },
    {
      "epoch": 0.04450765057543915,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0013929177095773776,
      "loss": 1.6644969940185548,
      "step": 79760,
      "train_bpb": 0.5872845240082668,
      "train_bytes": 28656.90625,
      "train_loss_nats_per_token": 1.664872146309802
    },
    {
      "epoch": 0.044639330015011455,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0013922956646658302,
      "loss": 1.6802038192749023,
      "step": 79780,
      "train_bpb": 0.5910167259922445,
      "train_bytes": 28349.5375,
      "train_loss_nats_per_token": 1.6808680617038083
    },
    {
      "epoch": 0.04477100945458376,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0013916736383719835,
      "loss": 1.6703662872314453,
      "step": 79800,
      "train_bpb": 0.5889116188888097,
      "train_bytes": 28491.0875,
      "train_loss_nats_per_token": 1.671024358420464
    },
    {
      "epoch": 0.044902688894156065,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0013910516308033602,
      "loss": 1.679056167602539,
      "step": 79820,
      "train_bpb": 0.5926902685903613,
      "train_bytes": 28328.2,
      "train_loss_nats_per_token": 1.6795398629107854
    },
    {
      "epoch": 0.04503436833372837,
      "grad_norm": 0.2490234375,
      "learning_rate": 0.0013904296420674796,
      "loss": 1.6793943405151368,
      "step": 79840,
      "train_bpb": 0.5983843771982787,
      "train_bytes": 27797.4875,
      "train_loss_nats_per_token": 1.6801233019855866
    },
    {
      "epoch": 0.045166047773300674,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0013898076722718591,
      "loss": 1.6926210403442383,
      "step": 79860,
      "train_bpb": 0.6006042006369546,
      "train_bytes": 28126.0875,
      "train_loss_nats_per_token": 1.6928463414786696
    },
    {
      "epoch": 0.04529772721287298,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0013891857215240118,
      "loss": 1.680025291442871,
      "step": 79880,
      "train_bpb": 0.595849913869838,
      "train_bytes": 28065.6125,
      "train_loss_nats_per_token": 1.680358208409025
    },
    {
      "epoch": 0.045429406652445284,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.001388563789931447,
      "loss": 1.6764389038085938,
      "step": 79900,
      "train_bpb": 0.5963227749351437,
      "train_bytes": 28168.28125,
      "train_loss_nats_per_token": 1.6767384037795787
    },
    {
      "epoch": 0.045561086092017596,
      "grad_norm": 0.171875,
      "learning_rate": 0.0013879418776016725,
      "loss": 1.6821218490600587,
      "step": 79920,
      "train_bpb": 0.5939690565535533,
      "train_bytes": 28461.70625,
      "train_loss_nats_per_token": 1.6828992098727726
    },
    {
      "epoch": 0.0456927655315899,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0013873199846421906,
      "loss": 1.687179946899414,
      "step": 79940,
      "train_bpb": 0.5993772565861241,
      "train_bytes": 28368.3625,
      "train_loss_nats_per_token": 1.6887570656018305
    },
    {
      "epoch": 0.045824444971162205,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.001386698111160502,
      "loss": 1.680758285522461,
      "step": 79960,
      "train_bpb": 0.5965812639362278,
      "train_bytes": 28311.725,
      "train_loss_nats_per_token": 1.6817041926596337
    },
    {
      "epoch": 0.04595612441073451,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.001386076257264103,
      "loss": 1.678987693786621,
      "step": 79980,
      "train_bpb": 0.5955642797109776,
      "train_bytes": 28201.5125,
      "train_loss_nats_per_token": 1.679466997702721
    },
    {
      "epoch": 0.046087803850306815,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0013854544230604872,
      "loss": 1.6729726791381836,
      "step": 80000,
      "train_bpb": 0.5916797092652399,
      "train_bytes": 28176.33125,
      "train_loss_nats_per_token": 1.6730188637416572
    },
    {
      "epoch": 0.046087803850306815,
      "eval_loss": 1.5209064483642578,
      "eval_runtime": 71.4534,
      "eval_samples_per_second": 13.995,
      "eval_steps_per_second": 13.995,
      "step": 80000,
      "train_bpb": 0.5367353585367478,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5284144347728283
    },
    {
      "epoch": 0.04621948328987912,
      "grad_norm": 0.15625,
      "learning_rate": 0.0013848326086571441,
      "loss": 1.6860183715820312,
      "step": 80020,
      "train_bpb": 0.5946114987665799,
      "train_bytes": 28447.6,
      "train_loss_nats_per_token": 1.6864669909420869
    },
    {
      "epoch": 0.046351162729451424,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.00138421081416156,
      "loss": 1.6734783172607421,
      "step": 80040,
      "train_bpb": 0.5951679702662056,
      "train_bytes": 28138.0625,
      "train_loss_nats_per_token": 1.6740616602680383
    },
    {
      "epoch": 0.04648284216902373,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0013835890396812186,
      "loss": 1.6799339294433593,
      "step": 80060,
      "train_bpb": 0.5949380919527701,
      "train_bytes": 28188.475,
      "train_loss_nats_per_token": 1.6804634384365604
    },
    {
      "epoch": 0.046614521608596034,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0013829672853235987,
      "loss": 1.689885139465332,
      "step": 80080,
      "train_bpb": 0.5977052738270994,
      "train_bytes": 28428.9375,
      "train_loss_nats_per_token": 1.6901001892351224
    },
    {
      "epoch": 0.04674620104816834,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0013823455511961765,
      "loss": 1.6892723083496093,
      "step": 80100,
      "train_bpb": 0.5962336668344478,
      "train_bytes": 28796.21875,
      "train_loss_nats_per_token": 1.6899928553453925
    },
    {
      "epoch": 0.04687788048774064,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0013817238374064244,
      "loss": 1.67318115234375,
      "step": 80120,
      "train_bpb": 0.5967875337112938,
      "train_bytes": 27954.30625,
      "train_loss_nats_per_token": 1.6743707437782804
    },
    {
      "epoch": 0.04700955992731295,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0013811021440618123,
      "loss": 1.6914354324340821,
      "step": 80140,
      "train_bpb": 0.5984233946177734,
      "train_bytes": 28397.3125,
      "train_loss_nats_per_token": 1.6912813742263877
    },
    {
      "epoch": 0.04714123936688525,
      "grad_norm": 0.134765625,
      "learning_rate": 0.001380480471269805,
      "loss": 1.697138786315918,
      "step": 80160,
      "train_bpb": 0.6027914522380069,
      "train_bytes": 28276.2625,
      "train_loss_nats_per_token": 1.6978453882109996
    },
    {
      "epoch": 0.04727291880645756,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0013798588191378642,
      "loss": 1.6861032485961913,
      "step": 80180,
      "train_bpb": 0.5953889154136831,
      "train_bytes": 28214.1875,
      "train_loss_nats_per_token": 1.6868173932912476
    },
    {
      "epoch": 0.04740459824602986,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0013792371877734489,
      "loss": 1.6891668319702149,
      "step": 80200,
      "train_bpb": 0.5994079279021219,
      "train_bytes": 28019.1,
      "train_loss_nats_per_token": 1.688540780247276
    },
    {
      "epoch": 0.04753627768560217,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0013786155772840137,
      "loss": 1.6806306838989258,
      "step": 80220,
      "train_bpb": 0.5934343357836933,
      "train_bytes": 28356.44375,
      "train_loss_nats_per_token": 1.6809400545528588
    },
    {
      "epoch": 0.04766795712517447,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0013779939877770094,
      "loss": 1.6796886444091796,
      "step": 80240,
      "train_bpb": 0.5985375327593334,
      "train_bytes": 28036.85625,
      "train_loss_nats_per_token": 1.6810979827573358
    },
    {
      "epoch": 0.047799636564746784,
      "grad_norm": 0.169921875,
      "learning_rate": 0.001377372419359884,
      "loss": 1.6750242233276367,
      "step": 80260,
      "train_bpb": 0.5911241770686695,
      "train_bytes": 28481.975,
      "train_loss_nats_per_token": 1.6752721127050278
    },
    {
      "epoch": 0.04793131600431909,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.001376750872140081,
      "loss": 1.6896203994750976,
      "step": 80280,
      "train_bpb": 0.5989535799353154,
      "train_bytes": 28213.5125,
      "train_loss_nats_per_token": 1.690844073407159
    },
    {
      "epoch": 0.04806299544389139,
      "grad_norm": 0.21484375,
      "learning_rate": 0.001376129346225041,
      "loss": 1.6883146286010742,
      "step": 80300,
      "train_bpb": 0.600050795954489,
      "train_bytes": 28479.06875,
      "train_loss_nats_per_token": 1.6894229449119351
    },
    {
      "epoch": 0.0481946748834637,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.0013755078417222004,
      "loss": 1.6962020874023438,
      "step": 80320,
      "train_bpb": 0.6017344848676283,
      "train_bytes": 28059.54375,
      "train_loss_nats_per_token": 1.6966601530999754
    },
    {
      "epoch": 0.048326354323036,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0013748863587389924,
      "loss": 1.693354034423828,
      "step": 80340,
      "train_bpb": 0.5989416606235669,
      "train_bytes": 28015.69375,
      "train_loss_nats_per_token": 1.6940142282618071
    },
    {
      "epoch": 0.04845803376260831,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0013742648973828458,
      "loss": 1.6880977630615235,
      "step": 80360,
      "train_bpb": 0.6029248897053727,
      "train_bytes": 27679.45,
      "train_loss_nats_per_token": 1.6898991259612708
    },
    {
      "epoch": 0.04858971320218061,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0013736434577611856,
      "loss": 1.6885290145874023,
      "step": 80380,
      "train_bpb": 0.6001658332200619,
      "train_bytes": 28196.69375,
      "train_loss_nats_per_token": 1.68926624367086
    },
    {
      "epoch": 0.04872139264175292,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0013730220399814344,
      "loss": 1.6964723587036132,
      "step": 80400,
      "train_bpb": 0.601671136022736,
      "train_bytes": 28386.06875,
      "train_loss_nats_per_token": 1.6978047369393328
    },
    {
      "epoch": 0.04885307208132522,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0013724006441510094,
      "loss": 1.6990602493286133,
      "step": 80420,
      "train_bpb": 0.6015881257220772,
      "train_bytes": 28622.6,
      "train_loss_nats_per_token": 1.6999054774124276
    },
    {
      "epoch": 0.048984751520897526,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0013717792703773245,
      "loss": 1.6963205337524414,
      "step": 80440,
      "train_bpb": 0.6001737860748544,
      "train_bytes": 28689.65625,
      "train_loss_nats_per_token": 1.6974976680156544
    },
    {
      "epoch": 0.04911643096046983,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0013711579187677908,
      "loss": 1.6800081253051757,
      "step": 80460,
      "train_bpb": 0.5948851709765165,
      "train_bytes": 28659.85,
      "train_loss_nats_per_token": 1.6805018213969727
    },
    {
      "epoch": 0.049248110400042136,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0013705365894298137,
      "loss": 1.6963815689086914,
      "step": 80480,
      "train_bpb": 0.5978450728350683,
      "train_bytes": 29004.68125,
      "train_loss_nats_per_token": 1.6976965769946
    },
    {
      "epoch": 0.04937978983961444,
      "grad_norm": 0.140625,
      "learning_rate": 0.0013699152824707967,
      "loss": 1.6943540573120117,
      "step": 80500,
      "train_bpb": 0.5955058370692187,
      "train_bytes": 28697.8625,
      "train_loss_nats_per_token": 1.6956686889964934
    },
    {
      "epoch": 0.049511469279186746,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0013692939979981378,
      "loss": 1.68204345703125,
      "step": 80520,
      "train_bpb": 0.5968977122664886,
      "train_bytes": 28152.4375,
      "train_loss_nats_per_token": 1.6831754972639685
    },
    {
      "epoch": 0.04964314871875905,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0013686727361192322,
      "loss": 1.6968557357788085,
      "step": 80540,
      "train_bpb": 0.6002532425312389,
      "train_bytes": 28229.29375,
      "train_loss_nats_per_token": 1.6970010822037747
    },
    {
      "epoch": 0.049774828158331355,
      "grad_norm": 0.154296875,
      "learning_rate": 0.001368051496941471,
      "loss": 1.6875120162963868,
      "step": 80560,
      "train_bpb": 0.5977573927661388,
      "train_bytes": 28112.6625,
      "train_loss_nats_per_token": 1.688334060952872
    },
    {
      "epoch": 0.04990650759790366,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0013674302805722406,
      "loss": 1.6913122177124023,
      "step": 80580,
      "train_bpb": 0.5993358393814886,
      "train_bytes": 28250.025,
      "train_loss_nats_per_token": 1.6916479430690678
    },
    {
      "epoch": 0.05003818703747597,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0013668090871189248,
      "loss": 1.6952655792236329,
      "step": 80600,
      "train_bpb": 0.5968788229343955,
      "train_bytes": 28150.6875,
      "train_loss_nats_per_token": 1.6954608215232683
    },
    {
      "epoch": 0.050169866477048276,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0013661879166889017,
      "loss": 1.7020614624023438,
      "step": 80620,
      "train_bpb": 0.6023287286107594,
      "train_bytes": 28237.7625,
      "train_loss_nats_per_token": 1.7030013203103769
    },
    {
      "epoch": 0.05030154591662058,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0013655667693895474,
      "loss": 1.6853736877441405,
      "step": 80640,
      "train_bpb": 0.5975943843936641,
      "train_bytes": 27951.25,
      "train_loss_nats_per_token": 1.6861898694470576
    },
    {
      "epoch": 0.050433225356192886,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0013649456453282325,
      "loss": 1.6925846099853517,
      "step": 80660,
      "train_bpb": 0.5990186094948922,
      "train_bytes": 28247.73125,
      "train_loss_nats_per_token": 1.6937586303225276
    },
    {
      "epoch": 0.05056490479576519,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0013643245446123243,
      "loss": 1.6992395401000977,
      "step": 80680,
      "train_bpb": 0.5998189241452371,
      "train_bytes": 28377.38125,
      "train_loss_nats_per_token": 1.6993486969437819
    },
    {
      "epoch": 0.050696584235337495,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0013637034673491858,
      "loss": 1.6985652923583985,
      "step": 80700,
      "train_bpb": 0.5995790608914646,
      "train_bytes": 28601.7875,
      "train_loss_nats_per_token": 1.6989992178266677
    },
    {
      "epoch": 0.0508282636749098,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0013630824136461757,
      "loss": 1.6913288116455079,
      "step": 80720,
      "train_bpb": 0.5979826704446489,
      "train_bytes": 28076.4375,
      "train_loss_nats_per_token": 1.6924823177521247
    },
    {
      "epoch": 0.050959943114482105,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0013624613836106495,
      "loss": 1.6821475982666017,
      "step": 80740,
      "train_bpb": 0.5925132347626911,
      "train_bytes": 28550.4875,
      "train_loss_nats_per_token": 1.683530505798816
    },
    {
      "epoch": 0.05109162255405441,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0013618403773499577,
      "loss": 1.6938121795654297,
      "step": 80760,
      "train_bpb": 0.5990967276548247,
      "train_bytes": 28356.73125,
      "train_loss_nats_per_token": 1.694706695733475
    },
    {
      "epoch": 0.051223301993626714,
      "grad_norm": 0.1484375,
      "learning_rate": 0.001361219394971447,
      "loss": 1.6846248626708984,
      "step": 80780,
      "train_bpb": 0.5960775838381449,
      "train_bytes": 28199.98125,
      "train_loss_nats_per_token": 1.684981325751396
    },
    {
      "epoch": 0.05135498143319902,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.00136059843658246,
      "loss": 1.6976186752319335,
      "step": 80800,
      "train_bpb": 0.601620435752007,
      "train_bytes": 28291.60625,
      "train_loss_nats_per_token": 1.6991491647498558
    },
    {
      "epoch": 0.051486660872771324,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.001359977502290335,
      "loss": 1.6946918487548828,
      "step": 80820,
      "train_bpb": 0.5991872231197406,
      "train_bytes": 28389.11875,
      "train_loss_nats_per_token": 1.6953020032534807
    },
    {
      "epoch": 0.05161834031234363,
      "grad_norm": 0.150390625,
      "learning_rate": 0.001359356592202407,
      "loss": 1.7050199508666992,
      "step": 80840,
      "train_bpb": 0.6023857361792035,
      "train_bytes": 28274.58125,
      "train_loss_nats_per_token": 1.7060578689392796
    },
    {
      "epoch": 0.051750019751915934,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0013587357064260052,
      "loss": 1.674178123474121,
      "step": 80860,
      "train_bpb": 0.592723131101603,
      "train_bytes": 28322.6125,
      "train_loss_nats_per_token": 1.6746162019806126
    },
    {
      "epoch": 0.05188169919148824,
      "grad_norm": 0.2001953125,
      "learning_rate": 0.0013581148450684563,
      "loss": 1.6762098312377929,
      "step": 80880,
      "train_bpb": 0.5929896870712654,
      "train_bytes": 28221.55,
      "train_loss_nats_per_token": 1.6761362684276158
    },
    {
      "epoch": 0.05201337863106054,
      "grad_norm": 0.2158203125,
      "learning_rate": 0.0013574940082370815,
      "loss": 1.681930923461914,
      "step": 80900,
      "train_bpb": 0.594372831783757,
      "train_bytes": 28057.475,
      "train_loss_nats_per_token": 1.6823402565193162
    },
    {
      "epoch": 0.052145058070632855,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0013568731960391983,
      "loss": 1.6799665451049806,
      "step": 80920,
      "train_bpb": 0.5963943234370802,
      "train_bytes": 28250.29375,
      "train_loss_nats_per_token": 1.6801026358927404
    },
    {
      "epoch": 0.05227673751020516,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00135625240858212,
      "loss": 1.6872684478759765,
      "step": 80940,
      "train_bpb": 0.5992679289309656,
      "train_bytes": 28408.0125,
      "train_loss_nats_per_token": 1.6870032790904523
    },
    {
      "epoch": 0.052408416949777464,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0013556316459731553,
      "loss": 1.6849714279174806,
      "step": 80960,
      "train_bpb": 0.598131704859762,
      "train_bytes": 28057.80625,
      "train_loss_nats_per_token": 1.685149811903678
    },
    {
      "epoch": 0.05254009638934977,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0013550109083196094,
      "loss": 1.6904672622680663,
      "step": 80980,
      "train_bpb": 0.6005591427305431,
      "train_bytes": 28163.6,
      "train_loss_nats_per_token": 1.6913564162179187
    },
    {
      "epoch": 0.052671775828922074,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0013543901957287822,
      "loss": 1.6871288299560547,
      "step": 81000,
      "train_bpb": 0.5991886744677525,
      "train_bytes": 28242.8875,
      "train_loss_nats_per_token": 1.686946580379365
    },
    {
      "epoch": 0.052671775828922074,
      "eval_loss": 1.5229506492614746,
      "eval_runtime": 73.2826,
      "eval_samples_per_second": 13.646,
      "eval_steps_per_second": 13.646,
      "step": 81000,
      "train_bpb": 0.537466004762407,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5304950322595583
    },
    {
      "epoch": 0.05280345526849438,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0013537695083079692,
      "loss": 1.6855363845825195,
      "step": 81020,
      "train_bpb": 0.5976324980058169,
      "train_bytes": 28474.7375,
      "train_loss_nats_per_token": 1.6866735188420399
    },
    {
      "epoch": 0.05293513470806668,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0013531488461644627,
      "loss": 1.6917591094970703,
      "step": 81040,
      "train_bpb": 0.5982721282178183,
      "train_bytes": 28385.625,
      "train_loss_nats_per_token": 1.6923448905370817
    },
    {
      "epoch": 0.05306681414763899,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0013525282094055495,
      "loss": 1.6802894592285156,
      "step": 81060,
      "train_bpb": 0.596359287487539,
      "train_bytes": 28148.76875,
      "train_loss_nats_per_token": 1.681963665465098
    },
    {
      "epoch": 0.05319849358721129,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.001351907598138513,
      "loss": 1.6837865829467773,
      "step": 81080,
      "train_bpb": 0.5976407131090891,
      "train_bytes": 27947.95625,
      "train_loss_nats_per_token": 1.684430728090636
    },
    {
      "epoch": 0.0533301730267836,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.001351287012470631,
      "loss": 1.6838973999023437,
      "step": 81100,
      "train_bpb": 0.5944631502256849,
      "train_bytes": 28322.93125,
      "train_loss_nats_per_token": 1.6844525394515266
    },
    {
      "epoch": 0.0534618524663559,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0013506664525091783,
      "loss": 1.6869863510131835,
      "step": 81120,
      "train_bpb": 0.5966706537629671,
      "train_bytes": 28130.7625,
      "train_loss_nats_per_token": 1.6876216537567497
    },
    {
      "epoch": 0.05359353190592821,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0013500459183614238,
      "loss": 1.6802406311035156,
      "step": 81140,
      "train_bpb": 0.5959067724003406,
      "train_bytes": 28193.49375,
      "train_loss_nats_per_token": 1.6806005071252768
    },
    {
      "epoch": 0.05372521134550051,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0013494254101346327,
      "loss": 1.6862648010253907,
      "step": 81160,
      "train_bpb": 0.5942828456153667,
      "train_bytes": 28455.43125,
      "train_loss_nats_per_token": 1.6873109471758496
    },
    {
      "epoch": 0.05385689078507282,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.001348804927936066,
      "loss": 1.6853225708007813,
      "step": 81180,
      "train_bpb": 0.599717286291604,
      "train_bytes": 28371.0625,
      "train_loss_nats_per_token": 1.6857710984639187
    },
    {
      "epoch": 0.05398857022464512,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0013481844718729796,
      "loss": 1.6798797607421876,
      "step": 81200,
      "train_bpb": 0.5957327907494773,
      "train_bytes": 28339.46875,
      "train_loss_nats_per_token": 1.6806442134228317
    },
    {
      "epoch": 0.054120249664217426,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0013475640420526252,
      "loss": 1.6787830352783204,
      "step": 81220,
      "train_bpb": 0.5952667684430958,
      "train_bytes": 28164.06875,
      "train_loss_nats_per_token": 1.6786392574734106
    },
    {
      "epoch": 0.05425192910378973,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.00134694363858225,
      "loss": 1.6826528549194335,
      "step": 81240,
      "train_bpb": 0.5957772839300642,
      "train_bytes": 28377.91875,
      "train_loss_nats_per_token": 1.6832979390755916
    },
    {
      "epoch": 0.05438360854336204,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.001346323261569096,
      "loss": 1.684334945678711,
      "step": 81260,
      "train_bpb": 0.5976851421418901,
      "train_bytes": 28286.10625,
      "train_loss_nats_per_token": 1.6854582107929614
    },
    {
      "epoch": 0.05451528798293435,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.001345702911120402,
      "loss": 1.6822532653808593,
      "step": 81280,
      "train_bpb": 0.5975568759812289,
      "train_bytes": 27937.8125,
      "train_loss_nats_per_token": 1.6843535595091805
    },
    {
      "epoch": 0.05464696742250665,
      "grad_norm": 0.2138671875,
      "learning_rate": 0.0013450825873434007,
      "loss": 1.691354751586914,
      "step": 81300,
      "train_bpb": 0.5953590408009239,
      "train_bytes": 28405.75,
      "train_loss_nats_per_token": 1.6918751791444007
    },
    {
      "epoch": 0.05477864686207896,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0013444622903453214,
      "loss": 1.6942440032958985,
      "step": 81320,
      "train_bpb": 0.5988450382001419,
      "train_bytes": 28572.51875,
      "train_loss_nats_per_token": 1.6947543112831562
    },
    {
      "epoch": 0.05491032630165126,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0013438420202333875,
      "loss": 1.6839797973632813,
      "step": 81340,
      "train_bpb": 0.5957952744820788,
      "train_bytes": 28148.85,
      "train_loss_nats_per_token": 1.6850407356172643
    },
    {
      "epoch": 0.05504200574122357,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.001343221777114819,
      "loss": 1.6895692825317383,
      "step": 81360,
      "train_bpb": 0.5988221716341364,
      "train_bytes": 28404.7625,
      "train_loss_nats_per_token": 1.690611580318598
    },
    {
      "epoch": 0.05517368518079587,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0013426015610968308,
      "loss": 1.68006591796875,
      "step": 81380,
      "train_bpb": 0.5956780641014299,
      "train_bytes": 28345.9375,
      "train_loss_nats_per_token": 1.6810544811329224
    },
    {
      "epoch": 0.055305364620368176,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0013419813722866326,
      "loss": 1.699898910522461,
      "step": 81400,
      "train_bpb": 0.6018253125018296,
      "train_bytes": 28597.46875,
      "train_loss_nats_per_token": 1.7002161461905652
    },
    {
      "epoch": 0.05543704405994048,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0013413612107914304,
      "loss": 1.686505126953125,
      "step": 81420,
      "train_bpb": 0.598405903191489,
      "train_bytes": 28306.25,
      "train_loss_nats_per_token": 1.6881931938589072
    },
    {
      "epoch": 0.055568723499512786,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0013407410767184243,
      "loss": 1.6890113830566407,
      "step": 81440,
      "train_bpb": 0.5968199309697839,
      "train_bytes": 28319.81875,
      "train_loss_nats_per_token": 1.6896096144990544
    },
    {
      "epoch": 0.05570040293908509,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0013401209701748107,
      "loss": 1.6768486022949218,
      "step": 81460,
      "train_bpb": 0.5940226841782027,
      "train_bytes": 28114.75,
      "train_loss_nats_per_token": 1.6784408195537357
    },
    {
      "epoch": 0.055832082378657395,
      "grad_norm": 0.15625,
      "learning_rate": 0.0013395008912677807,
      "loss": 1.6863988876342773,
      "step": 81480,
      "train_bpb": 0.5947682347454777,
      "train_bytes": 28355.88125,
      "train_loss_nats_per_token": 1.6864363728009542
    },
    {
      "epoch": 0.0559637618182297,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0013388808401045204,
      "loss": 1.6969097137451172,
      "step": 81500,
      "train_bpb": 0.5992723495054724,
      "train_bytes": 28433.29375,
      "train_loss_nats_per_token": 1.6974673153582236
    },
    {
      "epoch": 0.056095441257802005,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0013382608167922122,
      "loss": 1.6913000106811524,
      "step": 81520,
      "train_bpb": 0.594259564651408,
      "train_bytes": 28749.76875,
      "train_loss_nats_per_token": 1.6906971982251415
    },
    {
      "epoch": 0.05622712069737431,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0013376408214380324,
      "loss": 1.6933767318725585,
      "step": 81540,
      "train_bpb": 0.6002781752983344,
      "train_bytes": 28297.4,
      "train_loss_nats_per_token": 1.6936884531518033
    },
    {
      "epoch": 0.056358800136946614,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0013370208541491526,
      "loss": 1.6854103088378907,
      "step": 81560,
      "train_bpb": 0.5959389389264093,
      "train_bytes": 28104.83125,
      "train_loss_nats_per_token": 1.6867861034190954
    },
    {
      "epoch": 0.05649047957651892,
      "grad_norm": 0.1875,
      "learning_rate": 0.0013364009150327407,
      "loss": 1.6857458114624024,
      "step": 81580,
      "train_bpb": 0.5948978624013034,
      "train_bytes": 28469.76875,
      "train_loss_nats_per_token": 1.6872653160935398
    },
    {
      "epoch": 0.05662215901609123,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0013357810041959581,
      "loss": 1.6742776870727538,
      "step": 81600,
      "train_bpb": 0.5948578340545728,
      "train_bytes": 28243.8375,
      "train_loss_nats_per_token": 1.675004149899453
    },
    {
      "epoch": 0.056753838455663536,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0013351611217459634,
      "loss": 1.6849441528320312,
      "step": 81620,
      "train_bpb": 0.5991228895975063,
      "train_bytes": 28362.2125,
      "train_loss_nats_per_token": 1.6859486804140924
    },
    {
      "epoch": 0.05688551789523584,
      "grad_norm": 0.138671875,
      "learning_rate": 0.001334541267789908,
      "loss": 1.6755489349365233,
      "step": 81640,
      "train_bpb": 0.5931654382655243,
      "train_bytes": 28613.6375,
      "train_loss_nats_per_token": 1.6766343782636748
    },
    {
      "epoch": 0.057017197334808145,
      "grad_norm": 0.171875,
      "learning_rate": 0.00133392144243494,
      "loss": 1.6879549026489258,
      "step": 81660,
      "train_bpb": 0.5968774085729595,
      "train_bytes": 28350.79375,
      "train_loss_nats_per_token": 1.6881769591775924
    },
    {
      "epoch": 0.05714887677438045,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.001333301645788202,
      "loss": 1.6942483901977539,
      "step": 81680,
      "train_bpb": 0.5989629928717642,
      "train_bytes": 28231.9125,
      "train_loss_nats_per_token": 1.6947133270682553
    },
    {
      "epoch": 0.057280556213952755,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.001332681877956831,
      "loss": 1.6889476776123047,
      "step": 81700,
      "train_bpb": 0.5928232433476497,
      "train_bytes": 28512.875,
      "train_loss_nats_per_token": 1.689164881695767
    },
    {
      "epoch": 0.05741223565352506,
      "grad_norm": 0.2080078125,
      "learning_rate": 0.0013320621390479606,
      "loss": 1.6931175231933593,
      "step": 81720,
      "train_bpb": 0.5984241354618112,
      "train_bytes": 28113.8125,
      "train_loss_nats_per_token": 1.694015530165898
    },
    {
      "epoch": 0.057543915093097364,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0013314424291687174,
      "loss": 1.680215835571289,
      "step": 81740,
      "train_bpb": 0.5941480988635712,
      "train_bytes": 28572.825,
      "train_loss_nats_per_token": 1.6812305670458005
    },
    {
      "epoch": 0.05767559453266967,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0013308227484262252,
      "loss": 1.6919075012207032,
      "step": 81760,
      "train_bpb": 0.5988434987449595,
      "train_bytes": 28308.78125,
      "train_loss_nats_per_token": 1.6931163879822435
    },
    {
      "epoch": 0.057807273972241974,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0013302030969276008,
      "loss": 1.6816888809204102,
      "step": 81780,
      "train_bpb": 0.5964253280642374,
      "train_bytes": 28322.75,
      "train_loss_nats_per_token": 1.6834669838527856
    },
    {
      "epoch": 0.05793895341181428,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0013295834747799572,
      "loss": 1.691514015197754,
      "step": 81800,
      "train_bpb": 0.5986781310134969,
      "train_bytes": 28569.15625,
      "train_loss_nats_per_token": 1.692175339511676
    },
    {
      "epoch": 0.05807063285138658,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0013289638820904017,
      "loss": 1.6840764999389648,
      "step": 81820,
      "train_bpb": 0.5982880806549106,
      "train_bytes": 27968.225,
      "train_loss_nats_per_token": 1.6850603295375506
    },
    {
      "epoch": 0.05820231229095889,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.001328344318966036,
      "loss": 1.6829679489135743,
      "step": 81840,
      "train_bpb": 0.5932890439018076,
      "train_bytes": 28606.8,
      "train_loss_nats_per_token": 1.6842355084978893
    },
    {
      "epoch": 0.05833399173053119,
      "grad_norm": 0.140625,
      "learning_rate": 0.0013277247855139585,
      "loss": 1.6785261154174804,
      "step": 81860,
      "train_bpb": 0.596733608759056,
      "train_bytes": 28184.91875,
      "train_loss_nats_per_token": 1.6789193836144471
    },
    {
      "epoch": 0.0584656711701035,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0013271052818412611,
      "loss": 1.6809038162231444,
      "step": 81880,
      "train_bpb": 0.593515699883417,
      "train_bytes": 28208.3875,
      "train_loss_nats_per_token": 1.6809804884833577
    },
    {
      "epoch": 0.0585973506096758,
      "grad_norm": 0.130859375,
      "learning_rate": 0.00132648580805503,
      "loss": 1.687033462524414,
      "step": 81900,
      "train_bpb": 0.5959941162684491,
      "train_bytes": 28391.85,
      "train_loss_nats_per_token": 1.6881547467966032
    },
    {
      "epoch": 0.05872903004924811,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0013258663642623477,
      "loss": 1.6939193725585937,
      "step": 81920,
      "train_bpb": 0.598597624992096,
      "train_bytes": 28380.10625,
      "train_loss_nats_per_token": 1.6948766566475055
    },
    {
      "epoch": 0.05886070948882042,
      "grad_norm": 0.140625,
      "learning_rate": 0.001325246950570291,
      "loss": 1.6873512268066406,
      "step": 81940,
      "train_bpb": 0.5961164557683413,
      "train_bytes": 28369.0625,
      "train_loss_nats_per_token": 1.6876015013343544
    },
    {
      "epoch": 0.058992388928392724,
      "grad_norm": 0.1484375,
      "learning_rate": 0.001324627567085931,
      "loss": 1.681678009033203,
      "step": 81960,
      "train_bpb": 0.5912127588298427,
      "train_bytes": 28564.2875,
      "train_loss_nats_per_token": 1.681968025560681
    },
    {
      "epoch": 0.05912406836796503,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0013240082139163339,
      "loss": 1.6768363952636718,
      "step": 81980,
      "train_bpb": 0.5962517402448815,
      "train_bytes": 28129.9625,
      "train_loss_nats_per_token": 1.6769990996922117
    },
    {
      "epoch": 0.05925574780753733,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0013233888911685608,
      "loss": 1.6683378219604492,
      "step": 82000,
      "train_bpb": 0.5888887190488946,
      "train_bytes": 28659.0,
      "train_loss_nats_per_token": 1.6685223109770297
    },
    {
      "epoch": 0.05925574780753733,
      "eval_loss": 1.5222294330596924,
      "eval_runtime": 71.85,
      "eval_samples_per_second": 13.918,
      "eval_steps_per_second": 13.918,
      "step": 82000,
      "train_bpb": 0.537197435184247,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5297302501123173
    },
    {
      "epoch": 0.05938742724710964,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0013227695989496675,
      "loss": 1.6948740005493164,
      "step": 82020,
      "train_bpb": 0.5988969045329386,
      "train_bytes": 28266.16875,
      "train_loss_nats_per_token": 1.695329260245143
    },
    {
      "epoch": 0.05951910668668194,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.001322150337366704,
      "loss": 1.6843755722045899,
      "step": 82040,
      "train_bpb": 0.5953331716558405,
      "train_bytes": 28440.0,
      "train_loss_nats_per_token": 1.685396073231479
    },
    {
      "epoch": 0.05965078612625425,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0013215311065267162,
      "loss": 1.6795381546020507,
      "step": 82060,
      "train_bpb": 0.5893803144207287,
      "train_bytes": 28386.68125,
      "train_loss_nats_per_token": 1.6799298632233723
    },
    {
      "epoch": 0.05978246556582655,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0013209119065367432,
      "loss": 1.691514778137207,
      "step": 82080,
      "train_bpb": 0.5942651756762901,
      "train_bytes": 28714.9125,
      "train_loss_nats_per_token": 1.6922541068920434
    },
    {
      "epoch": 0.05991414500539886,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00132029273750382,
      "loss": 1.6803184509277345,
      "step": 82100,
      "train_bpb": 0.5953962020406735,
      "train_bytes": 28146.06875,
      "train_loss_nats_per_token": 1.680925382604465
    },
    {
      "epoch": 0.06004582444497116,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0013196735995349752,
      "loss": 1.6839012145996093,
      "step": 82120,
      "train_bpb": 0.5981912802481295,
      "train_bytes": 27993.54375,
      "train_loss_nats_per_token": 1.6848839469513253
    },
    {
      "epoch": 0.060177503884543466,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0013190544927372335,
      "loss": 1.6753915786743163,
      "step": 82140,
      "train_bpb": 0.5950881407724389,
      "train_bytes": 28225.65625,
      "train_loss_nats_per_token": 1.6760233917265432
    },
    {
      "epoch": 0.06030918332411577,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0013184354172176126,
      "loss": 1.6917684555053711,
      "step": 82160,
      "train_bpb": 0.5976267063729928,
      "train_bytes": 28151.125,
      "train_loss_nats_per_token": 1.6926695288643587
    },
    {
      "epoch": 0.060440862763688076,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0013178163730831255,
      "loss": 1.6925100326538085,
      "step": 82180,
      "train_bpb": 0.5979891743839685,
      "train_bytes": 28228.275,
      "train_loss_nats_per_token": 1.6933360137273372
    },
    {
      "epoch": 0.06057254220326038,
      "grad_norm": 0.177734375,
      "learning_rate": 0.00131719736044078,
      "loss": 1.6837270736694336,
      "step": 82200,
      "train_bpb": 0.5942153659786845,
      "train_bytes": 28424.24375,
      "train_loss_nats_per_token": 1.6844670418595575
    },
    {
      "epoch": 0.060704221642832686,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0013165783793975781,
      "loss": 1.6692951202392579,
      "step": 82220,
      "train_bpb": 0.5961725526060535,
      "train_bytes": 27804.93125,
      "train_loss_nats_per_token": 1.6696532572299319
    },
    {
      "epoch": 0.06083590108240499,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0013159594300605163,
      "loss": 1.6890762329101563,
      "step": 82240,
      "train_bpb": 0.5956621796178674,
      "train_bytes": 28257.63125,
      "train_loss_nats_per_token": 1.689640898011238
    },
    {
      "epoch": 0.060967580521977295,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.0013153405125365858,
      "loss": 1.6790447235107422,
      "step": 82260,
      "train_bpb": 0.5945637975504028,
      "train_bytes": 28189.4625,
      "train_loss_nats_per_token": 1.6792192495480778
    },
    {
      "epoch": 0.06109925996154961,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0013147216269327727,
      "loss": 1.6847352981567383,
      "step": 82280,
      "train_bpb": 0.5949678722683863,
      "train_bytes": 28536.825,
      "train_loss_nats_per_token": 1.6850629200897287
    },
    {
      "epoch": 0.06123093940112191,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.001314102773356057,
      "loss": 1.6775861740112306,
      "step": 82300,
      "train_bpb": 0.5935896945792616,
      "train_bytes": 28397.96875,
      "train_loss_nats_per_token": 1.6776273860846072
    },
    {
      "epoch": 0.061362618840694216,
      "grad_norm": 0.15625,
      "learning_rate": 0.001313483951913413,
      "loss": 1.680905532836914,
      "step": 82320,
      "train_bpb": 0.5942433124993103,
      "train_bytes": 28523.10625,
      "train_loss_nats_per_token": 1.6813907249631592
    },
    {
      "epoch": 0.06149429828026652,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00131286516271181,
      "loss": 1.67979736328125,
      "step": 82340,
      "train_bpb": 0.594896114137574,
      "train_bytes": 28318.91875,
      "train_loss_nats_per_token": 1.6799454935873213
    },
    {
      "epoch": 0.061625977719838826,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0013122464058582116,
      "loss": 1.669745635986328,
      "step": 82360,
      "train_bpb": 0.5908309718015442,
      "train_bytes": 28212.575,
      "train_loss_nats_per_token": 1.66958328446011
    },
    {
      "epoch": 0.06175765715941113,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.001311627681459575,
      "loss": 1.690546417236328,
      "step": 82380,
      "train_bpb": 0.6014464348582302,
      "train_bytes": 27861.35625,
      "train_loss_nats_per_token": 1.6914439927825105
    },
    {
      "epoch": 0.061889336598983435,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0013110089896228537,
      "loss": 1.6829914093017577,
      "step": 82400,
      "train_bpb": 0.5958913159029654,
      "train_bytes": 28348.16875,
      "train_loss_nats_per_token": 1.6835372656293528
    },
    {
      "epoch": 0.06202101603855574,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0013103903304549935,
      "loss": 1.6894540786743164,
      "step": 82420,
      "train_bpb": 0.5968151258693456,
      "train_bytes": 28414.6875,
      "train_loss_nats_per_token": 1.6898638482246988
    },
    {
      "epoch": 0.062152695478128045,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0013097717040629353,
      "loss": 1.6806690216064453,
      "step": 82440,
      "train_bpb": 0.5937028031978341,
      "train_bytes": 28297.40625,
      "train_loss_nats_per_token": 1.6815828786674487
    },
    {
      "epoch": 0.06228437491770035,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0013091531105536148,
      "loss": 1.682087516784668,
      "step": 82460,
      "train_bpb": 0.5972473032642082,
      "train_bytes": 28258.13125,
      "train_loss_nats_per_token": 1.6833673955413209
    },
    {
      "epoch": 0.062416054357272654,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0013085345500339618,
      "loss": 1.6807971954345704,
      "step": 82480,
      "train_bpb": 0.5980176509016794,
      "train_bytes": 28232.3875,
      "train_loss_nats_per_token": 1.6813824915967301
    },
    {
      "epoch": 0.06254773379684496,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.0013079160226109002,
      "loss": 1.6841974258422852,
      "step": 82500,
      "train_bpb": 0.5953580061951356,
      "train_bytes": 28161.9625,
      "train_loss_nats_per_token": 1.6854190482759874
    },
    {
      "epoch": 0.06267941323641726,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0013072975283913478,
      "loss": 1.6862964630126953,
      "step": 82520,
      "train_bpb": 0.5943516054519515,
      "train_bytes": 28554.025,
      "train_loss_nats_per_token": 1.6873112802804604
    },
    {
      "epoch": 0.06281109267598957,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0013066790674822178,
      "loss": 1.6802627563476562,
      "step": 82540,
      "train_bpb": 0.594431328153589,
      "train_bytes": 28320.99375,
      "train_loss_nats_per_token": 1.681489502200706
    },
    {
      "epoch": 0.06294277211556187,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0013060606399904167,
      "loss": 1.6897079467773437,
      "step": 82560,
      "train_bpb": 0.599003058746694,
      "train_bytes": 28137.0625,
      "train_loss_nats_per_token": 1.6910657582334425
    },
    {
      "epoch": 0.06307445155513418,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0013054422460228449,
      "loss": 1.6940731048583983,
      "step": 82580,
      "train_bpb": 0.5958966880422284,
      "train_bytes": 28787.95,
      "train_loss_nats_per_token": 1.6947679461757257
    },
    {
      "epoch": 0.06320613099470648,
      "grad_norm": 0.19140625,
      "learning_rate": 0.001304823885686398,
      "loss": 1.695199203491211,
      "step": 82600,
      "train_bpb": 0.6011735529612061,
      "train_bytes": 28353.45625,
      "train_loss_nats_per_token": 1.6964421442802085
    },
    {
      "epoch": 0.06333781043427879,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.001304205559087966,
      "loss": 1.6792963027954102,
      "step": 82620,
      "train_bpb": 0.5941510862470603,
      "train_bytes": 28292.175,
      "train_loss_nats_per_token": 1.6795129179622679
    },
    {
      "epoch": 0.06346948987385109,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0013035872663344314,
      "loss": 1.6851119995117188,
      "step": 82640,
      "train_bpb": 0.5927201308218829,
      "train_bytes": 28468.7,
      "train_loss_nats_per_token": 1.685608994505385
    },
    {
      "epoch": 0.0636011693134234,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0013029690075326721,
      "loss": 1.680490493774414,
      "step": 82660,
      "train_bpb": 0.5948308275463122,
      "train_bytes": 28303.175,
      "train_loss_nats_per_token": 1.6817729729710331
    },
    {
      "epoch": 0.0637328487529957,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0013023507827895605,
      "loss": 1.6868883132934571,
      "step": 82680,
      "train_bpb": 0.5940003318092067,
      "train_bytes": 28587.09375,
      "train_loss_nats_per_token": 1.6874637038447382
    },
    {
      "epoch": 0.06386452819256801,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.001301732592211962,
      "loss": 1.696150779724121,
      "step": 82700,
      "train_bpb": 0.5974968784973308,
      "train_bytes": 28412.225,
      "train_loss_nats_per_token": 1.6973483685223838
    },
    {
      "epoch": 0.06399620763214031,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0013011144359067366,
      "loss": 1.680854606628418,
      "step": 82720,
      "train_bpb": 0.5950394717700961,
      "train_bytes": 28118.5125,
      "train_loss_nats_per_token": 1.6817978954623674
    },
    {
      "epoch": 0.06412788707171262,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0013004963139807387,
      "loss": 1.6959037780761719,
      "step": 82740,
      "train_bpb": 0.6012350808210131,
      "train_bytes": 28108.9,
      "train_loss_nats_per_token": 1.6961423169331091
    },
    {
      "epoch": 0.06425956651128492,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0012998782265408162,
      "loss": 1.6861019134521484,
      "step": 82760,
      "train_bpb": 0.5959673887000397,
      "train_bytes": 28291.86875,
      "train_loss_nats_per_token": 1.6867593761219064
    },
    {
      "epoch": 0.06439124595085724,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0012992601736938113,
      "loss": 1.6949066162109374,
      "step": 82780,
      "train_bpb": 0.5992058293947558,
      "train_bytes": 28232.0625,
      "train_loss_nats_per_token": 1.6958707050726591
    },
    {
      "epoch": 0.06452292539042954,
      "grad_norm": 0.12890625,
      "learning_rate": 0.00129864215554656,
      "loss": 1.6782894134521484,
      "step": 82800,
      "train_bpb": 0.593490311937132,
      "train_bytes": 28235.76875,
      "train_loss_nats_per_token": 1.6788559225302655
    },
    {
      "epoch": 0.06465460483000185,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0012980241722058933,
      "loss": 1.692189598083496,
      "step": 82820,
      "train_bpb": 0.6002851141269571,
      "train_bytes": 28096.46875,
      "train_loss_nats_per_token": 1.692920252967665
    },
    {
      "epoch": 0.06478628426957415,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0012974062237786348,
      "loss": 1.675914192199707,
      "step": 82840,
      "train_bpb": 0.5962252878940378,
      "train_bytes": 27838.2875,
      "train_loss_nats_per_token": 1.6770589083347758
    },
    {
      "epoch": 0.06491796370914646,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0012967883103716022,
      "loss": 1.6768795013427735,
      "step": 82860,
      "train_bpb": 0.5939243774218711,
      "train_bytes": 28264.1375,
      "train_loss_nats_per_token": 1.6781834660547266
    },
    {
      "epoch": 0.06504964314871876,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0012961704320916087,
      "loss": 1.6827678680419922,
      "step": 82880,
      "train_bpb": 0.5946008034786568,
      "train_bytes": 28294.7125,
      "train_loss_nats_per_token": 1.6827556028083162
    },
    {
      "epoch": 0.06518132258829107,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0012955525890454597,
      "loss": 1.679183578491211,
      "step": 82900,
      "train_bpb": 0.593721729662134,
      "train_bytes": 28634.35,
      "train_loss_nats_per_token": 1.6797283016464732
    },
    {
      "epoch": 0.06531300202786337,
      "grad_norm": 0.126953125,
      "learning_rate": 0.001294934781339955,
      "loss": 1.6774284362792968,
      "step": 82920,
      "train_bpb": 0.592643021355426,
      "train_bytes": 28285.525,
      "train_loss_nats_per_token": 1.6781111551253303
    },
    {
      "epoch": 0.06544468146743568,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0012943170090818886,
      "loss": 1.6810956954956056,
      "step": 82940,
      "train_bpb": 0.5897408910224495,
      "train_bytes": 28600.675,
      "train_loss_nats_per_token": 1.6813025218492579
    },
    {
      "epoch": 0.06557636090700798,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0012936992723780487,
      "loss": 1.6848194122314453,
      "step": 82960,
      "train_bpb": 0.5914707940971033,
      "train_bytes": 28733.55625,
      "train_loss_nats_per_token": 1.6847560378412305
    },
    {
      "epoch": 0.06570804034658029,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0012930815713352159,
      "loss": 1.6838251113891602,
      "step": 82980,
      "train_bpb": 0.593735729293612,
      "train_bytes": 28220.78125,
      "train_loss_nats_per_token": 1.68448056564424
    },
    {
      "epoch": 0.06583971978615259,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.001292463906060166,
      "loss": 1.6758907318115235,
      "step": 83000,
      "train_bpb": 0.5922398548378753,
      "train_bytes": 28349.15625,
      "train_loss_nats_per_token": 1.6760083910422265
    },
    {
      "epoch": 0.06583971978615259,
      "eval_loss": 1.5229840278625488,
      "eval_runtime": 71.2674,
      "eval_samples_per_second": 14.032,
      "eval_steps_per_second": 14.032,
      "step": 83000,
      "train_bpb": 0.5374740805164678,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.530518028879554
    },
    {
      "epoch": 0.0659713992257249,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0012918462766596686,
      "loss": 1.6808719635009766,
      "step": 83020,
      "train_bpb": 0.5951421701363435,
      "train_bytes": 28196.6375,
      "train_loss_nats_per_token": 1.6820281737441887
    },
    {
      "epoch": 0.0661030786652972,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0012912286832404865,
      "loss": 1.678431510925293,
      "step": 83040,
      "train_bpb": 0.5981603362505087,
      "train_bytes": 27928.84375,
      "train_loss_nats_per_token": 1.6789165600657816
    },
    {
      "epoch": 0.0662347581048695,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0012906111259093761,
      "loss": 1.6820060729980468,
      "step": 83060,
      "train_bpb": 0.5941953216516153,
      "train_bytes": 28291.24375,
      "train_loss_nats_per_token": 1.6833451576290603
    },
    {
      "epoch": 0.06636643754444181,
      "grad_norm": 0.15625,
      "learning_rate": 0.001289993604773088,
      "loss": 1.6932994842529296,
      "step": 83080,
      "train_bpb": 0.597758100940032,
      "train_bytes": 28304.475,
      "train_loss_nats_per_token": 1.6936783797517863
    },
    {
      "epoch": 0.06649811698401412,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0012893761199383677,
      "loss": 1.6855173110961914,
      "step": 83100,
      "train_bpb": 0.5961728450592921,
      "train_bytes": 28189.30625,
      "train_loss_nats_per_token": 1.6857721918886852
    },
    {
      "epoch": 0.06662979642358642,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0012887586715119516,
      "loss": 1.681027603149414,
      "step": 83120,
      "train_bpb": 0.5951163724205054,
      "train_bytes": 28339.5375,
      "train_loss_nats_per_token": 1.6818243624762845
    },
    {
      "epoch": 0.06676147586315873,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0012881412596005722,
      "loss": 1.683403778076172,
      "step": 83140,
      "train_bpb": 0.5941547390454629,
      "train_bytes": 28117.2,
      "train_loss_nats_per_token": 1.6839303448624137
    },
    {
      "epoch": 0.06689315530273103,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.001287523884310955,
      "loss": 1.6665287017822266,
      "step": 83160,
      "train_bpb": 0.5886895295700251,
      "train_bytes": 28179.66875,
      "train_loss_nats_per_token": 1.6667956721099724
    },
    {
      "epoch": 0.06702483474230334,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0012869065457498188,
      "loss": 1.6845399856567382,
      "step": 83180,
      "train_bpb": 0.5980788373445097,
      "train_bytes": 28448.2875,
      "train_loss_nats_per_token": 1.6857837520199026
    },
    {
      "epoch": 0.06715651418187564,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0012862892440238761,
      "loss": 1.6857242584228516,
      "step": 83200,
      "train_bpb": 0.5931419407141494,
      "train_bytes": 28480.55625,
      "train_loss_nats_per_token": 1.6872846104498898
    },
    {
      "epoch": 0.06728819362144794,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0012856719792398341,
      "loss": 1.6749343872070312,
      "step": 83220,
      "train_bpb": 0.5988295839188081,
      "train_bytes": 27911.1,
      "train_loss_nats_per_token": 1.675248849053073
    },
    {
      "epoch": 0.06741987306102025,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0012850547515043922,
      "loss": 1.694100570678711,
      "step": 83240,
      "train_bpb": 0.5970350837320166,
      "train_bytes": 28410.3375,
      "train_loss_nats_per_token": 1.6955799035875339
    },
    {
      "epoch": 0.06755155250059255,
      "grad_norm": 0.18359375,
      "learning_rate": 0.001284437560924244,
      "loss": 1.6757266998291016,
      "step": 83260,
      "train_bpb": 0.5905948814400499,
      "train_bytes": 28219.4,
      "train_loss_nats_per_token": 1.676063515507613
    },
    {
      "epoch": 0.06768323194016486,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0012838204076060765,
      "loss": 1.6788202285766602,
      "step": 83280,
      "train_bpb": 0.5938739246152929,
      "train_bytes": 28268.975,
      "train_loss_nats_per_token": 1.6801804017189195
    },
    {
      "epoch": 0.06781491137973716,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0012832032916565712,
      "loss": 1.6686717987060546,
      "step": 83300,
      "train_bpb": 0.5918900790325045,
      "train_bytes": 28152.89375,
      "train_loss_nats_per_token": 1.6697105063647002
    },
    {
      "epoch": 0.06794659081930947,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0012825862131824013,
      "loss": 1.6725725173950194,
      "step": 83320,
      "train_bpb": 0.589486354460976,
      "train_bytes": 28317.7,
      "train_loss_nats_per_token": 1.6731980356781742
    },
    {
      "epoch": 0.06807827025888177,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.001281969172290235,
      "loss": 1.6723382949829102,
      "step": 83340,
      "train_bpb": 0.5932127911003476,
      "train_bytes": 28275.925,
      "train_loss_nats_per_token": 1.6734732083891644
    },
    {
      "epoch": 0.06820994969845408,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0012813521690867341,
      "loss": 1.6782133102416992,
      "step": 83360,
      "train_bpb": 0.5960903021801499,
      "train_bytes": 28398.6,
      "train_loss_nats_per_token": 1.680023856593821
    },
    {
      "epoch": 0.06834162913802638,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0012807352036785528,
      "loss": 1.6890600204467774,
      "step": 83380,
      "train_bpb": 0.5963652862675256,
      "train_bytes": 28356.76875,
      "train_loss_nats_per_token": 1.6890941655987808
    },
    {
      "epoch": 0.06847330857759869,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0012801182761723391,
      "loss": 1.6843774795532227,
      "step": 83400,
      "train_bpb": 0.5943256817531869,
      "train_bytes": 28156.6375,
      "train_loss_nats_per_token": 1.6861797138411125
    },
    {
      "epoch": 0.06860498801717099,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.001279501386674735,
      "loss": 1.6912076950073243,
      "step": 83420,
      "train_bpb": 0.596704489533858,
      "train_bytes": 28425.675,
      "train_loss_nats_per_token": 1.6919156579890873
    },
    {
      "epoch": 0.0687366674567433,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0012788845352923765,
      "loss": 1.687037467956543,
      "step": 83440,
      "train_bpb": 0.5965374942831492,
      "train_bytes": 28263.1875,
      "train_loss_nats_per_token": 1.6874406602463965
    },
    {
      "epoch": 0.06886834689631562,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0012782677221318906,
      "loss": 1.681224250793457,
      "step": 83460,
      "train_bpb": 0.5945458967725845,
      "train_bytes": 28277.21875,
      "train_loss_nats_per_token": 1.6831264528374157
    },
    {
      "epoch": 0.06900002633588792,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0012776509472998997,
      "loss": 1.6740007400512695,
      "step": 83480,
      "train_bpb": 0.5908404034728957,
      "train_bytes": 28383.65,
      "train_loss_nats_per_token": 1.674554251621095
    },
    {
      "epoch": 0.06913170577546023,
      "grad_norm": 0.171875,
      "learning_rate": 0.0012770342109030201,
      "loss": 1.6848039627075195,
      "step": 83500,
      "train_bpb": 0.595989390769744,
      "train_bytes": 28503.03125,
      "train_loss_nats_per_token": 1.6853146686615568
    },
    {
      "epoch": 0.06926338521503253,
      "grad_norm": 0.1328125,
      "learning_rate": 0.001276417513047859,
      "loss": 1.676354217529297,
      "step": 83520,
      "train_bpb": 0.5918365927719677,
      "train_bytes": 28570.24375,
      "train_loss_nats_per_token": 1.676460698225242
    },
    {
      "epoch": 0.06939506465460483,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0012758008538410188,
      "loss": 1.6712156295776368,
      "step": 83540,
      "train_bpb": 0.5890017427147779,
      "train_bytes": 28521.0375,
      "train_loss_nats_per_token": 1.6716333828919347
    },
    {
      "epoch": 0.06952674409417714,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0012751842333890955,
      "loss": 1.690862274169922,
      "step": 83560,
      "train_bpb": 0.5965732411175031,
      "train_bytes": 28401.225,
      "train_loss_nats_per_token": 1.6911822115734017
    },
    {
      "epoch": 0.06965842353374944,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0012745676517986772,
      "loss": 1.6791257858276367,
      "step": 83580,
      "train_bpb": 0.5932094963516304,
      "train_bytes": 28457.23125,
      "train_loss_nats_per_token": 1.6803305410361524
    },
    {
      "epoch": 0.06979010297332175,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0012739511091763454,
      "loss": 1.6868148803710938,
      "step": 83600,
      "train_bpb": 0.5949259049636684,
      "train_bytes": 28461.30625,
      "train_loss_nats_per_token": 1.6876816497561982
    },
    {
      "epoch": 0.06992178241289405,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0012733346056286756,
      "loss": 1.6835195541381835,
      "step": 83620,
      "train_bpb": 0.5964950584691431,
      "train_bytes": 28125.25,
      "train_loss_nats_per_token": 1.6847095593108046
    },
    {
      "epoch": 0.07005346185246636,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.001272718141262237,
      "loss": 1.6832542419433594,
      "step": 83640,
      "train_bpb": 0.5919468159845679,
      "train_bytes": 28371.2,
      "train_loss_nats_per_token": 1.6842222671318654
    },
    {
      "epoch": 0.07018514129203866,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00127210171618359,
      "loss": 1.6769060134887694,
      "step": 83660,
      "train_bpb": 0.595269486639258,
      "train_bytes": 28250.83125,
      "train_loss_nats_per_token": 1.677093940761592
    },
    {
      "epoch": 0.07031682073161097,
      "grad_norm": 0.2412109375,
      "learning_rate": 0.0012714853304992898,
      "loss": 1.6757551193237306,
      "step": 83680,
      "train_bpb": 0.5941648015492809,
      "train_bytes": 28053.89375,
      "train_loss_nats_per_token": 1.6757125644458537
    },
    {
      "epoch": 0.07044850017118327,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0012708689843158845,
      "loss": 1.6834287643432617,
      "step": 83700,
      "train_bpb": 0.5960613458911683,
      "train_bytes": 28093.25,
      "train_loss_nats_per_token": 1.6834639370779132
    },
    {
      "epoch": 0.07058017961075558,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0012702526777399158,
      "loss": 1.6747058868408202,
      "step": 83720,
      "train_bpb": 0.5940303066613284,
      "train_bytes": 28075.275,
      "train_loss_nats_per_token": 1.6745821094269884
    },
    {
      "epoch": 0.07071185905032788,
      "grad_norm": 0.14453125,
      "learning_rate": 0.001269636410877917,
      "loss": 1.682390022277832,
      "step": 83740,
      "train_bpb": 0.5907670527933492,
      "train_bytes": 28595.6375,
      "train_loss_nats_per_token": 1.6831416920767315
    },
    {
      "epoch": 0.07084353848990019,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0012690201838364162,
      "loss": 1.6782983779907226,
      "step": 83760,
      "train_bpb": 0.589918282266646,
      "train_bytes": 28378.65625,
      "train_loss_nats_per_token": 1.6785699581684503
    },
    {
      "epoch": 0.07097521792947249,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0012684039967219346,
      "loss": 1.6815803527832032,
      "step": 83780,
      "train_bpb": 0.5967180988954729,
      "train_bytes": 27905.84375,
      "train_loss_nats_per_token": 1.6816633162747923
    },
    {
      "epoch": 0.0711068973690448,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0012677878496409848,
      "loss": 1.6723861694335938,
      "step": 83800,
      "train_bpb": 0.594877614296969,
      "train_bytes": 27782.8125,
      "train_loss_nats_per_token": 1.6721579920266751
    },
    {
      "epoch": 0.0712385768086171,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0012671717427000736,
      "loss": 1.6775707244873046,
      "step": 83820,
      "train_bpb": 0.5946114378260482,
      "train_bytes": 27980.21875,
      "train_loss_nats_per_token": 1.678326475874152
    },
    {
      "epoch": 0.0713702562481894,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0012665556760057023,
      "loss": 1.6807912826538085,
      "step": 83840,
      "train_bpb": 0.5994407276111653,
      "train_bytes": 27823.38125,
      "train_loss_nats_per_token": 1.681480887813874
    },
    {
      "epoch": 0.07150193568776171,
      "grad_norm": 0.1640625,
      "learning_rate": 0.001265939649664362,
      "loss": 1.6715967178344726,
      "step": 83860,
      "train_bpb": 0.5910649106268651,
      "train_bytes": 28202.19375,
      "train_loss_nats_per_token": 1.6722455623573624
    },
    {
      "epoch": 0.07163361512733402,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0012653236637825392,
      "loss": 1.6835262298583984,
      "step": 83880,
      "train_bpb": 0.5925094641648339,
      "train_bytes": 28710.83125,
      "train_loss_nats_per_token": 1.6839744465743416
    },
    {
      "epoch": 0.07176529456690632,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0012647077184667133,
      "loss": 1.6812711715698243,
      "step": 83900,
      "train_bpb": 0.5933603709905616,
      "train_bytes": 28570.6375,
      "train_loss_nats_per_token": 1.681652595154694
    },
    {
      "epoch": 0.07189697400647863,
      "grad_norm": 0.19921875,
      "learning_rate": 0.001264091813823356,
      "loss": 1.6734937667846679,
      "step": 83920,
      "train_bpb": 0.5926739318823212,
      "train_bytes": 28275.7875,
      "train_loss_nats_per_token": 1.674163372391998
    },
    {
      "epoch": 0.07202865344605093,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0012634759499589318,
      "loss": 1.6619832992553711,
      "step": 83940,
      "train_bpb": 0.5895946150721557,
      "train_bytes": 28205.5125,
      "train_loss_nats_per_token": 1.6630425432345277
    },
    {
      "epoch": 0.07216033288562324,
      "grad_norm": 0.125,
      "learning_rate": 0.0012628601269798986,
      "loss": 1.6773136138916016,
      "step": 83960,
      "train_bpb": 0.5906999991724646,
      "train_bytes": 28183.35625,
      "train_loss_nats_per_token": 1.6779072924301437
    },
    {
      "epoch": 0.07229201232519554,
      "grad_norm": 0.158203125,
      "learning_rate": 0.001262244344992708,
      "loss": 1.6734001159667968,
      "step": 83980,
      "train_bpb": 0.5897667491007678,
      "train_bytes": 28346.0625,
      "train_loss_nats_per_token": 1.674122685041746
    },
    {
      "epoch": 0.07242369176476784,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0012616286041038025,
      "loss": 1.674492645263672,
      "step": 84000,
      "train_bpb": 0.5924047488616996,
      "train_bytes": 28069.0875,
      "train_loss_nats_per_token": 1.675223674318899
    },
    {
      "epoch": 0.07242369176476784,
      "eval_loss": 1.523167371749878,
      "eval_runtime": 71.9484,
      "eval_samples_per_second": 13.899,
      "eval_steps_per_second": 13.899,
      "step": 84000,
      "train_bpb": 0.5375585936490092,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5307586895508707
    },
    {
      "epoch": 0.07255537120434015,
      "grad_norm": 0.150390625,
      "learning_rate": 0.001261012904419619,
      "loss": 1.6673892974853515,
      "step": 84020,
      "train_bpb": 0.5875464471424601,
      "train_bytes": 28710.3,
      "train_loss_nats_per_token": 1.6677406948781348
    },
    {
      "epoch": 0.07268705064391245,
      "grad_norm": 0.19921875,
      "learning_rate": 0.0012603972460465873,
      "loss": 1.686715316772461,
      "step": 84040,
      "train_bpb": 0.5985901817205982,
      "train_bytes": 28181.46875,
      "train_loss_nats_per_token": 1.688363569211867
    },
    {
      "epoch": 0.07281873008348476,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0012597816290911296,
      "loss": 1.6639463424682617,
      "step": 84060,
      "train_bpb": 0.5899892542726182,
      "train_bytes": 28028.5875,
      "train_loss_nats_per_token": 1.665178019086396
    },
    {
      "epoch": 0.07295040952305706,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0012591660536596606,
      "loss": 1.6831945419311523,
      "step": 84080,
      "train_bpb": 0.5952576168897369,
      "train_bytes": 28073.58125,
      "train_loss_nats_per_token": 1.6843394100014972
    },
    {
      "epoch": 0.07308208896262937,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0012585505198585885,
      "loss": 1.6873212814331056,
      "step": 84100,
      "train_bpb": 0.5908116941197611,
      "train_bytes": 28722.8125,
      "train_loss_nats_per_token": 1.6877513065625733
    },
    {
      "epoch": 0.07321376840220169,
      "grad_norm": 0.130859375,
      "learning_rate": 0.001257935027794315,
      "loss": 1.6916494369506836,
      "step": 84120,
      "train_bpb": 0.5982441744758344,
      "train_bytes": 28480.24375,
      "train_loss_nats_per_token": 1.6920031327804812
    },
    {
      "epoch": 0.07334544784177399,
      "grad_norm": 0.16796875,
      "learning_rate": 0.001257319577573232,
      "loss": 1.6856298446655273,
      "step": 84140,
      "train_bpb": 0.5930342407565562,
      "train_bytes": 28564.575,
      "train_loss_nats_per_token": 1.687197327472941
    },
    {
      "epoch": 0.0734771272813463,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0012567041693017267,
      "loss": 1.6870885848999024,
      "step": 84160,
      "train_bpb": 0.5978531030310676,
      "train_bytes": 28300.73125,
      "train_loss_nats_per_token": 1.6881754177965875
    },
    {
      "epoch": 0.0736088067209186,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0012560888030861788,
      "loss": 1.6756889343261718,
      "step": 84180,
      "train_bpb": 0.5912883522307718,
      "train_bytes": 28500.50625,
      "train_loss_nats_per_token": 1.6760457263720592
    },
    {
      "epoch": 0.0737404861604909,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0012554734790329585,
      "loss": 1.6769149780273438,
      "step": 84200,
      "train_bpb": 0.5927747374604253,
      "train_bytes": 28250.55625,
      "train_loss_nats_per_token": 1.677473505841293
    },
    {
      "epoch": 0.07387216560006321,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0012548581972484313,
      "loss": 1.670102882385254,
      "step": 84220,
      "train_bpb": 0.5926570764280943,
      "train_bytes": 28389.075,
      "train_loss_nats_per_token": 1.671875972115792
    },
    {
      "epoch": 0.07400384503963552,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0012542429578389544,
      "loss": 1.688670539855957,
      "step": 84240,
      "train_bpb": 0.5932592184528639,
      "train_bytes": 28932.0875,
      "train_loss_nats_per_token": 1.688984853259599
    },
    {
      "epoch": 0.07413552447920782,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0012536277609108784,
      "loss": 1.6727073669433594,
      "step": 84260,
      "train_bpb": 0.588974025942274,
      "train_bytes": 28356.775,
      "train_loss_nats_per_token": 1.6732763492831586
    },
    {
      "epoch": 0.07426720391878013,
      "grad_norm": 0.2080078125,
      "learning_rate": 0.001253012606570544,
      "loss": 1.6787992477416993,
      "step": 84280,
      "train_bpb": 0.5912859816972108,
      "train_bytes": 28305.44375,
      "train_loss_nats_per_token": 1.6791139749043136
    },
    {
      "epoch": 0.07439888335835243,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0012523974949242877,
      "loss": 1.6806352615356446,
      "step": 84300,
      "train_bpb": 0.5945268514824742,
      "train_bytes": 28149.7625,
      "train_loss_nats_per_token": 1.6816038405438924
    },
    {
      "epoch": 0.07453056279792473,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0012517824260784376,
      "loss": 1.6816987991333008,
      "step": 84320,
      "train_bpb": 0.591592858695213,
      "train_bytes": 28494.41875,
      "train_loss_nats_per_token": 1.6820504604058737
    },
    {
      "epoch": 0.07466224223749704,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0012511674001393126,
      "loss": 1.6643041610717773,
      "step": 84340,
      "train_bpb": 0.5925685160803889,
      "train_bytes": 28236.93125,
      "train_loss_nats_per_token": 1.6656388709635255
    },
    {
      "epoch": 0.07479392167706934,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0012505524172132267,
      "loss": 1.6841060638427734,
      "step": 84360,
      "train_bpb": 0.5953845941865986,
      "train_bytes": 28527.8375,
      "train_loss_nats_per_token": 1.6839941089253132
    },
    {
      "epoch": 0.07492560111664165,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0012499374774064859,
      "loss": 1.6807327270507812,
      "step": 84380,
      "train_bpb": 0.594875959163365,
      "train_bytes": 28347.9375,
      "train_loss_nats_per_token": 1.681291004680984
    },
    {
      "epoch": 0.07505728055621395,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0012493225808253871,
      "loss": 1.6721319198608398,
      "step": 84400,
      "train_bpb": 0.5902476438375446,
      "train_bytes": 28173.3,
      "train_loss_nats_per_token": 1.673283882932179
    },
    {
      "epoch": 0.07518895999578626,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0012487077275762217,
      "loss": 1.6749977111816405,
      "step": 84420,
      "train_bpb": 0.5926085727139795,
      "train_bytes": 28093.50625,
      "train_loss_nats_per_token": 1.676187489453599
    },
    {
      "epoch": 0.07532063943535856,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0012480929177652723,
      "loss": 1.6766271591186523,
      "step": 84440,
      "train_bpb": 0.5955891409576829,
      "train_bytes": 28071.24375,
      "train_loss_nats_per_token": 1.6756254204311314
    },
    {
      "epoch": 0.07545231887493087,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0012474781514988157,
      "loss": 1.6743719100952148,
      "step": 84460,
      "train_bpb": 0.5908101177676746,
      "train_bytes": 28389.98125,
      "train_loss_nats_per_token": 1.675114907516307
    },
    {
      "epoch": 0.07558399831450317,
      "grad_norm": 0.216796875,
      "learning_rate": 0.0012468634288831184,
      "loss": 1.6838315963745116,
      "step": 84480,
      "train_bpb": 0.5924890754146738,
      "train_bytes": 28453.15625,
      "train_loss_nats_per_token": 1.6843338411000028
    },
    {
      "epoch": 0.07571567775407548,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0012462487500244416,
      "loss": 1.6785634994506835,
      "step": 84500,
      "train_bpb": 0.5947727504416801,
      "train_bytes": 28260.74375,
      "train_loss_nats_per_token": 1.6794235746040287
    },
    {
      "epoch": 0.07584735719364778,
      "grad_norm": 0.2177734375,
      "learning_rate": 0.0012456341150290393,
      "loss": 1.675738525390625,
      "step": 84520,
      "train_bpb": 0.5929915719200777,
      "train_bytes": 28216.26875,
      "train_loss_nats_per_token": 1.6768728557061134
    },
    {
      "epoch": 0.07597903663322009,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.001245019524003155,
      "loss": 1.6829137802124023,
      "step": 84540,
      "train_bpb": 0.5925635433997946,
      "train_bytes": 28656.70625,
      "train_loss_nats_per_token": 1.6822611236452318
    },
    {
      "epoch": 0.07611071607279239,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0012444049770530278,
      "loss": 1.6597244262695312,
      "step": 84560,
      "train_bpb": 0.5915203352913493,
      "train_bytes": 27962.15625,
      "train_loss_nats_per_token": 1.6603143554890798
    },
    {
      "epoch": 0.0762423955123647,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0012437904742848872,
      "loss": 1.6815107345581055,
      "step": 84580,
      "train_bpb": 0.5956666743937562,
      "train_bytes": 28524.29375,
      "train_loss_nats_per_token": 1.6817621624579864
    },
    {
      "epoch": 0.076374074951937,
      "grad_norm": 0.146484375,
      "learning_rate": 0.001243176015804957,
      "loss": 1.671140480041504,
      "step": 84600,
      "train_bpb": 0.5900261451795576,
      "train_bytes": 28241.6875,
      "train_loss_nats_per_token": 1.6718999242132873
    },
    {
      "epoch": 0.0765057543915093,
      "grad_norm": 0.15625,
      "learning_rate": 0.0012425616017194506,
      "loss": 1.6799137115478515,
      "step": 84620,
      "train_bpb": 0.5934216696037043,
      "train_bytes": 28327.88125,
      "train_loss_nats_per_token": 1.6809567523928985
    },
    {
      "epoch": 0.07663743383108161,
      "grad_norm": 0.125,
      "learning_rate": 0.0012419472321345754,
      "loss": 1.6734209060668945,
      "step": 84640,
      "train_bpb": 0.5883106470918749,
      "train_bytes": 28446.6625,
      "train_loss_nats_per_token": 1.674195797117619
    },
    {
      "epoch": 0.07676911327065392,
      "grad_norm": 0.171875,
      "learning_rate": 0.0012413329071565323,
      "loss": 1.6815177917480468,
      "step": 84660,
      "train_bpb": 0.5960256353483925,
      "train_bytes": 28450.275,
      "train_loss_nats_per_token": 1.6834029469453553
    },
    {
      "epoch": 0.07690079271022622,
      "grad_norm": 0.15625,
      "learning_rate": 0.0012407186268915114,
      "loss": 1.672262191772461,
      "step": 84680,
      "train_bpb": 0.5895014412136282,
      "train_bytes": 28407.8375,
      "train_loss_nats_per_token": 1.6731020357082371
    },
    {
      "epoch": 0.07703247214979853,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0012401043914456974,
      "loss": 1.6746007919311523,
      "step": 84700,
      "train_bpb": 0.5907673095872524,
      "train_bytes": 28535.11875,
      "train_loss_nats_per_token": 1.6753119228237547
    },
    {
      "epoch": 0.07716415158937083,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0012394902009252675,
      "loss": 1.667833137512207,
      "step": 84720,
      "train_bpb": 0.592816726757976,
      "train_bytes": 27906.83125,
      "train_loss_nats_per_token": 1.6682666496317375
    },
    {
      "epoch": 0.07729583102894313,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0012388760554363889,
      "loss": 1.673788833618164,
      "step": 84740,
      "train_bpb": 0.5953385392864536,
      "train_bytes": 28006.84375,
      "train_loss_nats_per_token": 1.6747798470998143
    },
    {
      "epoch": 0.07742751046851544,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0012382619550852228,
      "loss": 1.679342269897461,
      "step": 84760,
      "train_bpb": 0.5941178626965197,
      "train_bytes": 28408.6875,
      "train_loss_nats_per_token": 1.6794720641486032
    },
    {
      "epoch": 0.07755918990808774,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0012376478999779226,
      "loss": 1.6867462158203126,
      "step": 84780,
      "train_bpb": 0.5975794750309301,
      "train_bytes": 28398.15,
      "train_loss_nats_per_token": 1.68718671710135
    },
    {
      "epoch": 0.07769086934766006,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0012370338902206338,
      "loss": 1.6609907150268555,
      "step": 84800,
      "train_bpb": 0.5928093634447988,
      "train_bytes": 27646.9375,
      "train_loss_nats_per_token": 1.6621631506018195
    },
    {
      "epoch": 0.07782254878723237,
      "grad_norm": 0.193359375,
      "learning_rate": 0.0012364199259194922,
      "loss": 1.665089988708496,
      "step": 84820,
      "train_bpb": 0.5893341364621879,
      "train_bytes": 28646.2375,
      "train_loss_nats_per_token": 1.6658761401522786
    },
    {
      "epoch": 0.07795422822680467,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0012358060071806288,
      "loss": 1.6792274475097657,
      "step": 84840,
      "train_bpb": 0.5973282569213456,
      "train_bytes": 27847.46875,
      "train_loss_nats_per_token": 1.6799884353528058
    },
    {
      "epoch": 0.07808590766637698,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0012351921341101648,
      "loss": 1.68502197265625,
      "step": 84860,
      "train_bpb": 0.5938210187549081,
      "train_bytes": 28486.85,
      "train_loss_nats_per_token": 1.6862895392761592
    },
    {
      "epoch": 0.07821758710594928,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0012345783068142132,
      "loss": 1.6819900512695312,
      "step": 84880,
      "train_bpb": 0.5915177183360523,
      "train_bytes": 28545.7875,
      "train_loss_nats_per_token": 1.683056167328191
    },
    {
      "epoch": 0.07834926654552159,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0012339645253988802,
      "loss": 1.6831527709960938,
      "step": 84900,
      "train_bpb": 0.5946556315043824,
      "train_bytes": 28617.2875,
      "train_loss_nats_per_token": 1.6851451792124756
    },
    {
      "epoch": 0.07848094598509389,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0012333507899702644,
      "loss": 1.6625829696655274,
      "step": 84920,
      "train_bpb": 0.5898461447948572,
      "train_bytes": 28270.05625,
      "train_loss_nats_per_token": 1.6636932822530839
    },
    {
      "epoch": 0.0786126254246662,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0012327371006344546,
      "loss": 1.6766450881958008,
      "step": 84940,
      "train_bpb": 0.5928607697190058,
      "train_bytes": 28298.7625,
      "train_loss_nats_per_token": 1.6772121883191564
    },
    {
      "epoch": 0.0787443048642385,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.001232123457497533,
      "loss": 1.6801450729370118,
      "step": 84960,
      "train_bpb": 0.5947668369147635,
      "train_bytes": 28246.13125,
      "train_loss_nats_per_token": 1.6810343467653093
    },
    {
      "epoch": 0.0788759843038108,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0012315098606655737,
      "loss": 1.659465217590332,
      "step": 84980,
      "train_bpb": 0.5897708156756434,
      "train_bytes": 28286.625,
      "train_loss_nats_per_token": 1.6597150896052535
    },
    {
      "epoch": 0.07900766374338311,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0012308963102446433,
      "loss": 1.6826343536376953,
      "step": 85000,
      "train_bpb": 0.5951501662745441,
      "train_bytes": 28351.975,
      "train_loss_nats_per_token": 1.6825021934140068
    },
    {
      "epoch": 0.07900766374338311,
      "eval_loss": 1.5235505104064941,
      "eval_runtime": 71.4789,
      "eval_samples_per_second": 13.99,
      "eval_steps_per_second": 13.99,
      "step": 85000,
      "train_bpb": 0.5376893247709781,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5311309611568171
    },
    {
      "epoch": 0.07913934318295542,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0012302828063407985,
      "loss": 1.6908088684082032,
      "step": 85020,
      "train_bpb": 0.5964222849655463,
      "train_bytes": 28410.825,
      "train_loss_nats_per_token": 1.6920217059513292
    },
    {
      "epoch": 0.07927102262252772,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0012296693490600897,
      "loss": 1.668879508972168,
      "step": 85040,
      "train_bpb": 0.5899301987538331,
      "train_bytes": 28565.425,
      "train_loss_nats_per_token": 1.6695563108361078
    },
    {
      "epoch": 0.07940270206210003,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0012290559385085592,
      "loss": 1.679759407043457,
      "step": 85060,
      "train_bpb": 0.5939315239252698,
      "train_bytes": 28371.58125,
      "train_loss_nats_per_token": 1.680145675962945
    },
    {
      "epoch": 0.07953438150167233,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.00122844257479224,
      "loss": 1.68770751953125,
      "step": 85080,
      "train_bpb": 0.5938495691752068,
      "train_bytes": 28622.25,
      "train_loss_nats_per_token": 1.688969370875993
    },
    {
      "epoch": 0.07966606094124463,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0012278292580171581,
      "loss": 1.6801836013793945,
      "step": 85100,
      "train_bpb": 0.5963068059826565,
      "train_bytes": 28146.95,
      "train_loss_nats_per_token": 1.6811844922298713
    },
    {
      "epoch": 0.07979774038081694,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0012272159882893308,
      "loss": 1.6701616287231444,
      "step": 85120,
      "train_bpb": 0.5908099167792166,
      "train_bytes": 28372.25,
      "train_loss_nats_per_token": 1.6704218908438515
    },
    {
      "epoch": 0.07992941982038924,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0012266027657147683,
      "loss": 1.672623062133789,
      "step": 85140,
      "train_bpb": 0.5870581402485355,
      "train_bytes": 28467.06875,
      "train_loss_nats_per_token": 1.672751807248115
    },
    {
      "epoch": 0.08006109925996155,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0012259895903994707,
      "loss": 1.6716455459594726,
      "step": 85160,
      "train_bpb": 0.5897226528890281,
      "train_bytes": 28389.2875,
      "train_loss_nats_per_token": 1.67251942180053
    },
    {
      "epoch": 0.08019277869953385,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0012253764624494316,
      "loss": 1.6705068588256835,
      "step": 85180,
      "train_bpb": 0.5940902357733453,
      "train_bytes": 28201.21875,
      "train_loss_nats_per_token": 1.670883258267984
    },
    {
      "epoch": 0.08032445813910616,
      "grad_norm": 0.2080078125,
      "learning_rate": 0.0012247633819706366,
      "loss": 1.6725255966186523,
      "step": 85200,
      "train_bpb": 0.5924155087369697,
      "train_bytes": 28365.775,
      "train_loss_nats_per_token": 1.6726118806112633
    },
    {
      "epoch": 0.08045613757867846,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0012241503490690612,
      "loss": 1.6676738739013672,
      "step": 85220,
      "train_bpb": 0.5906681250347796,
      "train_bytes": 28225.5,
      "train_loss_nats_per_token": 1.6675684601681324
    },
    {
      "epoch": 0.08058781701825077,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0012235373638506742,
      "loss": 1.6812665939331055,
      "step": 85240,
      "train_bpb": 0.5937753857125508,
      "train_bytes": 28497.1375,
      "train_loss_nats_per_token": 1.6812379644717899
    },
    {
      "epoch": 0.08071949645782307,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0012229244264214368,
      "loss": 1.6727476119995117,
      "step": 85260,
      "train_bpb": 0.5902396102027905,
      "train_bytes": 28378.43125,
      "train_loss_nats_per_token": 1.674503860168765
    },
    {
      "epoch": 0.08085117589739538,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0012223115368872994,
      "loss": 1.6700342178344727,
      "step": 85280,
      "train_bpb": 0.5938037010359797,
      "train_bytes": 28040.00625,
      "train_loss_nats_per_token": 1.670524629145049
    },
    {
      "epoch": 0.08098285533696768,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0012216986953542063,
      "loss": 1.6690286636352538,
      "step": 85300,
      "train_bpb": 0.5931085115459174,
      "train_bytes": 28109.725,
      "train_loss_nats_per_token": 1.6699861360764903
    },
    {
      "epoch": 0.08111453477653999,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0012210859019280932,
      "loss": 1.6714973449707031,
      "step": 85320,
      "train_bpb": 0.5875501109666817,
      "train_bytes": 28787.425,
      "train_loss_nats_per_token": 1.672262510568209
    },
    {
      "epoch": 0.08124621421611229,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0012204731567148875,
      "loss": 1.667624282836914,
      "step": 85340,
      "train_bpb": 0.5908586723701972,
      "train_bytes": 28323.34375,
      "train_loss_nats_per_token": 1.6680938207357348
    },
    {
      "epoch": 0.0813778936556846,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0012198604598205069,
      "loss": 1.67711181640625,
      "step": 85360,
      "train_bpb": 0.5925807267991132,
      "train_bytes": 28293.6625,
      "train_loss_nats_per_token": 1.677431793244826
    },
    {
      "epoch": 0.0815095730952569,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0012192478113508623,
      "loss": 1.6809501647949219,
      "step": 85380,
      "train_bpb": 0.5941380907514697,
      "train_bytes": 28009.03125,
      "train_loss_nats_per_token": 1.6813707404448353
    },
    {
      "epoch": 0.0816412525348292,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0012186352114118564,
      "loss": 1.6726613998413087,
      "step": 85400,
      "train_bpb": 0.594566674059633,
      "train_bytes": 28269.0,
      "train_loss_nats_per_token": 1.6736672785452558
    },
    {
      "epoch": 0.08177293197440151,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0012180226601093812,
      "loss": 1.6713655471801758,
      "step": 85420,
      "train_bpb": 0.5916933510448498,
      "train_bytes": 28131.79375,
      "train_loss_nats_per_token": 1.6720347851989898
    },
    {
      "epoch": 0.08190461141397382,
      "grad_norm": 0.1328125,
      "learning_rate": 0.001217410157549323,
      "loss": 1.6762025833129883,
      "step": 85440,
      "train_bpb": 0.5949904159680873,
      "train_bytes": 28242.725,
      "train_loss_nats_per_token": 1.67629247262428
    },
    {
      "epoch": 0.08203629085354612,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0012167977038375591,
      "loss": 1.6801307678222657,
      "step": 85460,
      "train_bpb": 0.594004644421574,
      "train_bytes": 28255.6625,
      "train_loss_nats_per_token": 1.680574740888238
    },
    {
      "epoch": 0.08216797029311844,
      "grad_norm": 0.171875,
      "learning_rate": 0.0012161852990799566,
      "loss": 1.6685134887695312,
      "step": 85480,
      "train_bpb": 0.5939726849444842,
      "train_bytes": 28134.06875,
      "train_loss_nats_per_token": 1.669315955392979
    },
    {
      "epoch": 0.08229964973269074,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0012155729433823757,
      "loss": 1.6649171829223632,
      "step": 85500,
      "train_bpb": 0.5911812494651205,
      "train_bytes": 28068.825,
      "train_loss_nats_per_token": 1.665124460812262
    },
    {
      "epoch": 0.08243132917226305,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0012149606368506683,
      "loss": 1.6748270034790038,
      "step": 85520,
      "train_bpb": 0.5927270083237721,
      "train_bytes": 28329.9375,
      "train_loss_nats_per_token": 1.6747860933856356
    },
    {
      "epoch": 0.08256300861183535,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0012143483795906775,
      "loss": 1.6630218505859375,
      "step": 85540,
      "train_bpb": 0.5908288106446213,
      "train_bytes": 28159.74375,
      "train_loss_nats_per_token": 1.663021759673548
    },
    {
      "epoch": 0.08269468805140766,
      "grad_norm": 0.1796875,
      "learning_rate": 0.001213736171708237,
      "loss": 1.675245475769043,
      "step": 85560,
      "train_bpb": 0.591103195119461,
      "train_bytes": 28301.91875,
      "train_loss_nats_per_token": 1.6759797654230928
    },
    {
      "epoch": 0.08282636749097996,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0012131240133091724,
      "loss": 1.6899452209472656,
      "step": 85580,
      "train_bpb": 0.5967006886026424,
      "train_bytes": 28430.85,
      "train_loss_nats_per_token": 1.691301703801661
    },
    {
      "epoch": 0.08295804693055227,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0012125119044993024,
      "loss": 1.6766450881958008,
      "step": 85600,
      "train_bpb": 0.5915042404273273,
      "train_bytes": 28459.39375,
      "train_loss_nats_per_token": 1.6782590233211647
    },
    {
      "epoch": 0.08308972637012457,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0012118998453844341,
      "loss": 1.6763154983520507,
      "step": 85620,
      "train_bpb": 0.592111467385238,
      "train_bytes": 28320.75,
      "train_loss_nats_per_token": 1.6767052577519888
    },
    {
      "epoch": 0.08322140580969688,
      "grad_norm": 0.208984375,
      "learning_rate": 0.0012112878360703684,
      "loss": 1.677937889099121,
      "step": 85640,
      "train_bpb": 0.5931987387082387,
      "train_bytes": 28334.4875,
      "train_loss_nats_per_token": 1.6789429520260326
    },
    {
      "epoch": 0.08335308524926918,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0012106758766628968,
      "loss": 1.6800081253051757,
      "step": 85660,
      "train_bpb": 0.5920380847137522,
      "train_bytes": 28210.55,
      "train_loss_nats_per_token": 1.6802173607079522
    },
    {
      "epoch": 0.08348476468884149,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.001210063967267803,
      "loss": 1.6658079147338867,
      "step": 85680,
      "train_bpb": 0.5899573088042601,
      "train_bytes": 28303.0,
      "train_loss_nats_per_token": 1.665938438419021
    },
    {
      "epoch": 0.08361644412841379,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0012094521079908595,
      "loss": 1.6795059204101563,
      "step": 85700,
      "train_bpb": 0.5956184891469047,
      "train_bytes": 28272.70625,
      "train_loss_nats_per_token": 1.6798523941860721
    },
    {
      "epoch": 0.0837481235679861,
      "grad_norm": 0.1484375,
      "learning_rate": 0.001208840298937833,
      "loss": 1.6650808334350586,
      "step": 85720,
      "train_bpb": 0.5914760534617478,
      "train_bytes": 28337.20625,
      "train_loss_nats_per_token": 1.6652585456526077
    },
    {
      "epoch": 0.0838798030075584,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0012082285402144806,
      "loss": 1.6740131378173828,
      "step": 85740,
      "train_bpb": 0.5907107562632401,
      "train_bytes": 28327.89375,
      "train_loss_nats_per_token": 1.6748549810702633
    },
    {
      "epoch": 0.0840114824471307,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0012076168319265497,
      "loss": 1.6671234130859376,
      "step": 85760,
      "train_bpb": 0.5900825750648782,
      "train_bytes": 28233.69375,
      "train_loss_nats_per_token": 1.6680899311738824
    },
    {
      "epoch": 0.08414316188670301,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0012070051741797801,
      "loss": 1.6807863235473632,
      "step": 85780,
      "train_bpb": 0.5959680039573794,
      "train_bytes": 28380.71875,
      "train_loss_nats_per_token": 1.6813870600444307
    },
    {
      "epoch": 0.08427484132627532,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0012063935670799035,
      "loss": 1.6822565078735352,
      "step": 85800,
      "train_bpb": 0.5943128258364268,
      "train_bytes": 28464.28125,
      "train_loss_nats_per_token": 1.682655654380397
    },
    {
      "epoch": 0.08440652076584762,
      "grad_norm": 0.140625,
      "learning_rate": 0.0012057820107326403,
      "loss": 1.6690940856933594,
      "step": 85820,
      "train_bpb": 0.5890613349687244,
      "train_bytes": 28473.3,
      "train_loss_nats_per_token": 1.6701401955418054
    },
    {
      "epoch": 0.08453820020541992,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0012051705052437047,
      "loss": 1.6675775527954102,
      "step": 85840,
      "train_bpb": 0.5906866064057041,
      "train_bytes": 28044.4375,
      "train_loss_nats_per_token": 1.668804161168038
    },
    {
      "epoch": 0.08466987964499223,
      "grad_norm": 0.1484375,
      "learning_rate": 0.001204559050718801,
      "loss": 1.6713602066040039,
      "step": 85860,
      "train_bpb": 0.5902056205776579,
      "train_bytes": 28270.61875,
      "train_loss_nats_per_token": 1.6721248784634528
    },
    {
      "epoch": 0.08480155908456453,
      "grad_norm": 0.130859375,
      "learning_rate": 0.001203947647263625,
      "loss": 1.6638267517089844,
      "step": 85880,
      "train_bpb": 0.5892998723849271,
      "train_bytes": 28242.0,
      "train_loss_nats_per_token": 1.6644620210140446
    },
    {
      "epoch": 0.08493323852413684,
      "grad_norm": 0.134765625,
      "learning_rate": 0.001203336294983863,
      "loss": 1.6731527328491211,
      "step": 85900,
      "train_bpb": 0.5931703310675642,
      "train_bytes": 28103.29375,
      "train_loss_nats_per_token": 1.6741873665204134
    },
    {
      "epoch": 0.08506491796370914,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.001202724993985193,
      "loss": 1.6740076065063476,
      "step": 85920,
      "train_bpb": 0.5913031263504397,
      "train_bytes": 28378.10625,
      "train_loss_nats_per_token": 1.6746306987674473
    },
    {
      "epoch": 0.08519659740328145,
      "grad_norm": 0.140625,
      "learning_rate": 0.0012021137443732851,
      "loss": 1.667241668701172,
      "step": 85940,
      "train_bpb": 0.5891416018379186,
      "train_bytes": 28258.94375,
      "train_loss_nats_per_token": 1.6674914057772763
    },
    {
      "epoch": 0.08532827684285375,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.001201502546253798,
      "loss": 1.6643892288208009,
      "step": 85960,
      "train_bpb": 0.5879962884387508,
      "train_bytes": 28467.9625,
      "train_loss_nats_per_token": 1.6655159754235909
    },
    {
      "epoch": 0.08545995628242606,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0012008913997323838,
      "loss": 1.6715732574462892,
      "step": 85980,
      "train_bpb": 0.5940904284288531,
      "train_bytes": 28075.2125,
      "train_loss_nats_per_token": 1.6726020575151772
    },
    {
      "epoch": 0.08559163572199836,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.0012002803049146855,
      "loss": 1.663926887512207,
      "step": 86000,
      "train_bpb": 0.5882134628786162,
      "train_bytes": 28074.98125,
      "train_loss_nats_per_token": 1.6648019739293258
    },
    {
      "epoch": 0.08559163572199836,
      "eval_loss": 1.5227795839309692,
      "eval_runtime": 72.0208,
      "eval_samples_per_second": 13.885,
      "eval_steps_per_second": 13.885,
      "step": 86000,
      "train_bpb": 0.537440514119413,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5304224447802999
    },
    {
      "epoch": 0.08572331516157067,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.001199669261906335,
      "loss": 1.6782615661621094,
      "step": 86020,
      "train_bpb": 0.5940914428200406,
      "train_bytes": 28345.64375,
      "train_loss_nats_per_token": 1.6794406317724633
    },
    {
      "epoch": 0.08585499460114297,
      "grad_norm": 0.15625,
      "learning_rate": 0.0011990582708129578,
      "loss": 1.6656494140625,
      "step": 86040,
      "train_bpb": 0.5916539341638238,
      "train_bytes": 28285.34375,
      "train_loss_nats_per_token": 1.6656443450127325
    },
    {
      "epoch": 0.08598667404071528,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0011984473317401692,
      "loss": 1.6689268112182618,
      "step": 86060,
      "train_bpb": 0.5877007919183933,
      "train_bytes": 28702.93125,
      "train_loss_nats_per_token": 1.6696737268830995
    },
    {
      "epoch": 0.08611835348028758,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0011978364447935764,
      "loss": 1.6694953918457032,
      "step": 86080,
      "train_bpb": 0.5910402219580901,
      "train_bytes": 28154.44375,
      "train_loss_nats_per_token": 1.6697659026917289
    },
    {
      "epoch": 0.08625003291985989,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.0011972256100787756,
      "loss": 1.6693553924560547,
      "step": 86100,
      "train_bpb": 0.5925599584522491,
      "train_bytes": 28336.2625,
      "train_loss_nats_per_token": 1.6698089032440984
    },
    {
      "epoch": 0.08638171235943219,
      "grad_norm": 0.140625,
      "learning_rate": 0.001196614827701356,
      "loss": 1.6602701187133788,
      "step": 86120,
      "train_bpb": 0.5908016258812151,
      "train_bytes": 28315.8375,
      "train_loss_nats_per_token": 1.661310666630756
    },
    {
      "epoch": 0.08651339179900451,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0011960040977668977,
      "loss": 1.6746963500976562,
      "step": 86140,
      "train_bpb": 0.5914965766093578,
      "train_bytes": 28281.73125,
      "train_loss_nats_per_token": 1.6752375746134391
    },
    {
      "epoch": 0.08664507123857682,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0011953934203809697,
      "loss": 1.6590734481811524,
      "step": 86160,
      "train_bpb": 0.5850792135772475,
      "train_bytes": 28436.86875,
      "train_loss_nats_per_token": 1.6594241574667836
    },
    {
      "epoch": 0.08677675067814912,
      "grad_norm": 0.2314453125,
      "learning_rate": 0.001194782795649134,
      "loss": 1.663435173034668,
      "step": 86180,
      "train_bpb": 0.5893134175778533,
      "train_bytes": 28344.675,
      "train_loss_nats_per_token": 1.663868726308109
    },
    {
      "epoch": 0.08690843011772142,
      "grad_norm": 0.2080078125,
      "learning_rate": 0.001194172223676943,
      "loss": 1.661760139465332,
      "step": 86200,
      "train_bpb": 0.5871881441262362,
      "train_bytes": 28270.25625,
      "train_loss_nats_per_token": 1.6625316053089363
    },
    {
      "epoch": 0.08704010955729373,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0011935617045699398,
      "loss": 1.6620721817016602,
      "step": 86220,
      "train_bpb": 0.5901174482168987,
      "train_bytes": 28054.0125,
      "train_loss_nats_per_token": 1.6615667264679226
    },
    {
      "epoch": 0.08717178899686603,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0011929512384336578,
      "loss": 1.6815225601196289,
      "step": 86240,
      "train_bpb": 0.5927697097909344,
      "train_bytes": 28336.54375,
      "train_loss_nats_per_token": 1.6818966231722705
    },
    {
      "epoch": 0.08730346843643834,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0011923408253736217,
      "loss": 1.664779281616211,
      "step": 86260,
      "train_bpb": 0.5905377904018388,
      "train_bytes": 28365.725,
      "train_loss_nats_per_token": 1.6658138833453264
    },
    {
      "epoch": 0.08743514787601064,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0011917304654953484,
      "loss": 1.6672819137573243,
      "step": 86280,
      "train_bpb": 0.5911311110829436,
      "train_bytes": 28335.06875,
      "train_loss_nats_per_token": 1.6683137954236018
    },
    {
      "epoch": 0.08756682731558295,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0011911201589043425,
      "loss": 1.6781049728393556,
      "step": 86300,
      "train_bpb": 0.5917089879849037,
      "train_bytes": 28539.89375,
      "train_loss_nats_per_token": 1.6786828347427205
    },
    {
      "epoch": 0.08769850675515525,
      "grad_norm": 0.1953125,
      "learning_rate": 0.001190509905706102,
      "loss": 1.6516740798950196,
      "step": 86320,
      "train_bpb": 0.5843443333338219,
      "train_bytes": 28662.225,
      "train_loss_nats_per_token": 1.65222873173356
    },
    {
      "epoch": 0.08783018619472756,
      "grad_norm": 0.205078125,
      "learning_rate": 0.0011898997060061159,
      "loss": 1.6738998413085937,
      "step": 86340,
      "train_bpb": 0.5963187731403938,
      "train_bytes": 27623.2,
      "train_loss_nats_per_token": 1.6748882325657113
    },
    {
      "epoch": 0.08796186563429986,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.001189289559909861,
      "loss": 1.6831855773925781,
      "step": 86360,
      "train_bpb": 0.595199346773047,
      "train_bytes": 28259.25,
      "train_loss_nats_per_token": 1.6832264133306112
    },
    {
      "epoch": 0.08809354507387217,
      "grad_norm": 0.2138671875,
      "learning_rate": 0.0011886794675228076,
      "loss": 1.6758611679077149,
      "step": 86380,
      "train_bpb": 0.5921643755093319,
      "train_bytes": 28257.14375,
      "train_loss_nats_per_token": 1.676844314658929
    },
    {
      "epoch": 0.08822522451344447,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0011880694289504161,
      "loss": 1.673537254333496,
      "step": 86400,
      "train_bpb": 0.5941292838025998,
      "train_bytes": 28196.6625,
      "train_loss_nats_per_token": 1.6741691931154026
    },
    {
      "epoch": 0.08835690395301678,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0011874594442981376,
      "loss": 1.6710268020629884,
      "step": 86420,
      "train_bpb": 0.5888901188055795,
      "train_bytes": 28240.6375,
      "train_loss_nats_per_token": 1.6719177694578176
    },
    {
      "epoch": 0.08848858339258908,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.001186849513671413,
      "loss": 1.6709802627563477,
      "step": 86440,
      "train_bpb": 0.5913703020464732,
      "train_bytes": 28311.85,
      "train_loss_nats_per_token": 1.672643073109495
    },
    {
      "epoch": 0.08862026283216139,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0011862396371756744,
      "loss": 1.6747396469116211,
      "step": 86460,
      "train_bpb": 0.5927704574539034,
      "train_bytes": 28116.4375,
      "train_loss_nats_per_token": 1.6755163234965187
    },
    {
      "epoch": 0.08875194227173369,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0011856298149163459,
      "loss": 1.6689535140991212,
      "step": 86480,
      "train_bpb": 0.5872516066866604,
      "train_bytes": 28578.08125,
      "train_loss_nats_per_token": 1.670170625457925
    },
    {
      "epoch": 0.088883621711306,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0011850200469988395,
      "loss": 1.6686365127563476,
      "step": 86500,
      "train_bpb": 0.5907965660369233,
      "train_bytes": 28280.58125,
      "train_loss_nats_per_token": 1.6693251383209875
    },
    {
      "epoch": 0.0890153011508783,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0011844103335285597,
      "loss": 1.6618932723999023,
      "step": 86520,
      "train_bpb": 0.5901601736832646,
      "train_bytes": 28323.4,
      "train_loss_nats_per_token": 1.6634430356914522
    },
    {
      "epoch": 0.0891469805904506,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.0011838006746109013,
      "loss": 1.666280746459961,
      "step": 86540,
      "train_bpb": 0.5920253523426702,
      "train_bytes": 28339.13125,
      "train_loss_nats_per_token": 1.6666495302007487
    },
    {
      "epoch": 0.08927866003002291,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.00118319107035125,
      "loss": 1.6548843383789062,
      "step": 86560,
      "train_bpb": 0.5869139149592614,
      "train_bytes": 28161.75625,
      "train_loss_nats_per_token": 1.6549245917493063
    },
    {
      "epoch": 0.08941033946959522,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.001182581520854981,
      "loss": 1.6690713882446289,
      "step": 86580,
      "train_bpb": 0.590179230458443,
      "train_bytes": 28535.63125,
      "train_loss_nats_per_token": 1.6694332157949607
    },
    {
      "epoch": 0.08954201890916752,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0011819720262274604,
      "loss": 1.670628547668457,
      "step": 86600,
      "train_bpb": 0.5901989611521481,
      "train_bytes": 28510.725,
      "train_loss_nats_per_token": 1.6717345968875694
    },
    {
      "epoch": 0.08967369834873982,
      "grad_norm": 0.197265625,
      "learning_rate": 0.001181362586574046,
      "loss": 1.654507827758789,
      "step": 86620,
      "train_bpb": 0.5833624143658261,
      "train_bytes": 28579.9625,
      "train_loss_nats_per_token": 1.654944820477801
    },
    {
      "epoch": 0.08980537778831213,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0011807532020000843,
      "loss": 1.662590789794922,
      "step": 86640,
      "train_bpb": 0.5873625617511153,
      "train_bytes": 28374.81875,
      "train_loss_nats_per_token": 1.6626360599194927
    },
    {
      "epoch": 0.08993705722788443,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0011801438726109132,
      "loss": 1.6724597930908203,
      "step": 86660,
      "train_bpb": 0.5907188266243275,
      "train_bytes": 28283.0625,
      "train_loss_nats_per_token": 1.6721834647395504
    },
    {
      "epoch": 0.09006873666745674,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0011795345985118617,
      "loss": 1.6685310363769532,
      "step": 86680,
      "train_bpb": 0.5903954405827018,
      "train_bytes": 28156.95,
      "train_loss_nats_per_token": 1.6684248930004104
    },
    {
      "epoch": 0.09020041610702904,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0011789253798082473,
      "loss": 1.6731172561645509,
      "step": 86700,
      "train_bpb": 0.5920966599376674,
      "train_bytes": 28152.33125,
      "train_loss_nats_per_token": 1.6735137868121326
    },
    {
      "epoch": 0.09033209554660135,
      "grad_norm": 0.208984375,
      "learning_rate": 0.00117831621660538,
      "loss": 1.6728673934936524,
      "step": 86720,
      "train_bpb": 0.5908930450124514,
      "train_bytes": 28295.53125,
      "train_loss_nats_per_token": 1.6732766273524249
    },
    {
      "epoch": 0.09046377498617365,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.001177707109008559,
      "loss": 1.6715465545654298,
      "step": 86740,
      "train_bpb": 0.5925646585748746,
      "train_bytes": 28277.3875,
      "train_loss_nats_per_token": 1.6724248414780076
    },
    {
      "epoch": 0.09059545442574596,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0011770980571230748,
      "loss": 1.6690071105957032,
      "step": 86760,
      "train_bpb": 0.5907151887189706,
      "train_bytes": 28303.95,
      "train_loss_nats_per_token": 1.669553911646702
    },
    {
      "epoch": 0.09072713386531826,
      "grad_norm": 0.140625,
      "learning_rate": 0.0011764890610542072,
      "loss": 1.6716501235961914,
      "step": 86780,
      "train_bpb": 0.5910444777733592,
      "train_bytes": 28552.93125,
      "train_loss_nats_per_token": 1.672279992836953
    },
    {
      "epoch": 0.09085881330489057,
      "grad_norm": 0.1875,
      "learning_rate": 0.0011758801209072265,
      "loss": 1.6539083480834962,
      "step": 86800,
      "train_bpb": 0.5863033366743433,
      "train_bytes": 28446.34375,
      "train_loss_nats_per_token": 1.654885103575704
    },
    {
      "epoch": 0.09099049274446289,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0011752712367873951,
      "loss": 1.6672906875610352,
      "step": 86820,
      "train_bpb": 0.5943426335984605,
      "train_bytes": 28061.4125,
      "train_loss_nats_per_token": 1.668253866544174
    },
    {
      "epoch": 0.09112217218403519,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0011746624087999623,
      "loss": 1.6601408004760743,
      "step": 86840,
      "train_bpb": 0.5854840892994955,
      "train_bytes": 28299.50625,
      "train_loss_nats_per_token": 1.6612282276196673
    },
    {
      "epoch": 0.0912538516236075,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0011740536370501713,
      "loss": 1.657553482055664,
      "step": 86860,
      "train_bpb": 0.5896290353282763,
      "train_bytes": 28087.975,
      "train_loss_nats_per_token": 1.658019549226485
    },
    {
      "epoch": 0.0913855310631798,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.001173444921643254,
      "loss": 1.6715084075927735,
      "step": 86880,
      "train_bpb": 0.5908846003306149,
      "train_bytes": 28379.19375,
      "train_loss_nats_per_token": 1.6712656672231643
    },
    {
      "epoch": 0.0915172105027521,
      "grad_norm": 0.1640625,
      "learning_rate": 0.001172836262684431,
      "loss": 1.6624736785888672,
      "step": 86900,
      "train_bpb": 0.5897968357021074,
      "train_bytes": 28325.68125,
      "train_loss_nats_per_token": 1.6631855980592227
    },
    {
      "epoch": 0.09164888994232441,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0011722276602789158,
      "loss": 1.6552536010742187,
      "step": 86920,
      "train_bpb": 0.5836753592590206,
      "train_bytes": 28599.10625,
      "train_loss_nats_per_token": 1.6563828366576798
    },
    {
      "epoch": 0.09178056938189671,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.001171619114531911,
      "loss": 1.6659343719482422,
      "step": 86940,
      "train_bpb": 0.5892627194988175,
      "train_bytes": 28196.21875,
      "train_loss_nats_per_token": 1.6669488909902352
    },
    {
      "epoch": 0.09191224882146902,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.00117101062554861,
      "loss": 1.6610198974609376,
      "step": 86960,
      "train_bpb": 0.5890297349932965,
      "train_bytes": 28112.6875,
      "train_loss_nats_per_token": 1.6608290695373107
    },
    {
      "epoch": 0.09204392826104132,
      "grad_norm": 0.14453125,
      "learning_rate": 0.001170402193434194,
      "loss": 1.6731071472167969,
      "step": 86980,
      "train_bpb": 0.5899333527456573,
      "train_bytes": 28299.3375,
      "train_loss_nats_per_token": 1.6730513572602386
    },
    {
      "epoch": 0.09217560770061363,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0011697938182938378,
      "loss": 1.6603551864624024,
      "step": 87000,
      "train_bpb": 0.5905516181341522,
      "train_bytes": 28168.95625,
      "train_loss_nats_per_token": 1.6613119846615094
    },
    {
      "epoch": 0.09217560770061363,
      "eval_loss": 1.5225749015808105,
      "eval_runtime": 71.5423,
      "eval_samples_per_second": 13.978,
      "eval_steps_per_second": 13.978,
      "step": 87000,
      "train_bpb": 0.5373803621828489,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5302511553605491
    },
    {
      "epoch": 0.09230728714018593,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0011691855002327048,
      "loss": 1.6513519287109375,
      "step": 87020,
      "train_bpb": 0.5872617917298378,
      "train_bytes": 27999.6125,
      "train_loss_nats_per_token": 1.6522576618067302
    },
    {
      "epoch": 0.09243896657975824,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0011685772393559468,
      "loss": 1.669655990600586,
      "step": 87040,
      "train_bpb": 0.5905777350786428,
      "train_bytes": 28376.0625,
      "train_loss_nats_per_token": 1.6710513765799133
    },
    {
      "epoch": 0.09257064601933054,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0011679690357687088,
      "loss": 1.6695013046264648,
      "step": 87060,
      "train_bpb": 0.5897685133190157,
      "train_bytes": 28211.94375,
      "train_loss_nats_per_token": 1.669622825807522
    },
    {
      "epoch": 0.09270232545890285,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.001167360889576124,
      "loss": 1.6703397750854492,
      "step": 87080,
      "train_bpb": 0.5877191811008047,
      "train_bytes": 28522.83125,
      "train_loss_nats_per_token": 1.67078918953937
    },
    {
      "epoch": 0.09283400489847515,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0011667528008833168,
      "loss": 1.662278175354004,
      "step": 87100,
      "train_bpb": 0.5920153151141699,
      "train_bytes": 27652.8125,
      "train_loss_nats_per_token": 1.6632706355286224
    },
    {
      "epoch": 0.09296568433804746,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0011661447697954,
      "loss": 1.6753955841064454,
      "step": 87120,
      "train_bpb": 0.5937687004193959,
      "train_bytes": 28473.30625,
      "train_loss_nats_per_token": 1.6760231685816935
    },
    {
      "epoch": 0.09309736377761976,
      "grad_norm": 0.146484375,
      "learning_rate": 0.001165536796417478,
      "loss": 1.6666479110717773,
      "step": 87140,
      "train_bpb": 0.5906223540943617,
      "train_bytes": 28370.84375,
      "train_loss_nats_per_token": 1.6672366082683117
    },
    {
      "epoch": 0.09322904321719207,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0011649288808546448,
      "loss": 1.6621421813964843,
      "step": 87160,
      "train_bpb": 0.5864066598901386,
      "train_bytes": 28507.5625,
      "train_loss_nats_per_token": 1.6617718295995856
    },
    {
      "epoch": 0.09336072265676437,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0011643210232119838,
      "loss": 1.6809309005737305,
      "step": 87180,
      "train_bpb": 0.5896728608243755,
      "train_bytes": 28548.66875,
      "train_loss_nats_per_token": 1.6811281077325135
    },
    {
      "epoch": 0.09349240209633668,
      "grad_norm": 0.1796875,
      "learning_rate": 0.001163713223594569,
      "loss": 1.670355987548828,
      "step": 87200,
      "train_bpb": 0.5914737791122666,
      "train_bytes": 28477.4125,
      "train_loss_nats_per_token": 1.6708524850704918
    },
    {
      "epoch": 0.09362408153590898,
      "grad_norm": 0.166015625,
      "learning_rate": 0.001163105482107465,
      "loss": 1.6728534698486328,
      "step": 87220,
      "train_bpb": 0.5925013387944075,
      "train_bytes": 28206.9,
      "train_loss_nats_per_token": 1.6734527910594705
    },
    {
      "epoch": 0.09375576097548129,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0011624977988557248,
      "loss": 1.6686277389526367,
      "step": 87240,
      "train_bpb": 0.5943613786794346,
      "train_bytes": 28102.46875,
      "train_loss_nats_per_token": 1.669208735570949
    },
    {
      "epoch": 0.09388744041505359,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0011618901739443917,
      "loss": 1.6749322891235352,
      "step": 87260,
      "train_bpb": 0.5902472996675483,
      "train_bytes": 28641.34375,
      "train_loss_nats_per_token": 1.6756312964920372
    },
    {
      "epoch": 0.0940191198546259,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0011612826074785003,
      "loss": 1.6614625930786133,
      "step": 87280,
      "train_bpb": 0.5892264077570201,
      "train_bytes": 28037.7,
      "train_loss_nats_per_token": 1.6618289175289203
    },
    {
      "epoch": 0.0941507992941982,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0011606750995630743,
      "loss": 1.6734134674072265,
      "step": 87300,
      "train_bpb": 0.5899556772682547,
      "train_bytes": 28525.5,
      "train_loss_nats_per_token": 1.6737091643791515
    },
    {
      "epoch": 0.0942824787337705,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0011600676503031259,
      "loss": 1.6651338577270507,
      "step": 87320,
      "train_bpb": 0.5903399158002779,
      "train_bytes": 28059.30625,
      "train_loss_nats_per_token": 1.665631535008022
    },
    {
      "epoch": 0.09441415817334281,
      "grad_norm": 0.140625,
      "learning_rate": 0.0011594602598036592,
      "loss": 1.6758541107177733,
      "step": 87340,
      "train_bpb": 0.5907867721346829,
      "train_bytes": 28305.175,
      "train_loss_nats_per_token": 1.6774436847103413
    },
    {
      "epoch": 0.09454583761291512,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0011588529281696678,
      "loss": 1.676002311706543,
      "step": 87360,
      "train_bpb": 0.5922807723183089,
      "train_bytes": 28473.5,
      "train_loss_nats_per_token": 1.6769175750629075
    },
    {
      "epoch": 0.09467751705248742,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0011582456555061333,
      "loss": 1.6726741790771484,
      "step": 87380,
      "train_bpb": 0.592189951260311,
      "train_bytes": 28105.48125,
      "train_loss_nats_per_token": 1.6734932298324674
    },
    {
      "epoch": 0.09480919649205972,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0011576384419180292,
      "loss": 1.6744997024536132,
      "step": 87400,
      "train_bpb": 0.5916616243369641,
      "train_bytes": 28304.375,
      "train_loss_nats_per_token": 1.6754294273715817
    },
    {
      "epoch": 0.09494087593163203,
      "grad_norm": 0.1484375,
      "learning_rate": 0.001157031287510319,
      "loss": 1.669887161254883,
      "step": 87420,
      "train_bpb": 0.5890845771112801,
      "train_bytes": 28408.7125,
      "train_loss_nats_per_token": 1.6702250365786357
    },
    {
      "epoch": 0.09507255537120433,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.001156424192387953,
      "loss": 1.6832225799560547,
      "step": 87440,
      "train_bpb": 0.5947772087715947,
      "train_bytes": 28292.96875,
      "train_loss_nats_per_token": 1.683804091892692
    },
    {
      "epoch": 0.09520423481077664,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0011558171566558744,
      "loss": 1.693549346923828,
      "step": 87460,
      "train_bpb": 0.597174188052186,
      "train_bytes": 28268.325,
      "train_loss_nats_per_token": 1.6934335444603366
    },
    {
      "epoch": 0.09533591425034894,
      "grad_norm": 0.15234375,
      "learning_rate": 0.001155210180419015,
      "loss": 1.6674745559692383,
      "step": 87480,
      "train_bpb": 0.585341379160304,
      "train_bytes": 28383.16875,
      "train_loss_nats_per_token": 1.6676267961515077
    },
    {
      "epoch": 0.09546759368992126,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.001154603263782297,
      "loss": 1.6751436233520507,
      "step": 87500,
      "train_bpb": 0.5920685264972915,
      "train_bytes": 28097.05,
      "train_loss_nats_per_token": 1.6759408076309945
    },
    {
      "epoch": 0.09559927312949357,
      "grad_norm": 0.130859375,
      "learning_rate": 0.00115399640685063,
      "loss": 1.6819177627563477,
      "step": 87520,
      "train_bpb": 0.5967360402532546,
      "train_bytes": 28083.03125,
      "train_loss_nats_per_token": 1.682479379614896
    },
    {
      "epoch": 0.09573095256906587,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0011533896097289157,
      "loss": 1.6728656768798829,
      "step": 87540,
      "train_bpb": 0.5907564688579615,
      "train_bytes": 28569.3,
      "train_loss_nats_per_token": 1.6731199595602793
    },
    {
      "epoch": 0.09586263200863818,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0011527828725220455,
      "loss": 1.6761903762817383,
      "step": 87560,
      "train_bpb": 0.5898826292901429,
      "train_bytes": 28664.275,
      "train_loss_nats_per_token": 1.676161491694361
    },
    {
      "epoch": 0.09599431144821048,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.001152176195334898,
      "loss": 1.6733413696289063,
      "step": 87580,
      "train_bpb": 0.5892124396480541,
      "train_bytes": 28502.09375,
      "train_loss_nats_per_token": 1.674040610370329
    },
    {
      "epoch": 0.09612599088778279,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0011515695782723435,
      "loss": 1.6838531494140625,
      "step": 87600,
      "train_bpb": 0.5966900494594509,
      "train_bytes": 28188.85625,
      "train_loss_nats_per_token": 1.684567248798027
    },
    {
      "epoch": 0.09625767032735509,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.001150963021439242,
      "loss": 1.672330093383789,
      "step": 87620,
      "train_bpb": 0.5924168671875597,
      "train_bytes": 28154.26875,
      "train_loss_nats_per_token": 1.6739370117540995
    },
    {
      "epoch": 0.0963893497669274,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0011503565249404428,
      "loss": 1.670847511291504,
      "step": 87640,
      "train_bpb": 0.590174661088782,
      "train_bytes": 28101.98125,
      "train_loss_nats_per_token": 1.6717372243788167
    },
    {
      "epoch": 0.0965210292064997,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0011497500888807833,
      "loss": 1.6740524291992187,
      "step": 87660,
      "train_bpb": 0.5924013022159236,
      "train_bytes": 28399.025,
      "train_loss_nats_per_token": 1.6746585883990255
    },
    {
      "epoch": 0.096652708646072,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0011491437133650923,
      "loss": 1.674867820739746,
      "step": 87680,
      "train_bpb": 0.5943156936015702,
      "train_bytes": 28062.30625,
      "train_loss_nats_per_token": 1.675606381562483
    },
    {
      "epoch": 0.09678438808564431,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0011485373984981878,
      "loss": 1.6699365615844726,
      "step": 87700,
      "train_bpb": 0.5896574102108159,
      "train_bytes": 28530.225,
      "train_loss_nats_per_token": 1.6711379047913828
    },
    {
      "epoch": 0.09691606752521661,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0011479311443848763,
      "loss": 1.6781232833862305,
      "step": 87720,
      "train_bpb": 0.5930406917307421,
      "train_bytes": 28443.68125,
      "train_loss_nats_per_token": 1.6784363803577966
    },
    {
      "epoch": 0.09704774696478892,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0011473249511299547,
      "loss": 1.6656906127929687,
      "step": 87740,
      "train_bpb": 0.5862844833149374,
      "train_bytes": 28660.05,
      "train_loss_nats_per_token": 1.6657155687342344
    },
    {
      "epoch": 0.09717942640436122,
      "grad_norm": 0.140625,
      "learning_rate": 0.0011467188188382096,
      "loss": 1.6831552505493164,
      "step": 87760,
      "train_bpb": 0.5936163826694971,
      "train_bytes": 28397.95625,
      "train_loss_nats_per_token": 1.6831865479745642
    },
    {
      "epoch": 0.09731110584393353,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0011461127476144158,
      "loss": 1.6755645751953125,
      "step": 87780,
      "train_bpb": 0.5928827329682708,
      "train_bytes": 28519.2125,
      "train_loss_nats_per_token": 1.6758010753887072
    },
    {
      "epoch": 0.09744278528350583,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0011455067375633389,
      "loss": 1.674932861328125,
      "step": 87800,
      "train_bpb": 0.5937973909646829,
      "train_bytes": 28364.1125,
      "train_loss_nats_per_token": 1.67553248151111
    },
    {
      "epoch": 0.09757446472307814,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0011449007887897333,
      "loss": 1.6828474044799804,
      "step": 87820,
      "train_bpb": 0.5961538604016919,
      "train_bytes": 28176.15625,
      "train_loss_nats_per_token": 1.6831083887910854
    },
    {
      "epoch": 0.09770614416265044,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0011442949013983437,
      "loss": 1.6827945709228516,
      "step": 87840,
      "train_bpb": 0.5931952140155676,
      "train_bytes": 28056.80625,
      "train_loss_nats_per_token": 1.6832756937292872
    },
    {
      "epoch": 0.09783782360222275,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.001143689075493902,
      "loss": 1.6689956665039063,
      "step": 87860,
      "train_bpb": 0.5921351495222825,
      "train_bytes": 28258.45,
      "train_loss_nats_per_token": 1.66908097128645
    },
    {
      "epoch": 0.09796950304179505,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0011430833111811312,
      "loss": 1.6811298370361327,
      "step": 87880,
      "train_bpb": 0.5916456314079457,
      "train_bytes": 28835.4125,
      "train_loss_nats_per_token": 1.681950384239801
    },
    {
      "epoch": 0.09810118248136736,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0011424776085647444,
      "loss": 1.666173553466797,
      "step": 87900,
      "train_bpb": 0.5882483917819404,
      "train_bytes": 28431.00625,
      "train_loss_nats_per_token": 1.6666777835788757
    },
    {
      "epoch": 0.09823286192093966,
      "grad_norm": 0.125,
      "learning_rate": 0.0011418719677494416,
      "loss": 1.6630800247192383,
      "step": 87920,
      "train_bpb": 0.5895252480707829,
      "train_bytes": 28132.9625,
      "train_loss_nats_per_token": 1.6636898283727768
    },
    {
      "epoch": 0.09836454136051197,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0011412663888399142,
      "loss": 1.675796890258789,
      "step": 87940,
      "train_bpb": 0.5936669282033088,
      "train_bytes": 28222.89375,
      "train_loss_nats_per_token": 1.6766692447696259
    },
    {
      "epoch": 0.09849622080008427,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0011406608719408426,
      "loss": 1.66473388671875,
      "step": 87960,
      "train_bpb": 0.5901738357777141,
      "train_bytes": 28255.83125,
      "train_loss_nats_per_token": 1.664859825618771
    },
    {
      "epoch": 0.09862790023965658,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0011400554171568948,
      "loss": 1.6667367935180664,
      "step": 87980,
      "train_bpb": 0.5874148623817068,
      "train_bytes": 28478.8625,
      "train_loss_nats_per_token": 1.667295884021062
    },
    {
      "epoch": 0.09875957967922888,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.00113945002459273,
      "loss": 1.6639467239379884,
      "step": 88000,
      "train_bpb": 0.5871068885346096,
      "train_bytes": 28539.6875,
      "train_loss_nats_per_token": 1.6641662400609079
    },
    {
      "epoch": 0.09875957967922888,
      "eval_loss": 1.5232092142105103,
      "eval_runtime": 71.3113,
      "eval_samples_per_second": 14.023,
      "eval_steps_per_second": 14.023,
      "step": 88000,
      "train_bpb": 0.5375787564743064,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.530816105468674
    },
    {
      "epoch": 0.09889125911880119,
      "grad_norm": 0.140625,
      "learning_rate": 0.0011388446943529959,
      "loss": 1.6660057067871095,
      "step": 88020,
      "train_bpb": 0.5920057045668123,
      "train_bytes": 28359.0,
      "train_loss_nats_per_token": 1.6671922240034724
    },
    {
      "epoch": 0.09902293855837349,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0011382394265423302,
      "loss": 1.670516014099121,
      "step": 88040,
      "train_bpb": 0.5931207192903881,
      "train_bytes": 28164.1375,
      "train_loss_nats_per_token": 1.672040537636474
    },
    {
      "epoch": 0.0991546179979458,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0011376342212653578,
      "loss": 1.670589065551758,
      "step": 88060,
      "train_bpb": 0.5919112177597051,
      "train_bytes": 28011.75,
      "train_loss_nats_per_token": 1.6709158435680362
    },
    {
      "epoch": 0.0992862974375181,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0011370290786266947,
      "loss": 1.6789680480957032,
      "step": 88080,
      "train_bpb": 0.592042217287551,
      "train_bytes": 28384.45,
      "train_loss_nats_per_token": 1.6788178987478302
    },
    {
      "epoch": 0.0994179768770904,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0011364239987309462,
      "loss": 1.6681821823120118,
      "step": 88100,
      "train_bpb": 0.5895091808172432,
      "train_bytes": 28527.29375,
      "train_loss_nats_per_token": 1.6685241063640632
    },
    {
      "epoch": 0.09954965631666271,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0011358189816827046,
      "loss": 1.6635183334350585,
      "step": 88120,
      "train_bpb": 0.5869176696008002,
      "train_bytes": 28774.61875,
      "train_loss_nats_per_token": 1.6645908401295184
    },
    {
      "epoch": 0.09968133575623501,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0011352140275865533,
      "loss": 1.678643035888672,
      "step": 88140,
      "train_bpb": 0.5947880760812311,
      "train_bytes": 28341.68125,
      "train_loss_nats_per_token": 1.6793747721302703
    },
    {
      "epoch": 0.09981301519580732,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0011346091365470644,
      "loss": 1.663510513305664,
      "step": 88160,
      "train_bpb": 0.5903624748677182,
      "train_bytes": 28012.48125,
      "train_loss_nats_per_token": 1.6639821461469815
    },
    {
      "epoch": 0.09994469463537964,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0011340043086687998,
      "loss": 1.677802276611328,
      "step": 88180,
      "train_bpb": 0.5929164111708853,
      "train_bytes": 28302.79375,
      "train_loss_nats_per_token": 1.6783513710205413
    },
    {
      "epoch": 0.10007637407495194,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.001133399544056308,
      "loss": 1.6771831512451172,
      "step": 88200,
      "train_bpb": 0.5871929377450582,
      "train_bytes": 28783.18125,
      "train_loss_nats_per_token": 1.6781310916032768
    },
    {
      "epoch": 0.10020805351452425,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0011327948428141292,
      "loss": 1.6573051452636718,
      "step": 88220,
      "train_bpb": 0.5883915710555496,
      "train_bytes": 27962.15625,
      "train_loss_nats_per_token": 1.6584086794997654
    },
    {
      "epoch": 0.10033973295409655,
      "grad_norm": 0.17578125,
      "learning_rate": 0.001132190205046792,
      "loss": 1.6682657241821288,
      "step": 88240,
      "train_bpb": 0.5905583064236286,
      "train_bytes": 28425.575,
      "train_loss_nats_per_token": 1.6692130227810835
    },
    {
      "epoch": 0.10047141239366886,
      "grad_norm": 0.125,
      "learning_rate": 0.0011315856308588123,
      "loss": 1.6653078079223633,
      "step": 88260,
      "train_bpb": 0.5910518215366547,
      "train_bytes": 28167.5125,
      "train_loss_nats_per_token": 1.6656622239943997
    },
    {
      "epoch": 0.10060309183324116,
      "grad_norm": 0.1953125,
      "learning_rate": 0.0011309811203546976,
      "loss": 1.6746637344360351,
      "step": 88280,
      "train_bpb": 0.5904675822641714,
      "train_bytes": 28260.15625,
      "train_loss_nats_per_token": 1.6750284023796507
    },
    {
      "epoch": 0.10073477127281347,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0011303766736389434,
      "loss": 1.6878007888793944,
      "step": 88300,
      "train_bpb": 0.5956920057195322,
      "train_bytes": 28409.95,
      "train_loss_nats_per_token": 1.6881909281601888
    },
    {
      "epoch": 0.10086645071238577,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0011297722908160327,
      "loss": 1.6705928802490235,
      "step": 88320,
      "train_bpb": 0.5920251410017399,
      "train_bytes": 28440.35,
      "train_loss_nats_per_token": 1.6716464037765686
    },
    {
      "epoch": 0.10099813015195808,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0011291679719904397,
      "loss": 1.664925765991211,
      "step": 88340,
      "train_bpb": 0.5884356897606039,
      "train_bytes": 28534.025,
      "train_loss_nats_per_token": 1.665708844602257
    },
    {
      "epoch": 0.10112980959153038,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0011285637172666258,
      "loss": 1.6708709716796875,
      "step": 88360,
      "train_bpb": 0.5904967308082341,
      "train_bytes": 28352.14375,
      "train_loss_nats_per_token": 1.6712951671968233
    },
    {
      "epoch": 0.10126148903110269,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0011279595267490435,
      "loss": 1.6739240646362306,
      "step": 88380,
      "train_bpb": 0.5918634424439758,
      "train_bytes": 28258.06875,
      "train_loss_nats_per_token": 1.6746200830791484
    },
    {
      "epoch": 0.10139316847067499,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0011273554005421312,
      "loss": 1.667422103881836,
      "step": 88400,
      "train_bpb": 0.5882167638117516,
      "train_bytes": 28327.425,
      "train_loss_nats_per_token": 1.6676236717664983
    },
    {
      "epoch": 0.1015248479102473,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0011267513387503182,
      "loss": 1.6760103225708007,
      "step": 88420,
      "train_bpb": 0.5974019650358472,
      "train_bytes": 27944.50625,
      "train_loss_nats_per_token": 1.6780278628591652
    },
    {
      "epoch": 0.1016565273498196,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0011261473414780232,
      "loss": 1.6776603698730468,
      "step": 88440,
      "train_bpb": 0.5922281846422475,
      "train_bytes": 28329.70625,
      "train_loss_nats_per_token": 1.6793481736388922
    },
    {
      "epoch": 0.1017882067893919,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0011255434088296512,
      "loss": 1.6810413360595704,
      "step": 88460,
      "train_bpb": 0.5909777767314205,
      "train_bytes": 28632.5125,
      "train_loss_nats_per_token": 1.6810157618146253
    },
    {
      "epoch": 0.10191988622896421,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0011249395409095985,
      "loss": 1.6690944671630858,
      "step": 88480,
      "train_bpb": 0.5902942539105662,
      "train_bytes": 28436.8875,
      "train_loss_nats_per_token": 1.66947345162981
    },
    {
      "epoch": 0.10205156566853651,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0011243357378222499,
      "loss": 1.6702360153198241,
      "step": 88500,
      "train_bpb": 0.5911096715535124,
      "train_bytes": 28545.91875,
      "train_loss_nats_per_token": 1.6709399364686994
    },
    {
      "epoch": 0.10218324510810882,
      "grad_norm": 0.14453125,
      "learning_rate": 0.001123731999671977,
      "loss": 1.6735252380371093,
      "step": 88520,
      "train_bpb": 0.5873535847046618,
      "train_bytes": 28675.79375,
      "train_loss_nats_per_token": 1.6734871737226638
    },
    {
      "epoch": 0.10231492454768112,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.001123128326563142,
      "loss": 1.6624927520751953,
      "step": 88540,
      "train_bpb": 0.5883811499841098,
      "train_bytes": 28016.675,
      "train_loss_nats_per_token": 1.663558627866
    },
    {
      "epoch": 0.10244660398725343,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0011225247186000958,
      "loss": 1.6699954986572265,
      "step": 88560,
      "train_bpb": 0.5910153108528724,
      "train_bytes": 28121.39375,
      "train_loss_nats_per_token": 1.6708194391827291
    },
    {
      "epoch": 0.10257828342682573,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0011219211758871784,
      "loss": 1.6724540710449218,
      "step": 88580,
      "train_bpb": 0.5900050565707025,
      "train_bytes": 28354.6625,
      "train_loss_nats_per_token": 1.6730940402705798
    },
    {
      "epoch": 0.10270996286639804,
      "grad_norm": 0.2158203125,
      "learning_rate": 0.0011213176985287164,
      "loss": 1.6666105270385743,
      "step": 88600,
      "train_bpb": 0.5922440241006074,
      "train_bytes": 28313.6375,
      "train_loss_nats_per_token": 1.6672511859048136
    },
    {
      "epoch": 0.10284164230597034,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0011207142866290266,
      "loss": 1.6558797836303711,
      "step": 88620,
      "train_bpb": 0.5856061869814234,
      "train_bytes": 28384.5375,
      "train_loss_nats_per_token": 1.65642324574692
    },
    {
      "epoch": 0.10297332174554265,
      "grad_norm": 0.125,
      "learning_rate": 0.0011201109402924159,
      "loss": 1.6693668365478516,
      "step": 88640,
      "train_bpb": 0.5903386800182227,
      "train_bytes": 28312.025,
      "train_loss_nats_per_token": 1.66975207977184
    },
    {
      "epoch": 0.10310500118511495,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0011195076596231764,
      "loss": 1.6798166275024413,
      "step": 88660,
      "train_bpb": 0.5930591568641127,
      "train_bytes": 28359.4125,
      "train_loss_nats_per_token": 1.6806920363141657
    },
    {
      "epoch": 0.10323668062468726,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0011189044447255918,
      "loss": 1.6584182739257813,
      "step": 88680,
      "train_bpb": 0.5857698648752939,
      "train_bytes": 28416.35,
      "train_loss_nats_per_token": 1.6593947876281936
    },
    {
      "epoch": 0.10336836006425956,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0011183012957039335,
      "loss": 1.6567291259765624,
      "step": 88700,
      "train_bpb": 0.5842370218547066,
      "train_bytes": 28463.14375,
      "train_loss_nats_per_token": 1.6577078539078145
    },
    {
      "epoch": 0.10350003950383187,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.001117698212662462,
      "loss": 1.6554121017456054,
      "step": 88720,
      "train_bpb": 0.5819172150065689,
      "train_bytes": 28398.875,
      "train_loss_nats_per_token": 1.656391675668537
    },
    {
      "epoch": 0.10363171894340417,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0011170951957054241,
      "loss": 1.6594512939453125,
      "step": 88740,
      "train_bpb": 0.588434131239655,
      "train_bytes": 28204.2625,
      "train_loss_nats_per_token": 1.66024808312835
    },
    {
      "epoch": 0.10376339838297648,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.001116492244937058,
      "loss": 1.6700307846069335,
      "step": 88760,
      "train_bpb": 0.591729724742275,
      "train_bytes": 28111.35625,
      "train_loss_nats_per_token": 1.6703750966777653
    },
    {
      "epoch": 0.10389507782254878,
      "grad_norm": 0.189453125,
      "learning_rate": 0.00111588936046159,
      "loss": 1.6673368453979491,
      "step": 88780,
      "train_bpb": 0.5882449142364926,
      "train_bytes": 28342.35625,
      "train_loss_nats_per_token": 1.6684033519113883
    },
    {
      "epoch": 0.10402675726212109,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0011152865423832328,
      "loss": 1.6674606323242187,
      "step": 88800,
      "train_bpb": 0.5928432928993652,
      "train_bytes": 28353.4375,
      "train_loss_nats_per_token": 1.6680883872360552
    },
    {
      "epoch": 0.10415843670169339,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.00111468379080619,
      "loss": 1.662862205505371,
      "step": 88820,
      "train_bpb": 0.5917808149092678,
      "train_bytes": 27976.64375,
      "train_loss_nats_per_token": 1.6639769691949093
    },
    {
      "epoch": 0.10429011614126571,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0011140811058346536,
      "loss": 1.659226417541504,
      "step": 88840,
      "train_bpb": 0.5885903293432585,
      "train_bytes": 28380.45,
      "train_loss_nats_per_token": 1.6601655522163161
    },
    {
      "epoch": 0.10442179558083801,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0011134784875728016,
      "loss": 1.6600034713745118,
      "step": 88860,
      "train_bpb": 0.5907437201655246,
      "train_bytes": 28049.86875,
      "train_loss_nats_per_token": 1.66131033392486
    },
    {
      "epoch": 0.10455347502041032,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0011128759361248035,
      "loss": 1.684718132019043,
      "step": 88880,
      "train_bpb": 0.5938793311086437,
      "train_bytes": 28402.5,
      "train_loss_nats_per_token": 1.6845594914901714
    },
    {
      "epoch": 0.10468515445998262,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0011122734515948151,
      "loss": 1.6525197982788087,
      "step": 88900,
      "train_bpb": 0.5834636185966054,
      "train_bytes": 28589.94375,
      "train_loss_nats_per_token": 1.65407343579424
    },
    {
      "epoch": 0.10481683389955493,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0011116710340869826,
      "loss": 1.6791912078857423,
      "step": 88920,
      "train_bpb": 0.5935113682525719,
      "train_bytes": 28303.3375,
      "train_loss_nats_per_token": 1.6795043471009008
    },
    {
      "epoch": 0.10494851333912723,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0011110686837054384,
      "loss": 1.6636697769165039,
      "step": 88940,
      "train_bpb": 0.5843363624358435,
      "train_bytes": 28653.875,
      "train_loss_nats_per_token": 1.6651321016040084
    },
    {
      "epoch": 0.10508019277869954,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0011104664005543048,
      "loss": 1.6734809875488281,
      "step": 88960,
      "train_bpb": 0.5905379268654427,
      "train_bytes": 28377.58125,
      "train_loss_nats_per_token": 1.6739751485198788
    },
    {
      "epoch": 0.10521187221827184,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0011098641847376927,
      "loss": 1.6471677780151368,
      "step": 88980,
      "train_bpb": 0.584364114646745,
      "train_bytes": 28206.21875,
      "train_loss_nats_per_token": 1.648754094708039
    },
    {
      "epoch": 0.10534355165784415,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0011092620363596993,
      "loss": 1.66516170501709,
      "step": 89000,
      "train_bpb": 0.5887320425301442,
      "train_bytes": 28389.2125,
      "train_loss_nats_per_token": 1.665239936537273
    },
    {
      "epoch": 0.10534355165784415,
      "eval_loss": 1.522652506828308,
      "eval_runtime": 73.7991,
      "eval_samples_per_second": 13.55,
      "eval_steps_per_second": 13.55,
      "step": 89000,
      "train_bpb": 0.537400586627503,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5303087467464196
    },
    {
      "epoch": 0.10547523109741645,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0011086599555244123,
      "loss": 1.6778255462646485,
      "step": 89020,
      "train_bpb": 0.5952491045427132,
      "train_bytes": 28230.30625,
      "train_loss_nats_per_token": 1.6781661773670977
    },
    {
      "epoch": 0.10560691053698876,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0011080579423359074,
      "loss": 1.656721305847168,
      "step": 89040,
      "train_bpb": 0.5876909652394445,
      "train_bytes": 28069.29375,
      "train_loss_nats_per_token": 1.6578113511831054
    },
    {
      "epoch": 0.10573858997656106,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0011074559968982482,
      "loss": 1.6627983093261718,
      "step": 89060,
      "train_bpb": 0.5868443416030662,
      "train_bytes": 28454.475,
      "train_loss_nats_per_token": 1.6632666298266483
    },
    {
      "epoch": 0.10587026941613337,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0011068541193154857,
      "loss": 1.6662485122680664,
      "step": 89080,
      "train_bpb": 0.5871640599994792,
      "train_bytes": 28354.025,
      "train_loss_nats_per_token": 1.6668250608238924
    },
    {
      "epoch": 0.10600194885570567,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0011062523096916606,
      "loss": 1.6551013946533204,
      "step": 89100,
      "train_bpb": 0.5862718969085474,
      "train_bytes": 28396.7375,
      "train_loss_nats_per_token": 1.6575043503732503
    },
    {
      "epoch": 0.10613362829527798,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0011056505681308023,
      "loss": 1.6689748764038086,
      "step": 89120,
      "train_bpb": 0.5892921208633939,
      "train_bytes": 28449.425,
      "train_loss_nats_per_token": 1.66907586068156
    },
    {
      "epoch": 0.10626530773485028,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0011050488947369256,
      "loss": 1.661665916442871,
      "step": 89140,
      "train_bpb": 0.5908658795110948,
      "train_bytes": 28092.825,
      "train_loss_nats_per_token": 1.6623096126665335
    },
    {
      "epoch": 0.10639698717442259,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.001104447289614036,
      "loss": 1.668121910095215,
      "step": 89160,
      "train_bpb": 0.5881870416929063,
      "train_bytes": 28255.00625,
      "train_loss_nats_per_token": 1.6694834434057275
    },
    {
      "epoch": 0.10652866661399489,
      "grad_norm": 0.255859375,
      "learning_rate": 0.0011038457528661277,
      "loss": 1.6636648178100586,
      "step": 89180,
      "train_bpb": 0.5891182639461445,
      "train_bytes": 28431.96875,
      "train_loss_nats_per_token": 1.6646710187077751
    },
    {
      "epoch": 0.1066603460535672,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0011032442845971804,
      "loss": 1.6499715805053712,
      "step": 89200,
      "train_bpb": 0.5827789687315479,
      "train_bytes": 28450.075,
      "train_loss_nats_per_token": 1.650648328579086
    },
    {
      "epoch": 0.1067920254931395,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0011026428849111643,
      "loss": 1.6622631072998046,
      "step": 89220,
      "train_bpb": 0.586934922887601,
      "train_bytes": 28348.65,
      "train_loss_nats_per_token": 1.662597241830455
    },
    {
      "epoch": 0.1069237049327118,
      "grad_norm": 0.125,
      "learning_rate": 0.0011020415539120367,
      "loss": 1.67215576171875,
      "step": 89240,
      "train_bpb": 0.5890577900331692,
      "train_bytes": 28210.275,
      "train_loss_nats_per_token": 1.6733838133762748
    },
    {
      "epoch": 0.10705538437228411,
      "grad_norm": 0.138671875,
      "learning_rate": 0.001101440291703744,
      "loss": 1.6660005569458007,
      "step": 89260,
      "train_bpb": 0.5890759185516077,
      "train_bytes": 28413.275,
      "train_loss_nats_per_token": 1.6665827977280685
    },
    {
      "epoch": 0.10718706381185641,
      "grad_norm": 0.140625,
      "learning_rate": 0.0011008390983902189,
      "loss": 1.6613252639770508,
      "step": 89280,
      "train_bpb": 0.5867012045877086,
      "train_bytes": 28449.44375,
      "train_loss_nats_per_token": 1.6620757977287643
    },
    {
      "epoch": 0.10731874325142872,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0011002379740753837,
      "loss": 1.6564794540405274,
      "step": 89300,
      "train_bpb": 0.5884060752393176,
      "train_bytes": 28075.46875,
      "train_loss_nats_per_token": 1.6578951757646188
    },
    {
      "epoch": 0.10745042269100102,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0010996369188631487,
      "loss": 1.6716167449951171,
      "step": 89320,
      "train_bpb": 0.5894156998873664,
      "train_bytes": 28384.51875,
      "train_loss_nats_per_token": 1.6721046089897427
    },
    {
      "epoch": 0.10758210213057333,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0010990359328574115,
      "loss": 1.6585203170776368,
      "step": 89340,
      "train_bpb": 0.5896391025934544,
      "train_bytes": 27907.075,
      "train_loss_nats_per_token": 1.659855038766223
    },
    {
      "epoch": 0.10771378157014563,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0010984350161620578,
      "loss": 1.6747468948364257,
      "step": 89360,
      "train_bpb": 0.590492173507608,
      "train_bytes": 28420.4125,
      "train_loss_nats_per_token": 1.6758888135625485
    },
    {
      "epoch": 0.10784546100971794,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0010978341688809633,
      "loss": 1.6593469619750976,
      "step": 89380,
      "train_bpb": 0.585204054543259,
      "train_bytes": 28424.625,
      "train_loss_nats_per_token": 1.6602745260527618
    },
    {
      "epoch": 0.10797714044929024,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0010972333911179878,
      "loss": 1.675338363647461,
      "step": 89400,
      "train_bpb": 0.5902572039643922,
      "train_bytes": 28564.975,
      "train_loss_nats_per_token": 1.6756887915876015
    },
    {
      "epoch": 0.10810881988886255,
      "grad_norm": 0.1875,
      "learning_rate": 0.0010966326829769829,
      "loss": 1.6530267715454101,
      "step": 89420,
      "train_bpb": 0.5871545813463499,
      "train_bytes": 28101.65625,
      "train_loss_nats_per_token": 1.6538721977623618
    },
    {
      "epoch": 0.10824049932843485,
      "grad_norm": 0.140625,
      "learning_rate": 0.0010960320445617857,
      "loss": 1.669907569885254,
      "step": 89440,
      "train_bpb": 0.5908455023365329,
      "train_bytes": 28498.75,
      "train_loss_nats_per_token": 1.670561371918689
    },
    {
      "epoch": 0.10837217876800716,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0010954314759762233,
      "loss": 1.6656139373779297,
      "step": 89460,
      "train_bpb": 0.5911097774983396,
      "train_bytes": 27816.425,
      "train_loss_nats_per_token": 1.6671344583792544
    },
    {
      "epoch": 0.10850385820757946,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0010948309773241086,
      "loss": 1.6617572784423829,
      "step": 89480,
      "train_bpb": 0.5843226033664779,
      "train_bytes": 28607.5625,
      "train_loss_nats_per_token": 1.6624463704832106
    },
    {
      "epoch": 0.10863553764715177,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.0010942305487092437,
      "loss": 1.6547136306762695,
      "step": 89500,
      "train_bpb": 0.5850337568934968,
      "train_bytes": 28458.9375,
      "train_loss_nats_per_token": 1.6560865799097082
    },
    {
      "epoch": 0.10876721708672409,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0010936301902354187,
      "loss": 1.6432645797729493,
      "step": 89520,
      "train_bpb": 0.5860193686763963,
      "train_bytes": 27921.0375,
      "train_loss_nats_per_token": 1.6444659792496334
    },
    {
      "epoch": 0.10889889652629639,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0010930299020064102,
      "loss": 1.6694574356079102,
      "step": 89540,
      "train_bpb": 0.5881922643042964,
      "train_bytes": 28414.725,
      "train_loss_nats_per_token": 1.6693053932185613
    },
    {
      "epoch": 0.1090305759658687,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0010924296841259842,
      "loss": 1.6515174865722657,
      "step": 89560,
      "train_bpb": 0.5842110500076343,
      "train_bytes": 28030.7,
      "train_loss_nats_per_token": 1.6516798742560734
    },
    {
      "epoch": 0.109162255405441,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.001091829536697894,
      "loss": 1.664474868774414,
      "step": 89580,
      "train_bpb": 0.588089209125206,
      "train_bytes": 28606.28125,
      "train_loss_nats_per_token": 1.6657310960126934
    },
    {
      "epoch": 0.1092939348450133,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0010912294598258811,
      "loss": 1.6629226684570313,
      "step": 89600,
      "train_bpb": 0.589089674557064,
      "train_bytes": 28275.54375,
      "train_loss_nats_per_token": 1.6633800596720065
    },
    {
      "epoch": 0.10942561428458561,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0010906294536136734,
      "loss": 1.6579742431640625,
      "step": 89620,
      "train_bpb": 0.5876204629208254,
      "train_bytes": 28294.48125,
      "train_loss_nats_per_token": 1.6587700144956838
    },
    {
      "epoch": 0.10955729372415791,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0010900295181649885,
      "loss": 1.6551984786987304,
      "step": 89640,
      "train_bpb": 0.5823997344186411,
      "train_bytes": 28263.95,
      "train_loss_nats_per_token": 1.655777444928985
    },
    {
      "epoch": 0.10968897316373022,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0010894296535835307,
      "loss": 1.6659515380859375,
      "step": 89660,
      "train_bpb": 0.5883591523993668,
      "train_bytes": 28258.06875,
      "train_loss_nats_per_token": 1.6669790926816492
    },
    {
      "epoch": 0.10982065260330252,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0010888298599729916,
      "loss": 1.6679025650024415,
      "step": 89680,
      "train_bpb": 0.5913124794354866,
      "train_bytes": 28326.06875,
      "train_loss_nats_per_token": 1.6685473468011904
    },
    {
      "epoch": 0.10995233204287483,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0010882301374370512,
      "loss": 1.6624134063720704,
      "step": 89700,
      "train_bpb": 0.5867804272698629,
      "train_bytes": 28044.95625,
      "train_loss_nats_per_token": 1.663382046804579
    },
    {
      "epoch": 0.11008401148244713,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0010876304860793782,
      "loss": 1.6636173248291015,
      "step": 89720,
      "train_bpb": 0.5865176289696162,
      "train_bytes": 28495.31875,
      "train_loss_nats_per_token": 1.665292138199359
    },
    {
      "epoch": 0.11021569092201944,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0010870309060036266,
      "loss": 1.6581533432006836,
      "step": 89740,
      "train_bpb": 0.5899447074233422,
      "train_bytes": 27995.58125,
      "train_loss_nats_per_token": 1.659114586989539
    },
    {
      "epoch": 0.11034737036159174,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0010864313973134399,
      "loss": 1.6521268844604493,
      "step": 89760,
      "train_bpb": 0.5870963246529518,
      "train_bytes": 28134.3625,
      "train_loss_nats_per_token": 1.6535987348190244
    },
    {
      "epoch": 0.11047904980116405,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.001085831960112449,
      "loss": 1.6508611679077148,
      "step": 89780,
      "train_bpb": 0.5832598321039333,
      "train_bytes": 28416.63125,
      "train_loss_nats_per_token": 1.652034793147826
    },
    {
      "epoch": 0.11061072924073635,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0010852325945042724,
      "loss": 1.6609691619873046,
      "step": 89800,
      "train_bpb": 0.5888162883029695,
      "train_bytes": 28319.9,
      "train_loss_nats_per_token": 1.6611000575880157
    },
    {
      "epoch": 0.11074240868030866,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.0010846333005925153,
      "loss": 1.6481319427490235,
      "step": 89820,
      "train_bpb": 0.5852126297909334,
      "train_bytes": 28297.85,
      "train_loss_nats_per_token": 1.6488960332205918
    },
    {
      "epoch": 0.11087408811988096,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0010840340784807716,
      "loss": 1.6731859207153321,
      "step": 89840,
      "train_bpb": 0.5909579104152864,
      "train_bytes": 28261.525,
      "train_loss_nats_per_token": 1.6734016213383864
    },
    {
      "epoch": 0.11100576755945327,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0010834349282726236,
      "loss": 1.651585578918457,
      "step": 89860,
      "train_bpb": 0.5858642188139039,
      "train_bytes": 28104.71875,
      "train_loss_nats_per_token": 1.6513606603511302
    },
    {
      "epoch": 0.11113744699902557,
      "grad_norm": 0.19140625,
      "learning_rate": 0.0010828358500716377,
      "loss": 1.6715526580810547,
      "step": 89880,
      "train_bpb": 0.5880612212996698,
      "train_bytes": 28303.9125,
      "train_loss_nats_per_token": 1.6722516847991575
    },
    {
      "epoch": 0.11126912643859788,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0010822368439813716,
      "loss": 1.6533075332641602,
      "step": 89900,
      "train_bpb": 0.5849700279663773,
      "train_bytes": 28174.3875,
      "train_loss_nats_per_token": 1.6544371344463733
    },
    {
      "epoch": 0.11140080587817018,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0010816379101053694,
      "loss": 1.6472387313842773,
      "step": 89920,
      "train_bpb": 0.5847842147060774,
      "train_bytes": 28303.18125,
      "train_loss_nats_per_token": 1.6477804026258294
    },
    {
      "epoch": 0.11153248531774249,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0010810390485471615,
      "loss": 1.652421760559082,
      "step": 89940,
      "train_bpb": 0.5878813335743586,
      "train_bytes": 28277.93125,
      "train_loss_nats_per_token": 1.6526543420098738
    },
    {
      "epoch": 0.11166416475731479,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.001080440259410267,
      "loss": 1.657526397705078,
      "step": 89960,
      "train_bpb": 0.5886703396526802,
      "train_bytes": 28294.2,
      "train_loss_nats_per_token": 1.6594122958115094
    },
    {
      "epoch": 0.1117958441968871,
      "grad_norm": 0.2001953125,
      "learning_rate": 0.0010798415427981922,
      "loss": 1.6534955978393555,
      "step": 89980,
      "train_bpb": 0.5850407828183196,
      "train_bytes": 28190.025,
      "train_loss_nats_per_token": 1.65323617101865
    },
    {
      "epoch": 0.1119275236364594,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0010792428988144317,
      "loss": 1.6621776580810548,
      "step": 90000,
      "train_bpb": 0.5891929993417674,
      "train_bytes": 28351.91875,
      "train_loss_nats_per_token": 1.6626889850795932
    },
    {
      "epoch": 0.1119275236364594,
      "eval_loss": 1.5228278636932373,
      "eval_runtime": 72.6025,
      "eval_samples_per_second": 13.774,
      "eval_steps_per_second": 13.774,
      "step": 90000,
      "train_bpb": 0.5374662917020719,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5304958493525993
    },
    {
      "epoch": 0.1120592030760317,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0010786443275624653,
      "loss": 1.6574762344360352,
      "step": 90020,
      "train_bpb": 0.587173127265416,
      "train_bytes": 28236.9,
      "train_loss_nats_per_token": 1.6583724415854477
    },
    {
      "epoch": 0.11219088251560401,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0010780458291457626,
      "loss": 1.661838912963867,
      "step": 90040,
      "train_bpb": 0.5844756269747167,
      "train_bytes": 28691.4875,
      "train_loss_nats_per_token": 1.6622923639970442
    },
    {
      "epoch": 0.11232256195517631,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0010774474036677798,
      "loss": 1.6563871383666993,
      "step": 90060,
      "train_bpb": 0.5857611804900611,
      "train_bytes": 28462.99375,
      "train_loss_nats_per_token": 1.6578051689190325
    },
    {
      "epoch": 0.11245424139474862,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.001076849051231959,
      "loss": 1.649925994873047,
      "step": 90080,
      "train_bpb": 0.5814873891771153,
      "train_bytes": 28356.8625,
      "train_loss_nats_per_token": 1.6502750904585077
    },
    {
      "epoch": 0.11258592083432092,
      "grad_norm": 0.146484375,
      "learning_rate": 0.001076250771941732,
      "loss": 1.6581758499145507,
      "step": 90100,
      "train_bpb": 0.5868758633738443,
      "train_bytes": 28288.55,
      "train_loss_nats_per_token": 1.6585842546807987
    },
    {
      "epoch": 0.11271760027389323,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0010756525659005168,
      "loss": 1.6517101287841798,
      "step": 90120,
      "train_bpb": 0.5853603305814605,
      "train_bytes": 28359.2875,
      "train_loss_nats_per_token": 1.6532210418144593
    },
    {
      "epoch": 0.11284927971346553,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.00107505443321172,
      "loss": 1.6518478393554688,
      "step": 90140,
      "train_bpb": 0.5861648525279737,
      "train_bytes": 28475.43125,
      "train_loss_nats_per_token": 1.6527598331809923
    },
    {
      "epoch": 0.11298095915303784,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0010744563739787322,
      "loss": 1.6658706665039062,
      "step": 90160,
      "train_bpb": 0.587154702188427,
      "train_bytes": 28736.64375,
      "train_loss_nats_per_token": 1.667016672465329
    },
    {
      "epoch": 0.11311263859261014,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0010738583883049349,
      "loss": 1.6608993530273437,
      "step": 90180,
      "train_bpb": 0.5877200032195374,
      "train_bytes": 28268.675,
      "train_loss_nats_per_token": 1.66151367807448
    },
    {
      "epoch": 0.11324431803218246,
      "grad_norm": 0.140625,
      "learning_rate": 0.001073260476293696,
      "loss": 1.6506391525268556,
      "step": 90200,
      "train_bpb": 0.5825129893794291,
      "train_bytes": 28395.8875,
      "train_loss_nats_per_token": 1.6508663943302366
    },
    {
      "epoch": 0.11337599747175477,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.001072662638048369,
      "loss": 1.6690750122070312,
      "step": 90220,
      "train_bpb": 0.5831621598927319,
      "train_bytes": 29032.28125,
      "train_loss_nats_per_token": 1.6701168780967426
    },
    {
      "epoch": 0.11350767691132707,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.001072064873672296,
      "loss": 1.6388826370239258,
      "step": 90240,
      "train_bpb": 0.583194261139292,
      "train_bytes": 28016.43125,
      "train_loss_nats_per_token": 1.6393351637489708
    },
    {
      "epoch": 0.11363935635089938,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.001071467183268807,
      "loss": 1.6682119369506836,
      "step": 90260,
      "train_bpb": 0.5883330032433876,
      "train_bytes": 28436.41875,
      "train_loss_nats_per_token": 1.6690420541815543
    },
    {
      "epoch": 0.11377103579047168,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0010708695669412174,
      "loss": 1.6473520278930665,
      "step": 90280,
      "train_bpb": 0.5848838124847613,
      "train_bytes": 28013.05,
      "train_loss_nats_per_token": 1.6482999193166614
    },
    {
      "epoch": 0.11390271523004399,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0010702720247928313,
      "loss": 1.6670915603637695,
      "step": 90300,
      "train_bpb": 0.5863917931150495,
      "train_bytes": 28647.525,
      "train_loss_nats_per_token": 1.6673908729814675
    },
    {
      "epoch": 0.11403439466961629,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.001069674556926939,
      "loss": 1.6557035446166992,
      "step": 90320,
      "train_bpb": 0.5865451673804563,
      "train_bytes": 28289.20625,
      "train_loss_nats_per_token": 1.6555882624166116
    },
    {
      "epoch": 0.1141660741091886,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0010690771634468198,
      "loss": 1.6614673614501954,
      "step": 90340,
      "train_bpb": 0.5883901568664452,
      "train_bytes": 28502.675,
      "train_loss_nats_per_token": 1.662600367209226
    },
    {
      "epoch": 0.1142977535487609,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0010684798444557367,
      "loss": 1.651504135131836,
      "step": 90360,
      "train_bpb": 0.5865137687208294,
      "train_bytes": 28312.55,
      "train_loss_nats_per_token": 1.6522902139573032
    },
    {
      "epoch": 0.1144294329883332,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.001067882600056943,
      "loss": 1.6665563583374023,
      "step": 90380,
      "train_bpb": 0.5892318240485168,
      "train_bytes": 28402.25,
      "train_loss_nats_per_token": 1.6668978414143232
    },
    {
      "epoch": 0.11456111242790551,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0010672854303536783,
      "loss": 1.6474483489990235,
      "step": 90400,
      "train_bpb": 0.5832452001567723,
      "train_bytes": 28006.85,
      "train_loss_nats_per_token": 1.6481592677961712
    },
    {
      "epoch": 0.11469279186747781,
      "grad_norm": 0.140625,
      "learning_rate": 0.001066688335449168,
      "loss": 1.6552967071533202,
      "step": 90420,
      "train_bpb": 0.5888413813585475,
      "train_bytes": 27743.0875,
      "train_loss_nats_per_token": 1.6567869484948672
    },
    {
      "epoch": 0.11482447130705012,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.001066091315446626,
      "loss": 1.6467727661132812,
      "step": 90440,
      "train_bpb": 0.5811149010225511,
      "train_bytes": 28399.8125,
      "train_loss_nats_per_token": 1.6465746417254306
    },
    {
      "epoch": 0.11495615074662242,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0010654943704492533,
      "loss": 1.648568344116211,
      "step": 90460,
      "train_bpb": 0.582564257825266,
      "train_bytes": 28429.175,
      "train_loss_nats_per_token": 1.6489375702271305
    },
    {
      "epoch": 0.11508783018619473,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0010648975005602365,
      "loss": 1.6417482376098633,
      "step": 90480,
      "train_bpb": 0.5818995236571685,
      "train_bytes": 28401.2125,
      "train_loss_nats_per_token": 1.643106876942397
    },
    {
      "epoch": 0.11521950962576703,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0010643007058827503,
      "loss": 1.6518083572387696,
      "step": 90500,
      "train_bpb": 0.5857076441393908,
      "train_bytes": 28029.14375,
      "train_loss_nats_per_token": 1.6524018078058846
    },
    {
      "epoch": 0.11535118906533934,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0010637039865199567,
      "loss": 1.6452848434448242,
      "step": 90520,
      "train_bpb": 0.5837556807308801,
      "train_bytes": 28337.04375,
      "train_loss_nats_per_token": 1.6460153692549935
    },
    {
      "epoch": 0.11548286850491164,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0010631073425750044,
      "loss": 1.646645736694336,
      "step": 90540,
      "train_bpb": 0.5849079471840434,
      "train_bytes": 28314.74375,
      "train_loss_nats_per_token": 1.647257092567346
    },
    {
      "epoch": 0.11561454794448395,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.001062510774151028,
      "loss": 1.6453697204589843,
      "step": 90560,
      "train_bpb": 0.5826291206878806,
      "train_bytes": 28426.9875,
      "train_loss_nats_per_token": 1.6460653918901056
    },
    {
      "epoch": 0.11574622738405625,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0010619142813511502,
      "loss": 1.6562952041625976,
      "step": 90580,
      "train_bpb": 0.5831069287103126,
      "train_bytes": 28589.625,
      "train_loss_nats_per_token": 1.6570098455091458
    },
    {
      "epoch": 0.11587790682362856,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0010613178642784815,
      "loss": 1.6438983917236327,
      "step": 90600,
      "train_bpb": 0.5845642976351509,
      "train_bytes": 28020.2625,
      "train_loss_nats_per_token": 1.643583468969861
    },
    {
      "epoch": 0.11600958626320086,
      "grad_norm": 0.109375,
      "learning_rate": 0.001060721523036116,
      "loss": 1.6561412811279297,
      "step": 90620,
      "train_bpb": 0.5852359683786,
      "train_bytes": 28341.4125,
      "train_loss_nats_per_token": 1.6562988921385908
    },
    {
      "epoch": 0.11614126570277317,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0010601252577271385,
      "loss": 1.6522760391235352,
      "step": 90640,
      "train_bpb": 0.5854504511882686,
      "train_bytes": 28436.2875,
      "train_loss_nats_per_token": 1.653405425388383
    },
    {
      "epoch": 0.11627294514234547,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0010595290684546185,
      "loss": 1.6517108917236327,
      "step": 90660,
      "train_bpb": 0.5860690562776982,
      "train_bytes": 28154.06875,
      "train_loss_nats_per_token": 1.6525552302440272
    },
    {
      "epoch": 0.11640462458191778,
      "grad_norm": 0.22265625,
      "learning_rate": 0.001058932955321614,
      "loss": 1.654214859008789,
      "step": 90680,
      "train_bpb": 0.5855123945109753,
      "train_bytes": 28220.7875,
      "train_loss_nats_per_token": 1.6556755543251431
    },
    {
      "epoch": 0.11653630402149008,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0010583369184311666,
      "loss": 1.6468927383422851,
      "step": 90700,
      "train_bpb": 0.5846098532105093,
      "train_bytes": 28002.29375,
      "train_loss_nats_per_token": 1.647700866278732
    },
    {
      "epoch": 0.11666798346106239,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.001057740957886308,
      "loss": 1.6476205825805663,
      "step": 90720,
      "train_bpb": 0.5847652243109156,
      "train_bytes": 28246.125,
      "train_loss_nats_per_token": 1.6481676975425823
    },
    {
      "epoch": 0.11679966290063469,
      "grad_norm": 0.130859375,
      "learning_rate": 0.001057145073790056,
      "loss": 1.6521656036376953,
      "step": 90740,
      "train_bpb": 0.5841784068042842,
      "train_bytes": 28218.625,
      "train_loss_nats_per_token": 1.6533781744180842
    },
    {
      "epoch": 0.116931342340207,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.001056549266245414,
      "loss": 1.646739387512207,
      "step": 90760,
      "train_bpb": 0.5858444548506669,
      "train_bytes": 28179.1625,
      "train_loss_nats_per_token": 1.6476938236373897
    },
    {
      "epoch": 0.1170630217797793,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0010559535353553728,
      "loss": 1.6551105499267578,
      "step": 90780,
      "train_bpb": 0.5841773811815894,
      "train_bytes": 28158.95,
      "train_loss_nats_per_token": 1.6556990621854741
    },
    {
      "epoch": 0.1171947012193516,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.001055357881222911,
      "loss": 1.6580238342285156,
      "step": 90800,
      "train_bpb": 0.587600424775524,
      "train_bytes": 28313.5125,
      "train_loss_nats_per_token": 1.6583730798776515
    },
    {
      "epoch": 0.11732638065892391,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0010547623039509919,
      "loss": 1.6495582580566406,
      "step": 90820,
      "train_bpb": 0.5838079592505561,
      "train_bytes": 28444.29375,
      "train_loss_nats_per_token": 1.65025362881951
    },
    {
      "epoch": 0.11745806009849621,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0010541668036425668,
      "loss": 1.6473817825317383,
      "step": 90840,
      "train_bpb": 0.5838823640969559,
      "train_bytes": 28220.075,
      "train_loss_nats_per_token": 1.6481595424831903
    },
    {
      "epoch": 0.11758973953806853,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0010535713804005738,
      "loss": 1.6497270584106445,
      "step": 90860,
      "train_bpb": 0.5846277290018957,
      "train_bytes": 28294.2875,
      "train_loss_nats_per_token": 1.6513965988398522
    },
    {
      "epoch": 0.11772141897764084,
      "grad_norm": 0.21875,
      "learning_rate": 0.001052976034327938,
      "loss": 1.6420907974243164,
      "step": 90880,
      "train_bpb": 0.5878883231441615,
      "train_bytes": 27916.78125,
      "train_loss_nats_per_token": 1.6421137099131982
    },
    {
      "epoch": 0.11785309841721314,
      "grad_norm": 0.15625,
      "learning_rate": 0.0010523807655275692,
      "loss": 1.661549186706543,
      "step": 90900,
      "train_bpb": 0.5870892740955699,
      "train_bytes": 28428.4375,
      "train_loss_nats_per_token": 1.6617009465301362
    },
    {
      "epoch": 0.11798477785678545,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0010517855741023656,
      "loss": 1.648881721496582,
      "step": 90920,
      "train_bpb": 0.5855012267133245,
      "train_bytes": 27993.16875,
      "train_loss_nats_per_token": 1.6493655639223543
    },
    {
      "epoch": 0.11811645729635775,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0010511904601552128,
      "loss": 1.6557029724121093,
      "step": 90940,
      "train_bpb": 0.5849738693838089,
      "train_bytes": 28334.2125,
      "train_loss_nats_per_token": 1.6563012750681865
    },
    {
      "epoch": 0.11824813673593006,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0010505954237889797,
      "loss": 1.6528173446655274,
      "step": 90960,
      "train_bpb": 0.5851157999579718,
      "train_bytes": 28458.88125,
      "train_loss_nats_per_token": 1.6536992164518582
    },
    {
      "epoch": 0.11837981617550236,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.0010500004651065255,
      "loss": 1.648295783996582,
      "step": 90980,
      "train_bpb": 0.580884207894658,
      "train_bytes": 28455.61875,
      "train_loss_nats_per_token": 1.6491166760576232
    },
    {
      "epoch": 0.11851149561507467,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0010494055842106935,
      "loss": 1.6522750854492188,
      "step": 91000,
      "train_bpb": 0.5870190021912879,
      "train_bytes": 28272.33125,
      "train_loss_nats_per_token": 1.6529348719370043
    },
    {
      "epoch": 0.11851149561507467,
      "eval_loss": 1.521862506866455,
      "eval_runtime": 82.5642,
      "eval_samples_per_second": 12.112,
      "eval_steps_per_second": 12.112,
      "step": 91000,
      "train_bpb": 0.5371164472088245,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.52949962772326
    },
    {
      "epoch": 0.11864317505464697,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0010488107812043155,
      "loss": 1.645880126953125,
      "step": 91020,
      "train_bpb": 0.5865931099949696,
      "train_bytes": 28072.68125,
      "train_loss_nats_per_token": 1.6469520758013476
    },
    {
      "epoch": 0.11877485449421928,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0010482160561902075,
      "loss": 1.657891082763672,
      "step": 91040,
      "train_bpb": 0.5877305348637194,
      "train_bytes": 28149.35625,
      "train_loss_nats_per_token": 1.6577262449464165
    },
    {
      "epoch": 0.11890653393379158,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0010476214092711737,
      "loss": 1.6586149215698243,
      "step": 91060,
      "train_bpb": 0.5898520026586876,
      "train_bytes": 28011.08125,
      "train_loss_nats_per_token": 1.658839215003954
    },
    {
      "epoch": 0.11903821337336389,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0010470268405500052,
      "loss": 1.6572265625,
      "step": 91080,
      "train_bpb": 0.58626535702402,
      "train_bytes": 28215.7875,
      "train_loss_nats_per_token": 1.6569600996198464
    },
    {
      "epoch": 0.11916989281293619,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0010464323501294772,
      "loss": 1.6656496047973632,
      "step": 91100,
      "train_bpb": 0.5867760747829633,
      "train_bytes": 28587.70625,
      "train_loss_nats_per_token": 1.6655720324460586
    },
    {
      "epoch": 0.1193015722525085,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0010458379381123538,
      "loss": 1.65361328125,
      "step": 91120,
      "train_bpb": 0.5827205875354701,
      "train_bytes": 28411.95625,
      "train_loss_nats_per_token": 1.6541279372305273
    },
    {
      "epoch": 0.1194332516920808,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0010452436046013854,
      "loss": 1.641702651977539,
      "step": 91140,
      "train_bpb": 0.5840219151109761,
      "train_bytes": 27877.625,
      "train_loss_nats_per_token": 1.6429485879867793
    },
    {
      "epoch": 0.1195649311316531,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0010446493496993068,
      "loss": 1.654024887084961,
      "step": 91160,
      "train_bpb": 0.588612674290273,
      "train_bytes": 28026.64375,
      "train_loss_nats_per_token": 1.6539602027986482
    },
    {
      "epoch": 0.11969661057122541,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0010440551735088409,
      "loss": 1.6348705291748047,
      "step": 91180,
      "train_bpb": 0.578834436366552,
      "train_bytes": 28035.525,
      "train_loss_nats_per_token": 1.6358533095740326
    },
    {
      "epoch": 0.11982829001079771,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0010434610761326968,
      "loss": 1.6603376388549804,
      "step": 91200,
      "train_bpb": 0.5872686661589783,
      "train_bytes": 28447.76875,
      "train_loss_nats_per_token": 1.661802374249847
    },
    {
      "epoch": 0.11995996945037002,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00104286705767357,
      "loss": 1.644371795654297,
      "step": 91220,
      "train_bpb": 0.5807616568402534,
      "train_bytes": 28411.73125,
      "train_loss_nats_per_token": 1.6449295545346045
    },
    {
      "epoch": 0.12009164888994232,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0010422731182341415,
      "loss": 1.6593709945678712,
      "step": 91240,
      "train_bpb": 0.5853407681355968,
      "train_bytes": 28521.1375,
      "train_loss_nats_per_token": 1.6605190213101992
    },
    {
      "epoch": 0.12022332832951463,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0010416792579170796,
      "loss": 1.640230369567871,
      "step": 91260,
      "train_bpb": 0.5815402245638428,
      "train_bytes": 28157.75625,
      "train_loss_nats_per_token": 1.6414897186412227
    },
    {
      "epoch": 0.12035500776908693,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0010410854768250395,
      "loss": 1.6439857482910156,
      "step": 91280,
      "train_bpb": 0.5804004653573588,
      "train_bytes": 28336.34375,
      "train_loss_nats_per_token": 1.6444180574023475
    },
    {
      "epoch": 0.12048668720865924,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0010404917750606601,
      "loss": 1.6515663146972657,
      "step": 91300,
      "train_bpb": 0.585642816059866,
      "train_bytes": 28131.6125,
      "train_loss_nats_per_token": 1.6513694178685485
    },
    {
      "epoch": 0.12061836664823154,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0010398981527265693,
      "loss": 1.6533763885498047,
      "step": 91320,
      "train_bpb": 0.5891904748187162,
      "train_bytes": 27902.76875,
      "train_loss_nats_per_token": 1.6542436676894396
    },
    {
      "epoch": 0.12075004608780385,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0010393046099253806,
      "loss": 1.6411430358886718,
      "step": 91340,
      "train_bpb": 0.5828059233510019,
      "train_bytes": 27656.89375,
      "train_loss_nats_per_token": 1.6421395364031872
    },
    {
      "epoch": 0.12088172552737615,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0010387111467596922,
      "loss": 1.6356903076171876,
      "step": 91360,
      "train_bpb": 0.5814878057874309,
      "train_bytes": 28220.8625,
      "train_loss_nats_per_token": 1.635621963785558
    },
    {
      "epoch": 0.12101340496694846,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.0010381177633320905,
      "loss": 1.6414079666137695,
      "step": 91380,
      "train_bpb": 0.5815364266562216,
      "train_bytes": 28260.1,
      "train_loss_nats_per_token": 1.6418814013897085
    },
    {
      "epoch": 0.12114508440652076,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0010375244597451475,
      "loss": 1.6544837951660156,
      "step": 91400,
      "train_bpb": 0.5860388760680877,
      "train_bytes": 27971.55625,
      "train_loss_nats_per_token": 1.6549937048132703
    },
    {
      "epoch": 0.12127676384609307,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0010369312361014215,
      "loss": 1.6572708129882812,
      "step": 91420,
      "train_bpb": 0.5878155092044447,
      "train_bytes": 28161.05,
      "train_loss_nats_per_token": 1.6583160613093941
    },
    {
      "epoch": 0.12140844328566537,
      "grad_norm": 0.1640625,
      "learning_rate": 0.001036338092503456,
      "loss": 1.6482627868652344,
      "step": 91440,
      "train_bpb": 0.5856679899892359,
      "train_bytes": 28191.26875,
      "train_loss_nats_per_token": 1.6489454458899462
    },
    {
      "epoch": 0.12154012272523768,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0010357450290537815,
      "loss": 1.6509916305541992,
      "step": 91460,
      "train_bpb": 0.5820183489221818,
      "train_bytes": 28476.675,
      "train_loss_nats_per_token": 1.651077207089373
    },
    {
      "epoch": 0.12167180216480998,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0010351520458549157,
      "loss": 1.6348960876464844,
      "step": 91480,
      "train_bpb": 0.5854197217189405,
      "train_bytes": 27719.5625,
      "train_loss_nats_per_token": 1.6358820918233667
    },
    {
      "epoch": 0.12180348160438229,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0010345591430093598,
      "loss": 1.6566421508789062,
      "step": 91500,
      "train_bpb": 0.5827018831681295,
      "train_bytes": 28637.5125,
      "train_loss_nats_per_token": 1.657216935297194
    },
    {
      "epoch": 0.12193516104395459,
      "grad_norm": 0.1484375,
      "learning_rate": 0.001033966320619603,
      "loss": 1.6411352157592773,
      "step": 91520,
      "train_bpb": 0.5829075956289864,
      "train_bytes": 28192.6,
      "train_loss_nats_per_token": 1.641638391492971
    },
    {
      "epoch": 0.12206684048352691,
      "grad_norm": 0.15625,
      "learning_rate": 0.0010333735787881205,
      "loss": 1.655182456970215,
      "step": 91540,
      "train_bpb": 0.586763296526593,
      "train_bytes": 28406.60625,
      "train_loss_nats_per_token": 1.6559395391264247
    },
    {
      "epoch": 0.12219851992309921,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0010327809176173739,
      "loss": 1.6573646545410157,
      "step": 91560,
      "train_bpb": 0.5864108845467635,
      "train_bytes": 28387.49375,
      "train_loss_nats_per_token": 1.6593805749501942
    },
    {
      "epoch": 0.12233019936267152,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0010321883372098085,
      "loss": 1.6444171905517577,
      "step": 91580,
      "train_bpb": 0.5833297249436047,
      "train_bytes": 28176.85625,
      "train_loss_nats_per_token": 1.6448551975934746
    },
    {
      "epoch": 0.12246187880224382,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0010315958376678586,
      "loss": 1.6493080139160157,
      "step": 91600,
      "train_bpb": 0.5838307087941732,
      "train_bytes": 28298.63125,
      "train_loss_nats_per_token": 1.6500296953908171
    },
    {
      "epoch": 0.12259355824181613,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0010310034190939437,
      "loss": 1.6472599029541015,
      "step": 91620,
      "train_bpb": 0.5861741881462754,
      "train_bytes": 27971.40625,
      "train_loss_nats_per_token": 1.6479872568408505
    },
    {
      "epoch": 0.12272523768138843,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0010304110815904674,
      "loss": 1.653843879699707,
      "step": 91640,
      "train_bpb": 0.5884185947860047,
      "train_bytes": 28018.91875,
      "train_loss_nats_per_token": 1.6547826961711016
    },
    {
      "epoch": 0.12285691712096074,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0010298188252598215,
      "loss": 1.6434820175170899,
      "step": 91660,
      "train_bpb": 0.5830293036439663,
      "train_bytes": 27948.45,
      "train_loss_nats_per_token": 1.6438639430848272
    },
    {
      "epoch": 0.12298859656053304,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0010292266502043839,
      "loss": 1.6417724609375,
      "step": 91680,
      "train_bpb": 0.5785073645878682,
      "train_bytes": 28480.8125,
      "train_loss_nats_per_token": 1.6420311734420339
    },
    {
      "epoch": 0.12312027600010535,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.001028634556526516,
      "loss": 1.6439151763916016,
      "step": 91700,
      "train_bpb": 0.5826697899092181,
      "train_bytes": 28306.63125,
      "train_loss_nats_per_token": 1.644208517678234
    },
    {
      "epoch": 0.12325195543967765,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0010280425443285677,
      "loss": 1.6410257339477539,
      "step": 91720,
      "train_bpb": 0.581948936964297,
      "train_bytes": 28272.74375,
      "train_loss_nats_per_token": 1.6426506363770474
    },
    {
      "epoch": 0.12338363487924996,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0010274506137128736,
      "loss": 1.6514909744262696,
      "step": 91740,
      "train_bpb": 0.5805428179917033,
      "train_bytes": 28348.14375,
      "train_loss_nats_per_token": 1.6520568394804913
    },
    {
      "epoch": 0.12351531431882226,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.001026858764781755,
      "loss": 1.636714553833008,
      "step": 91760,
      "train_bpb": 0.5801250432235538,
      "train_bytes": 28190.8125,
      "train_loss_nats_per_token": 1.637936995471826
    },
    {
      "epoch": 0.12364699375839457,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0010262669976375174,
      "loss": 1.624227523803711,
      "step": 91780,
      "train_bpb": 0.5772802686043076,
      "train_bytes": 28156.44375,
      "train_loss_nats_per_token": 1.624483597757378
    },
    {
      "epoch": 0.12377867319796687,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0010256753123824539,
      "loss": 1.6495384216308593,
      "step": 91800,
      "train_bpb": 0.5809515116614027,
      "train_bytes": 28332.19375,
      "train_loss_nats_per_token": 1.6501343911344433
    },
    {
      "epoch": 0.12391035263753918,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0010250837091188433,
      "loss": 1.6442987442016601,
      "step": 91820,
      "train_bpb": 0.5831027483929362,
      "train_bytes": 28104.1375,
      "train_loss_nats_per_token": 1.6443400154342664
    },
    {
      "epoch": 0.12404203207711148,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0010244921879489488,
      "loss": 1.626523780822754,
      "step": 91840,
      "train_bpb": 0.5800672940741292,
      "train_bytes": 27984.94375,
      "train_loss_nats_per_token": 1.6278189229354831
    },
    {
      "epoch": 0.12417371151668379,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0010239007489750208,
      "loss": 1.6402912139892578,
      "step": 91860,
      "train_bpb": 0.580718825075893,
      "train_bytes": 28301.83125,
      "train_loss_nats_per_token": 1.640903256900264
    },
    {
      "epoch": 0.12430539095625609,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0010233093922992957,
      "loss": 1.6546222686767578,
      "step": 91880,
      "train_bpb": 0.5835111313785493,
      "train_bytes": 28450.03125,
      "train_loss_nats_per_token": 1.6556742439215613
    },
    {
      "epoch": 0.1244370703958284,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0010227181180239937,
      "loss": 1.6489694595336915,
      "step": 91900,
      "train_bpb": 0.585746816364184,
      "train_bytes": 28326.8875,
      "train_loss_nats_per_token": 1.6494054921157857
    },
    {
      "epoch": 0.1245687498354007,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0010221269262513228,
      "loss": 1.6461067199707031,
      "step": 91920,
      "train_bpb": 0.581553831941518,
      "train_bytes": 28438.15,
      "train_loss_nats_per_token": 1.6464745566980568
    },
    {
      "epoch": 0.124700429274973,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0010215358170834758,
      "loss": 1.6406911849975585,
      "step": 91940,
      "train_bpb": 0.5828963671875778,
      "train_bytes": 28540.975,
      "train_loss_nats_per_token": 1.6416721380957275
    },
    {
      "epoch": 0.12483210871454531,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0010209447906226327,
      "loss": 1.6498516082763672,
      "step": 91960,
      "train_bpb": 0.5865418598013566,
      "train_bytes": 27965.375,
      "train_loss_nats_per_token": 1.651081995732319
    },
    {
      "epoch": 0.12496378815411761,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.001020353846970956,
      "loss": 1.6619569778442382,
      "step": 91980,
      "train_bpb": 0.5894401308568842,
      "train_bytes": 28145.09375,
      "train_loss_nats_per_token": 1.6621673540754596
    },
    {
      "epoch": 0.12509546759368992,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0010197629862305966,
      "loss": 1.6550022125244142,
      "step": 92000,
      "train_bpb": 0.5867121922159895,
      "train_bytes": 28104.95,
      "train_loss_nats_per_token": 1.65541414235722
    },
    {
      "epoch": 0.12509546759368992,
      "eval_loss": 1.522338628768921,
      "eval_runtime": 73.3994,
      "eval_samples_per_second": 13.624,
      "eval_steps_per_second": 13.624,
      "step": 92000,
      "train_bpb": 0.5372711531547546,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5299401703429023
    },
    {
      "epoch": 0.12522714703326224,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0010191722085036911,
      "loss": 1.6580673217773438,
      "step": 92020,
      "train_bpb": 0.5850030312026216,
      "train_bytes": 28328.48125,
      "train_loss_nats_per_token": 1.6593306144343476
    },
    {
      "epoch": 0.12535882647283453,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0010185815138923594,
      "loss": 1.6375835418701172,
      "step": 92040,
      "train_bpb": 0.5806068498686554,
      "train_bytes": 28294.4875,
      "train_loss_nats_per_token": 1.638596047911913
    },
    {
      "epoch": 0.12549050591240685,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0010179909024987097,
      "loss": 1.6411016464233399,
      "step": 92060,
      "train_bpb": 0.5820415476979947,
      "train_bytes": 28063.05625,
      "train_loss_nats_per_token": 1.6425013773230634
    },
    {
      "epoch": 0.12562218535197914,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0010174003744248344,
      "loss": 1.6421289443969727,
      "step": 92080,
      "train_bpb": 0.5817530506654525,
      "train_bytes": 28563.45625,
      "train_loss_nats_per_token": 1.6429426701190124
    },
    {
      "epoch": 0.12575386479155146,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0010168099297728124,
      "loss": 1.654477882385254,
      "step": 92100,
      "train_bpb": 0.5873392657560141,
      "train_bytes": 28013.85625,
      "train_loss_nats_per_token": 1.655629378178846
    },
    {
      "epoch": 0.12588554423112375,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0010162195686447063,
      "loss": 1.6448280334472656,
      "step": 92120,
      "train_bpb": 0.5819698473088883,
      "train_bytes": 28245.45625,
      "train_loss_nats_per_token": 1.6457851913100412
    },
    {
      "epoch": 0.12601722367069607,
      "grad_norm": 0.1875,
      "learning_rate": 0.001015629291142566,
      "loss": 1.6584793090820313,
      "step": 92140,
      "train_bpb": 0.5896889831196455,
      "train_bytes": 28013.53125,
      "train_loss_nats_per_token": 1.6595985284594337
    },
    {
      "epoch": 0.12614890311026836,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.001015039097368428,
      "loss": 1.6620304107666015,
      "step": 92160,
      "train_bpb": 0.5886654084029571,
      "train_bytes": 28353.7625,
      "train_loss_nats_per_token": 1.662899086074153
    },
    {
      "epoch": 0.12628058254984068,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0010144489874243103,
      "loss": 1.6496299743652343,
      "step": 92180,
      "train_bpb": 0.5845815450872199,
      "train_bytes": 28347.63125,
      "train_loss_nats_per_token": 1.6503221829136225
    },
    {
      "epoch": 0.12641226198941297,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0010138589614122203,
      "loss": 1.6673938751220703,
      "step": 92200,
      "train_bpb": 0.5908894016444372,
      "train_bytes": 28213.53125,
      "train_loss_nats_per_token": 1.6682305414866971
    },
    {
      "epoch": 0.12654394142898528,
      "grad_norm": 0.166015625,
      "learning_rate": 0.00101326901943415,
      "loss": 1.639183235168457,
      "step": 92220,
      "train_bpb": 0.5827560665892347,
      "train_bytes": 28112.13125,
      "train_loss_nats_per_token": 1.6392528917918443
    },
    {
      "epoch": 0.12667562086855758,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0010126791615920753,
      "loss": 1.646695327758789,
      "step": 92240,
      "train_bpb": 0.5855711699642648,
      "train_bytes": 28057.175,
      "train_loss_nats_per_token": 1.6470424602741693
    },
    {
      "epoch": 0.1268073003081299,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0010120893879879587,
      "loss": 1.6496477127075195,
      "step": 92260,
      "train_bpb": 0.582351742179787,
      "train_bytes": 28686.85625,
      "train_loss_nats_per_token": 1.650911447136993
    },
    {
      "epoch": 0.12693897974770219,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0010114996987237485,
      "loss": 1.6464349746704101,
      "step": 92280,
      "train_bpb": 0.58224742481962,
      "train_bytes": 28506.16875,
      "train_loss_nats_per_token": 1.647724723864236
    },
    {
      "epoch": 0.1270706591872745,
      "grad_norm": 0.15625,
      "learning_rate": 0.0010109100939013785,
      "loss": 1.6493152618408202,
      "step": 92300,
      "train_bpb": 0.5855280080820345,
      "train_bytes": 28336.98125,
      "train_loss_nats_per_token": 1.6508032782288216
    },
    {
      "epoch": 0.1272023386268468,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0010103205736227664,
      "loss": 1.6397415161132813,
      "step": 92320,
      "train_bpb": 0.5823164747361175,
      "train_bytes": 28103.05,
      "train_loss_nats_per_token": 1.6412662345878821
    },
    {
      "epoch": 0.1273340180664191,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0010097311379898162,
      "loss": 1.6342203140258789,
      "step": 92340,
      "train_bpb": 0.5797116274374416,
      "train_bytes": 28387.50625,
      "train_loss_nats_per_token": 1.6347452634377144
    },
    {
      "epoch": 0.1274656975059914,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0010091417871044193,
      "loss": 1.6467384338378905,
      "step": 92360,
      "train_bpb": 0.5823511787093468,
      "train_bytes": 28404.50625,
      "train_loss_nats_per_token": 1.647355213669335
    },
    {
      "epoch": 0.12759737694556372,
      "grad_norm": 0.13671875,
      "learning_rate": 0.001008552521068448,
      "loss": 1.6459430694580077,
      "step": 92380,
      "train_bpb": 0.5814755739853568,
      "train_bytes": 28365.71875,
      "train_loss_nats_per_token": 1.6462229099161112
    },
    {
      "epoch": 0.12772905638513601,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0010079633399837633,
      "loss": 1.6440181732177734,
      "step": 92400,
      "train_bpb": 0.5793609554000012,
      "train_bytes": 28136.4,
      "train_loss_nats_per_token": 1.644601932882426
    },
    {
      "epoch": 0.12786073582470833,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0010073742439522118,
      "loss": 1.6473649978637694,
      "step": 92420,
      "train_bpb": 0.5816171374551621,
      "train_bytes": 28600.5125,
      "train_loss_nats_per_token": 1.6478541811234686
    },
    {
      "epoch": 0.12799241526428062,
      "grad_norm": 0.125,
      "learning_rate": 0.0010067852330756224,
      "loss": 1.6511117935180664,
      "step": 92440,
      "train_bpb": 0.5878780828290148,
      "train_bytes": 28058.35625,
      "train_loss_nats_per_token": 1.652631236494141
    },
    {
      "epoch": 0.12812409470385294,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0010061963074558123,
      "loss": 1.656130027770996,
      "step": 92460,
      "train_bpb": 0.5822628486147736,
      "train_bytes": 28344.2625,
      "train_loss_nats_per_token": 1.6563004156205432
    },
    {
      "epoch": 0.12825577414342523,
      "grad_norm": 0.19921875,
      "learning_rate": 0.0010056074671945824,
      "loss": 1.6581230163574219,
      "step": 92480,
      "train_bpb": 0.5857220978238068,
      "train_bytes": 28311.00625,
      "train_loss_nats_per_token": 1.6586771401873701
    },
    {
      "epoch": 0.12838745358299755,
      "grad_norm": 0.373046875,
      "learning_rate": 0.00100501871239372,
      "loss": 1.6487703323364258,
      "step": 92500,
      "train_bpb": 0.5850451124854932,
      "train_bytes": 28206.23125,
      "train_loss_nats_per_token": 1.6493373709991574
    },
    {
      "epoch": 0.12851913302256984,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0010044300431549958,
      "loss": 1.647343635559082,
      "step": 92520,
      "train_bpb": 0.5865196023492127,
      "train_bytes": 27931.1125,
      "train_loss_nats_per_token": 1.648520752665985
    },
    {
      "epoch": 0.12865081246214216,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.001003841459580167,
      "loss": 1.6419862747192382,
      "step": 92540,
      "train_bpb": 0.5794826690586964,
      "train_bytes": 28542.40625,
      "train_loss_nats_per_token": 1.6423175465511581
    },
    {
      "epoch": 0.12878249190171448,
      "grad_norm": 0.171875,
      "learning_rate": 0.0010032529617709765,
      "loss": 1.6520280838012695,
      "step": 92560,
      "train_bpb": 0.5857527933379937,
      "train_bytes": 28425.2125,
      "train_loss_nats_per_token": 1.6536224123537175
    },
    {
      "epoch": 0.12891417134128677,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0010026645498291509,
      "loss": 1.6473070144653321,
      "step": 92580,
      "train_bpb": 0.5841889617893415,
      "train_bytes": 28320.99375,
      "train_loss_nats_per_token": 1.6481347979576146
    },
    {
      "epoch": 0.1290458507808591,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0010020762238564026,
      "loss": 1.6555234909057617,
      "step": 92600,
      "train_bpb": 0.5853905834834779,
      "train_bytes": 28473.21875,
      "train_loss_nats_per_token": 1.6559158259010953
    },
    {
      "epoch": 0.12917753022043138,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0010014879839544298,
      "loss": 1.6432493209838868,
      "step": 92620,
      "train_bpb": 0.5810656120108406,
      "train_bytes": 28368.9125,
      "train_loss_nats_per_token": 1.6440623445544527
    },
    {
      "epoch": 0.1293092096600037,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0010008998302249155,
      "loss": 1.6284215927124024,
      "step": 92640,
      "train_bpb": 0.5768845424064629,
      "train_bytes": 28107.7375,
      "train_loss_nats_per_token": 1.6287697392851967
    },
    {
      "epoch": 0.129440889099576,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0010003117627695266,
      "loss": 1.6383855819702149,
      "step": 92660,
      "train_bpb": 0.5826340593301126,
      "train_bytes": 27999.64375,
      "train_loss_nats_per_token": 1.6391591949540998
    },
    {
      "epoch": 0.1295725685391483,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0009997237816899162,
      "loss": 1.637952995300293,
      "step": 92680,
      "train_bpb": 0.5823662193573275,
      "train_bytes": 27972.075,
      "train_loss_nats_per_token": 1.6393991615032328
    },
    {
      "epoch": 0.1297042479787206,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0009991358870877238,
      "loss": 1.638491439819336,
      "step": 92700,
      "train_bpb": 0.5812296527438975,
      "train_bytes": 28417.075,
      "train_loss_nats_per_token": 1.639311859971331
    },
    {
      "epoch": 0.12983592741829292,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0009985480790645701,
      "loss": 1.6516521453857422,
      "step": 92720,
      "train_bpb": 0.5833148069976498,
      "train_bytes": 28118.0375,
      "train_loss_nats_per_token": 1.6524149348295354
    },
    {
      "epoch": 0.1299676068578652,
      "grad_norm": 0.1484375,
      "learning_rate": 0.000997960357722065,
      "loss": 1.6462972640991211,
      "step": 92740,
      "train_bpb": 0.5821874616949917,
      "train_bytes": 28743.11875,
      "train_loss_nats_per_token": 1.6460088975219027
    },
    {
      "epoch": 0.13009928629743753,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0009973727231618016,
      "loss": 1.6584039688110352,
      "step": 92760,
      "train_bpb": 0.5874727316897969,
      "train_bytes": 28487.5,
      "train_loss_nats_per_token": 1.6596983799000156
    },
    {
      "epoch": 0.13023096573700982,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0009967851754853565,
      "loss": 1.6425460815429687,
      "step": 92780,
      "train_bpb": 0.5815478792227785,
      "train_bytes": 28385.2125,
      "train_loss_nats_per_token": 1.6434872494239412
    },
    {
      "epoch": 0.13036264517658214,
      "grad_norm": 0.171875,
      "learning_rate": 0.0009961977147942942,
      "loss": 1.6402830123901366,
      "step": 92800,
      "train_bpb": 0.584378815863361,
      "train_bytes": 27757.10625,
      "train_loss_nats_per_token": 1.6404817321954228
    },
    {
      "epoch": 0.13049432461615443,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0009956103411901621,
      "loss": 1.6379425048828125,
      "step": 92820,
      "train_bpb": 0.5799695682229095,
      "train_bytes": 28019.85625,
      "train_loss_nats_per_token": 1.639179273386418
    },
    {
      "epoch": 0.13062600405572675,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0009950230547744944,
      "loss": 1.6393461227416992,
      "step": 92840,
      "train_bpb": 0.5786610190727342,
      "train_bytes": 28344.9125,
      "train_loss_nats_per_token": 1.6395317598748844
    },
    {
      "epoch": 0.13075768349529904,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0009944358556488071,
      "loss": 1.6446569442749024,
      "step": 92860,
      "train_bpb": 0.5834375266376292,
      "train_bytes": 28292.95,
      "train_loss_nats_per_token": 1.6455792267786662
    },
    {
      "epoch": 0.13088936293487136,
      "grad_norm": 0.2060546875,
      "learning_rate": 0.0009938487439146045,
      "loss": 1.6358503341674804,
      "step": 92880,
      "train_bpb": 0.5823934116195845,
      "train_bytes": 28244.81875,
      "train_loss_nats_per_token": 1.6364155601692552
    },
    {
      "epoch": 0.13102104237444365,
      "grad_norm": 0.2734375,
      "learning_rate": 0.0009932617196733743,
      "loss": 1.6528505325317382,
      "step": 92900,
      "train_bpb": 0.5865781790738925,
      "train_bytes": 28275.79375,
      "train_loss_nats_per_token": 1.6540095602635156
    },
    {
      "epoch": 0.13115272181401597,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.000992674783026588,
      "loss": 1.6361690521240235,
      "step": 92920,
      "train_bpb": 0.578369390011873,
      "train_bytes": 28226.6375,
      "train_loss_nats_per_token": 1.6365908461342598
    },
    {
      "epoch": 0.13128440125358826,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.000992087934075704,
      "loss": 1.638452911376953,
      "step": 92940,
      "train_bpb": 0.58164641690255,
      "train_bytes": 28157.26875,
      "train_loss_nats_per_token": 1.6386313894395546
    },
    {
      "epoch": 0.13141608069316058,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0009915011729221652,
      "loss": 1.6468685150146485,
      "step": 92960,
      "train_bpb": 0.5830232766323954,
      "train_bytes": 28062.775,
      "train_loss_nats_per_token": 1.6475962384124214
    },
    {
      "epoch": 0.13154776013273287,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0009909144996673972,
      "loss": 1.6452344894409179,
      "step": 92980,
      "train_bpb": 0.5810415241240457,
      "train_bytes": 28528.35,
      "train_loss_nats_per_token": 1.6462502851614411
    },
    {
      "epoch": 0.13167943957230518,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0009903279144128127,
      "loss": 1.635453224182129,
      "step": 93000,
      "train_bpb": 0.5784049805697793,
      "train_bytes": 28515.0,
      "train_loss_nats_per_token": 1.6357706925477409
    },
    {
      "epoch": 0.13167943957230518,
      "eval_loss": 1.5207418203353882,
      "eval_runtime": 80.0508,
      "eval_samples_per_second": 12.492,
      "eval_steps_per_second": 12.492,
      "step": 93000,
      "train_bpb": 0.5367355285909665,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5284149190213863
    },
    {
      "epoch": 0.13181111901187748,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0009897414172598084,
      "loss": 1.6389251708984376,
      "step": 93020,
      "train_bpb": 0.5806699239516309,
      "train_bytes": 28311.29375,
      "train_loss_nats_per_token": 1.6391091293686033
    },
    {
      "epoch": 0.1319427984514498,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0009891550083097665,
      "loss": 1.6327501296997071,
      "step": 93040,
      "train_bpb": 0.5802127121826929,
      "train_bytes": 27947.44375,
      "train_loss_nats_per_token": 1.6334312756720575
    },
    {
      "epoch": 0.13207447789102209,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0009885686876640522,
      "loss": 1.6320764541625976,
      "step": 93060,
      "train_bpb": 0.5767750046563411,
      "train_bytes": 28518.14375,
      "train_loss_nats_per_token": 1.6332220278737124
    },
    {
      "epoch": 0.1322061573305944,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.0009879824554240172,
      "loss": 1.6448060989379882,
      "step": 93080,
      "train_bpb": 0.5803251523490913,
      "train_bytes": 28448.75625,
      "train_loss_nats_per_token": 1.6452347805811334
    },
    {
      "epoch": 0.1323378367701667,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0009873963116909972,
      "loss": 1.6353975296020509,
      "step": 93100,
      "train_bpb": 0.5784559883481786,
      "train_bytes": 28518.9125,
      "train_loss_nats_per_token": 1.635643546622248
    },
    {
      "epoch": 0.132469516209739,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0009868102565663119,
      "loss": 1.635685348510742,
      "step": 93120,
      "train_bpb": 0.5792868328933861,
      "train_bytes": 28380.9875,
      "train_loss_nats_per_token": 1.6368228727280694
    },
    {
      "epoch": 0.1326011956493113,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0009862242901512667,
      "loss": 1.6253679275512696,
      "step": 93140,
      "train_bpb": 0.5769854645942761,
      "train_bytes": 28206.5,
      "train_loss_nats_per_token": 1.6271490551713332
    },
    {
      "epoch": 0.13273287508888362,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0009856384125471519,
      "loss": 1.640757942199707,
      "step": 93160,
      "train_bpb": 0.582381778913299,
      "train_bytes": 28191.725,
      "train_loss_nats_per_token": 1.6410644837538055
    },
    {
      "epoch": 0.1328645545284559,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.000985052623855242,
      "loss": 1.6425949096679688,
      "step": 93180,
      "train_bpb": 0.5824267749970886,
      "train_bytes": 27982.99375,
      "train_loss_nats_per_token": 1.643290423597291
    },
    {
      "epoch": 0.13299623396802823,
      "grad_norm": 0.17578125,
      "learning_rate": 0.000984466924176795,
      "loss": 1.6372903823852538,
      "step": 93200,
      "train_bpb": 0.5797969313757145,
      "train_bytes": 28349.44375,
      "train_loss_nats_per_token": 1.6378548694412585
    },
    {
      "epoch": 0.13312791340760055,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0009838813136130553,
      "loss": 1.6193014144897462,
      "step": 93220,
      "train_bpb": 0.5726662391871643,
      "train_bytes": 27928.5125,
      "train_loss_nats_per_token": 1.6207853088652464
    },
    {
      "epoch": 0.13325959284717284,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0009832957922652516,
      "loss": 1.648630142211914,
      "step": 93240,
      "train_bpb": 0.5839657894336068,
      "train_bytes": 28135.1625,
      "train_loss_nats_per_token": 1.6500323235451622
    },
    {
      "epoch": 0.13339127228674516,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0009827103602345955,
      "loss": 1.6273887634277344,
      "step": 93260,
      "train_bpb": 0.5781067827595611,
      "train_bytes": 28201.29375,
      "train_loss_nats_per_token": 1.6292113577505956
    },
    {
      "epoch": 0.13352295172631745,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0009821250176222853,
      "loss": 1.6445343017578125,
      "step": 93280,
      "train_bpb": 0.5828774592794409,
      "train_bytes": 28381.13125,
      "train_loss_nats_per_token": 1.645136668590566
    },
    {
      "epoch": 0.13365463116588977,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.000981539764529503,
      "loss": 1.6506120681762695,
      "step": 93300,
      "train_bpb": 0.5834675841563435,
      "train_bytes": 28300.0375,
      "train_loss_nats_per_token": 1.6509899008519302
    },
    {
      "epoch": 0.13378631060546206,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0009809546010574144,
      "loss": 1.6416360855102539,
      "step": 93320,
      "train_bpb": 0.5787879460790857,
      "train_bytes": 28697.0125,
      "train_loss_nats_per_token": 1.643511363507727
    },
    {
      "epoch": 0.13391799004503438,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0009803695273071706,
      "loss": 1.63317813873291,
      "step": 93340,
      "train_bpb": 0.5757509868615652,
      "train_bytes": 28444.45,
      "train_loss_nats_per_token": 1.6340725488169843
    },
    {
      "epoch": 0.13404966948460667,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0009797845433799075,
      "loss": 1.6400566101074219,
      "step": 93360,
      "train_bpb": 0.578785754834215,
      "train_bytes": 28388.375,
      "train_loss_nats_per_token": 1.640861599207988
    },
    {
      "epoch": 0.134181348924179,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0009791996493767453,
      "loss": 1.6507171630859374,
      "step": 93380,
      "train_bpb": 0.5834919362065605,
      "train_bytes": 28507.63125,
      "train_loss_nats_per_token": 1.6518873143294504
    },
    {
      "epoch": 0.13431302836375128,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0009786148453987875,
      "loss": 1.6291337966918946,
      "step": 93400,
      "train_bpb": 0.5746167399537062,
      "train_bytes": 28277.6875,
      "train_loss_nats_per_token": 1.629843895204561
    },
    {
      "epoch": 0.1344447078033236,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0009780301315471232,
      "loss": 1.6478610992431642,
      "step": 93420,
      "train_bpb": 0.5827818020891855,
      "train_bytes": 28440.8375,
      "train_loss_nats_per_token": 1.6479369217411703
    },
    {
      "epoch": 0.1345763872428959,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0009774455079228264,
      "loss": 1.6364181518554688,
      "step": 93440,
      "train_bpb": 0.5763510245082653,
      "train_bytes": 28495.86875,
      "train_loss_nats_per_token": 1.6369828636925599
    },
    {
      "epoch": 0.1347080666824682,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0009768609746269538,
      "loss": 1.642169189453125,
      "step": 93460,
      "train_bpb": 0.5816555936933006,
      "train_bytes": 28142.25625,
      "train_loss_nats_per_token": 1.6437464333751652
    },
    {
      "epoch": 0.1348397461220405,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0009762765317605477,
      "loss": 1.631119728088379,
      "step": 93480,
      "train_bpb": 0.579568136894769,
      "train_bytes": 28195.15,
      "train_loss_nats_per_token": 1.632451712181244
    },
    {
      "epoch": 0.13497142556161282,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0009756921794246346,
      "loss": 1.6327531814575196,
      "step": 93500,
      "train_bpb": 0.5774567033538031,
      "train_bytes": 28424.25625,
      "train_loss_nats_per_token": 1.6339339031538578
    },
    {
      "epoch": 0.1351031050011851,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0009751079177202257,
      "loss": 1.6300512313842774,
      "step": 93520,
      "train_bpb": 0.5783315471306939,
      "train_bytes": 28035.8875,
      "train_loss_nats_per_token": 1.6317938331044164
    },
    {
      "epoch": 0.13523478444075743,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.000974523746748315,
      "loss": 1.6323549270629882,
      "step": 93540,
      "train_bpb": 0.5789512766677246,
      "train_bytes": 28442.24375,
      "train_loss_nats_per_token": 1.6322943139505484
    },
    {
      "epoch": 0.13536646388032972,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0009739396666098829,
      "loss": 1.6309940338134765,
      "step": 93560,
      "train_bpb": 0.5787885218099181,
      "train_bytes": 27910.2625,
      "train_loss_nats_per_token": 1.6318369942298707
    },
    {
      "epoch": 0.13549814331990204,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0009733556774058932,
      "loss": 1.6235204696655274,
      "step": 93580,
      "train_bpb": 0.5768135571034496,
      "train_bytes": 28061.35,
      "train_loss_nats_per_token": 1.6246749710274968
    },
    {
      "epoch": 0.13562982275947433,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.000972771779237293,
      "loss": 1.6406171798706055,
      "step": 93600,
      "train_bpb": 0.5810520584686545,
      "train_bytes": 28372.51875,
      "train_loss_nats_per_token": 1.6413251755304346
    },
    {
      "epoch": 0.13576150219904665,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.0009721879722050149,
      "loss": 1.6364131927490235,
      "step": 93620,
      "train_bpb": 0.577771386787828,
      "train_bytes": 28680.55625,
      "train_loss_nats_per_token": 1.6368827986176986
    },
    {
      "epoch": 0.13589318163861894,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0009716042564099761,
      "loss": 1.6365921020507812,
      "step": 93640,
      "train_bpb": 0.5793314372930005,
      "train_bytes": 28391.575,
      "train_loss_nats_per_token": 1.63817280888874
    },
    {
      "epoch": 0.13602486107819126,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0009710206319530763,
      "loss": 1.6317621231079102,
      "step": 93660,
      "train_bpb": 0.5790224746254963,
      "train_bytes": 28262.5125,
      "train_loss_nats_per_token": 1.6331915435909514
    },
    {
      "epoch": 0.13615654051776355,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0009704370989352006,
      "loss": 1.6481637954711914,
      "step": 93680,
      "train_bpb": 0.5828419828809386,
      "train_bytes": 28270.0625,
      "train_loss_nats_per_token": 1.6485489164176546
    },
    {
      "epoch": 0.13628821995733587,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0009698536574572186,
      "loss": 1.6413450241088867,
      "step": 93700,
      "train_bpb": 0.5834136148188732,
      "train_bytes": 28017.7875,
      "train_loss_nats_per_token": 1.6419399299187103
    },
    {
      "epoch": 0.13641989939690816,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0009692703076199841,
      "loss": 1.6358396530151367,
      "step": 93720,
      "train_bpb": 0.5819108772970815,
      "train_bytes": 28166.45,
      "train_loss_nats_per_token": 1.636855664130564
    },
    {
      "epoch": 0.13655157883648048,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0009686870495243331,
      "loss": 1.6390090942382813,
      "step": 93740,
      "train_bpb": 0.5794541838084171,
      "train_bytes": 28539.38125,
      "train_loss_nats_per_token": 1.6392712396179216
    },
    {
      "epoch": 0.13668325827605277,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0009681038832710882,
      "loss": 1.642421340942383,
      "step": 93760,
      "train_bpb": 0.5834062757638023,
      "train_bytes": 28260.50625,
      "train_loss_nats_per_token": 1.6429824363832295
    },
    {
      "epoch": 0.13681493771562508,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0009675208089610553,
      "loss": 1.6314905166625977,
      "step": 93780,
      "train_bpb": 0.579912423650107,
      "train_bytes": 28158.51875,
      "train_loss_nats_per_token": 1.6334710763165432
    },
    {
      "epoch": 0.13694661715519738,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0009669378266950236,
      "loss": 1.641739273071289,
      "step": 93800,
      "train_bpb": 0.5840175613888222,
      "train_bytes": 28080.525,
      "train_loss_nats_per_token": 1.641591275314427
    },
    {
      "epoch": 0.1370782965947697,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0009663549365737669,
      "loss": 1.634971809387207,
      "step": 93820,
      "train_bpb": 0.5811269590876371,
      "train_bytes": 28030.7625,
      "train_loss_nats_per_token": 1.6356085316724702
    },
    {
      "epoch": 0.13720997603434198,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0009657721386980444,
      "loss": 1.6410818099975586,
      "step": 93840,
      "train_bpb": 0.5796207371891192,
      "train_bytes": 28834.81875,
      "train_loss_nats_per_token": 1.6413060298108042
    },
    {
      "epoch": 0.1373416554739143,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0009651894331685969,
      "loss": 1.6431255340576172,
      "step": 93860,
      "train_bpb": 0.5799740617559795,
      "train_bytes": 28460.84375,
      "train_loss_nats_per_token": 1.6445700874186084
    },
    {
      "epoch": 0.1374733349134866,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0009646068200861509,
      "loss": 1.6331262588500977,
      "step": 93880,
      "train_bpb": 0.5752311236535829,
      "train_bytes": 28528.94375,
      "train_loss_nats_per_token": 1.6333350414347365
    },
    {
      "epoch": 0.1376050143530589,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0009640242995514165,
      "loss": 1.627403450012207,
      "step": 93900,
      "train_bpb": 0.5760421295768119,
      "train_bytes": 28268.0,
      "train_loss_nats_per_token": 1.6277944585866193
    },
    {
      "epoch": 0.13773669379263123,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0009634418716650885,
      "loss": 1.6177848815917968,
      "step": 93920,
      "train_bpb": 0.5755559392429177,
      "train_bytes": 28391.33125,
      "train_loss_nats_per_token": 1.6184613156995207
    },
    {
      "epoch": 0.13786837323220352,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0009628595365278438,
      "loss": 1.6467573165893554,
      "step": 93940,
      "train_bpb": 0.5853665566712498,
      "train_bytes": 28304.09375,
      "train_loss_nats_per_token": 1.647610898349199
    },
    {
      "epoch": 0.13800005267177584,
      "grad_norm": 0.12890625,
      "learning_rate": 0.000962277294240345,
      "loss": 1.6300529479980468,
      "step": 93960,
      "train_bpb": 0.5744025918495883,
      "train_bytes": 28648.625,
      "train_loss_nats_per_token": 1.6303057628529254
    },
    {
      "epoch": 0.13813173211134813,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0009616951449032388,
      "loss": 1.6398902893066407,
      "step": 93980,
      "train_bpb": 0.5781655156498361,
      "train_bytes": 28536.28125,
      "train_loss_nats_per_token": 1.6403228091956785
    },
    {
      "epoch": 0.13826341155092045,
      "grad_norm": 0.140625,
      "learning_rate": 0.0009611130886171539,
      "loss": 1.6317173004150392,
      "step": 94000,
      "train_bpb": 0.574574292379005,
      "train_bytes": 28564.70625,
      "train_loss_nats_per_token": 1.633182431916836
    },
    {
      "epoch": 0.13826341155092045,
      "eval_loss": 1.5201064348220825,
      "eval_runtime": 74.5416,
      "eval_samples_per_second": 13.415,
      "eval_steps_per_second": 13.415,
      "step": 94000,
      "train_bpb": 0.5364954144986799,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.527731167077703
    },
    {
      "epoch": 0.13839509099049274,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0009605311254827047,
      "loss": 1.6452445983886719,
      "step": 94020,
      "train_bpb": 0.5814499974424292,
      "train_bytes": 28338.86875,
      "train_loss_nats_per_token": 1.6454334091449847
    },
    {
      "epoch": 0.13852677043006506,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0009599492556004892,
      "loss": 1.6378572463989258,
      "step": 94040,
      "train_bpb": 0.5787217272914725,
      "train_bytes": 28258.60625,
      "train_loss_nats_per_token": 1.6392024943994792
    },
    {
      "epoch": 0.13865844986963735,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0009593674790710895,
      "loss": 1.6251667022705079,
      "step": 94060,
      "train_bpb": 0.5768404858188622,
      "train_bytes": 28059.825,
      "train_loss_nats_per_token": 1.6263154791280021
    },
    {
      "epoch": 0.13879012930920967,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0009587857959950697,
      "loss": 1.6249204635620118,
      "step": 94080,
      "train_bpb": 0.5771408185973715,
      "train_bytes": 28018.88125,
      "train_loss_nats_per_token": 1.6258425882789178
    },
    {
      "epoch": 0.13892180874878196,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0009582042064729801,
      "loss": 1.6348556518554687,
      "step": 94100,
      "train_bpb": 0.5788737317790832,
      "train_bytes": 28406.125,
      "train_loss_nats_per_token": 1.6355656112289305
    },
    {
      "epoch": 0.13905348818835428,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0009576227106053545,
      "loss": 1.6264799118041993,
      "step": 94120,
      "train_bpb": 0.5765611364546928,
      "train_bytes": 28406.0125,
      "train_loss_nats_per_token": 1.6269312290543483
    },
    {
      "epoch": 0.13918516762792657,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0009570413084927083,
      "loss": 1.6275951385498046,
      "step": 94140,
      "train_bpb": 0.5774498153573433,
      "train_bytes": 28344.60625,
      "train_loss_nats_per_token": 1.6287598692851253
    },
    {
      "epoch": 0.1393168470674989,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.000956460000235543,
      "loss": 1.6336198806762696,
      "step": 94160,
      "train_bpb": 0.578213901499634,
      "train_bytes": 28359.3125,
      "train_loss_nats_per_token": 1.6344171164196335
    },
    {
      "epoch": 0.13944852650707118,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0009558787859343438,
      "loss": 1.620503044128418,
      "step": 94180,
      "train_bpb": 0.5732237403525589,
      "train_bytes": 28345.08125,
      "train_loss_nats_per_token": 1.621950364210261
    },
    {
      "epoch": 0.1395802059466435,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0009552976656895785,
      "loss": 1.6453435897827149,
      "step": 94200,
      "train_bpb": 0.5783235998398304,
      "train_bytes": 28560.875,
      "train_loss_nats_per_token": 1.6469350498230875
    },
    {
      "epoch": 0.1397118853862158,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0009547166396016986,
      "loss": 1.6301456451416017,
      "step": 94220,
      "train_bpb": 0.580562368368489,
      "train_bytes": 28165.61875,
      "train_loss_nats_per_token": 1.6304273274721293
    },
    {
      "epoch": 0.1398435648257881,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0009541357077711408,
      "loss": 1.6377435684204102,
      "step": 94240,
      "train_bpb": 0.5808743730524928,
      "train_bytes": 28127.19375,
      "train_loss_nats_per_token": 1.6386837738623723
    },
    {
      "epoch": 0.1399752442653604,
      "grad_norm": 0.15625,
      "learning_rate": 0.0009535548702983245,
      "loss": 1.6416109085083008,
      "step": 94260,
      "train_bpb": 0.5797294337228468,
      "train_bytes": 28511.50625,
      "train_loss_nats_per_token": 1.6424763782109293
    },
    {
      "epoch": 0.14010692370493272,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0009529741272836521,
      "loss": 1.6242132186889648,
      "step": 94280,
      "train_bpb": 0.5793786176937249,
      "train_bytes": 28324.975,
      "train_loss_nats_per_token": 1.6250313587233782
    },
    {
      "epoch": 0.140238603144505,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.000952393478827511,
      "loss": 1.6468502044677735,
      "step": 94300,
      "train_bpb": 0.5847262955680409,
      "train_bytes": 28057.36875,
      "train_loss_nats_per_token": 1.6470103565618632
    },
    {
      "epoch": 0.14037028258407733,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.000951812925030272,
      "loss": 1.63560791015625,
      "step": 94320,
      "train_bpb": 0.5785422667777658,
      "train_bytes": 28029.39375,
      "train_loss_nats_per_token": 1.6364519825872534
    },
    {
      "epoch": 0.14050196202364962,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0009512324659922889,
      "loss": 1.6382070541381837,
      "step": 94340,
      "train_bpb": 0.5810224012943462,
      "train_bytes": 28235.75,
      "train_loss_nats_per_token": 1.6386623892422343
    },
    {
      "epoch": 0.14063364146322194,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0009506521018138992,
      "loss": 1.638409423828125,
      "step": 94360,
      "train_bpb": 0.5817802252541879,
      "train_bytes": 28482.33125,
      "train_loss_nats_per_token": 1.6397812266192162
    },
    {
      "epoch": 0.14076532090279423,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0009500718325954248,
      "loss": 1.646524429321289,
      "step": 94380,
      "train_bpb": 0.585400537523431,
      "train_bytes": 28148.9875,
      "train_loss_nats_per_token": 1.648002967519645
    },
    {
      "epoch": 0.14089700034236655,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0009494916584371706,
      "loss": 1.641836929321289,
      "step": 94400,
      "train_bpb": 0.5838281169366494,
      "train_bytes": 28270.275,
      "train_loss_nats_per_token": 1.6425559034311574
    },
    {
      "epoch": 0.14102867978193884,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0009489115794394244,
      "loss": 1.6435583114624024,
      "step": 94420,
      "train_bpb": 0.5814839128576863,
      "train_bytes": 28307.80625,
      "train_loss_nats_per_token": 1.6436073119309822
    },
    {
      "epoch": 0.14116035922151116,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0009483315957024589,
      "loss": 1.6572757720947267,
      "step": 94440,
      "train_bpb": 0.5826237804398884,
      "train_bytes": 28506.99375,
      "train_loss_nats_per_token": 1.6580678011383583
    },
    {
      "epoch": 0.14129203866108345,
      "grad_norm": 0.1875,
      "learning_rate": 0.0009477517073265304,
      "loss": 1.657562828063965,
      "step": 94460,
      "train_bpb": 0.5868807252162528,
      "train_bytes": 28348.06875,
      "train_loss_nats_per_token": 1.6586287851108887
    },
    {
      "epoch": 0.14142371810065577,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0009471719144118764,
      "loss": 1.6481866836547852,
      "step": 94480,
      "train_bpb": 0.5818581275074746,
      "train_bytes": 28600.03125,
      "train_loss_nats_per_token": 1.6489599282044827
    },
    {
      "epoch": 0.14155539754022806,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0009465922170587202,
      "loss": 1.652492141723633,
      "step": 94500,
      "train_bpb": 0.580178674023091,
      "train_bytes": 28580.0375,
      "train_loss_nats_per_token": 1.6538066730321732
    },
    {
      "epoch": 0.14168707697980037,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0009460126153672687,
      "loss": 1.6488662719726563,
      "step": 94520,
      "train_bpb": 0.5837774962052865,
      "train_bytes": 28406.7375,
      "train_loss_nats_per_token": 1.6502082525086632
    },
    {
      "epoch": 0.14181875641937267,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0009454331094377107,
      "loss": 1.645648193359375,
      "step": 94540,
      "train_bpb": 0.5854806746706581,
      "train_bytes": 28184.975,
      "train_loss_nats_per_token": 1.646519872210184
    },
    {
      "epoch": 0.14195043585894498,
      "grad_norm": 0.1875,
      "learning_rate": 0.000944853699370219,
      "loss": 1.6547765731811523,
      "step": 94560,
      "train_bpb": 0.5879692145038149,
      "train_bytes": 28104.70625,
      "train_loss_nats_per_token": 1.6549988689473816
    },
    {
      "epoch": 0.1420821152985173,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0009442743852649504,
      "loss": 1.658769416809082,
      "step": 94580,
      "train_bpb": 0.5826363517404373,
      "train_bytes": 28442.0375,
      "train_loss_nats_per_token": 1.659264729260412
    },
    {
      "epoch": 0.1422137947380896,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0009436951672220454,
      "loss": 1.650425910949707,
      "step": 94600,
      "train_bpb": 0.5802360411389957,
      "train_bytes": 28493.65,
      "train_loss_nats_per_token": 1.6514079471978633
    },
    {
      "epoch": 0.1423454741776619,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0009431160453416258,
      "loss": 1.6406333923339844,
      "step": 94620,
      "train_bpb": 0.5799737341041015,
      "train_bytes": 28315.83125,
      "train_loss_nats_per_token": 1.6415696010532645
    },
    {
      "epoch": 0.1424771536172342,
      "grad_norm": 0.171875,
      "learning_rate": 0.0009425370197237992,
      "loss": 1.6406391143798829,
      "step": 94640,
      "train_bpb": 0.5829979995510036,
      "train_bytes": 28197.96875,
      "train_loss_nats_per_token": 1.64121076941771
    },
    {
      "epoch": 0.14260883305680652,
      "grad_norm": 0.15625,
      "learning_rate": 0.0009419580904686556,
      "loss": 1.6436012268066407,
      "step": 94660,
      "train_bpb": 0.5789030652062247,
      "train_bytes": 28339.3375,
      "train_loss_nats_per_token": 1.6436132829790788
    },
    {
      "epoch": 0.1427405124963788,
      "grad_norm": 0.15625,
      "learning_rate": 0.000941379257676268,
      "loss": 1.6536056518554687,
      "step": 94680,
      "train_bpb": 0.5828618853074086,
      "train_bytes": 28500.6,
      "train_loss_nats_per_token": 1.6532911155742551
    },
    {
      "epoch": 0.14287219193595113,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0009408005214466931,
      "loss": 1.6452133178710937,
      "step": 94700,
      "train_bpb": 0.5810898274024865,
      "train_bytes": 28053.91875,
      "train_loss_nats_per_token": 1.6461940942977464
    },
    {
      "epoch": 0.14300387137552342,
      "grad_norm": 0.16015625,
      "learning_rate": 0.000940221881879971,
      "loss": 1.6496334075927734,
      "step": 94720,
      "train_bpb": 0.5820344879135189,
      "train_bytes": 28592.13125,
      "train_loss_nats_per_token": 1.649723409436925
    },
    {
      "epoch": 0.14313555081509574,
      "grad_norm": 0.146484375,
      "learning_rate": 0.000939643339076125,
      "loss": 1.6464969635009765,
      "step": 94740,
      "train_bpb": 0.5817830791593307,
      "train_bytes": 28228.23125,
      "train_loss_nats_per_token": 1.6472755076889178
    },
    {
      "epoch": 0.14326723025466803,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0009390648931351613,
      "loss": 1.6474292755126954,
      "step": 94760,
      "train_bpb": 0.5818918219128865,
      "train_bytes": 28308.94375,
      "train_loss_nats_per_token": 1.648025441420733
    },
    {
      "epoch": 0.14339890969424035,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0009384865441570699,
      "loss": 1.6391420364379883,
      "step": 94780,
      "train_bpb": 0.5811307837592442,
      "train_bytes": 28350.3,
      "train_loss_nats_per_token": 1.6395143663317902
    },
    {
      "epoch": 0.14353058913381264,
      "grad_norm": 0.15625,
      "learning_rate": 0.0009379082922418244,
      "loss": 1.6482124328613281,
      "step": 94800,
      "train_bpb": 0.5796551863757474,
      "train_bytes": 28799.4875,
      "train_loss_nats_per_token": 1.6498981773012507
    },
    {
      "epoch": 0.14366226857338496,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0009373301374893798,
      "loss": 1.6547229766845704,
      "step": 94820,
      "train_bpb": 0.5837216623609425,
      "train_bytes": 28373.71875,
      "train_loss_nats_per_token": 1.6545339362294806
    },
    {
      "epoch": 0.14379394801295725,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0009367520799996763,
      "loss": 1.652752685546875,
      "step": 94840,
      "train_bpb": 0.5845069830939993,
      "train_bytes": 28185.64375,
      "train_loss_nats_per_token": 1.6533599835151256
    },
    {
      "epoch": 0.14392562745252957,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0009361741198726367,
      "loss": 1.6343219757080079,
      "step": 94860,
      "train_bpb": 0.5796965977680681,
      "train_bytes": 28196.325,
      "train_loss_nats_per_token": 1.6342727343537997
    },
    {
      "epoch": 0.14405730689210186,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0009355962572081665,
      "loss": 1.6453245162963868,
      "step": 94880,
      "train_bpb": 0.5808457071978539,
      "train_bytes": 28208.8625,
      "train_loss_nats_per_token": 1.646481640841963
    },
    {
      "epoch": 0.14418898633167418,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0009350184921061544,
      "loss": 1.6669956207275392,
      "step": 94900,
      "train_bpb": 0.5884002951629231,
      "train_bytes": 28755.925,
      "train_loss_nats_per_token": 1.6678776666625004
    },
    {
      "epoch": 0.14432066577124647,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0009344408246664733,
      "loss": 1.6457817077636718,
      "step": 94920,
      "train_bpb": 0.5818863948329532,
      "train_bytes": 28508.29375,
      "train_loss_nats_per_token": 1.6478051998897199
    },
    {
      "epoch": 0.1444523452108188,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0009338632549889778,
      "loss": 1.6459692001342774,
      "step": 94940,
      "train_bpb": 0.5826005658931654,
      "train_bytes": 28511.8375,
      "train_loss_nats_per_token": 1.646650829200346
    },
    {
      "epoch": 0.14458402465039108,
      "grad_norm": 0.1875,
      "learning_rate": 0.0009332857831735059,
      "loss": 1.652542495727539,
      "step": 94960,
      "train_bpb": 0.584409893052382,
      "train_bytes": 28372.88125,
      "train_loss_nats_per_token": 1.654234975696103
    },
    {
      "epoch": 0.1447157040899634,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0009327084093198796,
      "loss": 1.6424230575561523,
      "step": 94980,
      "train_bpb": 0.5848413183766428,
      "train_bytes": 27965.99375,
      "train_loss_nats_per_token": 1.642665335657267
    },
    {
      "epoch": 0.1448473835295357,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0009321311335279035,
      "loss": 1.651705551147461,
      "step": 95000,
      "train_bpb": 0.5801450956293323,
      "train_bytes": 28618.1375,
      "train_loss_nats_per_token": 1.6523254108516074
    },
    {
      "epoch": 0.1448473835295357,
      "eval_loss": 1.519939661026001,
      "eval_runtime": 79.3422,
      "eval_samples_per_second": 12.604,
      "eval_steps_per_second": 12.604,
      "step": 95000,
      "train_bpb": 0.536461736588248,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5276352654315404
    },
    {
      "epoch": 0.144979062969108,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0009315539558973648,
      "loss": 1.7190870285034179,
      "step": 95020,
      "train_bpb": 0.6093398436978615,
      "train_bytes": 27880.66875,
      "train_loss_nats_per_token": 1.721135540678568
    },
    {
      "epoch": 0.1451107424086803,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0009309768765280339,
      "loss": 1.7090419769287108,
      "step": 95040,
      "train_bpb": 0.6041477714139694,
      "train_bytes": 28201.7625,
      "train_loss_nats_per_token": 1.7098769798619307
    },
    {
      "epoch": 0.14524242184825262,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0009303998955196644,
      "loss": 1.718939208984375,
      "step": 95060,
      "train_bpb": 0.6039750055969214,
      "train_bytes": 28556.88125,
      "train_loss_nats_per_token": 1.7188924435925514
    },
    {
      "epoch": 0.1453741012878249,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0009298230129719932,
      "loss": 1.7042245864868164,
      "step": 95080,
      "train_bpb": 0.604459977115566,
      "train_bytes": 28125.21875,
      "train_loss_nats_per_token": 1.7052208233832642
    },
    {
      "epoch": 0.14550578072739723,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0009292462289847396,
      "loss": 1.7076333999633788,
      "step": 95100,
      "train_bpb": 0.6025392525074811,
      "train_bytes": 28126.91875,
      "train_loss_nats_per_token": 1.708306964476866
    },
    {
      "epoch": 0.14563746016696952,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.0009286695436576056,
      "loss": 1.7039628982543946,
      "step": 95120,
      "train_bpb": 0.6046959062536605,
      "train_bytes": 28116.5,
      "train_loss_nats_per_token": 1.7043294368518065
    },
    {
      "epoch": 0.14576913960654184,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.000928092957090278,
      "loss": 1.704327392578125,
      "step": 95140,
      "train_bpb": 0.6025830234964993,
      "train_bytes": 28197.75625,
      "train_loss_nats_per_token": 1.7049222414936713
    },
    {
      "epoch": 0.14590081904611413,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0009275164693824237,
      "loss": 1.7044149398803712,
      "step": 95160,
      "train_bpb": 0.6007837658753755,
      "train_bytes": 28236.925,
      "train_loss_nats_per_token": 1.7053525796109748
    },
    {
      "epoch": 0.14603249848568645,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0009269400806336947,
      "loss": 1.7088064193725585,
      "step": 95180,
      "train_bpb": 0.6031776911840173,
      "train_bytes": 28378.55,
      "train_loss_nats_per_token": 1.7092471634776847
    },
    {
      "epoch": 0.14616417792525874,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0009263637909437253,
      "loss": 1.7114120483398438,
      "step": 95200,
      "train_bpb": 0.603227533466954,
      "train_bytes": 28402.48125,
      "train_loss_nats_per_token": 1.7117022131647939
    },
    {
      "epoch": 0.14629585736483106,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0009257876004121323,
      "loss": 1.7003568649291991,
      "step": 95220,
      "train_bpb": 0.6043115244430842,
      "train_bytes": 27937.80625,
      "train_loss_nats_per_token": 1.7021461843510886
    },
    {
      "epoch": 0.14642753680440337,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0009252115091385154,
      "loss": 1.7040756225585938,
      "step": 95240,
      "train_bpb": 0.6069188233331904,
      "train_bytes": 27873.28125,
      "train_loss_nats_per_token": 1.7044773550992087
    },
    {
      "epoch": 0.14655921624397567,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0009246355172224578,
      "loss": 1.7194852828979492,
      "step": 95260,
      "train_bpb": 0.606944995420696,
      "train_bytes": 28249.525,
      "train_loss_nats_per_token": 1.719125608092493
    },
    {
      "epoch": 0.14669089568354798,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0009240596247635252,
      "loss": 1.7120227813720703,
      "step": 95280,
      "train_bpb": 0.6024481643391331,
      "train_bytes": 28509.575,
      "train_loss_nats_per_token": 1.712650265861199
    },
    {
      "epoch": 0.14682257512312027,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0009234838318612651,
      "loss": 1.7068548202514648,
      "step": 95300,
      "train_bpb": 0.6044527552130733,
      "train_bytes": 28106.76875,
      "train_loss_nats_per_token": 1.7070275614663823
    },
    {
      "epoch": 0.1469542545626926,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0009229081386152094,
      "loss": 1.7022993087768554,
      "step": 95320,
      "train_bpb": 0.6048014081934504,
      "train_bytes": 27973.9375,
      "train_loss_nats_per_token": 1.7022674313030728
    },
    {
      "epoch": 0.14708593400226488,
      "grad_norm": 0.1943359375,
      "learning_rate": 0.0009223325451248722,
      "loss": 1.7011146545410156,
      "step": 95340,
      "train_bpb": 0.6017001280078664,
      "train_bytes": 28369.2375,
      "train_loss_nats_per_token": 1.7015652121006293
    },
    {
      "epoch": 0.1472176134418372,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0009217570514897499,
      "loss": 1.7209228515625,
      "step": 95360,
      "train_bpb": 0.6051625192227708,
      "train_bytes": 28491.05,
      "train_loss_nats_per_token": 1.7208920210000527
    },
    {
      "epoch": 0.1473492928814095,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0009211816578093219,
      "loss": 1.7098424911499024,
      "step": 95380,
      "train_bpb": 0.6030886548630252,
      "train_bytes": 28421.7375,
      "train_loss_nats_per_token": 1.7111904598482326
    },
    {
      "epoch": 0.1474809723209818,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0009206063641830504,
      "loss": 1.7157215118408202,
      "step": 95400,
      "train_bpb": 0.6057670401035388,
      "train_bytes": 28298.6875,
      "train_loss_nats_per_token": 1.7163405153506792
    },
    {
      "epoch": 0.1476126517605541,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0009200311707103805,
      "loss": 1.696200180053711,
      "step": 95420,
      "train_bpb": 0.6029178710661375,
      "train_bytes": 28211.69375,
      "train_loss_nats_per_token": 1.6966002636933408
    },
    {
      "epoch": 0.14774433120012642,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0009194560774907393,
      "loss": 1.709141731262207,
      "step": 95440,
      "train_bpb": 0.6027867672920109,
      "train_bytes": 28760.21875,
      "train_loss_nats_per_token": 1.7093699689545967
    },
    {
      "epoch": 0.1478760106396987,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0009188810846235372,
      "loss": 1.7037792205810547,
      "step": 95460,
      "train_bpb": 0.6016007308799143,
      "train_bytes": 28113.36875,
      "train_loss_nats_per_token": 1.7037976985647338
    },
    {
      "epoch": 0.14800769007927103,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.0009183061922081677,
      "loss": 1.703654670715332,
      "step": 95480,
      "train_bpb": 0.6014951956956033,
      "train_bytes": 28196.0,
      "train_loss_nats_per_token": 1.7043104008983683
    },
    {
      "epoch": 0.14813936951884332,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0009177314003440059,
      "loss": 1.711880874633789,
      "step": 95500,
      "train_bpb": 0.6060015120063672,
      "train_bytes": 27951.14375,
      "train_loss_nats_per_token": 1.7118295031936206
    },
    {
      "epoch": 0.14827104895841564,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0009171567091304094,
      "loss": 1.7037710189819335,
      "step": 95520,
      "train_bpb": 0.6029902762821988,
      "train_bytes": 28516.65,
      "train_loss_nats_per_token": 1.7050110581802866
    },
    {
      "epoch": 0.14840272839798793,
      "grad_norm": 0.2216796875,
      "learning_rate": 0.0009165821186667198,
      "loss": 1.707832145690918,
      "step": 95540,
      "train_bpb": 0.6003159573354963,
      "train_bytes": 28524.8375,
      "train_loss_nats_per_token": 1.7091651455270944
    },
    {
      "epoch": 0.14853440783756025,
      "grad_norm": 0.169921875,
      "learning_rate": 0.00091600762905226,
      "loss": 1.7060239791870118,
      "step": 95560,
      "train_bpb": 0.6032677126523251,
      "train_bytes": 28212.21875,
      "train_loss_nats_per_token": 1.7062051189663858
    },
    {
      "epoch": 0.14866608727713254,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0009154332403863356,
      "loss": 1.715734100341797,
      "step": 95580,
      "train_bpb": 0.6033015060307502,
      "train_bytes": 28664.2125,
      "train_loss_nats_per_token": 1.7157968431775452
    },
    {
      "epoch": 0.14879776671670486,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0009148589527682358,
      "loss": 1.71593017578125,
      "step": 95600,
      "train_bpb": 0.6054203952631815,
      "train_bytes": 28444.48125,
      "train_loss_nats_per_token": 1.7162519757206107
    },
    {
      "epoch": 0.14892944615627715,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0009142847662972312,
      "loss": 1.6997478485107422,
      "step": 95620,
      "train_bpb": 0.6022188145391113,
      "train_bytes": 28048.78125,
      "train_loss_nats_per_token": 1.7007631820559337
    },
    {
      "epoch": 0.14906112559584947,
      "grad_norm": 0.10546875,
      "learning_rate": 0.0009137106810725748,
      "loss": 1.7125518798828125,
      "step": 95640,
      "train_bpb": 0.6074833296361932,
      "train_bytes": 28026.99375,
      "train_loss_nats_per_token": 1.712518161131948
    },
    {
      "epoch": 0.14919280503542176,
      "grad_norm": 0.125,
      "learning_rate": 0.0009131366971935034,
      "loss": 1.7098072052001954,
      "step": 95660,
      "train_bpb": 0.6033933552938161,
      "train_bytes": 28333.70625,
      "train_loss_nats_per_token": 1.7097487609905606
    },
    {
      "epoch": 0.14932448447499408,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0009125628147592352,
      "loss": 1.724881362915039,
      "step": 95680,
      "train_bpb": 0.6047435795170906,
      "train_bytes": 28754.51875,
      "train_loss_nats_per_token": 1.7251962488311847
    },
    {
      "epoch": 0.14945616391456637,
      "grad_norm": 0.1953125,
      "learning_rate": 0.0009119890338689709,
      "loss": 1.7066352844238282,
      "step": 95700,
      "train_bpb": 0.6020684650221082,
      "train_bytes": 28427.98125,
      "train_loss_nats_per_token": 1.7065373493182814
    },
    {
      "epoch": 0.1495878433541387,
      "grad_norm": 0.12890625,
      "learning_rate": 0.000911415354621894,
      "loss": 1.7076074600219726,
      "step": 95720,
      "train_bpb": 0.606861603950306,
      "train_bytes": 28210.0125,
      "train_loss_nats_per_token": 1.708430431438823
    },
    {
      "epoch": 0.14971952279371098,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0009108417771171707,
      "loss": 1.714594841003418,
      "step": 95740,
      "train_bpb": 0.6052636527208017,
      "train_bytes": 28418.81875,
      "train_loss_nats_per_token": 1.7143597826855599
    },
    {
      "epoch": 0.1498512022332833,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0009102683014539488,
      "loss": 1.702469825744629,
      "step": 95760,
      "train_bpb": 0.6032826137453359,
      "train_bytes": 28148.4875,
      "train_loss_nats_per_token": 1.7028669598915154
    },
    {
      "epoch": 0.1499828816728556,
      "grad_norm": 0.15625,
      "learning_rate": 0.0009096949277313588,
      "loss": 1.719025230407715,
      "step": 95780,
      "train_bpb": 0.6063180067553239,
      "train_bytes": 28704.01875,
      "train_loss_nats_per_token": 1.7198389428174723
    },
    {
      "epoch": 0.1501145611124279,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0009091216560485143,
      "loss": 1.7078948974609376,
      "step": 95800,
      "train_bpb": 0.6060050866714778,
      "train_bytes": 28095.81875,
      "train_loss_nats_per_token": 1.7088327379287203
    },
    {
      "epoch": 0.1502462405520002,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0009085484865045098,
      "loss": 1.7067127227783203,
      "step": 95820,
      "train_bpb": 0.6012260196312187,
      "train_bytes": 28513.59375,
      "train_loss_nats_per_token": 1.7070808123446104
    },
    {
      "epoch": 0.15037791999157252,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0009079754191984239,
      "loss": 1.6988542556762696,
      "step": 95840,
      "train_bpb": 0.5999672096031596,
      "train_bytes": 28547.86875,
      "train_loss_nats_per_token": 1.6989293734979778
    },
    {
      "epoch": 0.1505095994311448,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0009074024542293159,
      "loss": 1.7154918670654298,
      "step": 95860,
      "train_bpb": 0.6080230643400178,
      "train_bytes": 28127.91875,
      "train_loss_nats_per_token": 1.7167460852362775
    },
    {
      "epoch": 0.15064127887071713,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0009068295916962289,
      "loss": 1.7033304214477538,
      "step": 95880,
      "train_bpb": 0.5986086942747421,
      "train_bytes": 28587.2375,
      "train_loss_nats_per_token": 1.702841971803792
    },
    {
      "epoch": 0.15077295831028942,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0009062568316981871,
      "loss": 1.7053024291992187,
      "step": 95900,
      "train_bpb": 0.6027677403061455,
      "train_bytes": 28222.88125,
      "train_loss_nats_per_token": 1.7063917008438354
    },
    {
      "epoch": 0.15090463774986174,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0009056841743341971,
      "loss": 1.7176174163818358,
      "step": 95920,
      "train_bpb": 0.6068557391141673,
      "train_bytes": 28444.58125,
      "train_loss_nats_per_token": 1.7178446729135783
    },
    {
      "epoch": 0.15103631718943406,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0009051116197032489,
      "loss": 1.7154226303100586,
      "step": 95940,
      "train_bpb": 0.6080434121678144,
      "train_bytes": 28380.6375,
      "train_loss_nats_per_token": 1.7156331425657831
    },
    {
      "epoch": 0.15116799662900635,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0009045391679043131,
      "loss": 1.7020042419433594,
      "step": 95960,
      "train_bpb": 0.6033364804956379,
      "train_bytes": 28177.96875,
      "train_loss_nats_per_token": 1.7033995423499535
    },
    {
      "epoch": 0.15129967606857866,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0009039668190363435,
      "loss": 1.7243108749389648,
      "step": 95980,
      "train_bpb": 0.6056138892694818,
      "train_bytes": 28773.3625,
      "train_loss_nats_per_token": 1.724319402950242
    },
    {
      "epoch": 0.15143135550815096,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0009033945731982761,
      "loss": 1.7056566238403321,
      "step": 96000,
      "train_bpb": 0.6028767622741884,
      "train_bytes": 28477.1625,
      "train_loss_nats_per_token": 1.7059092342583861
    },
    {
      "epoch": 0.15143135550815096,
      "eval_loss": 1.528017282485962,
      "eval_runtime": 74.8411,
      "eval_samples_per_second": 13.362,
      "eval_steps_per_second": 13.362,
      "step": 96000,
      "train_bpb": 0.5392267356371999,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.535508911896853
    },
    {
      "epoch": 0.15156303494772327,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0009028224304890294,
      "loss": 1.706231689453125,
      "step": 96020,
      "train_bpb": 0.6060019204918843,
      "train_bytes": 28387.98125,
      "train_loss_nats_per_token": 1.707734317273735
    },
    {
      "epoch": 0.15169471438729557,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0009022503910075031,
      "loss": 1.7198251724243163,
      "step": 96040,
      "train_bpb": 0.6074794700145093,
      "train_bytes": 28474.01875,
      "train_loss_nats_per_token": 1.7206518929846455
    },
    {
      "epoch": 0.15182639382686788,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0009016784548525794,
      "loss": 1.6977010726928712,
      "step": 96060,
      "train_bpb": 0.6014357874273449,
      "train_bytes": 28119.36875,
      "train_loss_nats_per_token": 1.6990779264767395
    },
    {
      "epoch": 0.15195807326644017,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0009011066221231232,
      "loss": 1.719192123413086,
      "step": 96080,
      "train_bpb": 0.6048781047917404,
      "train_bytes": 28546.94375,
      "train_loss_nats_per_token": 1.7190546590620919
    },
    {
      "epoch": 0.1520897527060125,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0009005348929179811,
      "loss": 1.7105871200561524,
      "step": 96100,
      "train_bpb": 0.6052223111212718,
      "train_bytes": 28158.56875,
      "train_loss_nats_per_token": 1.710938734980096
    },
    {
      "epoch": 0.15222143214558478,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0008999632673359813,
      "loss": 1.7072542190551758,
      "step": 96120,
      "train_bpb": 0.6035046026838982,
      "train_bytes": 28463.2375,
      "train_loss_nats_per_token": 1.7082235903492253
    },
    {
      "epoch": 0.1523531115851571,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0008993917454759355,
      "loss": 1.71112060546875,
      "step": 96140,
      "train_bpb": 0.6059590838598097,
      "train_bytes": 28385.43125,
      "train_loss_nats_per_token": 1.7117195670538536
    },
    {
      "epoch": 0.1524847910247294,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0008988203274366356,
      "loss": 1.6971397399902344,
      "step": 96160,
      "train_bpb": 0.603632027882881,
      "train_bytes": 28130.36875,
      "train_loss_nats_per_token": 1.6974154441617888
    },
    {
      "epoch": 0.1526164704643017,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0008982490133168574,
      "loss": 1.7221086502075196,
      "step": 96180,
      "train_bpb": 0.6058118429543412,
      "train_bytes": 28517.55625,
      "train_loss_nats_per_token": 1.7223230699458898
    },
    {
      "epoch": 0.152748149903874,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0008976778032153574,
      "loss": 1.7185670852661132,
      "step": 96200,
      "train_bpb": 0.6061443231201026,
      "train_bytes": 28597.8,
      "train_loss_nats_per_token": 1.7187714136538892
    },
    {
      "epoch": 0.15287982934344632,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.000897106697230875,
      "loss": 1.7045419692993165,
      "step": 96220,
      "train_bpb": 0.6045240271013362,
      "train_bytes": 27843.35,
      "train_loss_nats_per_token": 1.7050883946336055
    },
    {
      "epoch": 0.1530115087830186,
      "grad_norm": 0.2041015625,
      "learning_rate": 0.0008965356954621309,
      "loss": 1.7011941909790038,
      "step": 96240,
      "train_bpb": 0.6009061911394035,
      "train_bytes": 28124.7875,
      "train_loss_nats_per_token": 1.7019298212570997
    },
    {
      "epoch": 0.15314318822259093,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0008959647980078277,
      "loss": 1.710463523864746,
      "step": 96260,
      "train_bpb": 0.6035575734180033,
      "train_bytes": 28454.475,
      "train_loss_nats_per_token": 1.7116523490788396
    },
    {
      "epoch": 0.15327486766216322,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0008953940049666512,
      "loss": 1.7112716674804687,
      "step": 96280,
      "train_bpb": 0.6038570128749241,
      "train_bytes": 28446.5875,
      "train_loss_nats_per_token": 1.7120883846556707
    },
    {
      "epoch": 0.15340654710173554,
      "grad_norm": 0.142578125,
      "learning_rate": 0.000894823316437268,
      "loss": 1.6881231307983398,
      "step": 96300,
      "train_bpb": 0.5978223964588301,
      "train_bytes": 28309.91875,
      "train_loss_nats_per_token": 1.6888068969582133
    },
    {
      "epoch": 0.15353822654130783,
      "grad_norm": 0.66015625,
      "learning_rate": 0.0008942527325183264,
      "loss": 1.7038166046142578,
      "step": 96320,
      "train_bpb": 0.6028835691462051,
      "train_bytes": 28224.70625,
      "train_loss_nats_per_token": 1.7048069897237983
    },
    {
      "epoch": 0.15366990598088015,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.000893682253308458,
      "loss": 1.7030439376831055,
      "step": 96340,
      "train_bpb": 0.5999882936619342,
      "train_bytes": 28276.9,
      "train_loss_nats_per_token": 1.7028061349301482
    },
    {
      "epoch": 0.15380158542045244,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0008931118789062748,
      "loss": 1.7082128524780273,
      "step": 96360,
      "train_bpb": 0.6061821979016546,
      "train_bytes": 28144.8625,
      "train_loss_nats_per_token": 1.7087083494356845
    },
    {
      "epoch": 0.15393326486002476,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0008925416094103718,
      "loss": 1.7086477279663086,
      "step": 96380,
      "train_bpb": 0.606055289736087,
      "train_bytes": 28250.74375,
      "train_loss_nats_per_token": 1.7088069354658362
    },
    {
      "epoch": 0.15406494429959705,
      "grad_norm": 0.12890625,
      "learning_rate": 0.000891971444919325,
      "loss": 1.7105846405029297,
      "step": 96400,
      "train_bpb": 0.6023958296198173,
      "train_bytes": 28636.58125,
      "train_loss_nats_per_token": 1.7108318711304142
    },
    {
      "epoch": 0.15419662373916937,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0008914013855316932,
      "loss": 1.7035861968994142,
      "step": 96420,
      "train_bpb": 0.6025552046110522,
      "train_bytes": 28397.0375,
      "train_loss_nats_per_token": 1.704703117343184
    },
    {
      "epoch": 0.15432830317874166,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0008908314313460161,
      "loss": 1.714505386352539,
      "step": 96440,
      "train_bpb": 0.6036161307088231,
      "train_bytes": 28633.75,
      "train_loss_nats_per_token": 1.714586539924221
    },
    {
      "epoch": 0.15445998261831398,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0008902615824608153,
      "loss": 1.7022619247436523,
      "step": 96460,
      "train_bpb": 0.601748721131171,
      "train_bytes": 28264.39375,
      "train_loss_nats_per_token": 1.7034027547309105
    },
    {
      "epoch": 0.15459166205788627,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0008896918389745952,
      "loss": 1.7073741912841798,
      "step": 96480,
      "train_bpb": 0.6057777951494951,
      "train_bytes": 28072.44375,
      "train_loss_nats_per_token": 1.707238779288468
    },
    {
      "epoch": 0.1547233414974586,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0008891222009858407,
      "loss": 1.7049394607543946,
      "step": 96500,
      "train_bpb": 0.604837940476666,
      "train_bytes": 28200.34375,
      "train_loss_nats_per_token": 1.7054584720351453
    },
    {
      "epoch": 0.15485502093703088,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0008885526685930196,
      "loss": 1.715053176879883,
      "step": 96520,
      "train_bpb": 0.6088905035527877,
      "train_bytes": 28176.06875,
      "train_loss_nats_per_token": 1.7155436687629781
    },
    {
      "epoch": 0.1549867003766032,
      "grad_norm": 0.25,
      "learning_rate": 0.0008879832418945803,
      "loss": 1.7019725799560548,
      "step": 96540,
      "train_bpb": 0.6013591330646739,
      "train_bytes": 28205.4625,
      "train_loss_nats_per_token": 1.7027917725124964
    },
    {
      "epoch": 0.1551183798161755,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0008874139209889541,
      "loss": 1.707000732421875,
      "step": 96560,
      "train_bpb": 0.6018915109406643,
      "train_bytes": 28393.4875,
      "train_loss_nats_per_token": 1.7074380125532977
    },
    {
      "epoch": 0.1552500592557478,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0008868447059745533,
      "loss": 1.7087844848632812,
      "step": 96580,
      "train_bpb": 0.6031222530440157,
      "train_bytes": 28608.50625,
      "train_loss_nats_per_token": 1.7091980894162164
    },
    {
      "epoch": 0.15538173869532013,
      "grad_norm": 0.20703125,
      "learning_rate": 0.0008862755969497713,
      "loss": 1.7120590209960938,
      "step": 96600,
      "train_bpb": 0.6046083744577644,
      "train_bytes": 28232.54375,
      "train_loss_nats_per_token": 1.7130358858264132
    },
    {
      "epoch": 0.15551341813489242,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0008857065940129848,
      "loss": 1.7057601928710937,
      "step": 96620,
      "train_bpb": 0.6035767295982335,
      "train_bytes": 28275.1,
      "train_loss_nats_per_token": 1.7070263397561918
    },
    {
      "epoch": 0.15564509757446474,
      "grad_norm": 0.140625,
      "learning_rate": 0.000885137697262551,
      "loss": 1.7131263732910156,
      "step": 96640,
      "train_bpb": 0.6043166280101848,
      "train_bytes": 28260.75625,
      "train_loss_nats_per_token": 1.713664412344133
    },
    {
      "epoch": 0.15577677701403703,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.000884568906796809,
      "loss": 1.719533348083496,
      "step": 96660,
      "train_bpb": 0.6073879069115424,
      "train_bytes": 28278.8,
      "train_loss_nats_per_token": 1.7202748816840758
    },
    {
      "epoch": 0.15590845645360935,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0008840002227140796,
      "loss": 1.709517478942871,
      "step": 96680,
      "train_bpb": 0.6050861343886899,
      "train_bytes": 28215.1125,
      "train_loss_nats_per_token": 1.7108122693164034
    },
    {
      "epoch": 0.15604013589318164,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0008834316451126647,
      "loss": 1.7221654891967773,
      "step": 96700,
      "train_bpb": 0.6086486728742158,
      "train_bytes": 28276.9625,
      "train_loss_nats_per_token": 1.7222467859825585
    },
    {
      "epoch": 0.15617181533275395,
      "grad_norm": 0.142578125,
      "learning_rate": 0.000882863174090849,
      "loss": 1.734154510498047,
      "step": 96720,
      "train_bpb": 0.6128983120659469,
      "train_bytes": 28350.1125,
      "train_loss_nats_per_token": 1.734993087915622
    },
    {
      "epoch": 0.15630349477232625,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0008822948097468975,
      "loss": 1.714712142944336,
      "step": 96740,
      "train_bpb": 0.6072384522388864,
      "train_bytes": 28334.85,
      "train_loss_nats_per_token": 1.7162100546288224
    },
    {
      "epoch": 0.15643517421189856,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0008817265521790576,
      "loss": 1.7266813278198243,
      "step": 96760,
      "train_bpb": 0.6110520394466599,
      "train_bytes": 28435.4,
      "train_loss_nats_per_token": 1.7273700217312844
    },
    {
      "epoch": 0.15656685365147086,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0008811584014855578,
      "loss": 1.7203510284423829,
      "step": 96780,
      "train_bpb": 0.6085557880613963,
      "train_bytes": 28400.925,
      "train_loss_nats_per_token": 1.7206878889078276
    },
    {
      "epoch": 0.15669853309104317,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0008805903577646079,
      "loss": 1.7153970718383789,
      "step": 96800,
      "train_bpb": 0.606132494727477,
      "train_bytes": 28329.78125,
      "train_loss_nats_per_token": 1.7163344409025598
    },
    {
      "epoch": 0.15683021253061546,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0008800224211144001,
      "loss": 1.7110122680664062,
      "step": 96820,
      "train_bpb": 0.6065627877714972,
      "train_bytes": 28484.8,
      "train_loss_nats_per_token": 1.7110201995574381
    },
    {
      "epoch": 0.15696189197018778,
      "grad_norm": 0.11328125,
      "learning_rate": 0.0008794545916331072,
      "loss": 1.7139879226684571,
      "step": 96840,
      "train_bpb": 0.606811151138485,
      "train_bytes": 27930.3375,
      "train_loss_nats_per_token": 1.713660448184724
    },
    {
      "epoch": 0.15709357140976007,
      "grad_norm": 0.193359375,
      "learning_rate": 0.0008788868694188841,
      "loss": 1.7143594741821289,
      "step": 96860,
      "train_bpb": 0.6049792489466518,
      "train_bytes": 28333.96875,
      "train_loss_nats_per_token": 1.7154200592725601
    },
    {
      "epoch": 0.1572252508493324,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0008783192545698669,
      "loss": 1.71708984375,
      "step": 96880,
      "train_bpb": 0.6090678745403622,
      "train_bytes": 28155.80625,
      "train_loss_nats_per_token": 1.7178110272128917
    },
    {
      "epoch": 0.15735693028890468,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0008777517471841726,
      "loss": 1.7153387069702148,
      "step": 96900,
      "train_bpb": 0.6057009470217097,
      "train_bytes": 28571.99375,
      "train_loss_nats_per_token": 1.7152849579287663
    },
    {
      "epoch": 0.157488609728477,
      "grad_norm": 0.14453125,
      "learning_rate": 0.000877184347359901,
      "loss": 1.711075210571289,
      "step": 96920,
      "train_bpb": 0.6043619921329981,
      "train_bytes": 28514.25,
      "train_loss_nats_per_token": 1.712243929863769
    },
    {
      "epoch": 0.1576202891680493,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0008766170551951316,
      "loss": 1.7041330337524414,
      "step": 96940,
      "train_bpb": 0.6021504251481438,
      "train_bytes": 28048.675,
      "train_loss_nats_per_token": 1.7048742960502872
    },
    {
      "epoch": 0.1577519686076216,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0008760498707879267,
      "loss": 1.7087915420532227,
      "step": 96960,
      "train_bpb": 0.6052154186715671,
      "train_bytes": 28252.86875,
      "train_loss_nats_per_token": 1.7094547642228997
    },
    {
      "epoch": 0.1578836480471939,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0008754827942363294,
      "loss": 1.7256914138793946,
      "step": 96980,
      "train_bpb": 0.6114879569626126,
      "train_bytes": 28244.14375,
      "train_loss_nats_per_token": 1.7266259616808284
    },
    {
      "epoch": 0.15801532748676622,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0008749158256383636,
      "loss": 1.705889892578125,
      "step": 97000,
      "train_bpb": 0.606905341734699,
      "train_bytes": 28004.7,
      "train_loss_nats_per_token": 1.7063754618635965
    },
    {
      "epoch": 0.15801532748676622,
      "eval_loss": 1.53024160861969,
      "eval_runtime": 75.2645,
      "eval_samples_per_second": 13.286,
      "eval_steps_per_second": 13.286,
      "step": 97000,
      "train_bpb": 0.5400173376330201,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5377602401973882
    },
    {
      "epoch": 0.1581470069263385,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0008743489650920363,
      "loss": 1.7304197311401368,
      "step": 97020,
      "train_bpb": 0.6089478310690976,
      "train_bytes": 28671.1875,
      "train_loss_nats_per_token": 1.730615188071403
    },
    {
      "epoch": 0.15827868636591083,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0008737822126953335,
      "loss": 1.7193201065063477,
      "step": 97040,
      "train_bpb": 0.6073023971280734,
      "train_bytes": 28385.4125,
      "train_loss_nats_per_token": 1.720925546352208
    },
    {
      "epoch": 0.15841036580548312,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0008732155685462238,
      "loss": 1.7161243438720704,
      "step": 97060,
      "train_bpb": 0.6059167214593474,
      "train_bytes": 28408.21875,
      "train_loss_nats_per_token": 1.7167141560421255
    },
    {
      "epoch": 0.15854204524505544,
      "grad_norm": 0.171875,
      "learning_rate": 0.0008726490327426573,
      "loss": 1.7215957641601562,
      "step": 97080,
      "train_bpb": 0.6066088777475446,
      "train_bytes": 28712.125,
      "train_loss_nats_per_token": 1.7213554662610446
    },
    {
      "epoch": 0.15867372468462773,
      "grad_norm": 0.2001953125,
      "learning_rate": 0.0008720826053825654,
      "loss": 1.716069221496582,
      "step": 97100,
      "train_bpb": 0.605365865632852,
      "train_bytes": 28448.24375,
      "train_loss_nats_per_token": 1.7166282863563818
    },
    {
      "epoch": 0.15880540412420005,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0008715162865638591,
      "loss": 1.7131099700927734,
      "step": 97120,
      "train_bpb": 0.6055197074311008,
      "train_bytes": 28483.6875,
      "train_loss_nats_per_token": 1.7144884847748862
    },
    {
      "epoch": 0.15893708356377234,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0008709500763844329,
      "loss": 1.7149831771850585,
      "step": 97140,
      "train_bpb": 0.6085789566606653,
      "train_bytes": 28200.11875,
      "train_loss_nats_per_token": 1.7161604078810815
    },
    {
      "epoch": 0.15906876300334466,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0008703839749421617,
      "loss": 1.7182350158691406,
      "step": 97160,
      "train_bpb": 0.6078313590106988,
      "train_bytes": 28178.69375,
      "train_loss_nats_per_token": 1.7188920785687671
    },
    {
      "epoch": 0.15920044244291695,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0008698179823349006,
      "loss": 1.709062385559082,
      "step": 97180,
      "train_bpb": 0.6059577123704843,
      "train_bytes": 28143.6,
      "train_loss_nats_per_token": 1.709836425553916
    },
    {
      "epoch": 0.15933212188248927,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0008692520986604869,
      "loss": 1.714494514465332,
      "step": 97200,
      "train_bpb": 0.6031134924454211,
      "train_bytes": 28488.7875,
      "train_loss_nats_per_token": 1.7143458373452691
    },
    {
      "epoch": 0.15946380132206156,
      "grad_norm": 0.140625,
      "learning_rate": 0.0008686863240167398,
      "loss": 1.7206430435180664,
      "step": 97220,
      "train_bpb": 0.6076559678718548,
      "train_bytes": 28225.26875,
      "train_loss_nats_per_token": 1.7223818850175112
    },
    {
      "epoch": 0.15959548076163388,
      "grad_norm": 0.125,
      "learning_rate": 0.0008681206585014571,
      "loss": 1.7276628494262696,
      "step": 97240,
      "train_bpb": 0.6048243631957532,
      "train_bytes": 28562.5125,
      "train_loss_nats_per_token": 1.7289076399125527
    },
    {
      "epoch": 0.1597271602012062,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0008675551022124206,
      "loss": 1.7185007095336915,
      "step": 97260,
      "train_bpb": 0.6068075067571587,
      "train_bytes": 28477.26875,
      "train_loss_nats_per_token": 1.7182789695527936
    },
    {
      "epoch": 0.1598588396407785,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0008669896552473916,
      "loss": 1.712815284729004,
      "step": 97280,
      "train_bpb": 0.6005482670425654,
      "train_bytes": 28508.4625,
      "train_loss_nats_per_token": 1.7129519191164697
    },
    {
      "epoch": 0.1599905190803508,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0008664243177041135,
      "loss": 1.7164598464965821,
      "step": 97300,
      "train_bpb": 0.6050641201685442,
      "train_bytes": 28485.7875,
      "train_loss_nats_per_token": 1.7172729569392775
    },
    {
      "epoch": 0.1601221985199231,
      "grad_norm": 0.1796875,
      "learning_rate": 0.000865859089680309,
      "loss": 1.7093982696533203,
      "step": 97320,
      "train_bpb": 0.605378493229492,
      "train_bytes": 28232.00625,
      "train_loss_nats_per_token": 1.7090425121480628
    },
    {
      "epoch": 0.16025387795949542,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0008652939712736836,
      "loss": 1.7183828353881836,
      "step": 97340,
      "train_bpb": 0.6086618942601485,
      "train_bytes": 28305.73125,
      "train_loss_nats_per_token": 1.7197025763257117
    },
    {
      "epoch": 0.1603855573990677,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0008647289625819242,
      "loss": 1.7179227828979493,
      "step": 97360,
      "train_bpb": 0.6059676420068923,
      "train_bytes": 28309.9125,
      "train_loss_nats_per_token": 1.7181467725496582
    },
    {
      "epoch": 0.16051723683864003,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0008641640637026962,
      "loss": 1.7142154693603515,
      "step": 97380,
      "train_bpb": 0.6068293114673411,
      "train_bytes": 28319.3,
      "train_loss_nats_per_token": 1.7143144866485434
    },
    {
      "epoch": 0.16064891627821232,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0008635992747336483,
      "loss": 1.7133317947387696,
      "step": 97400,
      "train_bpb": 0.6073147627810324,
      "train_bytes": 28616.14375,
      "train_loss_nats_per_token": 1.7138099392505834
    },
    {
      "epoch": 0.16078059571778464,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0008630345957724104,
      "loss": 1.7180124282836915,
      "step": 97420,
      "train_bpb": 0.607110802717712,
      "train_bytes": 28559.59375,
      "train_loss_nats_per_token": 1.7184115894817322
    },
    {
      "epoch": 0.16091227515735693,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.0008624700269165914,
      "loss": 1.7253122329711914,
      "step": 97440,
      "train_bpb": 0.6075854193226277,
      "train_bytes": 28439.74375,
      "train_loss_nats_per_token": 1.7257205164488465
    },
    {
      "epoch": 0.16104395459692925,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0008619055682637825,
      "loss": 1.719085693359375,
      "step": 97460,
      "train_bpb": 0.6099790475847431,
      "train_bytes": 28302.075,
      "train_loss_nats_per_token": 1.7198779892521807
    },
    {
      "epoch": 0.16117563403650154,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0008613412199115559,
      "loss": 1.7108814239501953,
      "step": 97480,
      "train_bpb": 0.6052774524330268,
      "train_bytes": 28393.26875,
      "train_loss_nats_per_token": 1.7120759206975484
    },
    {
      "epoch": 0.16130731347607385,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0008607769819574652,
      "loss": 1.7275371551513672,
      "step": 97500,
      "train_bpb": 0.6055734211460039,
      "train_bytes": 28502.91875,
      "train_loss_nats_per_token": 1.7288584032081205
    },
    {
      "epoch": 0.16143899291564615,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0008602128544990428,
      "loss": 1.7254179000854493,
      "step": 97520,
      "train_bpb": 0.6109427265713658,
      "train_bytes": 28243.95,
      "train_loss_nats_per_token": 1.7263600183602148
    },
    {
      "epoch": 0.16157067235521846,
      "grad_norm": 0.125,
      "learning_rate": 0.000859648837633804,
      "loss": 1.7186656951904298,
      "step": 97540,
      "train_bpb": 0.6032398198089199,
      "train_bytes": 28335.7125,
      "train_loss_nats_per_token": 1.7186419494544023
    },
    {
      "epoch": 0.16170235179479076,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0008590849314592448,
      "loss": 1.7144151687622071,
      "step": 97560,
      "train_bpb": 0.6058987004429919,
      "train_bytes": 28112.8125,
      "train_loss_nats_per_token": 1.7145826460520721
    },
    {
      "epoch": 0.16183403123436307,
      "grad_norm": 0.126953125,
      "learning_rate": 0.000858521136072841,
      "loss": 1.7218326568603515,
      "step": 97580,
      "train_bpb": 0.6065899456609052,
      "train_bytes": 28904.35625,
      "train_loss_nats_per_token": 1.7224192158746152
    },
    {
      "epoch": 0.16196571067393536,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0008579574515720502,
      "loss": 1.7106002807617187,
      "step": 97600,
      "train_bpb": 0.6073743519243663,
      "train_bytes": 28128.39375,
      "train_loss_nats_per_token": 1.7113961859878613
    },
    {
      "epoch": 0.16209739011350768,
      "grad_norm": 0.11083984375,
      "learning_rate": 0.0008573938780543103,
      "loss": 1.7176782608032226,
      "step": 97620,
      "train_bpb": 0.6081547906346477,
      "train_bytes": 28235.74375,
      "train_loss_nats_per_token": 1.7189271110438356
    },
    {
      "epoch": 0.16222906955307997,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0008568304156170413,
      "loss": 1.729477882385254,
      "step": 97640,
      "train_bpb": 0.6078741829931008,
      "train_bytes": 28599.9625,
      "train_loss_nats_per_token": 1.7296724890486916
    },
    {
      "epoch": 0.1623607489926523,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0008562670643576413,
      "loss": 1.719893455505371,
      "step": 97660,
      "train_bpb": 0.6082893848251069,
      "train_bytes": 28359.5625,
      "train_loss_nats_per_token": 1.7201737555984198
    },
    {
      "epoch": 0.16249242843222458,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0008557038243734915,
      "loss": 1.719468116760254,
      "step": 97680,
      "train_bpb": 0.6093339776202332,
      "train_bytes": 28189.7,
      "train_loss_nats_per_token": 1.7200679275291082
    },
    {
      "epoch": 0.1626241078717969,
      "grad_norm": 0.140625,
      "learning_rate": 0.0008551406957619541,
      "loss": 1.714750099182129,
      "step": 97700,
      "train_bpb": 0.6049638027921196,
      "train_bytes": 28590.125,
      "train_loss_nats_per_token": 1.7151751891452074
    },
    {
      "epoch": 0.1627557873113692,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0008545776786203699,
      "loss": 1.7083047866821288,
      "step": 97720,
      "train_bpb": 0.6039110102574048,
      "train_bytes": 28165.26875,
      "train_loss_nats_per_token": 1.7084498988338368
    },
    {
      "epoch": 0.1628874667509415,
      "grad_norm": 0.171875,
      "learning_rate": 0.000854014773046062,
      "loss": 1.7159622192382813,
      "step": 97740,
      "train_bpb": 0.6024185436535728,
      "train_bytes": 28903.1,
      "train_loss_nats_per_token": 1.7161810777329545
    },
    {
      "epoch": 0.1630191461905138,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0008534519791363345,
      "loss": 1.7127613067626952,
      "step": 97760,
      "train_bpb": 0.6034982592877165,
      "train_bytes": 28747.89375,
      "train_loss_nats_per_token": 1.7131581973689043
    },
    {
      "epoch": 0.16315082563008612,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0008528892969884706,
      "loss": 1.7211023330688477,
      "step": 97780,
      "train_bpb": 0.6101770802844396,
      "train_bytes": 28121.49375,
      "train_loss_nats_per_token": 1.7218016738728692
    },
    {
      "epoch": 0.1632825050696584,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0008523267266997358,
      "loss": 1.727752113342285,
      "step": 97800,
      "train_bpb": 0.6110869480679767,
      "train_bytes": 28174.73125,
      "train_loss_nats_per_token": 1.7278602744560423
    },
    {
      "epoch": 0.16341418450923073,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0008517642683673753,
      "loss": 1.7160152435302733,
      "step": 97820,
      "train_bpb": 0.6069225719298871,
      "train_bytes": 28374.80625,
      "train_loss_nats_per_token": 1.716572075346511
    },
    {
      "epoch": 0.16354586394880302,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0008512019220886162,
      "loss": 1.710062026977539,
      "step": 97840,
      "train_bpb": 0.6054451668792462,
      "train_bytes": 28085.35,
      "train_loss_nats_per_token": 1.7106442990689996
    },
    {
      "epoch": 0.16367754338837534,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0008506396879606638,
      "loss": 1.7125869750976563,
      "step": 97860,
      "train_bpb": 0.6054585653727305,
      "train_bytes": 28281.94375,
      "train_loss_nats_per_token": 1.7128850211978395
    },
    {
      "epoch": 0.16380922282794763,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0008500775660807064,
      "loss": 1.7142269134521484,
      "step": 97880,
      "train_bpb": 0.6058449924057004,
      "train_bytes": 28290.15625,
      "train_loss_nats_per_token": 1.715220093818438
    },
    {
      "epoch": 0.16394090226751995,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0008495155565459125,
      "loss": 1.7292591094970704,
      "step": 97900,
      "train_bpb": 0.6100066882648117,
      "train_bytes": 28348.34375,
      "train_loss_nats_per_token": 1.7289958645609826
    },
    {
      "epoch": 0.16407258170709224,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0008489536594534297,
      "loss": 1.7186697006225586,
      "step": 97920,
      "train_bpb": 0.605013457976008,
      "train_bytes": 28521.99375,
      "train_loss_nats_per_token": 1.7189750303595905
    },
    {
      "epoch": 0.16420426114666456,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0008483918749003874,
      "loss": 1.7016262054443358,
      "step": 97940,
      "train_bpb": 0.6016839415011968,
      "train_bytes": 28179.3125,
      "train_loss_nats_per_token": 1.7027716031611484
    },
    {
      "epoch": 0.16433594058623688,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0008478302029838955,
      "loss": 1.7287675857543945,
      "step": 97960,
      "train_bpb": 0.6090060524180191,
      "train_bytes": 28627.775,
      "train_loss_nats_per_token": 1.7296935820304289
    },
    {
      "epoch": 0.16446762002580917,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0008472686438010448,
      "loss": 1.721968650817871,
      "step": 97980,
      "train_bpb": 0.6119809389707593,
      "train_bytes": 28268.7625,
      "train_loss_nats_per_token": 1.7223805678943196
    },
    {
      "epoch": 0.1645992994653815,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0008467071974489049,
      "loss": 1.712474250793457,
      "step": 98000,
      "train_bpb": 0.6052010390536308,
      "train_bytes": 28070.2875,
      "train_loss_nats_per_token": 1.7126385936938964
    },
    {
      "epoch": 0.1645992994653815,
      "eval_loss": 1.5297398567199707,
      "eval_runtime": 74.5553,
      "eval_samples_per_second": 13.413,
      "eval_steps_per_second": 13.413,
      "step": 98000,
      "train_bpb": 0.5398403867554646,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5372563526274727
    },
    {
      "epoch": 0.16473097890495378,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0008461458640245277,
      "loss": 1.7190916061401367,
      "step": 98020,
      "train_bpb": 0.6104271158757798,
      "train_bytes": 28499.1375,
      "train_loss_nats_per_token": 1.7193516434281033
    },
    {
      "epoch": 0.1648626583445261,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0008455846436249453,
      "loss": 1.7086805343627929,
      "step": 98040,
      "train_bpb": 0.6063408966210563,
      "train_bytes": 28196.35625,
      "train_loss_nats_per_token": 1.7098122344233149
    },
    {
      "epoch": 0.1649943377840984,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0008450235363471688,
      "loss": 1.729271125793457,
      "step": 98060,
      "train_bpb": 0.6090577869327821,
      "train_bytes": 28476.31875,
      "train_loss_nats_per_token": 1.730559467912518
    },
    {
      "epoch": 0.1651260172236707,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0008444625422881915,
      "loss": 1.7039188385009765,
      "step": 98080,
      "train_bpb": 0.5980624507249618,
      "train_bytes": 28443.44375,
      "train_loss_nats_per_token": 1.7044266836625392
    },
    {
      "epoch": 0.165257696663243,
      "grad_norm": 0.126953125,
      "learning_rate": 0.000843901661544987,
      "loss": 1.7096662521362305,
      "step": 98100,
      "train_bpb": 0.6037787196127933,
      "train_bytes": 28539.2875,
      "train_loss_nats_per_token": 1.7109514830304329
    },
    {
      "epoch": 0.16538937610281532,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0008433408942145076,
      "loss": 1.7116558074951171,
      "step": 98120,
      "train_bpb": 0.6039160551058638,
      "train_bytes": 28529.54375,
      "train_loss_nats_per_token": 1.7116145322589837
    },
    {
      "epoch": 0.1655210555423876,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0008427802403936879,
      "loss": 1.7148588180541993,
      "step": 98140,
      "train_bpb": 0.6046702581802893,
      "train_bytes": 28461.75625,
      "train_loss_nats_per_token": 1.7157061322467473
    },
    {
      "epoch": 0.16565273498195993,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0008422197001794419,
      "loss": 1.7082500457763672,
      "step": 98160,
      "train_bpb": 0.6081443888780335,
      "train_bytes": 28111.48125,
      "train_loss_nats_per_token": 1.708606415646093
    },
    {
      "epoch": 0.16578441442153222,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0008416592736686649,
      "loss": 1.7296689987182616,
      "step": 98180,
      "train_bpb": 0.6121502645207342,
      "train_bytes": 28290.55,
      "train_loss_nats_per_token": 1.7301523148110614
    },
    {
      "epoch": 0.16591609386110454,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0008410989609582307,
      "loss": 1.7085638046264648,
      "step": 98200,
      "train_bpb": 0.6009996012724732,
      "train_bytes": 28638.25625,
      "train_loss_nats_per_token": 1.7082587135792815
    },
    {
      "epoch": 0.16604777330067683,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0008405387621449952,
      "loss": 1.7172508239746094,
      "step": 98220,
      "train_bpb": 0.6064729146303544,
      "train_bytes": 28325.45,
      "train_loss_nats_per_token": 1.717170468368065
    },
    {
      "epoch": 0.16617945274024915,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0008399786773257946,
      "loss": 1.7191389083862305,
      "step": 98240,
      "train_bpb": 0.6062303989861966,
      "train_bytes": 28793.3375,
      "train_loss_nats_per_token": 1.719450250156713
    },
    {
      "epoch": 0.16631113217982144,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0008394187065974436,
      "loss": 1.7204935073852539,
      "step": 98260,
      "train_bpb": 0.6088064192412769,
      "train_bytes": 28030.24375,
      "train_loss_nats_per_token": 1.7211064255090154
    },
    {
      "epoch": 0.16644281161939375,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0008388588500567387,
      "loss": 1.7084901809692383,
      "step": 98280,
      "train_bpb": 0.6084975950430904,
      "train_bytes": 27850.30625,
      "train_loss_nats_per_token": 1.7093024185269714
    },
    {
      "epoch": 0.16657449105896605,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0008382991078004574,
      "loss": 1.7106107711791991,
      "step": 98300,
      "train_bpb": 0.6019544676799934,
      "train_bytes": 28701.59375,
      "train_loss_nats_per_token": 1.711144393432094
    },
    {
      "epoch": 0.16670617049853836,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0008377394799253549,
      "loss": 1.713780975341797,
      "step": 98320,
      "train_bpb": 0.6035670569783621,
      "train_bytes": 28649.4875,
      "train_loss_nats_per_token": 1.714452424117823
    },
    {
      "epoch": 0.16683784993811066,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0008371799665281687,
      "loss": 1.7124006271362304,
      "step": 98340,
      "train_bpb": 0.6043942447667431,
      "train_bytes": 28311.2875,
      "train_loss_nats_per_token": 1.7131267875016363
    },
    {
      "epoch": 0.16696952937768297,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.000836620567705616,
      "loss": 1.7109600067138673,
      "step": 98360,
      "train_bpb": 0.6034976574511436,
      "train_bytes": 28286.49375,
      "train_loss_nats_per_token": 1.7111789977088023
    },
    {
      "epoch": 0.16710120881725526,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0008360612835543945,
      "loss": 1.7092672348022462,
      "step": 98380,
      "train_bpb": 0.604774791477713,
      "train_bytes": 28070.475,
      "train_loss_nats_per_token": 1.709566531610457
    },
    {
      "epoch": 0.16723288825682758,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0008355021141711809,
      "loss": 1.7256145477294922,
      "step": 98400,
      "train_bpb": 0.6093451163859357,
      "train_bytes": 28110.5375,
      "train_loss_nats_per_token": 1.7259032978432784
    },
    {
      "epoch": 0.16736456769639987,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0008349430596526332,
      "loss": 1.6995357513427733,
      "step": 98420,
      "train_bpb": 0.6034598918984085,
      "train_bytes": 28041.26875,
      "train_loss_nats_per_token": 1.7000010574771023
    },
    {
      "epoch": 0.1674962471359722,
      "grad_norm": 0.193359375,
      "learning_rate": 0.0008343841200953897,
      "loss": 1.7269229888916016,
      "step": 98440,
      "train_bpb": 0.6127785937802478,
      "train_bytes": 28048.58125,
      "train_loss_nats_per_token": 1.7277718169668634
    },
    {
      "epoch": 0.16762792657554448,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0008338252955960672,
      "loss": 1.712315559387207,
      "step": 98460,
      "train_bpb": 0.6079628024282007,
      "train_bytes": 28285.675,
      "train_loss_nats_per_token": 1.712605903668746
    },
    {
      "epoch": 0.1677596060151168,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0008332665862512643,
      "loss": 1.7140884399414062,
      "step": 98480,
      "train_bpb": 0.6072736135331545,
      "train_bytes": 28154.2125,
      "train_loss_nats_per_token": 1.71452736784942
    },
    {
      "epoch": 0.1678912854546891,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0008327079921575594,
      "loss": 1.6971839904785155,
      "step": 98500,
      "train_bpb": 0.6014676693873197,
      "train_bytes": 27996.78125,
      "train_loss_nats_per_token": 1.6973017705106908
    },
    {
      "epoch": 0.1680229648942614,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0008321495134115108,
      "loss": 1.720326042175293,
      "step": 98520,
      "train_bpb": 0.6064183704207896,
      "train_bytes": 28440.2,
      "train_loss_nats_per_token": 1.720628272409161
    },
    {
      "epoch": 0.1681546443338337,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0008315911501096562,
      "loss": 1.7241512298583985,
      "step": 98540,
      "train_bpb": 0.611213144455298,
      "train_bytes": 28206.94375,
      "train_loss_nats_per_token": 1.7256348928056753
    },
    {
      "epoch": 0.16828632377340602,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0008310329023485139,
      "loss": 1.7302265167236328,
      "step": 98560,
      "train_bpb": 0.6118881640924947,
      "train_bytes": 28624.1625,
      "train_loss_nats_per_token": 1.7306066602594214
    },
    {
      "epoch": 0.1684180032129783,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0008304747702245833,
      "loss": 1.7084726333618163,
      "step": 98580,
      "train_bpb": 0.604419066020422,
      "train_bytes": 28400.1875,
      "train_loss_nats_per_token": 1.7084899154929776
    },
    {
      "epoch": 0.16854968265255063,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0008299167538343414,
      "loss": 1.7023855209350587,
      "step": 98600,
      "train_bpb": 0.6053132862010078,
      "train_bytes": 27824.5375,
      "train_loss_nats_per_token": 1.703512656033544
    },
    {
      "epoch": 0.16868136209212295,
      "grad_norm": 0.130859375,
      "learning_rate": 0.000829358853274247,
      "loss": 1.7268369674682618,
      "step": 98620,
      "train_bpb": 0.6137233236895598,
      "train_bytes": 28086.26875,
      "train_loss_nats_per_token": 1.7275244413271749
    },
    {
      "epoch": 0.16881304153169524,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0008288010686407392,
      "loss": 1.7012250900268555,
      "step": 98640,
      "train_bpb": 0.603841259495656,
      "train_bytes": 28009.90625,
      "train_loss_nats_per_token": 1.7016484099588054
    },
    {
      "epoch": 0.16894472097126756,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0008282434000302351,
      "loss": 1.7132694244384765,
      "step": 98660,
      "train_bpb": 0.6022827731910395,
      "train_bytes": 28388.0625,
      "train_loss_nats_per_token": 1.7138569645428656
    },
    {
      "epoch": 0.16907640041083985,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0008276858475391333,
      "loss": 1.715043830871582,
      "step": 98680,
      "train_bpb": 0.6058273285037009,
      "train_bytes": 28544.06875,
      "train_loss_nats_per_token": 1.7161639091177037
    },
    {
      "epoch": 0.16920807985041217,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0008271284112638121,
      "loss": 1.7066448211669922,
      "step": 98700,
      "train_bpb": 0.6088710556534132,
      "train_bytes": 28045.34375,
      "train_loss_nats_per_token": 1.7066987603481278
    },
    {
      "epoch": 0.16933975928998446,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0008265710913006303,
      "loss": 1.7165369033813476,
      "step": 98720,
      "train_bpb": 0.6065299813050671,
      "train_bytes": 28157.50625,
      "train_loss_nats_per_token": 1.7175793453212211
    },
    {
      "epoch": 0.16947143872955678,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0008260138877459242,
      "loss": 1.7040063858032226,
      "step": 98740,
      "train_bpb": 0.6014568133039462,
      "train_bytes": 28331.65625,
      "train_loss_nats_per_token": 1.7043533189302982
    },
    {
      "epoch": 0.16960311816912907,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0008254568006960129,
      "loss": 1.705516242980957,
      "step": 98760,
      "train_bpb": 0.6075804376057444,
      "train_bytes": 28441.8375,
      "train_loss_nats_per_token": 1.7060303830370838
    },
    {
      "epoch": 0.1697347976087014,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0008248998302471943,
      "loss": 1.7089920043945312,
      "step": 98780,
      "train_bpb": 0.6092485795245847,
      "train_bytes": 27916.95625,
      "train_loss_nats_per_token": 1.7095606063161493
    },
    {
      "epoch": 0.16986647704827368,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0008243429764957447,
      "loss": 1.703347396850586,
      "step": 98800,
      "train_bpb": 0.6040586417847622,
      "train_bytes": 28189.59375,
      "train_loss_nats_per_token": 1.7042558746725247
    },
    {
      "epoch": 0.169998156487846,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0008237862395379222,
      "loss": 1.7211017608642578,
      "step": 98820,
      "train_bpb": 0.6076486490437619,
      "train_bytes": 28254.4,
      "train_loss_nats_per_token": 1.7213938781106368
    },
    {
      "epoch": 0.1701298359274183,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0008232296194699647,
      "loss": 1.7207374572753906,
      "step": 98840,
      "train_bpb": 0.6107331764057847,
      "train_bytes": 28127.63125,
      "train_loss_nats_per_token": 1.7205898914532773
    },
    {
      "epoch": 0.1702615153669906,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0008226731163880878,
      "loss": 1.7189855575561523,
      "step": 98860,
      "train_bpb": 0.6041564987680224,
      "train_bytes": 28546.78125,
      "train_loss_nats_per_token": 1.7195533664777962
    },
    {
      "epoch": 0.1703931948065629,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0008221167303884889,
      "loss": 1.7200967788696289,
      "step": 98880,
      "train_bpb": 0.6057926668848268,
      "train_bytes": 28439.3625,
      "train_loss_nats_per_token": 1.7205048012202193
    },
    {
      "epoch": 0.17052487424613522,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0008215604615673444,
      "loss": 1.7122968673706054,
      "step": 98900,
      "train_bpb": 0.6078475476100818,
      "train_bytes": 28235.575,
      "train_loss_nats_per_token": 1.7130680347228098
    },
    {
      "epoch": 0.1706565536857075,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0008210043100208114,
      "loss": 1.7143131256103517,
      "step": 98920,
      "train_bpb": 0.6067811330986341,
      "train_bytes": 28102.7875,
      "train_loss_nats_per_token": 1.7150462601366303
    },
    {
      "epoch": 0.17078823312527983,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0008204482758450244,
      "loss": 1.7140438079833984,
      "step": 98940,
      "train_bpb": 0.6030771184109527,
      "train_bytes": 28810.89375,
      "train_loss_nats_per_token": 1.7142933704776697
    },
    {
      "epoch": 0.17091991256485212,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0008198923591361,
      "loss": 1.7017852783203125,
      "step": 98960,
      "train_bpb": 0.6011919959965029,
      "train_bytes": 28232.4625,
      "train_loss_nats_per_token": 1.7034207279882794
    },
    {
      "epoch": 0.17105159200442444,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0008193365599901342,
      "loss": 1.7145067214965821,
      "step": 98980,
      "train_bpb": 0.606152321462534,
      "train_bytes": 28244.34375,
      "train_loss_nats_per_token": 1.7137983113444877
    },
    {
      "epoch": 0.17118327144399673,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0008187808785032005,
      "loss": 1.7050457000732422,
      "step": 99000,
      "train_bpb": 0.6023341794205778,
      "train_bytes": 28331.71875,
      "train_loss_nats_per_token": 1.7056127088392297
    },
    {
      "epoch": 0.17118327144399673,
      "eval_loss": 1.530900478363037,
      "eval_runtime": 75.4804,
      "eval_samples_per_second": 13.248,
      "eval_steps_per_second": 13.248,
      "step": 99000,
      "train_bpb": 0.5402451347589203,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5384089181909038
    },
    {
      "epoch": 0.17131495088356904,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0008182253147713546,
      "loss": 1.7091960906982422,
      "step": 99020,
      "train_bpb": 0.6057592577314406,
      "train_bytes": 28354.41875,
      "train_loss_nats_per_token": 1.7088792187802775
    },
    {
      "epoch": 0.17144663032314134,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0008176698688906307,
      "loss": 1.7110626220703125,
      "step": 99040,
      "train_bpb": 0.6044245212759899,
      "train_bytes": 28589.55625,
      "train_loss_nats_per_token": 1.7108065955226808
    },
    {
      "epoch": 0.17157830976271365,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0008171145409570435,
      "loss": 1.701681137084961,
      "step": 99060,
      "train_bpb": 0.6014767325381046,
      "train_bytes": 28239.3625,
      "train_loss_nats_per_token": 1.7020318046549936
    },
    {
      "epoch": 0.17170998920228595,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0008165593310665854,
      "loss": 1.7253955841064452,
      "step": 99080,
      "train_bpb": 0.6095285829526143,
      "train_bytes": 28208.4875,
      "train_loss_nats_per_token": 1.7261561384299353
    },
    {
      "epoch": 0.17184166864185826,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0008160042393152301,
      "loss": 1.715919303894043,
      "step": 99100,
      "train_bpb": 0.6063656594684945,
      "train_bytes": 28473.68125,
      "train_loss_nats_per_token": 1.7165590893220612
    },
    {
      "epoch": 0.17197334808143055,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0008154492657989311,
      "loss": 1.7088752746582032,
      "step": 99120,
      "train_bpb": 0.6053602814436779,
      "train_bytes": 28028.6375,
      "train_loss_nats_per_token": 1.7100126627106906
    },
    {
      "epoch": 0.17210502752100287,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0008148944106136196,
      "loss": 1.7083175659179688,
      "step": 99140,
      "train_bpb": 0.6057815799842475,
      "train_bytes": 28451.3875,
      "train_loss_nats_per_token": 1.7084780159511843
    },
    {
      "epoch": 0.17223670696057516,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.000814339673855208,
      "loss": 1.7116609573364259,
      "step": 99160,
      "train_bpb": 0.6052672812300031,
      "train_bytes": 28483.33125,
      "train_loss_nats_per_token": 1.7135081211833276
    },
    {
      "epoch": 0.17236838640014748,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0008137850556195885,
      "loss": 1.7148239135742187,
      "step": 99180,
      "train_bpb": 0.6044595967613621,
      "train_bytes": 28304.44375,
      "train_loss_nats_per_token": 1.7149879229074305
    },
    {
      "epoch": 0.17250006583971977,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0008132305560026305,
      "loss": 1.7144504547119142,
      "step": 99200,
      "train_bpb": 0.6034975197712241,
      "train_bytes": 28230.65625,
      "train_loss_nats_per_token": 1.715142622119932
    },
    {
      "epoch": 0.1726317452792921,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.0008126761751001855,
      "loss": 1.7072307586669921,
      "step": 99220,
      "train_bpb": 0.6079550013784721,
      "train_bytes": 27998.48125,
      "train_loss_nats_per_token": 1.7087491107654298
    },
    {
      "epoch": 0.17276342471886438,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0008121219130080829,
      "loss": 1.7078535079956054,
      "step": 99240,
      "train_bpb": 0.6089162570926345,
      "train_bytes": 28098.04375,
      "train_loss_nats_per_token": 1.7089038800139977
    },
    {
      "epoch": 0.1728951041584367,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0008115677698221334,
      "loss": 1.7227144241333008,
      "step": 99260,
      "train_bpb": 0.6073538015374224,
      "train_bytes": 28174.575,
      "train_loss_nats_per_token": 1.7230689397627048
    },
    {
      "epoch": 0.17302678359800902,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0008110137456381237,
      "loss": 1.7179483413696288,
      "step": 99280,
      "train_bpb": 0.6084423846582191,
      "train_bytes": 28295.1,
      "train_loss_nats_per_token": 1.7184877545053507
    },
    {
      "epoch": 0.1731584630375813,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0008104598405518233,
      "loss": 1.7095626831054687,
      "step": 99300,
      "train_bpb": 0.6044836680194127,
      "train_bytes": 28443.03125,
      "train_loss_nats_per_token": 1.7099641874158718
    },
    {
      "epoch": 0.17329014247715363,
      "grad_norm": 0.1806640625,
      "learning_rate": 0.0008099060546589802,
      "loss": 1.7113285064697266,
      "step": 99320,
      "train_bpb": 0.6035274829464526,
      "train_bytes": 28616.7875,
      "train_loss_nats_per_token": 1.712078707344503
    },
    {
      "epoch": 0.17342182191672592,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0008093523880553207,
      "loss": 1.7171113967895508,
      "step": 99340,
      "train_bpb": 0.6043304217629453,
      "train_bytes": 28411.20625,
      "train_loss_nats_per_token": 1.7167828828516625
    },
    {
      "epoch": 0.17355350135629824,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0008087988408365511,
      "loss": 1.7121242523193358,
      "step": 99360,
      "train_bpb": 0.6054240963351649,
      "train_bytes": 28220.5625,
      "train_loss_nats_per_token": 1.7126540100281031
    },
    {
      "epoch": 0.17368518079587053,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0008082454130983584,
      "loss": 1.7110103607177733,
      "step": 99380,
      "train_bpb": 0.6036020518098235,
      "train_bytes": 28284.33125,
      "train_loss_nats_per_token": 1.7113952048626426
    },
    {
      "epoch": 0.17381686023544285,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0008076921049364064,
      "loss": 1.7133922576904297,
      "step": 99400,
      "train_bpb": 0.6047075308923741,
      "train_bytes": 28072.025,
      "train_loss_nats_per_token": 1.7142744617974457
    },
    {
      "epoch": 0.17394853967501514,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0008071389164463401,
      "loss": 1.7145389556884765,
      "step": 99420,
      "train_bpb": 0.6044095239768192,
      "train_bytes": 28584.875,
      "train_loss_nats_per_token": 1.715571643506368
    },
    {
      "epoch": 0.17408021911458746,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0008065858477237834,
      "loss": 1.7208654403686523,
      "step": 99440,
      "train_bpb": 0.6056168393202078,
      "train_bytes": 28627.01875,
      "train_loss_nats_per_token": 1.7218227316069044
    },
    {
      "epoch": 0.17421189855415975,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0008060328988643398,
      "loss": 1.7255056381225586,
      "step": 99460,
      "train_bpb": 0.6104983007859154,
      "train_bytes": 28183.60625,
      "train_loss_nats_per_token": 1.7255060983044688
    },
    {
      "epoch": 0.17434357799373207,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0008054800699635907,
      "loss": 1.7176660537719726,
      "step": 99480,
      "train_bpb": 0.6057379174273597,
      "train_bytes": 28487.2875,
      "train_loss_nats_per_token": 1.7193268071791916
    },
    {
      "epoch": 0.17447525743330436,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0008049273611170983,
      "loss": 1.7256902694702148,
      "step": 99500,
      "train_bpb": 0.6079764889756745,
      "train_bytes": 28508.34375,
      "train_loss_nats_per_token": 1.7268414937940475
    },
    {
      "epoch": 0.17460693687287668,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0008043747724204041,
      "loss": 1.7310585021972655,
      "step": 99520,
      "train_bpb": 0.6099261337118278,
      "train_bytes": 28358.91875,
      "train_loss_nats_per_token": 1.7313354320536487
    },
    {
      "epoch": 0.17473861631244897,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0008038223039690268,
      "loss": 1.7150506973266602,
      "step": 99540,
      "train_bpb": 0.6073338987465435,
      "train_bytes": 28104.8375,
      "train_loss_nats_per_token": 1.7153307065557193
    },
    {
      "epoch": 0.1748702957520213,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0008032699558584665,
      "loss": 1.7215845108032226,
      "step": 99560,
      "train_bpb": 0.6089029973440017,
      "train_bytes": 28332.9,
      "train_loss_nats_per_token": 1.7222210705935017
    },
    {
      "epoch": 0.17500197519159358,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0008027177281842018,
      "loss": 1.7115571975708008,
      "step": 99580,
      "train_bpb": 0.6043789884582225,
      "train_bytes": 28320.9125,
      "train_loss_nats_per_token": 1.7118949122471212
    },
    {
      "epoch": 0.1751336546311659,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0008021656210416908,
      "loss": 1.7133068084716796,
      "step": 99600,
      "train_bpb": 0.6057264148683396,
      "train_bytes": 28391.09375,
      "train_loss_nats_per_token": 1.7133654287959923
    },
    {
      "epoch": 0.1752653340707382,
      "grad_norm": 0.2177734375,
      "learning_rate": 0.0008016136345263696,
      "loss": 1.7150516510009766,
      "step": 99620,
      "train_bpb": 0.6086218439053479,
      "train_bytes": 28013.06875,
      "train_loss_nats_per_token": 1.7151256672514643
    },
    {
      "epoch": 0.1753970135103105,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0008010617687336544,
      "loss": 1.7149160385131836,
      "step": 99640,
      "train_bpb": 0.6047765429605888,
      "train_bytes": 28347.9375,
      "train_loss_nats_per_token": 1.715084046404025
    },
    {
      "epoch": 0.1755286929498828,
      "grad_norm": 0.13671875,
      "learning_rate": 0.000800510023758941,
      "loss": 1.7164749145507812,
      "step": 99660,
      "train_bpb": 0.6088968079244084,
      "train_bytes": 28263.03125,
      "train_loss_nats_per_token": 1.7168223272377927
    },
    {
      "epoch": 0.17566037238945512,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0007999583996976025,
      "loss": 1.7179901123046875,
      "step": 99680,
      "train_bpb": 0.6058905089790146,
      "train_bytes": 28426.04375,
      "train_loss_nats_per_token": 1.717828237494739
    },
    {
      "epoch": 0.1757920518290274,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.000799406896644993,
      "loss": 1.7204532623291016,
      "step": 99700,
      "train_bpb": 0.6050513480757647,
      "train_bytes": 28666.13125,
      "train_loss_nats_per_token": 1.7213074662352394
    },
    {
      "epoch": 0.17592373126859973,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0007988555146964454,
      "loss": 1.7150892257690429,
      "step": 99720,
      "train_bpb": 0.6094905325916885,
      "train_bytes": 28097.325,
      "train_loss_nats_per_token": 1.7162464525218954
    },
    {
      "epoch": 0.17605541070817202,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0007983042539472704,
      "loss": 1.7249879837036133,
      "step": 99740,
      "train_bpb": 0.6064728862925047,
      "train_bytes": 28770.63125,
      "train_loss_nats_per_token": 1.7257712905348732
    },
    {
      "epoch": 0.17618709014774434,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0007977531144927586,
      "loss": 1.708460807800293,
      "step": 99760,
      "train_bpb": 0.604793767403788,
      "train_bytes": 28356.725,
      "train_loss_nats_per_token": 1.7087761509619652
    },
    {
      "epoch": 0.17631876958731663,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00079720209642818,
      "loss": 1.7178485870361329,
      "step": 99780,
      "train_bpb": 0.6075798956016781,
      "train_bytes": 28326.26875,
      "train_loss_nats_per_token": 1.7182869522359443
    },
    {
      "epoch": 0.17645044902688894,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0007966511998487837,
      "loss": 1.707650375366211,
      "step": 99800,
      "train_bpb": 0.6039461621522236,
      "train_bytes": 28442.3875,
      "train_loss_nats_per_token": 1.7081767785422843
    },
    {
      "epoch": 0.17658212846646124,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0007961004248497961,
      "loss": 1.7125701904296875,
      "step": 99820,
      "train_bpb": 0.6028096167930157,
      "train_bytes": 28808.8375,
      "train_loss_nats_per_token": 1.7142860781182827
    },
    {
      "epoch": 0.17671380790603355,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0007955497715264244,
      "loss": 1.7273519515991211,
      "step": 99840,
      "train_bpb": 0.6117953790529017,
      "train_bytes": 28551.53125,
      "train_loss_nats_per_token": 1.728363533898167
    },
    {
      "epoch": 0.17684548734560585,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0007949992399738547,
      "loss": 1.713421630859375,
      "step": 99860,
      "train_bpb": 0.6049298910466941,
      "train_bytes": 28268.59375,
      "train_loss_nats_per_token": 1.7142723394966803
    },
    {
      "epoch": 0.17697716678517816,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0007944488302872506,
      "loss": 1.7085237503051758,
      "step": 99880,
      "train_bpb": 0.6022646480293598,
      "train_bytes": 28393.6,
      "train_loss_nats_per_token": 1.7085663994650657
    },
    {
      "epoch": 0.17710884622475045,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0007938985425617557,
      "loss": 1.711512565612793,
      "step": 99900,
      "train_bpb": 0.6045365318417923,
      "train_bytes": 28198.03125,
      "train_loss_nats_per_token": 1.7118646665438977
    },
    {
      "epoch": 0.17724052566432277,
      "grad_norm": 0.26953125,
      "learning_rate": 0.000793348376892493,
      "loss": 1.702057456970215,
      "step": 99920,
      "train_bpb": 0.6059675497360395,
      "train_bytes": 27836.68125,
      "train_loss_nats_per_token": 1.7033098463389584
    },
    {
      "epoch": 0.17737220510389506,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007927983333745638,
      "loss": 1.7062755584716798,
      "step": 99940,
      "train_bpb": 0.6049542504042826,
      "train_bytes": 28091.54375,
      "train_loss_nats_per_token": 1.7070497813647456
    },
    {
      "epoch": 0.17750388454346738,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0007922484121030473,
      "loss": 1.708414649963379,
      "step": 99960,
      "train_bpb": 0.603045859818299,
      "train_bytes": 28237.4625,
      "train_loss_nats_per_token": 1.7095268161159607
    },
    {
      "epoch": 0.1776355639830397,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0007916986131730032,
      "loss": 1.7200492858886718,
      "step": 99980,
      "train_bpb": 0.6085166724102589,
      "train_bytes": 28435.0375,
      "train_loss_nats_per_token": 1.7211559958523552
    },
    {
      "epoch": 0.177767243422612,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0007911489366794695,
      "loss": 1.7172561645507813,
      "step": 100000,
      "train_bpb": 0.60895531518613,
      "train_bytes": 28061.09375,
      "train_loss_nats_per_token": 1.7185428553433704
    },
    {
      "epoch": 0.177767243422612,
      "eval_loss": 1.5313209295272827,
      "eval_runtime": 75.2915,
      "eval_samples_per_second": 13.282,
      "eval_steps_per_second": 13.282,
      "step": 100000,
      "train_bpb": 0.540398401627917,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5388453628767704
    },
    {
      "epoch": 0.1778989228621843,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0007905993827174624,
      "loss": 1.7167840957641602,
      "step": 100020,
      "train_bpb": 0.6087509776466988,
      "train_bytes": 28538.15625,
      "train_loss_nats_per_token": 1.7178706152883163
    },
    {
      "epoch": 0.1780306023017566,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0007900499513819777,
      "loss": 1.7204923629760742,
      "step": 100040,
      "train_bpb": 0.6076193010405541,
      "train_bytes": 28356.125,
      "train_loss_nats_per_token": 1.7207347432113693
    },
    {
      "epoch": 0.17816228174132892,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0007895006427679905,
      "loss": 1.7271757125854492,
      "step": 100060,
      "train_bpb": 0.6134097617666999,
      "train_bytes": 28193.54375,
      "train_loss_nats_per_token": 1.7282079014240461
    },
    {
      "epoch": 0.1782939611809012,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0007889514569704523,
      "loss": 1.7135433197021483,
      "step": 100080,
      "train_bpb": 0.6067895321651354,
      "train_bytes": 28406.60625,
      "train_loss_nats_per_token": 1.7138385167808292
    },
    {
      "epoch": 0.17842564062047353,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007884023940842961,
      "loss": 1.7119754791259765,
      "step": 100100,
      "train_bpb": 0.6009350016588786,
      "train_bytes": 28632.43125,
      "train_loss_nats_per_token": 1.7129320371794252
    },
    {
      "epoch": 0.17855732006004582,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007878534542044323,
      "loss": 1.717801856994629,
      "step": 100120,
      "train_bpb": 0.6048367710659845,
      "train_bytes": 28445.3625,
      "train_loss_nats_per_token": 1.718033262398536
    },
    {
      "epoch": 0.17868899949961814,
      "grad_norm": 0.30078125,
      "learning_rate": 0.000787304637425751,
      "loss": 1.7179838180541993,
      "step": 100140,
      "train_bpb": 0.6059939383748452,
      "train_bytes": 28187.8875,
      "train_loss_nats_per_token": 1.7181996522258747
    },
    {
      "epoch": 0.17882067893919043,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0007867559438431189,
      "loss": 1.724494743347168,
      "step": 100160,
      "train_bpb": 0.608252595621462,
      "train_bytes": 28581.2375,
      "train_loss_nats_per_token": 1.7253734896598656
    },
    {
      "epoch": 0.17895235837876275,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0007862073735513835,
      "loss": 1.7229093551635741,
      "step": 100180,
      "train_bpb": 0.6088777568697064,
      "train_bytes": 28148.71875,
      "train_loss_nats_per_token": 1.7234177679157536
    },
    {
      "epoch": 0.17908403781833504,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.000785658926645371,
      "loss": 1.7163408279418946,
      "step": 100200,
      "train_bpb": 0.6050391539410881,
      "train_bytes": 28617.71875,
      "train_loss_nats_per_token": 1.7173085452271082
    },
    {
      "epoch": 0.17921571725790736,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0007851106032198843,
      "loss": 1.718695831298828,
      "step": 100220,
      "train_bpb": 0.6087430105232009,
      "train_bytes": 28359.5375,
      "train_loss_nats_per_token": 1.7191858757712688
    },
    {
      "epoch": 0.17934739669747965,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0007845624033697068,
      "loss": 1.702712059020996,
      "step": 100240,
      "train_bpb": 0.6041480538244751,
      "train_bytes": 28200.36875,
      "train_loss_nats_per_token": 1.7038511294115493
    },
    {
      "epoch": 0.17947907613705197,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0007840143271896006,
      "loss": 1.71789608001709,
      "step": 100260,
      "train_bpb": 0.6087132299269226,
      "train_bytes": 28388.7625,
      "train_loss_nats_per_token": 1.7184892620779058
    },
    {
      "epoch": 0.17961075557662426,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0007834663747743045,
      "loss": 1.7133708953857423,
      "step": 100280,
      "train_bpb": 0.6047421858381642,
      "train_bytes": 28527.3125,
      "train_loss_nats_per_token": 1.7146405335893726
    },
    {
      "epoch": 0.17974243501619658,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.000782918546218538,
      "loss": 1.701887321472168,
      "step": 100300,
      "train_bpb": 0.604405530369458,
      "train_bytes": 27955.3875,
      "train_loss_nats_per_token": 1.7010606533848491
    },
    {
      "epoch": 0.17987411445576887,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0007823708416169983,
      "loss": 1.701384925842285,
      "step": 100320,
      "train_bpb": 0.6000763057623231,
      "train_bytes": 28339.625,
      "train_loss_nats_per_token": 1.7026163864443613
    },
    {
      "epoch": 0.1800057938953412,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0007818232610643618,
      "loss": 1.708353042602539,
      "step": 100340,
      "train_bpb": 0.6036893684604004,
      "train_bytes": 28280.74375,
      "train_loss_nats_per_token": 1.709009715636542
    },
    {
      "epoch": 0.18013747333491348,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0007812758046552816,
      "loss": 1.707010269165039,
      "step": 100360,
      "train_bpb": 0.6072503466349823,
      "train_bytes": 28058.15,
      "train_loss_nats_per_token": 1.7075913124246562
    },
    {
      "epoch": 0.1802691527744858,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0007807284724843915,
      "loss": 1.7193119049072265,
      "step": 100380,
      "train_bpb": 0.6053270411526699,
      "train_bytes": 28496.50625,
      "train_loss_nats_per_token": 1.7198615445028678
    },
    {
      "epoch": 0.1804008322140581,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0007801812646463034,
      "loss": 1.715644645690918,
      "step": 100400,
      "train_bpb": 0.6077208034391661,
      "train_bytes": 28307.5125,
      "train_loss_nats_per_token": 1.7164410557100531
    },
    {
      "epoch": 0.1805325116536304,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0007796341812356062,
      "loss": 1.721978759765625,
      "step": 100420,
      "train_bpb": 0.6085072393112133,
      "train_bytes": 28341.8,
      "train_loss_nats_per_token": 1.7233036741884498
    },
    {
      "epoch": 0.1806641910932027,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0007790872223468689,
      "loss": 1.7172998428344726,
      "step": 100440,
      "train_bpb": 0.6048863629085223,
      "train_bytes": 28359.24375,
      "train_loss_nats_per_token": 1.7169252554003829
    },
    {
      "epoch": 0.18079587053277502,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0007785403880746384,
      "loss": 1.7093582153320312,
      "step": 100460,
      "train_bpb": 0.6068713381440044,
      "train_bytes": 28050.05,
      "train_loss_nats_per_token": 1.7097812631886633
    },
    {
      "epoch": 0.1809275499723473,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0007779936785134407,
      "loss": 1.7161571502685546,
      "step": 100480,
      "train_bpb": 0.6071857432589549,
      "train_bytes": 28414.74375,
      "train_loss_nats_per_token": 1.7167740710615094
    },
    {
      "epoch": 0.18105922941191963,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0007774470937577788,
      "loss": 1.7126033782958985,
      "step": 100500,
      "train_bpb": 0.6069997329000859,
      "train_bytes": 28247.74375,
      "train_loss_nats_per_token": 1.7132770465920957
    },
    {
      "epoch": 0.18119090885149192,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0007769006339021352,
      "loss": 1.7201522827148437,
      "step": 100520,
      "train_bpb": 0.6049429640319315,
      "train_bytes": 28647.775,
      "train_loss_nats_per_token": 1.72073628393931
    },
    {
      "epoch": 0.18132258829106424,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0007763542990409713,
      "loss": 1.7055246353149414,
      "step": 100540,
      "train_bpb": 0.6026149577503892,
      "train_bytes": 28239.6,
      "train_loss_nats_per_token": 1.7058064886953692
    },
    {
      "epoch": 0.18145426773063653,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.000775808089268725,
      "loss": 1.7298852920532226,
      "step": 100560,
      "train_bpb": 0.609495375818016,
      "train_bytes": 28056.28125,
      "train_loss_nats_per_token": 1.7309433021446947
    },
    {
      "epoch": 0.18158594717020884,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0007752620046798142,
      "loss": 1.7199722290039063,
      "step": 100580,
      "train_bpb": 0.6061471926489606,
      "train_bytes": 28416.73125,
      "train_loss_nats_per_token": 1.7202517356929
    },
    {
      "epoch": 0.18171762660978114,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0007747160453686357,
      "loss": 1.7168107986450196,
      "step": 100600,
      "train_bpb": 0.6061097935845976,
      "train_bytes": 28209.14375,
      "train_loss_nats_per_token": 1.7172961619734572
    },
    {
      "epoch": 0.18184930604935345,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0007741702114295621,
      "loss": 1.6997573852539063,
      "step": 100620,
      "train_bpb": 0.6003078016813199,
      "train_bytes": 28197.59375,
      "train_loss_nats_per_token": 1.700293354121843
    },
    {
      "epoch": 0.18198098548892577,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0007736245029569469,
      "loss": 1.7111143112182616,
      "step": 100640,
      "train_bpb": 0.6027624031699893,
      "train_bytes": 28568.91875,
      "train_loss_nats_per_token": 1.7120752009767415
    },
    {
      "epoch": 0.18211266492849806,
      "grad_norm": 0.15625,
      "learning_rate": 0.0007730789200451201,
      "loss": 1.7217378616333008,
      "step": 100660,
      "train_bpb": 0.6104458040593962,
      "train_bytes": 28096.64375,
      "train_loss_nats_per_token": 1.7229052975764283
    },
    {
      "epoch": 0.18224434436807038,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007725334627883924,
      "loss": 1.7051570892333985,
      "step": 100680,
      "train_bpb": 0.6043936393801195,
      "train_bytes": 28221.875,
      "train_loss_nats_per_token": 1.7054115283119002
    },
    {
      "epoch": 0.18237602380764267,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0007719881312810494,
      "loss": 1.7110530853271484,
      "step": 100700,
      "train_bpb": 0.6048827158059636,
      "train_bytes": 28397.30625,
      "train_loss_nats_per_token": 1.7113533249658515
    },
    {
      "epoch": 0.182507703247215,
      "grad_norm": 0.125,
      "learning_rate": 0.0007714429256173575,
      "loss": 1.7117719650268555,
      "step": 100720,
      "train_bpb": 0.6060149460217931,
      "train_bytes": 28294.4625,
      "train_loss_nats_per_token": 1.7119398107681634
    },
    {
      "epoch": 0.18263938268678728,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0007708978458915611,
      "loss": 1.7203380584716796,
      "step": 100740,
      "train_bpb": 0.6092638517720573,
      "train_bytes": 28324.43125,
      "train_loss_nats_per_token": 1.7206620593407969
    },
    {
      "epoch": 0.1827710621263596,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0007703528921978813,
      "loss": 1.7069183349609376,
      "step": 100760,
      "train_bpb": 0.6029282077485699,
      "train_bytes": 28403.39375,
      "train_loss_nats_per_token": 1.707729521072192
    },
    {
      "epoch": 0.1829027415659319,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0007698080646305187,
      "loss": 1.7230745315551759,
      "step": 100780,
      "train_bpb": 0.6078989598986054,
      "train_bytes": 28432.2,
      "train_loss_nats_per_token": 1.723001788610142
    },
    {
      "epoch": 0.1830344210055042,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0007692633632836529,
      "loss": 1.7144948959350585,
      "step": 100800,
      "train_bpb": 0.6047152682472555,
      "train_bytes": 28497.51875,
      "train_loss_nats_per_token": 1.7143406226565807
    },
    {
      "epoch": 0.1831661004450765,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0007687187882514389,
      "loss": 1.7212818145751954,
      "step": 100820,
      "train_bpb": 0.6089392045307271,
      "train_bytes": 28158.3125,
      "train_loss_nats_per_token": 1.7230886509852377
    },
    {
      "epoch": 0.18329777988464882,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0007681743396280126,
      "loss": 1.7066696166992188,
      "step": 100840,
      "train_bpb": 0.6051763213305867,
      "train_bytes": 28105.53125,
      "train_loss_nats_per_token": 1.7074132555006678
    },
    {
      "epoch": 0.1834294593242211,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007676300175074867,
      "loss": 1.7161203384399415,
      "step": 100860,
      "train_bpb": 0.6080161935981793,
      "train_bytes": 28212.8625,
      "train_loss_nats_per_token": 1.7171303320765563
    },
    {
      "epoch": 0.18356113876379343,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0007670858219839535,
      "loss": 1.7093839645385742,
      "step": 100880,
      "train_bpb": 0.6066414504984289,
      "train_bytes": 27829.76875,
      "train_loss_nats_per_token": 1.709865298357382
    },
    {
      "epoch": 0.18369281820336572,
      "grad_norm": 0.1953125,
      "learning_rate": 0.0007665417531514804,
      "loss": 1.7147165298461915,
      "step": 100900,
      "train_bpb": 0.6062294162963617,
      "train_bytes": 28581.8625,
      "train_loss_nats_per_token": 1.7148182390148097
    },
    {
      "epoch": 0.18382449764293804,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0007659978111041159,
      "loss": 1.7103452682495117,
      "step": 100920,
      "train_bpb": 0.6026542524844247,
      "train_bytes": 28548.3,
      "train_loss_nats_per_token": 1.7111876277963867
    },
    {
      "epoch": 0.18395617708251033,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0007654539959358859,
      "loss": 1.7162370681762695,
      "step": 100940,
      "train_bpb": 0.6062686566970112,
      "train_bytes": 28456.16875,
      "train_loss_nats_per_token": 1.7175446442046631
    },
    {
      "epoch": 0.18408785652208265,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0007649103077407923,
      "loss": 1.709036636352539,
      "step": 100960,
      "train_bpb": 0.6036176572400315,
      "train_bytes": 28372.30625,
      "train_loss_nats_per_token": 1.7094757404628018
    },
    {
      "epoch": 0.18421953596165494,
      "grad_norm": 0.177734375,
      "learning_rate": 0.000764366746612818,
      "loss": 1.7167409896850585,
      "step": 100980,
      "train_bpb": 0.6080483883057795,
      "train_bytes": 28320.1125,
      "train_loss_nats_per_token": 1.7171358287479197
    },
    {
      "epoch": 0.18435121540122726,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0007638233126459222,
      "loss": 1.710921859741211,
      "step": 101000,
      "train_bpb": 0.6054110889723385,
      "train_bytes": 28385.85625,
      "train_loss_nats_per_token": 1.7115717597740232
    },
    {
      "epoch": 0.18435121540122726,
      "eval_loss": 1.5314311981201172,
      "eval_runtime": 74.3001,
      "eval_samples_per_second": 13.459,
      "eval_steps_per_second": 13.459,
      "step": 101000,
      "train_bpb": 0.5404305075946781,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.538936788236137
    },
    {
      "epoch": 0.18448289484079955,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0007632800059340434,
      "loss": 1.7148937225341796,
      "step": 101020,
      "train_bpb": 0.607134492519035,
      "train_bytes": 28436.0625,
      "train_loss_nats_per_token": 1.715280798939403
    },
    {
      "epoch": 0.18461457428037187,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0007627368265710955,
      "loss": 1.720450210571289,
      "step": 101040,
      "train_bpb": 0.6061260275672,
      "train_bytes": 28362.2875,
      "train_loss_nats_per_token": 1.7212717173018248
    },
    {
      "epoch": 0.18474625371994416,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0007621937746509733,
      "loss": 1.7155824661254884,
      "step": 101060,
      "train_bpb": 0.6081273418197752,
      "train_bytes": 28384.0375,
      "train_loss_nats_per_token": 1.7160862852166692
    },
    {
      "epoch": 0.18487793315951648,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007616508502675488,
      "loss": 1.7204551696777344,
      "step": 101080,
      "train_bpb": 0.6042092303335671,
      "train_bytes": 28544.4875,
      "train_loss_nats_per_token": 1.721241116947048
    },
    {
      "epoch": 0.18500961259908877,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0007611080535146705,
      "loss": 1.7104314804077148,
      "step": 101100,
      "train_bpb": 0.6059352133964923,
      "train_bytes": 28190.2625,
      "train_loss_nats_per_token": 1.710657630532288
    },
    {
      "epoch": 0.1851412920386611,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007605653844861662,
      "loss": 1.7129598617553712,
      "step": 101120,
      "train_bpb": 0.6042042926390901,
      "train_bytes": 28342.10625,
      "train_loss_nats_per_token": 1.7130469304590787
    },
    {
      "epoch": 0.18527297147823338,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0007600228432758418,
      "loss": 1.7103961944580077,
      "step": 101140,
      "train_bpb": 0.6039560323453663,
      "train_bytes": 28176.5,
      "train_loss_nats_per_token": 1.7112454554420025
    },
    {
      "epoch": 0.1854046509178057,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0007594804299774799,
      "loss": 1.7158056259155274,
      "step": 101160,
      "train_bpb": 0.6118472728787063,
      "train_bytes": 28274.3875,
      "train_loss_nats_per_token": 1.7164577349802843
    },
    {
      "epoch": 0.185536330357378,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0007589381446848421,
      "loss": 1.7085783004760742,
      "step": 101180,
      "train_bpb": 0.6090588152018285,
      "train_bytes": 28055.6625,
      "train_loss_nats_per_token": 1.7086231144697017
    },
    {
      "epoch": 0.1856680097969503,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0007583959874916674,
      "loss": 1.7189952850341796,
      "step": 101200,
      "train_bpb": 0.6052411619984478,
      "train_bytes": 28554.96875,
      "train_loss_nats_per_token": 1.7190978096657896
    },
    {
      "epoch": 0.1857996892365226,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0007578539584916732,
      "loss": 1.7140005111694336,
      "step": 101220,
      "train_bpb": 0.6058836524071983,
      "train_bytes": 28024.5375,
      "train_loss_nats_per_token": 1.7143783652894171
    },
    {
      "epoch": 0.18593136867609492,
      "grad_norm": 0.107421875,
      "learning_rate": 0.0007573120577785535,
      "loss": 1.7036310195922852,
      "step": 101240,
      "train_bpb": 0.6055692573534937,
      "train_bytes": 28251.525,
      "train_loss_nats_per_token": 1.7045094720715293
    },
    {
      "epoch": 0.1860630481156672,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0007567702854459811,
      "loss": 1.7151338577270507,
      "step": 101260,
      "train_bpb": 0.6034848195357201,
      "train_bytes": 28665.4375,
      "train_loss_nats_per_token": 1.7161336052862648
    },
    {
      "epoch": 0.18619472755523953,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0007562286415876071,
      "loss": 1.718790054321289,
      "step": 101280,
      "train_bpb": 0.6062004908904984,
      "train_bytes": 28701.65625,
      "train_loss_nats_per_token": 1.7198805296279798
    },
    {
      "epoch": 0.18632640699481182,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0007556871262970588,
      "loss": 1.6980707168579101,
      "step": 101300,
      "train_bpb": 0.6035859018793474,
      "train_bytes": 28265.65625,
      "train_loss_nats_per_token": 1.6989402836932508
    },
    {
      "epoch": 0.18645808643438413,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0007551457396679423,
      "loss": 1.7057640075683593,
      "step": 101320,
      "train_bpb": 0.6054096356558485,
      "train_bytes": 28126.1875,
      "train_loss_nats_per_token": 1.706961394087102
    },
    {
      "epoch": 0.18658976587395645,
      "grad_norm": 0.140625,
      "learning_rate": 0.0007546044817938422,
      "loss": 1.716883659362793,
      "step": 101340,
      "train_bpb": 0.6030784429296852,
      "train_bytes": 28245.8875,
      "train_loss_nats_per_token": 1.7178187697387175
    },
    {
      "epoch": 0.18672144531352874,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0007540633527683186,
      "loss": 1.7117462158203125,
      "step": 101360,
      "train_bpb": 0.6053762800776109,
      "train_bytes": 28324.59375,
      "train_loss_nats_per_token": 1.71203693165897
    },
    {
      "epoch": 0.18685312475310106,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0007535223526849116,
      "loss": 1.7209459304809571,
      "step": 101380,
      "train_bpb": 0.6071346000537007,
      "train_bytes": 28242.625,
      "train_loss_nats_per_token": 1.720511874802937
    },
    {
      "epoch": 0.18698480419267335,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0007529814816371382,
      "loss": 1.7212331771850586,
      "step": 101400,
      "train_bpb": 0.6104621918909413,
      "train_bytes": 28113.6625,
      "train_loss_nats_per_token": 1.7218756685922123
    },
    {
      "epoch": 0.18711648363224567,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0007524407397184931,
      "loss": 1.7140430450439452,
      "step": 101420,
      "train_bpb": 0.6067359908360517,
      "train_bytes": 28206.5625,
      "train_loss_nats_per_token": 1.7151416301018898
    },
    {
      "epoch": 0.18724816307181796,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0007519001270224477,
      "loss": 1.706991958618164,
      "step": 101440,
      "train_bpb": 0.6057243797665296,
      "train_bytes": 27928.78125,
      "train_loss_nats_per_token": 1.7072942615557163
    },
    {
      "epoch": 0.18737984251139028,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0007513596436424524,
      "loss": 1.710467529296875,
      "step": 101460,
      "train_bpb": 0.6038170236734401,
      "train_bytes": 28437.9,
      "train_loss_nats_per_token": 1.7109001776676205
    },
    {
      "epoch": 0.18751152195096257,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0007508192896719359,
      "loss": 1.707286262512207,
      "step": 101480,
      "train_bpb": 0.606030577124552,
      "train_bytes": 28073.13125,
      "train_loss_nats_per_token": 1.7072651034112225
    },
    {
      "epoch": 0.1876432013905349,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0007502790652043017,
      "loss": 1.7179557800292968,
      "step": 101500,
      "train_bpb": 0.605659644794833,
      "train_bytes": 28275.93125,
      "train_loss_nats_per_token": 1.7188351352344622
    },
    {
      "epoch": 0.18777488083010718,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0007497389703329332,
      "loss": 1.7233518600463866,
      "step": 101520,
      "train_bpb": 0.6077103028659134,
      "train_bytes": 28650.33125,
      "train_loss_nats_per_token": 1.7239173645258419
    },
    {
      "epoch": 0.1879065602696795,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0007491990051511913,
      "loss": 1.7143999099731446,
      "step": 101540,
      "train_bpb": 0.6050381872064668,
      "train_bytes": 28414.725,
      "train_loss_nats_per_token": 1.71581193210899
    },
    {
      "epoch": 0.1880382397092518,
      "grad_norm": 0.140625,
      "learning_rate": 0.0007486591697524141,
      "loss": 1.7236177444458007,
      "step": 101560,
      "train_bpb": 0.6075306354785486,
      "train_bytes": 28445.68125,
      "train_loss_nats_per_token": 1.7245150851100455
    },
    {
      "epoch": 0.1881699191488241,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0007481194642299164,
      "loss": 1.7167608261108398,
      "step": 101580,
      "train_bpb": 0.6098565445028956,
      "train_bytes": 27739.66875,
      "train_loss_nats_per_token": 1.7180376432772415
    },
    {
      "epoch": 0.1883015985883964,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0007475798886769919,
      "loss": 1.7191995620727538,
      "step": 101600,
      "train_bpb": 0.60738248594047,
      "train_bytes": 28233.88125,
      "train_loss_nats_per_token": 1.719765016288172
    },
    {
      "epoch": 0.18843327802796872,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0007470404431869116,
      "loss": 1.7210487365722655,
      "step": 101620,
      "train_bpb": 0.6121478218511257,
      "train_bytes": 27902.46875,
      "train_loss_nats_per_token": 1.7213239963290985
    },
    {
      "epoch": 0.188564957467541,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0007465011278529228,
      "loss": 1.7060169219970702,
      "step": 101640,
      "train_bpb": 0.6051671161079962,
      "train_bytes": 28330.475,
      "train_loss_nats_per_token": 1.7070764870667792
    },
    {
      "epoch": 0.18869663690711333,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0007459619427682517,
      "loss": 1.7150041580200195,
      "step": 101660,
      "train_bpb": 0.6062458062808286,
      "train_bytes": 28264.9,
      "train_loss_nats_per_token": 1.715770076707731
    },
    {
      "epoch": 0.18882831634668562,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0007454228880261013,
      "loss": 1.7230854034423828,
      "step": 101680,
      "train_bpb": 0.6087339727035197,
      "train_bytes": 28520.9125,
      "train_loss_nats_per_token": 1.7234506984277114
    },
    {
      "epoch": 0.18895999578625794,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0007448839637196529,
      "loss": 1.7103206634521484,
      "step": 101700,
      "train_bpb": 0.6025555305608601,
      "train_bytes": 28561.21875,
      "train_loss_nats_per_token": 1.71121500664248
    },
    {
      "epoch": 0.18909167522583023,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0007443451699420636,
      "loss": 1.7139127731323243,
      "step": 101720,
      "train_bpb": 0.6080487434267616,
      "train_bytes": 28129.1625,
      "train_loss_nats_per_token": 1.7152426551770596
    },
    {
      "epoch": 0.18922335466540255,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0007438065067864695,
      "loss": 1.7171886444091797,
      "step": 101740,
      "train_bpb": 0.609776747449147,
      "train_bytes": 28064.9,
      "train_loss_nats_per_token": 1.7176925419953455
    },
    {
      "epoch": 0.18935503410497484,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0007432679743459838,
      "loss": 1.718115234375,
      "step": 101760,
      "train_bpb": 0.6113600776599364,
      "train_bytes": 28029.15,
      "train_loss_nats_per_token": 1.7191374171037126
    },
    {
      "epoch": 0.18948671354454716,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0007427295727136962,
      "loss": 1.7171085357666016,
      "step": 101780,
      "train_bpb": 0.6058168624991035,
      "train_bytes": 28267.05,
      "train_loss_nats_per_token": 1.7174916333500823
    },
    {
      "epoch": 0.18961839298411945,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.0007421913019826748,
      "loss": 1.7200515747070313,
      "step": 101800,
      "train_bpb": 0.6077505935703752,
      "train_bytes": 28265.08125,
      "train_loss_nats_per_token": 1.7214308983590012
    },
    {
      "epoch": 0.18975007242369177,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0007416531622459651,
      "loss": 1.7230222702026368,
      "step": 101820,
      "train_bpb": 0.6087167314403233,
      "train_bytes": 28347.60625,
      "train_loss_nats_per_token": 1.7237857418843752
    },
    {
      "epoch": 0.18988175186326406,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007411151535965889,
      "loss": 1.715103530883789,
      "step": 101840,
      "train_bpb": 0.6069493766086392,
      "train_bytes": 28376.04375,
      "train_loss_nats_per_token": 1.7157897804475504
    },
    {
      "epoch": 0.19001343130283638,
      "grad_norm": 0.1484375,
      "learning_rate": 0.000740577276127546,
      "loss": 1.7217174530029298,
      "step": 101860,
      "train_bpb": 0.6100865690071415,
      "train_bytes": 28354.40625,
      "train_loss_nats_per_token": 1.722091768559236
    },
    {
      "epoch": 0.19014511074240867,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.000740039529931814,
      "loss": 1.7194478988647461,
      "step": 101880,
      "train_bpb": 0.6080690433341271,
      "train_bytes": 28426.5875,
      "train_loss_nats_per_token": 1.7200282849386708
    },
    {
      "epoch": 0.190276790181981,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.0007395019151023479,
      "loss": 1.7060201644897461,
      "step": 101900,
      "train_bpb": 0.602270520860317,
      "train_bytes": 28328.125,
      "train_loss_nats_per_token": 1.7062879121752714
    },
    {
      "epoch": 0.19040846962155328,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.000738964431732078,
      "loss": 1.7265584945678711,
      "step": 101920,
      "train_bpb": 0.6130172372750888,
      "train_bytes": 28397.56875,
      "train_loss_nats_per_token": 1.7275505831960916
    },
    {
      "epoch": 0.1905401490611256,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0007384270799139141,
      "loss": 1.7235631942749023,
      "step": 101940,
      "train_bpb": 0.6090238744149794,
      "train_bytes": 28623.28125,
      "train_loss_nats_per_token": 1.7247405989360767
    },
    {
      "epoch": 0.1906718285006979,
      "grad_norm": 0.19140625,
      "learning_rate": 0.000737889859740743,
      "loss": 1.72259521484375,
      "step": 101960,
      "train_bpb": 0.609468077768853,
      "train_bytes": 28060.28125,
      "train_loss_nats_per_token": 1.7228914389617602
    },
    {
      "epoch": 0.1908035079402702,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0007373527713054272,
      "loss": 1.7236173629760743,
      "step": 101980,
      "train_bpb": 0.609370735449265,
      "train_bytes": 28279.6625,
      "train_loss_nats_per_token": 1.7240687985412864
    },
    {
      "epoch": 0.19093518737984252,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0007368158147008078,
      "loss": 1.719095230102539,
      "step": 102000,
      "train_bpb": 0.6085234587287591,
      "train_bytes": 28564.24375,
      "train_loss_nats_per_token": 1.7199423472729853
    },
    {
      "epoch": 0.19093518737984252,
      "eval_loss": 1.5306872129440308,
      "eval_runtime": 73.6021,
      "eval_samples_per_second": 13.587,
      "eval_steps_per_second": 13.587,
      "step": 102000,
      "train_bpb": 0.5401898278896708,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5382514256464748
    },
    {
      "epoch": 0.19106686681941482,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.000736278990019703,
      "loss": 1.7185522079467774,
      "step": 102020,
      "train_bpb": 0.6083120693549544,
      "train_bytes": 28112.85625,
      "train_loss_nats_per_token": 1.7189730194878872
    },
    {
      "epoch": 0.19119854625898713,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0007357422973549085,
      "loss": 1.7193540573120116,
      "step": 102040,
      "train_bpb": 0.605364115753269,
      "train_bytes": 28524.2,
      "train_loss_nats_per_token": 1.720339244777864
    },
    {
      "epoch": 0.19133022569855943,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0007352057367991953,
      "loss": 1.7257465362548827,
      "step": 102060,
      "train_bpb": 0.6105480363012236,
      "train_bytes": 28533.7625,
      "train_loss_nats_per_token": 1.7264911425400686
    },
    {
      "epoch": 0.19146190513813174,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0007346693084453136,
      "loss": 1.7207143783569336,
      "step": 102080,
      "train_bpb": 0.607606930741742,
      "train_bytes": 28435.575,
      "train_loss_nats_per_token": 1.7210279834634012
    },
    {
      "epoch": 0.19159358457770403,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0007341330123859908,
      "loss": 1.7139589309692382,
      "step": 102100,
      "train_bpb": 0.6054684521724346,
      "train_bytes": 28034.9375,
      "train_loss_nats_per_token": 1.7142185437623843
    },
    {
      "epoch": 0.19172526401727635,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0007335968487139291,
      "loss": 1.7184564590454101,
      "step": 102120,
      "train_bpb": 0.6084236577784033,
      "train_bytes": 28167.59375,
      "train_loss_nats_per_token": 1.7189031463927855
    },
    {
      "epoch": 0.19185694345684864,
      "grad_norm": 0.125,
      "learning_rate": 0.0007330608175218103,
      "loss": 1.7206663131713866,
      "step": 102140,
      "train_bpb": 0.6085196118510904,
      "train_bytes": 28045.7125,
      "train_loss_nats_per_token": 1.7214328341347749
    },
    {
      "epoch": 0.19198862289642096,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.000732524918902293,
      "loss": 1.7207138061523437,
      "step": 102160,
      "train_bpb": 0.613257623340907,
      "train_bytes": 28134.3375,
      "train_loss_nats_per_token": 1.7210107343088652
    },
    {
      "epoch": 0.19212030233599325,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0007319891529480111,
      "loss": 1.707444953918457,
      "step": 102180,
      "train_bpb": 0.6014512677622896,
      "train_bytes": 28383.15,
      "train_loss_nats_per_token": 1.7080345129649126
    },
    {
      "epoch": 0.19225198177556557,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0007314535197515771,
      "loss": 1.7200021743774414,
      "step": 102200,
      "train_bpb": 0.6065868726859733,
      "train_bytes": 28249.725,
      "train_loss_nats_per_token": 1.720135744040184
    },
    {
      "epoch": 0.19238366121513786,
      "grad_norm": 0.205078125,
      "learning_rate": 0.0007309180194055804,
      "loss": 1.7091594696044923,
      "step": 102220,
      "train_bpb": 0.6051033189290767,
      "train_bytes": 28341.0125,
      "train_loss_nats_per_token": 1.7097986777900702
    },
    {
      "epoch": 0.19251534065471018,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0007303826520025878,
      "loss": 1.7148029327392578,
      "step": 102240,
      "train_bpb": 0.6065890371186491,
      "train_bytes": 28231.9625,
      "train_loss_nats_per_token": 1.7152674657575033
    },
    {
      "epoch": 0.19264702009428247,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0007298474176351413,
      "loss": 1.7288322448730469,
      "step": 102260,
      "train_bpb": 0.6119001685023105,
      "train_bytes": 28459.59375,
      "train_loss_nats_per_token": 1.7297260057827228
    },
    {
      "epoch": 0.1927786995338548,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0007293123163957619,
      "loss": 1.7183267593383789,
      "step": 102280,
      "train_bpb": 0.6046080241574715,
      "train_bytes": 28416.5,
      "train_loss_nats_per_token": 1.718951572678671
    },
    {
      "epoch": 0.19291037897342708,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0007287773483769473,
      "loss": 1.7218795776367188,
      "step": 102300,
      "train_bpb": 0.6081578900193235,
      "train_bytes": 28477.7625,
      "train_loss_nats_per_token": 1.722060100616185
    },
    {
      "epoch": 0.1930420584129994,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0007282425136711705,
      "loss": 1.7273109436035157,
      "step": 102320,
      "train_bpb": 0.6120484183059112,
      "train_bytes": 27986.58125,
      "train_loss_nats_per_token": 1.7274632735525532
    },
    {
      "epoch": 0.1931737378525717,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0007277078123708835,
      "loss": 1.7139926910400392,
      "step": 102340,
      "train_bpb": 0.6088628743465007,
      "train_bytes": 27878.05625,
      "train_loss_nats_per_token": 1.7148164920492506
    },
    {
      "epoch": 0.193305417292144,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0007271732445685142,
      "loss": 1.7051162719726562,
      "step": 102360,
      "train_bpb": 0.6051917263147639,
      "train_bytes": 28258.35625,
      "train_loss_nats_per_token": 1.7065661825379213
    },
    {
      "epoch": 0.1934370967317163,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0007266388103564684,
      "loss": 1.7125562667846679,
      "step": 102380,
      "train_bpb": 0.6074828066139109,
      "train_bytes": 28126.925,
      "train_loss_nats_per_token": 1.714030662226259
    },
    {
      "epoch": 0.19356877617128862,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0007261045098271266,
      "loss": 1.725687026977539,
      "step": 102400,
      "train_bpb": 0.6080548273788506,
      "train_bytes": 28289.7125,
      "train_loss_nats_per_token": 1.7261955339956898
    },
    {
      "epoch": 0.1937004556108609,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0007255703430728485,
      "loss": 1.7319965362548828,
      "step": 102420,
      "train_bpb": 0.6121249576111496,
      "train_bytes": 28343.53125,
      "train_loss_nats_per_token": 1.7324281353091113
    },
    {
      "epoch": 0.19383213505043323,
      "grad_norm": 0.171875,
      "learning_rate": 0.0007250363101859702,
      "loss": 1.7245916366577148,
      "step": 102440,
      "train_bpb": 0.6072693344492199,
      "train_bytes": 28507.68125,
      "train_loss_nats_per_token": 1.7252888482016489
    },
    {
      "epoch": 0.19396381449000552,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0007245024112588033,
      "loss": 1.7202871322631836,
      "step": 102460,
      "train_bpb": 0.6079257641999327,
      "train_bytes": 28403.125,
      "train_loss_nats_per_token": 1.720125173867799
    },
    {
      "epoch": 0.19409549392957784,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0007239686463836376,
      "loss": 1.7192422866821289,
      "step": 102480,
      "train_bpb": 0.6054386509595725,
      "train_bytes": 28933.0625,
      "train_loss_nats_per_token": 1.7199357410410747
    },
    {
      "epoch": 0.19422717336915013,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0007234350156527394,
      "loss": 1.7156333923339844,
      "step": 102500,
      "train_bpb": 0.6096551534848579,
      "train_bytes": 28346.5125,
      "train_loss_nats_per_token": 1.7163461019796937
    },
    {
      "epoch": 0.19435885280872245,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0007229015191583522,
      "loss": 1.7248836517333985,
      "step": 102520,
      "train_bpb": 0.6045489441115219,
      "train_bytes": 28630.8125,
      "train_loss_nats_per_token": 1.72504679962578
    },
    {
      "epoch": 0.19449053224829474,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0007223681569926952,
      "loss": 1.7164220809936523,
      "step": 102540,
      "train_bpb": 0.6044932504121596,
      "train_bytes": 28227.53125,
      "train_loss_nats_per_token": 1.7164174702015313
    },
    {
      "epoch": 0.19462221168786706,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0007218349292479651,
      "loss": 1.7081613540649414,
      "step": 102560,
      "train_bpb": 0.6024691369412569,
      "train_bytes": 28358.775,
      "train_loss_nats_per_token": 1.7095375405306816
    },
    {
      "epoch": 0.19475389112743935,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0007213018360163362,
      "loss": 1.7229911804199218,
      "step": 102580,
      "train_bpb": 0.6084461047863258,
      "train_bytes": 28244.76875,
      "train_loss_nats_per_token": 1.7240224101531758
    },
    {
      "epoch": 0.19488557056701167,
      "grad_norm": 0.171875,
      "learning_rate": 0.0007207688773899571,
      "loss": 1.7168275833129882,
      "step": 102600,
      "train_bpb": 0.6069604094860346,
      "train_bytes": 28137.9375,
      "train_loss_nats_per_token": 1.7170353909911522
    },
    {
      "epoch": 0.19501725000658396,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0007202360534609557,
      "loss": 1.7192375183105468,
      "step": 102620,
      "train_bpb": 0.6073204129804627,
      "train_bytes": 28372.74375,
      "train_loss_nats_per_token": 1.71938821796426
    },
    {
      "epoch": 0.19514892944615628,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0007197033643214358,
      "loss": 1.706901741027832,
      "step": 102640,
      "train_bpb": 0.6050181783000032,
      "train_bytes": 28207.3,
      "train_loss_nats_per_token": 1.7070862395811255
    },
    {
      "epoch": 0.1952806088857286,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0007191708100634767,
      "loss": 1.7222261428833008,
      "step": 102660,
      "train_bpb": 0.6082340187745324,
      "train_bytes": 28323.98125,
      "train_loss_nats_per_token": 1.7223780523826449
    },
    {
      "epoch": 0.1954122883253009,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.000718638390779136,
      "loss": 1.7033349990844726,
      "step": 102680,
      "train_bpb": 0.6046999271284683,
      "train_bytes": 28062.625,
      "train_loss_nats_per_token": 1.7028296545323873
    },
    {
      "epoch": 0.1955439677648732,
      "grad_norm": 0.125,
      "learning_rate": 0.0007181061065604474,
      "loss": 1.7013946533203126,
      "step": 102700,
      "train_bpb": 0.6039046404425162,
      "train_bytes": 28082.2875,
      "train_loss_nats_per_token": 1.702180577285521
    },
    {
      "epoch": 0.1956756472044455,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0007175739574994214,
      "loss": 1.7230838775634765,
      "step": 102720,
      "train_bpb": 0.6075020063651007,
      "train_bytes": 28794.33125,
      "train_loss_nats_per_token": 1.7240719284136525
    },
    {
      "epoch": 0.19580732664401782,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0007170419436880442,
      "loss": 1.706479263305664,
      "step": 102740,
      "train_bpb": 0.6016132193956905,
      "train_bytes": 28423.63125,
      "train_loss_nats_per_token": 1.7075835723660011
    },
    {
      "epoch": 0.1959390060835901,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0007165100652182798,
      "loss": 1.7293163299560548,
      "step": 102760,
      "train_bpb": 0.6092731253039964,
      "train_bytes": 28612.3,
      "train_loss_nats_per_token": 1.729925259232246
    },
    {
      "epoch": 0.19607068552316242,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.000715978322182069,
      "loss": 1.7105342864990234,
      "step": 102780,
      "train_bpb": 0.5999461152241912,
      "train_bytes": 28489.38125,
      "train_loss_nats_per_token": 1.7104902422089663
    },
    {
      "epoch": 0.19620236496273472,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0007154467146713272,
      "loss": 1.6999525070190429,
      "step": 102800,
      "train_bpb": 0.6008359043341676,
      "train_bytes": 28368.9875,
      "train_loss_nats_per_token": 1.7010878352282799
    },
    {
      "epoch": 0.19633404440230703,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0007149152427779486,
      "loss": 1.7083927154541017,
      "step": 102820,
      "train_bpb": 0.6060183024251159,
      "train_bytes": 28301.49375,
      "train_loss_nats_per_token": 1.7097734576544037
    },
    {
      "epoch": 0.19646572384187933,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0007143839065938026,
      "loss": 1.702183151245117,
      "step": 102840,
      "train_bpb": 0.6050160597535598,
      "train_bytes": 28345.84375,
      "train_loss_nats_per_token": 1.703253226169085
    },
    {
      "epoch": 0.19659740328145164,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0007138527062107367,
      "loss": 1.7228355407714844,
      "step": 102860,
      "train_bpb": 0.6090192374444691,
      "train_bytes": 28315.4875,
      "train_loss_nats_per_token": 1.7234738414048596
    },
    {
      "epoch": 0.19672908272102393,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0007133216417205726,
      "loss": 1.7064359664916993,
      "step": 102880,
      "train_bpb": 0.6029894134067182,
      "train_bytes": 28617.35625,
      "train_loss_nats_per_token": 1.7067936678945899
    },
    {
      "epoch": 0.19686076216059625,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0007127907132151101,
      "loss": 1.713540267944336,
      "step": 102900,
      "train_bpb": 0.6063865375087095,
      "train_bytes": 28206.4625,
      "train_loss_nats_per_token": 1.7127222591787332
    },
    {
      "epoch": 0.19699244160016854,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0007122599207861256,
      "loss": 1.7130853652954101,
      "step": 102920,
      "train_bpb": 0.6076551314133077,
      "train_bytes": 28037.6125,
      "train_loss_nats_per_token": 1.7139001968694045
    },
    {
      "epoch": 0.19712412103974086,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0007117292645253709,
      "loss": 1.7076053619384766,
      "step": 102940,
      "train_bpb": 0.6029636170703139,
      "train_bytes": 28279.46875,
      "train_loss_nats_per_token": 1.7076874117573622
    },
    {
      "epoch": 0.19725580047931315,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0007111987445245751,
      "loss": 1.717823600769043,
      "step": 102960,
      "train_bpb": 0.6066338143047146,
      "train_bytes": 28281.61875,
      "train_loss_nats_per_token": 1.7178542626919886
    },
    {
      "epoch": 0.19738747991888547,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0007106683608754443,
      "loss": 1.7014097213745116,
      "step": 102980,
      "train_bpb": 0.6013842993990529,
      "train_bytes": 28502.5625,
      "train_loss_nats_per_token": 1.7018636551104334
    },
    {
      "epoch": 0.19751915935845776,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0007101381136696591,
      "loss": 1.7156618118286133,
      "step": 103000,
      "train_bpb": 0.6047483511121482,
      "train_bytes": 28700.05625,
      "train_loss_nats_per_token": 1.7166395962739844
    },
    {
      "epoch": 0.19751915935845776,
      "eval_loss": 1.5306203365325928,
      "eval_runtime": 74.5015,
      "eval_samples_per_second": 13.423,
      "eval_steps_per_second": 13.423,
      "step": 103000,
      "train_bpb": 0.5401546188289561,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5381511638772885
    },
    {
      "epoch": 0.19765083879803008,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0007096080029988781,
      "loss": 1.7155488967895507,
      "step": 103020,
      "train_bpb": 0.6074612145832063,
      "train_bytes": 28064.41875,
      "train_loss_nats_per_token": 1.7155624604654172
    },
    {
      "epoch": 0.19778251823760237,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0007090780289547361,
      "loss": 1.7178308486938476,
      "step": 103040,
      "train_bpb": 0.6079052993850431,
      "train_bytes": 28214.125,
      "train_loss_nats_per_token": 1.7187571483405417
    },
    {
      "epoch": 0.1979141976771747,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0007085481916288442,
      "loss": 1.7202323913574218,
      "step": 103060,
      "train_bpb": 0.6083923683689451,
      "train_bytes": 28461.08125,
      "train_loss_nats_per_token": 1.7205518325120395
    },
    {
      "epoch": 0.19804587711674698,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0007080184911127891,
      "loss": 1.7058071136474608,
      "step": 103080,
      "train_bpb": 0.6047446463267262,
      "train_bytes": 28063.40625,
      "train_loss_nats_per_token": 1.7060594400833942
    },
    {
      "epoch": 0.1981775565563193,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0007074889274981349,
      "loss": 1.7167318344116211,
      "step": 103100,
      "train_bpb": 0.6050829053732203,
      "train_bytes": 28525.875,
      "train_loss_nats_per_token": 1.717223747514268
    },
    {
      "epoch": 0.1983092359958916,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0007069595008764219,
      "loss": 1.7140432357788087,
      "step": 103120,
      "train_bpb": 0.6055800947901985,
      "train_bytes": 28477.7,
      "train_loss_nats_per_token": 1.7142683015691424
    },
    {
      "epoch": 0.1984409154354639,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0007064302113391656,
      "loss": 1.7097154617309571,
      "step": 103140,
      "train_bpb": 0.6005480783494379,
      "train_bytes": 28461.0375,
      "train_loss_nats_per_token": 1.7103486939762464
    },
    {
      "epoch": 0.1985725948750362,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0007059010589778593,
      "loss": 1.7053361892700196,
      "step": 103160,
      "train_bpb": 0.603053721960013,
      "train_bytes": 28425.725,
      "train_loss_nats_per_token": 1.70599800061727
    },
    {
      "epoch": 0.19870427431460852,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0007053720438839717,
      "loss": 1.7084068298339843,
      "step": 103180,
      "train_bpb": 0.6076843011275863,
      "train_bytes": 27996.08125,
      "train_loss_nats_per_token": 1.708740486432896
    },
    {
      "epoch": 0.1988359537541808,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0007048431661489485,
      "loss": 1.7117549896240234,
      "step": 103200,
      "train_bpb": 0.6038128152805657,
      "train_bytes": 28329.46875,
      "train_loss_nats_per_token": 1.7122285205238497
    },
    {
      "epoch": 0.19896763319375313,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0007043144258642103,
      "loss": 1.7198904037475586,
      "step": 103220,
      "train_bpb": 0.6072129309261574,
      "train_bytes": 28442.9875,
      "train_loss_nats_per_token": 1.7204453212786261
    },
    {
      "epoch": 0.19909931263332542,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0007037858231211551,
      "loss": 1.7174942016601562,
      "step": 103240,
      "train_bpb": 0.6025771778120981,
      "train_bytes": 28607.38125,
      "train_loss_nats_per_token": 1.7183916495876312
    },
    {
      "epoch": 0.19923099207289774,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0007032573580111574,
      "loss": 1.7153289794921875,
      "step": 103260,
      "train_bpb": 0.6053020835812325,
      "train_bytes": 28542.175,
      "train_loss_nats_per_token": 1.715934991793844
    },
    {
      "epoch": 0.19936267151247003,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0007027290306255666,
      "loss": 1.7134557723999024,
      "step": 103280,
      "train_bpb": 0.6031459214220974,
      "train_bytes": 28443.5875,
      "train_loss_nats_per_token": 1.713692885088924
    },
    {
      "epoch": 0.19949435095204235,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0007022008410557091,
      "loss": 1.7105209350585937,
      "step": 103300,
      "train_bpb": 0.6053136419630294,
      "train_bytes": 28649.575,
      "train_loss_nats_per_token": 1.711375854538351
    },
    {
      "epoch": 0.19962603039161464,
      "grad_norm": 0.140625,
      "learning_rate": 0.0007016727893928882,
      "loss": 1.7139432907104493,
      "step": 103320,
      "train_bpb": 0.6012759306222347,
      "train_bytes": 28712.225,
      "train_loss_nats_per_token": 1.714735127802312
    },
    {
      "epoch": 0.19975770983118696,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0007011448757283812,
      "loss": 1.7138561248779296,
      "step": 103340,
      "train_bpb": 0.6050024624321598,
      "train_bytes": 28439.88125,
      "train_loss_nats_per_token": 1.714458614792112
    },
    {
      "epoch": 0.19988938927075928,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0007006171001534439,
      "loss": 1.7115362167358399,
      "step": 103360,
      "train_bpb": 0.603043754043916,
      "train_bytes": 28691.3125,
      "train_loss_nats_per_token": 1.7124018635605094
    },
    {
      "epoch": 0.20002106871033157,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.000700089462759307,
      "loss": 1.7078405380249024,
      "step": 103380,
      "train_bpb": 0.6049369827964926,
      "train_bytes": 28298.75,
      "train_loss_nats_per_token": 1.708707517569081
    },
    {
      "epoch": 0.2001527481499039,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.000699561963637178,
      "loss": 1.7175397872924805,
      "step": 103400,
      "train_bpb": 0.6096211770461013,
      "train_bytes": 28209.36875,
      "train_loss_nats_per_token": 1.7184376835830364
    },
    {
      "epoch": 0.20028442758947618,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.000699034602878239,
      "loss": 1.708108901977539,
      "step": 103420,
      "train_bpb": 0.6065117600624069,
      "train_bytes": 28231.34375,
      "train_loss_nats_per_token": 1.7082232615355182
    },
    {
      "epoch": 0.2004161070290485,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.00069850738057365,
      "loss": 1.7204010009765625,
      "step": 103440,
      "train_bpb": 0.6080630280744149,
      "train_bytes": 28322.76875,
      "train_loss_nats_per_token": 1.7203452635688297
    },
    {
      "epoch": 0.2005477864686208,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0006979802968145465,
      "loss": 1.7097557067871094,
      "step": 103460,
      "train_bpb": 0.6066873722065615,
      "train_bytes": 28742.65625,
      "train_loss_nats_per_token": 1.7104828176824212
    },
    {
      "epoch": 0.2006794659081931,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.0006974533516920391,
      "loss": 1.7121332168579102,
      "step": 103480,
      "train_bpb": 0.6064329832955832,
      "train_bytes": 27884.3625,
      "train_loss_nats_per_token": 1.7119819966314427
    },
    {
      "epoch": 0.2008111453477654,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0006969265452972155,
      "loss": 1.7142446517944336,
      "step": 103500,
      "train_bpb": 0.6050284211741322,
      "train_bytes": 28533.45625,
      "train_loss_nats_per_token": 1.7139598924679023
    },
    {
      "epoch": 0.20094282478733771,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0006963998777211393,
      "loss": 1.710982894897461,
      "step": 103520,
      "train_bpb": 0.6074754100438411,
      "train_bytes": 28161.36875,
      "train_loss_nats_per_token": 1.7124424131434077
    },
    {
      "epoch": 0.20107450422691,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0006958733490548501,
      "loss": 1.7028034210205079,
      "step": 103540,
      "train_bpb": 0.6029463742949375,
      "train_bytes": 28304.6875,
      "train_loss_nats_per_token": 1.7026131622190204
    },
    {
      "epoch": 0.20120618366648232,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0006953469593893624,
      "loss": 1.7154083251953125,
      "step": 103560,
      "train_bpb": 0.6053119617483256,
      "train_bytes": 28133.25625,
      "train_loss_nats_per_token": 1.7161180413442998
    },
    {
      "epoch": 0.20133786310605462,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0006948207088156683,
      "loss": 1.7193016052246093,
      "step": 103580,
      "train_bpb": 0.6059165659599106,
      "train_bytes": 28351.3,
      "train_loss_nats_per_token": 1.719581821608848
    },
    {
      "epoch": 0.20146954254562693,
      "grad_norm": 0.1123046875,
      "learning_rate": 0.0006942945974247357,
      "loss": 1.7105047225952148,
      "step": 103600,
      "train_bpb": 0.6053837985244046,
      "train_bytes": 28253.0125,
      "train_loss_nats_per_token": 1.7117506010483077
    },
    {
      "epoch": 0.20160122198519922,
      "grad_norm": 0.1875,
      "learning_rate": 0.0006937686253075064,
      "loss": 1.7307844161987305,
      "step": 103620,
      "train_bpb": 0.6106317457018873,
      "train_bytes": 28392.825,
      "train_loss_nats_per_token": 1.731330821591347
    },
    {
      "epoch": 0.20173290142477154,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0006932427925549002,
      "loss": 1.7252349853515625,
      "step": 103640,
      "train_bpb": 0.6091486281576841,
      "train_bytes": 28410.625,
      "train_loss_nats_per_token": 1.7258045228885563
    },
    {
      "epoch": 0.20186458086434383,
      "grad_norm": 0.181640625,
      "learning_rate": 0.0006927170992578125,
      "loss": 1.7224868774414062,
      "step": 103660,
      "train_bpb": 0.6069122359651079,
      "train_bytes": 28696.59375,
      "train_loss_nats_per_token": 1.722323791296876
    },
    {
      "epoch": 0.20199626030391615,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0006921915455071147,
      "loss": 1.7186338424682617,
      "step": 103680,
      "train_bpb": 0.611267357531768,
      "train_bytes": 28010.00625,
      "train_loss_nats_per_token": 1.7189793565199305
    },
    {
      "epoch": 0.20212793974348844,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0006916661313936524,
      "loss": 1.714025115966797,
      "step": 103700,
      "train_bpb": 0.6062561237851368,
      "train_bytes": 28307.36875,
      "train_loss_nats_per_token": 1.7151949979312244
    },
    {
      "epoch": 0.20225961918306076,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0006911408570082493,
      "loss": 1.7145961761474608,
      "step": 103720,
      "train_bpb": 0.6032132248070152,
      "train_bytes": 28554.74375,
      "train_loss_nats_per_token": 1.715861999517201
    },
    {
      "epoch": 0.20239129862263305,
      "grad_norm": 0.125,
      "learning_rate": 0.0006906157224417041,
      "loss": 1.709646224975586,
      "step": 103740,
      "train_bpb": 0.6086615834106438,
      "train_bytes": 28028.9125,
      "train_loss_nats_per_token": 1.710222862154311
    },
    {
      "epoch": 0.20252297806220537,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0006900907277847905,
      "loss": 1.7176027297973633,
      "step": 103760,
      "train_bpb": 0.6052586613055835,
      "train_bytes": 28361.55625,
      "train_loss_nats_per_token": 1.718064717241358
    },
    {
      "epoch": 0.20265465750177766,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0006895658731282588,
      "loss": 1.7175174713134767,
      "step": 103780,
      "train_bpb": 0.6093336425545787,
      "train_bytes": 28109.6,
      "train_loss_nats_per_token": 1.7181254838975186
    },
    {
      "epoch": 0.20278633694134998,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.000689041158562836,
      "loss": 1.7052240371704102,
      "step": 103800,
      "train_bpb": 0.6049837212918893,
      "train_bytes": 28412.95,
      "train_loss_nats_per_token": 1.706599760989407
    },
    {
      "epoch": 0.20291801638092227,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0006885165841792227,
      "loss": 1.7289678573608398,
      "step": 103820,
      "train_bpb": 0.6132602327332884,
      "train_bytes": 28261.08125,
      "train_loss_nats_per_token": 1.7307605747255326
    },
    {
      "epoch": 0.2030496958204946,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0006879921500680971,
      "loss": 1.7188825607299805,
      "step": 103840,
      "train_bpb": 0.6068935201638576,
      "train_bytes": 28412.675,
      "train_loss_nats_per_token": 1.7199184407113084
    },
    {
      "epoch": 0.20318137526006688,
      "grad_norm": 0.140625,
      "learning_rate": 0.0006874678563201121,
      "loss": 1.7216541290283203,
      "step": 103860,
      "train_bpb": 0.6124243295688392,
      "train_bytes": 28062.6875,
      "train_loss_nats_per_token": 1.7218278126620408
    },
    {
      "epoch": 0.2033130546996392,
      "grad_norm": 0.150390625,
      "learning_rate": 0.000686943703025898,
      "loss": 1.7143280029296875,
      "step": 103880,
      "train_bpb": 0.6058383942619678,
      "train_bytes": 28626.7125,
      "train_loss_nats_per_token": 1.7148465248125602
    },
    {
      "epoch": 0.2034447341392115,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0006864196902760578,
      "loss": 1.7168281555175782,
      "step": 103900,
      "train_bpb": 0.6097049924141285,
      "train_bytes": 28331.28125,
      "train_loss_nats_per_token": 1.716737119392785
    },
    {
      "epoch": 0.2035764135787838,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0006858958181611732,
      "loss": 1.7144258499145508,
      "step": 103920,
      "train_bpb": 0.6099020852411239,
      "train_bytes": 28222.2,
      "train_loss_nats_per_token": 1.7149633025113442
    },
    {
      "epoch": 0.2037080930183561,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0006853720867718004,
      "loss": 1.720533561706543,
      "step": 103940,
      "train_bpb": 0.6109750110556137,
      "train_bytes": 28208.60625,
      "train_loss_nats_per_token": 1.7211017176856487
    },
    {
      "epoch": 0.20383977245792842,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0006848484961984707,
      "loss": 1.7235706329345704,
      "step": 103960,
      "train_bpb": 0.6097549419058147,
      "train_bytes": 28699.46875,
      "train_loss_nats_per_token": 1.7244244820150365
    },
    {
      "epoch": 0.2039714518975007,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0006843250465316915,
      "loss": 1.720595932006836,
      "step": 103980,
      "train_bpb": 0.6132840560066368,
      "train_bytes": 28040.33125,
      "train_loss_nats_per_token": 1.7215851378584128
    },
    {
      "epoch": 0.20410313133707303,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0006838017378619466,
      "loss": 1.7180959701538085,
      "step": 104000,
      "train_bpb": 0.6068652904785317,
      "train_bytes": 28430.9875,
      "train_loss_nats_per_token": 1.7181006371842489
    },
    {
      "epoch": 0.20410313133707303,
      "eval_loss": 1.5309079885482788,
      "eval_runtime": 78.2416,
      "eval_samples_per_second": 12.781,
      "eval_steps_per_second": 12.781,
      "step": 104000,
      "train_bpb": 0.5402574849634552,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.538444086790447
    },
    {
      "epoch": 0.20423481077664535,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.000683278570279695,
      "loss": 1.7238460540771485,
      "step": 104020,
      "train_bpb": 0.6089925405542901,
      "train_bytes": 28356.2,
      "train_loss_nats_per_token": 1.7236844836873
    },
    {
      "epoch": 0.20436649021621764,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.0006827555438753701,
      "loss": 1.7282581329345703,
      "step": 104040,
      "train_bpb": 0.6121796072631843,
      "train_bytes": 28403.23125,
      "train_loss_nats_per_token": 1.7285842577223705
    },
    {
      "epoch": 0.20449816965578996,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0006822326587393826,
      "loss": 1.717844581604004,
      "step": 104060,
      "train_bpb": 0.6070943397322817,
      "train_bytes": 28354.075,
      "train_loss_nats_per_token": 1.718469203151316
    },
    {
      "epoch": 0.20462984909536225,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0006817099149621182,
      "loss": 1.722024917602539,
      "step": 104080,
      "train_bpb": 0.6107176696748081,
      "train_bytes": 28038.43125,
      "train_loss_nats_per_token": 1.7226993776856738
    },
    {
      "epoch": 0.20476152853493457,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0006811873126339372,
      "loss": 1.717633056640625,
      "step": 104100,
      "train_bpb": 0.6065206737175516,
      "train_bytes": 28323.31875,
      "train_loss_nats_per_token": 1.7186852927397682
    },
    {
      "epoch": 0.20489320797450686,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0006806648518451768,
      "loss": 1.7206892013549804,
      "step": 104120,
      "train_bpb": 0.606998040022982,
      "train_bytes": 28357.41875,
      "train_loss_nats_per_token": 1.7209770596040905
    },
    {
      "epoch": 0.20502488741407918,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00068014253268615,
      "loss": 1.717148208618164,
      "step": 104140,
      "train_bpb": 0.6028512401781222,
      "train_bytes": 28643.86875,
      "train_loss_nats_per_token": 1.7177807153502116
    },
    {
      "epoch": 0.20515656685365147,
      "grad_norm": 0.10986328125,
      "learning_rate": 0.0006796203552471431,
      "loss": 1.7080574035644531,
      "step": 104160,
      "train_bpb": 0.6039673340134251,
      "train_bytes": 28523.31875,
      "train_loss_nats_per_token": 1.7084908737158724
    },
    {
      "epoch": 0.2052882462932238,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0006790983196184203,
      "loss": 1.7154590606689453,
      "step": 104180,
      "train_bpb": 0.6072924625675665,
      "train_bytes": 28294.5875,
      "train_loss_nats_per_token": 1.7164370304386645
    },
    {
      "epoch": 0.20541992573279608,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0006785764258902198,
      "loss": 1.7157169342041017,
      "step": 104200,
      "train_bpb": 0.6069901089238923,
      "train_bytes": 28274.59375,
      "train_loss_nats_per_token": 1.7160116299423764
    },
    {
      "epoch": 0.2055516051723684,
      "grad_norm": 0.224609375,
      "learning_rate": 0.0006780546741527568,
      "loss": 1.7221359252929687,
      "step": 104220,
      "train_bpb": 0.6085568123285146,
      "train_bytes": 28461.4125,
      "train_loss_nats_per_token": 1.7223856577115264
    },
    {
      "epoch": 0.2056832846119407,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0006775330644962197,
      "loss": 1.7172903060913085,
      "step": 104240,
      "train_bpb": 0.6088111822625069,
      "train_bytes": 28060.11875,
      "train_loss_nats_per_token": 1.7176355969445585
    },
    {
      "epoch": 0.205814964051513,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0006770115970107743,
      "loss": 1.7254148483276368,
      "step": 104260,
      "train_bpb": 0.6070506986250913,
      "train_bytes": 28538.2375,
      "train_loss_nats_per_token": 1.7260323699065119
    },
    {
      "epoch": 0.2059466434910853,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0006764902717865613,
      "loss": 1.7135713577270508,
      "step": 104280,
      "train_bpb": 0.6060553312440886,
      "train_bytes": 28219.375,
      "train_loss_nats_per_token": 1.713415099034022
    },
    {
      "epoch": 0.20607832293065761,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.0006759690889136959,
      "loss": 1.723668098449707,
      "step": 104300,
      "train_bpb": 0.608797068477937,
      "train_bytes": 28260.6375,
      "train_loss_nats_per_token": 1.7239572171973694
    },
    {
      "epoch": 0.2062100023702299,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0006754480484822698,
      "loss": 1.7208890914916992,
      "step": 104320,
      "train_bpb": 0.6085331743865959,
      "train_bytes": 28185.9375,
      "train_loss_nats_per_token": 1.7222682373859655
    },
    {
      "epoch": 0.20634168180980222,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0006749271505823498,
      "loss": 1.7136833190917968,
      "step": 104340,
      "train_bpb": 0.6052423230915466,
      "train_bytes": 28379.3,
      "train_loss_nats_per_token": 1.7140396699401839
    },
    {
      "epoch": 0.20647336124937452,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0006744063953039784,
      "loss": 1.717932891845703,
      "step": 104360,
      "train_bpb": 0.6077435772854018,
      "train_bytes": 28645.525,
      "train_loss_nats_per_token": 1.7182246952743998
    },
    {
      "epoch": 0.20660504068894683,
      "grad_norm": 0.142578125,
      "learning_rate": 0.000673885782737172,
      "loss": 1.715194320678711,
      "step": 104380,
      "train_bpb": 0.6106789635782507,
      "train_bytes": 28133.35,
      "train_loss_nats_per_token": 1.7159394704400501
    },
    {
      "epoch": 0.20673672012851912,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.000673365312971924,
      "loss": 1.7236183166503907,
      "step": 104400,
      "train_bpb": 0.6095250477196936,
      "train_bytes": 28468.9625,
      "train_loss_nats_per_token": 1.7244039083025056
    },
    {
      "epoch": 0.20686839956809144,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0006728449860982028,
      "loss": 1.7086633682250976,
      "step": 104420,
      "train_bpb": 0.6052335881405524,
      "train_bytes": 28072.7875,
      "train_loss_nats_per_token": 1.709272295590271
    },
    {
      "epoch": 0.20700007900766373,
      "grad_norm": 0.15625,
      "learning_rate": 0.0006723248022059507,
      "loss": 1.718056869506836,
      "step": 104440,
      "train_bpb": 0.6078974375375952,
      "train_bytes": 28265.81875,
      "train_loss_nats_per_token": 1.718895696920498
    },
    {
      "epoch": 0.20713175844723605,
      "grad_norm": 0.140625,
      "learning_rate": 0.000671804761385087,
      "loss": 1.7203866958618164,
      "step": 104460,
      "train_bpb": 0.6078529911733598,
      "train_bytes": 27976.875,
      "train_loss_nats_per_token": 1.7212910760658438
    },
    {
      "epoch": 0.20726343788680834,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0006712848637255061,
      "loss": 1.7186227798461915,
      "step": 104480,
      "train_bpb": 0.6083234649409834,
      "train_bytes": 28354.45,
      "train_loss_nats_per_token": 1.7196538675449065
    },
    {
      "epoch": 0.20739511732638066,
      "grad_norm": 0.13671875,
      "learning_rate": 0.000670765109317076,
      "loss": 1.714698600769043,
      "step": 104500,
      "train_bpb": 0.6035145284074829,
      "train_bytes": 28288.73125,
      "train_loss_nats_per_token": 1.7134886047982798
    },
    {
      "epoch": 0.20752679676595295,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0006702454982496419,
      "loss": 1.7108556747436523,
      "step": 104520,
      "train_bpb": 0.6068797012674186,
      "train_bytes": 28136.6125,
      "train_loss_nats_per_token": 1.7117234696275039
    },
    {
      "epoch": 0.20765847620552527,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0006697260306130231,
      "loss": 1.7162343978881835,
      "step": 104540,
      "train_bpb": 0.6081083874660773,
      "train_bytes": 28079.6125,
      "train_loss_nats_per_token": 1.7173148801739393
    },
    {
      "epoch": 0.20779015564509756,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0006692067064970153,
      "loss": 1.711166763305664,
      "step": 104560,
      "train_bpb": 0.6082720758211148,
      "train_bytes": 28037.41875,
      "train_loss_nats_per_token": 1.7115825511977243
    },
    {
      "epoch": 0.20792183508466988,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0006686875259913871,
      "loss": 1.7155162811279296,
      "step": 104580,
      "train_bpb": 0.6070974937047426,
      "train_bytes": 28130.28125,
      "train_loss_nats_per_token": 1.7161546767046099
    },
    {
      "epoch": 0.20805351452424217,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0006681684891858845,
      "loss": 1.7250349044799804,
      "step": 104600,
      "train_bpb": 0.610328567804427,
      "train_bytes": 28492.9375,
      "train_loss_nats_per_token": 1.724828466642419
    },
    {
      "epoch": 0.2081851939638145,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0006676495961702283,
      "loss": 1.6998973846435548,
      "step": 104620,
      "train_bpb": 0.6033033120833902,
      "train_bytes": 28116.3375,
      "train_loss_nats_per_token": 1.7002516721467196
    },
    {
      "epoch": 0.20831687340338678,
      "grad_norm": 0.1103515625,
      "learning_rate": 0.0006671308470341128,
      "loss": 1.7127973556518554,
      "step": 104640,
      "train_bpb": 0.6121916854357496,
      "train_bytes": 27891.8375,
      "train_loss_nats_per_token": 1.7133416493494427
    },
    {
      "epoch": 0.2084485528429591,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0006666122418672097,
      "loss": 1.7132017135620117,
      "step": 104660,
      "train_bpb": 0.6068427515776822,
      "train_bytes": 28560.4875,
      "train_loss_nats_per_token": 1.7153429044289568
    },
    {
      "epoch": 0.20858023228253142,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0006660937807591639,
      "loss": 1.7190113067626953,
      "step": 104680,
      "train_bpb": 0.609068703296414,
      "train_bytes": 28368.88125,
      "train_loss_nats_per_token": 1.7197427972041113
    },
    {
      "epoch": 0.2087119117221037,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0006655754637995975,
      "loss": 1.7126890182495118,
      "step": 104700,
      "train_bpb": 0.6072472766064669,
      "train_bytes": 28196.9375,
      "train_loss_nats_per_token": 1.7128003695131824
    },
    {
      "epoch": 0.20884359116167603,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.000665057291078105,
      "loss": 1.7365146636962892,
      "step": 104720,
      "train_bpb": 0.6112608092293279,
      "train_bytes": 28503.71875,
      "train_loss_nats_per_token": 1.7370571266114427
    },
    {
      "epoch": 0.20897527060124832,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0006645392626842579,
      "loss": 1.7157608032226563,
      "step": 104740,
      "train_bpb": 0.6098014247885815,
      "train_bytes": 27991.96875,
      "train_loss_nats_per_token": 1.7161886964428417
    },
    {
      "epoch": 0.20910695004082064,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.000664021378707603,
      "loss": 1.7124557495117188,
      "step": 104760,
      "train_bpb": 0.6035940475963457,
      "train_bytes": 28422.9875,
      "train_loss_nats_per_token": 1.7136561652459605
    },
    {
      "epoch": 0.20923862948039293,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.0006635036392376605,
      "loss": 1.7110015869140625,
      "step": 104780,
      "train_bpb": 0.6069905420956747,
      "train_bytes": 28255.5375,
      "train_loss_nats_per_token": 1.7117143144150704
    },
    {
      "epoch": 0.20937030891996525,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0006629860443639263,
      "loss": 1.718672752380371,
      "step": 104800,
      "train_bpb": 0.6116703699718589,
      "train_bytes": 28148.575,
      "train_loss_nats_per_token": 1.7196522410700374
    },
    {
      "epoch": 0.20950198835953754,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0006624685941758722,
      "loss": 1.7159822463989258,
      "step": 104820,
      "train_bpb": 0.6066963746058776,
      "train_bytes": 28430.2125,
      "train_loss_nats_per_token": 1.7171608516383947
    },
    {
      "epoch": 0.20963366779910986,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0006619512887629446,
      "loss": 1.715170669555664,
      "step": 104840,
      "train_bpb": 0.6054655550393834,
      "train_bytes": 28412.56875,
      "train_loss_nats_per_token": 1.7154178028597358
    },
    {
      "epoch": 0.20976534723868215,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0006614341282145633,
      "loss": 1.7151885986328126,
      "step": 104860,
      "train_bpb": 0.6085420389914445,
      "train_bytes": 28365.89375,
      "train_loss_nats_per_token": 1.7154129936916387
    },
    {
      "epoch": 0.20989702667825447,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0006609171126201253,
      "loss": 1.72290096282959,
      "step": 104880,
      "train_bpb": 0.6055309238460435,
      "train_bytes": 28502.05625,
      "train_loss_nats_per_token": 1.7234548217023935
    },
    {
      "epoch": 0.21002870611782676,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0006604002420690018,
      "loss": 1.712907600402832,
      "step": 104900,
      "train_bpb": 0.6108188925679475,
      "train_bytes": 28154.7875,
      "train_loss_nats_per_token": 1.7141501233435483
    },
    {
      "epoch": 0.21016038555739908,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0006598835166505376,
      "loss": 1.7082342147827148,
      "step": 104920,
      "train_bpb": 0.6023674546594125,
      "train_bytes": 28409.83125,
      "train_loss_nats_per_token": 1.708531201782115
    },
    {
      "epoch": 0.21029206499697137,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0006593669364540544,
      "loss": 1.7257144927978516,
      "step": 104940,
      "train_bpb": 0.6105565197668571,
      "train_bytes": 27991.9625,
      "train_loss_nats_per_token": 1.726430101962802
    },
    {
      "epoch": 0.21042374443654369,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.000658850501568848,
      "loss": 1.7068458557128907,
      "step": 104960,
      "train_bpb": 0.6073382947377641,
      "train_bytes": 27970.61875,
      "train_loss_nats_per_token": 1.707422003861775
    },
    {
      "epoch": 0.21055542387611598,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0006583342120841881,
      "loss": 1.7119768142700196,
      "step": 104980,
      "train_bpb": 0.609153224164252,
      "train_bytes": 28228.0375,
      "train_loss_nats_per_token": 1.713918636385722
    },
    {
      "epoch": 0.2106871033156883,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0006578180680893208,
      "loss": 1.7148191452026367,
      "step": 105000,
      "train_bpb": 0.6056955443810895,
      "train_bytes": 28595.81875,
      "train_loss_nats_per_token": 1.7155607845218437
    },
    {
      "epoch": 0.2106871033156883,
      "eval_loss": 1.5313955545425415,
      "eval_runtime": 74.3996,
      "eval_samples_per_second": 13.441,
      "eval_steps_per_second": 13.441,
      "step": 105000,
      "train_bpb": 0.5404281661694634,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5389301207639081
    },
    {
      "epoch": 0.2108187827552606,
      "grad_norm": 0.201171875,
      "learning_rate": 0.0006573020696734663,
      "loss": 1.7184391021728516,
      "step": 105020,
      "train_bpb": 0.6086307189959699,
      "train_bytes": 28439.05,
      "train_loss_nats_per_token": 1.7196978443963595
    },
    {
      "epoch": 0.2109504621948329,
      "grad_norm": 0.125,
      "learning_rate": 0.0006567862169258203,
      "loss": 1.7126445770263672,
      "step": 105040,
      "train_bpb": 0.6079568485149714,
      "train_bytes": 28053.7875,
      "train_loss_nats_per_token": 1.713207415769085
    },
    {
      "epoch": 0.2110821416344052,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0006562705099355517,
      "loss": 1.7167131423950195,
      "step": 105060,
      "train_bpb": 0.6070053249284664,
      "train_bytes": 28248.1,
      "train_loss_nats_per_token": 1.7158497589245594
    },
    {
      "epoch": 0.21121382107397751,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0006557549487918058,
      "loss": 1.7219013214111327,
      "step": 105080,
      "train_bpb": 0.6080148895837352,
      "train_bytes": 28326.89375,
      "train_loss_nats_per_token": 1.7225557167672598
    },
    {
      "epoch": 0.2113455005135498,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0006552395335837028,
      "loss": 1.7111183166503907,
      "step": 105100,
      "train_bpb": 0.6082243649619615,
      "train_bytes": 28119.88125,
      "train_loss_nats_per_token": 1.7105934532208475
    },
    {
      "epoch": 0.21147717995312212,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0006547242644003357,
      "loss": 1.6997434616088867,
      "step": 105120,
      "train_bpb": 0.6037413146662185,
      "train_bytes": 28203.3,
      "train_loss_nats_per_token": 1.7003786236208807
    },
    {
      "epoch": 0.21160885939269442,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0006542091413307744,
      "loss": 1.7173521041870117,
      "step": 105140,
      "train_bpb": 0.6069357281238695,
      "train_bytes": 28271.15,
      "train_loss_nats_per_token": 1.7172079160795533
    },
    {
      "epoch": 0.21174053883226673,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.0006536941644640626,
      "loss": 1.7219736099243164,
      "step": 105160,
      "train_bpb": 0.6090795156389665,
      "train_bytes": 28336.58125,
      "train_loss_nats_per_token": 1.7222930458611219
    },
    {
      "epoch": 0.21187221827183902,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0006531793338892195,
      "loss": 1.7104660034179688,
      "step": 105180,
      "train_bpb": 0.606042367057452,
      "train_bytes": 28472.375,
      "train_loss_nats_per_token": 1.7108107419451464
    },
    {
      "epoch": 0.21200389771141134,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0006526646496952371,
      "loss": 1.7128587722778321,
      "step": 105200,
      "train_bpb": 0.6081928125377599,
      "train_bytes": 28376.63125,
      "train_loss_nats_per_token": 1.7136939868966279
    },
    {
      "epoch": 0.21213557715098363,
      "grad_norm": 0.2119140625,
      "learning_rate": 0.000652150111971084,
      "loss": 1.7185462951660155,
      "step": 105220,
      "train_bpb": 0.6140818601255534,
      "train_bytes": 27814.45625,
      "train_loss_nats_per_token": 1.7191879477874983
    },
    {
      "epoch": 0.21226725659055595,
      "grad_norm": 0.353515625,
      "learning_rate": 0.0006516357208057036,
      "loss": 1.7107929229736327,
      "step": 105240,
      "train_bpb": 0.6046300156473288,
      "train_bytes": 28517.6125,
      "train_loss_nats_per_token": 1.7117483992844575
    },
    {
      "epoch": 0.21239893603012824,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0006511214762880119,
      "loss": 1.7106115341186523,
      "step": 105260,
      "train_bpb": 0.6060287779184297,
      "train_bytes": 28441.875,
      "train_loss_nats_per_token": 1.7104138138768605
    },
    {
      "epoch": 0.21253061546970056,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0006506073785069015,
      "loss": 1.7227563858032227,
      "step": 105280,
      "train_bpb": 0.6112777008692447,
      "train_bytes": 28274.95625,
      "train_loss_nats_per_token": 1.7240129325723368
    },
    {
      "epoch": 0.21266229490927285,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0006500934275512393,
      "loss": 1.7121034622192384,
      "step": 105300,
      "train_bpb": 0.6096932805614155,
      "train_bytes": 28198.80625,
      "train_loss_nats_per_token": 1.7134030336168302
    },
    {
      "epoch": 0.21279397434884517,
      "grad_norm": 0.109375,
      "learning_rate": 0.0006495796235098659,
      "loss": 1.7229665756225585,
      "step": 105320,
      "train_bpb": 0.6061609834205433,
      "train_bytes": 28647.26875,
      "train_loss_nats_per_token": 1.7228454617363234
    },
    {
      "epoch": 0.21292565378841746,
      "grad_norm": 0.1728515625,
      "learning_rate": 0.0006490659664715973,
      "loss": 1.7228912353515624,
      "step": 105340,
      "train_bpb": 0.6072038100708202,
      "train_bytes": 28397.4375,
      "train_loss_nats_per_token": 1.7234679193947244
    },
    {
      "epoch": 0.21305733322798978,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0006485524565252239,
      "loss": 1.7223651885986329,
      "step": 105360,
      "train_bpb": 0.6090724734775174,
      "train_bytes": 28408.48125,
      "train_loss_nats_per_token": 1.7230184321199014
    },
    {
      "epoch": 0.2131890126675621,
      "grad_norm": 0.11376953125,
      "learning_rate": 0.0006480390937595116,
      "loss": 1.723835563659668,
      "step": 105380,
      "train_bpb": 0.605258354602039,
      "train_bytes": 28717.78125,
      "train_loss_nats_per_token": 1.7238989532144982
    },
    {
      "epoch": 0.2133206921071344,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0006475258782631986,
      "loss": 1.7193965911865234,
      "step": 105400,
      "train_bpb": 0.6076250216372088,
      "train_bytes": 28214.0625,
      "train_loss_nats_per_token": 1.7193389089886302
    },
    {
      "epoch": 0.2134523715467067,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0006470128101249995,
      "loss": 1.720841407775879,
      "step": 105420,
      "train_bpb": 0.6092799076181684,
      "train_bytes": 28474.95625,
      "train_loss_nats_per_token": 1.7207628649016107
    },
    {
      "epoch": 0.213584050986279,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0006464998894336035,
      "loss": 1.7101099014282226,
      "step": 105440,
      "train_bpb": 0.6052083214562404,
      "train_bytes": 28417.4125,
      "train_loss_nats_per_token": 1.7097820935272705
    },
    {
      "epoch": 0.21371573042585132,
      "grad_norm": 0.125,
      "learning_rate": 0.0006459871162776728,
      "loss": 1.7060575485229492,
      "step": 105460,
      "train_bpb": 0.6044234652939419,
      "train_bytes": 28131.56875,
      "train_loss_nats_per_token": 1.7065554277112114
    },
    {
      "epoch": 0.2138474098654236,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0006454744907458454,
      "loss": 1.7157722473144532,
      "step": 105480,
      "train_bpb": 0.6110600751074083,
      "train_bytes": 28156.0875,
      "train_loss_nats_per_token": 1.716788551562358
    },
    {
      "epoch": 0.21397908930499593,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0006449620129267335,
      "loss": 1.7314117431640625,
      "step": 105500,
      "train_bpb": 0.6106394668002549,
      "train_bytes": 28315.6125,
      "train_loss_nats_per_token": 1.732416066208423
    },
    {
      "epoch": 0.21411076874456822,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0006444496829089239,
      "loss": 1.7174560546875,
      "step": 105520,
      "train_bpb": 0.6061648479544343,
      "train_bytes": 28383.09375,
      "train_loss_nats_per_token": 1.7168848223896285
    },
    {
      "epoch": 0.21424244818414054,
      "grad_norm": 0.13671875,
      "learning_rate": 0.000643937500780977,
      "loss": 1.716644287109375,
      "step": 105540,
      "train_bpb": 0.6081077960167981,
      "train_bytes": 28216.04375,
      "train_loss_nats_per_token": 1.7179703422502166
    },
    {
      "epoch": 0.21437412762371283,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0006434254666314288,
      "loss": 1.709830093383789,
      "step": 105560,
      "train_bpb": 0.6041178154529663,
      "train_bytes": 28737.86875,
      "train_loss_nats_per_token": 1.7111620838066743
    },
    {
      "epoch": 0.21450580706328515,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0006429135805487893,
      "loss": 1.7154211044311523,
      "step": 105580,
      "train_bpb": 0.6076422746422718,
      "train_bytes": 28264.9375,
      "train_loss_nats_per_token": 1.7154128139232139
    },
    {
      "epoch": 0.21463748650285744,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0006424018426215422,
      "loss": 1.711317253112793,
      "step": 105600,
      "train_bpb": 0.6070095470424127,
      "train_bytes": 28460.01875,
      "train_loss_nats_per_token": 1.7118286750896383
    },
    {
      "epoch": 0.21476916594242976,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0006418902529381464,
      "loss": 1.712839698791504,
      "step": 105620,
      "train_bpb": 0.6055628010261959,
      "train_bytes": 28353.85,
      "train_loss_nats_per_token": 1.7129900835658927
    },
    {
      "epoch": 0.21490084538200205,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0006413788115870355,
      "loss": 1.7264759063720703,
      "step": 105640,
      "train_bpb": 0.6083656257340693,
      "train_bytes": 28401.2,
      "train_loss_nats_per_token": 1.7271204481983646
    },
    {
      "epoch": 0.21503252482157437,
      "grad_norm": 0.1162109375,
      "learning_rate": 0.0006408675186566161,
      "loss": 1.72425537109375,
      "step": 105660,
      "train_bpb": 0.6109913165496357,
      "train_bytes": 28191.69375,
      "train_loss_nats_per_token": 1.7246628002478475
    },
    {
      "epoch": 0.21516420426114666,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0006403563742352701,
      "loss": 1.715690803527832,
      "step": 105680,
      "train_bpb": 0.6049433436277353,
      "train_bytes": 28624.58125,
      "train_loss_nats_per_token": 1.7166761268786432
    },
    {
      "epoch": 0.21529588370071898,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0006398453784113539,
      "loss": 1.7067975997924805,
      "step": 105700,
      "train_bpb": 0.6029457543103154,
      "train_bytes": 28481.0875,
      "train_loss_nats_per_token": 1.7071716798916388
    },
    {
      "epoch": 0.21542756314029127,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0006393345312731983,
      "loss": 1.7249105453491211,
      "step": 105720,
      "train_bpb": 0.6109932766083809,
      "train_bytes": 28297.3375,
      "train_loss_nats_per_token": 1.7257000517086654
    },
    {
      "epoch": 0.21555924257986359,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.000638823832909107,
      "loss": 1.7119312286376953,
      "step": 105740,
      "train_bpb": 0.6090670554617084,
      "train_bytes": 28121.0625,
      "train_loss_nats_per_token": 1.7126909902277765
    },
    {
      "epoch": 0.21569092201943588,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0006383132834073596,
      "loss": 1.719743537902832,
      "step": 105760,
      "train_bpb": 0.6045159677561331,
      "train_bytes": 28302.3625,
      "train_loss_nats_per_token": 1.721238448525203
    },
    {
      "epoch": 0.2158226014590082,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0006378028828562095,
      "loss": 1.7168598175048828,
      "step": 105780,
      "train_bpb": 0.6100905827269262,
      "train_bytes": 28041.50625,
      "train_loss_nats_per_token": 1.7179012249157832
    },
    {
      "epoch": 0.2159542808985805,
      "grad_norm": 0.140625,
      "learning_rate": 0.0006372926313438834,
      "loss": 1.7206666946411133,
      "step": 105800,
      "train_bpb": 0.6099645880202011,
      "train_bytes": 28167.71875,
      "train_loss_nats_per_token": 1.7212218849039085
    },
    {
      "epoch": 0.2160859603381528,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0006367825289585839,
      "loss": 1.710561752319336,
      "step": 105820,
      "train_bpb": 0.6076137635343887,
      "train_bytes": 28290.7875,
      "train_loss_nats_per_token": 1.7110209171987005
    },
    {
      "epoch": 0.2162176397777251,
      "grad_norm": 0.1796875,
      "learning_rate": 0.0006362725757884866,
      "loss": 1.7135860443115234,
      "step": 105840,
      "train_bpb": 0.6014844738336333,
      "train_bytes": 28755.325,
      "train_loss_nats_per_token": 1.7134698290421209
    },
    {
      "epoch": 0.21634931921729741,
      "grad_norm": 0.10791015625,
      "learning_rate": 0.000635762771921742,
      "loss": 1.7174055099487304,
      "step": 105860,
      "train_bpb": 0.6068942187857979,
      "train_bytes": 28774.24375,
      "train_loss_nats_per_token": 1.7174609954882507
    },
    {
      "epoch": 0.2164809986568697,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0006352531174464737,
      "loss": 1.721501350402832,
      "step": 105880,
      "train_bpb": 0.6073871293481453,
      "train_bytes": 28512.675,
      "train_loss_nats_per_token": 1.721936707307247
    },
    {
      "epoch": 0.21661267809644202,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0006347436124507806,
      "loss": 1.7168485641479492,
      "step": 105900,
      "train_bpb": 0.6048907399023281,
      "train_bytes": 28574.74375,
      "train_loss_nats_per_token": 1.7170915649272134
    },
    {
      "epoch": 0.21674435753601431,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0006342342570227363,
      "loss": 1.722616958618164,
      "step": 105920,
      "train_bpb": 0.6092673348694064,
      "train_bytes": 28382.55625,
      "train_loss_nats_per_token": 1.723854201638231
    },
    {
      "epoch": 0.21687603697558663,
      "grad_norm": 0.125,
      "learning_rate": 0.0006337250512503861,
      "loss": 1.7116640090942383,
      "step": 105940,
      "train_bpb": 0.6042283001021773,
      "train_bytes": 28423.68125,
      "train_loss_nats_per_token": 1.712783678650915
    },
    {
      "epoch": 0.21700771641515892,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0006332159952217518,
      "loss": 1.7148097991943358,
      "step": 105960,
      "train_bpb": 0.6096080958853621,
      "train_bytes": 27890.125,
      "train_loss_nats_per_token": 1.7149778161514353
    },
    {
      "epoch": 0.21713939585473124,
      "grad_norm": 0.140625,
      "learning_rate": 0.0006327070890248285,
      "loss": 1.7235742568969727,
      "step": 105980,
      "train_bpb": 0.6137337629750743,
      "train_bytes": 27810.5875,
      "train_loss_nats_per_token": 1.7246800459808125
    },
    {
      "epoch": 0.21727107529430353,
      "grad_norm": 0.125,
      "learning_rate": 0.0006321983327475855,
      "loss": 1.7198221206665039,
      "step": 106000,
      "train_bpb": 0.6071684165187033,
      "train_bytes": 28369.275,
      "train_loss_nats_per_token": 1.720044606062221
    },
    {
      "epoch": 0.21727107529430353,
      "eval_loss": 1.5308316946029663,
      "eval_runtime": 73.4018,
      "eval_samples_per_second": 13.624,
      "eval_steps_per_second": 13.624,
      "step": 106000,
      "train_bpb": 0.5402446834069345,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5384076329152505
    },
    {
      "epoch": 0.21740275473387585,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0006316897264779656,
      "loss": 1.7347684860229493,
      "step": 106020,
      "train_bpb": 0.6151869426617734,
      "train_bytes": 27904.575,
      "train_loss_nats_per_token": 1.7348400816292058
    },
    {
      "epoch": 0.21753443417344817,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0006311812703038863,
      "loss": 1.711555290222168,
      "step": 106040,
      "train_bpb": 0.6058534206586828,
      "train_bytes": 28481.09375,
      "train_loss_nats_per_token": 1.712440666800259
    },
    {
      "epoch": 0.21766611361302046,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0006306729643132398,
      "loss": 1.7205759048461915,
      "step": 106060,
      "train_bpb": 0.6086457724682703,
      "train_bytes": 28269.975,
      "train_loss_nats_per_token": 1.7211141659192837
    },
    {
      "epoch": 0.21779779305259278,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0006301648085938902,
      "loss": 1.703644371032715,
      "step": 106080,
      "train_bpb": 0.6055725420112517,
      "train_bytes": 28171.00625,
      "train_loss_nats_per_token": 1.7042983228005564
    },
    {
      "epoch": 0.21792947249216507,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0006296568032336777,
      "loss": 1.7131624221801758,
      "step": 106100,
      "train_bpb": 0.6079760804339409,
      "train_bytes": 28073.16875,
      "train_loss_nats_per_token": 1.7133186217659988
    },
    {
      "epoch": 0.2180611519317374,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0006291489483204161,
      "loss": 1.7178789138793946,
      "step": 106120,
      "train_bpb": 0.6069951088512392,
      "train_bytes": 28192.23125,
      "train_loss_nats_per_token": 1.7181292513345352
    },
    {
      "epoch": 0.21819283137130968,
      "grad_norm": 0.125,
      "learning_rate": 0.000628641243941892,
      "loss": 1.7275550842285157,
      "step": 106140,
      "train_bpb": 0.608776426280242,
      "train_bytes": 28524.35625,
      "train_loss_nats_per_token": 1.7272389964155155
    },
    {
      "epoch": 0.218324510810882,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0006281336901858672,
      "loss": 1.7102323532104493,
      "step": 106160,
      "train_bpb": 0.6066366002081836,
      "train_bytes": 28049.24375,
      "train_loss_nats_per_token": 1.7104696335635023
    },
    {
      "epoch": 0.2184561902504543,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0006276262871400772,
      "loss": 1.7197237014770508,
      "step": 106180,
      "train_bpb": 0.6079175675017356,
      "train_bytes": 28063.64375,
      "train_loss_nats_per_token": 1.7198688085562717
    },
    {
      "epoch": 0.2185878696900266,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.000627119034892232,
      "loss": 1.710196304321289,
      "step": 106200,
      "train_bpb": 0.6037517366121394,
      "train_bytes": 28349.80625,
      "train_loss_nats_per_token": 1.7113305017276688
    },
    {
      "epoch": 0.2187195491295989,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0006266119335300136,
      "loss": 1.7328311920166015,
      "step": 106220,
      "train_bpb": 0.6105676536970011,
      "train_bytes": 28369.48125,
      "train_loss_nats_per_token": 1.7334890669780711
    },
    {
      "epoch": 0.21885122856917122,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.00062610498314108,
      "loss": 1.72625675201416,
      "step": 106240,
      "train_bpb": 0.6129195573099393,
      "train_bytes": 28253.00625,
      "train_loss_nats_per_token": 1.726891538563
    },
    {
      "epoch": 0.2189829080087435,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0006255981838130625,
      "loss": 1.721968650817871,
      "step": 106260,
      "train_bpb": 0.6077645010534413,
      "train_bytes": 28514.6625,
      "train_loss_nats_per_token": 1.7228378181962571
    },
    {
      "epoch": 0.21911458744831583,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0006250915356335653,
      "loss": 1.7184537887573241,
      "step": 106280,
      "train_bpb": 0.6054653154570201,
      "train_bytes": 28524.275,
      "train_loss_nats_per_token": 1.718840210632454
    },
    {
      "epoch": 0.21924626688788812,
      "grad_norm": 0.125,
      "learning_rate": 0.0006245850386901678,
      "loss": 1.7221229553222657,
      "step": 106300,
      "train_bpb": 0.6103876212546021,
      "train_bytes": 28269.3125,
      "train_loss_nats_per_token": 1.7225037145203446
    },
    {
      "epoch": 0.21937794632746044,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0006240786930704226,
      "loss": 1.708241081237793,
      "step": 106320,
      "train_bpb": 0.604222090445046,
      "train_bytes": 28377.44375,
      "train_loss_nats_per_token": 1.7095340660651195
    },
    {
      "epoch": 0.21950962576703273,
      "grad_norm": 0.1142578125,
      "learning_rate": 0.0006235724988618566,
      "loss": 1.7204246520996094,
      "step": 106340,
      "train_bpb": 0.6093290332871003,
      "train_bytes": 28130.55625,
      "train_loss_nats_per_token": 1.7216684673505858
    },
    {
      "epoch": 0.21964130520660505,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0006230664561519697,
      "loss": 1.7261377334594727,
      "step": 106360,
      "train_bpb": 0.6074104875946648,
      "train_bytes": 28156.4875,
      "train_loss_nats_per_token": 1.7260963204621371
    },
    {
      "epoch": 0.21977298464617734,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0006225605650282363,
      "loss": 1.7128728866577148,
      "step": 106380,
      "train_bpb": 0.604355954848753,
      "train_bytes": 28492.69375,
      "train_loss_nats_per_token": 1.7132347520800588
    },
    {
      "epoch": 0.21990466408574966,
      "grad_norm": 0.193359375,
      "learning_rate": 0.0006220548255781049,
      "loss": 1.7114456176757813,
      "step": 106400,
      "train_bpb": 0.6055712126459405,
      "train_bytes": 28166.4625,
      "train_loss_nats_per_token": 1.71196639651339
    },
    {
      "epoch": 0.22003634352532195,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0006215492378889964,
      "loss": 1.7033273696899414,
      "step": 106420,
      "train_bpb": 0.6056192543834268,
      "train_bytes": 28259.0125,
      "train_loss_nats_per_token": 1.7042124787158712
    },
    {
      "epoch": 0.22016802296489427,
      "grad_norm": 0.15625,
      "learning_rate": 0.0006210438020483067,
      "loss": 1.7130559921264648,
      "step": 106440,
      "train_bpb": 0.6090153907446476,
      "train_bytes": 27938.86875,
      "train_loss_nats_per_token": 1.714159486140583
    },
    {
      "epoch": 0.22029970240446656,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0006205385181434056,
      "loss": 1.7182605743408204,
      "step": 106460,
      "train_bpb": 0.6097897650513732,
      "train_bytes": 28477.7125,
      "train_loss_nats_per_token": 1.7189137550718085
    },
    {
      "epoch": 0.22043138184403888,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0006200333862616353,
      "loss": 1.7306060791015625,
      "step": 106480,
      "train_bpb": 0.6099628792540449,
      "train_bytes": 28501.25625,
      "train_loss_nats_per_token": 1.7317291547329328
    },
    {
      "epoch": 0.22056306128361117,
      "grad_norm": 0.1748046875,
      "learning_rate": 0.000619528406490313,
      "loss": 1.707911491394043,
      "step": 106500,
      "train_bpb": 0.6052353329971495,
      "train_bytes": 28184.65625,
      "train_loss_nats_per_token": 1.708585714361312
    },
    {
      "epoch": 0.22069474072318349,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0006190235789167292,
      "loss": 1.7175241470336915,
      "step": 106520,
      "train_bpb": 0.6097032135124539,
      "train_bytes": 28034.7,
      "train_loss_nats_per_token": 1.7188059732248262
    },
    {
      "epoch": 0.22082642016275578,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0006185189036281485,
      "loss": 1.7095260620117188,
      "step": 106540,
      "train_bpb": 0.6067560206995969,
      "train_bytes": 28391.86875,
      "train_loss_nats_per_token": 1.710710638033565
    },
    {
      "epoch": 0.2209580996023281,
      "grad_norm": 0.1171875,
      "learning_rate": 0.0006180143807118076,
      "loss": 1.7262996673583983,
      "step": 106560,
      "train_bpb": 0.6102336609639822,
      "train_bytes": 28681.475,
      "train_loss_nats_per_token": 1.7264419917169043
    },
    {
      "epoch": 0.22108977904190039,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0006175100102549188,
      "loss": 1.7179931640625,
      "step": 106580,
      "train_bpb": 0.6088277351096554,
      "train_bytes": 28170.9375,
      "train_loss_nats_per_token": 1.7185066845343284
    },
    {
      "epoch": 0.2212214584814727,
      "grad_norm": 0.1708984375,
      "learning_rate": 0.0006170057923446675,
      "loss": 1.7212869644165039,
      "step": 106600,
      "train_bpb": 0.6067967041741974,
      "train_bytes": 28217.5875,
      "train_loss_nats_per_token": 1.7224697566757865
    },
    {
      "epoch": 0.221353137921045,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.0006165017270682117,
      "loss": 1.7265918731689454,
      "step": 106620,
      "train_bpb": 0.6112220027860229,
      "train_bytes": 28324.05,
      "train_loss_nats_per_token": 1.7273336861531026
    },
    {
      "epoch": 0.22148481736061731,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0006159978145126842,
      "loss": 1.7138168334960937,
      "step": 106640,
      "train_bpb": 0.6056051158247945,
      "train_bytes": 28276.7375,
      "train_loss_nats_per_token": 1.7145229372856476
    },
    {
      "epoch": 0.2216164968001896,
      "grad_norm": 0.125,
      "learning_rate": 0.0006154940547651907,
      "loss": 1.7207897186279297,
      "step": 106660,
      "train_bpb": 0.6074348516955906,
      "train_bytes": 28376.0,
      "train_loss_nats_per_token": 1.7208912113359254
    },
    {
      "epoch": 0.22174817623976192,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0006149904479128118,
      "loss": 1.731637191772461,
      "step": 106680,
      "train_bpb": 0.6131447182261865,
      "train_bytes": 28450.81875,
      "train_loss_nats_per_token": 1.7316566354479104
    },
    {
      "epoch": 0.22187985567933424,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0006144869940425992,
      "loss": 1.7256509780883789,
      "step": 106700,
      "train_bpb": 0.6121164208705747,
      "train_bytes": 28169.3125,
      "train_loss_nats_per_token": 1.7262456356931302
    },
    {
      "epoch": 0.22201153511890653,
      "grad_norm": 0.11279296875,
      "learning_rate": 0.0006139836932415805,
      "loss": 1.718500518798828,
      "step": 106720,
      "train_bpb": 0.6064402317399992,
      "train_bytes": 28555.5875,
      "train_loss_nats_per_token": 1.7197121280779255
    },
    {
      "epoch": 0.22214321455847885,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.000613480545596756,
      "loss": 1.7180770874023437,
      "step": 106740,
      "train_bpb": 0.6074739942458086,
      "train_bytes": 27997.25625,
      "train_loss_nats_per_token": 1.7189123373941972
    },
    {
      "epoch": 0.22227489399805114,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0006129775511950988,
      "loss": 1.71234188079834,
      "step": 106760,
      "train_bpb": 0.6041945466364951,
      "train_bytes": 28504.43125,
      "train_loss_nats_per_token": 1.7128343203236955
    },
    {
      "epoch": 0.22240657343762346,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0006124747101235566,
      "loss": 1.7257137298583984,
      "step": 106780,
      "train_bpb": 0.6075921156423437,
      "train_bytes": 28485.7625,
      "train_loss_nats_per_token": 1.7262188986605083
    },
    {
      "epoch": 0.22253825287719575,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0006119720224690501,
      "loss": 1.7173526763916016,
      "step": 106800,
      "train_bpb": 0.6124320772139579,
      "train_bytes": 28212.34375,
      "train_loss_nats_per_token": 1.7175520180227075
    },
    {
      "epoch": 0.22266993231676807,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0006114694883184741,
      "loss": 1.7169116973876952,
      "step": 106820,
      "train_bpb": 0.6085430334505443,
      "train_bytes": 28260.95625,
      "train_loss_nats_per_token": 1.7179272376606627
    },
    {
      "epoch": 0.22280161175634036,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0006109671077586953,
      "loss": 1.7261365890502929,
      "step": 106840,
      "train_bpb": 0.6115298997091132,
      "train_bytes": 28155.5125,
      "train_loss_nats_per_token": 1.726214997730113
    },
    {
      "epoch": 0.22293329119591268,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0006104648808765555,
      "loss": 1.7346275329589844,
      "step": 106860,
      "train_bpb": 0.6118842821029548,
      "train_bytes": 28357.46875,
      "train_loss_nats_per_token": 1.7354188514778202
    },
    {
      "epoch": 0.22306497063548497,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0006099628077588698,
      "loss": 1.7280017852783203,
      "step": 106880,
      "train_bpb": 0.6096177656283989,
      "train_bytes": 28242.6125,
      "train_loss_nats_per_token": 1.7288321185618616
    },
    {
      "epoch": 0.2231966500750573,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0006094608884924251,
      "loss": 1.7108613967895507,
      "step": 106900,
      "train_bpb": 0.605892091127154,
      "train_bytes": 28205.7375,
      "train_loss_nats_per_token": 1.711934015191058
    },
    {
      "epoch": 0.22332832951462958,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0006089591231639832,
      "loss": 1.7271146774291992,
      "step": 106920,
      "train_bpb": 0.6101337373549888,
      "train_bytes": 28421.475,
      "train_loss_nats_per_token": 1.727165266746005
    },
    {
      "epoch": 0.2234600089542019,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0006084575118602797,
      "loss": 1.7289831161499023,
      "step": 106940,
      "train_bpb": 0.6101922042458854,
      "train_bytes": 28378.63125,
      "train_loss_nats_per_token": 1.7288601058789903
    },
    {
      "epoch": 0.2235916883937742,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0006079560546680216,
      "loss": 1.7106246948242188,
      "step": 106960,
      "train_bpb": 0.6035044602010841,
      "train_bytes": 28382.25,
      "train_loss_nats_per_token": 1.7111803329856288
    },
    {
      "epoch": 0.2237233678333465,
      "grad_norm": 0.2138671875,
      "learning_rate": 0.0006074547516738912,
      "loss": 1.7230567932128906,
      "step": 106980,
      "train_bpb": 0.6079578620518418,
      "train_bytes": 28486.175,
      "train_loss_nats_per_token": 1.7235444158536908
    },
    {
      "epoch": 0.2238550472729188,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0006069536029645431,
      "loss": 1.7194774627685547,
      "step": 107000,
      "train_bpb": 0.6075480830228257,
      "train_bytes": 28310.7,
      "train_loss_nats_per_token": 1.719380921402944
    },
    {
      "epoch": 0.2238550472729188,
      "eval_loss": 1.531146764755249,
      "eval_runtime": 75.127,
      "eval_samples_per_second": 13.311,
      "eval_steps_per_second": 13.311,
      "step": 107000,
      "train_bpb": 0.540344575315829,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5386920863857438
    },
    {
      "epoch": 0.22398672671249112,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0006064526086266064,
      "loss": 1.7129722595214845,
      "step": 107020,
      "train_bpb": 0.6060238557524759,
      "train_bytes": 28135.03125,
      "train_loss_nats_per_token": 1.7130473064767502
    },
    {
      "epoch": 0.2241184061520634,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0006059517687466813,
      "loss": 1.7201192855834961,
      "step": 107040,
      "train_bpb": 0.6079625460037821,
      "train_bytes": 28230.4375,
      "train_loss_nats_per_token": 1.720604671090603
    },
    {
      "epoch": 0.22425008559163573,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0006054510834113434,
      "loss": 1.7103466033935546,
      "step": 107060,
      "train_bpb": 0.607794346500623,
      "train_bytes": 28033.75625,
      "train_loss_nats_per_token": 1.7112645929477281
    },
    {
      "epoch": 0.22438176503120802,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0006049505527071413,
      "loss": 1.7159767150878906,
      "step": 107080,
      "train_bpb": 0.6081059605831196,
      "train_bytes": 28303.49375,
      "train_loss_nats_per_token": 1.7171043140756421
    },
    {
      "epoch": 0.22451344447078034,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0006044501767205951,
      "loss": 1.7224729537963868,
      "step": 107100,
      "train_bpb": 0.6078284181384068,
      "train_bytes": 28415.66875,
      "train_loss_nats_per_token": 1.722179429712414
    },
    {
      "epoch": 0.22464512391035263,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0006039499555382002,
      "loss": 1.7145904541015624,
      "step": 107120,
      "train_bpb": 0.6058403815774877,
      "train_bytes": 28539.9,
      "train_loss_nats_per_token": 1.7150094879641484
    },
    {
      "epoch": 0.22477680334992495,
      "grad_norm": 0.2265625,
      "learning_rate": 0.0006034498892464247,
      "loss": 1.7223855972290039,
      "step": 107140,
      "train_bpb": 0.6056501649054393,
      "train_bytes": 28475.9875,
      "train_loss_nats_per_token": 1.7220671699448793
    },
    {
      "epoch": 0.22490848278949724,
      "grad_norm": 0.123046875,
      "learning_rate": 0.00060294997793171,
      "loss": 1.7098611831665038,
      "step": 107160,
      "train_bpb": 0.6102240863854983,
      "train_bytes": 28222.2875,
      "train_loss_nats_per_token": 1.7106841430060542
    },
    {
      "epoch": 0.22504016222906956,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0006024502216804694,
      "loss": 1.7237878799438477,
      "step": 107180,
      "train_bpb": 0.6083927388948163,
      "train_bytes": 28805.725,
      "train_loss_nats_per_token": 1.7242229680920023
    },
    {
      "epoch": 0.22517184166864185,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0006019506205790909,
      "loss": 1.7231651306152345,
      "step": 107200,
      "train_bpb": 0.6061589437193531,
      "train_bytes": 28497.14375,
      "train_loss_nats_per_token": 1.7224247252806535
    },
    {
      "epoch": 0.22530352110821417,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.000601451174713936,
      "loss": 1.7151824951171875,
      "step": 107220,
      "train_bpb": 0.6086879165901974,
      "train_bytes": 28071.70625,
      "train_loss_nats_per_token": 1.7156390150808274
    },
    {
      "epoch": 0.22543520054778646,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0006009518841713373,
      "loss": 1.718593215942383,
      "step": 107240,
      "train_bpb": 0.6054656303407289,
      "train_bytes": 28490.33125,
      "train_loss_nats_per_token": 1.7194311665231992
    },
    {
      "epoch": 0.22556687998735878,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0006004527490376026,
      "loss": 1.7067354202270508,
      "step": 107260,
      "train_bpb": 0.6046234112281871,
      "train_bytes": 28122.48125,
      "train_loss_nats_per_token": 1.707150945472496
    },
    {
      "epoch": 0.22569855942693107,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0005999537693990123,
      "loss": 1.7171665191650392,
      "step": 107280,
      "train_bpb": 0.6060240860803633,
      "train_bytes": 28418.4375,
      "train_loss_nats_per_token": 1.7174629953824865
    },
    {
      "epoch": 0.22583023886650339,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0005994549453418188,
      "loss": 1.7168817520141602,
      "step": 107300,
      "train_bpb": 0.6030987828244794,
      "train_bytes": 28653.56875,
      "train_loss_nats_per_token": 1.717107342553484
    },
    {
      "epoch": 0.22596191830607568,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0005989562769522491,
      "loss": 1.719235610961914,
      "step": 107320,
      "train_bpb": 0.6092735278575014,
      "train_bytes": 28326.21875,
      "train_loss_nats_per_token": 1.7193217794442646
    },
    {
      "epoch": 0.226093597745648,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0005984577643165025,
      "loss": 1.7027622222900392,
      "step": 107340,
      "train_bpb": 0.6050358915370603,
      "train_bytes": 28639.35,
      "train_loss_nats_per_token": 1.7031542736191638
    },
    {
      "epoch": 0.22622527718522029,
      "grad_norm": 0.15625,
      "learning_rate": 0.0005979594075207522,
      "loss": 1.7075950622558593,
      "step": 107360,
      "train_bpb": 0.6088067512904561,
      "train_bytes": 28142.0625,
      "train_loss_nats_per_token": 1.708268593345014
    },
    {
      "epoch": 0.2263569566247926,
      "grad_norm": 0.16796875,
      "learning_rate": 0.0005974612066511427,
      "loss": 1.718845748901367,
      "step": 107380,
      "train_bpb": 0.6085645092050039,
      "train_bytes": 28370.4,
      "train_loss_nats_per_token": 1.7197925315061189
    },
    {
      "epoch": 0.22648863606436492,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0005969631617937934,
      "loss": 1.7165342330932618,
      "step": 107400,
      "train_bpb": 0.6055407316630953,
      "train_bytes": 28280.25,
      "train_loss_nats_per_token": 1.7180900000223331
    },
    {
      "epoch": 0.22662031550393721,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0005964652730347963,
      "loss": 1.7166091918945312,
      "step": 107420,
      "train_bpb": 0.6129294616458203,
      "train_bytes": 27741.5375,
      "train_loss_nats_per_token": 1.7162793412880808
    },
    {
      "epoch": 0.22675199494350953,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0005959675404602151,
      "loss": 1.7068731307983398,
      "step": 107440,
      "train_bpb": 0.6038066317490574,
      "train_bytes": 28234.7625,
      "train_loss_nats_per_token": 1.7077416626642623
    },
    {
      "epoch": 0.22688367438308182,
      "grad_norm": 0.15625,
      "learning_rate": 0.0005954699641560881,
      "loss": 1.7078624725341798,
      "step": 107460,
      "train_bpb": 0.6047255800550129,
      "train_bytes": 28407.15,
      "train_loss_nats_per_token": 1.708712690142789
    },
    {
      "epoch": 0.22701535382265414,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0005949725442084262,
      "loss": 1.7164400100708008,
      "step": 107480,
      "train_bpb": 0.6098975208029825,
      "train_bytes": 28160.68125,
      "train_loss_nats_per_token": 1.7171730882298255
    },
    {
      "epoch": 0.22714703326222643,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0005944752807032132,
      "loss": 1.7211917877197265,
      "step": 107500,
      "train_bpb": 0.6068826875791076,
      "train_bytes": 28493.9125,
      "train_loss_nats_per_token": 1.7224134563381002
    },
    {
      "epoch": 0.22727871270179875,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.000593978173726405,
      "loss": 1.7207727432250977,
      "step": 107520,
      "train_bpb": 0.6116564977814066,
      "train_bytes": 28068.675,
      "train_loss_nats_per_token": 1.7208085632866184
    },
    {
      "epoch": 0.22741039214137104,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0005934812233639316,
      "loss": 1.7179883956909179,
      "step": 107540,
      "train_bpb": 0.6059928906042447,
      "train_bytes": 28369.49375,
      "train_loss_nats_per_token": 1.7181989661148813
    },
    {
      "epoch": 0.22754207158094336,
      "grad_norm": 0.1904296875,
      "learning_rate": 0.0005929844297016962,
      "loss": 1.713050079345703,
      "step": 107560,
      "train_bpb": 0.6085750697151927,
      "train_bytes": 28198.675,
      "train_loss_nats_per_token": 1.7139536470411103
    },
    {
      "epoch": 0.22767375102051565,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0005924877928255729,
      "loss": 1.7252456665039062,
      "step": 107580,
      "train_bpb": 0.6072617898368788,
      "train_bytes": 28725.73125,
      "train_loss_nats_per_token": 1.7260860088075647
    },
    {
      "epoch": 0.22780543046008797,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0005919913128214107,
      "loss": 1.7135778427124024,
      "step": 107600,
      "train_bpb": 0.6070662453091336,
      "train_bytes": 28506.575,
      "train_loss_nats_per_token": 1.7137861656949074
    },
    {
      "epoch": 0.22793710989966026,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0005914949897750314,
      "loss": 1.7124414443969727,
      "step": 107620,
      "train_bpb": 0.6038411765846315,
      "train_bytes": 28388.325,
      "train_loss_nats_per_token": 1.7118292193383782
    },
    {
      "epoch": 0.22806878933923258,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0005909988237722279,
      "loss": 1.7223165512084961,
      "step": 107640,
      "train_bpb": 0.6090400390603576,
      "train_bytes": 28237.9375,
      "train_loss_nats_per_token": 1.7224631533944927
    },
    {
      "epoch": 0.22820046877880487,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0005905028148987676,
      "loss": 1.7325305938720703,
      "step": 107660,
      "train_bpb": 0.6159602584590637,
      "train_bytes": 28116.41875,
      "train_loss_nats_per_token": 1.7326275008592351
    },
    {
      "epoch": 0.2283321482183772,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0005900069632403906,
      "loss": 1.7155315399169921,
      "step": 107680,
      "train_bpb": 0.6079603679558379,
      "train_bytes": 28196.85,
      "train_loss_nats_per_token": 1.7164955148356025
    },
    {
      "epoch": 0.22846382765794948,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0005895112688828096,
      "loss": 1.7189224243164063,
      "step": 107700,
      "train_bpb": 0.6089833534259643,
      "train_bytes": 28080.7375,
      "train_loss_nats_per_token": 1.7195974088639538
    },
    {
      "epoch": 0.2285955070975218,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0005890157319117091,
      "loss": 1.7288179397583008,
      "step": 107720,
      "train_bpb": 0.6095050895906096,
      "train_bytes": 28456.875,
      "train_loss_nats_per_token": 1.7299584137519874
    },
    {
      "epoch": 0.2287271865370941,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0005885203524127479,
      "loss": 1.711867904663086,
      "step": 107740,
      "train_bpb": 0.6066732688103676,
      "train_bytes": 28125.6125,
      "train_loss_nats_per_token": 1.7128487683690088
    },
    {
      "epoch": 0.2288588659766664,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0005880251304715574,
      "loss": 1.7254388809204102,
      "step": 107760,
      "train_bpb": 0.6082794181148593,
      "train_bytes": 28500.125,
      "train_loss_nats_per_token": 1.7261431501856435
    },
    {
      "epoch": 0.2289905454162387,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0005875300661737405,
      "loss": 1.719898796081543,
      "step": 107780,
      "train_bpb": 0.6074391365873276,
      "train_bytes": 28291.7125,
      "train_loss_nats_per_token": 1.7206135288586537
    },
    {
      "epoch": 0.22912222485581102,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0005870351596048737,
      "loss": 1.7326292037963866,
      "step": 107800,
      "train_bpb": 0.609294305866522,
      "train_bytes": 28665.35625,
      "train_loss_nats_per_token": 1.734155385762589
    },
    {
      "epoch": 0.2292539042953833,
      "grad_norm": 0.185546875,
      "learning_rate": 0.0005865404108505067,
      "loss": 1.7211341857910156,
      "step": 107820,
      "train_bpb": 0.6075700369599729,
      "train_bytes": 28279.06875,
      "train_loss_nats_per_token": 1.7225057417247807
    },
    {
      "epoch": 0.22938558373495563,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.000586045819996162,
      "loss": 1.715780258178711,
      "step": 107840,
      "train_bpb": 0.6087886054437333,
      "train_bytes": 28068.975,
      "train_loss_nats_per_token": 1.7159593781733486
    },
    {
      "epoch": 0.22951726317452792,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0005855513871273329,
      "loss": 1.7156642913818358,
      "step": 107860,
      "train_bpb": 0.6071478446958256,
      "train_bytes": 28289.21875,
      "train_loss_nats_per_token": 1.7157428582221166
    },
    {
      "epoch": 0.22964894261410024,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0005850571123294876,
      "loss": 1.7107980728149415,
      "step": 107880,
      "train_bpb": 0.603932756369833,
      "train_bytes": 28597.7375,
      "train_loss_nats_per_token": 1.7111992342235725
    },
    {
      "epoch": 0.22978062205367253,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0005845629956880661,
      "loss": 1.7211137771606446,
      "step": 107900,
      "train_bpb": 0.6100204717062945,
      "train_bytes": 28319.625,
      "train_loss_nats_per_token": 1.7222758318814695
    },
    {
      "epoch": 0.22991230149324485,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0005840690372884806,
      "loss": 1.7104034423828125,
      "step": 107920,
      "train_bpb": 0.6041417167822587,
      "train_bytes": 28395.6875,
      "train_loss_nats_per_token": 1.7105670404424986
    },
    {
      "epoch": 0.23004398093281714,
      "grad_norm": 0.232421875,
      "learning_rate": 0.0005835752372161166,
      "loss": 1.7140724182128906,
      "step": 107940,
      "train_bpb": 0.610530208757801,
      "train_bytes": 28021.5125,
      "train_loss_nats_per_token": 1.7146360658885633
    },
    {
      "epoch": 0.23017566037238946,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0005830815955563322,
      "loss": 1.7130626678466796,
      "step": 107960,
      "train_bpb": 0.6074073854426076,
      "train_bytes": 28261.65625,
      "train_loss_nats_per_token": 1.7134673645299414
    },
    {
      "epoch": 0.23030733981196175,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0005825881123944585,
      "loss": 1.7238788604736328,
      "step": 107980,
      "train_bpb": 0.6113896450891583,
      "train_bytes": 28229.05625,
      "train_loss_nats_per_token": 1.7243691786153388
    },
    {
      "epoch": 0.23043901925153407,
      "grad_norm": 0.140625,
      "learning_rate": 0.0005820947878157974,
      "loss": 1.7218250274658202,
      "step": 108000,
      "train_bpb": 0.6058042966263257,
      "train_bytes": 28561.51875,
      "train_loss_nats_per_token": 1.7226107551756518
    },
    {
      "epoch": 0.23043901925153407,
      "eval_loss": 1.5308159589767456,
      "eval_runtime": 71.7721,
      "eval_samples_per_second": 13.933,
      "eval_steps_per_second": 13.933,
      "step": 108000,
      "train_bpb": 0.5402142248871843,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5383208988469441
    },
    {
      "epoch": 0.23057069869110636,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0005816016219056256,
      "loss": 1.7173263549804687,
      "step": 108020,
      "train_bpb": 0.6072077613773211,
      "train_bytes": 28415.55,
      "train_loss_nats_per_token": 1.7184839594495505
    },
    {
      "epoch": 0.23070237813067868,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0005811086147491916,
      "loss": 1.7182907104492187,
      "step": 108040,
      "train_bpb": 0.6095841830846398,
      "train_bytes": 28399.3,
      "train_loss_nats_per_token": 1.7184864957330104
    },
    {
      "epoch": 0.230834057570251,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0005806157664317155,
      "loss": 1.7292242050170898,
      "step": 108060,
      "train_bpb": 0.6057822205502007,
      "train_bytes": 28847.48125,
      "train_loss_nats_per_token": 1.7290781423205301
    },
    {
      "epoch": 0.23096573700982329,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0005801230770383912,
      "loss": 1.7089189529418944,
      "step": 108080,
      "train_bpb": 0.6059709434681935,
      "train_bytes": 28132.2625,
      "train_loss_nats_per_token": 1.7101140981752456
    },
    {
      "epoch": 0.2310974164493956,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0005796305466543848,
      "loss": 1.7077478408813476,
      "step": 108100,
      "train_bpb": 0.6094238820947753,
      "train_bytes": 27791.7125,
      "train_loss_nats_per_token": 1.70845658809599
    },
    {
      "epoch": 0.2312290958889679,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0005791381753648342,
      "loss": 1.7094461441040039,
      "step": 108120,
      "train_bpb": 0.6017607942084623,
      "train_bytes": 28659.475,
      "train_loss_nats_per_token": 1.7103334594832227
    },
    {
      "epoch": 0.2313607753285402,
      "grad_norm": 0.19921875,
      "learning_rate": 0.0005786459632548508,
      "loss": 1.7104011535644532,
      "step": 108140,
      "train_bpb": 0.6091945700452628,
      "train_bytes": 27961.00625,
      "train_loss_nats_per_token": 1.710765153996331
    },
    {
      "epoch": 0.2314924547681125,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0005781539104095179,
      "loss": 1.7173738479614258,
      "step": 108160,
      "train_bpb": 0.6101465068036765,
      "train_bytes": 28397.51875,
      "train_loss_nats_per_token": 1.7182398183832976
    },
    {
      "epoch": 0.23162413420768482,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0005776620169138921,
      "loss": 1.7154508590698243,
      "step": 108180,
      "train_bpb": 0.6065248901910016,
      "train_bytes": 28079.03125,
      "train_loss_nats_per_token": 1.7157560570647878
    },
    {
      "epoch": 0.23175581364725711,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0005771702828530009,
      "loss": 1.7011613845825195,
      "step": 108200,
      "train_bpb": 0.60558275143236,
      "train_bytes": 28214.59375,
      "train_loss_nats_per_token": 1.7018957106798343
    },
    {
      "epoch": 0.23188749308682943,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0005766787083118452,
      "loss": 1.7159107208251954,
      "step": 108220,
      "train_bpb": 0.6046435532523695,
      "train_bytes": 28513.9875,
      "train_loss_nats_per_token": 1.715926755845011
    },
    {
      "epoch": 0.23201917252640172,
      "grad_norm": 0.1845703125,
      "learning_rate": 0.0005761872933753992,
      "loss": 1.7060308456420898,
      "step": 108240,
      "train_bpb": 0.6067610908793791,
      "train_bytes": 28148.78125,
      "train_loss_nats_per_token": 1.70593498295387
    },
    {
      "epoch": 0.23215085196597404,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0005756960381286073,
      "loss": 1.7148975372314452,
      "step": 108260,
      "train_bpb": 0.6051350457845498,
      "train_bytes": 28440.1875,
      "train_loss_nats_per_token": 1.7154508034797862
    },
    {
      "epoch": 0.23228253140554633,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0005752049426563879,
      "loss": 1.705033302307129,
      "step": 108280,
      "train_bpb": 0.6026312502221415,
      "train_bytes": 28162.95,
      "train_loss_nats_per_token": 1.705495424861854
    },
    {
      "epoch": 0.23241421084511865,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0005747140070436324,
      "loss": 1.7145246505737304,
      "step": 108300,
      "train_bpb": 0.6052001290212886,
      "train_bytes": 28357.10625,
      "train_loss_nats_per_token": 1.7155266033971246
    },
    {
      "epoch": 0.23254589028469094,
      "grad_norm": 0.15625,
      "learning_rate": 0.0005742232313752023,
      "loss": 1.7111528396606446,
      "step": 108320,
      "train_bpb": 0.6059250454150029,
      "train_bytes": 28243.51875,
      "train_loss_nats_per_token": 1.7111210303339806
    },
    {
      "epoch": 0.23267756972426326,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0005737326157359334,
      "loss": 1.7131898880004883,
      "step": 108340,
      "train_bpb": 0.6068219245859537,
      "train_bytes": 28137.51875,
      "train_loss_nats_per_token": 1.714327360564871
    },
    {
      "epoch": 0.23280924916383555,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.000573242160210633,
      "loss": 1.72103271484375,
      "step": 108360,
      "train_bpb": 0.6122234404633434,
      "train_bytes": 28024.75625,
      "train_loss_nats_per_token": 1.7217080444223718
    },
    {
      "epoch": 0.23294092860340787,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0005727518648840816,
      "loss": 1.7125701904296875,
      "step": 108380,
      "train_bpb": 0.6076829190864393,
      "train_bytes": 28171.8125,
      "train_loss_nats_per_token": 1.7128139212548084
    },
    {
      "epoch": 0.23307260804298016,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0005722617298410301,
      "loss": 1.7110614776611328,
      "step": 108400,
      "train_bpb": 0.6054458713536132,
      "train_bytes": 28343.56875,
      "train_loss_nats_per_token": 1.711741109382912
    },
    {
      "epoch": 0.23320428748255248,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0005717717551662037,
      "loss": 1.7074930191040039,
      "step": 108420,
      "train_bpb": 0.6061113365174405,
      "train_bytes": 28114.65,
      "train_loss_nats_per_token": 1.7088628661858374
    },
    {
      "epoch": 0.23333596692212477,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0005712819409442996,
      "loss": 1.7253852844238282,
      "step": 108440,
      "train_bpb": 0.6109757383584711,
      "train_bytes": 28330.4125,
      "train_loss_nats_per_token": 1.7256749065693529
    },
    {
      "epoch": 0.2334676463616971,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0005707922872599854,
      "loss": 1.7128511428833009,
      "step": 108460,
      "train_bpb": 0.6052645900664183,
      "train_bytes": 28469.88125,
      "train_loss_nats_per_token": 1.7133040239590551
    },
    {
      "epoch": 0.23359932580126938,
      "grad_norm": 0.134765625,
      "learning_rate": 0.0005703027941979033,
      "loss": 1.7154953002929687,
      "step": 108480,
      "train_bpb": 0.6051578801589896,
      "train_bytes": 28356.075,
      "train_loss_nats_per_token": 1.7165667625244325
    },
    {
      "epoch": 0.2337310052408417,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0005698134618426663,
      "loss": 1.709852409362793,
      "step": 108500,
      "train_bpb": 0.605264649403771,
      "train_bytes": 28336.70625,
      "train_loss_nats_per_token": 1.7103513954848835
    },
    {
      "epoch": 0.233862684680414,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0005693242902788609,
      "loss": 1.7178932189941407,
      "step": 108520,
      "train_bpb": 0.6081595082688723,
      "train_bytes": 28324.0375,
      "train_loss_nats_per_token": 1.7183983963193308
    },
    {
      "epoch": 0.2339943641199863,
      "grad_norm": 0.189453125,
      "learning_rate": 0.0005688352795910439,
      "loss": 1.7187591552734376,
      "step": 108540,
      "train_bpb": 0.6083165034968333,
      "train_bytes": 28097.70625,
      "train_loss_nats_per_token": 1.7193900251531884
    },
    {
      "epoch": 0.2341260435595586,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0005683464298637458,
      "loss": 1.7163597106933595,
      "step": 108560,
      "train_bpb": 0.6125618401940811,
      "train_bytes": 28065.19375,
      "train_loss_nats_per_token": 1.7171484128383314
    },
    {
      "epoch": 0.23425772299913092,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0005678577411814691,
      "loss": 1.713827896118164,
      "step": 108580,
      "train_bpb": 0.6082228544941611,
      "train_bytes": 28177.45,
      "train_loss_nats_per_token": 1.714507154125086
    },
    {
      "epoch": 0.2343894024387032,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0005673692136286877,
      "loss": 1.7264226913452148,
      "step": 108600,
      "train_bpb": 0.6038096250535533,
      "train_bytes": 28816.6125,
      "train_loss_nats_per_token": 1.7272234833670739
    },
    {
      "epoch": 0.23452108187827553,
      "grad_norm": 0.15625,
      "learning_rate": 0.0005668808472898484,
      "loss": 1.713286781311035,
      "step": 108620,
      "train_bpb": 0.6090534227946098,
      "train_bytes": 28168.625,
      "train_loss_nats_per_token": 1.7143220498331737
    },
    {
      "epoch": 0.23465276131784782,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0005663926422493703,
      "loss": 1.7214759826660155,
      "step": 108640,
      "train_bpb": 0.608533088372957,
      "train_bytes": 28047.74375,
      "train_loss_nats_per_token": 1.7230734492056647
    },
    {
      "epoch": 0.23478444075742014,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0005659045985916433,
      "loss": 1.709956169128418,
      "step": 108660,
      "train_bpb": 0.607498283230322,
      "train_bytes": 28188.3875,
      "train_loss_nats_per_token": 1.7103107316241728
    },
    {
      "epoch": 0.23491612019699243,
      "grad_norm": 0.2080078125,
      "learning_rate": 0.0005654167164010307,
      "loss": 1.717238998413086,
      "step": 108680,
      "train_bpb": 0.6066118507865654,
      "train_bytes": 28281.18125,
      "train_loss_nats_per_token": 1.7181532972496496
    },
    {
      "epoch": 0.23504779963656475,
      "grad_norm": 0.140625,
      "learning_rate": 0.0005649289957618678,
      "loss": 1.7175399780273437,
      "step": 108700,
      "train_bpb": 0.607216629541048,
      "train_bytes": 28186.73125,
      "train_loss_nats_per_token": 1.7178237851986786
    },
    {
      "epoch": 0.23517947907613707,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0005644414367584618,
      "loss": 1.7257413864135742,
      "step": 108720,
      "train_bpb": 0.6090097547058116,
      "train_bytes": 28323.2875,
      "train_loss_nats_per_token": 1.7264998212179314
    },
    {
      "epoch": 0.23531115851570936,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0005639540394750911,
      "loss": 1.7251644134521484,
      "step": 108740,
      "train_bpb": 0.6076879210028496,
      "train_bytes": 28545.9375,
      "train_loss_nats_per_token": 1.7255031062477242
    },
    {
      "epoch": 0.23544283795528168,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0005634668039960071,
      "loss": 1.7279243469238281,
      "step": 108760,
      "train_bpb": 0.6112872053472159,
      "train_bytes": 28490.05625,
      "train_loss_nats_per_token": 1.7289803621162514
    },
    {
      "epoch": 0.23557451739485397,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0005629797304054339,
      "loss": 1.722403335571289,
      "step": 108780,
      "train_bpb": 0.60815141567272,
      "train_bytes": 28315.825,
      "train_loss_nats_per_token": 1.7226995978687742
    },
    {
      "epoch": 0.23570619683442628,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0005624928187875654,
      "loss": 1.7309446334838867,
      "step": 108800,
      "train_bpb": 0.6121280061689626,
      "train_bytes": 28302.375,
      "train_loss_nats_per_token": 1.7308733639353655
    },
    {
      "epoch": 0.23583787627399858,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0005620060692265692,
      "loss": 1.7324323654174805,
      "step": 108820,
      "train_bpb": 0.607170844267972,
      "train_bytes": 28826.84375,
      "train_loss_nats_per_token": 1.7323938153886993
    },
    {
      "epoch": 0.2359695557135709,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0005615194818065852,
      "loss": 1.7320568084716796,
      "step": 108840,
      "train_bpb": 0.6129035802732662,
      "train_bytes": 28347.36875,
      "train_loss_nats_per_token": 1.732466043058138
    },
    {
      "epoch": 0.23610123515314319,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0005610330566117237,
      "loss": 1.7351024627685547,
      "step": 108860,
      "train_bpb": 0.6153305061955024,
      "train_bytes": 28162.19375,
      "train_loss_nats_per_token": 1.7364528430350818
    },
    {
      "epoch": 0.2362329145927155,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0005605467937260684,
      "loss": 1.7100460052490234,
      "step": 108880,
      "train_bpb": 0.605591040179151,
      "train_bytes": 28022.5875,
      "train_loss_nats_per_token": 1.710724327413153
    },
    {
      "epoch": 0.2363645940322878,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0005600606932336739,
      "loss": 1.7270521163940429,
      "step": 108900,
      "train_bpb": 0.6093298895231944,
      "train_bytes": 28425.4625,
      "train_loss_nats_per_token": 1.7270193126118214
    },
    {
      "epoch": 0.2364962734718601,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0005595747552185682,
      "loss": 1.729203224182129,
      "step": 108920,
      "train_bpb": 0.6117371685307816,
      "train_bytes": 28133.58125,
      "train_loss_nats_per_token": 1.7297896691066685
    },
    {
      "epoch": 0.2366279529114324,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.000559088979764749,
      "loss": 1.7223403930664063,
      "step": 108940,
      "train_bpb": 0.6073718522980915,
      "train_bytes": 28357.2125,
      "train_loss_nats_per_token": 1.722771425179999
    },
    {
      "epoch": 0.23675963235100472,
      "grad_norm": 0.1640625,
      "learning_rate": 0.0005586033669561876,
      "loss": 1.7291631698608398,
      "step": 108960,
      "train_bpb": 0.6120785469775205,
      "train_bytes": 28626.8375,
      "train_loss_nats_per_token": 1.7296118241443563
    },
    {
      "epoch": 0.236891311790577,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0005581179168768275,
      "loss": 1.7303485870361328,
      "step": 108980,
      "train_bpb": 0.6174768851376106,
      "train_bytes": 27836.525,
      "train_loss_nats_per_token": 1.7313244598539426
    },
    {
      "epoch": 0.23702299123014933,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0005576326296105818,
      "loss": 1.7170703887939454,
      "step": 109000,
      "train_bpb": 0.6083259711237545,
      "train_bytes": 28022.09375,
      "train_loss_nats_per_token": 1.717846854498328
    },
    {
      "epoch": 0.23702299123014933,
      "eval_loss": 1.5313774347305298,
      "eval_runtime": 72.5584,
      "eval_samples_per_second": 13.782,
      "eval_steps_per_second": 13.782,
      "step": 109000,
      "train_bpb": 0.5404262383322985,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5389246310302742
    },
    {
      "epoch": 0.23715467066972162,
      "grad_norm": 0.1875,
      "learning_rate": 0.0005571475052413378,
      "loss": 1.7128055572509766,
      "step": 109020,
      "train_bpb": 0.6105407389480004,
      "train_bytes": 28062.1,
      "train_loss_nats_per_token": 1.7140311292358776
    },
    {
      "epoch": 0.23728635010929394,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0005566625438529536,
      "loss": 1.725593376159668,
      "step": 109040,
      "train_bpb": 0.6111737781415142,
      "train_bytes": 28273.725,
      "train_loss_nats_per_token": 1.7260245922257986
    },
    {
      "epoch": 0.23741802954886623,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00055617774552926,
      "loss": 1.7176311492919922,
      "step": 109060,
      "train_bpb": 0.6078887249883094,
      "train_bytes": 28104.31875,
      "train_loss_nats_per_token": 1.7175259236701523
    },
    {
      "epoch": 0.23754970898843855,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0005556931103540577,
      "loss": 1.7079097747802734,
      "step": 109080,
      "train_bpb": 0.6050685075253177,
      "train_bytes": 28192.36875,
      "train_loss_nats_per_token": 1.7081533056119687
    },
    {
      "epoch": 0.23768138842801084,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.000555208638411121,
      "loss": 1.7133865356445312,
      "step": 109100,
      "train_bpb": 0.6076207278439555,
      "train_bytes": 28116.93125,
      "train_loss_nats_per_token": 1.7133457369539553
    },
    {
      "epoch": 0.23781306786758316,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.000554724329784196,
      "loss": 1.7165584564208984,
      "step": 109120,
      "train_bpb": 0.6054530517088031,
      "train_bytes": 28719.84375,
      "train_loss_nats_per_token": 1.716677230204716
    },
    {
      "epoch": 0.23794474730715545,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0005542401845569989,
      "loss": 1.7200868606567383,
      "step": 109140,
      "train_bpb": 0.6095539795728666,
      "train_bytes": 28069.48125,
      "train_loss_nats_per_token": 1.7212585062384516
    },
    {
      "epoch": 0.23807642674672777,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.000553756202813219,
      "loss": 1.7180652618408203,
      "step": 109160,
      "train_bpb": 0.6073368326212663,
      "train_bytes": 28485.4875,
      "train_loss_nats_per_token": 1.7186115470077226
    },
    {
      "epoch": 0.23820810618630006,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0005532723846365181,
      "loss": 1.7237993240356446,
      "step": 109180,
      "train_bpb": 0.6095170493862317,
      "train_bytes": 28358.70625,
      "train_loss_nats_per_token": 1.7235298750872112
    },
    {
      "epoch": 0.23833978562587238,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0005527887301105271,
      "loss": 1.7325426101684571,
      "step": 109200,
      "train_bpb": 0.6095913572981506,
      "train_bytes": 28609.125,
      "train_loss_nats_per_token": 1.733894820651007
    },
    {
      "epoch": 0.23847146506544467,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0005523052393188509,
      "loss": 1.716433334350586,
      "step": 109220,
      "train_bpb": 0.6084774624656242,
      "train_bytes": 28227.85625,
      "train_loss_nats_per_token": 1.7169055429222229
    },
    {
      "epoch": 0.238603144505017,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0005518219123450656,
      "loss": 1.7214109420776367,
      "step": 109240,
      "train_bpb": 0.6064499681138903,
      "train_bytes": 28254.575,
      "train_loss_nats_per_token": 1.7221689475674973
    },
    {
      "epoch": 0.23873482394458928,
      "grad_norm": 0.1220703125,
      "learning_rate": 0.0005513387492727191,
      "loss": 1.7091569900512695,
      "step": 109260,
      "train_bpb": 0.6083731421558803,
      "train_bytes": 28143.05,
      "train_loss_nats_per_token": 1.709988170257561
    },
    {
      "epoch": 0.2388665033841616,
      "grad_norm": 0.1630859375,
      "learning_rate": 0.0005508557501853298,
      "loss": 1.7224344253540038,
      "step": 109280,
      "train_bpb": 0.6073568128284016,
      "train_bytes": 28252.9625,
      "train_loss_nats_per_token": 1.723608554016573
    },
    {
      "epoch": 0.2389981828237339,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0005503729151663889,
      "loss": 1.7244766235351563,
      "step": 109300,
      "train_bpb": 0.6087336109725513,
      "train_bytes": 28553.04375,
      "train_loss_nats_per_token": 1.7258655384190535
    },
    {
      "epoch": 0.2391298622633062,
      "grad_norm": 0.119140625,
      "learning_rate": 0.0005498902442993595,
      "loss": 1.7285398483276366,
      "step": 109320,
      "train_bpb": 0.6079736485916568,
      "train_bytes": 28590.6625,
      "train_loss_nats_per_token": 1.729560545184019
    },
    {
      "epoch": 0.2392615417028785,
      "grad_norm": 0.115234375,
      "learning_rate": 0.0005494077376676749,
      "loss": 1.7117595672607422,
      "step": 109340,
      "train_bpb": 0.6073705021374244,
      "train_bytes": 28244.93125,
      "train_loss_nats_per_token": 1.7119734069172472
    },
    {
      "epoch": 0.23939322114245082,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0005489253953547413,
      "loss": 1.7290210723876953,
      "step": 109360,
      "train_bpb": 0.6121000878594647,
      "train_bytes": 28393.34375,
      "train_loss_nats_per_token": 1.7291499581874212
    },
    {
      "epoch": 0.2395249005820231,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0005484432174439359,
      "loss": 1.7199954986572266,
      "step": 109380,
      "train_bpb": 0.6070791057127853,
      "train_bytes": 28465.89375,
      "train_loss_nats_per_token": 1.72130759711535
    },
    {
      "epoch": 0.23965658002159543,
      "grad_norm": 0.12890625,
      "learning_rate": 0.0005479612040186085,
      "loss": 1.7253049850463866,
      "step": 109400,
      "train_bpb": 0.6113526255610107,
      "train_bytes": 28183.29375,
      "train_loss_nats_per_token": 1.7256483430864382
    },
    {
      "epoch": 0.23978825946116775,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0005474793551620783,
      "loss": 1.7236288070678711,
      "step": 109420,
      "train_bpb": 0.6068547870305595,
      "train_bytes": 28283.65,
      "train_loss_nats_per_token": 1.7244355311420279
    },
    {
      "epoch": 0.23991993890074004,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0005469976709576381,
      "loss": 1.7152029037475587,
      "step": 109440,
      "train_bpb": 0.6060642581440288,
      "train_bytes": 28111.8125,
      "train_loss_nats_per_token": 1.715060186846877
    },
    {
      "epoch": 0.24005161834031236,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0005465161514885519,
      "loss": 1.719838523864746,
      "step": 109460,
      "train_bpb": 0.6091773772910223,
      "train_bytes": 28242.89375,
      "train_loss_nats_per_token": 1.7210029437373333
    },
    {
      "epoch": 0.24018329777988465,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0005460347968380538,
      "loss": 1.71539363861084,
      "step": 109480,
      "train_bpb": 0.6059653364983935,
      "train_bytes": 28412.4875,
      "train_loss_nats_per_token": 1.7163443123971907
    },
    {
      "epoch": 0.24031497721945697,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.000545553607089351,
      "loss": 1.7126758575439454,
      "step": 109500,
      "train_bpb": 0.6022333280013685,
      "train_bytes": 28340.325,
      "train_loss_nats_per_token": 1.7132315356369232
    },
    {
      "epoch": 0.24044665665902926,
      "grad_norm": 0.1923828125,
      "learning_rate": 0.0005450725823256223,
      "loss": 1.7121431350708007,
      "step": 109520,
      "train_bpb": 0.6050907627769729,
      "train_bytes": 28231.26875,
      "train_loss_nats_per_token": 1.7116209508190472
    },
    {
      "epoch": 0.24057833609860158,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0005445917226300159,
      "loss": 1.716716194152832,
      "step": 109540,
      "train_bpb": 0.6079718244511343,
      "train_bytes": 28139.86875,
      "train_loss_nats_per_token": 1.7167060661280713
    },
    {
      "epoch": 0.24071001553817387,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0005441110280856539,
      "loss": 1.7298473358154296,
      "step": 109560,
      "train_bpb": 0.608670266303459,
      "train_bytes": 28571.69375,
      "train_loss_nats_per_token": 1.7305394170576287
    },
    {
      "epoch": 0.24084169497774618,
      "grad_norm": 0.15625,
      "learning_rate": 0.0005436304987756286,
      "loss": 1.7012443542480469,
      "step": 109580,
      "train_bpb": 0.6041231063880577,
      "train_bytes": 27878.7,
      "train_loss_nats_per_token": 1.7017747691710405
    },
    {
      "epoch": 0.24097337441731848,
      "grad_norm": 0.1884765625,
      "learning_rate": 0.0005431501347830044,
      "loss": 1.729555320739746,
      "step": 109600,
      "train_bpb": 0.6134653519441224,
      "train_bytes": 27880.60625,
      "train_loss_nats_per_token": 1.7296765626909203
    },
    {
      "epoch": 0.2411050538568908,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0005426699361908157,
      "loss": 1.7238964080810546,
      "step": 109620,
      "train_bpb": 0.6092220760773847,
      "train_bytes": 28178.7625,
      "train_loss_nats_per_token": 1.7240647040497041
    },
    {
      "epoch": 0.24123673329646309,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.00054218990308207,
      "loss": 1.7256725311279297,
      "step": 109640,
      "train_bpb": 0.6136817649520175,
      "train_bytes": 27857.56875,
      "train_loss_nats_per_token": 1.7260397357414685
    },
    {
      "epoch": 0.2413684127360354,
      "grad_norm": 0.140625,
      "learning_rate": 0.000541710035539746,
      "loss": 1.712358283996582,
      "step": 109660,
      "train_bpb": 0.6052372982872503,
      "train_bytes": 28186.8125,
      "train_loss_nats_per_token": 1.7123515903022144
    },
    {
      "epoch": 0.2415000921756077,
      "grad_norm": 0.150390625,
      "learning_rate": 0.0005412303336467921,
      "loss": 1.7171276092529297,
      "step": 109680,
      "train_bpb": 0.6066501353607419,
      "train_bytes": 28162.325,
      "train_loss_nats_per_token": 1.7172558839802399
    },
    {
      "epoch": 0.24163177161518,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0005407507974861299,
      "loss": 1.7205854415893556,
      "step": 109700,
      "train_bpb": 0.6062772031990313,
      "train_bytes": 28462.5875,
      "train_loss_nats_per_token": 1.7215137157450378
    },
    {
      "epoch": 0.2417634510547523,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0005402714271406521,
      "loss": 1.7230989456176757,
      "step": 109720,
      "train_bpb": 0.6084869047519769,
      "train_bytes": 28431.7125,
      "train_loss_nats_per_token": 1.7235944387656332
    },
    {
      "epoch": 0.24189513049432462,
      "grad_norm": 0.23046875,
      "learning_rate": 0.0005397922226932214,
      "loss": 1.737328338623047,
      "step": 109740,
      "train_bpb": 0.612565096918293,
      "train_bytes": 28207.36875,
      "train_loss_nats_per_token": 1.737290917735993
    },
    {
      "epoch": 0.2420268099338969,
      "grad_norm": 0.171875,
      "learning_rate": 0.0005393131842266731,
      "loss": 1.716399574279785,
      "step": 109760,
      "train_bpb": 0.6071459700176074,
      "train_bytes": 28592.6,
      "train_loss_nats_per_token": 1.716697227133653
    },
    {
      "epoch": 0.24215848937346923,
      "grad_norm": 0.12255859375,
      "learning_rate": 0.0005388343118238138,
      "loss": 1.7300724029541015,
      "step": 109780,
      "train_bpb": 0.6115139981472779,
      "train_bytes": 28444.80625,
      "train_loss_nats_per_token": 1.7305803196722498
    },
    {
      "epoch": 0.24229016881304152,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0005383556055674214,
      "loss": 1.723931884765625,
      "step": 109800,
      "train_bpb": 0.6150461742296308,
      "train_bytes": 27780.025,
      "train_loss_nats_per_token": 1.725591422752
    },
    {
      "epoch": 0.24242184825261384,
      "grad_norm": 0.1982421875,
      "learning_rate": 0.0005378770655402438,
      "loss": 1.7160707473754884,
      "step": 109820,
      "train_bpb": 0.6056610568983268,
      "train_bytes": 28198.2875,
      "train_loss_nats_per_token": 1.7174018621602514
    },
    {
      "epoch": 0.24255352769218613,
      "grad_norm": 0.224609375,
      "learning_rate": 0.0005373986918250015,
      "loss": 1.7079870223999023,
      "step": 109840,
      "train_bpb": 0.6065989323429885,
      "train_bytes": 28277.1375,
      "train_loss_nats_per_token": 1.7082988029088013
    },
    {
      "epoch": 0.24268520713175845,
      "grad_norm": 0.140625,
      "learning_rate": 0.0005369204845043862,
      "loss": 1.732179832458496,
      "step": 109860,
      "train_bpb": 0.6108098849707743,
      "train_bytes": 28245.43125,
      "train_loss_nats_per_token": 1.7327105105902925
    },
    {
      "epoch": 0.24281688657133074,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00053644244366106,
      "loss": 1.7092731475830079,
      "step": 109880,
      "train_bpb": 0.6080609336653494,
      "train_bytes": 27826.04375,
      "train_loss_nats_per_token": 1.7095324858354546
    },
    {
      "epoch": 0.24294856601090306,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0005359645693776569,
      "loss": 1.714297103881836,
      "step": 109900,
      "train_bpb": 0.6033411106360044,
      "train_bytes": 28477.0625,
      "train_loss_nats_per_token": 1.7139493156696202
    },
    {
      "epoch": 0.24308024545047535,
      "grad_norm": 0.158203125,
      "learning_rate": 0.0005354868617367819,
      "loss": 1.7032630920410157,
      "step": 109920,
      "train_bpb": 0.6020015281280506,
      "train_bytes": 28414.03125,
      "train_loss_nats_per_token": 1.7033632829651428
    },
    {
      "epoch": 0.24321192489004767,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0005350093208210119,
      "loss": 1.7061830520629884,
      "step": 109940,
      "train_bpb": 0.6013642794347593,
      "train_bytes": 28580.91875,
      "train_loss_nats_per_token": 1.706705450531454
    },
    {
      "epoch": 0.24334360432961996,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.0005345319467128932,
      "loss": 1.7176799774169922,
      "step": 109960,
      "train_bpb": 0.605517754257116,
      "train_bytes": 28566.81875,
      "train_loss_nats_per_token": 1.7181885057759678
    },
    {
      "epoch": 0.24347528376919228,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0005340547394949448,
      "loss": 1.7289253234863282,
      "step": 109980,
      "train_bpb": 0.6077271014104756,
      "train_bytes": 28598.85625,
      "train_loss_nats_per_token": 1.7286939325561648
    },
    {
      "epoch": 0.24360696320876457,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.000533577699249657,
      "loss": 1.7311487197875977,
      "step": 110000,
      "train_bpb": 0.6088833164003673,
      "train_bytes": 28588.975,
      "train_loss_nats_per_token": 1.73154589592979
    },
    {
      "epoch": 0.24360696320876457,
      "eval_loss": 1.530983805656433,
      "eval_runtime": 78.3465,
      "eval_samples_per_second": 12.764,
      "eval_steps_per_second": 12.764,
      "step": 110000,
      "train_bpb": 0.540296077532811,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5385539834818787
    },
    {
      "epoch": 0.2437386426483369,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.0005331008260594897,
      "loss": 1.7133548736572266,
      "step": 110020,
      "train_bpb": 0.6066081751264738,
      "train_bytes": 28323.5625,
      "train_loss_nats_per_token": 1.7135454224409692
    },
    {
      "epoch": 0.24387032208790918,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0005326241200068754,
      "loss": 1.7273595809936524,
      "step": 110040,
      "train_bpb": 0.6109324760750152,
      "train_bytes": 28381.56875,
      "train_loss_nats_per_token": 1.7271671422742536
    },
    {
      "epoch": 0.2440020015274815,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0005321475811742177,
      "loss": 1.717525100708008,
      "step": 110060,
      "train_bpb": 0.6084373826531937,
      "train_bytes": 28274.8875,
      "train_loss_nats_per_token": 1.7177902723286715
    },
    {
      "epoch": 0.24413368096705382,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0005316712096438898,
      "loss": 1.7221260070800781,
      "step": 110080,
      "train_bpb": 0.6063191999165831,
      "train_bytes": 28637.96875,
      "train_loss_nats_per_token": 1.722241129909338
    },
    {
      "epoch": 0.2442653604066261,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0005311950054982373,
      "loss": 1.7185951232910157,
      "step": 110100,
      "train_bpb": 0.6056242323274348,
      "train_bytes": 28556.7875,
      "train_loss_nats_per_token": 1.7189510387348754
    },
    {
      "epoch": 0.24439703984619843,
      "grad_norm": 0.1669921875,
      "learning_rate": 0.0005307189688195765,
      "loss": 1.71678466796875,
      "step": 110120,
      "train_bpb": 0.6042887568552291,
      "train_bytes": 28806.48125,
      "train_loss_nats_per_token": 1.716991781546025
    },
    {
      "epoch": 0.24452871928577072,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0005302430996901957,
      "loss": 1.7187725067138673,
      "step": 110140,
      "train_bpb": 0.6078782195433409,
      "train_bytes": 28285.35,
      "train_loss_nats_per_token": 1.7195792357483342
    },
    {
      "epoch": 0.24466039872534304,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0005297673981923519,
      "loss": 1.7172367095947265,
      "step": 110160,
      "train_bpb": 0.6139206138025446,
      "train_bytes": 27727.20625,
      "train_loss_nats_per_token": 1.7178320527724205
    },
    {
      "epoch": 0.24479207816491533,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0005292918644082752,
      "loss": 1.7222135543823243,
      "step": 110180,
      "train_bpb": 0.6086901923634849,
      "train_bytes": 28396.9625,
      "train_loss_nats_per_token": 1.723487340195575
    },
    {
      "epoch": 0.24492375760448765,
      "grad_norm": 0.12158203125,
      "learning_rate": 0.0005288164984201666,
      "loss": 1.732524299621582,
      "step": 110200,
      "train_bpb": 0.6128076869899601,
      "train_bytes": 28251.1875,
      "train_loss_nats_per_token": 1.733820723261662
    },
    {
      "epoch": 0.24505543704405994,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0005283413003101965,
      "loss": 1.7187496185302735,
      "step": 110220,
      "train_bpb": 0.6061815613840023,
      "train_bytes": 28493.05625,
      "train_loss_nats_per_token": 1.7198173548937796
    },
    {
      "epoch": 0.24518711648363226,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0005278662701605079,
      "loss": 1.7262392044067383,
      "step": 110240,
      "train_bpb": 0.6071882591588316,
      "train_bytes": 28726.60625,
      "train_loss_nats_per_token": 1.725384622561898
    },
    {
      "epoch": 0.24531879592320455,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0005273914080532145,
      "loss": 1.723727035522461,
      "step": 110260,
      "train_bpb": 0.6097961211492301,
      "train_bytes": 28424.2625,
      "train_loss_nats_per_token": 1.7247261174979744
    },
    {
      "epoch": 0.24545047536277687,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0005269167140703998,
      "loss": 1.7121734619140625,
      "step": 110280,
      "train_bpb": 0.6024508933054024,
      "train_bytes": 28409.0375,
      "train_loss_nats_per_token": 1.712517175175566
    },
    {
      "epoch": 0.24558215480234916,
      "grad_norm": 0.140625,
      "learning_rate": 0.0005264421882941197,
      "loss": 1.7159561157226562,
      "step": 110300,
      "train_bpb": 0.6117714240617652,
      "train_bytes": 27900.5125,
      "train_loss_nats_per_token": 1.7171116955631482
    },
    {
      "epoch": 0.24571383424192147,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0005259678308064003,
      "loss": 1.716153335571289,
      "step": 110320,
      "train_bpb": 0.6051164509882287,
      "train_bytes": 28394.63125,
      "train_loss_nats_per_token": 1.7169959224555715
    },
    {
      "epoch": 0.24584551368149377,
      "grad_norm": 0.12890625,
      "learning_rate": 0.000525493641689239,
      "loss": 1.7082075119018554,
      "step": 110340,
      "train_bpb": 0.6071661195318043,
      "train_bytes": 28314.51875,
      "train_loss_nats_per_token": 1.7086627040554847
    },
    {
      "epoch": 0.24597719312106608,
      "grad_norm": 0.11669921875,
      "learning_rate": 0.0005250196210246032,
      "loss": 1.7294771194458007,
      "step": 110360,
      "train_bpb": 0.6078692997395342,
      "train_bytes": 28511.8625,
      "train_loss_nats_per_token": 1.7302926662340463
    },
    {
      "epoch": 0.24610887256063838,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0005245457688944321,
      "loss": 1.712773895263672,
      "step": 110380,
      "train_bpb": 0.608340921063524,
      "train_bytes": 28148.64375,
      "train_loss_nats_per_token": 1.7135876807230783
    },
    {
      "epoch": 0.2462405520002107,
      "grad_norm": 0.14453125,
      "learning_rate": 0.000524072085380636,
      "loss": 1.7196598052978516,
      "step": 110400,
      "train_bpb": 0.6087383450720856,
      "train_bytes": 28281.78125,
      "train_loss_nats_per_token": 1.7196287567294788
    },
    {
      "epoch": 0.24637223143978298,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0005235985705650945,
      "loss": 1.7198509216308593,
      "step": 110420,
      "train_bpb": 0.6060156400826149,
      "train_bytes": 28456.90625,
      "train_loss_nats_per_token": 1.7197190346075473
    },
    {
      "epoch": 0.2465039108793553,
      "grad_norm": 0.1650390625,
      "learning_rate": 0.0005231252245296596,
      "loss": 1.70921630859375,
      "step": 110440,
      "train_bpb": 0.6028280071705281,
      "train_bytes": 28423.30625,
      "train_loss_nats_per_token": 1.7091913882609913
    },
    {
      "epoch": 0.2466355903189276,
      "grad_norm": 0.1826171875,
      "learning_rate": 0.0005226520473561535,
      "loss": 1.7107978820800782,
      "step": 110460,
      "train_bpb": 0.6047288768827188,
      "train_bytes": 28562.8375,
      "train_loss_nats_per_token": 1.7114464878363058
    },
    {
      "epoch": 0.2467672697584999,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0005221790391263698,
      "loss": 1.7111249923706056,
      "step": 110480,
      "train_bpb": 0.6089268908611674,
      "train_bytes": 28101.56875,
      "train_loss_nats_per_token": 1.7116901667161801
    },
    {
      "epoch": 0.2468989491980722,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0005217061999220716,
      "loss": 1.7026735305786134,
      "step": 110500,
      "train_bpb": 0.6023416435424647,
      "train_bytes": 28398.23125,
      "train_loss_nats_per_token": 1.7035431258954685
    },
    {
      "epoch": 0.24703062863764452,
      "grad_norm": 0.142578125,
      "learning_rate": 0.0005212335298249938,
      "loss": 1.7120479583740233,
      "step": 110520,
      "train_bpb": 0.6093728747918761,
      "train_bytes": 28199.15,
      "train_loss_nats_per_token": 1.7130607247698608
    },
    {
      "epoch": 0.2471623080772168,
      "grad_norm": 0.171875,
      "learning_rate": 0.0005207610289168426,
      "loss": 1.7153709411621094,
      "step": 110540,
      "train_bpb": 0.609455867818978,
      "train_bytes": 28101.4875,
      "train_loss_nats_per_token": 1.716454266885847
    },
    {
      "epoch": 0.24729398751678913,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0005202886972792928,
      "loss": 1.704010581970215,
      "step": 110560,
      "train_bpb": 0.605125089407355,
      "train_bytes": 28170.0375,
      "train_loss_nats_per_token": 1.7044913168867637
    },
    {
      "epoch": 0.24742566695636142,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0005198165349939922,
      "loss": 1.7194940567016601,
      "step": 110580,
      "train_bpb": 0.6087102664806792,
      "train_bytes": 28297.46875,
      "train_loss_nats_per_token": 1.7202708448126132
    },
    {
      "epoch": 0.24755734639593374,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.000519344542142559,
      "loss": 1.716411018371582,
      "step": 110600,
      "train_bpb": 0.6068147914152704,
      "train_bytes": 28465.34375,
      "train_loss_nats_per_token": 1.7162612985182566
    },
    {
      "epoch": 0.24768902583550603,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0005188727188065802,
      "loss": 1.7052494049072267,
      "step": 110620,
      "train_bpb": 0.6050404382393613,
      "train_bytes": 28150.70625,
      "train_loss_nats_per_token": 1.7057670452845866
    },
    {
      "epoch": 0.24782070527507835,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0005184010650676156,
      "loss": 1.718343734741211,
      "step": 110640,
      "train_bpb": 0.6068865515000433,
      "train_bytes": 28285.31875,
      "train_loss_nats_per_token": 1.719187488076989
    },
    {
      "epoch": 0.24795238471465064,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.000517929581007195,
      "loss": 1.7160953521728515,
      "step": 110660,
      "train_bpb": 0.6041811744691382,
      "train_bytes": 28409.39375,
      "train_loss_nats_per_token": 1.7173378763720086
    },
    {
      "epoch": 0.24808406415422296,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0005174582667068192,
      "loss": 1.7242042541503906,
      "step": 110680,
      "train_bpb": 0.6072884583073347,
      "train_bytes": 28305.5,
      "train_loss_nats_per_token": 1.7247291996603265
    },
    {
      "epoch": 0.24821574359379525,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0005169871222479585,
      "loss": 1.7200180053710938,
      "step": 110700,
      "train_bpb": 0.6110822441012885,
      "train_bytes": 28102.55625,
      "train_loss_nats_per_token": 1.7202871499579309
    },
    {
      "epoch": 0.24834742303336757,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0005165161477120549,
      "loss": 1.7208019256591798,
      "step": 110720,
      "train_bpb": 0.60875974476397,
      "train_bytes": 28429.51875,
      "train_loss_nats_per_token": 1.7222909703350535
    },
    {
      "epoch": 0.2484791024729399,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.0005160453431805212,
      "loss": 1.7147289276123048,
      "step": 110740,
      "train_bpb": 0.6071413445815153,
      "train_bytes": 28078.40625,
      "train_loss_nats_per_token": 1.714177607831115
    },
    {
      "epoch": 0.24861078191251218,
      "grad_norm": 0.2021484375,
      "learning_rate": 0.0005155747087347396,
      "loss": 1.7195318222045899,
      "step": 110760,
      "train_bpb": 0.6087504863404959,
      "train_bytes": 28189.34375,
      "train_loss_nats_per_token": 1.7203843224678343
    },
    {
      "epoch": 0.2487424613520845,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0005151042444560641,
      "loss": 1.7126714706420898,
      "step": 110780,
      "train_bpb": 0.6073245046606789,
      "train_bytes": 28604.53125,
      "train_loss_nats_per_token": 1.7130505578020179
    },
    {
      "epoch": 0.2488741407916568,
      "grad_norm": 0.11474609375,
      "learning_rate": 0.0005146339504258191,
      "loss": 1.711126708984375,
      "step": 110800,
      "train_bpb": 0.6064146114786898,
      "train_bytes": 28269.38125,
      "train_loss_nats_per_token": 1.711675118414096
    },
    {
      "epoch": 0.2490058202312291,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0005141638267252986,
      "loss": 1.7315103530883789,
      "step": 110820,
      "train_bpb": 0.6088363856777435,
      "train_bytes": 28461.16875,
      "train_loss_nats_per_token": 1.7315400065633573
    },
    {
      "epoch": 0.2491374996708014,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0005136938734357681,
      "loss": 1.7100589752197266,
      "step": 110840,
      "train_bpb": 0.60250674313937,
      "train_bytes": 28507.25,
      "train_loss_nats_per_token": 1.710746734289316
    },
    {
      "epoch": 0.24926917911037372,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0005132240906384639,
      "loss": 1.713458251953125,
      "step": 110860,
      "train_bpb": 0.603519589186872,
      "train_bytes": 28862.5875,
      "train_loss_nats_per_token": 1.7139785757084374
    },
    {
      "epoch": 0.249400858549946,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.0005127544784145922,
      "loss": 1.7034481048583985,
      "step": 110880,
      "train_bpb": 0.6054166980591544,
      "train_bytes": 28343.16875,
      "train_loss_nats_per_token": 1.7044171672853963
    },
    {
      "epoch": 0.24953253798951833,
      "grad_norm": 0.166015625,
      "learning_rate": 0.0005122850368453294,
      "loss": 1.7226051330566405,
      "step": 110900,
      "train_bpb": 0.6093577901235925,
      "train_bytes": 28371.075,
      "train_loss_nats_per_token": 1.7227329061418977
    },
    {
      "epoch": 0.24966421742909062,
      "grad_norm": 0.146484375,
      "learning_rate": 0.000511815766011823,
      "loss": 1.7265363693237306,
      "step": 110920,
      "train_bpb": 0.6060514527126212,
      "train_bytes": 28644.575,
      "train_loss_nats_per_token": 1.7271171242848702
    },
    {
      "epoch": 0.24979589686866294,
      "grad_norm": 0.197265625,
      "learning_rate": 0.0005113466659951916,
      "loss": 1.7234535217285156,
      "step": 110940,
      "train_bpb": 0.608114782698748,
      "train_bytes": 28271.975,
      "train_loss_nats_per_token": 1.722937041423689
    },
    {
      "epoch": 0.24992757630823523,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0005108777368765223,
      "loss": 1.7135459899902343,
      "step": 110960,
      "train_bpb": 0.6049404326240212,
      "train_bytes": 28498.6875,
      "train_loss_nats_per_token": 1.7144110889221391
    },
    {
      "epoch": 0.2500592557478075,
      "grad_norm": 0.1328125,
      "learning_rate": 0.0005104089787368746,
      "loss": 1.7162099838256837,
      "step": 110980,
      "train_bpb": 0.6036523965807274,
      "train_bytes": 28699.4875,
      "train_loss_nats_per_token": 1.716784908323214
    },
    {
      "epoch": 0.25019093518737984,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0005099403916572775,
      "loss": 1.697690200805664,
      "step": 111000,
      "train_bpb": 0.6018248440265141,
      "train_bytes": 27917.9375,
      "train_loss_nats_per_token": 1.6981957732269195
    },
    {
      "epoch": 0.25019093518737984,
      "eval_loss": 1.5307097434997559,
      "eval_runtime": 71.3307,
      "eval_samples_per_second": 14.019,
      "eval_steps_per_second": 14.019,
      "step": 111000,
      "train_bpb": 0.5401866277182448,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5382423127976117
    },
    {
      "epoch": 0.25032261462695216,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0005094719757187312,
      "loss": 1.7092775344848632,
      "step": 111020,
      "train_bpb": 0.603883591297479,
      "train_bytes": 28454.875,
      "train_loss_nats_per_token": 1.7094977121273571
    },
    {
      "epoch": 0.2504542940665245,
      "grad_norm": 0.16015625,
      "learning_rate": 0.0005090037310022051,
      "loss": 1.7200653076171875,
      "step": 111040,
      "train_bpb": 0.6090890038846021,
      "train_bytes": 27964.54375,
      "train_loss_nats_per_token": 1.7204837961950978
    },
    {
      "epoch": 0.25058597350609674,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0005085356575886397,
      "loss": 1.704643440246582,
      "step": 111060,
      "train_bpb": 0.6056886717822826,
      "train_bytes": 28160.875,
      "train_loss_nats_per_token": 1.7064331498352903
    },
    {
      "epoch": 0.25071765294566906,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0005080677555589466,
      "loss": 1.7218650817871093,
      "step": 111080,
      "train_bpb": 0.6088785595109363,
      "train_bytes": 28483.175,
      "train_loss_nats_per_token": 1.722880672359283
    },
    {
      "epoch": 0.2508493323852414,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.000507600024994006,
      "loss": 1.7146593093872071,
      "step": 111100,
      "train_bpb": 0.6060350462212769,
      "train_bytes": 27979.41875,
      "train_loss_nats_per_token": 1.7151908390925172
    },
    {
      "epoch": 0.2509810118248137,
      "grad_norm": 0.10595703125,
      "learning_rate": 0.00050713246597467,
      "loss": 1.720484161376953,
      "step": 111120,
      "train_bpb": 0.6102933895605267,
      "train_bytes": 28158.9125,
      "train_loss_nats_per_token": 1.72119773493167
    },
    {
      "epoch": 0.25111269126438596,
      "grad_norm": 0.1328125,
      "learning_rate": 0.000506665078581761,
      "loss": 1.7080904006958009,
      "step": 111140,
      "train_bpb": 0.6042704428741936,
      "train_bytes": 28240.5875,
      "train_loss_nats_per_token": 1.7085032963185138
    },
    {
      "epoch": 0.2512443707039583,
      "grad_norm": 0.12109375,
      "learning_rate": 0.0005061978628960704,
      "loss": 1.7175691604614258,
      "step": 111160,
      "train_bpb": 0.6085034340715189,
      "train_bytes": 28130.4125,
      "train_loss_nats_per_token": 1.7179281930742125
    },
    {
      "epoch": 0.2513760501435306,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0005057308189983612,
      "loss": 1.7134090423583985,
      "step": 111180,
      "train_bpb": 0.602618320937132,
      "train_bytes": 28810.325,
      "train_loss_nats_per_token": 1.7135163658607155
    },
    {
      "epoch": 0.2515077295831029,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0005052639469693664,
      "loss": 1.714719009399414,
      "step": 111200,
      "train_bpb": 0.6062915980596542,
      "train_bytes": 28234.14375,
      "train_loss_nats_per_token": 1.7162219706241761
    },
    {
      "epoch": 0.2516394090226752,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0005047972468897894,
      "loss": 1.7141870498657226,
      "step": 111220,
      "train_bpb": 0.6061759733593478,
      "train_bytes": 28331.5125,
      "train_loss_nats_per_token": 1.7139966347819533
    },
    {
      "epoch": 0.2517710884622475,
      "grad_norm": 0.154296875,
      "learning_rate": 0.000504330718840303,
      "loss": 1.715096664428711,
      "step": 111240,
      "train_bpb": 0.6081405137648035,
      "train_bytes": 28330.96875,
      "train_loss_nats_per_token": 1.7158835987984518
    },
    {
      "epoch": 0.2519027679018198,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0005038643629015513,
      "loss": 1.7130468368530274,
      "step": 111260,
      "train_bpb": 0.6084881321205253,
      "train_bytes": 28271.325,
      "train_loss_nats_per_token": 1.7136901094624934
    },
    {
      "epoch": 0.25203444734139213,
      "grad_norm": 0.11572265625,
      "learning_rate": 0.0005033981791541487,
      "loss": 1.7154264450073242,
      "step": 111280,
      "train_bpb": 0.6048890789747495,
      "train_bytes": 28604.85,
      "train_loss_nats_per_token": 1.7158435371681873
    },
    {
      "epoch": 0.2521661267809644,
      "grad_norm": 0.173828125,
      "learning_rate": 0.0005029321676786787,
      "loss": 1.7030086517333984,
      "step": 111300,
      "train_bpb": 0.6027850955657593,
      "train_bytes": 28546.575,
      "train_loss_nats_per_token": 1.7038263231798993
    },
    {
      "epoch": 0.2522978062205367,
      "grad_norm": 0.125,
      "learning_rate": 0.0005024663285556958,
      "loss": 1.728964614868164,
      "step": 111320,
      "train_bpb": 0.6118507295796471,
      "train_bytes": 28513.575,
      "train_loss_nats_per_token": 1.7290898320939496
    },
    {
      "epoch": 0.25242948566010903,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.000502000661865725,
      "loss": 1.6974784851074218,
      "step": 111340,
      "train_bpb": 0.5976690582747568,
      "train_bytes": 28603.725,
      "train_loss_nats_per_token": 1.6978073338058897
    },
    {
      "epoch": 0.25256116509968135,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0005015351676892615,
      "loss": 1.7113729476928712,
      "step": 111360,
      "train_bpb": 0.606200240027749,
      "train_bytes": 28346.825,
      "train_loss_nats_per_token": 1.711965913932118
    },
    {
      "epoch": 0.2526928445392536,
      "grad_norm": 0.154296875,
      "learning_rate": 0.0005010698461067692,
      "loss": 1.7072687149047852,
      "step": 111380,
      "train_bpb": 0.6014832676765579,
      "train_bytes": 28573.9,
      "train_loss_nats_per_token": 1.7076162193504103
    },
    {
      "epoch": 0.25282452397882593,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0005006046971986842,
      "loss": 1.7155813217163085,
      "step": 111400,
      "train_bpb": 0.6068864028704268,
      "train_bytes": 28304.06875,
      "train_loss_nats_per_token": 1.7173041819440324
    },
    {
      "epoch": 0.25295620341839825,
      "grad_norm": 0.126953125,
      "learning_rate": 0.0005001397210454119,
      "loss": 1.7128314971923828,
      "step": 111420,
      "train_bpb": 0.6034774117284287,
      "train_bytes": 28409.05,
      "train_loss_nats_per_token": 1.7125712569729106
    },
    {
      "epoch": 0.25308788285797057,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0004996749177273267,
      "loss": 1.715683937072754,
      "step": 111440,
      "train_bpb": 0.606667069059481,
      "train_bytes": 28639.25625,
      "train_loss_nats_per_token": 1.7164050993150701
    },
    {
      "epoch": 0.2532195622975429,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0004992102873247752,
      "loss": 1.7137563705444336,
      "step": 111460,
      "train_bpb": 0.6066630287763765,
      "train_bytes": 28207.36875,
      "train_loss_nats_per_token": 1.7136990600438895
    },
    {
      "epoch": 0.25335124173711515,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.0004987458299180733,
      "loss": 1.7006565093994142,
      "step": 111480,
      "train_bpb": 0.600215443644072,
      "train_bytes": 28520.69375,
      "train_loss_nats_per_token": 1.7011776468849154
    },
    {
      "epoch": 0.25348292117668747,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004982815455875056,
      "loss": 1.706118392944336,
      "step": 111500,
      "train_bpb": 0.6092782493958145,
      "train_bytes": 28301.1375,
      "train_loss_nats_per_token": 1.7078730053258293
    },
    {
      "epoch": 0.2536146006162598,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0004978174344133288,
      "loss": 1.7184642791748046,
      "step": 111520,
      "train_bpb": 0.6064487555674538,
      "train_bytes": 28667.35625,
      "train_loss_nats_per_token": 1.7188075259176985
    },
    {
      "epoch": 0.2537462800558321,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.0004973534964757688,
      "loss": 1.7124521255493164,
      "step": 111540,
      "train_bpb": 0.6063749236344043,
      "train_bytes": 27849.5375,
      "train_loss_nats_per_token": 1.7129263964547128
    },
    {
      "epoch": 0.25387795949540437,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0004968897318550221,
      "loss": 1.7051202774047851,
      "step": 111560,
      "train_bpb": 0.6061345297228422,
      "train_bytes": 28160.9125,
      "train_loss_nats_per_token": 1.7062032075464417
    },
    {
      "epoch": 0.2540096389349767,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004964261406312536,
      "loss": 1.715925407409668,
      "step": 111580,
      "train_bpb": 0.6051800429803977,
      "train_bytes": 28471.525,
      "train_loss_nats_per_token": 1.7161653550294662
    },
    {
      "epoch": 0.254141318374549,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0004959627228846,
      "loss": 1.7141046524047852,
      "step": 111600,
      "train_bpb": 0.6094199993532947,
      "train_bytes": 28009.15,
      "train_loss_nats_per_token": 1.7150220908442841
    },
    {
      "epoch": 0.2542729978141213,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.000495499478695168,
      "loss": 1.7100229263305664,
      "step": 111620,
      "train_bpb": 0.6035648117964871,
      "train_bytes": 28625.93125,
      "train_loss_nats_per_token": 1.7104154931107747
    },
    {
      "epoch": 0.2544046772536936,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0004950364081430322,
      "loss": 1.7176681518554688,
      "step": 111640,
      "train_bpb": 0.6075025349922759,
      "train_bytes": 28466.8875,
      "train_loss_nats_per_token": 1.7187215460579377
    },
    {
      "epoch": 0.2545363566932659,
      "grad_norm": 0.140625,
      "learning_rate": 0.0004945735113082398,
      "loss": 1.704604148864746,
      "step": 111660,
      "train_bpb": 0.6070905373322671,
      "train_bytes": 28123.36875,
      "train_loss_nats_per_token": 1.7050208588627238
    },
    {
      "epoch": 0.2546680361328382,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.0004941107882708069,
      "loss": 1.7247371673583984,
      "step": 111680,
      "train_bpb": 0.610460336083333,
      "train_bytes": 28340.2375,
      "train_loss_nats_per_token": 1.7242095417924008
    },
    {
      "epoch": 0.25479971557241055,
      "grad_norm": 0.123046875,
      "learning_rate": 0.0004936482391107187,
      "loss": 1.7220870971679687,
      "step": 111700,
      "train_bpb": 0.6104572777608394,
      "train_bytes": 28287.975,
      "train_loss_nats_per_token": 1.721901737728078
    },
    {
      "epoch": 0.2549313950119828,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0004931858639079318,
      "loss": 1.7015893936157227,
      "step": 111720,
      "train_bpb": 0.6055717020979836,
      "train_bytes": 27943.59375,
      "train_loss_nats_per_token": 1.702172804440681
    },
    {
      "epoch": 0.2550630744515551,
      "grad_norm": 0.1484375,
      "learning_rate": 0.0004927236627423715,
      "loss": 1.7236068725585938,
      "step": 111740,
      "train_bpb": 0.6085609365791291,
      "train_bytes": 28539.4125,
      "train_loss_nats_per_token": 1.723566203808513
    },
    {
      "epoch": 0.25519475389112745,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0004922616356939348,
      "loss": 1.7136573791503906,
      "step": 111760,
      "train_bpb": 0.6019120764318123,
      "train_bytes": 28665.94375,
      "train_loss_nats_per_token": 1.713716402998701
    },
    {
      "epoch": 0.25532643333069976,
      "grad_norm": 0.15625,
      "learning_rate": 0.0004917997828424858,
      "loss": 1.7164573669433594,
      "step": 111780,
      "train_bpb": 0.6086862311948097,
      "train_bytes": 28265.29375,
      "train_loss_nats_per_token": 1.7177257138911732
    },
    {
      "epoch": 0.25545811277027203,
      "grad_norm": 0.169921875,
      "learning_rate": 0.0004913381042678611,
      "loss": 1.704574966430664,
      "step": 111800,
      "train_bpb": 0.6029668668670033,
      "train_bytes": 28227.24375,
      "train_loss_nats_per_token": 1.7054669828493423
    },
    {
      "epoch": 0.25558979220984435,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0004908766000498661,
      "loss": 1.7036638259887695,
      "step": 111820,
      "train_bpb": 0.6038473825714913,
      "train_bytes": 28200.1625,
      "train_loss_nats_per_token": 1.7049524878187874
    },
    {
      "epoch": 0.25572147164941667,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0004904152702682754,
      "loss": 1.7162900924682618,
      "step": 111840,
      "train_bpb": 0.6073929639970091,
      "train_bytes": 28506.7,
      "train_loss_nats_per_token": 1.7172999344965998
    },
    {
      "epoch": 0.255853151088989,
      "grad_norm": 0.162109375,
      "learning_rate": 0.0004899541150028349,
      "loss": 1.6969039916992188,
      "step": 111860,
      "train_bpb": 0.602328065602685,
      "train_bytes": 28287.4125,
      "train_loss_nats_per_token": 1.6976818773825257
    },
    {
      "epoch": 0.25598483052856125,
      "grad_norm": 0.1552734375,
      "learning_rate": 0.0004894931343332593,
      "loss": 1.7218236923217773,
      "step": 111880,
      "train_bpb": 0.6088425845203558,
      "train_bytes": 28088.80625,
      "train_loss_nats_per_token": 1.7219421839280071
    },
    {
      "epoch": 0.25611650996813357,
      "grad_norm": 0.14453125,
      "learning_rate": 0.0004890323283392338,
      "loss": 1.7230216979980468,
      "step": 111900,
      "train_bpb": 0.6085402734181506,
      "train_bytes": 28439.875,
      "train_loss_nats_per_token": 1.7235669593167688
    },
    {
      "epoch": 0.2562481894077059,
      "grad_norm": 0.10888671875,
      "learning_rate": 0.0004885716971004124,
      "loss": 1.7022665023803711,
      "step": 111920,
      "train_bpb": 0.6030047785989947,
      "train_bytes": 28187.33125,
      "train_loss_nats_per_token": 1.70258879030865
    },
    {
      "epoch": 0.2563798688472782,
      "grad_norm": 0.123046875,
      "learning_rate": 0.00048811124069641956,
      "loss": 1.7130685806274415,
      "step": 111940,
      "train_bpb": 0.6082516679140363,
      "train_bytes": 27914.09375,
      "train_loss_nats_per_token": 1.713804592148435
    },
    {
      "epoch": 0.25651154828685047,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.00048765095920685035,
      "loss": 1.7236461639404297,
      "step": 111960,
      "train_bpb": 0.6062492082619527,
      "train_bytes": 28653.68125,
      "train_loss_nats_per_token": 1.724024815131069
    },
    {
      "epoch": 0.2566432277264228,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0004871908527112675,
      "loss": 1.7081489562988281,
      "step": 111980,
      "train_bpb": 0.6028321977596638,
      "train_bytes": 28279.90625,
      "train_loss_nats_per_token": 1.709411651885975
    },
    {
      "epoch": 0.2567749071659951,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0004867309212892053,
      "loss": 1.7143125534057617,
      "step": 112000,
      "train_bpb": 0.6068712834814083,
      "train_bytes": 28185.2125,
      "train_loss_nats_per_token": 1.7147012703252607
    },
    {
      "epoch": 0.2567749071659951,
      "eval_loss": 1.5305503606796265,
      "eval_runtime": 71.151,
      "eval_samples_per_second": 14.055,
      "eval_steps_per_second": 14.055,
      "step": 112000,
      "train_bpb": 0.540121442875897,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5380566916113263
    },
    {
      "epoch": 0.2569065866055674,
      "grad_norm": 0.140625,
      "learning_rate": 0.0004862711650201677,
      "loss": 1.7164249420166016,
      "step": 112020,
      "train_bpb": 0.6078136268400487,
      "train_bytes": 28189.05625,
      "train_loss_nats_per_token": 1.7169912113751766
    },
    {
      "epoch": 0.2570382660451397,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.000485811583983627,
      "loss": 1.706865119934082,
      "step": 112040,
      "train_bpb": 0.6037584335463797,
      "train_bytes": 28380.28125,
      "train_loss_nats_per_token": 1.7080429393908394
    },
    {
      "epoch": 0.257169945484712,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00048535217825902634,
      "loss": 1.7017990112304688,
      "step": 112060,
      "train_bpb": 0.6030641121000396,
      "train_bytes": 28288.49375,
      "train_loss_nats_per_token": 1.7033810600883172
    },
    {
      "epoch": 0.2573016249242843,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004848929479257783,
      "loss": 1.7214805603027343,
      "step": 112080,
      "train_bpb": 0.6081350041551078,
      "train_bytes": 28471.1,
      "train_loss_nats_per_token": 1.7221065726501585
    },
    {
      "epoch": 0.25743330436385664,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.00048443389306326574,
      "loss": 1.7167343139648437,
      "step": 112100,
      "train_bpb": 0.608112584373705,
      "train_bytes": 28503.71875,
      "train_loss_nats_per_token": 1.7164592125369549
    },
    {
      "epoch": 0.25756498380342896,
      "grad_norm": 0.12890625,
      "learning_rate": 0.00048397501375083965,
      "loss": 1.7172395706176757,
      "step": 112120,
      "train_bpb": 0.6092151974255695,
      "train_bytes": 28214.025,
      "train_loss_nats_per_token": 1.7180101701095107
    },
    {
      "epoch": 0.2576966632430012,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.00048351631006782187,
      "loss": 1.713209342956543,
      "step": 112140,
      "train_bpb": 0.6068497340827084,
      "train_bytes": 28364.425,
      "train_loss_nats_per_token": 1.7139942759441225
    },
    {
      "epoch": 0.25782834268257354,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.0004830577820935044,
      "loss": 1.7141101837158204,
      "step": 112160,
      "train_bpb": 0.6037449846805277,
      "train_bytes": 28405.3625,
      "train_loss_nats_per_token": 1.7140870258356051
    },
    {
      "epoch": 0.25796002212214586,
      "grad_norm": 0.177734375,
      "learning_rate": 0.0004825994299071469,
      "loss": 1.7246208190917969,
      "step": 112180,
      "train_bpb": 0.613724667943373,
      "train_bytes": 28008.1625,
      "train_loss_nats_per_token": 1.7250468041779816
    },
    {
      "epoch": 0.2580917015617182,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.0004821412535879802,
      "loss": 1.7081174850463867,
      "step": 112200,
      "train_bpb": 0.6033620261734498,
      "train_bytes": 28388.41875,
      "train_loss_nats_per_token": 1.708607258984731
    },
    {
      "epoch": 0.25822338100129044,
      "grad_norm": 0.1279296875,
      "learning_rate": 0.00048168325321520515,
      "loss": 1.7009901046752929,
      "step": 112220,
      "train_bpb": 0.6019252613962056,
      "train_bytes": 28446.05625,
      "train_loss_nats_per_token": 1.702140259913021
    },
    {
      "epoch": 0.25835506044086276,
      "grad_norm": 0.15625,
      "learning_rate": 0.00048122542886799044,
      "loss": 1.7024892807006835,
      "step": 112240,
      "train_bpb": 0.6055748821474783,
      "train_bytes": 28058.36875,
      "train_loss_nats_per_token": 1.702793153233347
    },
    {
      "epoch": 0.2584867398804351,
      "grad_norm": 0.15625,
      "learning_rate": 0.00048076778062547534,
      "loss": 1.718125343322754,
      "step": 112260,
      "train_bpb": 0.6076044347258466,
      "train_bytes": 28177.75625,
      "train_loss_nats_per_token": 1.718266269497969
    },
    {
      "epoch": 0.2586184193200074,
      "grad_norm": 0.1875,
      "learning_rate": 0.00048031030856676894,
      "loss": 1.7126972198486328,
      "step": 112280,
      "train_bpb": 0.6065993307118875,
      "train_bytes": 28167.33125,
      "train_loss_nats_per_token": 1.7133813861572853
    },
    {
      "epoch": 0.25875009875957966,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.00047985301277095,
      "loss": 1.7142704010009766,
      "step": 112300,
      "train_bpb": 0.607880849357419,
      "train_bytes": 28362.20625,
      "train_loss_nats_per_token": 1.7151537628889677
    },
    {
      "epoch": 0.258881778199152,
      "grad_norm": 0.146484375,
      "learning_rate": 0.0004793958933170655,
      "loss": 1.714838409423828,
      "step": 112320,
      "train_bpb": 0.6056600826183041,
      "train_bytes": 28391.6375,
      "train_loss_nats_per_token": 1.7149576899172854
    },
    {
      "epoch": 0.2590134576387243,
      "grad_norm": 0.12353515625,
      "learning_rate": 0.0004789389502841331,
      "loss": 1.7107749938964845,
      "step": 112340,
      "train_bpb": 0.6052112239289795,
      "train_bytes": 28210.96875,
      "train_loss_nats_per_token": 1.7105637879045545
    },
    {
      "epoch": 0.2591451370782966,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00047848218375114024,
      "loss": 1.7149339675903321,
      "step": 112360,
      "train_bpb": 0.6010943127128998,
      "train_bytes": 28716.19375,
      "train_loss_nats_per_token": 1.715675688374498
    },
    {
      "epoch": 0.2592768165178689,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004780255937970424,
      "loss": 1.7198221206665039,
      "step": 112380,
      "train_bpb": 0.6066426334068056,
      "train_bytes": 28566.85,
      "train_loss_nats_per_token": 1.7208736655033792
    },
    {
      "epoch": 0.2594084959574412,
      "grad_norm": 0.13671875,
      "learning_rate": 0.0004775691805007657,
      "loss": 1.7077987670898438,
      "step": 112400,
      "train_bpb": 0.6055475186456611,
      "train_bytes": 28338.7,
      "train_loss_nats_per_token": 1.7080106083102573
    },
    {
      "epoch": 0.2595401753970135,
      "grad_norm": 0.1796875,
      "learning_rate": 0.00047711294394120544,
      "loss": 1.7098669052124023,
      "step": 112420,
      "train_bpb": 0.6030356619505438,
      "train_bytes": 28319.66875,
      "train_loss_nats_per_token": 1.7098653143721545
    },
    {
      "epoch": 0.25967185483658584,
      "grad_norm": 0.16796875,
      "learning_rate": 0.00047665688419722726,
      "loss": 1.7018596649169921,
      "step": 112440,
      "train_bpb": 0.602684775647238,
      "train_bytes": 28435.9875,
      "train_loss_nats_per_token": 1.703121462976093
    },
    {
      "epoch": 0.2598035342761581,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00047620100134766383,
      "loss": 1.698480987548828,
      "step": 112460,
      "train_bpb": 0.6027962927228817,
      "train_bytes": 28445.2375,
      "train_loss_nats_per_token": 1.6984630236905232
    },
    {
      "epoch": 0.2599352137157304,
      "grad_norm": 0.1513671875,
      "learning_rate": 0.00047574529547131963,
      "loss": 1.723784828186035,
      "step": 112480,
      "train_bpb": 0.6106460399952878,
      "train_bytes": 28427.0375,
      "train_loss_nats_per_token": 1.7237769745326263
    },
    {
      "epoch": 0.26006689315530274,
      "grad_norm": 0.158203125,
      "learning_rate": 0.000475289766646968,
      "loss": 1.7148386001586915,
      "step": 112500,
      "train_bpb": 0.6088000731378229,
      "train_bytes": 28060.33125,
      "train_loss_nats_per_token": 1.7153685225670772
    },
    {
      "epoch": 0.26019857259487505,
      "grad_norm": 0.171875,
      "learning_rate": 0.0004748344149533504,
      "loss": 1.7064743041992188,
      "step": 112520,
      "train_bpb": 0.6002406646347535,
      "train_bytes": 28479.31875,
      "train_loss_nats_per_token": 1.7073623279655399
    },
    {
      "epoch": 0.2603302520344473,
      "grad_norm": 0.1484375,
      "learning_rate": 0.00047437924046917916,
      "loss": 1.712114143371582,
      "step": 112540,
      "train_bpb": 0.6074586178825491,
      "train_bytes": 28386.30625,
      "train_loss_nats_per_token": 1.7129235143154453
    },
    {
      "epoch": 0.26046193147401964,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00047392424327313593,
      "loss": 1.7185453414916991,
      "step": 112560,
      "train_bpb": 0.6036069714108798,
      "train_bytes": 28584.325,
      "train_loss_nats_per_token": 1.7188857918034322
    },
    {
      "epoch": 0.26059361091359196,
      "grad_norm": 0.17578125,
      "learning_rate": 0.0004734694234438701,
      "loss": 1.716623306274414,
      "step": 112580,
      "train_bpb": 0.6066239713880132,
      "train_bytes": 28549.10625,
      "train_loss_nats_per_token": 1.7164767336809623
    },
    {
      "epoch": 0.2607252903531643,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00047301478106000217,
      "loss": 1.7295755386352538,
      "step": 112600,
      "train_bpb": 0.6140024461126097,
      "train_bytes": 28136.36875,
      "train_loss_nats_per_token": 1.7308018415229487
    },
    {
      "epoch": 0.26085696979273654,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.0004725603162001214,
      "loss": 1.7179555892944336,
      "step": 112620,
      "train_bpb": 0.6084629429205003,
      "train_bytes": 28211.9,
      "train_loss_nats_per_token": 1.718829849429561
    },
    {
      "epoch": 0.26098864923230886,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0004721060289427865,
      "loss": 1.7047863006591797,
      "step": 112640,
      "train_bpb": 0.6081124763512991,
      "train_bytes": 27917.71875,
      "train_loss_nats_per_token": 1.7058024423639895
    },
    {
      "epoch": 0.2611203286718812,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.0004716519193665244,
      "loss": 1.719975471496582,
      "step": 112660,
      "train_bpb": 0.6081313961899354,
      "train_bytes": 28470.5125,
      "train_loss_nats_per_token": 1.7207043650465828
    },
    {
      "epoch": 0.2612520081114535,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.0004711979875498329,
      "loss": 1.7152769088745117,
      "step": 112680,
      "train_bpb": 0.609528783950375,
      "train_bytes": 28143.38125,
      "train_loss_nats_per_token": 1.715982512115283
    },
    {
      "epoch": 0.26138368755102576,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004707442335711785,
      "loss": 1.7118122100830078,
      "step": 112700,
      "train_bpb": 0.6042736501280693,
      "train_bytes": 28580.5125,
      "train_loss_nats_per_token": 1.7128247292152734
    },
    {
      "epoch": 0.2615153669905981,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.00047029065750899603,
      "loss": 1.7037208557128907,
      "step": 112720,
      "train_bpb": 0.6074102021873128,
      "train_bytes": 28317.81875,
      "train_loss_nats_per_token": 1.7042231633268177
    },
    {
      "epoch": 0.2616470464301704,
      "grad_norm": 0.1494140625,
      "learning_rate": 0.0004698372594416907,
      "loss": 1.7190345764160155,
      "step": 112740,
      "train_bpb": 0.6102740008651131,
      "train_bytes": 28117.83125,
      "train_loss_nats_per_token": 1.7202539639469057
    },
    {
      "epoch": 0.2617787258697427,
      "grad_norm": 0.123046875,
      "learning_rate": 0.00046938403944763737,
      "loss": 1.7114696502685547,
      "step": 112760,
      "train_bpb": 0.6051390401752262,
      "train_bytes": 28537.7125,
      "train_loss_nats_per_token": 1.711536615838769
    },
    {
      "epoch": 0.26191040530931503,
      "grad_norm": 0.11865234375,
      "learning_rate": 0.000468930997605178,
      "loss": 1.712263298034668,
      "step": 112780,
      "train_bpb": 0.6072760984506071,
      "train_bytes": 28318.525,
      "train_loss_nats_per_token": 1.7120616890384956
    },
    {
      "epoch": 0.2620420847488873,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.00046847813399262595,
      "loss": 1.7162817001342774,
      "step": 112800,
      "train_bpb": 0.6072636773829526,
      "train_bytes": 28283.18125,
      "train_loss_nats_per_token": 1.7172739362106746
    },
    {
      "epoch": 0.2621737641884596,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00046802544868826266,
      "loss": 1.7129501342773437,
      "step": 112820,
      "train_bpb": 0.6094690304649477,
      "train_bytes": 28157.05625,
      "train_loss_nats_per_token": 1.7134981997544083
    },
    {
      "epoch": 0.26230544362803193,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0004675729417703396,
      "loss": 1.720880126953125,
      "step": 112840,
      "train_bpb": 0.6050425873758317,
      "train_bytes": 28567.89375,
      "train_loss_nats_per_token": 1.7216139391665986
    },
    {
      "epoch": 0.26243712306760425,
      "grad_norm": 0.1416015625,
      "learning_rate": 0.0004671206133170758,
      "loss": 1.710806655883789,
      "step": 112860,
      "train_bpb": 0.6065351567335873,
      "train_bytes": 28112.71875,
      "train_loss_nats_per_token": 1.711526608406553
    },
    {
      "epoch": 0.2625688025071765,
      "grad_norm": 0.1298828125,
      "learning_rate": 0.00046666846340666097,
      "loss": 1.7081989288330077,
      "step": 112880,
      "train_bpb": 0.6070527992410725,
      "train_bytes": 28133.4,
      "train_loss_nats_per_token": 1.7086262400935721
    },
    {
      "epoch": 0.26270048194674883,
      "grad_norm": 0.12890625,
      "learning_rate": 0.000466216492117254,
      "loss": 1.6991611480712892,
      "step": 112900,
      "train_bpb": 0.604289872822359,
      "train_bytes": 28054.51875,
      "train_loss_nats_per_token": 1.700161947341647
    },
    {
      "epoch": 0.26283216138632115,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004657646995269813,
      "loss": 1.7058767318725585,
      "step": 112920,
      "train_bpb": 0.6058282116422978,
      "train_bytes": 28264.60625,
      "train_loss_nats_per_token": 1.7059206532079945
    },
    {
      "epoch": 0.26296384082589347,
      "grad_norm": 0.11962890625,
      "learning_rate": 0.00046531308571394,
      "loss": 1.7106138229370118,
      "step": 112940,
      "train_bpb": 0.6034655485006826,
      "train_bytes": 28361.05,
      "train_loss_nats_per_token": 1.7114970436000156
    },
    {
      "epoch": 0.26309552026546573,
      "grad_norm": 0.130859375,
      "learning_rate": 0.0004648616507561956,
      "loss": 1.7054399490356444,
      "step": 112960,
      "train_bpb": 0.6037052911064502,
      "train_bytes": 28316.55,
      "train_loss_nats_per_token": 1.7063855365560732
    },
    {
      "epoch": 0.26322719970503805,
      "grad_norm": 0.14453125,
      "learning_rate": 0.00046441039473178345,
      "loss": 1.7066024780273437,
      "step": 112980,
      "train_bpb": 0.6068884148860412,
      "train_bytes": 28415.41875,
      "train_loss_nats_per_token": 1.7072612741682422
    },
    {
      "epoch": 0.26335887914461037,
      "grad_norm": 0.1767578125,
      "learning_rate": 0.0004639593177187065,
      "loss": 1.714677619934082,
      "step": 113000,
      "train_bpb": 0.6069234714716806,
      "train_bytes": 27944.05625,
      "train_loss_nats_per_token": 1.7156954596072989
    },
    {
      "epoch": 0.26335887914461037,
      "eval_loss": 1.5305984020233154,
      "eval_runtime": 71.3986,
      "eval_samples_per_second": 14.006,
      "eval_steps_per_second": 14.006,
      "step": 113000,
      "train_bpb": 0.5401517620407228,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.538143028850775
    },
    {
      "epoch": 0.2634905585841827,
      "grad_norm": 0.1591796875,
      "learning_rate": 0.000463508419794938,
      "loss": 1.7135343551635742,
      "step": 113020,
      "train_bpb": 0.6068658256802405,
      "train_bytes": 28305.2,
      "train_loss_nats_per_token": 1.714714649027483
    },
    {
      "epoch": 0.26362223802375495,
      "grad_norm": 0.150390625,
      "learning_rate": 0.00046305770103842025,
      "loss": 1.7205350875854493,
      "step": 113040,
      "train_bpb": 0.6087808245669194,
      "train_bytes": 28335.5375,
      "train_loss_nats_per_token": 1.7215228912920213
    },
    {
      "epoch": 0.26375391746332727,
      "grad_norm": 0.138671875,
      "learning_rate": 0.0004626071615270634,
      "loss": 1.6999040603637696,
      "step": 113060,
      "train_bpb": 0.6012678551163559,
      "train_bytes": 28305.75,
      "train_loss_nats_per_token": 1.700166943325421
    },
    {
      "epoch": 0.2638855969028996,
      "grad_norm": 0.1328125,
      "learning_rate": 0.00046215680133874777,
      "loss": 1.7151802062988282,
      "step": 113080,
      "train_bpb": 0.6041728996450315,
      "train_bytes": 28361.9875,
      "train_loss_nats_per_token": 1.7151248777027377
    },
    {
      "epoch": 0.2640172763424719,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00046170662055132293,
      "loss": 1.709501075744629,
      "step": 113100,
      "train_bpb": 0.6045809635391713,
      "train_bytes": 28138.825,
      "train_loss_nats_per_token": 1.7099256245983965
    },
    {
      "epoch": 0.26414895578204417,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00046125661924260587,
      "loss": 1.7165925979614258,
      "step": 113120,
      "train_bpb": 0.6044648919706884,
      "train_bytes": 28589.1125,
      "train_loss_nats_per_token": 1.7171746950413338
    },
    {
      "epoch": 0.2642806352216165,
      "grad_norm": 0.1787109375,
      "learning_rate": 0.0004608067974903838,
      "loss": 1.7125377655029297,
      "step": 113140,
      "train_bpb": 0.607021641994782,
      "train_bytes": 28186.94375,
      "train_loss_nats_per_token": 1.713827928701598
    },
    {
      "epoch": 0.2644123146611888,
      "grad_norm": 0.134765625,
      "learning_rate": 0.00046035715537241285,
      "loss": 1.713718795776367,
      "step": 113160,
      "train_bpb": 0.6044116017877804,
      "train_bytes": 28511.85625,
      "train_loss_nats_per_token": 1.7142480187138136
    },
    {
      "epoch": 0.2645439941007611,
      "grad_norm": 0.15234375,
      "learning_rate": 0.0004599076929664182,
      "loss": 1.7082695007324218,
      "step": 113180,
      "train_bpb": 0.6008047397089323,
      "train_bytes": 28447.45,
      "train_loss_nats_per_token": 1.7088103813330124
    },
    {
      "epoch": 0.2646756735403334,
      "grad_norm": 0.138671875,
      "learning_rate": 0.00045945841035009255,
      "loss": 1.7149255752563477,
      "step": 113200,
      "train_bpb": 0.6092861318267723,
      "train_bytes": 27931.55625,
      "train_loss_nats_per_token": 1.7162936463557343
    },
    {
      "epoch": 0.2648073529799057,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.00045900930760109945,
      "loss": 1.7101079940795898,
      "step": 113220,
      "train_bpb": 0.6015243794537583,
      "train_bytes": 28886.28125,
      "train_loss_nats_per_token": 1.7097995673389386
    },
    {
      "epoch": 0.264939032419478,
      "grad_norm": 0.1865234375,
      "learning_rate": 0.00045856038479707056,
      "loss": 1.715529441833496,
      "step": 113240,
      "train_bpb": 0.6071814368791437,
      "train_bytes": 28046.04375,
      "train_loss_nats_per_token": 1.7159290228302058
    },
    {
      "epoch": 0.26507071185905035,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0004581116420156059,
      "loss": 1.7139863967895508,
      "step": 113260,
      "train_bpb": 0.6062421504683279,
      "train_bytes": 28404.54375,
      "train_loss_nats_per_token": 1.7152852945641712
    },
    {
      "epoch": 0.2652023912986226,
      "grad_norm": 0.1572265625,
      "learning_rate": 0.00045766307933427485,
      "loss": 1.7228612899780273,
      "step": 113280,
      "train_bpb": 0.6077173328906628,
      "train_bytes": 28572.76875,
      "train_loss_nats_per_token": 1.723142367235616
    },
    {
      "epoch": 0.2653340707381949,
      "grad_norm": 0.126953125,
      "learning_rate": 0.00045721469683061627,
      "loss": 1.7301368713378906,
      "step": 113300,
      "train_bpb": 0.6084341900184179,
      "train_bytes": 28427.31875,
      "train_loss_nats_per_token": 1.7306240966971547
    },
    {
      "epoch": 0.26546575017776725,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.00045676649458213664,
      "loss": 1.7133804321289063,
      "step": 113320,
      "train_bpb": 0.6081095458486077,
      "train_bytes": 28359.675,
      "train_loss_nats_per_token": 1.7138349273384883
    },
    {
      "epoch": 0.26559742961733956,
      "grad_norm": 0.181640625,
      "learning_rate": 0.000456318472666312,
      "loss": 1.7229740142822265,
      "step": 113340,
      "train_bpb": 0.6105089165353005,
      "train_bytes": 28172.775,
      "train_loss_nats_per_token": 1.7241168277932564
    },
    {
      "epoch": 0.2657291090569118,
      "grad_norm": 0.1201171875,
      "learning_rate": 0.0004558706311605872,
      "loss": 1.7158979415893554,
      "step": 113360,
      "train_bpb": 0.6090560712507856,
      "train_bytes": 27968.35625,
      "train_loss_nats_per_token": 1.716802217532059
    },
    {
      "epoch": 0.26586078849648415,
      "grad_norm": 0.13671875,
      "learning_rate": 0.00045542297014237635,
      "loss": 1.721734619140625,
      "step": 113380,
      "train_bpb": 0.6067241091338977,
      "train_bytes": 28407.83125,
      "train_loss_nats_per_token": 1.7225922791041404
    },
    {
      "epoch": 0.26599246793605646,
      "grad_norm": 0.1181640625,
      "learning_rate": 0.0004549754896890608,
      "loss": 1.7202964782714845,
      "step": 113400,
      "train_bpb": 0.6074856120060708,
      "train_bytes": 28357.075,
      "train_loss_nats_per_token": 1.7209422398077332
    },
    {
      "epoch": 0.2661241473756288,
      "grad_norm": 0.18359375,
      "learning_rate": 0.0004545281898779925,
      "loss": 1.7135528564453124,
      "step": 113420,
      "train_bpb": 0.6052719786513064,
      "train_bytes": 28373.28125,
      "train_loss_nats_per_token": 1.7138017546886177
    },
    {
      "epoch": 0.2662558268152011,
      "grad_norm": 0.1318359375,
      "learning_rate": 0.0004540810707864915,
      "loss": 1.7115644454956054,
      "step": 113440,
      "train_bpb": 0.6096391865654823,
      "train_bytes": 27977.01875,
      "train_loss_nats_per_token": 1.7134301900233024
    },
    {
      "epoch": 0.26638750625477337,
      "grad_norm": 0.1455078125,
      "learning_rate": 0.00045363413249184574,
      "loss": 1.7266912460327148,
      "step": 113460,
      "train_bpb": 0.606672275655334,
      "train_bytes": 28304.4875,
      "train_loss_nats_per_token": 1.7270534722282052
    },
    {
      "epoch": 0.2665191856943457,
      "grad_norm": 0.146484375,
      "learning_rate": 0.00045318737507131306,
      "loss": 1.7155292510986329,
      "step": 113480,
      "train_bpb": 0.6035953954846283,
      "train_bytes": 28560.80625,
      "train_loss_nats_per_token": 1.7155954108349447
    },
    {
      "epoch": 0.266650865133918,
      "grad_norm": 0.12890625,
      "learning_rate": 0.00045274079860211996,
      "loss": 1.712361717224121,
      "step": 113500,
      "train_bpb": 0.6058649571684436,
      "train_bytes": 28172.875,
      "train_loss_nats_per_token": 1.713633408334182
    },
    {
      "epoch": 0.2667825445734903,
      "grad_norm": 0.142578125,
      "learning_rate": 0.00045229440316146136,
      "loss": 1.711868667602539,
      "step": 113520,
      "train_bpb": 0.6080084028087712,
      "train_bytes": 28122.1875,
      "train_loss_nats_per_token": 1.712159864101532
    },
    {
      "epoch": 0.2669142240130626,
      "grad_norm": 0.1376953125,
      "learning_rate": 0.00045184818882650036,
      "loss": 1.7225061416625977,
      "step": 113540,
      "train_bpb": 0.6088244306338705,
      "train_bytes": 28372.1125,
      "train_loss_nats_per_token": 1.723818832982979
    },
    {
      "epoch": 0.2670459034526349,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.00045140215567436944,
      "loss": 1.7262224197387694,
      "step": 113560,
      "train_bpb": 0.6109339805455727,
      "train_bytes": 28117.45,
      "train_loss_nats_per_token": 1.7271963807835657
    },
    {
      "epoch": 0.2671775828922072,
      "grad_norm": 0.1259765625,
      "learning_rate": 0.00045095630378217023,
      "loss": 1.7202932357788085,
      "step": 113580,
      "train_bpb": 0.6104706339680245,
      "train_bytes": 28231.73125,
      "train_loss_nats_per_token": 1.7204895784013816
    },
    {
      "epoch": 0.26730926233177954,
      "grad_norm": 0.1337890625,
      "learning_rate": 0.0004505106332269713,
      "loss": 1.6991752624511718,
      "step": 113600,
      "train_bpb": 0.6076966382662468,
      "train_bytes": 27803.8,
      "train_loss_nats_per_token": 1.699022441903039
    },
    {
      "epoch": 0.2674409417713518,
      "grad_norm": 0.1962890625,
      "learning_rate": 0.0004500651440858116,
      "loss": 1.7134334564208984,
      "step": 113620,
      "train_bpb": 0.6069317062161361,
      "train_bytes": 28111.1375,
      "train_loss_nats_per_token": 1.7128655163227353
    },
    {
      "epoch": 0.2675726212109241,
      "grad_norm": 0.11767578125,
      "learning_rate": 0.00044961983643569865,
      "loss": 1.7225822448730468,
      "step": 113640,
      "train_bpb": 0.6124373999791285,
      "train_bytes": 28318.53125,
      "train_loss_nats_per_token": 1.7228749859482753
    },
    {
      "epoch": 0.26770430065049644,
      "grad_norm": 0.12060546875,
      "learning_rate": 0.0004491747103536068,
      "loss": 1.7134218215942383,
      "step": 113660,
      "train_bpb": 0.6060777485602616,
      "train_bytes": 28424.9625,
      "train_loss_nats_per_token": 1.7142683364093463
    },
    {
      "epoch": 0.26783598009006876,
      "grad_norm": 0.1611328125,
      "learning_rate": 0.0004487297659164808,
      "loss": 1.7203998565673828,
      "step": 113680,
      "train_bpb": 0.6077761547020325,
      "train_bytes": 28302.3,
      "train_loss_nats_per_token": 1.7209322442666701
    },
    {
      "epoch": 0.267967659529641,
      "grad_norm": 0.1474609375,
      "learning_rate": 0.0004482850032012338,
      "loss": 1.7139318466186524,
      "step": 113700,
      "train_bpb": 0.604821394056659,
      "train_bytes": 28571.38125,
      "train_loss_nats_per_token": 1.7140441814450458
    },
    {
      "epoch": 0.26809933896921334,
      "grad_norm": 0.16015625,
      "learning_rate": 0.00044784042228474707,
      "loss": 1.7165063858032226,
      "step": 113720,
      "train_bpb": 0.604090914441222,
      "train_bytes": 28267.06875,
      "train_loss_nats_per_token": 1.7167791320019845
    },
    {
      "epoch": 0.26823101840878566,
      "grad_norm": 0.1396484375,
      "learning_rate": 0.0004473960232438701,
      "loss": 1.7189950942993164,
      "step": 113740,
      "train_bpb": 0.6078830570675214,
      "train_bytes": 28394.76875,
      "train_loss_nats_per_token": 1.7197516479377037
    },
    {
      "epoch": 0.268362697848358,
      "grad_norm": 0.1435546875,
      "learning_rate": 0.0004469518061554218,
      "loss": 1.7304641723632812,
      "step": 113760,
      "train_bpb": 0.6120029651264004,
      "train_bytes": 28699.125,
      "train_loss_nats_per_token": 1.7313797256618875
    },
    {
      "epoch": 0.26849437728793024,
      "grad_norm": 0.1240234375,
      "learning_rate": 0.0004465077710961895,
      "loss": 1.7216175079345704,
      "step": 113780,
      "train_bpb": 0.6080754872041422,
      "train_bytes": 28424.18125,
      "train_loss_nats_per_token": 1.7225035580929
    },
    {
      "epoch": 0.26862605672750256,
      "grad_norm": 0.16015625,
      "learning_rate": 0.000446063918142928,
      "loss": 1.712874984741211,
      "step": 113800,
      "train_bpb": 0.607395172290262,
      "train_bytes": 28020.95,
      "train_loss_nats_per_token": 1.7135570726862321
    },
    {
      "epoch": 0.2687577361670749,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.0004456202473723621,
      "loss": 1.716609573364258,
      "step": 113820,
      "train_bpb": 0.606213193226077,
      "train_bytes": 28191.15625,
      "train_loss_nats_per_token": 1.7178972460123523
    },
    {
      "epoch": 0.2688894156066472,
      "grad_norm": 0.1640625,
      "learning_rate": 0.00044517675886118416,
      "loss": 1.7191993713378906,
      "step": 113840,
      "train_bpb": 0.6084580788064012,
      "train_bytes": 28040.65,
      "train_loss_nats_per_token": 1.720824419316272
    },
    {
      "epoch": 0.26902109504621946,
      "grad_norm": 0.162109375,
      "learning_rate": 0.00044473345268605574,
      "loss": 1.7162294387817383,
      "step": 113860,
      "train_bpb": 0.6053623360785676,
      "train_bytes": 28257.3,
      "train_loss_nats_per_token": 1.716537823632671
    },
    {
      "epoch": 0.2691527744857918,
      "grad_norm": 0.12451171875,
      "learning_rate": 0.0004442903289236058,
      "loss": 1.7215482711791992,
      "step": 113880,
      "train_bpb": 0.6094090903934227,
      "train_bytes": 28203.45625,
      "train_loss_nats_per_token": 1.7224273879086154
    },
    {
      "epoch": 0.2692844539253641,
      "grad_norm": 0.1533203125,
      "learning_rate": 0.00044384738765043257,
      "loss": 1.7140342712402343,
      "step": 113900,
      "train_bpb": 0.6067352118773068,
      "train_bytes": 28289.16875,
      "train_loss_nats_per_token": 1.7145058171535146
    },
    {
      "epoch": 0.2694161333649364,
      "grad_norm": 0.11181640625,
      "learning_rate": 0.00044340462894310327,
      "loss": 1.715474510192871,
      "step": 113920,
      "train_bpb": 0.6066727888684618,
      "train_bytes": 28278.5875,
      "train_loss_nats_per_token": 1.716450873661507
    },
    {
      "epoch": 0.2695478128045087,
      "grad_norm": 0.154296875,
      "learning_rate": 0.00044296205287815203,
      "loss": 1.7226783752441406,
      "step": 113940,
      "train_bpb": 0.6101393859078104,
      "train_bytes": 27968.66875,
      "train_loss_nats_per_token": 1.7240551813035707
    },
    {
      "epoch": 0.269679492244081,
      "grad_norm": 0.1689453125,
      "learning_rate": 0.0004425196595320825,
      "loss": 1.731062126159668,
      "step": 113960,
      "train_bpb": 0.6130780960461476,
      "train_bytes": 28228.58125,
      "train_loss_nats_per_token": 1.731573170187102
    },
    {
      "epoch": 0.2698111716836533,
      "grad_norm": 0.1357421875,
      "learning_rate": 0.000442077448981367,
      "loss": 1.7237489700317383,
      "step": 113980,
      "train_bpb": 0.6109245109502628,
      "train_bytes": 28196.06875,
      "train_loss_nats_per_token": 1.724765200455909
    },
    {
      "epoch": 0.26994285112322564,
      "grad_norm": 0.216796875,
      "learning_rate": 0.000441635421302445,
      "loss": 1.7099153518676757,
      "step": 114000,
      "train_bpb": 0.6055746093956114,
      "train_bytes": 28263.025,
      "train_loss_nats_per_token": 1.7111042386211976
    },
    {
      "epoch": 0.26994285112322564,
      "eval_loss": 1.530694603919983,
      "eval_runtime": 72.227,
      "eval_samples_per_second": 13.845,
      "eval_steps_per_second": 13.845,
      "step": 114000,
      "train_bpb": 0.5401847529394688,
      "train_bytes": 3589.691,
      "train_loss_nats_per_token": 1.5382369741537192
    }
  ],
  "logging_steps": 20,
  "max_steps": 151884,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.876203156799488e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}