{
  "best_global_step": 4500,
  "best_metric": 0.05329994,
  "best_model_checkpoint": "/workspace/ms-sw/swift-training-workflow/runs/qwen3-vl-4b-latex-ocr-full/v0-20260327-112050/checkpoint-4500",
  "epoch": 2.830188679245283,
  "eval_steps": 500,
  "global_step": 4500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0006289308176100629,
      "grad_norm": 45.34716768625641,
      "learning_rate": 8.368200836820084e-08,
      "loss": 0.7045165300369263,
      "step": 1,
      "token_acc": 0.9067005937234945
    },
    {
      "epoch": 0.006289308176100629,
      "grad_norm": 20.89321839694586,
      "learning_rate": 8.368200836820084e-07,
      "loss": 0.6605887942843967,
      "step": 10,
      "token_acc": 0.9088069388002669
    },
    {
      "epoch": 0.012578616352201259,
      "grad_norm": 8.70654604666805,
      "learning_rate": 1.6736401673640167e-06,
      "loss": 0.4156378746032715,
      "step": 20,
      "token_acc": 0.9231436001155735
    },
    {
      "epoch": 0.018867924528301886,
      "grad_norm": 4.519491564596013,
      "learning_rate": 2.510460251046025e-06,
      "loss": 0.19132455587387084,
      "step": 30,
      "token_acc": 0.9457678075855689
    },
    {
      "epoch": 0.025157232704402517,
      "grad_norm": 6.196919953968438,
      "learning_rate": 3.3472803347280334e-06,
      "loss": 0.1531017780303955,
      "step": 40,
      "token_acc": 0.9549960123048877
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 3.6004202812134203,
      "learning_rate": 4.184100418410042e-06,
      "loss": 0.15698001384735108,
      "step": 50,
      "token_acc": 0.9532592069655254
    },
    {
      "epoch": 0.03773584905660377,
      "grad_norm": 3.115974891699392,
      "learning_rate": 5.02092050209205e-06,
      "loss": 0.13419015407562257,
      "step": 60,
      "token_acc": 0.9603841536614646
    },
    {
      "epoch": 0.0440251572327044,
      "grad_norm": 3.53139646167736,
      "learning_rate": 5.8577405857740585e-06,
      "loss": 0.13604100942611694,
      "step": 70,
      "token_acc": 0.9596007184711876
    },
    {
      "epoch": 0.050314465408805034,
      "grad_norm": 3.0256135958409387,
      "learning_rate": 6.694560669456067e-06,
      "loss": 0.13651411533355712,
      "step": 80,
      "token_acc": 0.9602559188607483
    },
    {
      "epoch": 0.05660377358490566,
      "grad_norm": 3.811738769540206,
      "learning_rate": 7.531380753138075e-06,
      "loss": 0.12942522764205933,
      "step": 90,
      "token_acc": 0.96127742262525
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 3.074401932705618,
      "learning_rate": 8.368200836820084e-06,
      "loss": 0.13949081897735596,
      "step": 100,
      "token_acc": 0.959375
    },
    {
      "epoch": 0.06918238993710692,
      "grad_norm": 2.843989040291082,
      "learning_rate": 9.205020920502092e-06,
      "loss": 0.13196282386779784,
      "step": 110,
      "token_acc": 0.9607720869589864
    },
    {
      "epoch": 0.07547169811320754,
      "grad_norm": 3.508369382446498,
      "learning_rate": 1.00418410041841e-05,
      "loss": 0.13052265644073485,
      "step": 120,
      "token_acc": 0.9622684985166623
    },
    {
      "epoch": 0.08176100628930817,
      "grad_norm": 3.440200280061563,
      "learning_rate": 1.0878661087866109e-05,
      "loss": 0.1312979578971863,
      "step": 130,
      "token_acc": 0.9609211828075189
    },
    {
      "epoch": 0.0880503144654088,
      "grad_norm": 4.201954871869706,
      "learning_rate": 1.1715481171548117e-05,
      "loss": 0.13678085803985596,
      "step": 140,
      "token_acc": 0.959023862406743
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 3.2305103935546957,
      "learning_rate": 1.2552301255230125e-05,
      "loss": 0.13586316108703614,
      "step": 150,
      "token_acc": 0.9603519195612431
    },
    {
      "epoch": 0.10062893081761007,
      "grad_norm": 3.1732101058142845,
      "learning_rate": 1.3389121338912134e-05,
      "loss": 0.14661180973052979,
      "step": 160,
      "token_acc": 0.9567754424778762
    },
    {
      "epoch": 0.1069182389937107,
      "grad_norm": 3.0981254219351912,
      "learning_rate": 1.4225941422594142e-05,
      "loss": 0.1456087589263916,
      "step": 170,
      "token_acc": 0.9564570883483433
    },
    {
      "epoch": 0.11320754716981132,
      "grad_norm": 2.7844698698117694,
      "learning_rate": 1.506276150627615e-05,
      "loss": 0.13063851594924927,
      "step": 180,
      "token_acc": 0.9629532927183058
    },
    {
      "epoch": 0.11949685534591195,
      "grad_norm": 2.582043948631837,
      "learning_rate": 1.589958158995816e-05,
      "loss": 0.1434727668762207,
      "step": 190,
      "token_acc": 0.9585141315820336
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 2.5184581779739084,
      "learning_rate": 1.6736401673640167e-05,
      "loss": 0.13957881927490234,
      "step": 200,
      "token_acc": 0.9597336624286084
    },
    {
      "epoch": 0.1320754716981132,
      "grad_norm": 3.8266489180800014,
      "learning_rate": 1.7573221757322177e-05,
      "loss": 0.14468042850494384,
      "step": 210,
      "token_acc": 0.9577662246626555
    },
    {
      "epoch": 0.13836477987421383,
      "grad_norm": 2.9201805311014466,
      "learning_rate": 1.8410041841004184e-05,
      "loss": 0.13508262634277343,
      "step": 220,
      "token_acc": 0.9598491126796788
    },
    {
      "epoch": 0.14465408805031446,
      "grad_norm": 2.9676576361919316,
      "learning_rate": 1.9246861924686194e-05,
      "loss": 0.1461448550224304,
      "step": 230,
      "token_acc": 0.9563967275314205
    },
    {
      "epoch": 0.1509433962264151,
      "grad_norm": 2.9276617136601097,
      "learning_rate": 1.999999759629256e-05,
      "loss": 0.15765782594680786,
      "step": 240,
      "token_acc": 0.9533182566937488
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 2.6683614726041727,
      "learning_rate": 1.9999709152797814e-05,
      "loss": 0.15170300006866455,
      "step": 250,
      "token_acc": 0.9559203868210953
    },
    {
      "epoch": 0.16352201257861634,
      "grad_norm": 3.322241566248514,
      "learning_rate": 1.9998939983703623e-05,
      "loss": 0.15328752994537354,
      "step": 260,
      "token_acc": 0.9542057698992006
    },
    {
      "epoch": 0.16981132075471697,
      "grad_norm": 4.064876637718516,
      "learning_rate": 1.9997690125987e-05,
      "loss": 0.14633382558822633,
      "step": 270,
      "token_acc": 0.9561424123142
    },
    {
      "epoch": 0.1761006289308176,
      "grad_norm": 3.1878547590503694,
      "learning_rate": 1.9995959639733544e-05,
      "loss": 0.15381388664245604,
      "step": 280,
      "token_acc": 0.954816314723749
    },
    {
      "epoch": 0.18238993710691823,
      "grad_norm": 4.212001601278116,
      "learning_rate": 1.999374860813458e-05,
      "loss": 0.3573228597640991,
      "step": 290,
      "token_acc": 0.9125418060200668
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 2.6659838103160354,
      "learning_rate": 1.9991057137483156e-05,
      "loss": 0.1465816617012024,
      "step": 300,
      "token_acc": 0.9558244474463254
    },
    {
      "epoch": 0.1949685534591195,
      "grad_norm": 2.752984167443659,
      "learning_rate": 1.998788535716891e-05,
      "loss": 0.1465825319290161,
      "step": 310,
      "token_acc": 0.9560354159149574
    },
    {
      "epoch": 0.20125786163522014,
      "grad_norm": 2.519397967504351,
      "learning_rate": 1.998423341967188e-05,
      "loss": 0.15822311639785766,
      "step": 320,
      "token_acc": 0.953077512994575
    },
    {
      "epoch": 0.20754716981132076,
      "grad_norm": 2.5939942368942885,
      "learning_rate": 1.9980101500555155e-05,
      "loss": 0.15183796882629394,
      "step": 330,
      "token_acc": 0.9557981547954553
    },
    {
      "epoch": 0.2138364779874214,
      "grad_norm": 3.36249390155496,
      "learning_rate": 1.9975489798456443e-05,
      "loss": 0.14122436046600342,
      "step": 340,
      "token_acc": 0.9577983222051019
    },
    {
      "epoch": 0.22012578616352202,
      "grad_norm": 2.532046774055889,
      "learning_rate": 1.9970398535078518e-05,
      "loss": 0.13906241655349733,
      "step": 350,
      "token_acc": 0.959155496749698
    },
    {
      "epoch": 0.22641509433962265,
      "grad_norm": 2.8758559441861946,
      "learning_rate": 1.9964827955178566e-05,
      "loss": 0.14161758422851561,
      "step": 360,
      "token_acc": 0.9578302863056226
    },
    {
      "epoch": 0.23270440251572327,
      "grad_norm": 2.510119368927116,
      "learning_rate": 1.99587783265564e-05,
      "loss": 0.13756366968154907,
      "step": 370,
      "token_acc": 0.959952686718486
    },
    {
      "epoch": 0.2389937106918239,
      "grad_norm": 3.128855162887323,
      "learning_rate": 1.995224994004163e-05,
      "loss": 0.13279154300689697,
      "step": 380,
      "token_acc": 0.9616177452602851
    },
    {
      "epoch": 0.24528301886792453,
      "grad_norm": 3.7680233143063004,
      "learning_rate": 1.9945243109479626e-05,
      "loss": 0.14165163040161133,
      "step": 390,
      "token_acc": 0.9579345850999395
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 2.7668143237049803,
      "learning_rate": 1.9937758171716468e-05,
      "loss": 0.12863855361938475,
      "step": 400,
      "token_acc": 0.960668946727726
    },
    {
      "epoch": 0.2578616352201258,
      "grad_norm": 2.2895343375681247,
      "learning_rate": 1.9929795486582745e-05,
      "loss": 0.13654918670654298,
      "step": 410,
      "token_acc": 0.9608831657909922
    },
    {
      "epoch": 0.2641509433962264,
      "grad_norm": 2.4479950246869038,
      "learning_rate": 1.9921355436876242e-05,
      "loss": 0.1467982292175293,
      "step": 420,
      "token_acc": 0.9568501501769412
    },
    {
      "epoch": 0.27044025157232704,
      "grad_norm": 2.4852520566549514,
      "learning_rate": 1.9912438428343562e-05,
      "loss": 0.129807186126709,
      "step": 430,
      "token_acc": 0.9607044921267633
    },
    {
      "epoch": 0.27672955974842767,
      "grad_norm": 2.4052015257221973,
      "learning_rate": 1.9903044889660595e-05,
      "loss": 0.1309417724609375,
      "step": 440,
      "token_acc": 0.9612522150029533
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 2.13869320934904,
      "learning_rate": 1.989317527241193e-05,
      "loss": 0.13723220825195312,
      "step": 450,
      "token_acc": 0.959500215424386
    },
    {
      "epoch": 0.2893081761006289,
      "grad_norm": 2.594597791053857,
      "learning_rate": 1.9882830051069135e-05,
      "loss": 0.12807786464691162,
      "step": 460,
      "token_acc": 0.9608131343945865
    },
    {
      "epoch": 0.29559748427672955,
      "grad_norm": 2.5797553863526415,
      "learning_rate": 1.9872009722967943e-05,
      "loss": 0.12285339832305908,
      "step": 470,
      "token_acc": 0.9625046229125772
    },
    {
      "epoch": 0.3018867924528302,
      "grad_norm": 2.2886282297299254,
      "learning_rate": 1.9860714808284364e-05,
      "loss": 0.1279349684715271,
      "step": 480,
      "token_acc": 0.9615531023981728
    },
    {
      "epoch": 0.3081761006289308,
      "grad_norm": 2.137927259030621,
      "learning_rate": 1.984894585000964e-05,
      "loss": 0.1333387613296509,
      "step": 490,
      "token_acc": 0.9601997026898157
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 3.1442649494651236,
      "learning_rate": 1.983670341392419e-05,
      "loss": 0.13746428489685059,
      "step": 500,
      "token_acc": 0.959205850392082
    },
    {
      "epoch": 0.31446540880503143,
      "eval_loss": 0.14142681658267975,
      "eval_runtime": 733.8473,
      "eval_samples_per_second": 11.549,
      "eval_steps_per_second": 2.888,
      "eval_token_acc": 0.958135771134416,
      "step": 500
    },
    {
      "epoch": 0.32075471698113206,
      "grad_norm": 2.1089789407513644,
      "learning_rate": 1.9823988088570365e-05,
      "loss": 0.13168094158172608,
      "step": 510,
      "token_acc": 0.9609981625274139
    },
    {
      "epoch": 0.3270440251572327,
      "grad_norm": 2.0388968384076707,
      "learning_rate": 1.9810800485224196e-05,
      "loss": 0.12072486877441406,
      "step": 520,
      "token_acc": 0.9647903704134442
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 1.92859398458753,
      "learning_rate": 1.9797141237865963e-05,
      "loss": 0.13192942142486572,
      "step": 530,
      "token_acc": 0.9618026930475405
    },
    {
      "epoch": 0.33962264150943394,
      "grad_norm": 2.279806525825656,
      "learning_rate": 1.9783011003149757e-05,
      "loss": 0.12413182258605956,
      "step": 540,
      "token_acc": 0.9630478289874409
    },
    {
      "epoch": 0.34591194968553457,
      "grad_norm": 1.7101152951853034,
      "learning_rate": 1.9768410460371893e-05,
      "loss": 0.11435494422912598,
      "step": 550,
      "token_acc": 0.9657330665904544
    },
    {
      "epoch": 0.3522012578616352,
      "grad_norm": 2.894784399552811,
      "learning_rate": 1.9753340311438252e-05,
      "loss": 0.12679414749145507,
      "step": 560,
      "token_acc": 0.9626316419339397
    },
    {
      "epoch": 0.3584905660377358,
      "grad_norm": 2.6598023214353863,
      "learning_rate": 1.9737801280830537e-05,
      "loss": 0.11091006994247436,
      "step": 570,
      "token_acc": 0.9672726190830383
    },
    {
      "epoch": 0.36477987421383645,
      "grad_norm": 2.114552593288772,
      "learning_rate": 1.9721794115571464e-05,
      "loss": 0.11245570182800294,
      "step": 580,
      "token_acc": 0.9667317874232
    },
    {
      "epoch": 0.3710691823899371,
      "grad_norm": 1.6750696590671157,
      "learning_rate": 1.9705319585188823e-05,
      "loss": 0.12619128227233886,
      "step": 590,
      "token_acc": 0.9624073017683971
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 2.158345192241544,
      "learning_rate": 1.9688378481678495e-05,
      "loss": 0.11731394529342651,
      "step": 600,
      "token_acc": 0.9648311615524731
    },
    {
      "epoch": 0.3836477987421384,
      "grad_norm": 1.473095460953491,
      "learning_rate": 1.9670971619466388e-05,
      "loss": 0.11155827045440674,
      "step": 610,
      "token_acc": 0.9661841244560735
    },
    {
      "epoch": 0.389937106918239,
      "grad_norm": 2.150767674822309,
      "learning_rate": 1.9653099835369263e-05,
      "loss": 0.12145018577575684,
      "step": 620,
      "token_acc": 0.964341614730072
    },
    {
      "epoch": 0.39622641509433965,
      "grad_norm": 2.528324828102242,
      "learning_rate": 1.963476398855452e-05,
      "loss": 0.10442175865173339,
      "step": 630,
      "token_acc": 0.9676074124107956
    },
    {
      "epoch": 0.4025157232704403,
      "grad_norm": 1.76400511136344,
      "learning_rate": 1.96159649604989e-05,
      "loss": 0.11156104803085327,
      "step": 640,
      "token_acc": 0.9659989999705874
    },
    {
      "epoch": 0.4088050314465409,
      "grad_norm": 1.9005627857000333,
      "learning_rate": 1.9596703654946078e-05,
      "loss": 0.106688392162323,
      "step": 650,
      "token_acc": 0.9679768786127168
    },
    {
      "epoch": 0.41509433962264153,
      "grad_norm": 2.160201745367065,
      "learning_rate": 1.9576980997863258e-05,
      "loss": 0.11403628587722778,
      "step": 660,
      "token_acc": 0.9654064597556078
    },
    {
      "epoch": 0.42138364779874216,
      "grad_norm": 1.9608680807748942,
      "learning_rate": 1.9556797937396635e-05,
      "loss": 0.11484429836273194,
      "step": 670,
      "token_acc": 0.9659063910667617
    },
    {
      "epoch": 0.4276729559748428,
      "grad_norm": 1.98197499802808,
      "learning_rate": 1.9536155443825808e-05,
      "loss": 0.11815891265869141,
      "step": 680,
      "token_acc": 0.9646716786091859
    },
    {
      "epoch": 0.4339622641509434,
      "grad_norm": 1.742292937981462,
      "learning_rate": 1.951505450951715e-05,
      "loss": 0.10748194456100464,
      "step": 690,
      "token_acc": 0.9683494044242767
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 1.4750243938418341,
      "learning_rate": 1.94934961488761e-05,
      "loss": 0.1117070198059082,
      "step": 700,
      "token_acc": 0.9667868396400489
    },
    {
      "epoch": 0.44654088050314467,
      "grad_norm": 1.7692436842372372,
      "learning_rate": 1.9471481398298378e-05,
      "loss": 0.11048619747161866,
      "step": 710,
      "token_acc": 0.9665264476863868
    },
    {
      "epoch": 0.4528301886792453,
      "grad_norm": 1.6345113315469586,
      "learning_rate": 1.9449011316120187e-05,
      "loss": 0.1079734206199646,
      "step": 720,
      "token_acc": 0.9672420166633492
    },
    {
      "epoch": 0.4591194968553459,
      "grad_norm": 2.8816190351055835,
      "learning_rate": 1.942608698256731e-05,
      "loss": 0.11024671792984009,
      "step": 730,
      "token_acc": 0.9670435621000965
    },
    {
      "epoch": 0.46540880503144655,
      "grad_norm": 2.0528406854349623,
      "learning_rate": 1.9402709499703212e-05,
      "loss": 0.1140676498413086,
      "step": 740,
      "token_acc": 0.9654670094258784
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 1.8157760955063753,
      "learning_rate": 1.9378879991376012e-05,
      "loss": 0.10537893772125244,
      "step": 750,
      "token_acc": 0.9682530570067168
    },
    {
      "epoch": 0.4779874213836478,
      "grad_norm": 1.861892023047174,
      "learning_rate": 1.9354599603164508e-05,
      "loss": 0.1153944969177246,
      "step": 760,
      "token_acc": 0.9660055798156235
    },
    {
      "epoch": 0.48427672955974843,
      "grad_norm": 1.607561464821473,
      "learning_rate": 1.932986950232306e-05,
      "loss": 0.10485104322433472,
      "step": 770,
      "token_acc": 0.9696273095600701
    },
    {
      "epoch": 0.49056603773584906,
      "grad_norm": 2.0248199444328874,
      "learning_rate": 1.9304690877725506e-05,
      "loss": 0.10481438636779786,
      "step": 780,
      "token_acc": 0.9688417381194594
    },
    {
      "epoch": 0.4968553459119497,
      "grad_norm": 2.726276175870277,
      "learning_rate": 1.9279064939807986e-05,
      "loss": 0.11137734651565552,
      "step": 790,
      "token_acc": 0.9666876178504086
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 1.657566011087072,
      "learning_rate": 1.925299292051077e-05,
      "loss": 0.10481610298156738,
      "step": 800,
      "token_acc": 0.9691483936766956
    },
    {
      "epoch": 0.5094339622641509,
      "grad_norm": 1.8815105869017927,
      "learning_rate": 1.9226476073219026e-05,
      "loss": 0.10296165943145752,
      "step": 810,
      "token_acc": 0.9680598805816645
    },
    {
      "epoch": 0.5157232704402516,
      "grad_norm": 1.6804255612094785,
      "learning_rate": 1.9199515672702558e-05,
      "loss": 0.1030133843421936,
      "step": 820,
      "token_acc": 0.9712198685171658
    },
    {
      "epoch": 0.5220125786163522,
      "grad_norm": 1.5923137757409032,
      "learning_rate": 1.917211301505453e-05,
      "loss": 0.10893633365631103,
      "step": 830,
      "token_acc": 0.9680603782248113
    },
    {
      "epoch": 0.5283018867924528,
      "grad_norm": 1.8175548963275856,
      "learning_rate": 1.9144269417629165e-05,
      "loss": 0.11561510562896729,
      "step": 840,
      "token_acc": 0.9666006145716666
    },
    {
      "epoch": 0.5345911949685535,
      "grad_norm": 1.5087489763513091,
      "learning_rate": 1.9115986218978396e-05,
      "loss": 0.10352299213409424,
      "step": 850,
      "token_acc": 0.9696626669685307
    },
    {
      "epoch": 0.5408805031446541,
      "grad_norm": 1.5188944733060936,
      "learning_rate": 1.9087264778787534e-05,
      "loss": 0.10984259843826294,
      "step": 860,
      "token_acc": 0.9671633237822349
    },
    {
      "epoch": 0.5471698113207547,
      "grad_norm": 1.5446089445746696,
      "learning_rate": 1.9058106477809892e-05,
      "loss": 0.10156491994857789,
      "step": 870,
      "token_acc": 0.9691671877973844
    },
    {
      "epoch": 0.5534591194968553,
      "grad_norm": 1.4274372010097796,
      "learning_rate": 1.9028512717800414e-05,
      "loss": 0.10093884468078614,
      "step": 880,
      "token_acc": 0.9690456722345521
    },
    {
      "epoch": 0.559748427672956,
      "grad_norm": 2.015436306862073,
      "learning_rate": 1.8998484921448275e-05,
      "loss": 0.10100080966949462,
      "step": 890,
      "token_acc": 0.9691417278373801
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 1.7641447202728093,
      "learning_rate": 1.8968024532308505e-05,
      "loss": 0.10313496589660645,
      "step": 900,
      "token_acc": 0.9694502727654217
    },
    {
      "epoch": 0.5723270440251572,
      "grad_norm": 2.7256435866673385,
      "learning_rate": 1.893713301473258e-05,
      "loss": 0.09976035952568055,
      "step": 910,
      "token_acc": 0.9701510251312939
    },
    {
      "epoch": 0.5786163522012578,
      "grad_norm": 2.7322740850122265,
      "learning_rate": 1.8905811853798014e-05,
      "loss": 0.1044041395187378,
      "step": 920,
      "token_acc": 0.968413186560515
    },
    {
      "epoch": 0.5849056603773585,
      "grad_norm": 1.858448409286501,
      "learning_rate": 1.8874062555236994e-05,
      "loss": 0.11018127202987671,
      "step": 930,
      "token_acc": 0.9671611053905881
    },
    {
      "epoch": 0.5911949685534591,
      "grad_norm": 12.802923692002347,
      "learning_rate": 1.884188664536397e-05,
      "loss": 0.10684647560119628,
      "step": 940,
      "token_acc": 0.9699350180505415
    },
    {
      "epoch": 0.5974842767295597,
      "grad_norm": 1.7459169896484135,
      "learning_rate": 1.8809285671002288e-05,
      "loss": 0.1010131597518921,
      "step": 950,
      "token_acc": 0.9700858356859726
    },
    {
      "epoch": 0.6037735849056604,
      "grad_norm": 1.7594807067351037,
      "learning_rate": 1.877626119940982e-05,
      "loss": 0.10133538246154786,
      "step": 960,
      "token_acc": 0.969674964417463
    },
    {
      "epoch": 0.610062893081761,
      "grad_norm": 2.4724881080506638,
      "learning_rate": 1.874281481820364e-05,
      "loss": 0.10011661052703857,
      "step": 970,
      "token_acc": 0.9699815625720212
    },
    {
      "epoch": 0.6163522012578616,
      "grad_norm": 1.8152867273949589,
      "learning_rate": 1.8708948135283672e-05,
      "loss": 0.09960671663284301,
      "step": 980,
      "token_acc": 0.9696417901954648
    },
    {
      "epoch": 0.6226415094339622,
      "grad_norm": 1.7086563311997371,
      "learning_rate": 1.8674662778755416e-05,
      "loss": 0.09888969659805298,
      "step": 990,
      "token_acc": 0.9705755084379056
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 1.8413390867049504,
      "learning_rate": 1.8639960396851665e-05,
      "loss": 0.10456682443618774,
      "step": 1000,
      "token_acc": 0.9697974961056943
    },
    {
      "epoch": 0.6289308176100629,
      "eval_loss": 0.09391385316848755,
      "eval_runtime": 721.3125,
      "eval_samples_per_second": 11.749,
      "eval_steps_per_second": 2.938,
      "eval_token_acc": 0.9717736562791468,
      "step": 1000
    },
    {
      "epoch": 0.6352201257861635,
      "grad_norm": 1.4004622630825614,
      "learning_rate": 1.8604842657853282e-05,
      "loss": 0.09269505739212036,
      "step": 1010,
      "token_acc": 0.9738395415472779
    },
    {
      "epoch": 0.6415094339622641,
      "grad_norm": 1.6048419525402355,
      "learning_rate": 1.8569311250008976e-05,
      "loss": 0.09986767768859864,
      "step": 1020,
      "token_acc": 0.9705582573749963
    },
    {
      "epoch": 0.6477987421383647,
      "grad_norm": 1.551914657474502,
      "learning_rate": 1.853336788145416e-05,
      "loss": 0.08478968739509582,
      "step": 1030,
      "token_acc": 0.9746740725818056
    },
    {
      "epoch": 0.6540880503144654,
      "grad_norm": 1.4609089183474224,
      "learning_rate": 1.8497014280128836e-05,
      "loss": 0.0968370258808136,
      "step": 1040,
      "token_acc": 0.9714946296350493
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 1.6209543116867677,
      "learning_rate": 1.846025219369451e-05,
      "loss": 0.10038182735443116,
      "step": 1050,
      "token_acc": 0.971301524149431
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 1.6745782525801582,
      "learning_rate": 1.8423083389450184e-05,
      "loss": 0.09864873886108398,
      "step": 1060,
      "token_acc": 0.9706830841395485
    },
    {
      "epoch": 0.6729559748427673,
      "grad_norm": 1.91824919274026,
      "learning_rate": 1.8385509654247405e-05,
      "loss": 0.08868163824081421,
      "step": 1070,
      "token_acc": 0.9730788746687715
    },
    {
      "epoch": 0.6792452830188679,
      "grad_norm": 1.5103142849955444,
      "learning_rate": 1.834753279440434e-05,
      "loss": 0.08533757328987121,
      "step": 1080,
      "token_acc": 0.9747726440859307
    },
    {
      "epoch": 0.6855345911949685,
      "grad_norm": 1.6371011754605465,
      "learning_rate": 1.8309154635618967e-05,
      "loss": 0.09762944579124451,
      "step": 1090,
      "token_acc": 0.9705254515599343
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 1.5207551294053872,
      "learning_rate": 1.827037702288128e-05,
      "loss": 0.08945590853691102,
      "step": 1100,
      "token_acc": 0.973365268078721
    },
    {
      "epoch": 0.6981132075471698,
      "grad_norm": 1.2701931582388237,
      "learning_rate": 1.8231201820384615e-05,
      "loss": 0.08889811635017394,
      "step": 1110,
      "token_acc": 0.9730332580285396
    },
    {
      "epoch": 0.7044025157232704,
      "grad_norm": 1.539922447920576,
      "learning_rate": 1.819163091143602e-05,
      "loss": 0.08988333940505981,
      "step": 1120,
      "token_acc": 0.9729425695677916
    },
    {
      "epoch": 0.710691823899371,
      "grad_norm": 1.5525273016055445,
      "learning_rate": 1.815166619836571e-05,
      "loss": 0.08635894060134888,
      "step": 1130,
      "token_acc": 0.9734208416238529
    },
    {
      "epoch": 0.7169811320754716,
      "grad_norm": 1.6311000903415054,
      "learning_rate": 1.8111309602435624e-05,
      "loss": 0.0973101019859314,
      "step": 1140,
      "token_acc": 0.9719618030360201
    },
    {
      "epoch": 0.7232704402515723,
      "grad_norm": 1.2925549737286608,
      "learning_rate": 1.807056306374708e-05,
      "loss": 0.09141036868095398,
      "step": 1150,
      "token_acc": 0.9733011774719819
    },
    {
      "epoch": 0.7295597484276729,
      "grad_norm": 1.3801237808827327,
      "learning_rate": 1.802942854114747e-05,
      "loss": 0.10020852088928223,
      "step": 1160,
      "token_acc": 0.9716870401648965
    },
    {
      "epoch": 0.7358490566037735,
      "grad_norm": 1.6226969361595323,
      "learning_rate": 1.798790801213611e-05,
      "loss": 0.095181143283844,
      "step": 1170,
      "token_acc": 0.971852404453116
    },
    {
      "epoch": 0.7421383647798742,
      "grad_norm": 1.63651004169666,
      "learning_rate": 1.7946003472769175e-05,
      "loss": 0.09436390399932862,
      "step": 1180,
      "token_acc": 0.97247915302957
    },
    {
      "epoch": 0.7484276729559748,
      "grad_norm": 1.545200648396311,
      "learning_rate": 1.7903716937563737e-05,
      "loss": 0.08995594978332519,
      "step": 1190,
      "token_acc": 0.9729257883943911
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 1.4510430382097352,
      "learning_rate": 1.786105043940092e-05,
      "loss": 0.09402973651885986,
      "step": 1200,
      "token_acc": 0.9736291690220463
    },
    {
      "epoch": 0.7610062893081762,
      "grad_norm": 1.542152182505383,
      "learning_rate": 1.7818006029428175e-05,
      "loss": 0.08624240159988403,
      "step": 1210,
      "token_acc": 0.9744192715801548
    },
    {
      "epoch": 0.7672955974842768,
      "grad_norm": 1.5153951462525146,
      "learning_rate": 1.7774585776960663e-05,
      "loss": 0.09034521579742431,
      "step": 1220,
      "token_acc": 0.9742687735822293
    },
    {
      "epoch": 0.7735849056603774,
      "grad_norm": 1.584286377289857,
      "learning_rate": 1.7730791769381786e-05,
      "loss": 0.1029138445854187,
      "step": 1230,
      "token_acc": 0.9693110282117043
    },
    {
      "epoch": 0.779874213836478,
      "grad_norm": 1.1675776428936107,
      "learning_rate": 1.7686626112042827e-05,
      "loss": 0.0900284469127655,
      "step": 1240,
      "token_acc": 0.9731250174060768
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 1.7145950918174153,
      "learning_rate": 1.7642090928161754e-05,
      "loss": 0.0883266270160675,
      "step": 1250,
      "token_acc": 0.974335142266063
    },
    {
      "epoch": 0.7924528301886793,
      "grad_norm": 1.5572882087279911,
      "learning_rate": 1.7597188358721135e-05,
      "loss": 0.08593888878822327,
      "step": 1260,
      "token_acc": 0.974071787635594
    },
    {
      "epoch": 0.7987421383647799,
      "grad_norm": 1.587797606510951,
      "learning_rate": 1.7551920562365213e-05,
      "loss": 0.08575895428657532,
      "step": 1270,
      "token_acc": 0.9759103973451065
    },
    {
      "epoch": 0.8050314465408805,
      "grad_norm": 1.3799851052052912,
      "learning_rate": 1.7506289715296134e-05,
      "loss": 0.0863305687904358,
      "step": 1280,
      "token_acc": 0.9738367941504099
    },
    {
      "epoch": 0.8113207547169812,
      "grad_norm": 1.859574713074995,
      "learning_rate": 1.7460298011169333e-05,
      "loss": 0.08664470314979553,
      "step": 1290,
      "token_acc": 0.9729901581333629
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 1.3443565152248043,
      "learning_rate": 1.7413947660988067e-05,
      "loss": 0.08178459405899048,
      "step": 1300,
      "token_acc": 0.9748381649310441
    },
    {
      "epoch": 0.8238993710691824,
      "grad_norm": 1.5494391575974118,
      "learning_rate": 1.7367240892997137e-05,
      "loss": 0.08511111736297608,
      "step": 1310,
      "token_acc": 0.9751603261658106
    },
    {
      "epoch": 0.8301886792452831,
      "grad_norm": 1.7615625697478026,
      "learning_rate": 1.732017995257575e-05,
      "loss": 0.08447937965393067,
      "step": 1320,
      "token_acc": 0.9743189557321226
    },
    {
      "epoch": 0.8364779874213837,
      "grad_norm": 1.4089633030126132,
      "learning_rate": 1.7272767102129586e-05,
      "loss": 0.0828474998474121,
      "step": 1330,
      "token_acc": 0.9742417015067762
    },
    {
      "epoch": 0.8427672955974843,
      "grad_norm": 1.6592961075137602,
      "learning_rate": 1.7225004620982035e-05,
      "loss": 0.08787728548049926,
      "step": 1340,
      "token_acc": 0.973914992575159
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 1.5117166617272377,
      "learning_rate": 1.7176894805264616e-05,
      "loss": 0.0850176215171814,
      "step": 1350,
      "token_acc": 0.9755728811549521
    },
    {
      "epoch": 0.8553459119496856,
      "grad_norm": 1.5748027114678638,
      "learning_rate": 1.7128439967806598e-05,
      "loss": 0.0911492645740509,
      "step": 1360,
      "token_acc": 0.9742503907915305
    },
    {
      "epoch": 0.8616352201257862,
      "grad_norm": 1.4473190275546743,
      "learning_rate": 1.707964243802381e-05,
      "loss": 0.08639097213745117,
      "step": 1370,
      "token_acc": 0.9745522819179665
    },
    {
      "epoch": 0.8679245283018868,
      "grad_norm": 1.379465232062877,
      "learning_rate": 1.7030504561806657e-05,
      "loss": 0.087519371509552,
      "step": 1380,
      "token_acc": 0.973626242781449
    },
    {
      "epoch": 0.8742138364779874,
      "grad_norm": 1.3921602039818965,
      "learning_rate": 1.6981028701407342e-05,
      "loss": 0.07906639575958252,
      "step": 1390,
      "token_acc": 0.9765992799778455
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 1.3231345110824035,
      "learning_rate": 1.6931217235326313e-05,
      "loss": 0.08956987261772156,
      "step": 1400,
      "token_acc": 0.9740674539221145
    },
    {
      "epoch": 0.8867924528301887,
      "grad_norm": 1.2608029267930918,
      "learning_rate": 1.6881072558197904e-05,
      "loss": 0.0810630202293396,
      "step": 1410,
      "token_acc": 0.9758955958099427
    },
    {
      "epoch": 0.8930817610062893,
      "grad_norm": 1.4322673455962067,
      "learning_rate": 1.6830597080675222e-05,
      "loss": 0.0859791398048401,
      "step": 1420,
      "token_acc": 0.9743442575905412
    },
    {
      "epoch": 0.89937106918239,
      "grad_norm": 1.8746897805081975,
      "learning_rate": 1.6779793229314268e-05,
      "loss": 0.08041177988052368,
      "step": 1430,
      "token_acc": 0.9760162243188432
    },
    {
      "epoch": 0.9056603773584906,
      "grad_norm": 1.3274725836940766,
      "learning_rate": 1.672866344645726e-05,
      "loss": 0.0832511305809021,
      "step": 1440,
      "token_acc": 0.9765505920975496
    },
    {
      "epoch": 0.9119496855345912,
      "grad_norm": 1.6426325467290352,
      "learning_rate": 1.667721019011524e-05,
      "loss": 0.07603156566619873,
      "step": 1450,
      "token_acc": 0.9772056082805287
    },
    {
      "epoch": 0.9182389937106918,
      "grad_norm": 1.4993824513571612,
      "learning_rate": 1.6625435933849912e-05,
      "loss": 0.07640135288238525,
      "step": 1460,
      "token_acc": 0.9774820023438808
    },
    {
      "epoch": 0.9245283018867925,
      "grad_norm": 1.7011352962645643,
      "learning_rate": 1.657334316665469e-05,
      "loss": 0.08306714296340942,
      "step": 1470,
      "token_acc": 0.9748449545006665
    },
    {
      "epoch": 0.9308176100628931,
      "grad_norm": 1.4785906863778826,
      "learning_rate": 1.6520934392835096e-05,
      "loss": 0.08007491827011108,
      "step": 1480,
      "token_acc": 0.9753671523023814
    },
    {
      "epoch": 0.9371069182389937,
      "grad_norm": 1.215720614479762,
      "learning_rate": 1.6468212131888324e-05,
      "loss": 0.08180446624755859,
      "step": 1490,
      "token_acc": 0.9762308998302207
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 1.394799734230705,
      "learning_rate": 1.6415178918382146e-05,
      "loss": 0.08114392161369324,
      "step": 1500,
      "token_acc": 0.9746889624472093
    },
    {
      "epoch": 0.9433962264150944,
      "eval_loss": 0.08027087152004242,
      "eval_runtime": 722.6841,
      "eval_samples_per_second": 11.727,
      "eval_steps_per_second": 2.932,
      "eval_token_acc": 0.9762356642152163,
      "step": 1500
    },
    {
      "epoch": 0.949685534591195,
      "grad_norm": 1.3907727949176176,
      "learning_rate": 1.6361837301833043e-05,
      "loss": 0.0873852789402008,
      "step": 1510,
      "token_acc": 0.9735111006695641
    },
    {
      "epoch": 0.9559748427672956,
      "grad_norm": 1.6463543824854343,
      "learning_rate": 1.6308189846583658e-05,
      "loss": 0.07256322503089904,
      "step": 1520,
      "token_acc": 0.9783221138163163
    },
    {
      "epoch": 0.9622641509433962,
      "grad_norm": 1.5507672072885261,
      "learning_rate": 1.6254239131679516e-05,
      "loss": 0.09013139009475708,
      "step": 1530,
      "token_acc": 0.9730759166109383
    },
    {
      "epoch": 0.9685534591194969,
      "grad_norm": 1.5083161719535216,
      "learning_rate": 1.6199987750745022e-05,
      "loss": 0.08049647212028503,
      "step": 1540,
      "token_acc": 0.9760573226144704
    },
    {
      "epoch": 0.9748427672955975,
      "grad_norm": 1.9355118295883647,
      "learning_rate": 1.61454383118588e-05,
      "loss": 0.08110915422439575,
      "step": 1550,
      "token_acc": 0.9759386260606765
    },
    {
      "epoch": 0.9811320754716981,
      "grad_norm": 1.4453680164285594,
      "learning_rate": 1.6090593437428284e-05,
      "loss": 0.08374351263046265,
      "step": 1560,
      "token_acc": 0.9741173766912757
    },
    {
      "epoch": 0.9874213836477987,
      "grad_norm": 1.8392039122485897,
      "learning_rate": 1.603545576406368e-05,
      "loss": 0.07776378393173218,
      "step": 1570,
      "token_acc": 0.9767461939160216
    },
    {
      "epoch": 0.9937106918238994,
      "grad_norm": 1.8713989393813286,
      "learning_rate": 1.5980027942451185e-05,
      "loss": 0.07840796709060668,
      "step": 1580,
      "token_acc": 0.9758594831564374
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.7874037128786733,
      "learning_rate": 1.5924312637225582e-05,
      "loss": 0.08290352821350097,
      "step": 1590,
      "token_acc": 0.9754263655220097
    },
    {
      "epoch": 1.0062893081761006,
      "grad_norm": 1.0863843323954543,
      "learning_rate": 1.586831252684212e-05,
      "loss": 0.054670459032058714,
      "step": 1600,
      "token_acc": 0.9831636765276576
    },
    {
      "epoch": 1.0125786163522013,
      "grad_norm": 1.2355860009095903,
      "learning_rate": 1.581203030344776e-05,
      "loss": 0.06227320432662964,
      "step": 1610,
      "token_acc": 0.9807899784224406
    },
    {
      "epoch": 1.0188679245283019,
      "grad_norm": 1.0627219529639433,
      "learning_rate": 1.5755468672751762e-05,
      "loss": 0.06504143476486206,
      "step": 1620,
      "token_acc": 0.9811943349394072
    },
    {
      "epoch": 1.0251572327044025,
      "grad_norm": 1.2030481551752021,
      "learning_rate": 1.5698630353895588e-05,
      "loss": 0.06021591424942017,
      "step": 1630,
      "token_acc": 0.9817465792802116
    },
    {
      "epoch": 1.0314465408805031,
      "grad_norm": 0.9952274038110551,
      "learning_rate": 1.5641518079322198e-05,
      "loss": 0.059084004163742064,
      "step": 1640,
      "token_acc": 0.981861242685985
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 1.1029301430587275,
      "learning_rate": 1.55841345946447e-05,
      "loss": 0.062122118473052976,
      "step": 1650,
      "token_acc": 0.9809982194524817
    },
    {
      "epoch": 1.0440251572327044,
      "grad_norm": 1.0631570096219156,
      "learning_rate": 1.5526482658514325e-05,
      "loss": 0.06019622087478638,
      "step": 1660,
      "token_acc": 0.9813258905449732
    },
    {
      "epoch": 1.050314465408805,
      "grad_norm": 1.2452397083425584,
      "learning_rate": 1.5468565042487844e-05,
      "loss": 0.0565668523311615,
      "step": 1670,
      "token_acc": 0.98244754438552
    },
    {
      "epoch": 1.0566037735849056,
      "grad_norm": 1.2893720007251595,
      "learning_rate": 1.541038453089431e-05,
      "loss": 0.05943567156791687,
      "step": 1680,
      "token_acc": 0.9819022607901348
    },
    {
      "epoch": 1.0628930817610063,
      "grad_norm": 1.4405565971353904,
      "learning_rate": 1.5351943920701194e-05,
      "loss": 0.06467938423156738,
      "step": 1690,
      "token_acc": 0.9807256911344138
    },
    {
      "epoch": 1.069182389937107,
      "grad_norm": 1.6088355986810434,
      "learning_rate": 1.5293246021379952e-05,
      "loss": 0.05654835104942322,
      "step": 1700,
      "token_acc": 0.982938223602124
    },
    {
      "epoch": 1.0754716981132075,
      "grad_norm": 1.429067962964369,
      "learning_rate": 1.5234293654770942e-05,
      "loss": 0.05899702310562134,
      "step": 1710,
      "token_acc": 0.9819976934544739
    },
    {
      "epoch": 1.0817610062893082,
      "grad_norm": 1.4383654244646944,
      "learning_rate": 1.5175089654947765e-05,
      "loss": 0.056074118614196776,
      "step": 1720,
      "token_acc": 0.9819119189094722
    },
    {
      "epoch": 1.0880503144654088,
      "grad_norm": 1.224480893565136,
      "learning_rate": 1.5115636868081031e-05,
      "loss": 0.060695433616638185,
      "step": 1730,
      "token_acc": 0.9811226333787428
    },
    {
      "epoch": 1.0943396226415094,
      "grad_norm": 1.2853685041540284,
      "learning_rate": 1.5055938152301533e-05,
      "loss": 0.0551666259765625,
      "step": 1740,
      "token_acc": 0.9832683138274816
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 1.4811194652333022,
      "learning_rate": 1.4995996377562831e-05,
      "loss": 0.06323553323745727,
      "step": 1750,
      "token_acc": 0.9809037193503268
    },
    {
      "epoch": 1.1069182389937107,
      "grad_norm": 1.4583514142720422,
      "learning_rate": 1.4935814425503301e-05,
      "loss": 0.0589030921459198,
      "step": 1760,
      "token_acc": 0.9828337176530874
    },
    {
      "epoch": 1.1132075471698113,
      "grad_norm": 1.1852990258851064,
      "learning_rate": 1.4875395189307584e-05,
      "loss": 0.05278688669204712,
      "step": 1770,
      "token_acc": 0.9831741052331384
    },
    {
      "epoch": 1.119496855345912,
      "grad_norm": 1.569277335890507,
      "learning_rate": 1.4814741573567514e-05,
      "loss": 0.053292316198348996,
      "step": 1780,
      "token_acc": 0.9836212116843046
    },
    {
      "epoch": 1.1257861635220126,
      "grad_norm": 1.1753581015135361,
      "learning_rate": 1.475385649414248e-05,
      "loss": 0.059940147399902347,
      "step": 1790,
      "token_acc": 0.9815808802757218
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 1.3654141810969276,
      "learning_rate": 1.4692742878019238e-05,
      "loss": 0.062000393867492676,
      "step": 1800,
      "token_acc": 0.9807436329754193
    },
    {
      "epoch": 1.1383647798742138,
      "grad_norm": 1.1668245354459672,
      "learning_rate": 1.4631403663171215e-05,
      "loss": 0.06278970837593079,
      "step": 1810,
      "token_acc": 0.9808653680399931
    },
    {
      "epoch": 1.1446540880503144,
      "grad_norm": 1.108737675654481,
      "learning_rate": 1.4569841798417257e-05,
      "loss": 0.052794039249420166,
      "step": 1820,
      "token_acc": 0.9834860773577724
    },
    {
      "epoch": 1.150943396226415,
      "grad_norm": 1.3645629847734626,
      "learning_rate": 1.4508060243279878e-05,
      "loss": 0.054091382026672366,
      "step": 1830,
      "token_acc": 0.9830678432271373
    },
    {
      "epoch": 1.1572327044025157,
      "grad_norm": 1.075485335510402,
      "learning_rate": 1.4446061967842973e-05,
      "loss": 0.05640895962715149,
      "step": 1840,
      "token_acc": 0.9829748578796226
    },
    {
      "epoch": 1.1635220125786163,
      "grad_norm": 1.1673617984452092,
      "learning_rate": 1.4383849952609041e-05,
      "loss": 0.05560168027877808,
      "step": 1850,
      "token_acc": 0.9829094373394152
    },
    {
      "epoch": 1.169811320754717,
      "grad_norm": 1.178620225968234,
      "learning_rate": 1.4321427188355901e-05,
      "loss": 0.06277099251747131,
      "step": 1860,
      "token_acc": 0.9805737056462246
    },
    {
      "epoch": 1.1761006289308176,
      "grad_norm": 4.438647349024938,
      "learning_rate": 1.425879667599291e-05,
      "loss": 0.05321987271308899,
      "step": 1870,
      "token_acc": 0.9836384771868643
    },
    {
      "epoch": 1.1823899371069182,
      "grad_norm": 0.8089960660126764,
      "learning_rate": 1.4195961426416695e-05,
      "loss": 0.05596756935119629,
      "step": 1880,
      "token_acc": 0.982458692110775
    },
    {
      "epoch": 1.1886792452830188,
      "grad_norm": 1.098178154828662,
      "learning_rate": 1.4132924460366422e-05,
      "loss": 0.056302571296691896,
      "step": 1890,
      "token_acc": 0.9826823556622047
    },
    {
      "epoch": 1.1949685534591195,
      "grad_norm": 1.6098174552103235,
      "learning_rate": 1.4069688808278555e-05,
      "loss": 0.05291517972946167,
      "step": 1900,
      "token_acc": 0.9840382360898188
    },
    {
      "epoch": 1.20125786163522,
      "grad_norm": 1.1838009862882108,
      "learning_rate": 1.4006257510141185e-05,
      "loss": 0.05891250967979431,
      "step": 1910,
      "token_acc": 0.9826159865846842
    },
    {
      "epoch": 1.2075471698113207,
      "grad_norm": 1.0812253602188817,
      "learning_rate": 1.3942633615347894e-05,
      "loss": 0.05368814468383789,
      "step": 1920,
      "token_acc": 0.9829718640093786
    },
    {
      "epoch": 1.2138364779874213,
      "grad_norm": 1.0195075509939109,
      "learning_rate": 1.3878820182551128e-05,
      "loss": 0.058639413118362425,
      "step": 1930,
      "token_acc": 0.9828350271803357
    },
    {
      "epoch": 1.220125786163522,
      "grad_norm": 1.3170894091360277,
      "learning_rate": 1.3814820279515187e-05,
      "loss": 0.05739471912384033,
      "step": 1940,
      "token_acc": 0.982506542268745
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 1.0061802280737047,
      "learning_rate": 1.3750636982968737e-05,
      "loss": 0.052751028537750246,
      "step": 1950,
      "token_acc": 0.9834080717488789
    },
    {
      "epoch": 1.2327044025157232,
      "grad_norm": 1.1823477189308416,
      "learning_rate": 1.3686273378456887e-05,
      "loss": 0.05880435705184937,
      "step": 1960,
      "token_acc": 0.9824817101051669
    },
    {
      "epoch": 1.2389937106918238,
      "grad_norm": 1.22554558640998,
      "learning_rate": 1.3621732560192871e-05,
      "loss": 0.05745119452476501,
      "step": 1970,
      "token_acc": 0.9821925431274346
    },
    {
      "epoch": 1.2452830188679245,
      "grad_norm": 1.3279829643490002,
      "learning_rate": 1.3557017630909281e-05,
      "loss": 0.05679129958152771,
      "step": 1980,
      "token_acc": 0.9827537461125248
    },
    {
      "epoch": 1.251572327044025,
      "grad_norm": 1.307619119343479,
      "learning_rate": 1.3492131701708923e-05,
      "loss": 0.05341850519180298,
      "step": 1990,
      "token_acc": 0.9835128725319313
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 0.9629697375654898,
      "learning_rate": 1.342707789191524e-05,
      "loss": 0.0589292049407959,
      "step": 2000,
      "token_acc": 0.982400046088259
    },
    {
      "epoch": 1.2578616352201257,
      "eval_loss": 0.07104787975549698,
      "eval_runtime": 730.4468,
      "eval_samples_per_second": 11.602,
      "eval_steps_per_second": 2.901,
      "eval_token_acc": 0.978991514710327,
      "step": 2000
    },
    {
      "epoch": 1.2641509433962264,
      "grad_norm": 1.1453083559028814,
      "learning_rate": 1.3361859328922368e-05,
      "loss": 0.05606248378753662,
      "step": 2010,
      "token_acc": 0.9830185358986392
    },
    {
      "epoch": 1.270440251572327,
      "grad_norm": 1.204837950589622,
      "learning_rate": 1.3296479148044772e-05,
      "loss": 0.05325353145599365,
      "step": 2020,
      "token_acc": 0.9833405389825624
    },
    {
      "epoch": 1.2767295597484276,
      "grad_norm": 2.6011333489424926,
      "learning_rate": 1.323094049236653e-05,
      "loss": 0.05622377395629883,
      "step": 2030,
      "token_acc": 0.9828592371232838
    },
    {
      "epoch": 1.2830188679245282,
      "grad_norm": 1.232823511568605,
      "learning_rate": 1.3165246512590234e-05,
      "loss": 0.05152193307876587,
      "step": 2040,
      "token_acc": 0.9835936818221486
    },
    {
      "epoch": 1.2893081761006289,
      "grad_norm": 1.1466821617032141,
      "learning_rate": 1.3099400366885525e-05,
      "loss": 0.05622357130050659,
      "step": 2050,
      "token_acc": 0.9826379815927765
    },
    {
      "epoch": 1.2955974842767295,
      "grad_norm": 0.8472290173088248,
      "learning_rate": 1.3033405220737247e-05,
      "loss": 0.05588294267654419,
      "step": 2060,
      "token_acc": 0.982960565881039
    },
    {
      "epoch": 1.3018867924528301,
      "grad_norm": 0.8561022792687387,
      "learning_rate": 1.29672642467933e-05,
      "loss": 0.05827444791793823,
      "step": 2070,
      "token_acc": 0.9819563431522121
    },
    {
      "epoch": 1.3081761006289307,
      "grad_norm": 0.8843280358103902,
      "learning_rate": 1.2900980624712102e-05,
      "loss": 0.053099215030670166,
      "step": 2080,
      "token_acc": 0.9832570120206068
    },
    {
      "epoch": 1.3144654088050314,
      "grad_norm": 0.9595233250713293,
      "learning_rate": 1.283455754100972e-05,
      "loss": 0.054636824131011966,
      "step": 2090,
      "token_acc": 0.9834532583680484
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 1.0568180647703327,
      "learning_rate": 1.2767998188906707e-05,
      "loss": 0.0547333836555481,
      "step": 2100,
      "token_acc": 0.9825695735337697
    },
    {
      "epoch": 1.3270440251572326,
      "grad_norm": 1.0358582290569485,
      "learning_rate": 1.2701305768174568e-05,
      "loss": 0.04895033836364746,
      "step": 2110,
      "token_acc": 0.9852507374631269
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 1.165756196741812,
      "learning_rate": 1.2634483484981947e-05,
      "loss": 0.05529743432998657,
      "step": 2120,
      "token_acc": 0.9829623215931625
    },
    {
      "epoch": 1.3396226415094339,
      "grad_norm": 0.9398124420553513,
      "learning_rate": 1.2567534551740495e-05,
      "loss": 0.05000982284545898,
      "step": 2130,
      "token_acc": 0.9845889904457656
    },
    {
      "epoch": 1.3459119496855345,
      "grad_norm": 1.04779088754709,
      "learning_rate": 1.250046218695042e-05,
      "loss": 0.05204824805259704,
      "step": 2140,
      "token_acc": 0.9837851750171586
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 1.1911828992184066,
      "learning_rate": 1.2433269615045787e-05,
      "loss": 0.059820497035980226,
      "step": 2150,
      "token_acc": 0.9811458424539591
    },
    {
      "epoch": 1.3584905660377358,
      "grad_norm": 1.1318809840091484,
      "learning_rate": 1.2365960066239481e-05,
      "loss": 0.04887381792068481,
      "step": 2160,
      "token_acc": 0.9850779142117725
    },
    {
      "epoch": 1.3647798742138364,
      "grad_norm": 1.18575128110939,
      "learning_rate": 1.2298536776367939e-05,
      "loss": 0.05481389760971069,
      "step": 2170,
      "token_acc": 0.9830817516066358
    },
    {
      "epoch": 1.371069182389937,
      "grad_norm": 1.4705395863488706,
      "learning_rate": 1.2231002986735581e-05,
      "loss": 0.053452318906784056,
      "step": 2180,
      "token_acc": 0.9840079667271981
    },
    {
      "epoch": 1.3773584905660377,
      "grad_norm": 1.1381781742512012,
      "learning_rate": 1.2163361943958978e-05,
      "loss": 0.05643002986907959,
      "step": 2190,
      "token_acc": 0.9830008052250157
    },
    {
      "epoch": 1.3836477987421385,
      "grad_norm": 1.108236107230958,
      "learning_rate": 1.2095616899810791e-05,
      "loss": 0.05098185539245605,
      "step": 2200,
      "token_acc": 0.983818037795632
    },
    {
      "epoch": 1.389937106918239,
      "grad_norm": 1.3111447759496122,
      "learning_rate": 1.202777111106344e-05,
      "loss": 0.05649634599685669,
      "step": 2210,
      "token_acc": 0.9826166514181153
    },
    {
      "epoch": 1.3962264150943398,
      "grad_norm": 1.199633245486944,
      "learning_rate": 1.1959827839332534e-05,
      "loss": 0.05167834758758545,
      "step": 2220,
      "token_acc": 0.9843418784093158
    },
    {
      "epoch": 1.4025157232704402,
      "grad_norm": 0.8666899643086994,
      "learning_rate": 1.1891790350920078e-05,
      "loss": 0.050674861669540404,
      "step": 2230,
      "token_acc": 0.9848721508735393
    },
    {
      "epoch": 1.408805031446541,
      "grad_norm": 1.0104295447062845,
      "learning_rate": 1.1823661916657441e-05,
      "loss": 0.05399259328842163,
      "step": 2240,
      "token_acc": 0.9835477202251063
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 1.0519494154431606,
      "learning_rate": 1.175544581174812e-05,
      "loss": 0.05260412693023682,
      "step": 2250,
      "token_acc": 0.9835558595260043
    },
    {
      "epoch": 1.4213836477987423,
      "grad_norm": 1.068852503916311,
      "learning_rate": 1.168714531561029e-05,
      "loss": 0.05003952980041504,
      "step": 2260,
      "token_acc": 0.9841691670238606
    },
    {
      "epoch": 1.4276729559748427,
      "grad_norm": 1.3301770971060891,
      "learning_rate": 1.1618763711719146e-05,
      "loss": 0.05456411838531494,
      "step": 2270,
      "token_acc": 0.983095406360424
    },
    {
      "epoch": 1.4339622641509435,
      "grad_norm": 1.4290925190264645,
      "learning_rate": 1.155030428744905e-05,
      "loss": 0.05128795504570007,
      "step": 2280,
      "token_acc": 0.9836185750133928
    },
    {
      "epoch": 1.440251572327044,
      "grad_norm": 1.0237953855660737,
      "learning_rate": 1.1481770333915503e-05,
      "loss": 0.05018550753593445,
      "step": 2290,
      "token_acc": 0.9849615590413698
    },
    {
      "epoch": 1.4465408805031448,
      "grad_norm": 0.9772590029300261,
      "learning_rate": 1.1413165145816915e-05,
      "loss": 0.04862302839756012,
      "step": 2300,
      "token_acc": 0.9849491445705215
    },
    {
      "epoch": 1.4528301886792452,
      "grad_norm": 1.6101981573211892,
      "learning_rate": 1.1344492021276234e-05,
      "loss": 0.05587894320487976,
      "step": 2310,
      "token_acc": 0.982574528060135
    },
    {
      "epoch": 1.459119496855346,
      "grad_norm": 1.0729401899055404,
      "learning_rate": 1.1275754261682373e-05,
      "loss": 0.04833915829658508,
      "step": 2320,
      "token_acc": 0.9852789472216801
    },
    {
      "epoch": 1.4654088050314464,
      "grad_norm": 1.138669296120146,
      "learning_rate": 1.1206955171531509e-05,
      "loss": 0.05411947965621948,
      "step": 2330,
      "token_acc": 0.9835947825087629
    },
    {
      "epoch": 1.4716981132075473,
      "grad_norm": 0.9193723256228759,
      "learning_rate": 1.113809805826823e-05,
      "loss": 0.055597662925720215,
      "step": 2340,
      "token_acc": 0.9836160314387262
    },
    {
      "epoch": 1.4779874213836477,
      "grad_norm": 1.15943849177995,
      "learning_rate": 1.1069186232126512e-05,
      "loss": 0.05205290317535401,
      "step": 2350,
      "token_acc": 0.9842589513357721
    },
    {
      "epoch": 1.4842767295597485,
      "grad_norm": 1.0278093026309196,
      "learning_rate": 1.1000223005970603e-05,
      "loss": 0.047948446869850156,
      "step": 2360,
      "token_acc": 0.9846708246708247
    },
    {
      "epoch": 1.490566037735849,
      "grad_norm": 0.9090992104674461,
      "learning_rate": 1.0931211695135753e-05,
      "loss": 0.04863405227661133,
      "step": 2370,
      "token_acc": 0.9849871620601117
    },
    {
      "epoch": 1.4968553459119498,
      "grad_norm": 0.9630374367443844,
      "learning_rate": 1.086215561726883e-05,
      "loss": 0.051735520362854004,
      "step": 2380,
      "token_acc": 0.9836164352117526
    },
    {
      "epoch": 1.5031446540880502,
      "grad_norm": 0.9821758509138084,
      "learning_rate": 1.0793058092168833e-05,
      "loss": 0.05245034694671631,
      "step": 2390,
      "token_acc": 0.9844871432608081
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 1.961427641140477,
      "learning_rate": 1.07239224416273e-05,
      "loss": 0.052685320377349854,
      "step": 2400,
      "token_acc": 0.9839027595269383
    },
    {
      "epoch": 1.5157232704402515,
      "grad_norm": 1.023089649058638,
      "learning_rate": 1.0654751989268588e-05,
      "loss": 0.05271808505058288,
      "step": 2410,
      "token_acc": 0.984044630404463
    },
    {
      "epoch": 1.5220125786163523,
      "grad_norm": 1.1362717857926448,
      "learning_rate": 1.0585550060390141e-05,
      "loss": 0.045315057039260864,
      "step": 2420,
      "token_acc": 0.986439398339044
    },
    {
      "epoch": 1.5283018867924527,
      "grad_norm": 1.1540779189780639,
      "learning_rate": 1.0516319981802589e-05,
      "loss": 0.05070680379867554,
      "step": 2430,
      "token_acc": 0.9837349397590361
    },
    {
      "epoch": 1.5345911949685536,
      "grad_norm": 1.370378196938021,
      "learning_rate": 1.0447065081669834e-05,
      "loss": 0.05064605474472046,
      "step": 2440,
      "token_acc": 0.984635722679201
    },
    {
      "epoch": 1.540880503144654,
      "grad_norm": 1.0034426130585499,
      "learning_rate": 1.037778868934905e-05,
      "loss": 0.05282008647918701,
      "step": 2450,
      "token_acc": 0.9836287259753712
    },
    {
      "epoch": 1.5471698113207548,
      "grad_norm": 1.0534531097485125,
      "learning_rate": 1.0308494135230616e-05,
      "loss": 0.05066592693328857,
      "step": 2460,
      "token_acc": 0.9850426723758688
    },
    {
      "epoch": 1.5534591194968552,
      "grad_norm": 1.0764284563415414,
      "learning_rate": 1.023918475057803e-05,
      "loss": 0.046628701686859134,
      "step": 2470,
      "token_acc": 0.9858265424912689
    },
    {
      "epoch": 1.559748427672956,
      "grad_norm": 1.005389844740969,
      "learning_rate": 1.0169863867367755e-05,
      "loss": 0.0455363005399704,
      "step": 2480,
      "token_acc": 0.9856986332968111
    },
    {
      "epoch": 1.5660377358490565,
      "grad_norm": 1.5808354451192173,
      "learning_rate": 1.010053481812903e-05,
      "loss": 0.04960752725601196,
      "step": 2490,
      "token_acc": 0.9848381212950297
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.8146209429997286,
      "learning_rate": 1.003120093578366e-05,
      "loss": 0.049983108043670656,
      "step": 2500,
      "token_acc": 0.9847141288742505
    },
    {
      "epoch": 1.5723270440251573,
      "eval_loss": 0.06294354796409607,
      "eval_runtime": 720.1002,
      "eval_samples_per_second": 11.769,
      "eval_steps_per_second": 2.943,
      "eval_token_acc": 0.9814012589817003,
      "step": 2500
    },
    {
      "epoch": 1.5786163522012577,
      "grad_norm": 1.0779724413765672,
      "learning_rate": 9.961865553485808e-06,
      "loss": 0.05379664897918701,
      "step": 2510,
      "token_acc": 0.9826207228336039
    },
    {
      "epoch": 1.5849056603773586,
      "grad_norm": 1.250981230954497,
      "learning_rate": 9.892532004461746e-06,
      "loss": 0.0480253279209137,
      "step": 2520,
      "token_acc": 0.9850908663177086
    },
    {
      "epoch": 1.591194968553459,
      "grad_norm": 0.9282644695610834,
      "learning_rate": 9.8232036218496e-06,
      "loss": 0.04891929030418396,
      "step": 2530,
      "token_acc": 0.9845905700048965
    },
    {
      "epoch": 1.5974842767295598,
      "grad_norm": 0.8282640147021464,
      "learning_rate": 9.753883738539147e-06,
      "loss": 0.047617962956428526,
      "step": 2540,
      "token_acc": 0.9850088183421517
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 0.9724621116688568,
      "learning_rate": 9.684575687011546e-06,
      "loss": 0.05135564804077149,
      "step": 2550,
      "token_acc": 0.9850393473811652
    },
    {
      "epoch": 1.610062893081761,
      "grad_norm": 1.271084592116781,
      "learning_rate": 9.615282799179174e-06,
      "loss": 0.044367313385009766,
      "step": 2560,
      "token_acc": 0.9855293221629855
    },
    {
      "epoch": 1.6163522012578615,
      "grad_norm": 1.090033419311402,
      "learning_rate": 9.546008406225414e-06,
      "loss": 0.047433316707611084,
      "step": 2570,
      "token_acc": 0.9850305411373348
    },
    {
      "epoch": 1.6226415094339623,
      "grad_norm": 0.9385748626886379,
      "learning_rate": 9.476755838444542e-06,
      "loss": 0.04910559058189392,
      "step": 2580,
      "token_acc": 0.9851407424021532
    },
    {
      "epoch": 1.6289308176100628,
      "grad_norm": 0.8144990044132233,
      "learning_rate": 9.407528425081598e-06,
      "loss": 0.046424055099487306,
      "step": 2590,
      "token_acc": 0.9859498736907837
    },
    {
      "epoch": 1.6352201257861636,
      "grad_norm": 1.0093041374441951,
      "learning_rate": 9.338329494172362e-06,
      "loss": 0.044951969385147096,
      "step": 2600,
      "token_acc": 0.9868212816159756
    },
    {
      "epoch": 1.641509433962264,
      "grad_norm": 1.614006415230121,
      "learning_rate": 9.269162372383338e-06,
      "loss": 0.048131465911865234,
      "step": 2610,
      "token_acc": 0.9848029439029187
    },
    {
      "epoch": 1.6477987421383649,
      "grad_norm": 0.9563758266821378,
      "learning_rate": 9.200030384851851e-06,
      "loss": 0.05282438993453979,
      "step": 2620,
      "token_acc": 0.9838705072939564
    },
    {
      "epoch": 1.6540880503144653,
      "grad_norm": 0.7360361071435326,
      "learning_rate": 9.130936855026166e-06,
      "loss": 0.04689803123474121,
      "step": 2630,
      "token_acc": 0.985498807741363
    },
    {
      "epoch": 1.6603773584905661,
      "grad_norm": 1.1217754732714569,
      "learning_rate": 9.061885104505757e-06,
      "loss": 0.04997251033782959,
      "step": 2640,
      "token_acc": 0.9845242306331801
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 1.3727043728275967,
      "learning_rate": 8.992878452881584e-06,
      "loss": 0.05193988084793091,
      "step": 2650,
      "token_acc": 0.9841976602060416
    },
    {
      "epoch": 1.6729559748427674,
      "grad_norm": 1.0738630903011401,
      "learning_rate": 8.923920217576534e-06,
      "loss": 0.05103607177734375,
      "step": 2660,
      "token_acc": 0.9841919016127161
    },
    {
      "epoch": 1.6792452830188678,
      "grad_norm": 0.8980456588406888,
      "learning_rate": 8.855013713685915e-06,
      "loss": 0.04786647856235504,
      "step": 2670,
      "token_acc": 0.9857719560737374
    },
    {
      "epoch": 1.6855345911949686,
      "grad_norm": 1.6346416405736106,
      "learning_rate": 8.786162253818127e-06,
      "loss": 0.044231253862380984,
      "step": 2680,
      "token_acc": 0.9870336289108248
    },
    {
      "epoch": 1.691823899371069,
      "grad_norm": 0.6531463448084285,
      "learning_rate": 8.71736914793536e-06,
      "loss": 0.043486350774765016,
      "step": 2690,
      "token_acc": 0.9871552030306617
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 0.9826959501582727,
      "learning_rate": 8.648637703194515e-06,
      "loss": 0.04538998007774353,
      "step": 2700,
      "token_acc": 0.9856196890897206
    },
    {
      "epoch": 1.7044025157232703,
      "grad_norm": 0.9001332673994735,
      "learning_rate": 8.579971223788187e-06,
      "loss": 0.04760159850120545,
      "step": 2710,
      "token_acc": 0.9858047797293407
    },
    {
      "epoch": 1.7106918238993711,
      "grad_norm": 1.4201032363470754,
      "learning_rate": 8.511373010785837e-06,
      "loss": 0.047868388891220096,
      "step": 2720,
      "token_acc": 0.9851355025831596
    },
    {
      "epoch": 1.7169811320754715,
      "grad_norm": 0.6636891523950763,
      "learning_rate": 8.44284636197508e-06,
      "loss": 0.044393020868301394,
      "step": 2730,
      "token_acc": 0.986869384934347
    },
    {
      "epoch": 1.7232704402515724,
      "grad_norm": 0.9926167607852617,
      "learning_rate": 8.374394571703173e-06,
      "loss": 0.04762988090515137,
      "step": 2740,
      "token_acc": 0.9852822347916426
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 0.8607059886840823,
      "learning_rate": 8.306020930718608e-06,
      "loss": 0.040895530581474306,
      "step": 2750,
      "token_acc": 0.9872684168934938
    },
    {
      "epoch": 1.7358490566037736,
      "grad_norm": 0.7767146375653069,
      "learning_rate": 8.237728726012953e-06,
      "loss": 0.041637516021728514,
      "step": 2760,
      "token_acc": 0.986967220585108
    },
    {
      "epoch": 1.742138364779874,
      "grad_norm": 0.743008860732743,
      "learning_rate": 8.169521240662788e-06,
      "loss": 0.04418378472328186,
      "step": 2770,
      "token_acc": 0.9866983791402396
    },
    {
      "epoch": 1.748427672955975,
      "grad_norm": 0.8964151609893208,
      "learning_rate": 8.10140175367192e-06,
      "loss": 0.0516668438911438,
      "step": 2780,
      "token_acc": 0.9840028126922739
    },
    {
      "epoch": 1.7547169811320755,
      "grad_norm": 1.1166309562768149,
      "learning_rate": 8.033373539813707e-06,
      "loss": 0.046999433636665346,
      "step": 2790,
      "token_acc": 0.9857126397050351
    },
    {
      "epoch": 1.7610062893081762,
      "grad_norm": 0.7683616948363582,
      "learning_rate": 7.965439869473664e-06,
      "loss": 0.044093912839889525,
      "step": 2800,
      "token_acc": 0.9867604713984623
    },
    {
      "epoch": 1.7672955974842768,
      "grad_norm": 1.073793810583486,
      "learning_rate": 7.897604008492213e-06,
      "loss": 0.046621406078338624,
      "step": 2810,
      "token_acc": 0.9859511745739291
    },
    {
      "epoch": 1.7735849056603774,
      "grad_norm": 1.0453627363845124,
      "learning_rate": 7.829869218007698e-06,
      "loss": 0.04217245578765869,
      "step": 2820,
      "token_acc": 0.987036982725753
    },
    {
      "epoch": 1.779874213836478,
      "grad_norm": 1.1770300073955866,
      "learning_rate": 7.762238754299596e-06,
      "loss": 0.04043961763381958,
      "step": 2830,
      "token_acc": 0.9876352395672334
    },
    {
      "epoch": 1.7861635220125787,
      "grad_norm": 0.9601702623421838,
      "learning_rate": 7.694715868632002e-06,
      "loss": 0.04152000546455383,
      "step": 2840,
      "token_acc": 0.9877592655559334
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 0.818686382169386,
      "learning_rate": 7.627303807097285e-06,
      "loss": 0.04074779152870178,
      "step": 2850,
      "token_acc": 0.9871859578286227
    },
    {
      "epoch": 1.79874213836478,
      "grad_norm": 0.7321550298673272,
      "learning_rate": 7.5600058104600805e-06,
      "loss": 0.041087231040000914,
      "step": 2860,
      "token_acc": 0.9878531152691671
    },
    {
      "epoch": 1.8050314465408805,
      "grad_norm": 0.8721067816123668,
      "learning_rate": 7.492825114001456e-06,
      "loss": 0.044392266869544984,
      "step": 2870,
      "token_acc": 0.986461784531018
    },
    {
      "epoch": 1.8113207547169812,
      "grad_norm": 0.9327119914838725,
      "learning_rate": 7.425764947363409e-06,
      "loss": 0.04353358447551727,
      "step": 2880,
      "token_acc": 0.986205931729155
    },
    {
      "epoch": 1.8176100628930818,
      "grad_norm": 0.7411353160886415,
      "learning_rate": 7.358828534393574e-06,
      "loss": 0.043460381031036374,
      "step": 2890,
      "token_acc": 0.9866405055255792
    },
    {
      "epoch": 1.8238993710691824,
      "grad_norm": 0.9077406542379596,
      "learning_rate": 7.29201909299028e-06,
      "loss": 0.05039287209510803,
      "step": 2900,
      "token_acc": 0.9835132073221633
    },
    {
      "epoch": 1.830188679245283,
      "grad_norm": 0.8855725031776173,
      "learning_rate": 7.225339834947811e-06,
      "loss": 0.04200481176376343,
      "step": 2910,
      "token_acc": 0.9875679212540184
    },
    {
      "epoch": 1.8364779874213837,
      "grad_norm": 0.9207682012858668,
      "learning_rate": 7.1587939658020275e-06,
      "loss": 0.04500528573989868,
      "step": 2920,
      "token_acc": 0.9862918819081701
    },
    {
      "epoch": 1.8427672955974843,
      "grad_norm": 0.7168665613965597,
      "learning_rate": 7.092384684676263e-06,
      "loss": 0.04347011148929596,
      "step": 2930,
      "token_acc": 0.9861606148362664
    },
    {
      "epoch": 1.849056603773585,
      "grad_norm": 1.068918453766715,
      "learning_rate": 7.026115184127518e-06,
      "loss": 0.044189158082008365,
      "step": 2940,
      "token_acc": 0.9863891583619955
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 1.046246115613562,
      "learning_rate": 6.9599886499929844e-06,
      "loss": 0.04125094413757324,
      "step": 2950,
      "token_acc": 0.9870236182275718
    },
    {
      "epoch": 1.8616352201257862,
      "grad_norm": 0.7754228161073464,
      "learning_rate": 6.894008261236907e-06,
      "loss": 0.04711427390575409,
      "step": 2960,
      "token_acc": 0.9856841252855945
    },
    {
      "epoch": 1.8679245283018868,
      "grad_norm": 0.7882340241312703,
      "learning_rate": 6.8281771897977274e-06,
      "loss": 0.043092700839042666,
      "step": 2970,
      "token_acc": 0.9864337892019553
    },
    {
      "epoch": 1.8742138364779874,
      "grad_norm": 0.7679388139347542,
      "learning_rate": 6.762498600435624e-06,
      "loss": 0.044446122646331784,
      "step": 2980,
      "token_acc": 0.9866929945054945
    },
    {
      "epoch": 1.880503144654088,
      "grad_norm": 1.0341275463341217,
      "learning_rate": 6.696975650580347e-06,
      "loss": 0.0423536092042923,
      "step": 2990,
      "token_acc": 0.9872913616398243
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 1.4903969853297627,
      "learning_rate": 6.631611490179452e-06,
      "loss": 0.03887941539287567,
      "step": 3000,
      "token_acc": 0.9879372399055224
    },
    {
      "epoch": 1.8867924528301887,
      "eval_loss": 0.05534382909536362,
      "eval_runtime": 717.8792,
      "eval_samples_per_second": 11.806,
      "eval_steps_per_second": 2.952,
      "eval_token_acc": 0.9835461426217628,
      "step": 3000
    },
    {
      "epoch": 1.8930817610062893,
      "grad_norm": 0.9513668679732343,
      "learning_rate": 6.566409261546853e-06,
      "loss": 0.04525440037250519,
      "step": 3010,
      "token_acc": 0.9861626332214567
    },
    {
      "epoch": 1.89937106918239,
      "grad_norm": 0.9978635657308409,
      "learning_rate": 6.501372099211758e-06,
      "loss": 0.044601985812187196,
      "step": 3020,
      "token_acc": 0.9857926366765856
    },
    {
      "epoch": 1.9056603773584906,
      "grad_norm": 1.074608364002311,
      "learning_rate": 6.43650312976799e-06,
      "loss": 0.03974335789680481,
      "step": 3030,
      "token_acc": 0.9873547505126452
    },
    {
      "epoch": 1.9119496855345912,
      "grad_norm": 1.185976614312186,
      "learning_rate": 6.371805471723667e-06,
      "loss": 0.043976855278015134,
      "step": 3040,
      "token_acc": 0.9867405217690784
    },
    {
      "epoch": 1.9182389937106918,
      "grad_norm": 0.7831735892695594,
      "learning_rate": 6.307282235351302e-06,
      "loss": 0.04128652811050415,
      "step": 3050,
      "token_acc": 0.9875675361645269
    },
    {
      "epoch": 1.9245283018867925,
      "grad_norm": 0.7682687872810261,
      "learning_rate": 6.2429365225382565e-06,
      "loss": 0.04307092428207397,
      "step": 3060,
      "token_acc": 0.9865218029657906
    },
    {
      "epoch": 1.930817610062893,
      "grad_norm": 1.0309616532569668,
      "learning_rate": 6.17877142663764e-06,
      "loss": 0.04245474934577942,
      "step": 3070,
      "token_acc": 0.9868967809850961
    },
    {
      "epoch": 1.9371069182389937,
      "grad_norm": 1.1348771283347114,
      "learning_rate": 6.114790032319579e-06,
      "loss": 0.038907968997955324,
      "step": 3080,
      "token_acc": 0.9880084782671815
    },
    {
      "epoch": 1.9433962264150944,
      "grad_norm": 0.9427207864096449,
      "learning_rate": 6.05099541542296e-06,
      "loss": 0.04092976450920105,
      "step": 3090,
      "token_acc": 0.987441221998306
    },
    {
      "epoch": 1.949685534591195,
      "grad_norm": 0.7102412660886335,
      "learning_rate": 5.987390642807516e-06,
      "loss": 0.04200551807880402,
      "step": 3100,
      "token_acc": 0.986920228087708
    },
    {
      "epoch": 1.9559748427672956,
      "grad_norm": 0.7929334085206099,
      "learning_rate": 5.923978772206432e-06,
      "loss": 0.042564332485198975,
      "step": 3110,
      "token_acc": 0.986267625573362
    },
    {
      "epoch": 1.9622641509433962,
      "grad_norm": 1.2645997361167853,
      "learning_rate": 5.860762852079318e-06,
      "loss": 0.04022677540779114,
      "step": 3120,
      "token_acc": 0.9878765190347291
    },
    {
      "epoch": 1.9685534591194969,
      "grad_norm": 1.2869788893371823,
      "learning_rate": 5.797745921465674e-06,
      "loss": 0.044468334317207335,
      "step": 3130,
      "token_acc": 0.9864629444523776
    },
    {
      "epoch": 1.9748427672955975,
      "grad_norm": 0.7375696397453761,
      "learning_rate": 5.734931009838785e-06,
      "loss": 0.04471827745437622,
      "step": 3140,
      "token_acc": 0.9861083206566976
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 0.7104888528538673,
      "learning_rate": 5.67232113696009e-06,
      "loss": 0.046862339973449706,
      "step": 3150,
      "token_acc": 0.9855168251961264
    },
    {
      "epoch": 1.9874213836477987,
      "grad_norm": 0.6304186527459887,
      "learning_rate": 5.609919312733987e-06,
      "loss": 0.040326648950576784,
      "step": 3160,
      "token_acc": 0.9875191903110252
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.8206946118047505,
      "learning_rate": 5.5477285370631776e-06,
      "loss": 0.03785489201545715,
      "step": 3170,
      "token_acc": 0.9877561289240488
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.6185232611309048,
      "learning_rate": 5.485751799704394e-06,
      "loss": 0.03770981431007385,
      "step": 3180,
      "token_acc": 0.9875341530054644
    },
    {
      "epoch": 2.006289308176101,
      "grad_norm": 0.4508860979571277,
      "learning_rate": 5.423992080124732e-06,
      "loss": 0.027628958225250244,
      "step": 3190,
      "token_acc": 0.9909122482056032
    },
    {
      "epoch": 2.0125786163522013,
      "grad_norm": 0.7265688364646201,
      "learning_rate": 5.362452347358351e-06,
      "loss": 0.02676335573196411,
      "step": 3200,
      "token_acc": 0.9914142278509898
    },
    {
      "epoch": 2.018867924528302,
      "grad_norm": 0.7176493010340509,
      "learning_rate": 5.301135559863797e-06,
      "loss": 0.026176384091377257,
      "step": 3210,
      "token_acc": 0.9912505572893446
    },
    {
      "epoch": 2.0251572327044025,
      "grad_norm": 0.4203332470924716,
      "learning_rate": 5.240044665381746e-06,
      "loss": 0.026522904634475708,
      "step": 3220,
      "token_acc": 0.9912850258274897
    },
    {
      "epoch": 2.0314465408805034,
      "grad_norm": 0.49076741004829705,
      "learning_rate": 5.179182600793302e-06,
      "loss": 0.028085267543792723,
      "step": 3230,
      "token_acc": 0.9911830389512457
    },
    {
      "epoch": 2.0377358490566038,
      "grad_norm": 0.9034204793151985,
      "learning_rate": 5.118552291978813e-06,
      "loss": 0.02409302443265915,
      "step": 3240,
      "token_acc": 0.9916755174096507
    },
    {
      "epoch": 2.0440251572327046,
      "grad_norm": 0.9532940107352192,
      "learning_rate": 5.058156653677208e-06,
      "loss": 0.02635040283203125,
      "step": 3250,
      "token_acc": 0.9913842784693407
    },
    {
      "epoch": 2.050314465408805,
      "grad_norm": 1.3923378199552692,
      "learning_rate": 4.997998589345876e-06,
      "loss": 0.027022877335548402,
      "step": 3260,
      "token_acc": 0.9907599907599908
    },
    {
      "epoch": 2.056603773584906,
      "grad_norm": 0.7649477167428459,
      "learning_rate": 4.938080991021088e-06,
      "loss": 0.027101951837539672,
      "step": 3270,
      "token_acc": 0.990909090909091
    },
    {
      "epoch": 2.0628930817610063,
      "grad_norm": 0.8061149358869371,
      "learning_rate": 4.878406739178948e-06,
      "loss": 0.025409945845603944,
      "step": 3280,
      "token_acc": 0.9915269196822595
    },
    {
      "epoch": 2.069182389937107,
      "grad_norm": 0.9591310020250684,
      "learning_rate": 4.818978702596957e-06,
      "loss": 0.025764793157577515,
      "step": 3290,
      "token_acc": 0.9913785307328249
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.8738700593689223,
      "learning_rate": 4.759799738216048e-06,
      "loss": 0.02476211339235306,
      "step": 3300,
      "token_acc": 0.9920101738962152
    },
    {
      "epoch": 2.0817610062893084,
      "grad_norm": 0.8666216893328724,
      "learning_rate": 4.700872691003283e-06,
      "loss": 0.0245857834815979,
      "step": 3310,
      "token_acc": 0.9914979636731175
    },
    {
      "epoch": 2.088050314465409,
      "grad_norm": 0.7330006799490474,
      "learning_rate": 4.642200393815065e-06,
      "loss": 0.024617944657802582,
      "step": 3320,
      "token_acc": 0.9923877978380214
    },
    {
      "epoch": 2.0943396226415096,
      "grad_norm": 0.7358978404096974,
      "learning_rate": 4.583785667260953e-06,
      "loss": 0.025178509950637817,
      "step": 3330,
      "token_acc": 0.9915244431588082
    },
    {
      "epoch": 2.10062893081761,
      "grad_norm": 0.5429978769204982,
      "learning_rate": 4.525631319568067e-06,
      "loss": 0.022534635663032532,
      "step": 3340,
      "token_acc": 0.9924528301886792
    },
    {
      "epoch": 2.106918238993711,
      "grad_norm": 0.40319756075770474,
      "learning_rate": 4.467740146446086e-06,
      "loss": 0.02614726424217224,
      "step": 3350,
      "token_acc": 0.9916803803254708
    },
    {
      "epoch": 2.1132075471698113,
      "grad_norm": 0.826434146525126,
      "learning_rate": 4.410114930952844e-06,
      "loss": 0.024831095337867738,
      "step": 3360,
      "token_acc": 0.9915295464301045
    },
    {
      "epoch": 2.119496855345912,
      "grad_norm": 0.63976873943924,
      "learning_rate": 4.352758443360545e-06,
      "loss": 0.024512295424938203,
      "step": 3370,
      "token_acc": 0.9919333853760083
    },
    {
      "epoch": 2.1257861635220126,
      "grad_norm": 0.6064353901033486,
      "learning_rate": 4.29567344102256e-06,
      "loss": 0.023602843284606934,
      "step": 3380,
      "token_acc": 0.9924209170982939
    },
    {
      "epoch": 2.1320754716981134,
      "grad_norm": 0.7517058145196435,
      "learning_rate": 4.23886266824092e-06,
      "loss": 0.02254200279712677,
      "step": 3390,
      "token_acc": 0.9921487603305785
    },
    {
      "epoch": 2.138364779874214,
      "grad_norm": 0.9220898043923944,
      "learning_rate": 4.1823288561343236e-06,
      "loss": 0.02772650122642517,
      "step": 3400,
      "token_acc": 0.9909362025612537
    },
    {
      "epoch": 2.1446540880503147,
      "grad_norm": 2.5322794453411097,
      "learning_rate": 4.126074722506902e-06,
      "loss": 0.0227973073720932,
      "step": 3410,
      "token_acc": 0.992287006117202
    },
    {
      "epoch": 2.150943396226415,
      "grad_norm": 0.6637478582466627,
      "learning_rate": 4.070102971717512e-06,
      "loss": 0.023247498273849487,
      "step": 3420,
      "token_acc": 0.9925187032418953
    },
    {
      "epoch": 2.157232704402516,
      "grad_norm": 0.6855147878689783,
      "learning_rate": 4.0144162945497634e-06,
      "loss": 0.027481943368911743,
      "step": 3430,
      "token_acc": 0.9906596957058098
    },
    {
      "epoch": 2.1635220125786163,
      "grad_norm": 0.6951399810120154,
      "learning_rate": 3.9590173680826425e-06,
      "loss": 0.028112170100212098,
      "step": 3440,
      "token_acc": 0.9909493103547986
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 0.9163370475340076,
      "learning_rate": 3.9039088555618256e-06,
      "loss": 0.02361641824245453,
      "step": 3450,
      "token_acc": 0.9920540556216106
    },
    {
      "epoch": 2.1761006289308176,
      "grad_norm": 0.7173776749029771,
      "learning_rate": 3.849093406271634e-06,
      "loss": 0.02243003249168396,
      "step": 3460,
      "token_acc": 0.992677332078579
    },
    {
      "epoch": 2.1823899371069184,
      "grad_norm": 0.934931570917733,
      "learning_rate": 3.7945736554076917e-06,
      "loss": 0.029029443860054016,
      "step": 3470,
      "token_acc": 0.9900564906617477
    },
    {
      "epoch": 2.188679245283019,
      "grad_norm": 0.6207792354919172,
      "learning_rate": 3.7403522239502065e-06,
      "loss": 0.02085772156715393,
      "step": 3480,
      "token_acc": 0.9933544655766878
    },
    {
      "epoch": 2.1949685534591197,
      "grad_norm": 0.5912514975563645,
      "learning_rate": 3.6864317185380206e-06,
      "loss": 0.020345935225486757,
      "step": 3490,
      "token_acc": 0.9932405566600397
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 0.6823875562891651,
      "learning_rate": 3.6328147313432427e-06,
      "loss": 0.024584601819515228,
      "step": 3500,
      "token_acc": 0.9920973696354407
    },
    {
      "epoch": 2.20125786163522,
      "eval_loss": 0.056817859411239624,
      "eval_runtime": 715.3532,
      "eval_samples_per_second": 11.847,
      "eval_steps_per_second": 2.962,
      "eval_token_acc": 0.9844349694029099,
      "step": 3500
    },
    {
      "epoch": 2.207547169811321,
      "grad_norm": 0.5403476894342643,
      "learning_rate": 3.5795038399466832e-06,
      "loss": 0.021820729970932005,
      "step": 3510,
      "token_acc": 0.9930049964311206
    },
    {
      "epoch": 2.2138364779874213,
      "grad_norm": 0.5875323337715701,
      "learning_rate": 3.526501607213897e-06,
      "loss": 0.02411033511161804,
      "step": 3520,
      "token_acc": 0.9922319599362622
    },
    {
      "epoch": 2.220125786163522,
      "grad_norm": 1.242104157658184,
      "learning_rate": 3.4738105811720047e-06,
      "loss": 0.02497299760580063,
      "step": 3530,
      "token_acc": 0.9922028985507246
    },
    {
      "epoch": 2.2264150943396226,
      "grad_norm": 0.7858727250077854,
      "learning_rate": 3.4214332948871885e-06,
      "loss": 0.024555695056915284,
      "step": 3540,
      "token_acc": 0.9920254293999554
    },
    {
      "epoch": 2.2327044025157234,
      "grad_norm": 0.7457502910203632,
      "learning_rate": 3.3693722663429186e-06,
      "loss": 0.02635159492492676,
      "step": 3550,
      "token_acc": 0.9917847362781284
    },
    {
      "epoch": 2.238993710691824,
      "grad_norm": 0.8007577002256944,
      "learning_rate": 3.3176299983189007e-06,
      "loss": 0.021693721413612366,
      "step": 3560,
      "token_acc": 0.9924659684369739
    },
    {
      "epoch": 2.2452830188679247,
      "grad_norm": 0.8938253603482125,
      "learning_rate": 3.266208978270765e-06,
      "loss": 0.023349395394325255,
      "step": 3570,
      "token_acc": 0.9923470839260313
    },
    {
      "epoch": 2.251572327044025,
      "grad_norm": 0.8019690456329862,
      "learning_rate": 3.2151116782104785e-06,
      "loss": 0.025201982259750365,
      "step": 3580,
      "token_acc": 0.9916726696441435
    },
    {
      "epoch": 2.257861635220126,
      "grad_norm": 0.928512618313778,
      "learning_rate": 3.1643405545875074e-06,
      "loss": 0.022884893417358398,
      "step": 3590,
      "token_acc": 0.9926495824276399
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 1.8518694118902743,
      "learning_rate": 3.1138980481707157e-06,
      "loss": 0.022659416496753692,
      "step": 3600,
      "token_acc": 0.9921473526185115
    },
    {
      "epoch": 2.270440251572327,
      "grad_norm": 0.7157409797457132,
      "learning_rate": 3.063786583931061e-06,
      "loss": 0.021560478210449218,
      "step": 3610,
      "token_acc": 0.9924329154217743
    },
    {
      "epoch": 2.2767295597484276,
      "grad_norm": 0.683960093277271,
      "learning_rate": 3.0140085709249666e-06,
      "loss": 0.022942088544368744,
      "step": 3620,
      "token_acc": 0.9926606502120257
    },
    {
      "epoch": 2.2830188679245285,
      "grad_norm": 0.3807671208250297,
      "learning_rate": 2.9645664021785514e-06,
      "loss": 0.022927108407020568,
      "step": 3630,
      "token_acc": 0.9927953890489913
    },
    {
      "epoch": 2.289308176100629,
      "grad_norm": 0.5798902198831075,
      "learning_rate": 2.9154624545725687e-06,
      "loss": 0.021968087553977965,
      "step": 3640,
      "token_acc": 0.9928758279067179
    },
    {
      "epoch": 2.2955974842767297,
      "grad_norm": 0.8968069110950454,
      "learning_rate": 2.8666990887281376e-06,
      "loss": 0.024356882274150848,
      "step": 3650,
      "token_acc": 0.9921203845935028
    },
    {
      "epoch": 2.30188679245283,
      "grad_norm": 0.9101388836181897,
      "learning_rate": 2.818278648893271e-06,
      "loss": 0.022238627076148987,
      "step": 3660,
      "token_acc": 0.992795683243401
    },
    {
      "epoch": 2.308176100628931,
      "grad_norm": 0.6722990881265821,
      "learning_rate": 2.7702034628301654e-06,
      "loss": 0.02386687248945236,
      "step": 3670,
      "token_acc": 0.9920182111572153
    },
    {
      "epoch": 2.3144654088050314,
      "grad_norm": 0.58030818463452,
      "learning_rate": 2.722475841703306e-06,
      "loss": 0.021091663837432863,
      "step": 3680,
      "token_acc": 0.9926483257710643
    },
    {
      "epoch": 2.3207547169811322,
      "grad_norm": 1.011732851655928,
      "learning_rate": 2.6750980799683555e-06,
      "loss": 0.027016055583953858,
      "step": 3690,
      "token_acc": 0.991425876704107
    },
    {
      "epoch": 2.3270440251572326,
      "grad_norm": 0.5918250477640592,
      "learning_rate": 2.6280724552618408e-06,
      "loss": 0.023545855283737184,
      "step": 3700,
      "token_acc": 0.9922041105598866
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.48516060801912675,
      "learning_rate": 2.58140122829169e-06,
      "loss": 0.024050341546535493,
      "step": 3710,
      "token_acc": 0.9917407154249259
    },
    {
      "epoch": 2.339622641509434,
      "grad_norm": 0.526800520074625,
      "learning_rate": 2.5350866427285014e-06,
      "loss": 0.025029301643371582,
      "step": 3720,
      "token_acc": 0.9919972463643404
    },
    {
      "epoch": 2.3459119496855347,
      "grad_norm": 0.5899523312172913,
      "learning_rate": 2.489130925097737e-06,
      "loss": 0.022905370593070982,
      "step": 3730,
      "token_acc": 0.992318268770923
    },
    {
      "epoch": 2.352201257861635,
      "grad_norm": 0.7955845513016064,
      "learning_rate": 2.443536284672635e-06,
      "loss": 0.023500500619411467,
      "step": 3740,
      "token_acc": 0.992278659713619
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 0.6747119310051193,
      "learning_rate": 2.3983049133680336e-06,
      "loss": 0.020411881804466247,
      "step": 3750,
      "token_acc": 0.993043378273493
    },
    {
      "epoch": 2.3647798742138364,
      "grad_norm": 0.551637714502972,
      "learning_rate": 2.353438985634984e-06,
      "loss": 0.021443535387516022,
      "step": 3760,
      "token_acc": 0.9929708672086721
    },
    {
      "epoch": 2.3710691823899372,
      "grad_norm": 0.7071071838410758,
      "learning_rate": 2.30894065835622e-06,
      "loss": 0.021306820213794708,
      "step": 3770,
      "token_acc": 0.9928304327044732
    },
    {
      "epoch": 2.3773584905660377,
      "grad_norm": 0.6383595298983202,
      "learning_rate": 2.2648120707424657e-06,
      "loss": 0.01998791694641113,
      "step": 3780,
      "token_acc": 0.9933602358261069
    },
    {
      "epoch": 2.3836477987421385,
      "grad_norm": 0.4429036125000402,
      "learning_rate": 2.221055344229598e-06,
      "loss": 0.020099201798439027,
      "step": 3790,
      "token_acc": 0.9931795746208382
    },
    {
      "epoch": 2.389937106918239,
      "grad_norm": 0.5189498149492784,
      "learning_rate": 2.1776725823766474e-06,
      "loss": 0.02125477194786072,
      "step": 3800,
      "token_acc": 0.9930950215958172
    },
    {
      "epoch": 2.3962264150943398,
      "grad_norm": 1.231799709812971,
      "learning_rate": 2.134665870764705e-06,
      "loss": 0.02221446931362152,
      "step": 3810,
      "token_acc": 0.9928049402300153
    },
    {
      "epoch": 2.40251572327044,
      "grad_norm": 0.8890189048591385,
      "learning_rate": 2.0920372768966133e-06,
      "loss": 0.021392084658145905,
      "step": 3820,
      "token_acc": 0.9932545385224506
    },
    {
      "epoch": 2.408805031446541,
      "grad_norm": 0.6308759320373736,
      "learning_rate": 2.0497888500976217e-06,
      "loss": 0.023821352422237395,
      "step": 3830,
      "token_acc": 0.9918196384898659
    },
    {
      "epoch": 2.4150943396226414,
      "grad_norm": 0.7886414731779052,
      "learning_rate": 2.0079226214168233e-06,
      "loss": 0.024218729138374327,
      "step": 3840,
      "token_acc": 0.9919518856412343
    },
    {
      "epoch": 2.4213836477987423,
      "grad_norm": 0.37738056026913436,
      "learning_rate": 1.9664406035295493e-06,
      "loss": 0.021755827963352202,
      "step": 3850,
      "token_acc": 0.992849932705249
    },
    {
      "epoch": 2.4276729559748427,
      "grad_norm": 0.48646463365440634,
      "learning_rate": 1.92534479064059e-06,
      "loss": 0.024741561710834505,
      "step": 3860,
      "token_acc": 0.992130426198065
    },
    {
      "epoch": 2.4339622641509435,
      "grad_norm": 0.8159345779241457,
      "learning_rate": 1.884637158388335e-06,
      "loss": 0.023516392707824706,
      "step": 3870,
      "token_acc": 0.9920172722365842
    },
    {
      "epoch": 2.440251572327044,
      "grad_norm": 0.951255033160176,
      "learning_rate": 1.8443196637497952e-06,
      "loss": 0.025745093822479248,
      "step": 3880,
      "token_acc": 0.9912382164131629
    },
    {
      "epoch": 2.4465408805031448,
      "grad_norm": 0.4221558934805368,
      "learning_rate": 1.804394244946519e-06,
      "loss": 0.02262389361858368,
      "step": 3890,
      "token_acc": 0.992309198198954
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 0.5344840976706485,
      "learning_rate": 1.7648628213514219e-06,
      "loss": 0.021597489714622498,
      "step": 3900,
      "token_acc": 0.9930053625553742
    },
    {
      "epoch": 2.459119496855346,
      "grad_norm": 0.5725230961295943,
      "learning_rate": 1.7257272933965074e-06,
      "loss": 0.022534751892089845,
      "step": 3910,
      "token_acc": 0.9928057553956835
    },
    {
      "epoch": 2.4654088050314464,
      "grad_norm": 0.7446463840263802,
      "learning_rate": 1.6869895424815074e-06,
      "loss": 0.025363484025001527,
      "step": 3920,
      "token_acc": 0.9917138750317072
    },
    {
      "epoch": 2.4716981132075473,
      "grad_norm": 0.6731229772496609,
      "learning_rate": 1.64865143088344e-06,
      "loss": 0.021260468661785124,
      "step": 3930,
      "token_acc": 0.9929787607512726
    },
    {
      "epoch": 2.4779874213836477,
      "grad_norm": 0.554208376735791,
      "learning_rate": 1.6107148016670805e-06,
      "loss": 0.023889759182929994,
      "step": 3940,
      "token_acc": 0.9920448112504842
    },
    {
      "epoch": 2.4842767295597485,
      "grad_norm": 0.67523932744472,
      "learning_rate": 1.5731814785963474e-06,
      "loss": 0.02100624144077301,
      "step": 3950,
      "token_acc": 0.9932140130630248
    },
    {
      "epoch": 2.490566037735849,
      "grad_norm": 0.70634398052636,
      "learning_rate": 1.5360532660466521e-06,
      "loss": 0.02119089215993881,
      "step": 3960,
      "token_acc": 0.9931327394577898
    },
    {
      "epoch": 2.49685534591195,
      "grad_norm": 0.8873229135241124,
      "learning_rate": 1.4993319489181234e-06,
      "loss": 0.02107432782649994,
      "step": 3970,
      "token_acc": 0.9928768056304452
    },
    {
      "epoch": 2.50314465408805,
      "grad_norm": 0.793292346733035,
      "learning_rate": 1.4630192925498344e-06,
      "loss": 0.021246784925460817,
      "step": 3980,
      "token_acc": 0.9925140911225928
    },
    {
      "epoch": 2.509433962264151,
      "grad_norm": 0.7213692163823019,
      "learning_rate": 1.427117042634899e-06,
      "loss": 0.022090838849544527,
      "step": 3990,
      "token_acc": 0.9922656578394283
    },
    {
      "epoch": 2.5157232704402515,
      "grad_norm": 0.3399190872934864,
      "learning_rate": 1.391626925136581e-06,
      "loss": 0.023202145099639894,
      "step": 4000,
      "token_acc": 0.992874826399372
    },
    {
      "epoch": 2.5157232704402515,
      "eval_loss": 0.0549781434237957,
      "eval_runtime": 714.4039,
      "eval_samples_per_second": 11.863,
      "eval_steps_per_second": 2.966,
      "eval_token_acc": 0.9851726793821435,
      "step": 4000
    },
    {
      "epoch": 2.5220125786163523,
      "grad_norm": 0.6374524762426332,
      "learning_rate": 1.3565506462053036e-06,
      "loss": 0.02244551628828049,
      "step": 4010,
      "token_acc": 0.9929325538783701
    },
    {
      "epoch": 2.5283018867924527,
      "grad_norm": 0.38379286396884926,
      "learning_rate": 1.321889892096634e-06,
      "loss": 0.020600757002830504,
      "step": 4020,
      "token_acc": 0.9929416471395096
    },
    {
      "epoch": 2.5345911949685536,
      "grad_norm": 0.41557599954404484,
      "learning_rate": 1.287646329090214e-06,
      "loss": 0.020873698592185973,
      "step": 4030,
      "token_acc": 0.9927871772039181
    },
    {
      "epoch": 2.540880503144654,
      "grad_norm": 0.523620093961174,
      "learning_rate": 1.2538216034096606e-06,
      "loss": 0.02275868207216263,
      "step": 4040,
      "token_acc": 0.9927199368476449
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 0.706260703041499,
      "learning_rate": 1.2204173411434161e-06,
      "loss": 0.02068159282207489,
      "step": 4050,
      "token_acc": 0.9934766365883096
    },
    {
      "epoch": 2.5534591194968552,
      "grad_norm": 0.5067252058108148,
      "learning_rate": 1.1874351481665935e-06,
      "loss": 0.023539449274539947,
      "step": 4060,
      "token_acc": 0.9921726553415955
    },
    {
      "epoch": 2.559748427672956,
      "grad_norm": 0.5384145679125522,
      "learning_rate": 1.1548766100637522e-06,
      "loss": 0.022212520241737366,
      "step": 4070,
      "token_acc": 0.9930183982374317
    },
    {
      "epoch": 2.5660377358490565,
      "grad_norm": 0.3706695255239234,
      "learning_rate": 1.122743292052697e-06,
      "loss": 0.021440133452415466,
      "step": 4080,
      "token_acc": 0.9922975290248222
    },
    {
      "epoch": 2.5723270440251573,
      "grad_norm": 0.6377989611149152,
      "learning_rate": 1.091036738909208e-06,
      "loss": 0.020651139318943024,
      "step": 4090,
      "token_acc": 0.9925337658342888
    },
    {
      "epoch": 2.5786163522012577,
      "grad_norm": 0.6231957294617673,
      "learning_rate": 1.0597584748927957e-06,
      "loss": 0.020253558456897736,
      "step": 4100,
      "token_acc": 0.9932907255848926
    },
    {
      "epoch": 2.5849056603773586,
      "grad_norm": 0.5196778496757918,
      "learning_rate": 1.0289100036734179e-06,
      "loss": 0.018897609412670137,
      "step": 4110,
      "token_acc": 0.9933135004958209
    },
    {
      "epoch": 2.591194968553459,
      "grad_norm": 0.5911443597758588,
      "learning_rate": 9.984928082591883e-07,
      "loss": 0.021263836324214934,
      "step": 4120,
      "token_acc": 0.9930807719799857
    },
    {
      "epoch": 2.59748427672956,
      "grad_norm": 0.7333439191126617,
      "learning_rate": 9.68508350925086e-07,
      "loss": 0.021468062698841096,
      "step": 4130,
      "token_acc": 0.9936830866870311
    },
    {
      "epoch": 2.6037735849056602,
      "grad_norm": 0.6243357787231448,
      "learning_rate": 9.389580731426595e-07,
      "loss": 0.023629823327064516,
      "step": 4140,
      "token_acc": 0.9919061683516983
    },
    {
      "epoch": 2.610062893081761,
      "grad_norm": 0.3768531248387712,
      "learning_rate": 9.098433955107245e-07,
      "loss": 0.022948920726776123,
      "step": 4150,
      "token_acc": 0.9923590672402083
    },
    {
      "epoch": 2.6163522012578615,
      "grad_norm": 0.41655994069801966,
      "learning_rate": 8.811657176870803e-07,
      "loss": 0.017796531319618225,
      "step": 4160,
      "token_acc": 0.994064597919728
    },
    {
      "epoch": 2.6226415094339623,
      "grad_norm": 0.7498471436247828,
      "learning_rate": 8.529264183212028e-07,
      "loss": 0.02242247611284256,
      "step": 4170,
      "token_acc": 0.9927712917013282
    },
    {
      "epoch": 2.6289308176100628,
      "grad_norm": 0.7164144826549131,
      "learning_rate": 8.251268549879965e-07,
      "loss": 0.019840958714485168,
      "step": 4180,
      "token_acc": 0.9937063932792254
    },
    {
      "epoch": 2.6352201257861636,
      "grad_norm": 0.5219754899465664,
      "learning_rate": 7.977683641224998e-07,
      "loss": 0.020760795474052428,
      "step": 4190,
      "token_acc": 0.9930836902419261
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 0.7751634890953712,
      "learning_rate": 7.708522609556613e-07,
      "loss": 0.02397152632474899,
      "step": 4200,
      "token_acc": 0.9922862347717181
    },
    {
      "epoch": 2.647798742138365,
      "grad_norm": 0.8400838444811691,
      "learning_rate": 7.443798394510971e-07,
      "loss": 0.01996571719646454,
      "step": 4210,
      "token_acc": 0.9935876502894463
    },
    {
      "epoch": 2.6540880503144653,
      "grad_norm": 0.9743339975258191,
      "learning_rate": 7.183523722428909e-07,
      "loss": 0.021362194418907167,
      "step": 4220,
      "token_acc": 0.9933829611248967
    },
    {
      "epoch": 2.660377358490566,
      "grad_norm": 0.7468883773281543,
      "learning_rate": 6.927711105744139e-07,
      "loss": 0.023838293552398682,
      "step": 4230,
      "token_acc": 0.9920445967133151
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.7047724581879639,
      "learning_rate": 6.676372842381674e-07,
      "loss": 0.019323495030403138,
      "step": 4240,
      "token_acc": 0.993488812596188
    },
    {
      "epoch": 2.6729559748427674,
      "grad_norm": 0.5330074971260592,
      "learning_rate": 6.429521015166684e-07,
      "loss": 0.022193637490272523,
      "step": 4250,
      "token_acc": 0.9929509953194608
    },
    {
      "epoch": 2.6792452830188678,
      "grad_norm": 0.9456292680760856,
      "learning_rate": 6.187167491243596e-07,
      "loss": 0.018984417617321014,
      "step": 4260,
      "token_acc": 0.9938085216701742
    },
    {
      "epoch": 2.6855345911949686,
      "grad_norm": 0.6020208718543678,
      "learning_rate": 5.949323921505534e-07,
      "loss": 0.02061026692390442,
      "step": 4270,
      "token_acc": 0.9933976637887253
    },
    {
      "epoch": 2.691823899371069,
      "grad_norm": 0.4947514694208445,
      "learning_rate": 5.716001740034371e-07,
      "loss": 0.020732149481773376,
      "step": 4280,
      "token_acc": 0.9931953006862859
    },
    {
      "epoch": 2.69811320754717,
      "grad_norm": 0.4971731498284735,
      "learning_rate": 5.487212163550848e-07,
      "loss": 0.02166786640882492,
      "step": 4290,
      "token_acc": 0.9930354872269102
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 0.3896129637224116,
      "learning_rate": 5.262966190875551e-07,
      "loss": 0.023386237025260926,
      "step": 4300,
      "token_acc": 0.9923059705668401
    },
    {
      "epoch": 2.710691823899371,
      "grad_norm": 0.6445372165794033,
      "learning_rate": 5.043274602399939e-07,
      "loss": 0.019444110989570617,
      "step": 4310,
      "token_acc": 0.9937604110787995
    },
    {
      "epoch": 2.7169811320754715,
      "grad_norm": 0.8571882173501956,
      "learning_rate": 4.828147959568252e-07,
      "loss": 0.0215284526348114,
      "step": 4320,
      "token_acc": 0.9930044517125466
    },
    {
      "epoch": 2.7232704402515724,
      "grad_norm": 0.5971479826797512,
      "learning_rate": 4.617596604369734e-07,
      "loss": 0.02067880630493164,
      "step": 4330,
      "token_acc": 0.9930786753110216
    },
    {
      "epoch": 2.729559748427673,
      "grad_norm": 0.6192693288729766,
      "learning_rate": 4.411630658841415e-07,
      "loss": 0.02310236692428589,
      "step": 4340,
      "token_acc": 0.9923919146546883
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 0.5854199322009673,
      "learning_rate": 4.2102600245815473e-07,
      "loss": 0.020319417119026184,
      "step": 4350,
      "token_acc": 0.9934170636639238
    },
    {
      "epoch": 2.742138364779874,
      "grad_norm": 0.5060972177534601,
      "learning_rate": 4.0134943822735863e-07,
      "loss": 0.020632094144821166,
      "step": 4360,
      "token_acc": 0.9927046064400715
    },
    {
      "epoch": 2.748427672955975,
      "grad_norm": 0.6535832360286701,
      "learning_rate": 3.8213431912207613e-07,
      "loss": 0.02077610194683075,
      "step": 4370,
      "token_acc": 0.9931016862544711
    },
    {
      "epoch": 2.7547169811320753,
      "grad_norm": 0.5769076059018868,
      "learning_rate": 3.633815688891418e-07,
      "loss": 0.021984952688217162,
      "step": 4380,
      "token_acc": 0.9925909231766065
    },
    {
      "epoch": 2.761006289308176,
      "grad_norm": 0.6414019547548331,
      "learning_rate": 3.450920890474829e-07,
      "loss": 0.022785688936710357,
      "step": 4390,
      "token_acc": 0.9923757698977457
    },
    {
      "epoch": 2.767295597484277,
      "grad_norm": 0.5779301518428668,
      "learning_rate": 3.272667588447931e-07,
      "loss": 0.019667960703372955,
      "step": 4400,
      "token_acc": 0.993188854489164
    },
    {
      "epoch": 2.7735849056603774,
      "grad_norm": 0.5661347233531627,
      "learning_rate": 3.099064352152459e-07,
      "loss": 0.02017345428466797,
      "step": 4410,
      "token_acc": 0.9930435268480751
    },
    {
      "epoch": 2.779874213836478,
      "grad_norm": 0.5682142833519405,
      "learning_rate": 2.930119527383146e-07,
      "loss": 0.0193926140666008,
      "step": 4420,
      "token_acc": 0.9935703370527348
    },
    {
      "epoch": 2.7861635220125787,
      "grad_norm": 0.6129756155590131,
      "learning_rate": 2.7658412359864327e-07,
      "loss": 0.021811990439891814,
      "step": 4430,
      "token_acc": 0.9932381008137697
    },
    {
      "epoch": 2.7924528301886795,
      "grad_norm": 0.5869125647929773,
      "learning_rate": 2.606237375470033e-07,
      "loss": 0.02384023666381836,
      "step": 4440,
      "token_acc": 0.9924499811249529
    },
    {
      "epoch": 2.79874213836478,
      "grad_norm": 0.7772183292502256,
      "learning_rate": 2.4513156186232514e-07,
      "loss": 0.02438468337059021,
      "step": 4450,
      "token_acc": 0.9916802895843064
    },
    {
      "epoch": 2.8050314465408803,
      "grad_norm": 0.6624999458168028,
      "learning_rate": 2.301083413148142e-07,
      "loss": 0.020420333743095397,
      "step": 4460,
      "token_acc": 0.9931096958383661
    },
    {
      "epoch": 2.811320754716981,
      "grad_norm": 0.43571255257652386,
      "learning_rate": 2.1555479813014757e-07,
      "loss": 0.019135478138923644,
      "step": 4470,
      "token_acc": 0.9937029124030136
    },
    {
      "epoch": 2.817610062893082,
      "grad_norm": 0.5935762625557985,
      "learning_rate": 2.0147163195475161e-07,
      "loss": 0.02100401222705841,
      "step": 4480,
      "token_acc": 0.9933416910433208
    },
    {
      "epoch": 2.8238993710691824,
      "grad_norm": 0.5788237631502353,
      "learning_rate": 1.8785951982216445e-07,
      "loss": 0.021649350225925446,
      "step": 4490,
      "token_acc": 0.9924480074358081
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.8397049647839784,
      "learning_rate": 1.7471911612050085e-07,
      "loss": 0.01883925497531891,
      "step": 4500,
      "token_acc": 0.9934104471280296
    },
    {
      "epoch": 2.830188679245283,
      "eval_loss": 0.05329994112253189,
      "eval_runtime": 710.6066,
      "eval_samples_per_second": 11.926,
      "eval_steps_per_second": 2.982,
      "eval_token_acc": 0.9857121501158562,
      "step": 4500
    }
  ],
  "logging_steps": 10,
  "max_steps": 4770,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 320035629735936.0,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}