{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 108,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "entropy": 1.7206335216760635,
      "epoch": 0.027777777777777776,
      "grad_norm": 12.239052772521973,
      "learning_rate": 0.0,
      "loss": 2.2445,
      "mean_token_accuracy": 0.5234357602894306,
      "num_tokens": 3149.0,
      "step": 1
    },
    {
      "entropy": 1.7157433032989502,
      "epoch": 0.05555555555555555,
      "grad_norm": 12.409416198730469,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 2.2591,
      "mean_token_accuracy": 0.5426237694919109,
      "num_tokens": 6352.0,
      "step": 2
    },
    {
      "entropy": 1.7281817346811295,
      "epoch": 0.08333333333333333,
      "grad_norm": 12.136215209960938,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 2.2025,
      "mean_token_accuracy": 0.5330252423882484,
      "num_tokens": 9480.0,
      "step": 3
    },
    {
      "entropy": 1.7453971356153488,
      "epoch": 0.1111111111111111,
      "grad_norm": 12.494486808776855,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.2858,
      "mean_token_accuracy": 0.5153909176588058,
      "num_tokens": 12696.0,
      "step": 4
    },
    {
      "entropy": 1.7059053778648376,
      "epoch": 0.1388888888888889,
      "grad_norm": 12.915605545043945,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.2393,
      "mean_token_accuracy": 0.5214959308505058,
      "num_tokens": 15845.0,
      "step": 5
    },
    {
      "entropy": 1.7419584095478058,
      "epoch": 0.16666666666666666,
      "grad_norm": 11.967181205749512,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.2269,
      "mean_token_accuracy": 0.5283601805567741,
      "num_tokens": 18953.0,
      "step": 6
    },
    {
      "entropy": 1.7141934782266617,
      "epoch": 0.19444444444444445,
      "grad_norm": 11.107477188110352,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 2.1667,
      "mean_token_accuracy": 0.5402880758047104,
      "num_tokens": 22106.0,
      "step": 7
    },
    {
      "entropy": 1.7215740531682968,
      "epoch": 0.2222222222222222,
      "grad_norm": 10.586934089660645,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 2.1178,
      "mean_token_accuracy": 0.558077298104763,
      "num_tokens": 25320.0,
      "step": 8
    },
    {
      "entropy": 1.7224711626768112,
      "epoch": 0.25,
      "grad_norm": 10.22936725616455,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 1.9993,
      "mean_token_accuracy": 0.5679031684994698,
      "num_tokens": 28514.0,
      "step": 9
    },
    {
      "entropy": 1.7291602343320847,
      "epoch": 0.2777777777777778,
      "grad_norm": 10.116068840026855,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 2.002,
      "mean_token_accuracy": 0.5621381998062134,
      "num_tokens": 31645.0,
      "step": 10
    },
    {
      "entropy": 1.823133796453476,
      "epoch": 0.3055555555555556,
      "grad_norm": 8.050676345825195,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.8115,
      "mean_token_accuracy": 0.6078527420759201,
      "num_tokens": 34837.0,
      "step": 11
    },
    {
      "entropy": 1.8218258619308472,
      "epoch": 0.3333333333333333,
      "grad_norm": 8.03921127319336,
      "learning_rate": 2.2e-06,
      "loss": 1.7858,
      "mean_token_accuracy": 0.5981088206171989,
      "num_tokens": 37987.0,
      "step": 12
    },
    {
      "entropy": 1.7872179597616196,
      "epoch": 0.3611111111111111,
      "grad_norm": 7.6152801513671875,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 1.6719,
      "mean_token_accuracy": 0.6255820393562317,
      "num_tokens": 41170.0,
      "step": 13
    },
    {
      "entropy": 1.6922450810670853,
      "epoch": 0.3888888888888889,
      "grad_norm": 6.331522464752197,
      "learning_rate": 2.6e-06,
      "loss": 1.4919,
      "mean_token_accuracy": 0.6602419465780258,
      "num_tokens": 44250.0,
      "step": 14
    },
    {
      "entropy": 1.6007178127765656,
      "epoch": 0.4166666666666667,
      "grad_norm": 6.234885215759277,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 1.3792,
      "mean_token_accuracy": 0.6762521862983704,
      "num_tokens": 47446.0,
      "step": 15
    },
    {
      "entropy": 1.5817099809646606,
      "epoch": 0.4444444444444444,
      "grad_norm": 6.2869696617126465,
      "learning_rate": 3e-06,
      "loss": 1.3805,
      "mean_token_accuracy": 0.6645312234759331,
      "num_tokens": 50759.0,
      "step": 16
    },
    {
      "entropy": 1.412723958492279,
      "epoch": 0.4722222222222222,
      "grad_norm": 5.045506000518799,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 1.2176,
      "mean_token_accuracy": 0.7046914845705032,
      "num_tokens": 53926.0,
      "step": 17
    },
    {
      "entropy": 1.3380708545446396,
      "epoch": 0.5,
      "grad_norm": 4.068710803985596,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 1.2425,
      "mean_token_accuracy": 0.6927981376647949,
      "num_tokens": 57077.0,
      "step": 18
    },
    {
      "entropy": 1.2170532494783401,
      "epoch": 0.5277777777777778,
      "grad_norm": 3.499069929122925,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 1.189,
      "mean_token_accuracy": 0.705949991941452,
      "num_tokens": 60181.0,
      "step": 19
    },
    {
      "entropy": 1.2205703258514404,
      "epoch": 0.5555555555555556,
      "grad_norm": 22.419281005859375,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 1.2191,
      "mean_token_accuracy": 0.6920858323574066,
      "num_tokens": 63319.0,
      "step": 20
    },
    {
      "entropy": 1.132231056690216,
      "epoch": 0.5833333333333334,
      "grad_norm": 3.432077646255493,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.1323,
      "mean_token_accuracy": 0.7145579382777214,
      "num_tokens": 66437.0,
      "step": 21
    },
    {
      "entropy": 1.1085182875394821,
      "epoch": 0.6111111111111112,
      "grad_norm": 3.328181505203247,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 1.118,
      "mean_token_accuracy": 0.7220780923962593,
      "num_tokens": 69581.0,
      "step": 22
    },
    {
      "entropy": 1.1160272061824799,
      "epoch": 0.6388888888888888,
      "grad_norm": 2.7953662872314453,
      "learning_rate": 4.4e-06,
      "loss": 1.062,
      "mean_token_accuracy": 0.7369436100125313,
      "num_tokens": 72724.0,
      "step": 23
    },
    {
      "entropy": 1.1081153899431229,
      "epoch": 0.6666666666666666,
      "grad_norm": 2.8708362579345703,
      "learning_rate": 4.600000000000001e-06,
      "loss": 1.0555,
      "mean_token_accuracy": 0.7323780283331871,
      "num_tokens": 75861.0,
      "step": 24
    },
    {
      "entropy": 1.0733779296278954,
      "epoch": 0.6944444444444444,
      "grad_norm": 2.657761335372925,
      "learning_rate": 4.800000000000001e-06,
      "loss": 1.0127,
      "mean_token_accuracy": 0.7391822040081024,
      "num_tokens": 79168.0,
      "step": 25
    },
    {
      "entropy": 1.0697464495897293,
      "epoch": 0.7222222222222222,
      "grad_norm": 2.6338393688201904,
      "learning_rate": 5e-06,
      "loss": 1.0236,
      "mean_token_accuracy": 0.725149542093277,
      "num_tokens": 82378.0,
      "step": 26
    },
    {
      "entropy": 1.0295503363013268,
      "epoch": 0.75,
      "grad_norm": 2.6113545894622803,
      "learning_rate": 5.2e-06,
      "loss": 1.0343,
      "mean_token_accuracy": 0.7344670742750168,
      "num_tokens": 85511.0,
      "step": 27
    },
    {
      "entropy": 0.9814292415976524,
      "epoch": 0.7777777777777778,
      "grad_norm": 2.581925868988037,
      "learning_rate": 5.400000000000001e-06,
      "loss": 0.9725,
      "mean_token_accuracy": 0.7381312251091003,
      "num_tokens": 88660.0,
      "step": 28
    },
    {
      "entropy": 1.007811225950718,
      "epoch": 0.8055555555555556,
      "grad_norm": 2.579724073410034,
      "learning_rate": 5.600000000000001e-06,
      "loss": 1.0091,
      "mean_token_accuracy": 0.7310391962528229,
      "num_tokens": 91875.0,
      "step": 29
    },
    {
      "entropy": 0.9555631652474403,
      "epoch": 0.8333333333333334,
      "grad_norm": 2.651214599609375,
      "learning_rate": 5.8e-06,
      "loss": 0.9463,
      "mean_token_accuracy": 0.749499000608921,
      "num_tokens": 95039.0,
      "step": 30
    },
    {
      "entropy": 1.0038460940122604,
      "epoch": 0.8611111111111112,
      "grad_norm": 2.505073070526123,
      "learning_rate": 6e-06,
      "loss": 1.0151,
      "mean_token_accuracy": 0.7300070971250534,
      "num_tokens": 98259.0,
      "step": 31
    },
    {
      "entropy": 1.0190542414784431,
      "epoch": 0.8888888888888888,
      "grad_norm": 2.6049623489379883,
      "learning_rate": 6.200000000000001e-06,
      "loss": 0.9595,
      "mean_token_accuracy": 0.7381655722856522,
      "num_tokens": 101389.0,
      "step": 32
    },
    {
      "entropy": 1.0043294206261635,
      "epoch": 0.9166666666666666,
      "grad_norm": 2.5841352939605713,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.9548,
      "mean_token_accuracy": 0.7301696166396141,
      "num_tokens": 104591.0,
      "step": 33
    },
    {
      "entropy": 1.0013394877314568,
      "epoch": 0.9444444444444444,
      "grad_norm": 2.4673030376434326,
      "learning_rate": 6.600000000000001e-06,
      "loss": 0.9206,
      "mean_token_accuracy": 0.7463522851467133,
      "num_tokens": 107798.0,
      "step": 34
    },
    {
      "entropy": 0.9427972882986069,
      "epoch": 0.9722222222222222,
      "grad_norm": 2.594034433364868,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.8944,
      "mean_token_accuracy": 0.7576184198260307,
      "num_tokens": 110869.0,
      "step": 35
    },
    {
      "entropy": 0.9821532517671585,
      "epoch": 1.0,
      "grad_norm": 2.5436978340148926,
      "learning_rate": 7e-06,
      "loss": 0.9404,
      "mean_token_accuracy": 0.7452764138579369,
      "num_tokens": 112970.0,
      "step": 36
    },
    {
      "entropy": 0.948033794760704,
      "epoch": 1.0277777777777777,
      "grad_norm": 2.457702159881592,
      "learning_rate": 7.2000000000000005e-06,
      "loss": 0.818,
      "mean_token_accuracy": 0.7714106887578964,
      "num_tokens": 116138.0,
      "step": 37
    },
    {
      "entropy": 0.849656954407692,
      "epoch": 1.0555555555555556,
      "grad_norm": 2.256528377532959,
      "learning_rate": 7.4e-06,
      "loss": 0.7358,
      "mean_token_accuracy": 0.797737255692482,
      "num_tokens": 119378.0,
      "step": 38
    },
    {
      "entropy": 0.8578782826662064,
      "epoch": 1.0833333333333333,
      "grad_norm": 2.557046413421631,
      "learning_rate": 7.600000000000001e-06,
      "loss": 0.8029,
      "mean_token_accuracy": 0.7780278623104095,
      "num_tokens": 122490.0,
      "step": 39
    },
    {
      "entropy": 0.769713781774044,
      "epoch": 1.1111111111111112,
      "grad_norm": 2.5916965007781982,
      "learning_rate": 7.800000000000002e-06,
      "loss": 0.7244,
      "mean_token_accuracy": 0.7930369228124619,
      "num_tokens": 125646.0,
      "step": 40
    },
    {
      "entropy": 0.7162968814373016,
      "epoch": 1.1388888888888888,
      "grad_norm": 2.44321608543396,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.6747,
      "mean_token_accuracy": 0.8100233748555183,
      "num_tokens": 128738.0,
      "step": 41
    },
    {
      "entropy": 0.7322863712906837,
      "epoch": 1.1666666666666667,
      "grad_norm": 2.5348217487335205,
      "learning_rate": 8.2e-06,
      "loss": 0.7154,
      "mean_token_accuracy": 0.7932721599936485,
      "num_tokens": 131900.0,
      "step": 42
    },
    {
      "entropy": 0.7626946941018105,
      "epoch": 1.1944444444444444,
      "grad_norm": 2.7120513916015625,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.7576,
      "mean_token_accuracy": 0.7777007296681404,
      "num_tokens": 135078.0,
      "step": 43
    },
    {
      "entropy": 0.8006442189216614,
      "epoch": 1.2222222222222223,
      "grad_norm": 2.7017619609832764,
      "learning_rate": 8.6e-06,
      "loss": 0.7101,
      "mean_token_accuracy": 0.7920609414577484,
      "num_tokens": 138176.0,
      "step": 44
    },
    {
      "entropy": 0.7950561568140984,
      "epoch": 1.25,
      "grad_norm": 2.666325807571411,
      "learning_rate": 8.8e-06,
      "loss": 0.7427,
      "mean_token_accuracy": 0.8027379214763641,
      "num_tokens": 141349.0,
      "step": 45
    },
    {
      "entropy": 0.793344184756279,
      "epoch": 1.2777777777777777,
      "grad_norm": 2.776358127593994,
      "learning_rate": 9e-06,
      "loss": 0.6878,
      "mean_token_accuracy": 0.8111961334943771,
      "num_tokens": 144461.0,
      "step": 46
    },
    {
      "entropy": 0.7685084566473961,
      "epoch": 1.3055555555555556,
      "grad_norm": 2.5511598587036133,
      "learning_rate": 9.200000000000002e-06,
      "loss": 0.6777,
      "mean_token_accuracy": 0.8126928880810738,
      "num_tokens": 147642.0,
      "step": 47
    },
    {
      "entropy": 0.796379953622818,
      "epoch": 1.3333333333333333,
      "grad_norm": 2.5649895668029785,
      "learning_rate": 9.4e-06,
      "loss": 0.7652,
      "mean_token_accuracy": 0.7886124402284622,
      "num_tokens": 150894.0,
      "step": 48
    },
    {
      "entropy": 0.7704698145389557,
      "epoch": 1.3611111111111112,
      "grad_norm": 2.559164047241211,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.7427,
      "mean_token_accuracy": 0.7878011912107468,
      "num_tokens": 154057.0,
      "step": 49
    },
    {
      "entropy": 0.6923560723662376,
      "epoch": 1.3888888888888888,
      "grad_norm": 2.321601390838623,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.7077,
      "mean_token_accuracy": 0.804372176527977,
      "num_tokens": 157317.0,
      "step": 50
    },
    {
      "entropy": 0.7212042436003685,
      "epoch": 1.4166666666666667,
      "grad_norm": 2.2511632442474365,
      "learning_rate": 1e-05,
      "loss": 0.6909,
      "mean_token_accuracy": 0.799708366394043,
      "num_tokens": 160493.0,
      "step": 51
    },
    {
      "entropy": 0.6932398602366447,
      "epoch": 1.4444444444444444,
      "grad_norm": 2.192110776901245,
      "learning_rate": 9.99266706925562e-06,
      "loss": 0.6846,
      "mean_token_accuracy": 0.7934979721903801,
      "num_tokens": 163690.0,
      "step": 52
    },
    {
      "entropy": 0.7240612953901291,
      "epoch": 1.4722222222222223,
      "grad_norm": 2.0434861183166504,
      "learning_rate": 9.970689785771798e-06,
      "loss": 0.6793,
      "mean_token_accuracy": 0.8052623942494392,
      "num_tokens": 166925.0,
      "step": 53
    },
    {
      "entropy": 0.7377813756465912,
      "epoch": 1.5,
      "grad_norm": 1.978843092918396,
      "learning_rate": 9.934132612707631e-06,
      "loss": 0.6852,
      "mean_token_accuracy": 0.8033023402094841,
      "num_tokens": 170012.0,
      "step": 54
    },
    {
      "entropy": 0.706544779241085,
      "epoch": 1.5277777777777777,
      "grad_norm": 1.989449381828308,
      "learning_rate": 9.883102778550434e-06,
      "loss": 0.65,
      "mean_token_accuracy": 0.8123279586434364,
      "num_tokens": 173174.0,
      "step": 55
    },
    {
      "entropy": 0.689531534910202,
      "epoch": 1.5555555555555556,
      "grad_norm": 2.1900088787078857,
      "learning_rate": 9.817749962596115e-06,
      "loss": 0.7314,
      "mean_token_accuracy": 0.7898781001567841,
      "num_tokens": 176217.0,
      "step": 56
    },
    {
      "entropy": 0.7628979608416557,
      "epoch": 1.5833333333333335,
      "grad_norm": 2.2131078243255615,
      "learning_rate": 9.738265855914014e-06,
      "loss": 0.7844,
      "mean_token_accuracy": 0.7743321061134338,
      "num_tokens": 179345.0,
      "step": 57
    },
    {
      "entropy": 0.7197903394699097,
      "epoch": 1.6111111111111112,
      "grad_norm": 2.043686628341675,
      "learning_rate": 9.644883599083959e-06,
      "loss": 0.7147,
      "mean_token_accuracy": 0.7910747677087784,
      "num_tokens": 182534.0,
      "step": 58
    },
    {
      "entropy": 0.7264939621090889,
      "epoch": 1.6388888888888888,
      "grad_norm": 1.9586451053619385,
      "learning_rate": 9.537877098354787e-06,
      "loss": 0.6769,
      "mean_token_accuracy": 0.8056130409240723,
      "num_tokens": 185775.0,
      "step": 59
    },
    {
      "entropy": 0.7570189833641052,
      "epoch": 1.6666666666666665,
      "grad_norm": 2.0687789916992188,
      "learning_rate": 9.417560222230115e-06,
      "loss": 0.7804,
      "mean_token_accuracy": 0.7776985839009285,
      "num_tokens": 188983.0,
      "step": 60
    },
    {
      "entropy": 0.716331735253334,
      "epoch": 1.6944444444444444,
      "grad_norm": 1.974429965019226,
      "learning_rate": 9.284285880837947e-06,
      "loss": 0.7131,
      "mean_token_accuracy": 0.8018393665552139,
      "num_tokens": 192190.0,
      "step": 61
    },
    {
      "entropy": 0.7596073672175407,
      "epoch": 1.7222222222222223,
      "grad_norm": 2.0154805183410645,
      "learning_rate": 9.138444990784455e-06,
      "loss": 0.7203,
      "mean_token_accuracy": 0.799439825117588,
      "num_tokens": 195293.0,
      "step": 62
    },
    {
      "entropy": 0.7313190475106239,
      "epoch": 1.75,
      "grad_norm": 1.892073631286621,
      "learning_rate": 8.98046532852822e-06,
      "loss": 0.6366,
      "mean_token_accuracy": 0.8158354088664055,
      "num_tokens": 198464.0,
      "step": 63
    },
    {
      "entropy": 0.6996584385633469,
      "epoch": 1.7777777777777777,
      "grad_norm": 2.004039764404297,
      "learning_rate": 8.810810275638183e-06,
      "loss": 0.7021,
      "mean_token_accuracy": 0.7982058823108673,
      "num_tokens": 201581.0,
      "step": 64
    },
    {
      "entropy": 0.6911773085594177,
      "epoch": 1.8055555555555556,
      "grad_norm": 2.0379583835601807,
      "learning_rate": 8.629977459615655e-06,
      "loss": 0.6898,
      "mean_token_accuracy": 0.8074537888169289,
      "num_tokens": 204669.0,
      "step": 65
    },
    {
      "entropy": 0.6928622052073479,
      "epoch": 1.8333333333333335,
      "grad_norm": 2.0388221740722656,
      "learning_rate": 8.438497294267117e-06,
      "loss": 0.7297,
      "mean_token_accuracy": 0.7990831881761551,
      "num_tokens": 207828.0,
      "step": 66
    },
    {
      "entropy": 0.6930385380983353,
      "epoch": 1.8611111111111112,
      "grad_norm": 2.110360860824585,
      "learning_rate": 8.23693142390914e-06,
      "loss": 0.7039,
      "mean_token_accuracy": 0.7907350063323975,
      "num_tokens": 211003.0,
      "step": 67
    },
    {
      "entropy": 0.6812680587172508,
      "epoch": 1.8888888888888888,
      "grad_norm": 1.8298258781433105,
      "learning_rate": 8.025871075968828e-06,
      "loss": 0.6573,
      "mean_token_accuracy": 0.805952712893486,
      "num_tokens": 214219.0,
      "step": 68
    },
    {
      "entropy": 0.6855239048600197,
      "epoch": 1.9166666666666665,
      "grad_norm": 1.94973886013031,
      "learning_rate": 7.805935326811913e-06,
      "loss": 0.6789,
      "mean_token_accuracy": 0.8042339980602264,
      "num_tokens": 217381.0,
      "step": 69
    },
    {
      "entropy": 0.7017524242401123,
      "epoch": 1.9444444444444444,
      "grad_norm": 1.844860315322876,
      "learning_rate": 7.57776928588511e-06,
      "loss": 0.6916,
      "mean_token_accuracy": 0.7993432655930519,
      "num_tokens": 220575.0,
      "step": 70
    },
    {
      "entropy": 0.6723021641373634,
      "epoch": 1.9722222222222223,
      "grad_norm": 1.9816311597824097,
      "learning_rate": 7.342042203498952e-06,
      "loss": 0.6899,
      "mean_token_accuracy": 0.8051915913820267,
      "num_tokens": 223785.0,
      "step": 71
    },
    {
      "entropy": 0.7291323319077492,
      "epoch": 2.0,
      "grad_norm": 1.919629454612732,
      "learning_rate": 7.099445507801324e-06,
      "loss": 0.6998,
      "mean_token_accuracy": 0.7905183210968971,
      "num_tokens": 225909.0,
      "step": 72
    },
    {
      "entropy": 0.61961829662323,
      "epoch": 2.0277777777777777,
      "grad_norm": 1.885007381439209,
      "learning_rate": 6.850690776699574e-06,
      "loss": 0.448,
      "mean_token_accuracy": 0.868491642177105,
      "num_tokens": 229208.0,
      "step": 73
    },
    {
      "entropy": 0.6041020900011063,
      "epoch": 2.0555555555555554,
      "grad_norm": 1.7473875284194946,
      "learning_rate": 6.5965076506799e-06,
      "loss": 0.4395,
      "mean_token_accuracy": 0.8738656416535378,
      "num_tokens": 232368.0,
      "step": 74
    },
    {
      "entropy": 0.5625318512320518,
      "epoch": 2.0833333333333335,
      "grad_norm": 1.7840373516082764,
      "learning_rate": 6.337641692646106e-06,
      "loss": 0.4065,
      "mean_token_accuracy": 0.8808101117610931,
      "num_tokens": 235594.0,
      "step": 75
    },
    {
      "entropy": 0.5406039394438267,
      "epoch": 2.111111111111111,
      "grad_norm": 1.6404682397842407,
      "learning_rate": 6.074852201055121e-06,
      "loss": 0.4094,
      "mean_token_accuracy": 0.8790416568517685,
      "num_tokens": 238743.0,
      "step": 76
    },
    {
      "entropy": 0.4563685469329357,
      "epoch": 2.138888888888889,
      "grad_norm": 1.6808091402053833,
      "learning_rate": 5.808909982763825e-06,
      "loss": 0.3968,
      "mean_token_accuracy": 0.8820647746324539,
      "num_tokens": 241937.0,
      "step": 77
    },
    {
      "entropy": 0.41335177794098854,
      "epoch": 2.1666666666666665,
      "grad_norm": 1.589560866355896,
      "learning_rate": 5.540595092119709e-06,
      "loss": 0.3575,
      "mean_token_accuracy": 0.8966728746891022,
      "num_tokens": 245098.0,
      "step": 78
    },
    {
      "entropy": 0.39409590885043144,
      "epoch": 2.1944444444444446,
      "grad_norm": 1.8227696418762207,
      "learning_rate": 5.270694542927089e-06,
      "loss": 0.3687,
      "mean_token_accuracy": 0.8862845078110695,
      "num_tokens": 248263.0,
      "step": 79
    },
    {
      "entropy": 0.3667695038020611,
      "epoch": 2.2222222222222223,
      "grad_norm": 2.003453493118286,
      "learning_rate": 5e-06,
      "loss": 0.3747,
      "mean_token_accuracy": 0.8866413086652756,
      "num_tokens": 251385.0,
      "step": 80
    },
    {
      "entropy": 0.372307363897562,
      "epoch": 2.25,
      "grad_norm": 2.1508708000183105,
      "learning_rate": 4.729305457072913e-06,
      "loss": 0.3925,
      "mean_token_accuracy": 0.8885233998298645,
      "num_tokens": 254521.0,
      "step": 81
    },
    {
      "entropy": 0.3625626228749752,
      "epoch": 2.2777777777777777,
      "grad_norm": 1.9702527523040771,
      "learning_rate": 4.459404907880293e-06,
      "loss": 0.3527,
      "mean_token_accuracy": 0.8954600095748901,
      "num_tokens": 257696.0,
      "step": 82
    },
    {
      "entropy": 0.3951098136603832,
      "epoch": 2.3055555555555554,
      "grad_norm": 2.033423662185669,
      "learning_rate": 4.191090017236177e-06,
      "loss": 0.3693,
      "mean_token_accuracy": 0.8858270421624184,
      "num_tokens": 260926.0,
      "step": 83
    },
    {
      "entropy": 0.362130768597126,
      "epoch": 2.3333333333333335,
      "grad_norm": 2.009316921234131,
      "learning_rate": 3.92514779894488e-06,
      "loss": 0.3742,
      "mean_token_accuracy": 0.8943712711334229,
      "num_tokens": 264056.0,
      "step": 84
    },
    {
      "entropy": 0.4374544806778431,
      "epoch": 2.361111111111111,
      "grad_norm": 2.210038185119629,
      "learning_rate": 3.662358307353897e-06,
      "loss": 0.4179,
      "mean_token_accuracy": 0.8706259727478027,
      "num_tokens": 267312.0,
      "step": 85
    },
    {
      "entropy": 0.42751823738217354,
      "epoch": 2.388888888888889,
      "grad_norm": 2.0217461585998535,
      "learning_rate": 3.403492349320101e-06,
      "loss": 0.3789,
      "mean_token_accuracy": 0.8888283669948578,
      "num_tokens": 270471.0,
      "step": 86
    },
    {
      "entropy": 0.4250110536813736,
      "epoch": 2.4166666666666665,
      "grad_norm": 1.8796360492706299,
      "learning_rate": 3.149309223300428e-06,
      "loss": 0.3811,
      "mean_token_accuracy": 0.8900559470057487,
      "num_tokens": 273631.0,
      "step": 87
    },
    {
      "entropy": 0.43111632391810417,
      "epoch": 2.4444444444444446,
      "grad_norm": 1.7372019290924072,
      "learning_rate": 2.9005544921986774e-06,
      "loss": 0.3647,
      "mean_token_accuracy": 0.8927169218659401,
      "num_tokens": 276837.0,
      "step": 88
    },
    {
      "entropy": 0.4418785385787487,
      "epoch": 2.4722222222222223,
      "grad_norm": 1.9140700101852417,
      "learning_rate": 2.65795779650105e-06,
      "loss": 0.3596,
      "mean_token_accuracy": 0.890064924955368,
      "num_tokens": 280046.0,
      "step": 89
    },
    {
      "entropy": 0.44870807603001595,
      "epoch": 2.5,
      "grad_norm": 1.9130913019180298,
      "learning_rate": 2.422230714114891e-06,
      "loss": 0.403,
      "mean_token_accuracy": 0.8809220641851425,
      "num_tokens": 283146.0,
      "step": 90
    },
    {
      "entropy": 0.43657322600483894,
      "epoch": 2.5277777777777777,
      "grad_norm": 1.7632478475570679,
      "learning_rate": 2.1940646731880887e-06,
      "loss": 0.3667,
      "mean_token_accuracy": 0.8956960290670395,
      "num_tokens": 286244.0,
      "step": 91
    },
    {
      "entropy": 0.43981461971998215,
      "epoch": 2.5555555555555554,
      "grad_norm": 1.8031526803970337,
      "learning_rate": 1.9741289240311757e-06,
      "loss": 0.3486,
      "mean_token_accuracy": 0.8991769030690193,
      "num_tokens": 289470.0,
      "step": 92
    },
    {
      "entropy": 0.46824824810028076,
      "epoch": 2.5833333333333335,
      "grad_norm": 1.8402049541473389,
      "learning_rate": 1.7630685760908623e-06,
      "loss": 0.3879,
      "mean_token_accuracy": 0.8875113874673843,
      "num_tokens": 292592.0,
      "step": 93
    },
    {
      "entropy": 0.43110212683677673,
      "epoch": 2.611111111111111,
      "grad_norm": 1.7616424560546875,
      "learning_rate": 1.561502705732883e-06,
      "loss": 0.3538,
      "mean_token_accuracy": 0.893041767179966,
      "num_tokens": 295627.0,
      "step": 94
    },
    {
      "entropy": 0.4747701808810234,
      "epoch": 2.638888888888889,
      "grad_norm": 1.8937019109725952,
      "learning_rate": 1.370022540384347e-06,
      "loss": 0.396,
      "mean_token_accuracy": 0.8820464089512825,
      "num_tokens": 298854.0,
      "step": 95
    },
    {
      "entropy": 0.42968159168958664,
      "epoch": 2.6666666666666665,
      "grad_norm": 1.7988896369934082,
      "learning_rate": 1.1891897243618184e-06,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8860316947102547,
      "num_tokens": 302047.0,
      "step": 96
    },
    {
      "entropy": 0.41574111208319664,
      "epoch": 2.6944444444444446,
      "grad_norm": 1.8018600940704346,
      "learning_rate": 1.0195346714717813e-06,
      "loss": 0.3355,
      "mean_token_accuracy": 0.9010386317968369,
      "num_tokens": 305164.0,
      "step": 97
    },
    {
      "entropy": 0.44235648587346077,
      "epoch": 2.7222222222222223,
      "grad_norm": 1.782006025314331,
      "learning_rate": 8.615550092155478e-07,
      "loss": 0.3585,
      "mean_token_accuracy": 0.8858170583844185,
      "num_tokens": 308277.0,
      "step": 98
    },
    {
      "entropy": 0.4145592227578163,
      "epoch": 2.75,
      "grad_norm": 1.7571319341659546,
      "learning_rate": 7.157141191620548e-07,
      "loss": 0.3524,
      "mean_token_accuracy": 0.8962656855583191,
      "num_tokens": 311428.0,
      "step": 99
    },
    {
      "entropy": 0.44449520111083984,
      "epoch": 2.7777777777777777,
      "grad_norm": 1.6875940561294556,
      "learning_rate": 5.824397777698859e-07,
      "loss": 0.3692,
      "mean_token_accuracy": 0.8917764648795128,
      "num_tokens": 314558.0,
      "step": 100
    },
    {
      "entropy": 0.42087528854608536,
      "epoch": 2.8055555555555554,
      "grad_norm": 1.6768156290054321,
      "learning_rate": 4.6212290164521554e-07,
      "loss": 0.3383,
      "mean_token_accuracy": 0.895787350833416,
      "num_tokens": 317656.0,
      "step": 101
    },
    {
      "entropy": 0.42996737360954285,
      "epoch": 2.8333333333333335,
      "grad_norm": 1.751430630683899,
      "learning_rate": 3.5511640091604293e-07,
      "loss": 0.3595,
      "mean_token_accuracy": 0.8938910216093063,
      "num_tokens": 320851.0,
      "step": 102
    },
    {
      "entropy": 0.41593609377741814,
      "epoch": 2.861111111111111,
      "grad_norm": 1.8688925504684448,
      "learning_rate": 2.617341440859883e-07,
      "loss": 0.3478,
      "mean_token_accuracy": 0.892216868698597,
      "num_tokens": 324023.0,
      "step": 103
    },
    {
      "entropy": 0.46036434918642044,
      "epoch": 2.888888888888889,
      "grad_norm": 1.765503168106079,
      "learning_rate": 1.8225003740388546e-07,
      "loss": 0.3977,
      "mean_token_accuracy": 0.8861500397324562,
      "num_tokens": 327183.0,
      "step": 104
    },
    {
      "entropy": 0.4100455492734909,
      "epoch": 2.9166666666666665,
      "grad_norm": 1.7134056091308594,
      "learning_rate": 1.1689722144956672e-07,
      "loss": 0.3664,
      "mean_token_accuracy": 0.8897344619035721,
      "num_tokens": 330411.0,
      "step": 105
    },
    {
      "entropy": 0.4413677863776684,
      "epoch": 2.9444444444444446,
      "grad_norm": 1.9713915586471558,
      "learning_rate": 6.58673872923693e-08,
      "loss": 0.3903,
      "mean_token_accuracy": 0.8836557418107986,
      "num_tokens": 333553.0,
      "step": 106
    },
    {
      "entropy": 0.41848642379045486,
      "epoch": 2.9722222222222223,
      "grad_norm": 1.7020299434661865,
      "learning_rate": 2.9310214228202016e-08,
      "loss": 0.3578,
      "mean_token_accuracy": 0.8925286307930946,
      "num_tokens": 336750.0,
      "step": 107
    },
    {
      "entropy": 0.4184332340955734,
      "epoch": 3.0,
      "grad_norm": 1.7416514158248901,
      "learning_rate": 7.332930744380906e-09,
      "loss": 0.3532,
      "mean_token_accuracy": 0.8908478692173958,
      "num_tokens": 338921.0,
      "step": 108
    }
  ],
  "logging_steps": 1,
  "max_steps": 108,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5784577125646336.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}