{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 3,
  "global_step": 282,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.031914893617021274,
      "grad_norm": 219.1436309814453,
      "learning_rate": 4.687722050568071e-06,
      "loss": 1.6437,
      "step": 3
    },
    {
      "epoch": 0.031914893617021274,
      "eval_loss": 1.2238304615020752,
      "eval_runtime": 49.6348,
      "eval_samples_per_second": 6.044,
      "eval_steps_per_second": 0.201,
      "step": 3
    },
    {
      "epoch": 0.06382978723404255,
      "grad_norm": 151.3193817138672,
      "learning_rate": 4.637316437121103e-06,
      "loss": 1.5519,
      "step": 6
    },
    {
      "epoch": 0.06382978723404255,
      "eval_loss": 1.166587471961975,
      "eval_runtime": 43.6496,
      "eval_samples_per_second": 6.873,
      "eval_steps_per_second": 0.229,
      "step": 6
    },
    {
      "epoch": 0.09574468085106383,
      "grad_norm": 151.81333923339844,
      "learning_rate": 4.586910823674134e-06,
      "loss": 1.1065,
      "step": 9
    },
    {
      "epoch": 0.09574468085106383,
      "eval_loss": 1.167779803276062,
      "eval_runtime": 43.8309,
      "eval_samples_per_second": 6.844,
      "eval_steps_per_second": 0.228,
      "step": 9
    },
    {
      "epoch": 0.1276595744680851,
      "grad_norm": 216.54078674316406,
      "learning_rate": 4.536505210227166e-06,
      "loss": 1.0103,
      "step": 12
    },
    {
      "epoch": 0.1276595744680851,
      "eval_loss": 1.1300101280212402,
      "eval_runtime": 42.8188,
      "eval_samples_per_second": 7.006,
      "eval_steps_per_second": 0.234,
      "step": 12
    },
    {
      "epoch": 0.1595744680851064,
      "grad_norm": 96.97032928466797,
      "learning_rate": 4.486099596780197e-06,
      "loss": 1.1391,
      "step": 15
    },
    {
      "epoch": 0.1595744680851064,
      "eval_loss": 1.093359351158142,
      "eval_runtime": 43.7071,
      "eval_samples_per_second": 6.864,
      "eval_steps_per_second": 0.229,
      "step": 15
    },
    {
      "epoch": 0.19148936170212766,
      "grad_norm": 93.6313247680664,
      "learning_rate": 4.435693983333229e-06,
      "loss": 1.1749,
      "step": 18
    },
    {
      "epoch": 0.19148936170212766,
      "eval_loss": 1.0837420225143433,
      "eval_runtime": 42.6236,
      "eval_samples_per_second": 7.038,
      "eval_steps_per_second": 0.235,
      "step": 18
    },
    {
      "epoch": 0.22340425531914893,
      "grad_norm": 128.5612030029297,
      "learning_rate": 4.38528836988626e-06,
      "loss": 1.0878,
      "step": 21
    },
    {
      "epoch": 0.22340425531914893,
      "eval_loss": 1.0713236331939697,
      "eval_runtime": 43.1057,
      "eval_samples_per_second": 6.96,
      "eval_steps_per_second": 0.232,
      "step": 21
    },
    {
      "epoch": 0.2553191489361702,
      "grad_norm": 106.38141632080078,
      "learning_rate": 4.334882756439292e-06,
      "loss": 0.8829,
      "step": 24
    },
    {
      "epoch": 0.2553191489361702,
      "eval_loss": 1.0457205772399902,
      "eval_runtime": 43.4262,
      "eval_samples_per_second": 6.908,
      "eval_steps_per_second": 0.23,
      "step": 24
    },
    {
      "epoch": 0.2872340425531915,
      "grad_norm": 109.02174377441406,
      "learning_rate": 4.284477142992323e-06,
      "loss": 1.3004,
      "step": 27
    },
    {
      "epoch": 0.2872340425531915,
      "eval_loss": 1.015425205230713,
      "eval_runtime": 42.8448,
      "eval_samples_per_second": 7.002,
      "eval_steps_per_second": 0.233,
      "step": 27
    },
    {
      "epoch": 0.3191489361702128,
      "grad_norm": 93.8167495727539,
      "learning_rate": 4.234071529545355e-06,
      "loss": 1.0768,
      "step": 30
    },
    {
      "epoch": 0.3191489361702128,
      "eval_loss": 0.9899512529373169,
      "eval_runtime": 62.0193,
      "eval_samples_per_second": 4.837,
      "eval_steps_per_second": 0.161,
      "step": 30
    },
    {
      "epoch": 0.35106382978723405,
      "grad_norm": 117.0296401977539,
      "learning_rate": 4.183665916098386e-06,
      "loss": 0.8761,
      "step": 33
    },
    {
      "epoch": 0.35106382978723405,
      "eval_loss": 0.9733409881591797,
      "eval_runtime": 43.6195,
      "eval_samples_per_second": 6.878,
      "eval_steps_per_second": 0.229,
      "step": 33
    },
    {
      "epoch": 0.3829787234042553,
      "grad_norm": 97.06668090820312,
      "learning_rate": 4.1332603026514175e-06,
      "loss": 1.0648,
      "step": 36
    },
    {
      "epoch": 0.3829787234042553,
      "eval_loss": 0.9584157466888428,
      "eval_runtime": 43.1964,
      "eval_samples_per_second": 6.945,
      "eval_steps_per_second": 0.232,
      "step": 36
    },
    {
      "epoch": 0.4148936170212766,
      "grad_norm": 69.0722427368164,
      "learning_rate": 4.082854689204449e-06,
      "loss": 0.8425,
      "step": 39
    },
    {
      "epoch": 0.4148936170212766,
      "eval_loss": 0.9429272413253784,
      "eval_runtime": 43.4379,
      "eval_samples_per_second": 6.906,
      "eval_steps_per_second": 0.23,
      "step": 39
    },
    {
      "epoch": 0.44680851063829785,
      "grad_norm": 72.62019348144531,
      "learning_rate": 4.032449075757481e-06,
      "loss": 0.8714,
      "step": 42
    },
    {
      "epoch": 0.44680851063829785,
      "eval_loss": 0.9335389137268066,
      "eval_runtime": 42.713,
      "eval_samples_per_second": 7.024,
      "eval_steps_per_second": 0.234,
      "step": 42
    },
    {
      "epoch": 0.4787234042553192,
      "grad_norm": 91.91283416748047,
      "learning_rate": 3.982043462310513e-06,
      "loss": 0.9101,
      "step": 45
    },
    {
      "epoch": 0.4787234042553192,
      "eval_loss": 0.9222455024719238,
      "eval_runtime": 43.5273,
      "eval_samples_per_second": 6.892,
      "eval_steps_per_second": 0.23,
      "step": 45
    },
    {
      "epoch": 0.5106382978723404,
      "grad_norm": 71.0954360961914,
      "learning_rate": 3.931637848863544e-06,
      "loss": 1.0336,
      "step": 48
    },
    {
      "epoch": 0.5106382978723404,
      "eval_loss": 0.9134926199913025,
      "eval_runtime": 46.5111,
      "eval_samples_per_second": 6.45,
      "eval_steps_per_second": 0.215,
      "step": 48
    },
    {
      "epoch": 0.5425531914893617,
      "grad_norm": 92.62516021728516,
      "learning_rate": 3.881232235416575e-06,
      "loss": 0.941,
      "step": 51
    },
    {
      "epoch": 0.5425531914893617,
      "eval_loss": 0.9049246311187744,
      "eval_runtime": 43.2332,
      "eval_samples_per_second": 6.939,
      "eval_steps_per_second": 0.231,
      "step": 51
    },
    {
      "epoch": 0.574468085106383,
      "grad_norm": 87.83678436279297,
      "learning_rate": 3.830826621969607e-06,
      "loss": 1.1561,
      "step": 54
    },
    {
      "epoch": 0.574468085106383,
      "eval_loss": 0.9015933275222778,
      "eval_runtime": 43.95,
      "eval_samples_per_second": 6.826,
      "eval_steps_per_second": 0.228,
      "step": 54
    },
    {
      "epoch": 0.6063829787234043,
      "grad_norm": 81.34359741210938,
      "learning_rate": 3.780421008522638e-06,
      "loss": 0.9756,
      "step": 57
    },
    {
      "epoch": 0.6063829787234043,
      "eval_loss": 0.8971786499023438,
      "eval_runtime": 43.2014,
      "eval_samples_per_second": 6.944,
      "eval_steps_per_second": 0.231,
      "step": 57
    },
    {
      "epoch": 0.6382978723404256,
      "grad_norm": 64.89433288574219,
      "learning_rate": 3.73001539507567e-06,
      "loss": 0.749,
      "step": 60
    },
    {
      "epoch": 0.6382978723404256,
      "eval_loss": 0.8940334916114807,
      "eval_runtime": 50.073,
      "eval_samples_per_second": 5.991,
      "eval_steps_per_second": 0.2,
      "step": 60
    },
    {
      "epoch": 0.6702127659574468,
      "grad_norm": 92.25138854980469,
      "learning_rate": 3.679609781628701e-06,
      "loss": 0.9315,
      "step": 63
    },
    {
      "epoch": 0.6702127659574468,
      "eval_loss": 0.8884549736976624,
      "eval_runtime": 46.5461,
      "eval_samples_per_second": 6.445,
      "eval_steps_per_second": 0.215,
      "step": 63
    },
    {
      "epoch": 0.7021276595744681,
      "grad_norm": 86.6369400024414,
      "learning_rate": 3.629204168181733e-06,
      "loss": 1.1571,
      "step": 66
    },
    {
      "epoch": 0.7021276595744681,
      "eval_loss": 0.883185625076294,
      "eval_runtime": 43.0575,
      "eval_samples_per_second": 6.967,
      "eval_steps_per_second": 0.232,
      "step": 66
    },
    {
      "epoch": 0.7340425531914894,
      "grad_norm": 87.74919128417969,
      "learning_rate": 3.5787985547347642e-06,
      "loss": 1.3094,
      "step": 69
    },
    {
      "epoch": 0.7340425531914894,
      "eval_loss": 0.8721677660942078,
      "eval_runtime": 43.1156,
      "eval_samples_per_second": 6.958,
      "eval_steps_per_second": 0.232,
      "step": 69
    },
    {
      "epoch": 0.7659574468085106,
      "grad_norm": 81.59742736816406,
      "learning_rate": 3.5283929412877956e-06,
      "loss": 1.0689,
      "step": 72
    },
    {
      "epoch": 0.7659574468085106,
      "eval_loss": 0.8642645478248596,
      "eval_runtime": 43.5775,
      "eval_samples_per_second": 6.884,
      "eval_steps_per_second": 0.229,
      "step": 72
    },
    {
      "epoch": 0.7978723404255319,
      "grad_norm": 81.44770050048828,
      "learning_rate": 3.477987327840827e-06,
      "loss": 0.8599,
      "step": 75
    },
    {
      "epoch": 0.7978723404255319,
      "eval_loss": 0.8605104684829712,
      "eval_runtime": 43.731,
      "eval_samples_per_second": 6.86,
      "eval_steps_per_second": 0.229,
      "step": 75
    },
    {
      "epoch": 0.8297872340425532,
      "grad_norm": 74.29925537109375,
      "learning_rate": 3.427581714393859e-06,
      "loss": 1.0591,
      "step": 78
    },
    {
      "epoch": 0.8297872340425532,
      "eval_loss": 0.8572540283203125,
      "eval_runtime": 43.0132,
      "eval_samples_per_second": 6.975,
      "eval_steps_per_second": 0.232,
      "step": 78
    },
    {
      "epoch": 0.8617021276595744,
      "grad_norm": 80.95857238769531,
      "learning_rate": 3.37717610094689e-06,
      "loss": 0.8628,
      "step": 81
    },
    {
      "epoch": 0.8617021276595744,
      "eval_loss": 0.854705810546875,
      "eval_runtime": 43.5041,
      "eval_samples_per_second": 6.896,
      "eval_steps_per_second": 0.23,
      "step": 81
    },
    {
      "epoch": 0.8936170212765957,
      "grad_norm": 73.22078704833984,
      "learning_rate": 3.3267704874999216e-06,
      "loss": 0.8449,
      "step": 84
    },
    {
      "epoch": 0.8936170212765957,
      "eval_loss": 0.8551430106163025,
      "eval_runtime": 43.3103,
      "eval_samples_per_second": 6.927,
      "eval_steps_per_second": 0.231,
      "step": 84
    },
    {
      "epoch": 0.925531914893617,
      "grad_norm": 81.8653564453125,
      "learning_rate": 3.2763648740529535e-06,
      "loss": 0.8565,
      "step": 87
    },
    {
      "epoch": 0.925531914893617,
      "eval_loss": 0.8558651804924011,
      "eval_runtime": 43.4899,
      "eval_samples_per_second": 6.898,
      "eval_steps_per_second": 0.23,
      "step": 87
    },
    {
      "epoch": 0.9574468085106383,
      "grad_norm": 97.51951599121094,
      "learning_rate": 3.2259592606059844e-06,
      "loss": 0.8928,
      "step": 90
    },
    {
      "epoch": 0.9574468085106383,
      "eval_loss": 0.8567631244659424,
      "eval_runtime": 43.4864,
      "eval_samples_per_second": 6.899,
      "eval_steps_per_second": 0.23,
      "step": 90
    },
    {
      "epoch": 0.9893617021276596,
      "grad_norm": 60.12514114379883,
      "learning_rate": 3.1755536471590163e-06,
      "loss": 0.7314,
      "step": 93
    },
    {
      "epoch": 0.9893617021276596,
      "eval_loss": 0.8565493822097778,
      "eval_runtime": 22.6087,
      "eval_samples_per_second": 13.269,
      "eval_steps_per_second": 0.442,
      "step": 93
    },
    {
      "epoch": 1.0212765957446808,
      "grad_norm": 57.255775451660156,
      "learning_rate": 3.1251480337120472e-06,
      "loss": 0.7931,
      "step": 96
    },
    {
      "epoch": 1.0212765957446808,
      "eval_loss": 0.8531547784805298,
      "eval_runtime": 23.9019,
      "eval_samples_per_second": 12.551,
      "eval_steps_per_second": 0.418,
      "step": 96
    },
    {
      "epoch": 1.053191489361702,
      "grad_norm": 43.00119400024414,
      "learning_rate": 3.074742420265079e-06,
      "loss": 0.3847,
      "step": 99
    },
    {
      "epoch": 1.053191489361702,
      "eval_loss": 0.84489506483078,
      "eval_runtime": 23.0798,
      "eval_samples_per_second": 12.998,
      "eval_steps_per_second": 0.433,
      "step": 99
    },
    {
      "epoch": 1.0851063829787233,
      "grad_norm": 21.35872459411621,
      "learning_rate": 3.024336806818111e-06,
      "loss": 0.252,
      "step": 102
    },
    {
      "epoch": 1.0851063829787233,
      "eval_loss": 0.8377478122711182,
      "eval_runtime": 23.6666,
      "eval_samples_per_second": 12.676,
      "eval_steps_per_second": 0.423,
      "step": 102
    },
    {
      "epoch": 1.1170212765957448,
      "grad_norm": 42.683170318603516,
      "learning_rate": 2.973931193371142e-06,
      "loss": 0.2581,
      "step": 105
    },
    {
      "epoch": 1.1170212765957448,
      "eval_loss": 0.832615315914154,
      "eval_runtime": 22.1568,
      "eval_samples_per_second": 13.54,
      "eval_steps_per_second": 0.451,
      "step": 105
    },
    {
      "epoch": 1.148936170212766,
      "grad_norm": 36.07487487792969,
      "learning_rate": 2.9235255799241737e-06,
      "loss": 0.272,
      "step": 108
    },
    {
      "epoch": 1.148936170212766,
      "eval_loss": 0.8290462493896484,
      "eval_runtime": 23.4484,
      "eval_samples_per_second": 12.794,
      "eval_steps_per_second": 0.426,
      "step": 108
    },
    {
      "epoch": 1.1808510638297873,
      "grad_norm": 25.169939041137695,
      "learning_rate": 2.873119966477205e-06,
      "loss": 0.3822,
      "step": 111
    },
    {
      "epoch": 1.1808510638297873,
      "eval_loss": 0.8273450136184692,
      "eval_runtime": 22.2076,
      "eval_samples_per_second": 13.509,
      "eval_steps_per_second": 0.45,
      "step": 111
    },
    {
      "epoch": 1.2127659574468086,
      "grad_norm": 43.500083923339844,
      "learning_rate": 2.8227143530302365e-06,
      "loss": 0.3555,
      "step": 114
    },
    {
      "epoch": 1.2127659574468086,
      "eval_loss": 0.825173020362854,
      "eval_runtime": 23.9364,
      "eval_samples_per_second": 12.533,
      "eval_steps_per_second": 0.418,
      "step": 114
    },
    {
      "epoch": 1.2446808510638299,
      "grad_norm": 47.19260787963867,
      "learning_rate": 2.772308739583268e-06,
      "loss": 0.2687,
      "step": 117
    },
    {
      "epoch": 1.2446808510638299,
      "eval_loss": 0.8255329132080078,
      "eval_runtime": 22.128,
      "eval_samples_per_second": 13.558,
      "eval_steps_per_second": 0.452,
      "step": 117
    },
    {
      "epoch": 1.2765957446808511,
      "grad_norm": 47.747982025146484,
      "learning_rate": 2.7219031261362997e-06,
      "loss": 0.2614,
      "step": 120
    },
    {
      "epoch": 1.2765957446808511,
      "eval_loss": 0.8268000483512878,
      "eval_runtime": 23.832,
      "eval_samples_per_second": 12.588,
      "eval_steps_per_second": 0.42,
      "step": 120
    },
    {
      "epoch": 1.3085106382978724,
      "grad_norm": 52.21699142456055,
      "learning_rate": 2.671497512689331e-06,
      "loss": 0.3726,
      "step": 123
    },
    {
      "epoch": 1.3085106382978724,
      "eval_loss": 0.8281159400939941,
      "eval_runtime": 22.7023,
      "eval_samples_per_second": 13.215,
      "eval_steps_per_second": 0.44,
      "step": 123
    },
    {
      "epoch": 1.3404255319148937,
      "grad_norm": 43.84182357788086,
      "learning_rate": 2.6210918992423625e-06,
      "loss": 0.2081,
      "step": 126
    },
    {
      "epoch": 1.3404255319148937,
      "eval_loss": 0.8298842310905457,
      "eval_runtime": 23.5811,
      "eval_samples_per_second": 12.722,
      "eval_steps_per_second": 0.424,
      "step": 126
    },
    {
      "epoch": 1.372340425531915,
      "grad_norm": 35.47488021850586,
      "learning_rate": 2.570686285795394e-06,
      "loss": 0.2363,
      "step": 129
    },
    {
      "epoch": 1.372340425531915,
      "eval_loss": 0.8358010053634644,
      "eval_runtime": 22.1111,
      "eval_samples_per_second": 13.568,
      "eval_steps_per_second": 0.452,
      "step": 129
    },
    {
      "epoch": 1.4042553191489362,
      "grad_norm": 47.503395080566406,
      "learning_rate": 2.5202806723484253e-06,
      "loss": 0.3497,
      "step": 132
    },
    {
      "epoch": 1.4042553191489362,
      "eval_loss": 0.8382316827774048,
      "eval_runtime": 23.6168,
      "eval_samples_per_second": 12.703,
      "eval_steps_per_second": 0.423,
      "step": 132
    },
    {
      "epoch": 1.4361702127659575,
      "grad_norm": 43.59933853149414,
      "learning_rate": 2.469875058901457e-06,
      "loss": 0.2927,
      "step": 135
    },
    {
      "epoch": 1.4361702127659575,
      "eval_loss": 0.8399573564529419,
      "eval_runtime": 22.1313,
      "eval_samples_per_second": 13.555,
      "eval_steps_per_second": 0.452,
      "step": 135
    },
    {
      "epoch": 1.4680851063829787,
      "grad_norm": 46.75539779663086,
      "learning_rate": 2.419469445454488e-06,
      "loss": 0.3142,
      "step": 138
    },
    {
      "epoch": 1.4680851063829787,
      "eval_loss": 0.841060996055603,
      "eval_runtime": 24.0437,
      "eval_samples_per_second": 12.477,
      "eval_steps_per_second": 0.416,
      "step": 138
    },
    {
      "epoch": 1.5,
      "grad_norm": 47.08267593383789,
      "learning_rate": 2.36906383200752e-06,
      "loss": 0.2303,
      "step": 141
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.8422343730926514,
      "eval_runtime": 21.8892,
      "eval_samples_per_second": 13.705,
      "eval_steps_per_second": 0.457,
      "step": 141
    },
    {
      "epoch": 1.5319148936170213,
      "grad_norm": 59.37859344482422,
      "learning_rate": 2.3186582185605513e-06,
      "loss": 0.2628,
      "step": 144
    },
    {
      "epoch": 1.5319148936170213,
      "eval_loss": 0.843423068523407,
      "eval_runtime": 23.5581,
      "eval_samples_per_second": 12.735,
      "eval_steps_per_second": 0.424,
      "step": 144
    },
    {
      "epoch": 1.5638297872340425,
      "grad_norm": 66.62377166748047,
      "learning_rate": 2.268252605113583e-06,
      "loss": 0.2734,
      "step": 147
    },
    {
      "epoch": 1.5638297872340425,
      "eval_loss": 0.8458875417709351,
      "eval_runtime": 22.7309,
      "eval_samples_per_second": 13.198,
      "eval_steps_per_second": 0.44,
      "step": 147
    },
    {
      "epoch": 1.5957446808510638,
      "grad_norm": 65.94501495361328,
      "learning_rate": 2.2178469916666146e-06,
      "loss": 0.2989,
      "step": 150
    },
    {
      "epoch": 1.5957446808510638,
      "eval_loss": 0.8469556570053101,
      "eval_runtime": 23.6111,
      "eval_samples_per_second": 12.706,
      "eval_steps_per_second": 0.424,
      "step": 150
    },
    {
      "epoch": 1.627659574468085,
      "grad_norm": 99.2269287109375,
      "learning_rate": 2.167441378219646e-06,
      "loss": 0.3895,
      "step": 153
    },
    {
      "epoch": 1.627659574468085,
      "eval_loss": 0.8473237752914429,
      "eval_runtime": 22.2962,
      "eval_samples_per_second": 13.455,
      "eval_steps_per_second": 0.449,
      "step": 153
    },
    {
      "epoch": 1.6595744680851063,
      "grad_norm": 45.91320037841797,
      "learning_rate": 2.1170357647726774e-06,
      "loss": 0.3273,
      "step": 156
    },
    {
      "epoch": 1.6595744680851063,
      "eval_loss": 0.8420724868774414,
      "eval_runtime": 24.0122,
      "eval_samples_per_second": 12.494,
      "eval_steps_per_second": 0.416,
      "step": 156
    },
    {
      "epoch": 1.6914893617021276,
      "grad_norm": 52.12400817871094,
      "learning_rate": 2.0666301513257088e-06,
      "loss": 0.2508,
      "step": 159
    },
    {
      "epoch": 1.6914893617021276,
      "eval_loss": 0.8369884490966797,
      "eval_runtime": 22.131,
      "eval_samples_per_second": 13.556,
      "eval_steps_per_second": 0.452,
      "step": 159
    },
    {
      "epoch": 1.7234042553191489,
      "grad_norm": 59.420372009277344,
      "learning_rate": 2.0162245378787406e-06,
      "loss": 0.2143,
      "step": 162
    },
    {
      "epoch": 1.7234042553191489,
      "eval_loss": 0.8358878493309021,
      "eval_runtime": 23.6554,
      "eval_samples_per_second": 12.682,
      "eval_steps_per_second": 0.423,
      "step": 162
    },
    {
      "epoch": 1.7553191489361701,
      "grad_norm": 53.903350830078125,
      "learning_rate": 1.965818924431772e-06,
      "loss": 0.2337,
      "step": 165
    },
    {
      "epoch": 1.7553191489361701,
      "eval_loss": 0.8342965841293335,
      "eval_runtime": 22.1781,
      "eval_samples_per_second": 13.527,
      "eval_steps_per_second": 0.451,
      "step": 165
    },
    {
      "epoch": 1.7872340425531914,
      "grad_norm": 37.46884536743164,
      "learning_rate": 1.9154133109848034e-06,
      "loss": 0.1867,
      "step": 168
    },
    {
      "epoch": 1.7872340425531914,
      "eval_loss": 0.8319395184516907,
      "eval_runtime": 23.6724,
      "eval_samples_per_second": 12.673,
      "eval_steps_per_second": 0.422,
      "step": 168
    },
    {
      "epoch": 1.8191489361702127,
      "grad_norm": 73.817138671875,
      "learning_rate": 1.865007697537835e-06,
      "loss": 0.3836,
      "step": 171
    },
    {
      "epoch": 1.8191489361702127,
      "eval_loss": 0.8294512033462524,
      "eval_runtime": 22.7496,
      "eval_samples_per_second": 13.187,
      "eval_steps_per_second": 0.44,
      "step": 171
    },
    {
      "epoch": 1.851063829787234,
      "grad_norm": 66.78804016113281,
      "learning_rate": 1.8146020840908664e-06,
      "loss": 0.2939,
      "step": 174
    },
    {
      "epoch": 1.851063829787234,
      "eval_loss": 0.8301381468772888,
      "eval_runtime": 23.7921,
      "eval_samples_per_second": 12.609,
      "eval_steps_per_second": 0.42,
      "step": 174
    },
    {
      "epoch": 1.8829787234042552,
      "grad_norm": 27.3295841217041,
      "learning_rate": 1.7641964706438978e-06,
      "loss": 0.2702,
      "step": 177
    },
    {
      "epoch": 1.8829787234042552,
      "eval_loss": 0.8287379741668701,
      "eval_runtime": 22.3208,
      "eval_samples_per_second": 13.44,
      "eval_steps_per_second": 0.448,
      "step": 177
    },
    {
      "epoch": 1.9148936170212765,
      "grad_norm": 34.200496673583984,
      "learning_rate": 1.7137908571969294e-06,
      "loss": 0.1949,
      "step": 180
    },
    {
      "epoch": 1.9148936170212765,
      "eval_loss": 0.827938437461853,
      "eval_runtime": 23.7637,
      "eval_samples_per_second": 12.624,
      "eval_steps_per_second": 0.421,
      "step": 180
    },
    {
      "epoch": 1.9468085106382977,
      "grad_norm": 57.093162536621094,
      "learning_rate": 1.6633852437499608e-06,
      "loss": 0.2186,
      "step": 183
    },
    {
      "epoch": 1.9468085106382977,
      "eval_loss": 0.8297010064125061,
      "eval_runtime": 22.4348,
      "eval_samples_per_second": 13.372,
      "eval_steps_per_second": 0.446,
      "step": 183
    },
    {
      "epoch": 1.978723404255319,
      "grad_norm": 54.119667053222656,
      "learning_rate": 1.6129796303029922e-06,
      "loss": 0.2609,
      "step": 186
    },
    {
      "epoch": 1.978723404255319,
      "eval_loss": 0.8327953219413757,
      "eval_runtime": 24.1101,
      "eval_samples_per_second": 12.443,
      "eval_steps_per_second": 0.415,
      "step": 186
    },
    {
      "epoch": 2.0106382978723403,
      "grad_norm": 34.58624267578125,
      "learning_rate": 1.5625740168560236e-06,
      "loss": 0.1952,
      "step": 189
    },
    {
      "epoch": 2.0106382978723403,
      "eval_loss": 0.8341763615608215,
      "eval_runtime": 45.5245,
      "eval_samples_per_second": 6.59,
      "eval_steps_per_second": 0.22,
      "step": 189
    },
    {
      "epoch": 2.0425531914893615,
      "grad_norm": 35.64149856567383,
      "learning_rate": 1.5121684034090554e-06,
      "loss": 0.0993,
      "step": 192
    },
    {
      "epoch": 2.0425531914893615,
      "eval_loss": 0.837165355682373,
      "eval_runtime": 13.0218,
      "eval_samples_per_second": 23.038,
      "eval_steps_per_second": 0.768,
      "step": 192
    },
    {
      "epoch": 2.074468085106383,
      "grad_norm": 50.38720703125,
      "learning_rate": 1.4617627899620868e-06,
      "loss": 0.1409,
      "step": 195
    },
    {
      "epoch": 2.074468085106383,
      "eval_loss": 0.8386108875274658,
      "eval_runtime": 32.929,
      "eval_samples_per_second": 9.11,
      "eval_steps_per_second": 0.304,
      "step": 195
    },
    {
      "epoch": 2.106382978723404,
      "grad_norm": 28.053754806518555,
      "learning_rate": 1.4113571765151182e-06,
      "loss": 0.0522,
      "step": 198
    },
    {
      "epoch": 2.106382978723404,
      "eval_loss": 0.8413487076759338,
      "eval_runtime": 13.422,
      "eval_samples_per_second": 22.351,
      "eval_steps_per_second": 0.745,
      "step": 198
    },
    {
      "epoch": 2.1382978723404253,
      "grad_norm": 48.99291229248047,
      "learning_rate": 1.3609515630681499e-06,
      "loss": 0.123,
      "step": 201
    },
    {
      "epoch": 2.1382978723404253,
      "eval_loss": 0.8437539339065552,
      "eval_runtime": 32.9133,
      "eval_samples_per_second": 9.115,
      "eval_steps_per_second": 0.304,
      "step": 201
    },
    {
      "epoch": 2.1702127659574466,
      "grad_norm": 30.540855407714844,
      "learning_rate": 1.3105459496211813e-06,
      "loss": 0.0946,
      "step": 204
    },
    {
      "epoch": 2.1702127659574466,
      "eval_loss": 0.8469984531402588,
      "eval_runtime": 12.8212,
      "eval_samples_per_second": 23.399,
      "eval_steps_per_second": 0.78,
      "step": 204
    },
    {
      "epoch": 2.202127659574468,
      "grad_norm": 65.50384521484375,
      "learning_rate": 1.2601403361742127e-06,
      "loss": 0.2346,
      "step": 207
    },
    {
      "epoch": 2.202127659574468,
      "eval_loss": 0.8491196036338806,
      "eval_runtime": 33.0116,
      "eval_samples_per_second": 9.088,
      "eval_steps_per_second": 0.303,
      "step": 207
    },
    {
      "epoch": 2.2340425531914896,
      "grad_norm": 37.9770622253418,
      "learning_rate": 1.209734722727244e-06,
      "loss": 0.1445,
      "step": 210
    },
    {
      "epoch": 2.2340425531914896,
      "eval_loss": 0.8509886860847473,
      "eval_runtime": 12.9443,
      "eval_samples_per_second": 23.176,
      "eval_steps_per_second": 0.773,
      "step": 210
    },
    {
      "epoch": 2.2659574468085104,
      "grad_norm": 34.36880111694336,
      "learning_rate": 1.1593291092802757e-06,
      "loss": 0.1112,
      "step": 213
    },
    {
      "epoch": 2.2659574468085104,
      "eval_loss": 0.8533729910850525,
      "eval_runtime": 33.1181,
      "eval_samples_per_second": 9.058,
      "eval_steps_per_second": 0.302,
      "step": 213
    },
    {
      "epoch": 2.297872340425532,
      "grad_norm": 45.28392028808594,
      "learning_rate": 1.1089234958333073e-06,
      "loss": 0.113,
      "step": 216
    },
    {
      "epoch": 2.297872340425532,
      "eval_loss": 0.8565312623977661,
      "eval_runtime": 12.9159,
      "eval_samples_per_second": 23.227,
      "eval_steps_per_second": 0.774,
      "step": 216
    },
    {
      "epoch": 2.329787234042553,
      "grad_norm": 24.961444854736328,
      "learning_rate": 1.0585178823863387e-06,
      "loss": 0.0929,
      "step": 219
    },
    {
      "epoch": 2.329787234042553,
      "eval_loss": 0.859375,
      "eval_runtime": 32.6853,
      "eval_samples_per_second": 9.178,
      "eval_steps_per_second": 0.306,
      "step": 219
    },
    {
      "epoch": 2.3617021276595747,
      "grad_norm": 22.85188865661621,
      "learning_rate": 1.0081122689393703e-06,
      "loss": 0.1108,
      "step": 222
    },
    {
      "epoch": 2.3617021276595747,
      "eval_loss": 0.8612162470817566,
      "eval_runtime": 13.8881,
      "eval_samples_per_second": 21.601,
      "eval_steps_per_second": 0.72,
      "step": 222
    },
    {
      "epoch": 2.393617021276596,
      "grad_norm": 20.546833038330078,
      "learning_rate": 9.577066554924017e-07,
      "loss": 0.092,
      "step": 225
    },
    {
      "epoch": 2.393617021276596,
      "eval_loss": 0.8626397848129272,
      "eval_runtime": 32.9393,
      "eval_samples_per_second": 9.108,
      "eval_steps_per_second": 0.304,
      "step": 225
    },
    {
      "epoch": 2.425531914893617,
      "grad_norm": 36.019561767578125,
      "learning_rate": 9.073010420454332e-07,
      "loss": 0.0641,
      "step": 228
    },
    {
      "epoch": 2.425531914893617,
      "eval_loss": 0.8644223809242249,
      "eval_runtime": 12.991,
      "eval_samples_per_second": 23.093,
      "eval_steps_per_second": 0.77,
      "step": 228
    },
    {
      "epoch": 2.4574468085106385,
      "grad_norm": 15.907193183898926,
      "learning_rate": 8.568954285984647e-07,
      "loss": 0.0437,
      "step": 231
    },
    {
      "epoch": 2.4574468085106385,
      "eval_loss": 0.8656131029129028,
      "eval_runtime": 33.0386,
      "eval_samples_per_second": 9.08,
      "eval_steps_per_second": 0.303,
      "step": 231
    },
    {
      "epoch": 2.4893617021276597,
      "grad_norm": 30.25502586364746,
      "learning_rate": 8.064898151514961e-07,
      "loss": 0.097,
      "step": 234
    },
    {
      "epoch": 2.4893617021276597,
      "eval_loss": 0.8662587404251099,
      "eval_runtime": 12.9969,
      "eval_samples_per_second": 23.082,
      "eval_steps_per_second": 0.769,
      "step": 234
    },
    {
      "epoch": 2.521276595744681,
      "grad_norm": 34.13824462890625,
      "learning_rate": 7.560842017045277e-07,
      "loss": 0.1028,
      "step": 237
    },
    {
      "epoch": 2.521276595744681,
      "eval_loss": 0.8667454123497009,
      "eval_runtime": 33.0579,
      "eval_samples_per_second": 9.075,
      "eval_steps_per_second": 0.302,
      "step": 237
    },
    {
      "epoch": 2.5531914893617023,
      "grad_norm": 47.30521011352539,
      "learning_rate": 7.056785882575591e-07,
      "loss": 0.1503,
      "step": 240
    },
    {
      "epoch": 2.5531914893617023,
      "eval_loss": 0.8677744269371033,
      "eval_runtime": 12.8958,
      "eval_samples_per_second": 23.263,
      "eval_steps_per_second": 0.775,
      "step": 240
    },
    {
      "epoch": 2.5851063829787235,
      "grad_norm": 40.989444732666016,
      "learning_rate": 6.552729748105906e-07,
      "loss": 0.1286,
      "step": 243
    },
    {
      "epoch": 2.5851063829787235,
      "eval_loss": 0.8686262965202332,
      "eval_runtime": 33.1611,
      "eval_samples_per_second": 9.047,
      "eval_steps_per_second": 0.302,
      "step": 243
    },
    {
      "epoch": 2.617021276595745,
      "grad_norm": 58.63330841064453,
      "learning_rate": 6.04867361363622e-07,
      "loss": 0.1938,
      "step": 246
    },
    {
      "epoch": 2.617021276595745,
      "eval_loss": 0.8695897459983826,
      "eval_runtime": 13.2572,
      "eval_samples_per_second": 22.629,
      "eval_steps_per_second": 0.754,
      "step": 246
    },
    {
      "epoch": 2.648936170212766,
      "grad_norm": 25.59177017211914,
      "learning_rate": 5.544617479166536e-07,
      "loss": 0.0655,
      "step": 249
    },
    {
      "epoch": 2.648936170212766,
      "eval_loss": 0.8699460029602051,
      "eval_runtime": 33.2452,
      "eval_samples_per_second": 9.024,
      "eval_steps_per_second": 0.301,
      "step": 249
    },
    {
      "epoch": 2.6808510638297873,
      "grad_norm": 10.76020622253418,
      "learning_rate": 5.040561344696851e-07,
      "loss": 0.0638,
      "step": 252
    },
    {
      "epoch": 2.6808510638297873,
      "eval_loss": 0.8698618412017822,
      "eval_runtime": 13.4624,
      "eval_samples_per_second": 22.284,
      "eval_steps_per_second": 0.743,
      "step": 252
    },
    {
      "epoch": 2.7127659574468086,
      "grad_norm": 14.23372745513916,
      "learning_rate": 4.536505210227166e-07,
      "loss": 0.04,
      "step": 255
    },
    {
      "epoch": 2.7127659574468086,
      "eval_loss": 0.8700897097587585,
      "eval_runtime": 33.0408,
      "eval_samples_per_second": 9.08,
      "eval_steps_per_second": 0.303,
      "step": 255
    },
    {
      "epoch": 2.74468085106383,
      "grad_norm": 54.65508270263672,
      "learning_rate": 4.0324490757574806e-07,
      "loss": 0.158,
      "step": 258
    },
    {
      "epoch": 2.74468085106383,
      "eval_loss": 0.8706519603729248,
      "eval_runtime": 13.0406,
      "eval_samples_per_second": 23.005,
      "eval_steps_per_second": 0.767,
      "step": 258
    },
    {
      "epoch": 2.776595744680851,
      "grad_norm": 36.574066162109375,
      "learning_rate": 3.5283929412877956e-07,
      "loss": 0.0951,
      "step": 261
    },
    {
      "epoch": 2.776595744680851,
      "eval_loss": 0.871029794216156,
      "eval_runtime": 32.8987,
      "eval_samples_per_second": 9.119,
      "eval_steps_per_second": 0.304,
      "step": 261
    },
    {
      "epoch": 2.8085106382978724,
      "grad_norm": 32.9112434387207,
      "learning_rate": 3.02433680681811e-07,
      "loss": 0.0788,
      "step": 264
    },
    {
      "epoch": 2.8085106382978724,
      "eval_loss": 0.871411144733429,
      "eval_runtime": 12.964,
      "eval_samples_per_second": 23.141,
      "eval_steps_per_second": 0.771,
      "step": 264
    },
    {
      "epoch": 2.8404255319148937,
      "grad_norm": 26.687891006469727,
      "learning_rate": 2.520280672348426e-07,
      "loss": 0.0429,
      "step": 267
    },
    {
      "epoch": 2.8404255319148937,
      "eval_loss": 0.8717812299728394,
      "eval_runtime": 32.9027,
      "eval_samples_per_second": 9.118,
      "eval_steps_per_second": 0.304,
      "step": 267
    },
    {
      "epoch": 2.872340425531915,
      "grad_norm": 25.479427337646484,
      "learning_rate": 2.0162245378787403e-07,
      "loss": 0.0809,
      "step": 270
    },
    {
      "epoch": 2.872340425531915,
      "eval_loss": 0.8716564178466797,
      "eval_runtime": 12.9971,
      "eval_samples_per_second": 23.082,
      "eval_steps_per_second": 0.769,
      "step": 270
    },
    {
      "epoch": 2.904255319148936,
      "grad_norm": 14.731446266174316,
      "learning_rate": 1.512168403409055e-07,
      "loss": 0.0942,
      "step": 273
    },
    {
      "epoch": 2.904255319148936,
      "eval_loss": 0.8718379735946655,
      "eval_runtime": 33.0136,
      "eval_samples_per_second": 9.087,
      "eval_steps_per_second": 0.303,
      "step": 273
    },
    {
      "epoch": 2.9361702127659575,
      "grad_norm": 23.23249053955078,
      "learning_rate": 1.0081122689393701e-07,
      "loss": 0.1419,
      "step": 276
    },
    {
      "epoch": 2.9361702127659575,
      "eval_loss": 0.8720559477806091,
      "eval_runtime": 13.6301,
      "eval_samples_per_second": 22.01,
      "eval_steps_per_second": 0.734,
      "step": 276
    },
    {
      "epoch": 2.9680851063829787,
      "grad_norm": 42.01313400268555,
      "learning_rate": 5.040561344696851e-08,
      "loss": 0.1036,
      "step": 279
    },
    {
      "epoch": 2.9680851063829787,
      "eval_loss": 0.8721166253089905,
      "eval_runtime": 33.0448,
      "eval_samples_per_second": 9.079,
      "eval_steps_per_second": 0.303,
      "step": 279
    },
    {
      "epoch": 3.0,
      "grad_norm": 37.830604553222656,
      "learning_rate": 0.0,
      "loss": 0.0854,
      "step": 282
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.8720741868019104,
      "eval_runtime": 12.8271,
      "eval_samples_per_second": 23.388,
      "eval_steps_per_second": 0.78,
      "step": 282
    }
  ],
  "logging_steps": 3,
  "max_steps": 282,
  "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": 523328480700102.0,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": {
    "_wandb": {},
    "assignments": {},
    "decay": 0.01,
    "learning_rate": 4.73812766401504e-06,
    "metric": "eval/loss",
    "per_device_train_batch_size": 32
  }
}