{
  "best_metric": 0.8989499807357788,
  "best_model_checkpoint": "../../experiments_checkpoints/MAdAiLab/Qwen/Qwen1.5_1.8B_patent/checkpoint-750",
  "epoch": 3.0,
  "eval_steps": 50,
  "global_step": 1173,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "grad_norm": 117.48015594482422,
      "learning_rate": 4.9573742540494466e-06,
      "loss": 3.9523,
      "step": 10
    },
    {
      "epoch": 0.05,
      "grad_norm": 108.45614624023438,
      "learning_rate": 4.914748508098892e-06,
      "loss": 2.4902,
      "step": 20
    },
    {
      "epoch": 0.08,
      "grad_norm": 70.28152465820312,
      "learning_rate": 4.872122762148338e-06,
      "loss": 1.9918,
      "step": 30
    },
    {
      "epoch": 0.1,
      "grad_norm": 79.78052520751953,
      "learning_rate": 4.829497016197784e-06,
      "loss": 1.6672,
      "step": 40
    },
    {
      "epoch": 0.13,
      "grad_norm": 45.827728271484375,
      "learning_rate": 4.7868712702472295e-06,
      "loss": 1.4469,
      "step": 50
    },
    {
      "epoch": 0.13,
      "eval_accuracy": 0.528,
      "eval_f1_macro": 0.3841980536108302,
      "eval_f1_micro": 0.528,
      "eval_loss": 1.3521000146865845,
      "eval_runtime": 12.8673,
      "eval_samples_per_second": 388.581,
      "eval_steps_per_second": 6.14,
      "step": 50
    },
    {
      "epoch": 0.15,
      "grad_norm": 34.13399887084961,
      "learning_rate": 4.744245524296676e-06,
      "loss": 1.3984,
      "step": 60
    },
    {
      "epoch": 0.18,
      "grad_norm": 39.08888626098633,
      "learning_rate": 4.701619778346122e-06,
      "loss": 1.2457,
      "step": 70
    },
    {
      "epoch": 0.2,
      "grad_norm": 70.29698181152344,
      "learning_rate": 4.658994032395567e-06,
      "loss": 1.3391,
      "step": 80
    },
    {
      "epoch": 0.23,
      "grad_norm": 38.475982666015625,
      "learning_rate": 4.616368286445013e-06,
      "loss": 1.2006,
      "step": 90
    },
    {
      "epoch": 0.26,
      "grad_norm": 34.23910903930664,
      "learning_rate": 4.573742540494459e-06,
      "loss": 1.1465,
      "step": 100
    },
    {
      "epoch": 0.26,
      "eval_accuracy": 0.596,
      "eval_f1_macro": 0.4990873064564526,
      "eval_f1_micro": 0.596,
      "eval_loss": 1.1614124774932861,
      "eval_runtime": 12.7706,
      "eval_samples_per_second": 391.525,
      "eval_steps_per_second": 6.186,
      "step": 100
    },
    {
      "epoch": 0.28,
      "grad_norm": 54.537071228027344,
      "learning_rate": 4.531116794543905e-06,
      "loss": 1.1721,
      "step": 110
    },
    {
      "epoch": 0.31,
      "grad_norm": 80.3447494506836,
      "learning_rate": 4.488491048593351e-06,
      "loss": 1.1453,
      "step": 120
    },
    {
      "epoch": 0.33,
      "grad_norm": 37.36294937133789,
      "learning_rate": 4.445865302642797e-06,
      "loss": 1.0252,
      "step": 130
    },
    {
      "epoch": 0.36,
      "grad_norm": 65.64375305175781,
      "learning_rate": 4.403239556692242e-06,
      "loss": 1.0926,
      "step": 140
    },
    {
      "epoch": 0.38,
      "grad_norm": 76.81836700439453,
      "learning_rate": 4.3606138107416884e-06,
      "loss": 1.1717,
      "step": 150
    },
    {
      "epoch": 0.38,
      "eval_accuracy": 0.6286,
      "eval_f1_macro": 0.5523456257544254,
      "eval_f1_micro": 0.6286,
      "eval_loss": 1.0560749769210815,
      "eval_runtime": 13.0362,
      "eval_samples_per_second": 383.546,
      "eval_steps_per_second": 6.06,
      "step": 150
    },
    {
      "epoch": 0.41,
      "grad_norm": 26.359981536865234,
      "learning_rate": 4.317988064791135e-06,
      "loss": 0.9779,
      "step": 160
    },
    {
      "epoch": 0.43,
      "grad_norm": 34.32497024536133,
      "learning_rate": 4.27536231884058e-06,
      "loss": 1.0311,
      "step": 170
    },
    {
      "epoch": 0.46,
      "grad_norm": 38.703914642333984,
      "learning_rate": 4.232736572890026e-06,
      "loss": 0.9883,
      "step": 180
    },
    {
      "epoch": 0.49,
      "grad_norm": 40.25131607055664,
      "learning_rate": 4.190110826939472e-06,
      "loss": 1.0908,
      "step": 190
    },
    {
      "epoch": 0.51,
      "grad_norm": 34.45612335205078,
      "learning_rate": 4.1474850809889175e-06,
      "loss": 0.9861,
      "step": 200
    },
    {
      "epoch": 0.51,
      "eval_accuracy": 0.6682,
      "eval_f1_macro": 0.581301697324148,
      "eval_f1_micro": 0.6682,
      "eval_loss": 0.9591749906539917,
      "eval_runtime": 12.8147,
      "eval_samples_per_second": 390.178,
      "eval_steps_per_second": 6.165,
      "step": 200
    },
    {
      "epoch": 0.54,
      "grad_norm": 48.11906051635742,
      "learning_rate": 4.104859335038364e-06,
      "loss": 1.0355,
      "step": 210
    },
    {
      "epoch": 0.56,
      "grad_norm": 40.41707992553711,
      "learning_rate": 4.06223358908781e-06,
      "loss": 0.968,
      "step": 220
    },
    {
      "epoch": 0.59,
      "grad_norm": 34.4398307800293,
      "learning_rate": 4.019607843137255e-06,
      "loss": 0.9777,
      "step": 230
    },
    {
      "epoch": 0.61,
      "grad_norm": 52.77057647705078,
      "learning_rate": 3.976982097186701e-06,
      "loss": 0.9646,
      "step": 240
    },
    {
      "epoch": 0.64,
      "grad_norm": 37.42899703979492,
      "learning_rate": 3.934356351236147e-06,
      "loss": 0.9701,
      "step": 250
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.6658,
      "eval_f1_macro": 0.5948802716943312,
      "eval_f1_micro": 0.6658,
      "eval_loss": 0.957937479019165,
      "eval_runtime": 12.8026,
      "eval_samples_per_second": 390.547,
      "eval_steps_per_second": 6.171,
      "step": 250
    },
    {
      "epoch": 0.66,
      "grad_norm": 51.38178634643555,
      "learning_rate": 3.891730605285593e-06,
      "loss": 0.9826,
      "step": 260
    },
    {
      "epoch": 0.69,
      "grad_norm": 31.820234298706055,
      "learning_rate": 3.849104859335039e-06,
      "loss": 0.9791,
      "step": 270
    },
    {
      "epoch": 0.72,
      "grad_norm": 39.74692916870117,
      "learning_rate": 3.8064791133844846e-06,
      "loss": 0.9047,
      "step": 280
    },
    {
      "epoch": 0.74,
      "grad_norm": 28.091707229614258,
      "learning_rate": 3.7638533674339308e-06,
      "loss": 0.8908,
      "step": 290
    },
    {
      "epoch": 0.77,
      "grad_norm": 37.695556640625,
      "learning_rate": 3.7212276214833765e-06,
      "loss": 0.9389,
      "step": 300
    },
    {
      "epoch": 0.77,
      "eval_accuracy": 0.679,
      "eval_f1_macro": 0.6287401468652754,
      "eval_f1_micro": 0.679,
      "eval_loss": 0.9363999962806702,
      "eval_runtime": 12.797,
      "eval_samples_per_second": 390.715,
      "eval_steps_per_second": 6.173,
      "step": 300
    },
    {
      "epoch": 0.79,
      "grad_norm": 45.27925109863281,
      "learning_rate": 3.678601875532822e-06,
      "loss": 0.9262,
      "step": 310
    },
    {
      "epoch": 0.82,
      "grad_norm": 33.71603775024414,
      "learning_rate": 3.6359761295822684e-06,
      "loss": 0.9182,
      "step": 320
    },
    {
      "epoch": 0.84,
      "grad_norm": 40.315677642822266,
      "learning_rate": 3.593350383631714e-06,
      "loss": 0.9084,
      "step": 330
    },
    {
      "epoch": 0.87,
      "grad_norm": 38.41544723510742,
      "learning_rate": 3.55072463768116e-06,
      "loss": 1.0266,
      "step": 340
    },
    {
      "epoch": 0.9,
      "grad_norm": 43.862613677978516,
      "learning_rate": 3.508098891730606e-06,
      "loss": 0.9914,
      "step": 350
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.6756,
      "eval_f1_macro": 0.6115341208469629,
      "eval_f1_micro": 0.6756,
      "eval_loss": 0.924637496471405,
      "eval_runtime": 12.8058,
      "eval_samples_per_second": 390.447,
      "eval_steps_per_second": 6.169,
      "step": 350
    },
    {
      "epoch": 0.92,
      "grad_norm": 26.705493927001953,
      "learning_rate": 3.4654731457800517e-06,
      "loss": 0.9863,
      "step": 360
    },
    {
      "epoch": 0.95,
      "grad_norm": 31.23122215270996,
      "learning_rate": 3.4228473998294974e-06,
      "loss": 0.8785,
      "step": 370
    },
    {
      "epoch": 0.97,
      "grad_norm": 17.27379608154297,
      "learning_rate": 3.3802216538789436e-06,
      "loss": 0.8624,
      "step": 380
    },
    {
      "epoch": 1.0,
      "grad_norm": 29.550363540649414,
      "learning_rate": 3.3375959079283893e-06,
      "loss": 0.9793,
      "step": 390
    },
    {
      "epoch": 1.02,
      "grad_norm": 40.75907897949219,
      "learning_rate": 3.294970161977835e-06,
      "loss": 0.7508,
      "step": 400
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.6812,
      "eval_f1_macro": 0.6406138110649251,
      "eval_f1_micro": 0.6812,
      "eval_loss": 0.9046875238418579,
      "eval_runtime": 12.966,
      "eval_samples_per_second": 385.623,
      "eval_steps_per_second": 6.093,
      "step": 400
    },
    {
      "epoch": 1.05,
      "grad_norm": 38.431182861328125,
      "learning_rate": 3.252344416027281e-06,
      "loss": 0.7557,
      "step": 410
    },
    {
      "epoch": 1.07,
      "grad_norm": 20.32223892211914,
      "learning_rate": 3.2097186700767265e-06,
      "loss": 0.7158,
      "step": 420
    },
    {
      "epoch": 1.1,
      "grad_norm": 22.297561645507812,
      "learning_rate": 3.1670929241261722e-06,
      "loss": 0.6851,
      "step": 430
    },
    {
      "epoch": 1.13,
      "grad_norm": 23.710554122924805,
      "learning_rate": 3.124467178175618e-06,
      "loss": 0.6733,
      "step": 440
    },
    {
      "epoch": 1.15,
      "grad_norm": 19.360496520996094,
      "learning_rate": 3.081841432225064e-06,
      "loss": 0.6312,
      "step": 450
    },
    {
      "epoch": 1.15,
      "eval_accuracy": 0.6844,
      "eval_f1_macro": 0.641013130489217,
      "eval_f1_micro": 0.6844,
      "eval_loss": 0.9342374801635742,
      "eval_runtime": 12.8113,
      "eval_samples_per_second": 390.281,
      "eval_steps_per_second": 6.166,
      "step": 450
    },
    {
      "epoch": 1.18,
      "grad_norm": 31.652124404907227,
      "learning_rate": 3.03921568627451e-06,
      "loss": 0.7273,
      "step": 460
    },
    {
      "epoch": 1.2,
      "grad_norm": 27.453479766845703,
      "learning_rate": 2.9965899403239555e-06,
      "loss": 0.6845,
      "step": 470
    },
    {
      "epoch": 1.23,
      "grad_norm": 31.358789443969727,
      "learning_rate": 2.9539641943734017e-06,
      "loss": 0.6445,
      "step": 480
    },
    {
      "epoch": 1.25,
      "grad_norm": 32.35221862792969,
      "learning_rate": 2.9113384484228474e-06,
      "loss": 0.6479,
      "step": 490
    },
    {
      "epoch": 1.28,
      "grad_norm": 26.45619773864746,
      "learning_rate": 2.868712702472293e-06,
      "loss": 0.6436,
      "step": 500
    },
    {
      "epoch": 1.28,
      "eval_accuracy": 0.6848,
      "eval_f1_macro": 0.6409585523940933,
      "eval_f1_micro": 0.6848,
      "eval_loss": 0.9463624954223633,
      "eval_runtime": 12.8005,
      "eval_samples_per_second": 390.608,
      "eval_steps_per_second": 6.172,
      "step": 500
    },
    {
      "epoch": 1.3,
      "grad_norm": 29.645923614501953,
      "learning_rate": 2.8260869565217393e-06,
      "loss": 0.7069,
      "step": 510
    },
    {
      "epoch": 1.33,
      "grad_norm": 29.035226821899414,
      "learning_rate": 2.783461210571185e-06,
      "loss": 0.7149,
      "step": 520
    },
    {
      "epoch": 1.36,
      "grad_norm": 35.058223724365234,
      "learning_rate": 2.7408354646206308e-06,
      "loss": 0.628,
      "step": 530
    },
    {
      "epoch": 1.38,
      "grad_norm": 38.28398513793945,
      "learning_rate": 2.698209718670077e-06,
      "loss": 0.6717,
      "step": 540
    },
    {
      "epoch": 1.41,
      "grad_norm": 29.30170440673828,
      "learning_rate": 2.6555839727195226e-06,
      "loss": 0.6429,
      "step": 550
    },
    {
      "epoch": 1.41,
      "eval_accuracy": 0.6846,
      "eval_f1_macro": 0.6298638266651704,
      "eval_f1_micro": 0.6846,
      "eval_loss": 0.9365624785423279,
      "eval_runtime": 12.804,
      "eval_samples_per_second": 390.503,
      "eval_steps_per_second": 6.17,
      "step": 550
    },
    {
      "epoch": 1.43,
      "grad_norm": 42.15592956542969,
      "learning_rate": 2.6129582267689684e-06,
      "loss": 0.6613,
      "step": 560
    },
    {
      "epoch": 1.46,
      "grad_norm": 40.29536819458008,
      "learning_rate": 2.5703324808184145e-06,
      "loss": 0.6224,
      "step": 570
    },
    {
      "epoch": 1.48,
      "grad_norm": 31.332948684692383,
      "learning_rate": 2.5277067348678602e-06,
      "loss": 0.6877,
      "step": 580
    },
    {
      "epoch": 1.51,
      "grad_norm": 29.125625610351562,
      "learning_rate": 2.4850809889173064e-06,
      "loss": 0.5654,
      "step": 590
    },
    {
      "epoch": 1.53,
      "grad_norm": 27.54241180419922,
      "learning_rate": 2.442455242966752e-06,
      "loss": 0.6471,
      "step": 600
    },
    {
      "epoch": 1.53,
      "eval_accuracy": 0.6812,
      "eval_f1_macro": 0.6490292737514075,
      "eval_f1_micro": 0.6812,
      "eval_loss": 0.9346874952316284,
      "eval_runtime": 12.8127,
      "eval_samples_per_second": 390.237,
      "eval_steps_per_second": 6.166,
      "step": 600
    },
    {
      "epoch": 1.56,
      "grad_norm": 45.75101089477539,
      "learning_rate": 2.3998294970161983e-06,
      "loss": 0.6867,
      "step": 610
    },
    {
      "epoch": 1.59,
      "grad_norm": 24.49323844909668,
      "learning_rate": 2.357203751065644e-06,
      "loss": 0.6749,
      "step": 620
    },
    {
      "epoch": 1.61,
      "grad_norm": 36.94871139526367,
      "learning_rate": 2.3145780051150897e-06,
      "loss": 0.7851,
      "step": 630
    },
    {
      "epoch": 1.64,
      "grad_norm": 37.22614288330078,
      "learning_rate": 2.2719522591645355e-06,
      "loss": 0.6278,
      "step": 640
    },
    {
      "epoch": 1.66,
      "grad_norm": 25.08321189880371,
      "learning_rate": 2.229326513213981e-06,
      "loss": 0.7045,
      "step": 650
    },
    {
      "epoch": 1.66,
      "eval_accuracy": 0.6696,
      "eval_f1_macro": 0.6265175426669454,
      "eval_f1_micro": 0.6696,
      "eval_loss": 0.9456750154495239,
      "eval_runtime": 12.9615,
      "eval_samples_per_second": 385.758,
      "eval_steps_per_second": 6.095,
      "step": 650
    },
    {
      "epoch": 1.69,
      "grad_norm": 22.992191314697266,
      "learning_rate": 2.1867007672634273e-06,
      "loss": 0.6695,
      "step": 660
    },
    {
      "epoch": 1.71,
      "grad_norm": 30.21503257751465,
      "learning_rate": 2.144075021312873e-06,
      "loss": 0.7161,
      "step": 670
    },
    {
      "epoch": 1.74,
      "grad_norm": 31.841976165771484,
      "learning_rate": 2.101449275362319e-06,
      "loss": 0.7071,
      "step": 680
    },
    {
      "epoch": 1.76,
      "grad_norm": 23.938678741455078,
      "learning_rate": 2.058823529411765e-06,
      "loss": 0.6467,
      "step": 690
    },
    {
      "epoch": 1.79,
      "grad_norm": 58.5498161315918,
      "learning_rate": 2.0161977834612107e-06,
      "loss": 0.6311,
      "step": 700
    },
    {
      "epoch": 1.79,
      "eval_accuracy": 0.6924,
      "eval_f1_macro": 0.6303167192886554,
      "eval_f1_micro": 0.6924,
      "eval_loss": 0.9205750226974487,
      "eval_runtime": 12.809,
      "eval_samples_per_second": 390.35,
      "eval_steps_per_second": 6.168,
      "step": 700
    },
    {
      "epoch": 1.82,
      "grad_norm": 30.269773483276367,
      "learning_rate": 1.9735720375106564e-06,
      "loss": 0.6146,
      "step": 710
    },
    {
      "epoch": 1.84,
      "grad_norm": 41.80614471435547,
      "learning_rate": 1.9309462915601026e-06,
      "loss": 0.7465,
      "step": 720
    },
    {
      "epoch": 1.87,
      "grad_norm": 25.003660202026367,
      "learning_rate": 1.8883205456095483e-06,
      "loss": 0.6379,
      "step": 730
    },
    {
      "epoch": 1.89,
      "grad_norm": 29.400390625,
      "learning_rate": 1.8456947996589942e-06,
      "loss": 0.7268,
      "step": 740
    },
    {
      "epoch": 1.92,
      "grad_norm": 20.223796844482422,
      "learning_rate": 1.80306905370844e-06,
      "loss": 0.6659,
      "step": 750
    },
    {
      "epoch": 1.92,
      "eval_accuracy": 0.6976,
      "eval_f1_macro": 0.650699169639119,
      "eval_f1_micro": 0.6976,
      "eval_loss": 0.8989499807357788,
      "eval_runtime": 12.7898,
      "eval_samples_per_second": 390.935,
      "eval_steps_per_second": 6.177,
      "step": 750
    },
    {
      "epoch": 1.94,
      "grad_norm": 30.679563522338867,
      "learning_rate": 1.7604433077578859e-06,
      "loss": 0.6621,
      "step": 760
    },
    {
      "epoch": 1.97,
      "grad_norm": 19.45303726196289,
      "learning_rate": 1.7178175618073316e-06,
      "loss": 0.5979,
      "step": 770
    },
    {
      "epoch": 1.99,
      "grad_norm": 30.14684295654297,
      "learning_rate": 1.6751918158567776e-06,
      "loss": 0.653,
      "step": 780
    },
    {
      "epoch": 2.02,
      "grad_norm": 20.023653030395508,
      "learning_rate": 1.6325660699062235e-06,
      "loss": 0.3664,
      "step": 790
    },
    {
      "epoch": 2.05,
      "grad_norm": 22.883848190307617,
      "learning_rate": 1.5899403239556692e-06,
      "loss": 0.2872,
      "step": 800
    },
    {
      "epoch": 2.05,
      "eval_accuracy": 0.6888,
      "eval_f1_macro": 0.6524141804138828,
      "eval_f1_micro": 0.6888,
      "eval_loss": 1.0101499557495117,
      "eval_runtime": 12.9904,
      "eval_samples_per_second": 384.901,
      "eval_steps_per_second": 6.081,
      "step": 800
    },
    {
      "epoch": 2.07,
      "grad_norm": 23.88119888305664,
      "learning_rate": 1.5473145780051152e-06,
      "loss": 0.3037,
      "step": 810
    },
    {
      "epoch": 2.1,
      "grad_norm": 23.18170928955078,
      "learning_rate": 1.504688832054561e-06,
      "loss": 0.2481,
      "step": 820
    },
    {
      "epoch": 2.12,
      "grad_norm": 29.328716278076172,
      "learning_rate": 1.4620630861040068e-06,
      "loss": 0.2706,
      "step": 830
    },
    {
      "epoch": 2.15,
      "grad_norm": 30.443620681762695,
      "learning_rate": 1.4194373401534528e-06,
      "loss": 0.2617,
      "step": 840
    },
    {
      "epoch": 2.17,
      "grad_norm": 22.83024024963379,
      "learning_rate": 1.3768115942028987e-06,
      "loss": 0.2666,
      "step": 850
    },
    {
      "epoch": 2.17,
      "eval_accuracy": 0.6824,
      "eval_f1_macro": 0.638404500149207,
      "eval_f1_micro": 0.6824,
      "eval_loss": 1.1458874940872192,
      "eval_runtime": 12.7965,
      "eval_samples_per_second": 390.732,
      "eval_steps_per_second": 6.174,
      "step": 850
    },
    {
      "epoch": 2.2,
      "grad_norm": 32.89601135253906,
      "learning_rate": 1.3341858482523444e-06,
      "loss": 0.3211,
      "step": 860
    },
    {
      "epoch": 2.23,
      "grad_norm": 33.39738845825195,
      "learning_rate": 1.2915601023017904e-06,
      "loss": 0.2551,
      "step": 870
    },
    {
      "epoch": 2.25,
      "grad_norm": 27.57012939453125,
      "learning_rate": 1.2489343563512363e-06,
      "loss": 0.2599,
      "step": 880
    },
    {
      "epoch": 2.28,
      "grad_norm": 25.224939346313477,
      "learning_rate": 1.206308610400682e-06,
      "loss": 0.2926,
      "step": 890
    },
    {
      "epoch": 2.3,
      "grad_norm": 35.74443817138672,
      "learning_rate": 1.163682864450128e-06,
      "loss": 0.3211,
      "step": 900
    },
    {
      "epoch": 2.3,
      "eval_accuracy": 0.6704,
      "eval_f1_macro": 0.6361511796437018,
      "eval_f1_micro": 0.6704,
      "eval_loss": 1.1165000200271606,
      "eval_runtime": 12.7946,
      "eval_samples_per_second": 390.789,
      "eval_steps_per_second": 6.174,
      "step": 900
    },
    {
      "epoch": 2.33,
      "grad_norm": 51.319583892822266,
      "learning_rate": 1.121057118499574e-06,
      "loss": 0.2633,
      "step": 910
    },
    {
      "epoch": 2.35,
      "grad_norm": 30.21436309814453,
      "learning_rate": 1.0784313725490197e-06,
      "loss": 0.2217,
      "step": 920
    },
    {
      "epoch": 2.38,
      "grad_norm": 28.57087516784668,
      "learning_rate": 1.0358056265984656e-06,
      "loss": 0.1969,
      "step": 930
    },
    {
      "epoch": 2.4,
      "grad_norm": 27.59893798828125,
      "learning_rate": 9.931798806479115e-07,
      "loss": 0.2769,
      "step": 940
    },
    {
      "epoch": 2.43,
      "grad_norm": 30.17582893371582,
      "learning_rate": 9.505541346973573e-07,
      "loss": 0.2831,
      "step": 950
    },
    {
      "epoch": 2.43,
      "eval_accuracy": 0.6698,
      "eval_f1_macro": 0.63603849051412,
      "eval_f1_micro": 0.6698,
      "eval_loss": 1.1721500158309937,
      "eval_runtime": 12.7914,
      "eval_samples_per_second": 390.888,
      "eval_steps_per_second": 6.176,
      "step": 950
    },
    {
      "epoch": 2.46,
      "grad_norm": 25.04727554321289,
      "learning_rate": 9.079283887468031e-07,
      "loss": 0.2756,
      "step": 960
    },
    {
      "epoch": 2.48,
      "grad_norm": 25.095623016357422,
      "learning_rate": 8.653026427962489e-07,
      "loss": 0.2273,
      "step": 970
    },
    {
      "epoch": 2.51,
      "grad_norm": 22.775043487548828,
      "learning_rate": 8.226768968456949e-07,
      "loss": 0.2542,
      "step": 980
    },
    {
      "epoch": 2.53,
      "grad_norm": 25.727027893066406,
      "learning_rate": 7.800511508951407e-07,
      "loss": 0.2738,
      "step": 990
    },
    {
      "epoch": 2.56,
      "grad_norm": 36.530517578125,
      "learning_rate": 7.374254049445865e-07,
      "loss": 0.2545,
      "step": 1000
    },
    {
      "epoch": 2.56,
      "eval_accuracy": 0.6714,
      "eval_f1_macro": 0.6458556535317085,
      "eval_f1_micro": 0.6714,
      "eval_loss": 1.2072625160217285,
      "eval_runtime": 12.8049,
      "eval_samples_per_second": 390.476,
      "eval_steps_per_second": 6.17,
      "step": 1000
    },
    {
      "epoch": 2.58,
      "grad_norm": 27.99028205871582,
      "learning_rate": 6.947996589940325e-07,
      "loss": 0.2864,
      "step": 1010
    },
    {
      "epoch": 2.61,
      "grad_norm": 18.240211486816406,
      "learning_rate": 6.521739130434783e-07,
      "loss": 0.2131,
      "step": 1020
    },
    {
      "epoch": 2.63,
      "grad_norm": 22.41055679321289,
      "learning_rate": 6.095481670929241e-07,
      "loss": 0.2248,
      "step": 1030
    },
    {
      "epoch": 2.66,
      "grad_norm": 33.52485656738281,
      "learning_rate": 5.669224211423701e-07,
      "loss": 0.2354,
      "step": 1040
    },
    {
      "epoch": 2.69,
      "grad_norm": 20.489734649658203,
      "learning_rate": 5.242966751918159e-07,
      "loss": 0.2069,
      "step": 1050
    },
    {
      "epoch": 2.69,
      "eval_accuracy": 0.6798,
      "eval_f1_macro": 0.6437579698962621,
      "eval_f1_micro": 0.6798,
      "eval_loss": 1.183899998664856,
      "eval_runtime": 12.9456,
      "eval_samples_per_second": 386.231,
      "eval_steps_per_second": 6.102,
      "step": 1050
    },
    {
      "epoch": 2.71,
      "grad_norm": 26.00508689880371,
      "learning_rate": 4.816709292412618e-07,
      "loss": 0.2575,
      "step": 1060
    },
    {
      "epoch": 2.74,
      "grad_norm": 37.518768310546875,
      "learning_rate": 4.3904518329070764e-07,
      "loss": 0.2543,
      "step": 1070
    },
    {
      "epoch": 2.76,
      "grad_norm": 23.725826263427734,
      "learning_rate": 3.964194373401535e-07,
      "loss": 0.2508,
      "step": 1080
    },
    {
      "epoch": 2.79,
      "grad_norm": 20.33026885986328,
      "learning_rate": 3.5379369138959936e-07,
      "loss": 0.2212,
      "step": 1090
    },
    {
      "epoch": 2.81,
      "grad_norm": 17.433706283569336,
      "learning_rate": 3.111679454390452e-07,
      "loss": 0.2109,
      "step": 1100
    },
    {
      "epoch": 2.81,
      "eval_accuracy": 0.6778,
      "eval_f1_macro": 0.6442839841548837,
      "eval_f1_micro": 0.6778,
      "eval_loss": 1.167662501335144,
      "eval_runtime": 12.8001,
      "eval_samples_per_second": 390.622,
      "eval_steps_per_second": 6.172,
      "step": 1100
    },
    {
      "epoch": 2.84,
      "grad_norm": 32.85825729370117,
      "learning_rate": 2.685421994884911e-07,
      "loss": 0.2321,
      "step": 1110
    },
    {
      "epoch": 2.86,
      "grad_norm": 30.777835845947266,
      "learning_rate": 2.2591645353793694e-07,
      "loss": 0.2671,
      "step": 1120
    },
    {
      "epoch": 2.89,
      "grad_norm": 17.449962615966797,
      "learning_rate": 1.8329070758738277e-07,
      "loss": 0.2326,
      "step": 1130
    },
    {
      "epoch": 2.92,
      "grad_norm": 34.233821868896484,
      "learning_rate": 1.4066496163682866e-07,
      "loss": 0.2256,
      "step": 1140
    },
    {
      "epoch": 2.94,
      "grad_norm": 28.030902862548828,
      "learning_rate": 9.803921568627452e-08,
      "loss": 0.2383,
      "step": 1150
    },
    {
      "epoch": 2.94,
      "eval_accuracy": 0.6776,
      "eval_f1_macro": 0.6461639329918436,
      "eval_f1_micro": 0.6776,
      "eval_loss": 1.1807124614715576,
      "eval_runtime": 12.7825,
      "eval_samples_per_second": 391.16,
      "eval_steps_per_second": 6.18,
      "step": 1150
    },
    {
      "epoch": 2.97,
      "grad_norm": 22.684385299682617,
      "learning_rate": 5.541346973572038e-08,
      "loss": 0.2418,
      "step": 1160
    },
    {
      "epoch": 2.99,
      "grad_norm": 30.925676345825195,
      "learning_rate": 1.2787723785166243e-08,
      "loss": 0.2341,
      "step": 1170
    },
    {
      "epoch": 3.0,
      "step": 1173,
      "total_flos": 7.002335143054541e+16,
      "train_loss": 0.7074367981737532,
      "train_runtime": 1784.4626,
      "train_samples_per_second": 42.029,
      "train_steps_per_second": 0.657
    }
  ],
  "logging_steps": 10,
  "max_steps": 1173,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 50,
  "total_flos": 7.002335143054541e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}