{
  "best_metric": 0.757429718875502,
  "best_model_checkpoint": "./results_small_seed2CasedUnstripped/checkpoint-19000",
  "epoch": 0.7741199478487614,
  "global_step": 19000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 1.0183299389002038e-05,
      "loss": 1.0885,
      "step": 250
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.0366598778004075e-05,
      "loss": 0.9616,
      "step": 500
    },
    {
      "epoch": 0.02,
      "eval_accuracy": 0.6345381526104418,
      "eval_loss": 0.8497252464294434,
      "eval_runtime": 4.8371,
      "eval_samples_per_second": 514.77,
      "eval_steps_per_second": 16.125,
      "step": 500
    },
    {
      "epoch": 0.03,
      "learning_rate": 3.0549898167006117e-05,
      "loss": 0.8224,
      "step": 750
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.073319755600815e-05,
      "loss": 0.7802,
      "step": 1000
    },
    {
      "epoch": 0.04,
      "eval_accuracy": 0.6522088353413654,
      "eval_loss": 0.8134140968322754,
      "eval_runtime": 4.8807,
      "eval_samples_per_second": 510.173,
      "eval_steps_per_second": 15.981,
      "step": 1000
    },
    {
      "epoch": 0.05,
      "learning_rate": 5.0916496945010185e-05,
      "loss": 0.7951,
      "step": 1250
    },
    {
      "epoch": 0.06,
      "learning_rate": 6.109979633401223e-05,
      "loss": 0.7652,
      "step": 1500
    },
    {
      "epoch": 0.06,
      "eval_accuracy": 0.6859437751004016,
      "eval_loss": 0.762012243270874,
      "eval_runtime": 4.8074,
      "eval_samples_per_second": 517.952,
      "eval_steps_per_second": 16.225,
      "step": 1500
    },
    {
      "epoch": 0.07,
      "learning_rate": 7.128309572301426e-05,
      "loss": 0.758,
      "step": 1750
    },
    {
      "epoch": 0.08,
      "learning_rate": 8.14663951120163e-05,
      "loss": 0.7811,
      "step": 2000
    },
    {
      "epoch": 0.08,
      "eval_accuracy": 0.6746987951807228,
      "eval_loss": 0.7685323357582092,
      "eval_runtime": 4.7928,
      "eval_samples_per_second": 519.528,
      "eval_steps_per_second": 16.274,
      "step": 2000
    },
    {
      "epoch": 0.09,
      "learning_rate": 9.164969450101833e-05,
      "loss": 0.7824,
      "step": 2250
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.979627869075105e-05,
      "loss": 0.81,
      "step": 2500
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.6481927710843374,
      "eval_loss": 0.7982615232467651,
      "eval_runtime": 4.802,
      "eval_samples_per_second": 518.539,
      "eval_steps_per_second": 16.243,
      "step": 2500
    },
    {
      "epoch": 0.11,
      "learning_rate": 9.866449363936802e-05,
      "loss": 0.791,
      "step": 2750
    },
    {
      "epoch": 0.12,
      "learning_rate": 9.753270858798497e-05,
      "loss": 0.793,
      "step": 3000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.6365461847389559,
      "eval_loss": 0.8263759016990662,
      "eval_runtime": 4.8663,
      "eval_samples_per_second": 511.678,
      "eval_steps_per_second": 16.028,
      "step": 3000
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.640092353660193e-05,
      "loss": 0.785,
      "step": 3250
    },
    {
      "epoch": 0.14,
      "learning_rate": 9.526913848521888e-05,
      "loss": 0.7876,
      "step": 3500
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.6598393574297189,
      "eval_loss": 0.7989997267723083,
      "eval_runtime": 4.7903,
      "eval_samples_per_second": 519.799,
      "eval_steps_per_second": 16.283,
      "step": 3500
    },
    {
      "epoch": 0.15,
      "learning_rate": 9.413735343383585e-05,
      "loss": 0.7866,
      "step": 3750
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.30055683824528e-05,
      "loss": 0.7799,
      "step": 4000
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.6718875502008033,
      "eval_loss": 0.7617995142936707,
      "eval_runtime": 4.8098,
      "eval_samples_per_second": 517.691,
      "eval_steps_per_second": 16.217,
      "step": 4000
    },
    {
      "epoch": 0.17,
      "learning_rate": 9.187378333106976e-05,
      "loss": 0.7699,
      "step": 4250
    },
    {
      "epoch": 0.18,
      "learning_rate": 9.074199827968673e-05,
      "loss": 0.7487,
      "step": 4500
    },
    {
      "epoch": 0.18,
      "eval_accuracy": 0.6767068273092369,
      "eval_loss": 0.7875193357467651,
      "eval_runtime": 4.8759,
      "eval_samples_per_second": 510.674,
      "eval_steps_per_second": 15.997,
      "step": 4500
    },
    {
      "epoch": 0.19,
      "learning_rate": 8.961021322830368e-05,
      "loss": 0.7532,
      "step": 4750
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.847842817692065e-05,
      "loss": 0.7375,
      "step": 5000
    },
    {
      "epoch": 0.2,
      "eval_accuracy": 0.6726907630522089,
      "eval_loss": 0.7819647789001465,
      "eval_runtime": 4.7823,
      "eval_samples_per_second": 520.672,
      "eval_steps_per_second": 16.31,
      "step": 5000
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.73466431255376e-05,
      "loss": 0.7374,
      "step": 5250
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.621485807415456e-05,
      "loss": 0.7674,
      "step": 5500
    },
    {
      "epoch": 0.22,
      "eval_accuracy": 0.6751004016064257,
      "eval_loss": 0.7469549179077148,
      "eval_runtime": 4.7908,
      "eval_samples_per_second": 519.751,
      "eval_steps_per_second": 16.281,
      "step": 5500
    },
    {
      "epoch": 0.23,
      "learning_rate": 8.508307302277152e-05,
      "loss": 0.7531,
      "step": 5750
    },
    {
      "epoch": 0.24,
      "learning_rate": 8.395128797138848e-05,
      "loss": 0.74,
      "step": 6000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.6763052208835342,
      "eval_loss": 0.7755768895149231,
      "eval_runtime": 4.7885,
      "eval_samples_per_second": 519.996,
      "eval_steps_per_second": 16.289,
      "step": 6000
    },
    {
      "epoch": 0.25,
      "learning_rate": 8.281950292000543e-05,
      "loss": 0.7448,
      "step": 6250
    },
    {
      "epoch": 0.26,
      "learning_rate": 8.16877178686224e-05,
      "loss": 0.7444,
      "step": 6500
    },
    {
      "epoch": 0.26,
      "eval_accuracy": 0.6775100401606425,
      "eval_loss": 0.7682979702949524,
      "eval_runtime": 4.7945,
      "eval_samples_per_second": 519.348,
      "eval_steps_per_second": 16.269,
      "step": 6500
    },
    {
      "epoch": 0.28,
      "learning_rate": 8.055593281723935e-05,
      "loss": 0.7326,
      "step": 6750
    },
    {
      "epoch": 0.29,
      "learning_rate": 7.942414776585632e-05,
      "loss": 0.7431,
      "step": 7000
    },
    {
      "epoch": 0.29,
      "eval_accuracy": 0.6783132530120481,
      "eval_loss": 0.7539064288139343,
      "eval_runtime": 4.7996,
      "eval_samples_per_second": 518.797,
      "eval_steps_per_second": 16.251,
      "step": 7000
    },
    {
      "epoch": 0.3,
      "learning_rate": 7.829236271447327e-05,
      "loss": 0.7284,
      "step": 7250
    },
    {
      "epoch": 0.31,
      "learning_rate": 7.716057766309023e-05,
      "loss": 0.7155,
      "step": 7500
    },
    {
      "epoch": 0.31,
      "eval_accuracy": 0.6827309236947792,
      "eval_loss": 0.7247231602668762,
      "eval_runtime": 4.7845,
      "eval_samples_per_second": 520.426,
      "eval_steps_per_second": 16.302,
      "step": 7500
    },
    {
      "epoch": 0.32,
      "learning_rate": 7.60287926117072e-05,
      "loss": 0.7499,
      "step": 7750
    },
    {
      "epoch": 0.33,
      "learning_rate": 7.489700756032415e-05,
      "loss": 0.7087,
      "step": 8000
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.7116465863453815,
      "eval_loss": 0.7027077674865723,
      "eval_runtime": 4.7959,
      "eval_samples_per_second": 519.195,
      "eval_steps_per_second": 16.264,
      "step": 8000
    },
    {
      "epoch": 0.34,
      "learning_rate": 7.37652225089411e-05,
      "loss": 0.7144,
      "step": 8250
    },
    {
      "epoch": 0.35,
      "learning_rate": 7.263343745755806e-05,
      "loss": 0.719,
      "step": 8500
    },
    {
      "epoch": 0.35,
      "eval_accuracy": 0.7004016064257028,
      "eval_loss": 0.73233962059021,
      "eval_runtime": 4.8318,
      "eval_samples_per_second": 515.339,
      "eval_steps_per_second": 16.143,
      "step": 8500
    },
    {
      "epoch": 0.36,
      "learning_rate": 7.150165240617503e-05,
      "loss": 0.7301,
      "step": 8750
    },
    {
      "epoch": 0.37,
      "learning_rate": 7.036986735479198e-05,
      "loss": 0.7225,
      "step": 9000
    },
    {
      "epoch": 0.37,
      "eval_accuracy": 0.7156626506024096,
      "eval_loss": 0.7080035209655762,
      "eval_runtime": 4.8777,
      "eval_samples_per_second": 510.486,
      "eval_steps_per_second": 15.991,
      "step": 9000
    },
    {
      "epoch": 0.38,
      "learning_rate": 6.923808230340894e-05,
      "loss": 0.7216,
      "step": 9250
    },
    {
      "epoch": 0.39,
      "learning_rate": 6.810629725202589e-05,
      "loss": 0.707,
      "step": 9500
    },
    {
      "epoch": 0.39,
      "eval_accuracy": 0.7004016064257028,
      "eval_loss": 0.7323043942451477,
      "eval_runtime": 4.9132,
      "eval_samples_per_second": 506.797,
      "eval_steps_per_second": 15.876,
      "step": 9500
    },
    {
      "epoch": 0.4,
      "learning_rate": 6.697451220064286e-05,
      "loss": 0.7133,
      "step": 9750
    },
    {
      "epoch": 0.41,
      "learning_rate": 6.584272714925982e-05,
      "loss": 0.6908,
      "step": 10000
    },
    {
      "epoch": 0.41,
      "eval_accuracy": 0.706425702811245,
      "eval_loss": 0.7354196310043335,
      "eval_runtime": 4.8275,
      "eval_samples_per_second": 515.794,
      "eval_steps_per_second": 16.157,
      "step": 10000
    },
    {
      "epoch": 0.42,
      "learning_rate": 6.471094209787677e-05,
      "loss": 0.7053,
      "step": 10250
    },
    {
      "epoch": 0.43,
      "learning_rate": 6.357915704649372e-05,
      "loss": 0.705,
      "step": 10500
    },
    {
      "epoch": 0.43,
      "eval_accuracy": 0.7261044176706827,
      "eval_loss": 0.6736636161804199,
      "eval_runtime": 4.8326,
      "eval_samples_per_second": 515.256,
      "eval_steps_per_second": 16.141,
      "step": 10500
    },
    {
      "epoch": 0.44,
      "learning_rate": 6.24473719951107e-05,
      "loss": 0.7213,
      "step": 10750
    },
    {
      "epoch": 0.45,
      "learning_rate": 6.131558694372765e-05,
      "loss": 0.6961,
      "step": 11000
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.7208835341365462,
      "eval_loss": 0.6973076462745667,
      "eval_runtime": 4.901,
      "eval_samples_per_second": 508.056,
      "eval_steps_per_second": 15.915,
      "step": 11000
    },
    {
      "epoch": 0.46,
      "learning_rate": 6.01838018923446e-05,
      "loss": 0.7029,
      "step": 11250
    },
    {
      "epoch": 0.47,
      "learning_rate": 5.9052016840961564e-05,
      "loss": 0.6842,
      "step": 11500
    },
    {
      "epoch": 0.47,
      "eval_accuracy": 0.6855421686746987,
      "eval_loss": 0.7288179993629456,
      "eval_runtime": 4.8248,
      "eval_samples_per_second": 516.079,
      "eval_steps_per_second": 16.166,
      "step": 11500
    },
    {
      "epoch": 0.48,
      "learning_rate": 5.7920231789578526e-05,
      "loss": 0.6894,
      "step": 11750
    },
    {
      "epoch": 0.49,
      "learning_rate": 5.678844673819549e-05,
      "loss": 0.6879,
      "step": 12000
    },
    {
      "epoch": 0.49,
      "eval_accuracy": 0.7204819277108434,
      "eval_loss": 0.6706427931785583,
      "eval_runtime": 4.816,
      "eval_samples_per_second": 517.027,
      "eval_steps_per_second": 16.196,
      "step": 12000
    },
    {
      "epoch": 0.5,
      "learning_rate": 5.565666168681244e-05,
      "loss": 0.6962,
      "step": 12250
    },
    {
      "epoch": 0.51,
      "learning_rate": 5.4524876635429404e-05,
      "loss": 0.6782,
      "step": 12500
    },
    {
      "epoch": 0.51,
      "eval_accuracy": 0.7172690763052209,
      "eval_loss": 0.6858745813369751,
      "eval_runtime": 4.8781,
      "eval_samples_per_second": 510.448,
      "eval_steps_per_second": 15.99,
      "step": 12500
    },
    {
      "epoch": 0.52,
      "learning_rate": 5.339309158404636e-05,
      "loss": 0.6871,
      "step": 12750
    },
    {
      "epoch": 0.53,
      "learning_rate": 5.226130653266332e-05,
      "loss": 0.6606,
      "step": 13000
    },
    {
      "epoch": 0.53,
      "eval_accuracy": 0.7244979919678715,
      "eval_loss": 0.6812068819999695,
      "eval_runtime": 4.8174,
      "eval_samples_per_second": 516.876,
      "eval_steps_per_second": 16.191,
      "step": 13000
    },
    {
      "epoch": 0.54,
      "learning_rate": 5.1129521481280275e-05,
      "loss": 0.6948,
      "step": 13250
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.9997736429897236e-05,
      "loss": 0.6752,
      "step": 13500
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.7321285140562249,
      "eval_loss": 0.6839885115623474,
      "eval_runtime": 4.8713,
      "eval_samples_per_second": 511.155,
      "eval_steps_per_second": 16.012,
      "step": 13500
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.88659513785142e-05,
      "loss": 0.6692,
      "step": 13750
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.773416632713115e-05,
      "loss": 0.6502,
      "step": 14000
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.7240963855421687,
      "eval_loss": 0.6590266823768616,
      "eval_runtime": 4.9174,
      "eval_samples_per_second": 506.361,
      "eval_steps_per_second": 15.862,
      "step": 14000
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.6602381275748114e-05,
      "loss": 0.6626,
      "step": 14250
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.547059622436507e-05,
      "loss": 0.6677,
      "step": 14500
    },
    {
      "epoch": 0.59,
      "eval_accuracy": 0.7289156626506024,
      "eval_loss": 0.6861540079116821,
      "eval_runtime": 4.8286,
      "eval_samples_per_second": 515.672,
      "eval_steps_per_second": 16.154,
      "step": 14500
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.433881117298203e-05,
      "loss": 0.6559,
      "step": 14750
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.3207026121598985e-05,
      "loss": 0.6694,
      "step": 15000
    },
    {
      "epoch": 0.61,
      "eval_accuracy": 0.7261044176706827,
      "eval_loss": 0.6638771891593933,
      "eval_runtime": 4.8259,
      "eval_samples_per_second": 515.971,
      "eval_steps_per_second": 16.163,
      "step": 15000
    },
    {
      "epoch": 0.62,
      "learning_rate": 4.207524107021595e-05,
      "loss": 0.6623,
      "step": 15250
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.09434560188329e-05,
      "loss": 0.6653,
      "step": 15500
    },
    {
      "epoch": 0.63,
      "eval_accuracy": 0.7128514056224899,
      "eval_loss": 0.6940041184425354,
      "eval_runtime": 4.9313,
      "eval_samples_per_second": 504.94,
      "eval_steps_per_second": 15.817,
      "step": 15500
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.981167096744986e-05,
      "loss": 0.6343,
      "step": 15750
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.8679885916066825e-05,
      "loss": 0.6636,
      "step": 16000
    },
    {
      "epoch": 0.65,
      "eval_accuracy": 0.7397590361445783,
      "eval_loss": 0.6664681434631348,
      "eval_runtime": 4.8293,
      "eval_samples_per_second": 515.605,
      "eval_steps_per_second": 16.151,
      "step": 16000
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.754810086468378e-05,
      "loss": 0.6591,
      "step": 16250
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.641631581330074e-05,
      "loss": 0.6394,
      "step": 16500
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.7441767068273092,
      "eval_loss": 0.6528923511505127,
      "eval_runtime": 4.8454,
      "eval_samples_per_second": 513.891,
      "eval_steps_per_second": 16.098,
      "step": 16500
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.52845307619177e-05,
      "loss": 0.6483,
      "step": 16750
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.415274571053466e-05,
      "loss": 0.6545,
      "step": 17000
    },
    {
      "epoch": 0.69,
      "eval_accuracy": 0.7365461847389558,
      "eval_loss": 0.6435971260070801,
      "eval_runtime": 4.8617,
      "eval_samples_per_second": 512.163,
      "eval_steps_per_second": 16.044,
      "step": 17000
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.302096065915162e-05,
      "loss": 0.6298,
      "step": 17250
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.1889175607768574e-05,
      "loss": 0.6377,
      "step": 17500
    },
    {
      "epoch": 0.71,
      "eval_accuracy": 0.738152610441767,
      "eval_loss": 0.6470540165901184,
      "eval_runtime": 4.8301,
      "eval_samples_per_second": 515.517,
      "eval_steps_per_second": 16.149,
      "step": 17500
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.0757390556385535e-05,
      "loss": 0.6441,
      "step": 17750
    },
    {
      "epoch": 0.73,
      "learning_rate": 2.962560550500249e-05,
      "loss": 0.6305,
      "step": 18000
    },
    {
      "epoch": 0.73,
      "eval_accuracy": 0.7248995983935743,
      "eval_loss": 0.6817350387573242,
      "eval_runtime": 4.833,
      "eval_samples_per_second": 515.206,
      "eval_steps_per_second": 16.139,
      "step": 18000
    },
    {
      "epoch": 0.74,
      "learning_rate": 2.849382045361945e-05,
      "loss": 0.62,
      "step": 18250
    },
    {
      "epoch": 0.75,
      "learning_rate": 2.7362035402236406e-05,
      "loss": 0.6345,
      "step": 18500
    },
    {
      "epoch": 0.75,
      "eval_accuracy": 0.7493975903614458,
      "eval_loss": 0.6382141709327698,
      "eval_runtime": 4.8844,
      "eval_samples_per_second": 509.786,
      "eval_steps_per_second": 15.969,
      "step": 18500
    },
    {
      "epoch": 0.76,
      "learning_rate": 2.6230250350853368e-05,
      "loss": 0.6236,
      "step": 18750
    },
    {
      "epoch": 0.77,
      "learning_rate": 2.5098465299470326e-05,
      "loss": 0.62,
      "step": 19000
    },
    {
      "epoch": 0.77,
      "eval_accuracy": 0.757429718875502,
      "eval_loss": 0.6400189399719238,
      "eval_runtime": 4.8308,
      "eval_samples_per_second": 515.438,
      "eval_steps_per_second": 16.146,
      "step": 19000
    }
  ],
  "max_steps": 24544,
  "num_train_epochs": 1,
  "total_flos": 1.1363112060596352e+16,
  "trial_name": null,
  "trial_params": null
}