{
  "best_metric": 0.4919604957103729,
  "best_model_checkpoint": "./vit-lr-linear/checkpoint-800",
  "epoch": 5.607476635514018,
  "eval_steps": 100,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "grad_norm": 3.5967516899108887,
      "learning_rate": 9.997196261682243e-05,
      "loss": 1.2508,
      "step": 10
    },
    {
      "epoch": 0.06,
      "grad_norm": 4.702871799468994,
      "learning_rate": 9.994080996884736e-05,
      "loss": 0.7499,
      "step": 20
    },
    {
      "epoch": 0.09,
      "grad_norm": 6.171701431274414,
      "learning_rate": 9.990965732087227e-05,
      "loss": 0.8834,
      "step": 30
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.749275207519531,
      "learning_rate": 9.98785046728972e-05,
      "loss": 1.0502,
      "step": 40
    },
    {
      "epoch": 0.16,
      "grad_norm": 4.334569931030273,
      "learning_rate": 9.984735202492212e-05,
      "loss": 0.769,
      "step": 50
    },
    {
      "epoch": 0.19,
      "grad_norm": 4.809230327606201,
      "learning_rate": 9.981619937694705e-05,
      "loss": 0.9034,
      "step": 60
    },
    {
      "epoch": 0.22,
      "grad_norm": 3.983206272125244,
      "learning_rate": 9.978504672897196e-05,
      "loss": 0.8274,
      "step": 70
    },
    {
      "epoch": 0.25,
      "grad_norm": 3.574300765991211,
      "learning_rate": 9.975389408099689e-05,
      "loss": 0.6348,
      "step": 80
    },
    {
      "epoch": 0.28,
      "grad_norm": 5.728092670440674,
      "learning_rate": 9.972274143302182e-05,
      "loss": 0.6568,
      "step": 90
    },
    {
      "epoch": 0.31,
      "grad_norm": 4.649799823760986,
      "learning_rate": 9.969158878504672e-05,
      "loss": 0.6029,
      "step": 100
    },
    {
      "epoch": 0.31,
      "eval_accuracy": 0.7805131761442441,
      "eval_f1": 0.7528727438506023,
      "eval_loss": 0.6126354932785034,
      "eval_precision": 0.7602026831090402,
      "eval_recall": 0.7805131761442441,
      "eval_runtime": 39.2505,
      "eval_samples_per_second": 73.477,
      "eval_steps_per_second": 9.197,
      "step": 100
    },
    {
      "epoch": 0.34,
      "grad_norm": 5.011153221130371,
      "learning_rate": 9.966043613707165e-05,
      "loss": 0.6847,
      "step": 110
    },
    {
      "epoch": 0.37,
      "grad_norm": 7.656059741973877,
      "learning_rate": 9.962928348909658e-05,
      "loss": 0.6162,
      "step": 120
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.763223648071289,
      "learning_rate": 9.95981308411215e-05,
      "loss": 0.6389,
      "step": 130
    },
    {
      "epoch": 0.44,
      "grad_norm": 2.2059271335601807,
      "learning_rate": 9.956697819314643e-05,
      "loss": 0.5381,
      "step": 140
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.06673002243042,
      "learning_rate": 9.953582554517134e-05,
      "loss": 0.5674,
      "step": 150
    },
    {
      "epoch": 0.5,
      "grad_norm": 5.407893657684326,
      "learning_rate": 9.950467289719627e-05,
      "loss": 0.5598,
      "step": 160
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.515843391418457,
      "learning_rate": 9.947352024922119e-05,
      "loss": 0.5488,
      "step": 170
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.197587966918945,
      "learning_rate": 9.944236760124612e-05,
      "loss": 0.7078,
      "step": 180
    },
    {
      "epoch": 0.59,
      "grad_norm": 4.026732921600342,
      "learning_rate": 9.941121495327103e-05,
      "loss": 0.595,
      "step": 190
    },
    {
      "epoch": 0.62,
      "grad_norm": 5.807314872741699,
      "learning_rate": 9.938006230529595e-05,
      "loss": 0.5726,
      "step": 200
    },
    {
      "epoch": 0.62,
      "eval_accuracy": 0.7649098474341193,
      "eval_f1": 0.7177347277722194,
      "eval_loss": 0.6950196623802185,
      "eval_precision": 0.7613231169607695,
      "eval_recall": 0.7649098474341193,
      "eval_runtime": 39.3588,
      "eval_samples_per_second": 73.275,
      "eval_steps_per_second": 9.172,
      "step": 200
    },
    {
      "epoch": 0.65,
      "grad_norm": 5.1864399909973145,
      "learning_rate": 9.934890965732088e-05,
      "loss": 0.6681,
      "step": 210
    },
    {
      "epoch": 0.69,
      "grad_norm": 7.364803314208984,
      "learning_rate": 9.93177570093458e-05,
      "loss": 0.7502,
      "step": 220
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.9555000066757202,
      "learning_rate": 9.928660436137072e-05,
      "loss": 0.3976,
      "step": 230
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.5017478466033936,
      "learning_rate": 9.925545171339564e-05,
      "loss": 0.5505,
      "step": 240
    },
    {
      "epoch": 0.78,
      "grad_norm": 4.633613109588623,
      "learning_rate": 9.922429906542056e-05,
      "loss": 0.7452,
      "step": 250
    },
    {
      "epoch": 0.81,
      "grad_norm": 4.068040370941162,
      "learning_rate": 9.91931464174455e-05,
      "loss": 0.5837,
      "step": 260
    },
    {
      "epoch": 0.84,
      "grad_norm": 3.191166639328003,
      "learning_rate": 9.916199376947041e-05,
      "loss": 0.4488,
      "step": 270
    },
    {
      "epoch": 0.87,
      "grad_norm": 2.8749473094940186,
      "learning_rate": 9.913084112149534e-05,
      "loss": 0.4906,
      "step": 280
    },
    {
      "epoch": 0.9,
      "grad_norm": 7.547589302062988,
      "learning_rate": 9.909968847352025e-05,
      "loss": 0.6191,
      "step": 290
    },
    {
      "epoch": 0.93,
      "grad_norm": 5.1279096603393555,
      "learning_rate": 9.906853582554517e-05,
      "loss": 0.6521,
      "step": 300
    },
    {
      "epoch": 0.93,
      "eval_accuracy": 0.8124133148404993,
      "eval_f1": 0.8060484287452215,
      "eval_loss": 0.5102406740188599,
      "eval_precision": 0.8148843275891411,
      "eval_recall": 0.8124133148404993,
      "eval_runtime": 39.7509,
      "eval_samples_per_second": 72.552,
      "eval_steps_per_second": 9.082,
      "step": 300
    },
    {
      "epoch": 0.97,
      "grad_norm": 3.052353858947754,
      "learning_rate": 9.90373831775701e-05,
      "loss": 0.5616,
      "step": 310
    },
    {
      "epoch": 1.0,
      "grad_norm": 4.507020473480225,
      "learning_rate": 9.900623052959503e-05,
      "loss": 0.5192,
      "step": 320
    },
    {
      "epoch": 1.03,
      "grad_norm": 5.176488876342773,
      "learning_rate": 9.897507788161994e-05,
      "loss": 0.4461,
      "step": 330
    },
    {
      "epoch": 1.06,
      "grad_norm": 4.797460079193115,
      "learning_rate": 9.894392523364486e-05,
      "loss": 0.3651,
      "step": 340
    },
    {
      "epoch": 1.09,
      "grad_norm": 2.186629295349121,
      "learning_rate": 9.891277258566979e-05,
      "loss": 0.4004,
      "step": 350
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.1874918937683105,
      "learning_rate": 9.888161993769472e-05,
      "loss": 0.3475,
      "step": 360
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.8143486976623535,
      "learning_rate": 9.885046728971963e-05,
      "loss": 0.4073,
      "step": 370
    },
    {
      "epoch": 1.18,
      "grad_norm": 7.921704292297363,
      "learning_rate": 9.881931464174455e-05,
      "loss": 0.3765,
      "step": 380
    },
    {
      "epoch": 1.21,
      "grad_norm": 5.491533279418945,
      "learning_rate": 9.878816199376948e-05,
      "loss": 0.4707,
      "step": 390
    },
    {
      "epoch": 1.25,
      "grad_norm": 2.563472032546997,
      "learning_rate": 9.875700934579439e-05,
      "loss": 0.3803,
      "step": 400
    },
    {
      "epoch": 1.25,
      "eval_accuracy": 0.7843273231622746,
      "eval_f1": 0.7933752302092828,
      "eval_loss": 0.6124634146690369,
      "eval_precision": 0.8128379357321871,
      "eval_recall": 0.7843273231622746,
      "eval_runtime": 39.1164,
      "eval_samples_per_second": 73.729,
      "eval_steps_per_second": 9.229,
      "step": 400
    },
    {
      "epoch": 1.28,
      "grad_norm": 3.978813648223877,
      "learning_rate": 9.872585669781932e-05,
      "loss": 0.4527,
      "step": 410
    },
    {
      "epoch": 1.31,
      "grad_norm": 3.3713996410369873,
      "learning_rate": 9.869470404984425e-05,
      "loss": 0.4344,
      "step": 420
    },
    {
      "epoch": 1.34,
      "grad_norm": 3.685149669647217,
      "learning_rate": 9.866355140186917e-05,
      "loss": 0.5163,
      "step": 430
    },
    {
      "epoch": 1.37,
      "grad_norm": 3.9468307495117188,
      "learning_rate": 9.863239875389408e-05,
      "loss": 0.4187,
      "step": 440
    },
    {
      "epoch": 1.4,
      "grad_norm": 4.608047962188721,
      "learning_rate": 9.860124610591901e-05,
      "loss": 0.3643,
      "step": 450
    },
    {
      "epoch": 1.43,
      "grad_norm": 3.3221333026885986,
      "learning_rate": 9.857009345794394e-05,
      "loss": 0.3324,
      "step": 460
    },
    {
      "epoch": 1.46,
      "grad_norm": 3.251314640045166,
      "learning_rate": 9.853894080996885e-05,
      "loss": 0.3229,
      "step": 470
    },
    {
      "epoch": 1.5,
      "grad_norm": 7.8897552490234375,
      "learning_rate": 9.850778816199377e-05,
      "loss": 0.3684,
      "step": 480
    },
    {
      "epoch": 1.53,
      "grad_norm": 5.32474422454834,
      "learning_rate": 9.84766355140187e-05,
      "loss": 0.3567,
      "step": 490
    },
    {
      "epoch": 1.56,
      "grad_norm": 2.955794334411621,
      "learning_rate": 9.844548286604361e-05,
      "loss": 0.4048,
      "step": 500
    },
    {
      "epoch": 1.56,
      "eval_accuracy": 0.8214285714285714,
      "eval_f1": 0.80775004871548,
      "eval_loss": 0.5058895349502563,
      "eval_precision": 0.8156080811716645,
      "eval_recall": 0.8214285714285714,
      "eval_runtime": 39.0979,
      "eval_samples_per_second": 73.764,
      "eval_steps_per_second": 9.233,
      "step": 500
    },
    {
      "epoch": 1.59,
      "grad_norm": 3.655855894088745,
      "learning_rate": 9.841433021806854e-05,
      "loss": 0.4464,
      "step": 510
    },
    {
      "epoch": 1.62,
      "grad_norm": 3.6523778438568115,
      "learning_rate": 9.838317757009346e-05,
      "loss": 0.4154,
      "step": 520
    },
    {
      "epoch": 1.65,
      "grad_norm": 6.649527072906494,
      "learning_rate": 9.835202492211837e-05,
      "loss": 0.4337,
      "step": 530
    },
    {
      "epoch": 1.68,
      "grad_norm": 4.308875560760498,
      "learning_rate": 9.83208722741433e-05,
      "loss": 0.3551,
      "step": 540
    },
    {
      "epoch": 1.71,
      "grad_norm": 5.290976047515869,
      "learning_rate": 9.828971962616823e-05,
      "loss": 0.4329,
      "step": 550
    },
    {
      "epoch": 1.74,
      "grad_norm": 3.688100814819336,
      "learning_rate": 9.825856697819316e-05,
      "loss": 0.4306,
      "step": 560
    },
    {
      "epoch": 1.78,
      "grad_norm": 2.4794669151306152,
      "learning_rate": 9.822741433021808e-05,
      "loss": 0.3281,
      "step": 570
    },
    {
      "epoch": 1.81,
      "grad_norm": 3.5106523036956787,
      "learning_rate": 9.819626168224299e-05,
      "loss": 0.3184,
      "step": 580
    },
    {
      "epoch": 1.84,
      "grad_norm": 5.135136604309082,
      "learning_rate": 9.816510903426792e-05,
      "loss": 0.3644,
      "step": 590
    },
    {
      "epoch": 1.87,
      "grad_norm": 3.2593114376068115,
      "learning_rate": 9.813395638629284e-05,
      "loss": 0.2939,
      "step": 600
    },
    {
      "epoch": 1.87,
      "eval_accuracy": 0.7680305131761442,
      "eval_f1": 0.7817959651329371,
      "eval_loss": 0.6723023653030396,
      "eval_precision": 0.8366418813297796,
      "eval_recall": 0.7680305131761442,
      "eval_runtime": 39.3089,
      "eval_samples_per_second": 73.368,
      "eval_steps_per_second": 9.184,
      "step": 600
    },
    {
      "epoch": 1.9,
      "grad_norm": 7.719142913818359,
      "learning_rate": 9.810280373831777e-05,
      "loss": 0.5629,
      "step": 610
    },
    {
      "epoch": 1.93,
      "grad_norm": 4.130067348480225,
      "learning_rate": 9.807165109034268e-05,
      "loss": 0.4416,
      "step": 620
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.9364572763442993,
      "learning_rate": 9.80404984423676e-05,
      "loss": 0.2339,
      "step": 630
    },
    {
      "epoch": 1.99,
      "grad_norm": 2.4033007621765137,
      "learning_rate": 9.800934579439253e-05,
      "loss": 0.3284,
      "step": 640
    },
    {
      "epoch": 2.02,
      "grad_norm": 3.25533390045166,
      "learning_rate": 9.797819314641746e-05,
      "loss": 0.2307,
      "step": 650
    },
    {
      "epoch": 2.06,
      "grad_norm": 3.1277670860290527,
      "learning_rate": 9.794704049844237e-05,
      "loss": 0.2063,
      "step": 660
    },
    {
      "epoch": 2.09,
      "grad_norm": 2.923804521560669,
      "learning_rate": 9.791588785046729e-05,
      "loss": 0.1563,
      "step": 670
    },
    {
      "epoch": 2.12,
      "grad_norm": 4.745158672332764,
      "learning_rate": 9.788473520249222e-05,
      "loss": 0.2451,
      "step": 680
    },
    {
      "epoch": 2.15,
      "grad_norm": 5.691440582275391,
      "learning_rate": 9.785358255451714e-05,
      "loss": 0.2315,
      "step": 690
    },
    {
      "epoch": 2.18,
      "grad_norm": 7.0971550941467285,
      "learning_rate": 9.782242990654206e-05,
      "loss": 0.2138,
      "step": 700
    },
    {
      "epoch": 2.18,
      "eval_accuracy": 0.812760055478502,
      "eval_f1": 0.8169834700772869,
      "eval_loss": 0.635110080242157,
      "eval_precision": 0.8480203726240728,
      "eval_recall": 0.812760055478502,
      "eval_runtime": 39.4295,
      "eval_samples_per_second": 73.143,
      "eval_steps_per_second": 9.156,
      "step": 700
    },
    {
      "epoch": 2.21,
      "grad_norm": 2.342768430709839,
      "learning_rate": 9.779127725856699e-05,
      "loss": 0.2763,
      "step": 710
    },
    {
      "epoch": 2.24,
      "grad_norm": 4.194737434387207,
      "learning_rate": 9.77601246105919e-05,
      "loss": 0.3159,
      "step": 720
    },
    {
      "epoch": 2.27,
      "grad_norm": 4.488862991333008,
      "learning_rate": 9.772897196261682e-05,
      "loss": 0.2594,
      "step": 730
    },
    {
      "epoch": 2.31,
      "grad_norm": 3.612229347229004,
      "learning_rate": 9.769781931464175e-05,
      "loss": 0.1651,
      "step": 740
    },
    {
      "epoch": 2.34,
      "grad_norm": 4.490918159484863,
      "learning_rate": 9.766666666666668e-05,
      "loss": 0.1714,
      "step": 750
    },
    {
      "epoch": 2.37,
      "grad_norm": 1.2064989805221558,
      "learning_rate": 9.76355140186916e-05,
      "loss": 0.1423,
      "step": 760
    },
    {
      "epoch": 2.4,
      "grad_norm": 3.8275320529937744,
      "learning_rate": 9.760436137071651e-05,
      "loss": 0.2647,
      "step": 770
    },
    {
      "epoch": 2.43,
      "grad_norm": 6.258784294128418,
      "learning_rate": 9.757320872274144e-05,
      "loss": 0.1925,
      "step": 780
    },
    {
      "epoch": 2.46,
      "grad_norm": 6.503872394561768,
      "learning_rate": 9.754205607476637e-05,
      "loss": 0.3565,
      "step": 790
    },
    {
      "epoch": 2.49,
      "grad_norm": 4.768139362335205,
      "learning_rate": 9.751090342679128e-05,
      "loss": 0.2615,
      "step": 800
    },
    {
      "epoch": 2.49,
      "eval_accuracy": 0.8321775312066574,
      "eval_f1": 0.8323189306799796,
      "eval_loss": 0.4919604957103729,
      "eval_precision": 0.8400335349095535,
      "eval_recall": 0.8321775312066574,
      "eval_runtime": 39.8278,
      "eval_samples_per_second": 72.412,
      "eval_steps_per_second": 9.064,
      "step": 800
    },
    {
      "epoch": 2.52,
      "grad_norm": 3.3493497371673584,
      "learning_rate": 9.74797507788162e-05,
      "loss": 0.179,
      "step": 810
    },
    {
      "epoch": 2.55,
      "grad_norm": 6.209288597106934,
      "learning_rate": 9.744859813084113e-05,
      "loss": 0.1872,
      "step": 820
    },
    {
      "epoch": 2.59,
      "grad_norm": 3.6066982746124268,
      "learning_rate": 9.741744548286604e-05,
      "loss": 0.24,
      "step": 830
    },
    {
      "epoch": 2.62,
      "grad_norm": 2.330963611602783,
      "learning_rate": 9.738629283489097e-05,
      "loss": 0.177,
      "step": 840
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.31879034638404846,
      "learning_rate": 9.73551401869159e-05,
      "loss": 0.1846,
      "step": 850
    },
    {
      "epoch": 2.68,
      "grad_norm": 6.56818151473999,
      "learning_rate": 9.732398753894082e-05,
      "loss": 0.2731,
      "step": 860
    },
    {
      "epoch": 2.71,
      "grad_norm": 1.255943775177002,
      "learning_rate": 9.729283489096573e-05,
      "loss": 0.3217,
      "step": 870
    },
    {
      "epoch": 2.74,
      "grad_norm": 5.009977340698242,
      "learning_rate": 9.726168224299066e-05,
      "loss": 0.2576,
      "step": 880
    },
    {
      "epoch": 2.77,
      "grad_norm": 1.8944584131240845,
      "learning_rate": 9.723052959501559e-05,
      "loss": 0.2223,
      "step": 890
    },
    {
      "epoch": 2.8,
      "grad_norm": 3.543699026107788,
      "learning_rate": 9.71993769470405e-05,
      "loss": 0.2125,
      "step": 900
    },
    {
      "epoch": 2.8,
      "eval_accuracy": 0.8491678224687933,
      "eval_f1": 0.843224216265628,
      "eval_loss": 0.5596445202827454,
      "eval_precision": 0.8508841434558159,
      "eval_recall": 0.8491678224687933,
      "eval_runtime": 39.2828,
      "eval_samples_per_second": 73.416,
      "eval_steps_per_second": 9.19,
      "step": 900
    },
    {
      "epoch": 2.83,
      "grad_norm": 5.571394920349121,
      "learning_rate": 9.716822429906542e-05,
      "loss": 0.2826,
      "step": 910
    },
    {
      "epoch": 2.87,
      "grad_norm": 5.057092666625977,
      "learning_rate": 9.713707165109035e-05,
      "loss": 0.2647,
      "step": 920
    },
    {
      "epoch": 2.9,
      "grad_norm": 3.190361738204956,
      "learning_rate": 9.710591900311527e-05,
      "loss": 0.4015,
      "step": 930
    },
    {
      "epoch": 2.93,
      "grad_norm": 6.205695629119873,
      "learning_rate": 9.70747663551402e-05,
      "loss": 0.312,
      "step": 940
    },
    {
      "epoch": 2.96,
      "grad_norm": 4.805609226226807,
      "learning_rate": 9.704361370716511e-05,
      "loss": 0.2098,
      "step": 950
    },
    {
      "epoch": 2.99,
      "grad_norm": 3.2721455097198486,
      "learning_rate": 9.701246105919004e-05,
      "loss": 0.1959,
      "step": 960
    },
    {
      "epoch": 3.02,
      "grad_norm": 4.190710067749023,
      "learning_rate": 9.698130841121495e-05,
      "loss": 0.1048,
      "step": 970
    },
    {
      "epoch": 3.05,
      "grad_norm": 0.5176565051078796,
      "learning_rate": 9.695015576323988e-05,
      "loss": 0.0816,
      "step": 980
    },
    {
      "epoch": 3.08,
      "grad_norm": 0.1512940376996994,
      "learning_rate": 9.691900311526481e-05,
      "loss": 0.0327,
      "step": 990
    },
    {
      "epoch": 3.12,
      "grad_norm": 5.473361015319824,
      "learning_rate": 9.688785046728971e-05,
      "loss": 0.0768,
      "step": 1000
    },
    {
      "epoch": 3.12,
      "eval_accuracy": 0.8290568654646324,
      "eval_f1": 0.8235345080050543,
      "eval_loss": 0.823904275894165,
      "eval_precision": 0.8499722083747269,
      "eval_recall": 0.8290568654646324,
      "eval_runtime": 39.6039,
      "eval_samples_per_second": 72.821,
      "eval_steps_per_second": 9.115,
      "step": 1000
    },
    {
      "epoch": 3.15,
      "grad_norm": 2.461423635482788,
      "learning_rate": 9.685669781931464e-05,
      "loss": 0.1258,
      "step": 1010
    },
    {
      "epoch": 3.18,
      "grad_norm": 5.284664154052734,
      "learning_rate": 9.682554517133957e-05,
      "loss": 0.1124,
      "step": 1020
    },
    {
      "epoch": 3.21,
      "grad_norm": 1.3465226888656616,
      "learning_rate": 9.679439252336449e-05,
      "loss": 0.0986,
      "step": 1030
    },
    {
      "epoch": 3.24,
      "grad_norm": 3.916722297668457,
      "learning_rate": 9.676323987538942e-05,
      "loss": 0.0541,
      "step": 1040
    },
    {
      "epoch": 3.27,
      "grad_norm": 0.6409209966659546,
      "learning_rate": 9.673208722741433e-05,
      "loss": 0.1176,
      "step": 1050
    },
    {
      "epoch": 3.3,
      "grad_norm": 4.325451374053955,
      "learning_rate": 9.670093457943926e-05,
      "loss": 0.2169,
      "step": 1060
    },
    {
      "epoch": 3.33,
      "grad_norm": 0.5127350687980652,
      "learning_rate": 9.666978193146418e-05,
      "loss": 0.0739,
      "step": 1070
    },
    {
      "epoch": 3.36,
      "grad_norm": 7.527690410614014,
      "learning_rate": 9.66386292834891e-05,
      "loss": 0.184,
      "step": 1080
    },
    {
      "epoch": 3.4,
      "grad_norm": 2.938415765762329,
      "learning_rate": 9.660747663551402e-05,
      "loss": 0.1526,
      "step": 1090
    },
    {
      "epoch": 3.43,
      "grad_norm": 0.9604325890541077,
      "learning_rate": 9.657632398753894e-05,
      "loss": 0.0649,
      "step": 1100
    },
    {
      "epoch": 3.43,
      "eval_accuracy": 0.8366851595006934,
      "eval_f1": 0.8359690431594896,
      "eval_loss": 0.6827093958854675,
      "eval_precision": 0.8480904372623412,
      "eval_recall": 0.8366851595006934,
      "eval_runtime": 39.8316,
      "eval_samples_per_second": 72.405,
      "eval_steps_per_second": 9.063,
      "step": 1100
    },
    {
      "epoch": 3.46,
      "grad_norm": 0.6656555533409119,
      "learning_rate": 9.654517133956387e-05,
      "loss": 0.0866,
      "step": 1110
    },
    {
      "epoch": 3.49,
      "grad_norm": 1.8327091932296753,
      "learning_rate": 9.65140186915888e-05,
      "loss": 0.1117,
      "step": 1120
    },
    {
      "epoch": 3.52,
      "grad_norm": 6.639819145202637,
      "learning_rate": 9.648286604361371e-05,
      "loss": 0.0607,
      "step": 1130
    },
    {
      "epoch": 3.55,
      "grad_norm": 9.595551490783691,
      "learning_rate": 9.645171339563863e-05,
      "loss": 0.1603,
      "step": 1140
    },
    {
      "epoch": 3.58,
      "grad_norm": 4.852327346801758,
      "learning_rate": 9.642056074766356e-05,
      "loss": 0.1289,
      "step": 1150
    },
    {
      "epoch": 3.61,
      "grad_norm": 0.8141772150993347,
      "learning_rate": 9.638940809968848e-05,
      "loss": 0.2513,
      "step": 1160
    },
    {
      "epoch": 3.64,
      "grad_norm": 2.2528672218322754,
      "learning_rate": 9.63582554517134e-05,
      "loss": 0.1033,
      "step": 1170
    },
    {
      "epoch": 3.68,
      "grad_norm": 4.099298000335693,
      "learning_rate": 9.632710280373833e-05,
      "loss": 0.2042,
      "step": 1180
    },
    {
      "epoch": 3.71,
      "grad_norm": 2.307119131088257,
      "learning_rate": 9.629595015576324e-05,
      "loss": 0.0714,
      "step": 1190
    },
    {
      "epoch": 3.74,
      "grad_norm": 6.8174543380737305,
      "learning_rate": 9.626479750778816e-05,
      "loss": 0.1382,
      "step": 1200
    },
    {
      "epoch": 3.74,
      "eval_accuracy": 0.84500693481276,
      "eval_f1": 0.8398698444204201,
      "eval_loss": 0.6838334798812866,
      "eval_precision": 0.8466724060209081,
      "eval_recall": 0.84500693481276,
      "eval_runtime": 39.4271,
      "eval_samples_per_second": 73.148,
      "eval_steps_per_second": 9.156,
      "step": 1200
    },
    {
      "epoch": 3.77,
      "grad_norm": 3.4418959617614746,
      "learning_rate": 9.623364485981309e-05,
      "loss": 0.0921,
      "step": 1210
    },
    {
      "epoch": 3.8,
      "grad_norm": 3.6201603412628174,
      "learning_rate": 9.620249221183802e-05,
      "loss": 0.15,
      "step": 1220
    },
    {
      "epoch": 3.83,
      "grad_norm": 6.8857550621032715,
      "learning_rate": 9.617133956386293e-05,
      "loss": 0.0965,
      "step": 1230
    },
    {
      "epoch": 3.86,
      "grad_norm": 3.10553240776062,
      "learning_rate": 9.614018691588785e-05,
      "loss": 0.1674,
      "step": 1240
    },
    {
      "epoch": 3.89,
      "grad_norm": 9.131609916687012,
      "learning_rate": 9.610903426791278e-05,
      "loss": 0.1596,
      "step": 1250
    },
    {
      "epoch": 3.93,
      "grad_norm": 0.35134002566337585,
      "learning_rate": 9.607788161993771e-05,
      "loss": 0.1157,
      "step": 1260
    },
    {
      "epoch": 3.96,
      "grad_norm": 5.575935363769531,
      "learning_rate": 9.604672897196262e-05,
      "loss": 0.0822,
      "step": 1270
    },
    {
      "epoch": 3.99,
      "grad_norm": 0.6137746572494507,
      "learning_rate": 9.601557632398754e-05,
      "loss": 0.1316,
      "step": 1280
    },
    {
      "epoch": 4.02,
      "grad_norm": 0.34164953231811523,
      "learning_rate": 9.598442367601247e-05,
      "loss": 0.0739,
      "step": 1290
    },
    {
      "epoch": 4.05,
      "grad_norm": 4.076730728149414,
      "learning_rate": 9.595327102803738e-05,
      "loss": 0.0486,
      "step": 1300
    },
    {
      "epoch": 4.05,
      "eval_accuracy": 0.8578363384188626,
      "eval_f1": 0.8494463868744596,
      "eval_loss": 0.6367300748825073,
      "eval_precision": 0.8548000651229425,
      "eval_recall": 0.8578363384188626,
      "eval_runtime": 39.1728,
      "eval_samples_per_second": 73.623,
      "eval_steps_per_second": 9.216,
      "step": 1300
    },
    {
      "epoch": 4.08,
      "grad_norm": 0.4267037808895111,
      "learning_rate": 9.592211838006231e-05,
      "loss": 0.0499,
      "step": 1310
    },
    {
      "epoch": 4.11,
      "grad_norm": 8.932145118713379,
      "learning_rate": 9.589096573208724e-05,
      "loss": 0.058,
      "step": 1320
    },
    {
      "epoch": 4.14,
      "grad_norm": 7.81501579284668,
      "learning_rate": 9.585981308411214e-05,
      "loss": 0.0497,
      "step": 1330
    },
    {
      "epoch": 4.17,
      "grad_norm": 3.25376296043396,
      "learning_rate": 9.582866043613707e-05,
      "loss": 0.0613,
      "step": 1340
    },
    {
      "epoch": 4.21,
      "grad_norm": 0.009625586681067944,
      "learning_rate": 9.5797507788162e-05,
      "loss": 0.0882,
      "step": 1350
    },
    {
      "epoch": 4.24,
      "grad_norm": 8.644308090209961,
      "learning_rate": 9.576635514018693e-05,
      "loss": 0.0729,
      "step": 1360
    },
    {
      "epoch": 4.27,
      "grad_norm": 11.613913536071777,
      "learning_rate": 9.573520249221185e-05,
      "loss": 0.1285,
      "step": 1370
    },
    {
      "epoch": 4.3,
      "grad_norm": 0.9490543603897095,
      "learning_rate": 9.570404984423676e-05,
      "loss": 0.0408,
      "step": 1380
    },
    {
      "epoch": 4.33,
      "grad_norm": 2.557040214538574,
      "learning_rate": 9.567289719626169e-05,
      "loss": 0.0689,
      "step": 1390
    },
    {
      "epoch": 4.36,
      "grad_norm": 7.547731399536133,
      "learning_rate": 9.56417445482866e-05,
      "loss": 0.1122,
      "step": 1400
    },
    {
      "epoch": 4.36,
      "eval_accuracy": 0.8398058252427184,
      "eval_f1": 0.833035822565054,
      "eval_loss": 0.7330206036567688,
      "eval_precision": 0.836759139491613,
      "eval_recall": 0.8398058252427184,
      "eval_runtime": 39.5302,
      "eval_samples_per_second": 72.957,
      "eval_steps_per_second": 9.132,
      "step": 1400
    },
    {
      "epoch": 4.39,
      "grad_norm": 0.47195371985435486,
      "learning_rate": 9.561059190031153e-05,
      "loss": 0.0087,
      "step": 1410
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.02496817521750927,
      "learning_rate": 9.557943925233645e-05,
      "loss": 0.0678,
      "step": 1420
    },
    {
      "epoch": 4.45,
      "grad_norm": 0.044717635959386826,
      "learning_rate": 9.554828660436137e-05,
      "loss": 0.0409,
      "step": 1430
    },
    {
      "epoch": 4.49,
      "grad_norm": 2.304049015045166,
      "learning_rate": 9.55171339563863e-05,
      "loss": 0.0661,
      "step": 1440
    },
    {
      "epoch": 4.52,
      "grad_norm": 11.10191822052002,
      "learning_rate": 9.548598130841122e-05,
      "loss": 0.3254,
      "step": 1450
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.0031379794236272573,
      "learning_rate": 9.545482866043615e-05,
      "loss": 0.0655,
      "step": 1460
    },
    {
      "epoch": 4.58,
      "grad_norm": 1.050758719444275,
      "learning_rate": 9.542367601246105e-05,
      "loss": 0.0968,
      "step": 1470
    },
    {
      "epoch": 4.61,
      "grad_norm": 0.027871431782841682,
      "learning_rate": 9.539252336448598e-05,
      "loss": 0.1033,
      "step": 1480
    },
    {
      "epoch": 4.64,
      "grad_norm": 0.054837290197610855,
      "learning_rate": 9.536137071651091e-05,
      "loss": 0.0225,
      "step": 1490
    },
    {
      "epoch": 4.67,
      "grad_norm": 5.67630672454834,
      "learning_rate": 9.533021806853583e-05,
      "loss": 0.0302,
      "step": 1500
    },
    {
      "epoch": 4.67,
      "eval_accuracy": 0.84500693481276,
      "eval_f1": 0.8442384441117506,
      "eval_loss": 0.7136919498443604,
      "eval_precision": 0.8469740143199302,
      "eval_recall": 0.84500693481276,
      "eval_runtime": 39.1304,
      "eval_samples_per_second": 73.702,
      "eval_steps_per_second": 9.226,
      "step": 1500
    },
    {
      "epoch": 4.7,
      "grad_norm": 0.295539915561676,
      "learning_rate": 9.529906542056076e-05,
      "loss": 0.1178,
      "step": 1510
    },
    {
      "epoch": 4.74,
      "grad_norm": 0.0796700268983841,
      "learning_rate": 9.526791277258567e-05,
      "loss": 0.0481,
      "step": 1520
    },
    {
      "epoch": 4.77,
      "grad_norm": 0.1068115308880806,
      "learning_rate": 9.523676012461059e-05,
      "loss": 0.0282,
      "step": 1530
    },
    {
      "epoch": 4.8,
      "grad_norm": 1.0221561193466187,
      "learning_rate": 9.520560747663552e-05,
      "loss": 0.0538,
      "step": 1540
    },
    {
      "epoch": 4.83,
      "grad_norm": 7.369207859039307,
      "learning_rate": 9.517445482866045e-05,
      "loss": 0.1239,
      "step": 1550
    },
    {
      "epoch": 4.86,
      "grad_norm": 9.008218765258789,
      "learning_rate": 9.514330218068536e-05,
      "loss": 0.055,
      "step": 1560
    },
    {
      "epoch": 4.89,
      "grad_norm": 3.585855722427368,
      "learning_rate": 9.511214953271028e-05,
      "loss": 0.03,
      "step": 1570
    },
    {
      "epoch": 4.92,
      "grad_norm": 0.15154320001602173,
      "learning_rate": 9.50809968847352e-05,
      "loss": 0.0168,
      "step": 1580
    },
    {
      "epoch": 4.95,
      "grad_norm": 0.030903339385986328,
      "learning_rate": 9.504984423676014e-05,
      "loss": 0.1067,
      "step": 1590
    },
    {
      "epoch": 4.98,
      "grad_norm": 0.4652014672756195,
      "learning_rate": 9.501869158878505e-05,
      "loss": 0.0462,
      "step": 1600
    },
    {
      "epoch": 4.98,
      "eval_accuracy": 0.8515950069348127,
      "eval_f1": 0.8455611718307666,
      "eval_loss": 0.8198381066322327,
      "eval_precision": 0.8519412050125947,
      "eval_recall": 0.8515950069348127,
      "eval_runtime": 39.6856,
      "eval_samples_per_second": 72.671,
      "eval_steps_per_second": 9.096,
      "step": 1600
    },
    {
      "epoch": 5.02,
      "grad_norm": 14.000198364257812,
      "learning_rate": 9.498753894080997e-05,
      "loss": 0.0785,
      "step": 1610
    },
    {
      "epoch": 5.05,
      "grad_norm": 0.21171867847442627,
      "learning_rate": 9.49563862928349e-05,
      "loss": 0.019,
      "step": 1620
    },
    {
      "epoch": 5.08,
      "grad_norm": 0.004491983912885189,
      "learning_rate": 9.492523364485981e-05,
      "loss": 0.0111,
      "step": 1630
    },
    {
      "epoch": 5.11,
      "grad_norm": 0.016514340415596962,
      "learning_rate": 9.489408099688474e-05,
      "loss": 0.0897,
      "step": 1640
    },
    {
      "epoch": 5.14,
      "grad_norm": 8.40817928314209,
      "learning_rate": 9.486292834890967e-05,
      "loss": 0.0798,
      "step": 1650
    },
    {
      "epoch": 5.17,
      "grad_norm": 0.07949113100767136,
      "learning_rate": 9.483177570093458e-05,
      "loss": 0.0216,
      "step": 1660
    },
    {
      "epoch": 5.2,
      "grad_norm": 4.111806869506836,
      "learning_rate": 9.48006230529595e-05,
      "loss": 0.03,
      "step": 1670
    },
    {
      "epoch": 5.23,
      "grad_norm": 0.036615125834941864,
      "learning_rate": 9.476947040498443e-05,
      "loss": 0.0115,
      "step": 1680
    },
    {
      "epoch": 5.26,
      "grad_norm": 0.11661379039287567,
      "learning_rate": 9.473831775700936e-05,
      "loss": 0.01,
      "step": 1690
    },
    {
      "epoch": 5.3,
      "grad_norm": 0.16430974006652832,
      "learning_rate": 9.470716510903427e-05,
      "loss": 0.0109,
      "step": 1700
    },
    {
      "epoch": 5.3,
      "eval_accuracy": 0.8477808599167822,
      "eval_f1": 0.8378487316868508,
      "eval_loss": 0.8481851816177368,
      "eval_precision": 0.8383786998144745,
      "eval_recall": 0.8477808599167822,
      "eval_runtime": 39.3442,
      "eval_samples_per_second": 73.302,
      "eval_steps_per_second": 9.175,
      "step": 1700
    },
    {
      "epoch": 5.33,
      "grad_norm": 0.005747305229306221,
      "learning_rate": 9.467601246105919e-05,
      "loss": 0.0333,
      "step": 1710
    },
    {
      "epoch": 5.36,
      "grad_norm": 5.644034385681152,
      "learning_rate": 9.464485981308412e-05,
      "loss": 0.0148,
      "step": 1720
    },
    {
      "epoch": 5.39,
      "grad_norm": 0.002410849556326866,
      "learning_rate": 9.461370716510903e-05,
      "loss": 0.0518,
      "step": 1730
    },
    {
      "epoch": 5.42,
      "grad_norm": 0.002693226793780923,
      "learning_rate": 9.458255451713396e-05,
      "loss": 0.0508,
      "step": 1740
    },
    {
      "epoch": 5.45,
      "grad_norm": 10.301042556762695,
      "learning_rate": 9.455140186915888e-05,
      "loss": 0.071,
      "step": 1750
    },
    {
      "epoch": 5.48,
      "grad_norm": 0.06220326945185661,
      "learning_rate": 9.452024922118381e-05,
      "loss": 0.0568,
      "step": 1760
    },
    {
      "epoch": 5.51,
      "grad_norm": 0.07617553323507309,
      "learning_rate": 9.448909657320872e-05,
      "loss": 0.078,
      "step": 1770
    },
    {
      "epoch": 5.55,
      "grad_norm": 0.08313547819852829,
      "learning_rate": 9.445794392523365e-05,
      "loss": 0.0483,
      "step": 1780
    },
    {
      "epoch": 5.58,
      "grad_norm": 5.813291072845459,
      "learning_rate": 9.442679127725858e-05,
      "loss": 0.0369,
      "step": 1790
    },
    {
      "epoch": 5.61,
      "grad_norm": 6.673477649688721,
      "learning_rate": 9.43956386292835e-05,
      "loss": 0.0545,
      "step": 1800
    },
    {
      "epoch": 5.61,
      "eval_accuracy": 0.8498613037447988,
      "eval_f1": 0.8506454763787459,
      "eval_loss": 0.8046442270278931,
      "eval_precision": 0.8546625043916174,
      "eval_recall": 0.8498613037447988,
      "eval_runtime": 39.6817,
      "eval_samples_per_second": 72.678,
      "eval_steps_per_second": 9.097,
      "step": 1800
    },
    {
      "epoch": 5.61,
      "step": 1800,
      "total_flos": 2.2287694956200755e+18,
      "train_loss": 0.26933162180913817,
      "train_runtime": 1430.4649,
      "train_samples_per_second": 358.485,
      "train_steps_per_second": 22.44
    }
  ],
  "logging_steps": 10,
  "max_steps": 32100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 100,
  "save_steps": 100,
  "total_flos": 2.2287694956200755e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}