{
  "best_metric": 0.9347826086956522,
  "best_model_checkpoint": "vit-base-patch16-224-ve-U13b-R\\checkpoint-360",
  "epoch": 39.61165048543689,
  "eval_steps": 500,
  "global_step": 2040,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.19,
      "learning_rate": 5.392156862745098e-06,
      "loss": 1.3862,
      "step": 10
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.0784313725490196e-05,
      "loss": 1.3853,
      "step": 20
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.6176470588235296e-05,
      "loss": 1.3787,
      "step": 30
    },
    {
      "epoch": 0.78,
      "learning_rate": 2.156862745098039e-05,
      "loss": 1.3618,
      "step": 40
    },
    {
      "epoch": 0.97,
      "learning_rate": 2.696078431372549e-05,
      "loss": 1.3157,
      "step": 50
    },
    {
      "epoch": 0.99,
      "eval_accuracy": 0.34782608695652173,
      "eval_loss": 1.296690583229065,
      "eval_runtime": 0.977,
      "eval_samples_per_second": 47.083,
      "eval_steps_per_second": 6.141,
      "step": 51
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.235294117647059e-05,
      "loss": 1.2679,
      "step": 60
    },
    {
      "epoch": 1.36,
      "learning_rate": 3.774509803921569e-05,
      "loss": 1.1855,
      "step": 70
    },
    {
      "epoch": 1.55,
      "learning_rate": 4.313725490196078e-05,
      "loss": 1.1615,
      "step": 80
    },
    {
      "epoch": 1.75,
      "learning_rate": 4.8529411764705885e-05,
      "loss": 1.0505,
      "step": 90
    },
    {
      "epoch": 1.94,
      "learning_rate": 5.392156862745098e-05,
      "loss": 0.9801,
      "step": 100
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5869565217391305,
      "eval_loss": 0.9965764880180359,
      "eval_runtime": 0.7837,
      "eval_samples_per_second": 58.697,
      "eval_steps_per_second": 7.656,
      "step": 103
    },
    {
      "epoch": 2.14,
      "learning_rate": 5.477296181630547e-05,
      "loss": 0.95,
      "step": 110
    },
    {
      "epoch": 2.33,
      "learning_rate": 5.448916408668731e-05,
      "loss": 0.9522,
      "step": 120
    },
    {
      "epoch": 2.52,
      "learning_rate": 5.4205366357069146e-05,
      "loss": 0.8341,
      "step": 130
    },
    {
      "epoch": 2.72,
      "learning_rate": 5.392156862745098e-05,
      "loss": 0.7555,
      "step": 140
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.363777089783282e-05,
      "loss": 0.7385,
      "step": 150
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.717391304347826,
      "eval_loss": 0.7600191235542297,
      "eval_runtime": 0.7802,
      "eval_samples_per_second": 58.961,
      "eval_steps_per_second": 7.691,
      "step": 154
    },
    {
      "epoch": 3.11,
      "learning_rate": 5.3353973168214655e-05,
      "loss": 0.7079,
      "step": 160
    },
    {
      "epoch": 3.3,
      "learning_rate": 5.3070175438596496e-05,
      "loss": 0.5291,
      "step": 170
    },
    {
      "epoch": 3.5,
      "learning_rate": 5.278637770897833e-05,
      "loss": 0.5315,
      "step": 180
    },
    {
      "epoch": 3.69,
      "learning_rate": 5.2502579979360165e-05,
      "loss": 0.5752,
      "step": 190
    },
    {
      "epoch": 3.88,
      "learning_rate": 5.2218782249742006e-05,
      "loss": 0.572,
      "step": 200
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 0.6425113677978516,
      "eval_runtime": 0.7782,
      "eval_samples_per_second": 59.112,
      "eval_steps_per_second": 7.71,
      "step": 206
    },
    {
      "epoch": 4.08,
      "learning_rate": 5.193498452012384e-05,
      "loss": 0.4306,
      "step": 210
    },
    {
      "epoch": 4.27,
      "learning_rate": 5.165118679050568e-05,
      "loss": 0.4744,
      "step": 220
    },
    {
      "epoch": 4.47,
      "learning_rate": 5.1367389060887515e-05,
      "loss": 0.4993,
      "step": 230
    },
    {
      "epoch": 4.66,
      "learning_rate": 5.108359133126935e-05,
      "loss": 0.3998,
      "step": 240
    },
    {
      "epoch": 4.85,
      "learning_rate": 5.079979360165119e-05,
      "loss": 0.3646,
      "step": 250
    },
    {
      "epoch": 4.99,
      "eval_accuracy": 0.6956521739130435,
      "eval_loss": 0.768746554851532,
      "eval_runtime": 0.8062,
      "eval_samples_per_second": 57.057,
      "eval_steps_per_second": 7.442,
      "step": 257
    },
    {
      "epoch": 5.05,
      "learning_rate": 5.0515995872033025e-05,
      "loss": 0.327,
      "step": 260
    },
    {
      "epoch": 5.24,
      "learning_rate": 5.023219814241486e-05,
      "loss": 0.3172,
      "step": 270
    },
    {
      "epoch": 5.44,
      "learning_rate": 4.99484004127967e-05,
      "loss": 0.3136,
      "step": 280
    },
    {
      "epoch": 5.63,
      "learning_rate": 4.9664602683178534e-05,
      "loss": 0.4175,
      "step": 290
    },
    {
      "epoch": 5.83,
      "learning_rate": 4.9380804953560375e-05,
      "loss": 0.3033,
      "step": 300
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7391304347826086,
      "eval_loss": 0.6335754990577698,
      "eval_runtime": 0.7852,
      "eval_samples_per_second": 58.586,
      "eval_steps_per_second": 7.642,
      "step": 309
    },
    {
      "epoch": 6.02,
      "learning_rate": 4.909700722394221e-05,
      "loss": 0.2792,
      "step": 310
    },
    {
      "epoch": 6.21,
      "learning_rate": 4.8813209494324044e-05,
      "loss": 0.2537,
      "step": 320
    },
    {
      "epoch": 6.41,
      "learning_rate": 4.8529411764705885e-05,
      "loss": 0.3938,
      "step": 330
    },
    {
      "epoch": 6.6,
      "learning_rate": 4.824561403508772e-05,
      "loss": 0.3292,
      "step": 340
    },
    {
      "epoch": 6.8,
      "learning_rate": 4.796181630546956e-05,
      "loss": 0.2956,
      "step": 350
    },
    {
      "epoch": 6.99,
      "learning_rate": 4.7678018575851394e-05,
      "loss": 0.3073,
      "step": 360
    },
    {
      "epoch": 6.99,
      "eval_accuracy": 0.9347826086956522,
      "eval_loss": 0.35338208079338074,
      "eval_runtime": 0.7677,
      "eval_samples_per_second": 59.921,
      "eval_steps_per_second": 7.816,
      "step": 360
    },
    {
      "epoch": 7.18,
      "learning_rate": 4.739422084623323e-05,
      "loss": 0.2049,
      "step": 370
    },
    {
      "epoch": 7.38,
      "learning_rate": 4.711042311661507e-05,
      "loss": 0.2429,
      "step": 380
    },
    {
      "epoch": 7.57,
      "learning_rate": 4.6826625386996904e-05,
      "loss": 0.2084,
      "step": 390
    },
    {
      "epoch": 7.77,
      "learning_rate": 4.6542827657378745e-05,
      "loss": 0.3038,
      "step": 400
    },
    {
      "epoch": 7.96,
      "learning_rate": 4.625902992776058e-05,
      "loss": 0.1623,
      "step": 410
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6739130434782609,
      "eval_loss": 0.8559263348579407,
      "eval_runtime": 0.8107,
      "eval_samples_per_second": 56.742,
      "eval_steps_per_second": 7.401,
      "step": 412
    },
    {
      "epoch": 8.16,
      "learning_rate": 4.597523219814241e-05,
      "loss": 0.1462,
      "step": 420
    },
    {
      "epoch": 8.35,
      "learning_rate": 4.5691434468524254e-05,
      "loss": 0.124,
      "step": 430
    },
    {
      "epoch": 8.54,
      "learning_rate": 4.540763673890609e-05,
      "loss": 0.1477,
      "step": 440
    },
    {
      "epoch": 8.74,
      "learning_rate": 4.512383900928793e-05,
      "loss": 0.1319,
      "step": 450
    },
    {
      "epoch": 8.93,
      "learning_rate": 4.4840041279669764e-05,
      "loss": 0.1079,
      "step": 460
    },
    {
      "epoch": 8.99,
      "eval_accuracy": 0.7391304347826086,
      "eval_loss": 0.973010778427124,
      "eval_runtime": 0.8122,
      "eval_samples_per_second": 56.638,
      "eval_steps_per_second": 7.388,
      "step": 463
    },
    {
      "epoch": 9.13,
      "learning_rate": 4.45562435500516e-05,
      "loss": 0.1712,
      "step": 470
    },
    {
      "epoch": 9.32,
      "learning_rate": 4.427244582043344e-05,
      "loss": 0.1125,
      "step": 480
    },
    {
      "epoch": 9.51,
      "learning_rate": 4.398864809081527e-05,
      "loss": 0.1858,
      "step": 490
    },
    {
      "epoch": 9.71,
      "learning_rate": 4.3704850361197114e-05,
      "loss": 0.1768,
      "step": 500
    },
    {
      "epoch": 9.9,
      "learning_rate": 4.342105263157895e-05,
      "loss": 0.2703,
      "step": 510
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.7767899632453918,
      "eval_runtime": 0.7982,
      "eval_samples_per_second": 57.63,
      "eval_steps_per_second": 7.517,
      "step": 515
    },
    {
      "epoch": 10.1,
      "learning_rate": 4.313725490196078e-05,
      "loss": 0.0645,
      "step": 520
    },
    {
      "epoch": 10.29,
      "learning_rate": 4.2853457172342624e-05,
      "loss": 0.1355,
      "step": 530
    },
    {
      "epoch": 10.49,
      "learning_rate": 4.2569659442724465e-05,
      "loss": 0.2119,
      "step": 540
    },
    {
      "epoch": 10.68,
      "learning_rate": 4.22858617131063e-05,
      "loss": 0.2935,
      "step": 550
    },
    {
      "epoch": 10.87,
      "learning_rate": 4.200206398348813e-05,
      "loss": 0.178,
      "step": 560
    },
    {
      "epoch": 10.99,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 0.8520329594612122,
      "eval_runtime": 0.7822,
      "eval_samples_per_second": 58.809,
      "eval_steps_per_second": 7.671,
      "step": 566
    },
    {
      "epoch": 11.07,
      "learning_rate": 4.171826625386997e-05,
      "loss": 0.1782,
      "step": 570
    },
    {
      "epoch": 11.26,
      "learning_rate": 4.143446852425181e-05,
      "loss": 0.1341,
      "step": 580
    },
    {
      "epoch": 11.46,
      "learning_rate": 4.115067079463365e-05,
      "loss": 0.0238,
      "step": 590
    },
    {
      "epoch": 11.65,
      "learning_rate": 4.0866873065015484e-05,
      "loss": 0.1414,
      "step": 600
    },
    {
      "epoch": 11.84,
      "learning_rate": 4.058307533539732e-05,
      "loss": 0.2191,
      "step": 610
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7391304347826086,
      "eval_loss": 1.0049434900283813,
      "eval_runtime": 0.7802,
      "eval_samples_per_second": 58.961,
      "eval_steps_per_second": 7.691,
      "step": 618
    },
    {
      "epoch": 12.04,
      "learning_rate": 4.029927760577915e-05,
      "loss": 0.1217,
      "step": 620
    },
    {
      "epoch": 12.23,
      "learning_rate": 4.001547987616099e-05,
      "loss": 0.1584,
      "step": 630
    },
    {
      "epoch": 12.43,
      "learning_rate": 3.9731682146542834e-05,
      "loss": 0.1076,
      "step": 640
    },
    {
      "epoch": 12.62,
      "learning_rate": 3.944788441692467e-05,
      "loss": 0.0764,
      "step": 650
    },
    {
      "epoch": 12.82,
      "learning_rate": 3.91640866873065e-05,
      "loss": 0.0597,
      "step": 660
    },
    {
      "epoch": 12.99,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 0.833352267742157,
      "eval_runtime": 0.7732,
      "eval_samples_per_second": 59.495,
      "eval_steps_per_second": 7.76,
      "step": 669
    },
    {
      "epoch": 13.01,
      "learning_rate": 3.888028895768834e-05,
      "loss": 0.071,
      "step": 670
    },
    {
      "epoch": 13.2,
      "learning_rate": 3.859649122807018e-05,
      "loss": 0.0797,
      "step": 680
    },
    {
      "epoch": 13.4,
      "learning_rate": 3.831269349845202e-05,
      "loss": 0.1046,
      "step": 690
    },
    {
      "epoch": 13.59,
      "learning_rate": 3.802889576883385e-05,
      "loss": 0.1188,
      "step": 700
    },
    {
      "epoch": 13.79,
      "learning_rate": 3.774509803921569e-05,
      "loss": 0.0627,
      "step": 710
    },
    {
      "epoch": 13.98,
      "learning_rate": 3.746130030959752e-05,
      "loss": 0.0881,
      "step": 720
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 0.9984623193740845,
      "eval_runtime": 0.7912,
      "eval_samples_per_second": 58.141,
      "eval_steps_per_second": 7.584,
      "step": 721
    },
    {
      "epoch": 14.17,
      "learning_rate": 3.7177502579979356e-05,
      "loss": 0.1591,
      "step": 730
    },
    {
      "epoch": 14.37,
      "learning_rate": 3.6893704850361204e-05,
      "loss": 0.1445,
      "step": 740
    },
    {
      "epoch": 14.56,
      "learning_rate": 3.660990712074304e-05,
      "loss": 0.0678,
      "step": 750
    },
    {
      "epoch": 14.76,
      "learning_rate": 3.632610939112487e-05,
      "loss": 0.0799,
      "step": 760
    },
    {
      "epoch": 14.95,
      "learning_rate": 3.6042311661506706e-05,
      "loss": 0.1265,
      "step": 770
    },
    {
      "epoch": 14.99,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.9443035125732422,
      "eval_runtime": 0.8091,
      "eval_samples_per_second": 56.851,
      "eval_steps_per_second": 7.415,
      "step": 772
    },
    {
      "epoch": 15.15,
      "learning_rate": 3.575851393188854e-05,
      "loss": 0.1045,
      "step": 780
    },
    {
      "epoch": 15.34,
      "learning_rate": 3.547471620227039e-05,
      "loss": 0.1274,
      "step": 790
    },
    {
      "epoch": 15.53,
      "learning_rate": 3.519091847265222e-05,
      "loss": 0.105,
      "step": 800
    },
    {
      "epoch": 15.73,
      "learning_rate": 3.490712074303406e-05,
      "loss": 0.1277,
      "step": 810
    },
    {
      "epoch": 15.92,
      "learning_rate": 3.462332301341589e-05,
      "loss": 0.0696,
      "step": 820
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8260869565217391,
      "eval_loss": 0.9877853393554688,
      "eval_runtime": 0.808,
      "eval_samples_per_second": 56.928,
      "eval_steps_per_second": 7.425,
      "step": 824
    },
    {
      "epoch": 16.12,
      "learning_rate": 3.4339525283797725e-05,
      "loss": 0.1507,
      "step": 830
    },
    {
      "epoch": 16.31,
      "learning_rate": 3.405572755417957e-05,
      "loss": 0.1025,
      "step": 840
    },
    {
      "epoch": 16.5,
      "learning_rate": 3.377192982456141e-05,
      "loss": 0.0113,
      "step": 850
    },
    {
      "epoch": 16.7,
      "learning_rate": 3.348813209494324e-05,
      "loss": 0.0393,
      "step": 860
    },
    {
      "epoch": 16.89,
      "learning_rate": 3.3204334365325076e-05,
      "loss": 0.1198,
      "step": 870
    },
    {
      "epoch": 16.99,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.8784380555152893,
      "eval_runtime": 0.8375,
      "eval_samples_per_second": 54.923,
      "eval_steps_per_second": 7.164,
      "step": 875
    },
    {
      "epoch": 17.09,
      "learning_rate": 3.292053663570691e-05,
      "loss": 0.1388,
      "step": 880
    },
    {
      "epoch": 17.28,
      "learning_rate": 3.263673890608876e-05,
      "loss": 0.1075,
      "step": 890
    },
    {
      "epoch": 17.48,
      "learning_rate": 3.235294117647059e-05,
      "loss": 0.0449,
      "step": 900
    },
    {
      "epoch": 17.67,
      "learning_rate": 3.2069143446852426e-05,
      "loss": 0.0916,
      "step": 910
    },
    {
      "epoch": 17.86,
      "learning_rate": 3.178534571723426e-05,
      "loss": 0.1484,
      "step": 920
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 0.9595004320144653,
      "eval_runtime": 0.7957,
      "eval_samples_per_second": 57.814,
      "eval_steps_per_second": 7.541,
      "step": 927
    },
    {
      "epoch": 18.06,
      "learning_rate": 3.1501547987616095e-05,
      "loss": 0.0764,
      "step": 930
    },
    {
      "epoch": 18.25,
      "learning_rate": 3.121775025799794e-05,
      "loss": 0.0681,
      "step": 940
    },
    {
      "epoch": 18.45,
      "learning_rate": 3.093395252837978e-05,
      "loss": 0.064,
      "step": 950
    },
    {
      "epoch": 18.64,
      "learning_rate": 3.065015479876161e-05,
      "loss": 0.067,
      "step": 960
    },
    {
      "epoch": 18.83,
      "learning_rate": 3.0366357069143445e-05,
      "loss": 0.2887,
      "step": 970
    },
    {
      "epoch": 18.99,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 1.0562622547149658,
      "eval_runtime": 0.8164,
      "eval_samples_per_second": 56.345,
      "eval_steps_per_second": 7.349,
      "step": 978
    },
    {
      "epoch": 19.03,
      "learning_rate": 3.0082559339525283e-05,
      "loss": 0.1123,
      "step": 980
    },
    {
      "epoch": 19.22,
      "learning_rate": 2.9798761609907124e-05,
      "loss": 0.0329,
      "step": 990
    },
    {
      "epoch": 19.42,
      "learning_rate": 2.9514963880288958e-05,
      "loss": 0.0561,
      "step": 1000
    },
    {
      "epoch": 19.61,
      "learning_rate": 2.9231166150670796e-05,
      "loss": 0.1581,
      "step": 1010
    },
    {
      "epoch": 19.81,
      "learning_rate": 2.894736842105263e-05,
      "loss": 0.1128,
      "step": 1020
    },
    {
      "epoch": 20.0,
      "learning_rate": 2.8663570691434468e-05,
      "loss": 0.1423,
      "step": 1030
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 0.8549535870552063,
      "eval_runtime": 0.8176,
      "eval_samples_per_second": 56.263,
      "eval_steps_per_second": 7.339,
      "step": 1030
    },
    {
      "epoch": 20.19,
      "learning_rate": 2.837977296181631e-05,
      "loss": 0.0869,
      "step": 1040
    },
    {
      "epoch": 20.39,
      "learning_rate": 2.8095975232198143e-05,
      "loss": 0.0625,
      "step": 1050
    },
    {
      "epoch": 20.58,
      "learning_rate": 2.781217750257998e-05,
      "loss": 0.136,
      "step": 1060
    },
    {
      "epoch": 20.78,
      "learning_rate": 2.7528379772961815e-05,
      "loss": 0.2042,
      "step": 1070
    },
    {
      "epoch": 20.97,
      "learning_rate": 2.7244582043343656e-05,
      "loss": 0.083,
      "step": 1080
    },
    {
      "epoch": 20.99,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 0.9092756509780884,
      "eval_runtime": 0.8048,
      "eval_samples_per_second": 57.154,
      "eval_steps_per_second": 7.455,
      "step": 1081
    },
    {
      "epoch": 21.17,
      "learning_rate": 2.696078431372549e-05,
      "loss": 0.0607,
      "step": 1090
    },
    {
      "epoch": 21.36,
      "learning_rate": 2.6676986584107328e-05,
      "loss": 0.0796,
      "step": 1100
    },
    {
      "epoch": 21.55,
      "learning_rate": 2.6393188854489165e-05,
      "loss": 0.0582,
      "step": 1110
    },
    {
      "epoch": 21.75,
      "learning_rate": 2.6109391124871003e-05,
      "loss": 0.0274,
      "step": 1120
    },
    {
      "epoch": 21.94,
      "learning_rate": 2.582559339525284e-05,
      "loss": 0.0695,
      "step": 1130
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.6739130434782609,
      "eval_loss": 1.2757601737976074,
      "eval_runtime": 0.8066,
      "eval_samples_per_second": 57.031,
      "eval_steps_per_second": 7.439,
      "step": 1133
    },
    {
      "epoch": 22.14,
      "learning_rate": 2.5541795665634675e-05,
      "loss": 0.0193,
      "step": 1140
    },
    {
      "epoch": 22.33,
      "learning_rate": 2.5257997936016512e-05,
      "loss": 0.0856,
      "step": 1150
    },
    {
      "epoch": 22.52,
      "learning_rate": 2.497420020639835e-05,
      "loss": 0.0969,
      "step": 1160
    },
    {
      "epoch": 22.72,
      "learning_rate": 2.4690402476780188e-05,
      "loss": 0.1214,
      "step": 1170
    },
    {
      "epoch": 22.91,
      "learning_rate": 2.4406604747162022e-05,
      "loss": 0.0285,
      "step": 1180
    },
    {
      "epoch": 22.99,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 1.0851949453353882,
      "eval_runtime": 0.8033,
      "eval_samples_per_second": 57.263,
      "eval_steps_per_second": 7.469,
      "step": 1184
    },
    {
      "epoch": 23.11,
      "learning_rate": 2.412280701754386e-05,
      "loss": 0.1273,
      "step": 1190
    },
    {
      "epoch": 23.3,
      "learning_rate": 2.3839009287925697e-05,
      "loss": 0.0777,
      "step": 1200
    },
    {
      "epoch": 23.5,
      "learning_rate": 2.3555211558307535e-05,
      "loss": 0.0507,
      "step": 1210
    },
    {
      "epoch": 23.69,
      "learning_rate": 2.3271413828689372e-05,
      "loss": 0.1113,
      "step": 1220
    },
    {
      "epoch": 23.88,
      "learning_rate": 2.2987616099071207e-05,
      "loss": 0.0132,
      "step": 1230
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.6956521739130435,
      "eval_loss": 1.3340740203857422,
      "eval_runtime": 0.8554,
      "eval_samples_per_second": 53.774,
      "eval_steps_per_second": 7.014,
      "step": 1236
    },
    {
      "epoch": 24.08,
      "learning_rate": 2.2703818369453044e-05,
      "loss": 0.054,
      "step": 1240
    },
    {
      "epoch": 24.27,
      "learning_rate": 2.2420020639834882e-05,
      "loss": 0.0702,
      "step": 1250
    },
    {
      "epoch": 24.47,
      "learning_rate": 2.213622291021672e-05,
      "loss": 0.0461,
      "step": 1260
    },
    {
      "epoch": 24.66,
      "learning_rate": 2.1852425180598557e-05,
      "loss": 0.0837,
      "step": 1270
    },
    {
      "epoch": 24.85,
      "learning_rate": 2.156862745098039e-05,
      "loss": 0.0957,
      "step": 1280
    },
    {
      "epoch": 24.99,
      "eval_accuracy": 0.7391304347826086,
      "eval_loss": 1.1964709758758545,
      "eval_runtime": 0.8164,
      "eval_samples_per_second": 56.344,
      "eval_steps_per_second": 7.349,
      "step": 1287
    },
    {
      "epoch": 25.05,
      "learning_rate": 2.1284829721362232e-05,
      "loss": 0.0968,
      "step": 1290
    },
    {
      "epoch": 25.24,
      "learning_rate": 2.1001031991744067e-05,
      "loss": 0.1526,
      "step": 1300
    },
    {
      "epoch": 25.44,
      "learning_rate": 2.0717234262125904e-05,
      "loss": 0.0653,
      "step": 1310
    },
    {
      "epoch": 25.63,
      "learning_rate": 2.0433436532507742e-05,
      "loss": 0.1046,
      "step": 1320
    },
    {
      "epoch": 25.83,
      "learning_rate": 2.0149638802889576e-05,
      "loss": 0.0633,
      "step": 1330
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 1.1199151277542114,
      "eval_runtime": 0.8717,
      "eval_samples_per_second": 52.769,
      "eval_steps_per_second": 6.883,
      "step": 1339
    },
    {
      "epoch": 26.02,
      "learning_rate": 1.9865841073271417e-05,
      "loss": 0.008,
      "step": 1340
    },
    {
      "epoch": 26.21,
      "learning_rate": 1.958204334365325e-05,
      "loss": 0.123,
      "step": 1350
    },
    {
      "epoch": 26.41,
      "learning_rate": 1.929824561403509e-05,
      "loss": 0.0715,
      "step": 1360
    },
    {
      "epoch": 26.6,
      "learning_rate": 1.9014447884416927e-05,
      "loss": 0.0765,
      "step": 1370
    },
    {
      "epoch": 26.8,
      "learning_rate": 1.873065015479876e-05,
      "loss": 0.0991,
      "step": 1380
    },
    {
      "epoch": 26.99,
      "learning_rate": 1.8446852425180602e-05,
      "loss": 0.0705,
      "step": 1390
    },
    {
      "epoch": 26.99,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 1.0550963878631592,
      "eval_runtime": 0.8012,
      "eval_samples_per_second": 57.414,
      "eval_steps_per_second": 7.489,
      "step": 1390
    },
    {
      "epoch": 27.18,
      "learning_rate": 1.8163054695562436e-05,
      "loss": 0.0501,
      "step": 1400
    },
    {
      "epoch": 27.38,
      "learning_rate": 1.787925696594427e-05,
      "loss": 0.0841,
      "step": 1410
    },
    {
      "epoch": 27.57,
      "learning_rate": 1.759545923632611e-05,
      "loss": 0.0257,
      "step": 1420
    },
    {
      "epoch": 27.77,
      "learning_rate": 1.7311661506707946e-05,
      "loss": 0.0027,
      "step": 1430
    },
    {
      "epoch": 27.96,
      "learning_rate": 1.7027863777089787e-05,
      "loss": 0.0564,
      "step": 1440
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.7391304347826086,
      "eval_loss": 1.4332396984100342,
      "eval_runtime": 0.8092,
      "eval_samples_per_second": 56.846,
      "eval_steps_per_second": 7.415,
      "step": 1442
    },
    {
      "epoch": 28.16,
      "learning_rate": 1.674406604747162e-05,
      "loss": 0.1458,
      "step": 1450
    },
    {
      "epoch": 28.35,
      "learning_rate": 1.6460268317853455e-05,
      "loss": 0.0965,
      "step": 1460
    },
    {
      "epoch": 28.54,
      "learning_rate": 1.6176470588235296e-05,
      "loss": 0.0443,
      "step": 1470
    },
    {
      "epoch": 28.74,
      "learning_rate": 1.589267285861713e-05,
      "loss": 0.0277,
      "step": 1480
    },
    {
      "epoch": 28.93,
      "learning_rate": 1.560887512899897e-05,
      "loss": 0.0798,
      "step": 1490
    },
    {
      "epoch": 28.99,
      "eval_accuracy": 0.7391304347826086,
      "eval_loss": 1.3854602575302124,
      "eval_runtime": 0.7817,
      "eval_samples_per_second": 58.847,
      "eval_steps_per_second": 7.676,
      "step": 1493
    },
    {
      "epoch": 29.13,
      "learning_rate": 1.5325077399380806e-05,
      "loss": 0.0701,
      "step": 1500
    },
    {
      "epoch": 29.32,
      "learning_rate": 1.5041279669762642e-05,
      "loss": 0.0716,
      "step": 1510
    },
    {
      "epoch": 29.51,
      "learning_rate": 1.4757481940144479e-05,
      "loss": 0.0463,
      "step": 1520
    },
    {
      "epoch": 29.71,
      "learning_rate": 1.4473684210526315e-05,
      "loss": 0.0273,
      "step": 1530
    },
    {
      "epoch": 29.9,
      "learning_rate": 1.4189886480908154e-05,
      "loss": 0.0326,
      "step": 1540
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 1.0533674955368042,
      "eval_runtime": 0.8902,
      "eval_samples_per_second": 51.674,
      "eval_steps_per_second": 6.74,
      "step": 1545
    },
    {
      "epoch": 30.1,
      "learning_rate": 1.390608875128999e-05,
      "loss": 0.0074,
      "step": 1550
    },
    {
      "epoch": 30.29,
      "learning_rate": 1.3622291021671828e-05,
      "loss": 0.0241,
      "step": 1560
    },
    {
      "epoch": 30.49,
      "learning_rate": 1.3338493292053664e-05,
      "loss": 0.089,
      "step": 1570
    },
    {
      "epoch": 30.68,
      "learning_rate": 1.3054695562435501e-05,
      "loss": 0.0401,
      "step": 1580
    },
    {
      "epoch": 30.87,
      "learning_rate": 1.2770897832817337e-05,
      "loss": 0.092,
      "step": 1590
    },
    {
      "epoch": 30.99,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 1.17453932762146,
      "eval_runtime": 0.8884,
      "eval_samples_per_second": 51.778,
      "eval_steps_per_second": 6.754,
      "step": 1596
    },
    {
      "epoch": 31.07,
      "learning_rate": 1.2487100103199175e-05,
      "loss": 0.0336,
      "step": 1600
    },
    {
      "epoch": 31.26,
      "learning_rate": 1.2203302373581011e-05,
      "loss": 0.0161,
      "step": 1610
    },
    {
      "epoch": 31.46,
      "learning_rate": 1.1919504643962849e-05,
      "loss": 0.0428,
      "step": 1620
    },
    {
      "epoch": 31.65,
      "learning_rate": 1.1635706914344686e-05,
      "loss": 0.0379,
      "step": 1630
    },
    {
      "epoch": 31.84,
      "learning_rate": 1.1351909184726522e-05,
      "loss": 0.1243,
      "step": 1640
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 1.1341125965118408,
      "eval_runtime": 0.8712,
      "eval_samples_per_second": 52.801,
      "eval_steps_per_second": 6.887,
      "step": 1648
    },
    {
      "epoch": 32.04,
      "learning_rate": 1.106811145510836e-05,
      "loss": 0.0214,
      "step": 1650
    },
    {
      "epoch": 32.23,
      "learning_rate": 1.0784313725490196e-05,
      "loss": 0.0496,
      "step": 1660
    },
    {
      "epoch": 32.43,
      "learning_rate": 1.0500515995872033e-05,
      "loss": 0.0483,
      "step": 1670
    },
    {
      "epoch": 32.62,
      "learning_rate": 1.0216718266253871e-05,
      "loss": 0.0087,
      "step": 1680
    },
    {
      "epoch": 32.82,
      "learning_rate": 9.932920536635709e-06,
      "loss": 0.062,
      "step": 1690
    },
    {
      "epoch": 32.99,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 1.2647993564605713,
      "eval_runtime": 0.8321,
      "eval_samples_per_second": 55.281,
      "eval_steps_per_second": 7.211,
      "step": 1699
    },
    {
      "epoch": 33.01,
      "learning_rate": 9.649122807017545e-06,
      "loss": 0.1248,
      "step": 1700
    },
    {
      "epoch": 33.2,
      "learning_rate": 9.36532507739938e-06,
      "loss": 0.0765,
      "step": 1710
    },
    {
      "epoch": 33.4,
      "learning_rate": 9.081527347781218e-06,
      "loss": 0.0796,
      "step": 1720
    },
    {
      "epoch": 33.59,
      "learning_rate": 8.797729618163056e-06,
      "loss": 0.0274,
      "step": 1730
    },
    {
      "epoch": 33.79,
      "learning_rate": 8.513931888544893e-06,
      "loss": 0.0525,
      "step": 1740
    },
    {
      "epoch": 33.98,
      "learning_rate": 8.230134158926728e-06,
      "loss": 0.0941,
      "step": 1750
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 1.1236172914505005,
      "eval_runtime": 0.8314,
      "eval_samples_per_second": 55.331,
      "eval_steps_per_second": 7.217,
      "step": 1751
    },
    {
      "epoch": 34.17,
      "learning_rate": 7.946336429308565e-06,
      "loss": 0.0313,
      "step": 1760
    },
    {
      "epoch": 34.37,
      "learning_rate": 7.662538699690403e-06,
      "loss": 0.046,
      "step": 1770
    },
    {
      "epoch": 34.56,
      "learning_rate": 7.3787409700722396e-06,
      "loss": 0.0394,
      "step": 1780
    },
    {
      "epoch": 34.76,
      "learning_rate": 7.094943240454077e-06,
      "loss": 0.0209,
      "step": 1790
    },
    {
      "epoch": 34.95,
      "learning_rate": 6.811145510835914e-06,
      "loss": 0.0119,
      "step": 1800
    },
    {
      "epoch": 34.99,
      "eval_accuracy": 0.8043478260869565,
      "eval_loss": 1.1302545070648193,
      "eval_runtime": 0.8257,
      "eval_samples_per_second": 55.711,
      "eval_steps_per_second": 7.267,
      "step": 1802
    },
    {
      "epoch": 35.15,
      "learning_rate": 6.527347781217751e-06,
      "loss": 0.0719,
      "step": 1810
    },
    {
      "epoch": 35.34,
      "learning_rate": 6.2435500515995875e-06,
      "loss": 0.0428,
      "step": 1820
    },
    {
      "epoch": 35.53,
      "learning_rate": 5.959752321981424e-06,
      "loss": 0.0183,
      "step": 1830
    },
    {
      "epoch": 35.73,
      "learning_rate": 5.675954592363261e-06,
      "loss": 0.027,
      "step": 1840
    },
    {
      "epoch": 35.92,
      "learning_rate": 5.392156862745098e-06,
      "loss": 0.044,
      "step": 1850
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 1.1848399639129639,
      "eval_runtime": 0.8214,
      "eval_samples_per_second": 56.0,
      "eval_steps_per_second": 7.304,
      "step": 1854
    },
    {
      "epoch": 36.12,
      "learning_rate": 5.1083591331269355e-06,
      "loss": 0.0312,
      "step": 1860
    },
    {
      "epoch": 36.31,
      "learning_rate": 4.824561403508772e-06,
      "loss": 0.0166,
      "step": 1870
    },
    {
      "epoch": 36.5,
      "learning_rate": 4.540763673890609e-06,
      "loss": 0.0951,
      "step": 1880
    },
    {
      "epoch": 36.7,
      "learning_rate": 4.256965944272447e-06,
      "loss": 0.1538,
      "step": 1890
    },
    {
      "epoch": 36.89,
      "learning_rate": 3.973168214654283e-06,
      "loss": 0.0073,
      "step": 1900
    },
    {
      "epoch": 36.99,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 1.1795865297317505,
      "eval_runtime": 0.8136,
      "eval_samples_per_second": 56.539,
      "eval_steps_per_second": 7.375,
      "step": 1905
    },
    {
      "epoch": 37.09,
      "learning_rate": 3.6893704850361198e-06,
      "loss": 0.0617,
      "step": 1910
    },
    {
      "epoch": 37.28,
      "learning_rate": 3.405572755417957e-06,
      "loss": 0.0482,
      "step": 1920
    },
    {
      "epoch": 37.48,
      "learning_rate": 3.1217750257997938e-06,
      "loss": 0.0452,
      "step": 1930
    },
    {
      "epoch": 37.67,
      "learning_rate": 2.8379772961816305e-06,
      "loss": 0.0511,
      "step": 1940
    },
    {
      "epoch": 37.86,
      "learning_rate": 2.5541795665634677e-06,
      "loss": 0.0149,
      "step": 1950
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 1.2491459846496582,
      "eval_runtime": 0.8526,
      "eval_samples_per_second": 53.952,
      "eval_steps_per_second": 7.037,
      "step": 1957
    },
    {
      "epoch": 38.06,
      "learning_rate": 2.2703818369453045e-06,
      "loss": 0.0527,
      "step": 1960
    },
    {
      "epoch": 38.25,
      "learning_rate": 1.9865841073271413e-06,
      "loss": 0.0693,
      "step": 1970
    },
    {
      "epoch": 38.45,
      "learning_rate": 1.7027863777089785e-06,
      "loss": 0.0776,
      "step": 1980
    },
    {
      "epoch": 38.64,
      "learning_rate": 1.4189886480908153e-06,
      "loss": 0.0413,
      "step": 1990
    },
    {
      "epoch": 38.83,
      "learning_rate": 1.1351909184726523e-06,
      "loss": 0.0194,
      "step": 2000
    },
    {
      "epoch": 38.99,
      "eval_accuracy": 0.782608695652174,
      "eval_loss": 1.1812182664871216,
      "eval_runtime": 0.8162,
      "eval_samples_per_second": 56.361,
      "eval_steps_per_second": 7.351,
      "step": 2008
    },
    {
      "epoch": 39.03,
      "learning_rate": 8.513931888544892e-07,
      "loss": 0.0202,
      "step": 2010
    },
    {
      "epoch": 39.22,
      "learning_rate": 5.675954592363261e-07,
      "loss": 0.0311,
      "step": 2020
    },
    {
      "epoch": 39.42,
      "learning_rate": 2.8379772961816306e-07,
      "loss": 0.037,
      "step": 2030
    },
    {
      "epoch": 39.61,
      "learning_rate": 0.0,
      "loss": 0.0577,
      "step": 2040
    },
    {
      "epoch": 39.61,
      "eval_accuracy": 0.7608695652173914,
      "eval_loss": 1.177699327468872,
      "eval_runtime": 0.8677,
      "eval_samples_per_second": 53.015,
      "eval_steps_per_second": 6.915,
      "step": 2040
    },
    {
      "epoch": 39.61,
      "step": 2040,
      "total_flos": 2.5142726714989363e+18,
      "train_loss": 0.19644491698172892,
      "train_runtime": 666.0117,
      "train_samples_per_second": 49.188,
      "train_steps_per_second": 3.063
    }
  ],
  "logging_steps": 10,
  "max_steps": 2040,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "total_flos": 2.5142726714989363e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}