{
  "best_metric": 0.6208920187793427,
  "best_model_checkpoint": "large-algae-vit-wirs/checkpoint-120",
  "epoch": 30.0,
  "global_step": 3600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "learning_rate": 1.388888888888889e-06,
      "loss": 1.3612,
      "step": 10
    },
    {
      "epoch": 0.17,
      "learning_rate": 2.777777777777778e-06,
      "loss": 1.3039,
      "step": 20
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.3076,
      "step": 30
    },
    {
      "epoch": 0.33,
      "learning_rate": 5.555555555555556e-06,
      "loss": 1.2573,
      "step": 40
    },
    {
      "epoch": 0.42,
      "learning_rate": 6.944444444444445e-06,
      "loss": 1.2478,
      "step": 50
    },
    {
      "epoch": 0.5,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.2011,
      "step": 60
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.722222222222223e-06,
      "loss": 1.1795,
      "step": 70
    },
    {
      "epoch": 0.67,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 1.1731,
      "step": 80
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.25e-05,
      "loss": 1.2165,
      "step": 90
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.388888888888889e-05,
      "loss": 1.2024,
      "step": 100
    },
    {
      "epoch": 0.92,
      "learning_rate": 1.527777777777778e-05,
      "loss": 1.1816,
      "step": 110
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.1662,
      "step": 120
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6208920187793427,
      "eval_loss": 0.9128149747848511,
      "eval_runtime": 23.9101,
      "eval_samples_per_second": 71.267,
      "eval_steps_per_second": 2.258,
      "step": 120
    },
    {
      "epoch": 1.08,
      "learning_rate": 1.8055555555555555e-05,
      "loss": 1.1644,
      "step": 130
    },
    {
      "epoch": 1.17,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 1.1601,
      "step": 140
    },
    {
      "epoch": 1.25,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.117,
      "step": 150
    },
    {
      "epoch": 1.33,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 1.1592,
      "step": 160
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.361111111111111e-05,
      "loss": 1.1402,
      "step": 170
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.5e-05,
      "loss": 1.1438,
      "step": 180
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 1.1215,
      "step": 190
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.777777777777778e-05,
      "loss": 1.1412,
      "step": 200
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.916666666666667e-05,
      "loss": 1.1488,
      "step": 210
    },
    {
      "epoch": 1.83,
      "learning_rate": 3.055555555555556e-05,
      "loss": 1.1669,
      "step": 220
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.194444444444444e-05,
      "loss": 1.12,
      "step": 230
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.0885,
      "step": 240
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.613849765258216,
      "eval_loss": 0.9469314217567444,
      "eval_runtime": 23.792,
      "eval_samples_per_second": 71.621,
      "eval_steps_per_second": 2.27,
      "step": 240
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.472222222222222e-05,
      "loss": 1.0651,
      "step": 250
    },
    {
      "epoch": 2.17,
      "learning_rate": 3.611111111111111e-05,
      "loss": 1.108,
      "step": 260
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.1074,
      "step": 270
    },
    {
      "epoch": 2.33,
      "learning_rate": 3.888888888888889e-05,
      "loss": 1.0908,
      "step": 280
    },
    {
      "epoch": 2.42,
      "learning_rate": 4.027777777777778e-05,
      "loss": 1.0659,
      "step": 290
    },
    {
      "epoch": 2.5,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.1009,
      "step": 300
    },
    {
      "epoch": 2.58,
      "learning_rate": 4.305555555555556e-05,
      "loss": 1.0968,
      "step": 310
    },
    {
      "epoch": 2.67,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 1.0859,
      "step": 320
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 1.1342,
      "step": 330
    },
    {
      "epoch": 2.83,
      "learning_rate": 4.722222222222222e-05,
      "loss": 1.0747,
      "step": 340
    },
    {
      "epoch": 2.92,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.0587,
      "step": 350
    },
    {
      "epoch": 3.0,
      "learning_rate": 5e-05,
      "loss": 1.1315,
      "step": 360
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5757042253521126,
      "eval_loss": 1.0919097661972046,
      "eval_runtime": 23.8029,
      "eval_samples_per_second": 71.588,
      "eval_steps_per_second": 2.269,
      "step": 360
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.984567901234568e-05,
      "loss": 1.0413,
      "step": 370
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.969135802469136e-05,
      "loss": 1.0521,
      "step": 380
    },
    {
      "epoch": 3.25,
      "learning_rate": 4.9537037037037035e-05,
      "loss": 1.0694,
      "step": 390
    },
    {
      "epoch": 3.33,
      "learning_rate": 4.938271604938271e-05,
      "loss": 1.0888,
      "step": 400
    },
    {
      "epoch": 3.42,
      "learning_rate": 4.92283950617284e-05,
      "loss": 1.0717,
      "step": 410
    },
    {
      "epoch": 3.5,
      "learning_rate": 4.9074074074074075e-05,
      "loss": 1.0258,
      "step": 420
    },
    {
      "epoch": 3.58,
      "learning_rate": 4.891975308641975e-05,
      "loss": 1.078,
      "step": 430
    },
    {
      "epoch": 3.67,
      "learning_rate": 4.876543209876544e-05,
      "loss": 1.0596,
      "step": 440
    },
    {
      "epoch": 3.75,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 1.059,
      "step": 450
    },
    {
      "epoch": 3.83,
      "learning_rate": 4.845679012345679e-05,
      "loss": 1.0505,
      "step": 460
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.830246913580247e-05,
      "loss": 1.0184,
      "step": 470
    },
    {
      "epoch": 4.0,
      "learning_rate": 4.814814814814815e-05,
      "loss": 1.0542,
      "step": 480
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5598591549295775,
      "eval_loss": 1.229072093963623,
      "eval_runtime": 24.2826,
      "eval_samples_per_second": 70.174,
      "eval_steps_per_second": 2.224,
      "step": 480
    },
    {
      "epoch": 4.08,
      "learning_rate": 4.799382716049383e-05,
      "loss": 1.0394,
      "step": 490
    },
    {
      "epoch": 4.17,
      "learning_rate": 4.783950617283951e-05,
      "loss": 1.0383,
      "step": 500
    },
    {
      "epoch": 4.25,
      "learning_rate": 4.768518518518519e-05,
      "loss": 1.0385,
      "step": 510
    },
    {
      "epoch": 4.33,
      "learning_rate": 4.7530864197530866e-05,
      "loss": 1.0761,
      "step": 520
    },
    {
      "epoch": 4.42,
      "learning_rate": 4.7376543209876543e-05,
      "loss": 1.0378,
      "step": 530
    },
    {
      "epoch": 4.5,
      "learning_rate": 4.722222222222222e-05,
      "loss": 1.0378,
      "step": 540
    },
    {
      "epoch": 4.58,
      "learning_rate": 4.70679012345679e-05,
      "loss": 1.0333,
      "step": 550
    },
    {
      "epoch": 4.67,
      "learning_rate": 4.691358024691358e-05,
      "loss": 1.0729,
      "step": 560
    },
    {
      "epoch": 4.75,
      "learning_rate": 4.675925925925926e-05,
      "loss": 1.0342,
      "step": 570
    },
    {
      "epoch": 4.83,
      "learning_rate": 4.6604938271604945e-05,
      "loss": 1.0316,
      "step": 580
    },
    {
      "epoch": 4.92,
      "learning_rate": 4.645061728395062e-05,
      "loss": 1.0213,
      "step": 590
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.62962962962963e-05,
      "loss": 1.028,
      "step": 600
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5598591549295775,
      "eval_loss": 1.1931029558181763,
      "eval_runtime": 24.0021,
      "eval_samples_per_second": 70.994,
      "eval_steps_per_second": 2.25,
      "step": 600
    },
    {
      "epoch": 5.08,
      "learning_rate": 4.614197530864198e-05,
      "loss": 1.0657,
      "step": 610
    },
    {
      "epoch": 5.17,
      "learning_rate": 4.5987654320987656e-05,
      "loss": 1.0504,
      "step": 620
    },
    {
      "epoch": 5.25,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 1.0345,
      "step": 630
    },
    {
      "epoch": 5.33,
      "learning_rate": 4.567901234567901e-05,
      "loss": 1.0113,
      "step": 640
    },
    {
      "epoch": 5.42,
      "learning_rate": 4.5524691358024696e-05,
      "loss": 1.0316,
      "step": 650
    },
    {
      "epoch": 5.5,
      "learning_rate": 4.5370370370370374e-05,
      "loss": 1.0287,
      "step": 660
    },
    {
      "epoch": 5.58,
      "learning_rate": 4.521604938271605e-05,
      "loss": 1.064,
      "step": 670
    },
    {
      "epoch": 5.67,
      "learning_rate": 4.506172839506173e-05,
      "loss": 1.0079,
      "step": 680
    },
    {
      "epoch": 5.75,
      "learning_rate": 4.490740740740741e-05,
      "loss": 1.007,
      "step": 690
    },
    {
      "epoch": 5.83,
      "learning_rate": 4.4753086419753084e-05,
      "loss": 0.9965,
      "step": 700
    },
    {
      "epoch": 5.92,
      "learning_rate": 4.459876543209877e-05,
      "loss": 1.035,
      "step": 710
    },
    {
      "epoch": 6.0,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 1.0023,
      "step": 720
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5674882629107981,
      "eval_loss": 1.1548198461532593,
      "eval_runtime": 23.6784,
      "eval_samples_per_second": 71.964,
      "eval_steps_per_second": 2.281,
      "step": 720
    },
    {
      "epoch": 6.08,
      "learning_rate": 4.429012345679013e-05,
      "loss": 1.033,
      "step": 730
    },
    {
      "epoch": 6.17,
      "learning_rate": 4.413580246913581e-05,
      "loss": 1.0003,
      "step": 740
    },
    {
      "epoch": 6.25,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 1.0196,
      "step": 750
    },
    {
      "epoch": 6.33,
      "learning_rate": 4.3827160493827164e-05,
      "loss": 1.0211,
      "step": 760
    },
    {
      "epoch": 6.42,
      "learning_rate": 4.367283950617284e-05,
      "loss": 0.9964,
      "step": 770
    },
    {
      "epoch": 6.5,
      "learning_rate": 4.351851851851852e-05,
      "loss": 1.0339,
      "step": 780
    },
    {
      "epoch": 6.58,
      "learning_rate": 4.33641975308642e-05,
      "loss": 0.9788,
      "step": 790
    },
    {
      "epoch": 6.67,
      "learning_rate": 4.3209876543209875e-05,
      "loss": 1.0567,
      "step": 800
    },
    {
      "epoch": 6.75,
      "learning_rate": 4.305555555555556e-05,
      "loss": 0.994,
      "step": 810
    },
    {
      "epoch": 6.83,
      "learning_rate": 4.290123456790124e-05,
      "loss": 1.0389,
      "step": 820
    },
    {
      "epoch": 6.92,
      "learning_rate": 4.2746913580246915e-05,
      "loss": 0.9833,
      "step": 830
    },
    {
      "epoch": 7.0,
      "learning_rate": 4.259259259259259e-05,
      "loss": 1.0176,
      "step": 840
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5757042253521126,
      "eval_loss": 1.0932233333587646,
      "eval_runtime": 23.5291,
      "eval_samples_per_second": 72.421,
      "eval_steps_per_second": 2.295,
      "step": 840
    },
    {
      "epoch": 7.08,
      "learning_rate": 4.243827160493827e-05,
      "loss": 1.021,
      "step": 850
    },
    {
      "epoch": 7.17,
      "learning_rate": 4.2283950617283955e-05,
      "loss": 0.9636,
      "step": 860
    },
    {
      "epoch": 7.25,
      "learning_rate": 4.212962962962963e-05,
      "loss": 0.9965,
      "step": 870
    },
    {
      "epoch": 7.33,
      "learning_rate": 4.197530864197531e-05,
      "loss": 0.9829,
      "step": 880
    },
    {
      "epoch": 7.42,
      "learning_rate": 4.1820987654320994e-05,
      "loss": 1.0139,
      "step": 890
    },
    {
      "epoch": 7.5,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.0446,
      "step": 900
    },
    {
      "epoch": 7.58,
      "learning_rate": 4.151234567901235e-05,
      "loss": 0.9917,
      "step": 910
    },
    {
      "epoch": 7.67,
      "learning_rate": 4.135802469135803e-05,
      "loss": 0.9904,
      "step": 920
    },
    {
      "epoch": 7.75,
      "learning_rate": 4.1203703703703705e-05,
      "loss": 1.0068,
      "step": 930
    },
    {
      "epoch": 7.83,
      "learning_rate": 4.104938271604938e-05,
      "loss": 0.974,
      "step": 940
    },
    {
      "epoch": 7.92,
      "learning_rate": 4.089506172839506e-05,
      "loss": 1.0291,
      "step": 950
    },
    {
      "epoch": 8.0,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.992,
      "step": 960
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5751173708920188,
      "eval_loss": 1.1386847496032715,
      "eval_runtime": 23.4444,
      "eval_samples_per_second": 72.683,
      "eval_steps_per_second": 2.303,
      "step": 960
    },
    {
      "epoch": 8.08,
      "learning_rate": 4.058641975308642e-05,
      "loss": 0.9983,
      "step": 970
    },
    {
      "epoch": 8.17,
      "learning_rate": 4.04320987654321e-05,
      "loss": 0.964,
      "step": 980
    },
    {
      "epoch": 8.25,
      "learning_rate": 4.027777777777778e-05,
      "loss": 0.9514,
      "step": 990
    },
    {
      "epoch": 8.33,
      "learning_rate": 4.012345679012346e-05,
      "loss": 1.003,
      "step": 1000
    },
    {
      "epoch": 8.42,
      "learning_rate": 3.996913580246914e-05,
      "loss": 0.9575,
      "step": 1010
    },
    {
      "epoch": 8.5,
      "learning_rate": 3.981481481481482e-05,
      "loss": 0.9965,
      "step": 1020
    },
    {
      "epoch": 8.58,
      "learning_rate": 3.9660493827160496e-05,
      "loss": 1.0445,
      "step": 1030
    },
    {
      "epoch": 8.67,
      "learning_rate": 3.950617283950617e-05,
      "loss": 0.9638,
      "step": 1040
    },
    {
      "epoch": 8.75,
      "learning_rate": 3.935185185185186e-05,
      "loss": 1.024,
      "step": 1050
    },
    {
      "epoch": 8.83,
      "learning_rate": 3.9197530864197535e-05,
      "loss": 1.0393,
      "step": 1060
    },
    {
      "epoch": 8.92,
      "learning_rate": 3.904320987654321e-05,
      "loss": 0.9897,
      "step": 1070
    },
    {
      "epoch": 9.0,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.9891,
      "step": 1080
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.5463615023474179,
      "eval_loss": 1.238718867301941,
      "eval_runtime": 23.4716,
      "eval_samples_per_second": 72.598,
      "eval_steps_per_second": 2.301,
      "step": 1080
    },
    {
      "epoch": 9.08,
      "learning_rate": 3.873456790123457e-05,
      "loss": 1.0242,
      "step": 1090
    },
    {
      "epoch": 9.17,
      "learning_rate": 3.8580246913580246e-05,
      "loss": 0.9982,
      "step": 1100
    },
    {
      "epoch": 9.25,
      "learning_rate": 3.8425925925925924e-05,
      "loss": 0.9985,
      "step": 1110
    },
    {
      "epoch": 9.33,
      "learning_rate": 3.82716049382716e-05,
      "loss": 1.0509,
      "step": 1120
    },
    {
      "epoch": 9.42,
      "learning_rate": 3.8117283950617286e-05,
      "loss": 0.9643,
      "step": 1130
    },
    {
      "epoch": 9.5,
      "learning_rate": 3.7962962962962964e-05,
      "loss": 0.9595,
      "step": 1140
    },
    {
      "epoch": 9.58,
      "learning_rate": 3.780864197530865e-05,
      "loss": 1.0199,
      "step": 1150
    },
    {
      "epoch": 9.67,
      "learning_rate": 3.7654320987654326e-05,
      "loss": 0.948,
      "step": 1160
    },
    {
      "epoch": 9.75,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.9953,
      "step": 1170
    },
    {
      "epoch": 9.83,
      "learning_rate": 3.734567901234568e-05,
      "loss": 0.9848,
      "step": 1180
    },
    {
      "epoch": 9.92,
      "learning_rate": 3.719135802469136e-05,
      "loss": 0.9249,
      "step": 1190
    },
    {
      "epoch": 10.0,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.9635,
      "step": 1200
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.5428403755868545,
      "eval_loss": 1.3771986961364746,
      "eval_runtime": 23.4919,
      "eval_samples_per_second": 72.536,
      "eval_steps_per_second": 2.299,
      "step": 1200
    },
    {
      "epoch": 10.08,
      "learning_rate": 3.6882716049382714e-05,
      "loss": 0.9894,
      "step": 1210
    },
    {
      "epoch": 10.17,
      "learning_rate": 3.67283950617284e-05,
      "loss": 1.0025,
      "step": 1220
    },
    {
      "epoch": 10.25,
      "learning_rate": 3.6574074074074076e-05,
      "loss": 0.9747,
      "step": 1230
    },
    {
      "epoch": 10.33,
      "learning_rate": 3.6419753086419754e-05,
      "loss": 0.9813,
      "step": 1240
    },
    {
      "epoch": 10.42,
      "learning_rate": 3.626543209876543e-05,
      "loss": 1.0371,
      "step": 1250
    },
    {
      "epoch": 10.5,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.9673,
      "step": 1260
    },
    {
      "epoch": 10.58,
      "learning_rate": 3.5956790123456794e-05,
      "loss": 0.9378,
      "step": 1270
    },
    {
      "epoch": 10.67,
      "learning_rate": 3.580246913580247e-05,
      "loss": 0.9659,
      "step": 1280
    },
    {
      "epoch": 10.75,
      "learning_rate": 3.564814814814815e-05,
      "loss": 0.9527,
      "step": 1290
    },
    {
      "epoch": 10.83,
      "learning_rate": 3.5493827160493834e-05,
      "loss": 0.9805,
      "step": 1300
    },
    {
      "epoch": 10.92,
      "learning_rate": 3.533950617283951e-05,
      "loss": 0.9974,
      "step": 1310
    },
    {
      "epoch": 11.0,
      "learning_rate": 3.518518518518519e-05,
      "loss": 0.9764,
      "step": 1320
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.5258215962441315,
      "eval_loss": 1.4329094886779785,
      "eval_runtime": 23.2917,
      "eval_samples_per_second": 73.159,
      "eval_steps_per_second": 2.318,
      "step": 1320
    },
    {
      "epoch": 11.08,
      "learning_rate": 3.503086419753087e-05,
      "loss": 1.0079,
      "step": 1330
    },
    {
      "epoch": 11.17,
      "learning_rate": 3.4876543209876545e-05,
      "loss": 0.99,
      "step": 1340
    },
    {
      "epoch": 11.25,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.9836,
      "step": 1350
    },
    {
      "epoch": 11.33,
      "learning_rate": 3.45679012345679e-05,
      "loss": 0.9996,
      "step": 1360
    },
    {
      "epoch": 11.42,
      "learning_rate": 3.441358024691358e-05,
      "loss": 0.9794,
      "step": 1370
    },
    {
      "epoch": 11.5,
      "learning_rate": 3.425925925925926e-05,
      "loss": 0.9639,
      "step": 1380
    },
    {
      "epoch": 11.58,
      "learning_rate": 3.410493827160494e-05,
      "loss": 0.9945,
      "step": 1390
    },
    {
      "epoch": 11.67,
      "learning_rate": 3.395061728395062e-05,
      "loss": 0.9832,
      "step": 1400
    },
    {
      "epoch": 11.75,
      "learning_rate": 3.3796296296296295e-05,
      "loss": 0.9606,
      "step": 1410
    },
    {
      "epoch": 11.83,
      "learning_rate": 3.364197530864198e-05,
      "loss": 0.9581,
      "step": 1420
    },
    {
      "epoch": 11.92,
      "learning_rate": 3.348765432098766e-05,
      "loss": 0.9784,
      "step": 1430
    },
    {
      "epoch": 12.0,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.9375,
      "step": 1440
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.5522300469483568,
      "eval_loss": 1.2830231189727783,
      "eval_runtime": 23.496,
      "eval_samples_per_second": 72.523,
      "eval_steps_per_second": 2.298,
      "step": 1440
    },
    {
      "epoch": 12.08,
      "learning_rate": 3.317901234567901e-05,
      "loss": 0.9912,
      "step": 1450
    },
    {
      "epoch": 12.17,
      "learning_rate": 3.30246913580247e-05,
      "loss": 0.9722,
      "step": 1460
    },
    {
      "epoch": 12.25,
      "learning_rate": 3.2870370370370375e-05,
      "loss": 1.0089,
      "step": 1470
    },
    {
      "epoch": 12.33,
      "learning_rate": 3.271604938271605e-05,
      "loss": 0.954,
      "step": 1480
    },
    {
      "epoch": 12.42,
      "learning_rate": 3.256172839506173e-05,
      "loss": 0.9823,
      "step": 1490
    },
    {
      "epoch": 12.5,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.9813,
      "step": 1500
    },
    {
      "epoch": 12.58,
      "learning_rate": 3.2253086419753086e-05,
      "loss": 0.9827,
      "step": 1510
    },
    {
      "epoch": 12.67,
      "learning_rate": 3.209876543209876e-05,
      "loss": 0.9573,
      "step": 1520
    },
    {
      "epoch": 12.75,
      "learning_rate": 3.194444444444444e-05,
      "loss": 1.002,
      "step": 1530
    },
    {
      "epoch": 12.83,
      "learning_rate": 3.1790123456790125e-05,
      "loss": 0.9474,
      "step": 1540
    },
    {
      "epoch": 12.92,
      "learning_rate": 3.16358024691358e-05,
      "loss": 0.9435,
      "step": 1550
    },
    {
      "epoch": 13.0,
      "learning_rate": 3.148148148148148e-05,
      "loss": 0.9574,
      "step": 1560
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.522887323943662,
      "eval_loss": 1.400294542312622,
      "eval_runtime": 23.5276,
      "eval_samples_per_second": 72.426,
      "eval_steps_per_second": 2.295,
      "step": 1560
    },
    {
      "epoch": 13.08,
      "learning_rate": 3.1327160493827165e-05,
      "loss": 0.9727,
      "step": 1570
    },
    {
      "epoch": 13.17,
      "learning_rate": 3.117283950617284e-05,
      "loss": 0.9709,
      "step": 1580
    },
    {
      "epoch": 13.25,
      "learning_rate": 3.101851851851852e-05,
      "loss": 0.9753,
      "step": 1590
    },
    {
      "epoch": 13.33,
      "learning_rate": 3.08641975308642e-05,
      "loss": 0.9267,
      "step": 1600
    },
    {
      "epoch": 13.42,
      "learning_rate": 3.0709876543209876e-05,
      "loss": 0.9905,
      "step": 1610
    },
    {
      "epoch": 13.5,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.9499,
      "step": 1620
    },
    {
      "epoch": 13.58,
      "learning_rate": 3.0401234567901238e-05,
      "loss": 1.0304,
      "step": 1630
    },
    {
      "epoch": 13.67,
      "learning_rate": 3.0246913580246916e-05,
      "loss": 1.0021,
      "step": 1640
    },
    {
      "epoch": 13.75,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.9771,
      "step": 1650
    },
    {
      "epoch": 13.83,
      "learning_rate": 2.993827160493827e-05,
      "loss": 0.9278,
      "step": 1660
    },
    {
      "epoch": 13.92,
      "learning_rate": 2.9783950617283952e-05,
      "loss": 0.9555,
      "step": 1670
    },
    {
      "epoch": 14.0,
      "learning_rate": 2.962962962962963e-05,
      "loss": 0.9907,
      "step": 1680
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.5422535211267606,
      "eval_loss": 1.344726324081421,
      "eval_runtime": 23.9181,
      "eval_samples_per_second": 71.243,
      "eval_steps_per_second": 2.258,
      "step": 1680
    },
    {
      "epoch": 14.08,
      "learning_rate": 2.9475308641975308e-05,
      "loss": 0.9437,
      "step": 1690
    },
    {
      "epoch": 14.17,
      "learning_rate": 2.9320987654320992e-05,
      "loss": 0.9556,
      "step": 1700
    },
    {
      "epoch": 14.25,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.9587,
      "step": 1710
    },
    {
      "epoch": 14.33,
      "learning_rate": 2.9012345679012347e-05,
      "loss": 0.9754,
      "step": 1720
    },
    {
      "epoch": 14.42,
      "learning_rate": 2.8858024691358025e-05,
      "loss": 0.9724,
      "step": 1730
    },
    {
      "epoch": 14.5,
      "learning_rate": 2.8703703703703706e-05,
      "loss": 0.9733,
      "step": 1740
    },
    {
      "epoch": 14.58,
      "learning_rate": 2.8549382716049384e-05,
      "loss": 0.9725,
      "step": 1750
    },
    {
      "epoch": 14.67,
      "learning_rate": 2.839506172839506e-05,
      "loss": 0.9776,
      "step": 1760
    },
    {
      "epoch": 14.75,
      "learning_rate": 2.824074074074074e-05,
      "loss": 0.95,
      "step": 1770
    },
    {
      "epoch": 14.83,
      "learning_rate": 2.8086419753086424e-05,
      "loss": 1.006,
      "step": 1780
    },
    {
      "epoch": 14.92,
      "learning_rate": 2.79320987654321e-05,
      "loss": 0.9395,
      "step": 1790
    },
    {
      "epoch": 15.0,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.9507,
      "step": 1800
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.5604460093896714,
      "eval_loss": 1.290692925453186,
      "eval_runtime": 23.8671,
      "eval_samples_per_second": 71.395,
      "eval_steps_per_second": 2.263,
      "step": 1800
    },
    {
      "epoch": 15.08,
      "learning_rate": 2.762345679012346e-05,
      "loss": 0.9754,
      "step": 1810
    },
    {
      "epoch": 15.17,
      "learning_rate": 2.7469135802469138e-05,
      "loss": 0.9558,
      "step": 1820
    },
    {
      "epoch": 15.25,
      "learning_rate": 2.7314814814814816e-05,
      "loss": 0.9572,
      "step": 1830
    },
    {
      "epoch": 15.33,
      "learning_rate": 2.7160493827160493e-05,
      "loss": 0.9987,
      "step": 1840
    },
    {
      "epoch": 15.42,
      "learning_rate": 2.700617283950617e-05,
      "loss": 0.9429,
      "step": 1850
    },
    {
      "epoch": 15.5,
      "learning_rate": 2.6851851851851855e-05,
      "loss": 0.9161,
      "step": 1860
    },
    {
      "epoch": 15.58,
      "learning_rate": 2.6697530864197533e-05,
      "loss": 0.9153,
      "step": 1870
    },
    {
      "epoch": 15.67,
      "learning_rate": 2.654320987654321e-05,
      "loss": 1.0111,
      "step": 1880
    },
    {
      "epoch": 15.75,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 0.9889,
      "step": 1890
    },
    {
      "epoch": 15.83,
      "learning_rate": 2.623456790123457e-05,
      "loss": 0.9616,
      "step": 1900
    },
    {
      "epoch": 15.92,
      "learning_rate": 2.6080246913580247e-05,
      "loss": 0.9414,
      "step": 1910
    },
    {
      "epoch": 16.0,
      "learning_rate": 2.5925925925925925e-05,
      "loss": 0.9866,
      "step": 1920
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.5393192488262911,
      "eval_loss": 1.4577730894088745,
      "eval_runtime": 23.9597,
      "eval_samples_per_second": 71.119,
      "eval_steps_per_second": 2.254,
      "step": 1920
    },
    {
      "epoch": 16.08,
      "learning_rate": 2.5771604938271603e-05,
      "loss": 0.9638,
      "step": 1930
    },
    {
      "epoch": 16.17,
      "learning_rate": 2.5617283950617287e-05,
      "loss": 0.9562,
      "step": 1940
    },
    {
      "epoch": 16.25,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.9464,
      "step": 1950
    },
    {
      "epoch": 16.33,
      "learning_rate": 2.5308641975308646e-05,
      "loss": 0.9533,
      "step": 1960
    },
    {
      "epoch": 16.42,
      "learning_rate": 2.5154320987654324e-05,
      "loss": 0.9763,
      "step": 1970
    },
    {
      "epoch": 16.5,
      "learning_rate": 2.5e-05,
      "loss": 0.9795,
      "step": 1980
    },
    {
      "epoch": 16.58,
      "learning_rate": 2.484567901234568e-05,
      "loss": 0.9827,
      "step": 1990
    },
    {
      "epoch": 16.67,
      "learning_rate": 2.4691358024691357e-05,
      "loss": 0.9269,
      "step": 2000
    },
    {
      "epoch": 16.75,
      "learning_rate": 2.4537037037037038e-05,
      "loss": 0.9397,
      "step": 2010
    },
    {
      "epoch": 16.83,
      "learning_rate": 2.438271604938272e-05,
      "loss": 0.9639,
      "step": 2020
    },
    {
      "epoch": 16.92,
      "learning_rate": 2.4228395061728396e-05,
      "loss": 0.9626,
      "step": 2030
    },
    {
      "epoch": 17.0,
      "learning_rate": 2.4074074074074074e-05,
      "loss": 0.9297,
      "step": 2040
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.528169014084507,
      "eval_loss": 1.477900743484497,
      "eval_runtime": 23.9974,
      "eval_samples_per_second": 71.008,
      "eval_steps_per_second": 2.25,
      "step": 2040
    },
    {
      "epoch": 17.08,
      "learning_rate": 2.3919753086419755e-05,
      "loss": 0.9454,
      "step": 2050
    },
    {
      "epoch": 17.17,
      "learning_rate": 2.3765432098765433e-05,
      "loss": 0.9679,
      "step": 2060
    },
    {
      "epoch": 17.25,
      "learning_rate": 2.361111111111111e-05,
      "loss": 0.9748,
      "step": 2070
    },
    {
      "epoch": 17.33,
      "learning_rate": 2.345679012345679e-05,
      "loss": 0.9272,
      "step": 2080
    },
    {
      "epoch": 17.42,
      "learning_rate": 2.3302469135802473e-05,
      "loss": 0.9831,
      "step": 2090
    },
    {
      "epoch": 17.5,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.9593,
      "step": 2100
    },
    {
      "epoch": 17.58,
      "learning_rate": 2.2993827160493828e-05,
      "loss": 0.9239,
      "step": 2110
    },
    {
      "epoch": 17.67,
      "learning_rate": 2.2839506172839506e-05,
      "loss": 0.9255,
      "step": 2120
    },
    {
      "epoch": 17.75,
      "learning_rate": 2.2685185185185187e-05,
      "loss": 0.9606,
      "step": 2130
    },
    {
      "epoch": 17.83,
      "learning_rate": 2.2530864197530865e-05,
      "loss": 0.9407,
      "step": 2140
    },
    {
      "epoch": 17.92,
      "learning_rate": 2.2376543209876542e-05,
      "loss": 0.9578,
      "step": 2150
    },
    {
      "epoch": 18.0,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.9385,
      "step": 2160
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.5469483568075117,
      "eval_loss": 1.3874293565750122,
      "eval_runtime": 24.2732,
      "eval_samples_per_second": 70.201,
      "eval_steps_per_second": 2.225,
      "step": 2160
    },
    {
      "epoch": 18.08,
      "learning_rate": 2.2067901234567904e-05,
      "loss": 0.9642,
      "step": 2170
    },
    {
      "epoch": 18.17,
      "learning_rate": 2.1913580246913582e-05,
      "loss": 0.9263,
      "step": 2180
    },
    {
      "epoch": 18.25,
      "learning_rate": 2.175925925925926e-05,
      "loss": 0.9209,
      "step": 2190
    },
    {
      "epoch": 18.33,
      "learning_rate": 2.1604938271604937e-05,
      "loss": 0.9415,
      "step": 2200
    },
    {
      "epoch": 18.42,
      "learning_rate": 2.145061728395062e-05,
      "loss": 1.0009,
      "step": 2210
    },
    {
      "epoch": 18.5,
      "learning_rate": 2.1296296296296296e-05,
      "loss": 0.9318,
      "step": 2220
    },
    {
      "epoch": 18.58,
      "learning_rate": 2.1141975308641977e-05,
      "loss": 0.9256,
      "step": 2230
    },
    {
      "epoch": 18.67,
      "learning_rate": 2.0987654320987655e-05,
      "loss": 0.9422,
      "step": 2240
    },
    {
      "epoch": 18.75,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.9357,
      "step": 2250
    },
    {
      "epoch": 18.83,
      "learning_rate": 2.0679012345679014e-05,
      "loss": 0.9546,
      "step": 2260
    },
    {
      "epoch": 18.92,
      "learning_rate": 2.052469135802469e-05,
      "loss": 0.9144,
      "step": 2270
    },
    {
      "epoch": 19.0,
      "learning_rate": 2.037037037037037e-05,
      "loss": 0.9951,
      "step": 2280
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.5586854460093896,
      "eval_loss": 1.2975717782974243,
      "eval_runtime": 23.895,
      "eval_samples_per_second": 71.312,
      "eval_steps_per_second": 2.26,
      "step": 2280
    },
    {
      "epoch": 19.08,
      "learning_rate": 2.021604938271605e-05,
      "loss": 0.9519,
      "step": 2290
    },
    {
      "epoch": 19.17,
      "learning_rate": 2.006172839506173e-05,
      "loss": 0.9362,
      "step": 2300
    },
    {
      "epoch": 19.25,
      "learning_rate": 1.990740740740741e-05,
      "loss": 0.9226,
      "step": 2310
    },
    {
      "epoch": 19.33,
      "learning_rate": 1.9753086419753087e-05,
      "loss": 0.9176,
      "step": 2320
    },
    {
      "epoch": 19.42,
      "learning_rate": 1.9598765432098768e-05,
      "loss": 0.9464,
      "step": 2330
    },
    {
      "epoch": 19.5,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.9369,
      "step": 2340
    },
    {
      "epoch": 19.58,
      "learning_rate": 1.9290123456790123e-05,
      "loss": 0.9732,
      "step": 2350
    },
    {
      "epoch": 19.67,
      "learning_rate": 1.91358024691358e-05,
      "loss": 0.9639,
      "step": 2360
    },
    {
      "epoch": 19.75,
      "learning_rate": 1.8981481481481482e-05,
      "loss": 0.9821,
      "step": 2370
    },
    {
      "epoch": 19.83,
      "learning_rate": 1.8827160493827163e-05,
      "loss": 0.9472,
      "step": 2380
    },
    {
      "epoch": 19.92,
      "learning_rate": 1.867283950617284e-05,
      "loss": 0.9674,
      "step": 2390
    },
    {
      "epoch": 20.0,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.9794,
      "step": 2400
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.556924882629108,
      "eval_loss": 1.3109557628631592,
      "eval_runtime": 23.3371,
      "eval_samples_per_second": 73.017,
      "eval_steps_per_second": 2.314,
      "step": 2400
    },
    {
      "epoch": 20.08,
      "learning_rate": 1.83641975308642e-05,
      "loss": 0.959,
      "step": 2410
    },
    {
      "epoch": 20.17,
      "learning_rate": 1.8209876543209877e-05,
      "loss": 0.9208,
      "step": 2420
    },
    {
      "epoch": 20.25,
      "learning_rate": 1.8055555555555555e-05,
      "loss": 0.8842,
      "step": 2430
    },
    {
      "epoch": 20.33,
      "learning_rate": 1.7901234567901236e-05,
      "loss": 0.9423,
      "step": 2440
    },
    {
      "epoch": 20.42,
      "learning_rate": 1.7746913580246917e-05,
      "loss": 0.909,
      "step": 2450
    },
    {
      "epoch": 20.5,
      "learning_rate": 1.7592592592592595e-05,
      "loss": 0.93,
      "step": 2460
    },
    {
      "epoch": 20.58,
      "learning_rate": 1.7438271604938272e-05,
      "loss": 0.94,
      "step": 2470
    },
    {
      "epoch": 20.67,
      "learning_rate": 1.728395061728395e-05,
      "loss": 0.9199,
      "step": 2480
    },
    {
      "epoch": 20.75,
      "learning_rate": 1.712962962962963e-05,
      "loss": 0.9436,
      "step": 2490
    },
    {
      "epoch": 20.83,
      "learning_rate": 1.697530864197531e-05,
      "loss": 0.9872,
      "step": 2500
    },
    {
      "epoch": 20.92,
      "learning_rate": 1.682098765432099e-05,
      "loss": 0.9622,
      "step": 2510
    },
    {
      "epoch": 21.0,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.9974,
      "step": 2520
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.5275821596244131,
      "eval_loss": 1.3648896217346191,
      "eval_runtime": 23.4473,
      "eval_samples_per_second": 72.674,
      "eval_steps_per_second": 2.303,
      "step": 2520
    },
    {
      "epoch": 21.08,
      "learning_rate": 1.651234567901235e-05,
      "loss": 0.9337,
      "step": 2530
    },
    {
      "epoch": 21.17,
      "learning_rate": 1.6358024691358026e-05,
      "loss": 0.9923,
      "step": 2540
    },
    {
      "epoch": 21.25,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.9443,
      "step": 2550
    },
    {
      "epoch": 21.33,
      "learning_rate": 1.604938271604938e-05,
      "loss": 0.9586,
      "step": 2560
    },
    {
      "epoch": 21.42,
      "learning_rate": 1.5895061728395063e-05,
      "loss": 0.9377,
      "step": 2570
    },
    {
      "epoch": 21.5,
      "learning_rate": 1.574074074074074e-05,
      "loss": 0.9484,
      "step": 2580
    },
    {
      "epoch": 21.58,
      "learning_rate": 1.558641975308642e-05,
      "loss": 0.926,
      "step": 2590
    },
    {
      "epoch": 21.67,
      "learning_rate": 1.54320987654321e-05,
      "loss": 0.9426,
      "step": 2600
    },
    {
      "epoch": 21.75,
      "learning_rate": 1.527777777777778e-05,
      "loss": 0.9252,
      "step": 2610
    },
    {
      "epoch": 21.83,
      "learning_rate": 1.5123456790123458e-05,
      "loss": 0.9228,
      "step": 2620
    },
    {
      "epoch": 21.92,
      "learning_rate": 1.4969135802469136e-05,
      "loss": 0.9403,
      "step": 2630
    },
    {
      "epoch": 22.0,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 0.9284,
      "step": 2640
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.5363849765258216,
      "eval_loss": 1.3712799549102783,
      "eval_runtime": 23.3686,
      "eval_samples_per_second": 72.918,
      "eval_steps_per_second": 2.311,
      "step": 2640
    },
    {
      "epoch": 22.08,
      "learning_rate": 1.4660493827160496e-05,
      "loss": 0.9516,
      "step": 2650
    },
    {
      "epoch": 22.17,
      "learning_rate": 1.4506172839506174e-05,
      "loss": 0.932,
      "step": 2660
    },
    {
      "epoch": 22.25,
      "learning_rate": 1.4351851851851853e-05,
      "loss": 0.939,
      "step": 2670
    },
    {
      "epoch": 22.33,
      "learning_rate": 1.419753086419753e-05,
      "loss": 0.9343,
      "step": 2680
    },
    {
      "epoch": 22.42,
      "learning_rate": 1.4043209876543212e-05,
      "loss": 0.968,
      "step": 2690
    },
    {
      "epoch": 22.5,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.9169,
      "step": 2700
    },
    {
      "epoch": 22.58,
      "learning_rate": 1.3734567901234569e-05,
      "loss": 0.9549,
      "step": 2710
    },
    {
      "epoch": 22.67,
      "learning_rate": 1.3580246913580247e-05,
      "loss": 0.9106,
      "step": 2720
    },
    {
      "epoch": 22.75,
      "learning_rate": 1.3425925925925928e-05,
      "loss": 0.9678,
      "step": 2730
    },
    {
      "epoch": 22.83,
      "learning_rate": 1.3271604938271605e-05,
      "loss": 0.9296,
      "step": 2740
    },
    {
      "epoch": 22.92,
      "learning_rate": 1.3117283950617285e-05,
      "loss": 0.9827,
      "step": 2750
    },
    {
      "epoch": 23.0,
      "learning_rate": 1.2962962962962962e-05,
      "loss": 0.9144,
      "step": 2760
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.534037558685446,
      "eval_loss": 1.4117422103881836,
      "eval_runtime": 23.4898,
      "eval_samples_per_second": 72.542,
      "eval_steps_per_second": 2.299,
      "step": 2760
    },
    {
      "epoch": 23.08,
      "learning_rate": 1.2808641975308644e-05,
      "loss": 0.946,
      "step": 2770
    },
    {
      "epoch": 23.17,
      "learning_rate": 1.2654320987654323e-05,
      "loss": 0.9184,
      "step": 2780
    },
    {
      "epoch": 23.25,
      "learning_rate": 1.25e-05,
      "loss": 0.8966,
      "step": 2790
    },
    {
      "epoch": 23.33,
      "learning_rate": 1.2345679012345678e-05,
      "loss": 0.9174,
      "step": 2800
    },
    {
      "epoch": 23.42,
      "learning_rate": 1.219135802469136e-05,
      "loss": 0.9383,
      "step": 2810
    },
    {
      "epoch": 23.5,
      "learning_rate": 1.2037037037037037e-05,
      "loss": 0.942,
      "step": 2820
    },
    {
      "epoch": 23.58,
      "learning_rate": 1.1882716049382716e-05,
      "loss": 0.9137,
      "step": 2830
    },
    {
      "epoch": 23.67,
      "learning_rate": 1.1728395061728396e-05,
      "loss": 0.933,
      "step": 2840
    },
    {
      "epoch": 23.75,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.9348,
      "step": 2850
    },
    {
      "epoch": 23.83,
      "learning_rate": 1.1419753086419753e-05,
      "loss": 0.8839,
      "step": 2860
    },
    {
      "epoch": 23.92,
      "learning_rate": 1.1265432098765432e-05,
      "loss": 0.9548,
      "step": 2870
    },
    {
      "epoch": 24.0,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.9771,
      "step": 2880
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.5357981220657277,
      "eval_loss": 1.383649230003357,
      "eval_runtime": 23.4971,
      "eval_samples_per_second": 72.52,
      "eval_steps_per_second": 2.298,
      "step": 2880
    },
    {
      "epoch": 24.08,
      "learning_rate": 1.0956790123456791e-05,
      "loss": 0.9594,
      "step": 2890
    },
    {
      "epoch": 24.17,
      "learning_rate": 1.0802469135802469e-05,
      "loss": 0.9391,
      "step": 2900
    },
    {
      "epoch": 24.25,
      "learning_rate": 1.0648148148148148e-05,
      "loss": 0.9376,
      "step": 2910
    },
    {
      "epoch": 24.33,
      "learning_rate": 1.0493827160493827e-05,
      "loss": 0.9284,
      "step": 2920
    },
    {
      "epoch": 24.42,
      "learning_rate": 1.0339506172839507e-05,
      "loss": 0.9186,
      "step": 2930
    },
    {
      "epoch": 24.5,
      "learning_rate": 1.0185185185185185e-05,
      "loss": 0.8916,
      "step": 2940
    },
    {
      "epoch": 24.58,
      "learning_rate": 1.0030864197530866e-05,
      "loss": 0.9302,
      "step": 2950
    },
    {
      "epoch": 24.67,
      "learning_rate": 9.876543209876543e-06,
      "loss": 0.9572,
      "step": 2960
    },
    {
      "epoch": 24.75,
      "learning_rate": 9.722222222222223e-06,
      "loss": 0.8967,
      "step": 2970
    },
    {
      "epoch": 24.83,
      "learning_rate": 9.5679012345679e-06,
      "loss": 0.9554,
      "step": 2980
    },
    {
      "epoch": 24.92,
      "learning_rate": 9.413580246913581e-06,
      "loss": 0.9231,
      "step": 2990
    },
    {
      "epoch": 25.0,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.8994,
      "step": 3000
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.528169014084507,
      "eval_loss": 1.5077193975448608,
      "eval_runtime": 23.4692,
      "eval_samples_per_second": 72.606,
      "eval_steps_per_second": 2.301,
      "step": 3000
    },
    {
      "epoch": 25.08,
      "learning_rate": 9.104938271604939e-06,
      "loss": 0.9479,
      "step": 3010
    },
    {
      "epoch": 25.17,
      "learning_rate": 8.950617283950618e-06,
      "loss": 0.9191,
      "step": 3020
    },
    {
      "epoch": 25.25,
      "learning_rate": 8.796296296296297e-06,
      "loss": 0.9563,
      "step": 3030
    },
    {
      "epoch": 25.33,
      "learning_rate": 8.641975308641975e-06,
      "loss": 0.9932,
      "step": 3040
    },
    {
      "epoch": 25.42,
      "learning_rate": 8.487654320987654e-06,
      "loss": 0.9327,
      "step": 3050
    },
    {
      "epoch": 25.5,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.9056,
      "step": 3060
    },
    {
      "epoch": 25.58,
      "learning_rate": 8.179012345679013e-06,
      "loss": 0.9228,
      "step": 3070
    },
    {
      "epoch": 25.67,
      "learning_rate": 8.02469135802469e-06,
      "loss": 0.9059,
      "step": 3080
    },
    {
      "epoch": 25.75,
      "learning_rate": 7.87037037037037e-06,
      "loss": 0.9224,
      "step": 3090
    },
    {
      "epoch": 25.83,
      "learning_rate": 7.71604938271605e-06,
      "loss": 0.8891,
      "step": 3100
    },
    {
      "epoch": 25.92,
      "learning_rate": 7.561728395061729e-06,
      "loss": 0.9142,
      "step": 3110
    },
    {
      "epoch": 26.0,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.9061,
      "step": 3120
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.5328638497652582,
      "eval_loss": 1.4622243642807007,
      "eval_runtime": 23.8116,
      "eval_samples_per_second": 71.562,
      "eval_steps_per_second": 2.268,
      "step": 3120
    },
    {
      "epoch": 26.08,
      "learning_rate": 7.253086419753087e-06,
      "loss": 0.9487,
      "step": 3130
    },
    {
      "epoch": 26.17,
      "learning_rate": 7.098765432098765e-06,
      "loss": 0.9639,
      "step": 3140
    },
    {
      "epoch": 26.25,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.9248,
      "step": 3150
    },
    {
      "epoch": 26.33,
      "learning_rate": 6.790123456790123e-06,
      "loss": 0.901,
      "step": 3160
    },
    {
      "epoch": 26.42,
      "learning_rate": 6.635802469135803e-06,
      "loss": 0.9158,
      "step": 3170
    },
    {
      "epoch": 26.5,
      "learning_rate": 6.481481481481481e-06,
      "loss": 0.9161,
      "step": 3180
    },
    {
      "epoch": 26.58,
      "learning_rate": 6.3271604938271615e-06,
      "loss": 0.9121,
      "step": 3190
    },
    {
      "epoch": 26.67,
      "learning_rate": 6.172839506172839e-06,
      "loss": 0.9275,
      "step": 3200
    },
    {
      "epoch": 26.75,
      "learning_rate": 6.0185185185185185e-06,
      "loss": 0.9336,
      "step": 3210
    },
    {
      "epoch": 26.83,
      "learning_rate": 5.864197530864198e-06,
      "loss": 0.9263,
      "step": 3220
    },
    {
      "epoch": 26.92,
      "learning_rate": 5.7098765432098764e-06,
      "loss": 0.932,
      "step": 3230
    },
    {
      "epoch": 27.0,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.9071,
      "step": 3240
    },
    {
      "epoch": 27.0,
      "eval_accuracy": 0.5393192488262911,
      "eval_loss": 1.430330753326416,
      "eval_runtime": 23.403,
      "eval_samples_per_second": 72.811,
      "eval_steps_per_second": 2.307,
      "step": 3240
    },
    {
      "epoch": 27.08,
      "learning_rate": 5.401234567901234e-06,
      "loss": 0.9078,
      "step": 3250
    },
    {
      "epoch": 27.17,
      "learning_rate": 5.246913580246914e-06,
      "loss": 0.9895,
      "step": 3260
    },
    {
      "epoch": 27.25,
      "learning_rate": 5.092592592592592e-06,
      "loss": 0.8999,
      "step": 3270
    },
    {
      "epoch": 27.33,
      "learning_rate": 4.938271604938272e-06,
      "loss": 0.9041,
      "step": 3280
    },
    {
      "epoch": 27.42,
      "learning_rate": 4.78395061728395e-06,
      "loss": 0.958,
      "step": 3290
    },
    {
      "epoch": 27.5,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.9329,
      "step": 3300
    },
    {
      "epoch": 27.58,
      "learning_rate": 4.475308641975309e-06,
      "loss": 0.902,
      "step": 3310
    },
    {
      "epoch": 27.67,
      "learning_rate": 4.3209876543209875e-06,
      "loss": 0.9415,
      "step": 3320
    },
    {
      "epoch": 27.75,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.8901,
      "step": 3330
    },
    {
      "epoch": 27.83,
      "learning_rate": 4.012345679012345e-06,
      "loss": 0.9495,
      "step": 3340
    },
    {
      "epoch": 27.92,
      "learning_rate": 3.858024691358025e-06,
      "loss": 0.9019,
      "step": 3350
    },
    {
      "epoch": 28.0,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.9288,
      "step": 3360
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.5328638497652582,
      "eval_loss": 1.4555658102035522,
      "eval_runtime": 23.4988,
      "eval_samples_per_second": 72.514,
      "eval_steps_per_second": 2.298,
      "step": 3360
    },
    {
      "epoch": 28.08,
      "learning_rate": 3.5493827160493827e-06,
      "loss": 0.9258,
      "step": 3370
    },
    {
      "epoch": 28.17,
      "learning_rate": 3.3950617283950617e-06,
      "loss": 0.9017,
      "step": 3380
    },
    {
      "epoch": 28.25,
      "learning_rate": 3.2407407407407406e-06,
      "loss": 0.8841,
      "step": 3390
    },
    {
      "epoch": 28.33,
      "learning_rate": 3.0864197530864196e-06,
      "loss": 0.8903,
      "step": 3400
    },
    {
      "epoch": 28.42,
      "learning_rate": 2.932098765432099e-06,
      "loss": 0.9242,
      "step": 3410
    },
    {
      "epoch": 28.5,
      "learning_rate": 2.777777777777778e-06,
      "loss": 0.9272,
      "step": 3420
    },
    {
      "epoch": 28.58,
      "learning_rate": 2.623456790123457e-06,
      "loss": 0.8952,
      "step": 3430
    },
    {
      "epoch": 28.67,
      "learning_rate": 2.469135802469136e-06,
      "loss": 0.9263,
      "step": 3440
    },
    {
      "epoch": 28.75,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.9586,
      "step": 3450
    },
    {
      "epoch": 28.83,
      "learning_rate": 2.1604938271604937e-06,
      "loss": 0.929,
      "step": 3460
    },
    {
      "epoch": 28.92,
      "learning_rate": 2.0061728395061727e-06,
      "loss": 0.9409,
      "step": 3470
    },
    {
      "epoch": 29.0,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.9285,
      "step": 3480
    },
    {
      "epoch": 29.0,
      "eval_accuracy": 0.5446009389671361,
      "eval_loss": 1.3899754285812378,
      "eval_runtime": 23.3569,
      "eval_samples_per_second": 72.955,
      "eval_steps_per_second": 2.312,
      "step": 3480
    },
    {
      "epoch": 29.08,
      "learning_rate": 1.6975308641975308e-06,
      "loss": 0.9286,
      "step": 3490
    },
    {
      "epoch": 29.17,
      "learning_rate": 1.5432098765432098e-06,
      "loss": 0.9462,
      "step": 3500
    },
    {
      "epoch": 29.25,
      "learning_rate": 1.388888888888889e-06,
      "loss": 0.9352,
      "step": 3510
    },
    {
      "epoch": 29.33,
      "learning_rate": 1.234567901234568e-06,
      "loss": 0.9116,
      "step": 3520
    },
    {
      "epoch": 29.42,
      "learning_rate": 1.0802469135802469e-06,
      "loss": 0.8827,
      "step": 3530
    },
    {
      "epoch": 29.5,
      "learning_rate": 9.259259259259259e-07,
      "loss": 0.9091,
      "step": 3540
    },
    {
      "epoch": 29.58,
      "learning_rate": 7.716049382716049e-07,
      "loss": 0.9104,
      "step": 3550
    },
    {
      "epoch": 29.67,
      "learning_rate": 6.17283950617284e-07,
      "loss": 0.9446,
      "step": 3560
    },
    {
      "epoch": 29.75,
      "learning_rate": 4.6296296296296297e-07,
      "loss": 0.9111,
      "step": 3570
    },
    {
      "epoch": 29.83,
      "learning_rate": 3.08641975308642e-07,
      "loss": 0.9165,
      "step": 3580
    },
    {
      "epoch": 29.92,
      "learning_rate": 1.54320987654321e-07,
      "loss": 0.9068,
      "step": 3590
    },
    {
      "epoch": 30.0,
      "learning_rate": 0.0,
      "loss": 0.8955,
      "step": 3600
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.5387323943661971,
      "eval_loss": 1.4081873893737793,
      "eval_runtime": 24.0466,
      "eval_samples_per_second": 70.862,
      "eval_steps_per_second": 2.246,
      "step": 3600
    },
    {
      "epoch": 30.0,
      "step": 3600,
      "total_flos": 3.5642168535562445e+19,
      "train_loss": 0.9839045113987392,
      "train_runtime": 19975.6243,
      "train_samples_per_second": 23.025,
      "train_steps_per_second": 0.18
    }
  ],
  "max_steps": 3600,
  "num_train_epochs": 30,
  "total_flos": 3.5642168535562445e+19,
  "trial_name": null,
  "trial_params": null
}