{
  "best_metric": 0.8068181818181818,
  "best_model_checkpoint": "/Modelos/Swin-DA2-final-AMD-Wet\\checkpoint-651",
  "epoch": 77.41935483870968,
  "eval_steps": 500,
  "global_step": 1200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.65,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.6093,
      "step": 10
    },
    {
      "epoch": 0.97,
      "eval_accuracy": 0.20454545454545456,
      "eval_loss": 1.6090493202209473,
      "eval_runtime": 1.6312,
      "eval_samples_per_second": 53.948,
      "eval_steps_per_second": 3.678,
      "step": 15
    },
    {
      "epoch": 1.29,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.6084,
      "step": 20
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.25e-05,
      "loss": 1.6042,
      "step": 30
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.20454545454545456,
      "eval_loss": 1.605498194694519,
      "eval_runtime": 1.0762,
      "eval_samples_per_second": 81.766,
      "eval_steps_per_second": 5.575,
      "step": 31
    },
    {
      "epoch": 2.58,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.586,
      "step": 40
    },
    {
      "epoch": 2.97,
      "eval_accuracy": 0.19318181818181818,
      "eval_loss": 1.5730183124542236,
      "eval_runtime": 1.0502,
      "eval_samples_per_second": 83.791,
      "eval_steps_per_second": 5.713,
      "step": 46
    },
    {
      "epoch": 3.23,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.5425,
      "step": 50
    },
    {
      "epoch": 3.87,
      "learning_rate": 2.5e-05,
      "loss": 1.4855,
      "step": 60
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.26136363636363635,
      "eval_loss": 1.4935040473937988,
      "eval_runtime": 1.0352,
      "eval_samples_per_second": 85.005,
      "eval_steps_per_second": 5.796,
      "step": 62
    },
    {
      "epoch": 4.52,
      "learning_rate": 2.916666666666667e-05,
      "loss": 1.3738,
      "step": 70
    },
    {
      "epoch": 4.97,
      "eval_accuracy": 0.5,
      "eval_loss": 1.3272513151168823,
      "eval_runtime": 1.0512,
      "eval_samples_per_second": 83.711,
      "eval_steps_per_second": 5.708,
      "step": 77
    },
    {
      "epoch": 5.16,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.2795,
      "step": 80
    },
    {
      "epoch": 5.81,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.1557,
      "step": 90
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.625,
      "eval_loss": 1.0827698707580566,
      "eval_runtime": 1.0852,
      "eval_samples_per_second": 81.088,
      "eval_steps_per_second": 5.529,
      "step": 93
    },
    {
      "epoch": 6.45,
      "learning_rate": 4.166666666666667e-05,
      "loss": 1.0441,
      "step": 100
    },
    {
      "epoch": 6.97,
      "eval_accuracy": 0.6022727272727273,
      "eval_loss": 0.8971154689788818,
      "eval_runtime": 1.0662,
      "eval_samples_per_second": 82.533,
      "eval_steps_per_second": 5.627,
      "step": 108
    },
    {
      "epoch": 7.1,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 0.9639,
      "step": 110
    },
    {
      "epoch": 7.74,
      "learning_rate": 5e-05,
      "loss": 0.9358,
      "step": 120
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6818181818181818,
      "eval_loss": 0.7635229229927063,
      "eval_runtime": 1.0302,
      "eval_samples_per_second": 85.418,
      "eval_steps_per_second": 5.824,
      "step": 124
    },
    {
      "epoch": 8.39,
      "learning_rate": 4.9537037037037035e-05,
      "loss": 0.7707,
      "step": 130
    },
    {
      "epoch": 8.97,
      "eval_accuracy": 0.6477272727272727,
      "eval_loss": 0.8717986941337585,
      "eval_runtime": 1.0472,
      "eval_samples_per_second": 84.031,
      "eval_steps_per_second": 5.729,
      "step": 139
    },
    {
      "epoch": 9.03,
      "learning_rate": 4.9074074074074075e-05,
      "loss": 0.8861,
      "step": 140
    },
    {
      "epoch": 9.68,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.7161,
      "step": 150
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.6902639269828796,
      "eval_runtime": 1.0472,
      "eval_samples_per_second": 84.031,
      "eval_steps_per_second": 5.729,
      "step": 155
    },
    {
      "epoch": 10.32,
      "learning_rate": 4.814814814814815e-05,
      "loss": 0.7251,
      "step": 160
    },
    {
      "epoch": 10.97,
      "learning_rate": 4.768518518518519e-05,
      "loss": 0.6799,
      "step": 170
    },
    {
      "epoch": 10.97,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.8108208179473877,
      "eval_runtime": 1.0902,
      "eval_samples_per_second": 80.716,
      "eval_steps_per_second": 5.503,
      "step": 170
    },
    {
      "epoch": 11.61,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.6402,
      "step": 180
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.7146868109703064,
      "eval_runtime": 1.0392,
      "eval_samples_per_second": 84.678,
      "eval_steps_per_second": 5.773,
      "step": 186
    },
    {
      "epoch": 12.26,
      "learning_rate": 4.675925925925926e-05,
      "loss": 0.5816,
      "step": 190
    },
    {
      "epoch": 12.9,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.5131,
      "step": 200
    },
    {
      "epoch": 12.97,
      "eval_accuracy": 0.75,
      "eval_loss": 0.7520684599876404,
      "eval_runtime": 1.1293,
      "eval_samples_per_second": 77.928,
      "eval_steps_per_second": 5.313,
      "step": 201
    },
    {
      "epoch": 13.55,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 0.5262,
      "step": 210
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.7967298626899719,
      "eval_runtime": 1.0822,
      "eval_samples_per_second": 81.313,
      "eval_steps_per_second": 5.544,
      "step": 217
    },
    {
      "epoch": 14.19,
      "learning_rate": 4.5370370370370374e-05,
      "loss": 0.5323,
      "step": 220
    },
    {
      "epoch": 14.84,
      "learning_rate": 4.490740740740741e-05,
      "loss": 0.4757,
      "step": 230
    },
    {
      "epoch": 14.97,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 0.7084075808525085,
      "eval_runtime": 1.0422,
      "eval_samples_per_second": 84.434,
      "eval_steps_per_second": 5.757,
      "step": 232
    },
    {
      "epoch": 15.48,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.4758,
      "step": 240
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.7528924345970154,
      "eval_runtime": 1.0702,
      "eval_samples_per_second": 82.224,
      "eval_steps_per_second": 5.606,
      "step": 248
    },
    {
      "epoch": 16.13,
      "learning_rate": 4.3981481481481486e-05,
      "loss": 0.4203,
      "step": 250
    },
    {
      "epoch": 16.77,
      "learning_rate": 4.351851851851852e-05,
      "loss": 0.4184,
      "step": 260
    },
    {
      "epoch": 16.97,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.7768625020980835,
      "eval_runtime": 1.0432,
      "eval_samples_per_second": 84.353,
      "eval_steps_per_second": 5.751,
      "step": 263
    },
    {
      "epoch": 17.42,
      "learning_rate": 4.305555555555556e-05,
      "loss": 0.398,
      "step": 270
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 0.8496245741844177,
      "eval_runtime": 1.0792,
      "eval_samples_per_second": 81.539,
      "eval_steps_per_second": 5.559,
      "step": 279
    },
    {
      "epoch": 18.06,
      "learning_rate": 4.259259259259259e-05,
      "loss": 0.3623,
      "step": 280
    },
    {
      "epoch": 18.71,
      "learning_rate": 4.212962962962963e-05,
      "loss": 0.3591,
      "step": 290
    },
    {
      "epoch": 18.97,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.8204294443130493,
      "eval_runtime": 1.0712,
      "eval_samples_per_second": 82.148,
      "eval_steps_per_second": 5.601,
      "step": 294
    },
    {
      "epoch": 19.35,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.3449,
      "step": 300
    },
    {
      "epoch": 20.0,
      "learning_rate": 4.1203703703703705e-05,
      "loss": 0.3536,
      "step": 310
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 0.8589130640029907,
      "eval_runtime": 1.0792,
      "eval_samples_per_second": 81.539,
      "eval_steps_per_second": 5.559,
      "step": 310
    },
    {
      "epoch": 20.65,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.2589,
      "step": 320
    },
    {
      "epoch": 20.97,
      "eval_accuracy": 0.7045454545454546,
      "eval_loss": 0.9754377007484436,
      "eval_runtime": 1.0472,
      "eval_samples_per_second": 84.031,
      "eval_steps_per_second": 5.729,
      "step": 325
    },
    {
      "epoch": 21.29,
      "learning_rate": 4.027777777777778e-05,
      "loss": 0.286,
      "step": 330
    },
    {
      "epoch": 21.94,
      "learning_rate": 3.981481481481482e-05,
      "loss": 0.3218,
      "step": 340
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.7159090909090909,
      "eval_loss": 1.0231382846832275,
      "eval_runtime": 1.0582,
      "eval_samples_per_second": 83.157,
      "eval_steps_per_second": 5.67,
      "step": 341
    },
    {
      "epoch": 22.58,
      "learning_rate": 3.935185185185186e-05,
      "loss": 0.3151,
      "step": 350
    },
    {
      "epoch": 22.97,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 0.9173128604888916,
      "eval_runtime": 1.0472,
      "eval_samples_per_second": 84.031,
      "eval_steps_per_second": 5.729,
      "step": 356
    },
    {
      "epoch": 23.23,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.3135,
      "step": 360
    },
    {
      "epoch": 23.87,
      "learning_rate": 3.8425925925925924e-05,
      "loss": 0.2708,
      "step": 370
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 0.9598014950752258,
      "eval_runtime": 1.0532,
      "eval_samples_per_second": 83.552,
      "eval_steps_per_second": 5.697,
      "step": 372
    },
    {
      "epoch": 24.52,
      "learning_rate": 3.7962962962962964e-05,
      "loss": 0.2802,
      "step": 380
    },
    {
      "epoch": 24.97,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 0.904956579208374,
      "eval_runtime": 1.0442,
      "eval_samples_per_second": 84.272,
      "eval_steps_per_second": 5.746,
      "step": 387
    },
    {
      "epoch": 25.16,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.2795,
      "step": 390
    },
    {
      "epoch": 25.81,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.3114,
      "step": 400
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.8725055456161499,
      "eval_runtime": 1.0662,
      "eval_samples_per_second": 82.533,
      "eval_steps_per_second": 5.627,
      "step": 403
    },
    {
      "epoch": 26.45,
      "learning_rate": 3.6574074074074076e-05,
      "loss": 0.2794,
      "step": 410
    },
    {
      "epoch": 26.97,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 0.9578754901885986,
      "eval_runtime": 1.0662,
      "eval_samples_per_second": 82.533,
      "eval_steps_per_second": 5.627,
      "step": 418
    },
    {
      "epoch": 27.1,
      "learning_rate": 3.611111111111111e-05,
      "loss": 0.2289,
      "step": 420
    },
    {
      "epoch": 27.74,
      "learning_rate": 3.564814814814815e-05,
      "loss": 0.26,
      "step": 430
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 1.0064443349838257,
      "eval_runtime": 1.0862,
      "eval_samples_per_second": 81.013,
      "eval_steps_per_second": 5.524,
      "step": 434
    },
    {
      "epoch": 28.39,
      "learning_rate": 3.518518518518519e-05,
      "loss": 0.2961,
      "step": 440
    },
    {
      "epoch": 28.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.1055842638015747,
      "eval_runtime": 1.1363,
      "eval_samples_per_second": 77.447,
      "eval_steps_per_second": 5.281,
      "step": 449
    },
    {
      "epoch": 29.03,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.2547,
      "step": 450
    },
    {
      "epoch": 29.68,
      "learning_rate": 3.425925925925926e-05,
      "loss": 0.297,
      "step": 460
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 0.8761419653892517,
      "eval_runtime": 1.1092,
      "eval_samples_per_second": 79.333,
      "eval_steps_per_second": 5.409,
      "step": 465
    },
    {
      "epoch": 30.32,
      "learning_rate": 3.3796296296296295e-05,
      "loss": 0.2583,
      "step": 470
    },
    {
      "epoch": 30.97,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.2044,
      "step": 480
    },
    {
      "epoch": 30.97,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 1.0461256504058838,
      "eval_runtime": 1.1173,
      "eval_samples_per_second": 78.765,
      "eval_steps_per_second": 5.37,
      "step": 480
    },
    {
      "epoch": 31.61,
      "learning_rate": 3.2870370370370375e-05,
      "loss": 0.1884,
      "step": 490
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.75,
      "eval_loss": 0.9888725280761719,
      "eval_runtime": 1.0972,
      "eval_samples_per_second": 80.201,
      "eval_steps_per_second": 5.468,
      "step": 496
    },
    {
      "epoch": 32.26,
      "learning_rate": 3.240740740740741e-05,
      "loss": 0.1857,
      "step": 500
    },
    {
      "epoch": 32.9,
      "learning_rate": 3.194444444444444e-05,
      "loss": 0.2156,
      "step": 510
    },
    {
      "epoch": 32.97,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 1.0186409950256348,
      "eval_runtime": 1.1423,
      "eval_samples_per_second": 77.04,
      "eval_steps_per_second": 5.253,
      "step": 511
    },
    {
      "epoch": 33.55,
      "learning_rate": 3.148148148148148e-05,
      "loss": 0.194,
      "step": 520
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 1.0899921655654907,
      "eval_runtime": 1.1843,
      "eval_samples_per_second": 74.308,
      "eval_steps_per_second": 5.066,
      "step": 527
    },
    {
      "epoch": 34.19,
      "learning_rate": 3.101851851851852e-05,
      "loss": 0.181,
      "step": 530
    },
    {
      "epoch": 34.84,
      "learning_rate": 3.055555555555556e-05,
      "loss": 0.2085,
      "step": 540
    },
    {
      "epoch": 34.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.0762078762054443,
      "eval_runtime": 1.1223,
      "eval_samples_per_second": 78.414,
      "eval_steps_per_second": 5.346,
      "step": 542
    },
    {
      "epoch": 35.48,
      "learning_rate": 3.0092592592592593e-05,
      "loss": 0.1909,
      "step": 550
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.7840909090909091,
      "eval_loss": 1.0324878692626953,
      "eval_runtime": 1.1523,
      "eval_samples_per_second": 76.372,
      "eval_steps_per_second": 5.207,
      "step": 558
    },
    {
      "epoch": 36.13,
      "learning_rate": 2.962962962962963e-05,
      "loss": 0.1864,
      "step": 560
    },
    {
      "epoch": 36.77,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.1551,
      "step": 570
    },
    {
      "epoch": 36.97,
      "eval_accuracy": 0.7045454545454546,
      "eval_loss": 1.149728775024414,
      "eval_runtime": 1.1653,
      "eval_samples_per_second": 75.519,
      "eval_steps_per_second": 5.149,
      "step": 573
    },
    {
      "epoch": 37.42,
      "learning_rate": 2.8703703703703706e-05,
      "loss": 0.2106,
      "step": 580
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 1.0304036140441895,
      "eval_runtime": 1.1102,
      "eval_samples_per_second": 79.261,
      "eval_steps_per_second": 5.404,
      "step": 589
    },
    {
      "epoch": 38.06,
      "learning_rate": 2.824074074074074e-05,
      "loss": 0.1922,
      "step": 590
    },
    {
      "epoch": 38.71,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.1771,
      "step": 600
    },
    {
      "epoch": 38.97,
      "eval_accuracy": 0.7840909090909091,
      "eval_loss": 1.0794123411178589,
      "eval_runtime": 1.1483,
      "eval_samples_per_second": 76.638,
      "eval_steps_per_second": 5.225,
      "step": 604
    },
    {
      "epoch": 39.35,
      "learning_rate": 2.7314814814814816e-05,
      "loss": 0.2051,
      "step": 610
    },
    {
      "epoch": 40.0,
      "learning_rate": 2.6851851851851855e-05,
      "loss": 0.1567,
      "step": 620
    },
    {
      "epoch": 40.0,
      "eval_accuracy": 0.7954545454545454,
      "eval_loss": 1.0634194612503052,
      "eval_runtime": 1.1643,
      "eval_samples_per_second": 75.584,
      "eval_steps_per_second": 5.153,
      "step": 620
    },
    {
      "epoch": 40.65,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 0.1856,
      "step": 630
    },
    {
      "epoch": 40.97,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 1.0716369152069092,
      "eval_runtime": 1.1173,
      "eval_samples_per_second": 78.765,
      "eval_steps_per_second": 5.37,
      "step": 635
    },
    {
      "epoch": 41.29,
      "learning_rate": 2.5925925925925925e-05,
      "loss": 0.1765,
      "step": 640
    },
    {
      "epoch": 41.94,
      "learning_rate": 2.5462962962962965e-05,
      "loss": 0.185,
      "step": 650
    },
    {
      "epoch": 42.0,
      "eval_accuracy": 0.8068181818181818,
      "eval_loss": 1.0321924686431885,
      "eval_runtime": 1.1433,
      "eval_samples_per_second": 76.973,
      "eval_steps_per_second": 5.248,
      "step": 651
    },
    {
      "epoch": 42.58,
      "learning_rate": 2.5e-05,
      "loss": 0.1239,
      "step": 660
    },
    {
      "epoch": 42.97,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 1.1515936851501465,
      "eval_runtime": 1.1523,
      "eval_samples_per_second": 76.372,
      "eval_steps_per_second": 5.207,
      "step": 666
    },
    {
      "epoch": 43.23,
      "learning_rate": 2.4537037037037038e-05,
      "loss": 0.2364,
      "step": 670
    },
    {
      "epoch": 43.87,
      "learning_rate": 2.4074074074074074e-05,
      "loss": 0.1617,
      "step": 680
    },
    {
      "epoch": 44.0,
      "eval_accuracy": 0.7840909090909091,
      "eval_loss": 1.0321751832962036,
      "eval_runtime": 1.1253,
      "eval_samples_per_second": 78.205,
      "eval_steps_per_second": 5.332,
      "step": 682
    },
    {
      "epoch": 44.52,
      "learning_rate": 2.361111111111111e-05,
      "loss": 0.1221,
      "step": 690
    },
    {
      "epoch": 44.97,
      "eval_accuracy": 0.8068181818181818,
      "eval_loss": 1.0553076267242432,
      "eval_runtime": 1.1333,
      "eval_samples_per_second": 77.652,
      "eval_steps_per_second": 5.294,
      "step": 697
    },
    {
      "epoch": 45.16,
      "learning_rate": 2.314814814814815e-05,
      "loss": 0.1953,
      "step": 700
    },
    {
      "epoch": 45.81,
      "learning_rate": 2.2685185185185187e-05,
      "loss": 0.1433,
      "step": 710
    },
    {
      "epoch": 46.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 1.0596939325332642,
      "eval_runtime": 1.1092,
      "eval_samples_per_second": 79.333,
      "eval_steps_per_second": 5.409,
      "step": 713
    },
    {
      "epoch": 46.45,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.216,
      "step": 720
    },
    {
      "epoch": 46.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.1585843563079834,
      "eval_runtime": 1.1223,
      "eval_samples_per_second": 78.414,
      "eval_steps_per_second": 5.346,
      "step": 728
    },
    {
      "epoch": 47.1,
      "learning_rate": 2.175925925925926e-05,
      "loss": 0.17,
      "step": 730
    },
    {
      "epoch": 47.74,
      "learning_rate": 2.1296296296296296e-05,
      "loss": 0.1807,
      "step": 740
    },
    {
      "epoch": 48.0,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 1.0872578620910645,
      "eval_runtime": 1.1243,
      "eval_samples_per_second": 78.274,
      "eval_steps_per_second": 5.337,
      "step": 744
    },
    {
      "epoch": 48.39,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.185,
      "step": 750
    },
    {
      "epoch": 48.97,
      "eval_accuracy": 0.7727272727272727,
      "eval_loss": 1.2490220069885254,
      "eval_runtime": 1.1133,
      "eval_samples_per_second": 79.048,
      "eval_steps_per_second": 5.39,
      "step": 759
    },
    {
      "epoch": 49.03,
      "learning_rate": 2.037037037037037e-05,
      "loss": 0.1114,
      "step": 760
    },
    {
      "epoch": 49.68,
      "learning_rate": 1.990740740740741e-05,
      "loss": 0.1554,
      "step": 770
    },
    {
      "epoch": 50.0,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 1.222277045249939,
      "eval_runtime": 1.2013,
      "eval_samples_per_second": 73.256,
      "eval_steps_per_second": 4.995,
      "step": 775
    },
    {
      "epoch": 50.32,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.1378,
      "step": 780
    },
    {
      "epoch": 50.97,
      "learning_rate": 1.8981481481481482e-05,
      "loss": 0.1359,
      "step": 790
    },
    {
      "epoch": 50.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.2345318794250488,
      "eval_runtime": 1.1193,
      "eval_samples_per_second": 78.624,
      "eval_steps_per_second": 5.361,
      "step": 790
    },
    {
      "epoch": 51.61,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.0929,
      "step": 800
    },
    {
      "epoch": 52.0,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 1.1833394765853882,
      "eval_runtime": 1.1123,
      "eval_samples_per_second": 79.119,
      "eval_steps_per_second": 5.394,
      "step": 806
    },
    {
      "epoch": 52.26,
      "learning_rate": 1.8055555555555555e-05,
      "loss": 0.1513,
      "step": 810
    },
    {
      "epoch": 52.9,
      "learning_rate": 1.7592592592592595e-05,
      "loss": 0.1379,
      "step": 820
    },
    {
      "epoch": 52.97,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 1.2581446170806885,
      "eval_runtime": 1.1193,
      "eval_samples_per_second": 78.624,
      "eval_steps_per_second": 5.361,
      "step": 821
    },
    {
      "epoch": 53.55,
      "learning_rate": 1.712962962962963e-05,
      "loss": 0.145,
      "step": 830
    },
    {
      "epoch": 54.0,
      "eval_accuracy": 0.75,
      "eval_loss": 1.3023442029953003,
      "eval_runtime": 1.1203,
      "eval_samples_per_second": 78.554,
      "eval_steps_per_second": 5.356,
      "step": 837
    },
    {
      "epoch": 54.19,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.1646,
      "step": 840
    },
    {
      "epoch": 54.84,
      "learning_rate": 1.6203703703703704e-05,
      "loss": 0.134,
      "step": 850
    },
    {
      "epoch": 54.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.2468658685684204,
      "eval_runtime": 1.1072,
      "eval_samples_per_second": 79.476,
      "eval_steps_per_second": 5.419,
      "step": 852
    },
    {
      "epoch": 55.48,
      "learning_rate": 1.574074074074074e-05,
      "loss": 0.1974,
      "step": 860
    },
    {
      "epoch": 56.0,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 1.2670570611953735,
      "eval_runtime": 1.1002,
      "eval_samples_per_second": 79.982,
      "eval_steps_per_second": 5.453,
      "step": 868
    },
    {
      "epoch": 56.13,
      "learning_rate": 1.527777777777778e-05,
      "loss": 0.1158,
      "step": 870
    },
    {
      "epoch": 56.77,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 0.122,
      "step": 880
    },
    {
      "epoch": 56.97,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 1.267562747001648,
      "eval_runtime": 1.0852,
      "eval_samples_per_second": 81.088,
      "eval_steps_per_second": 5.529,
      "step": 883
    },
    {
      "epoch": 57.42,
      "learning_rate": 1.4351851851851853e-05,
      "loss": 0.1487,
      "step": 890
    },
    {
      "epoch": 58.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 1.2845935821533203,
      "eval_runtime": 1.1173,
      "eval_samples_per_second": 78.765,
      "eval_steps_per_second": 5.37,
      "step": 899
    },
    {
      "epoch": 58.06,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.1473,
      "step": 900
    },
    {
      "epoch": 58.71,
      "learning_rate": 1.3425925925925928e-05,
      "loss": 0.1282,
      "step": 910
    },
    {
      "epoch": 58.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.1780353784561157,
      "eval_runtime": 1.0612,
      "eval_samples_per_second": 82.922,
      "eval_steps_per_second": 5.654,
      "step": 914
    },
    {
      "epoch": 59.35,
      "learning_rate": 1.2962962962962962e-05,
      "loss": 0.1006,
      "step": 920
    },
    {
      "epoch": 60.0,
      "learning_rate": 1.25e-05,
      "loss": 0.0989,
      "step": 930
    },
    {
      "epoch": 60.0,
      "eval_accuracy": 0.75,
      "eval_loss": 1.231979489326477,
      "eval_runtime": 1.0982,
      "eval_samples_per_second": 80.128,
      "eval_steps_per_second": 5.463,
      "step": 930
    },
    {
      "epoch": 60.65,
      "learning_rate": 1.2037037037037037e-05,
      "loss": 0.0997,
      "step": 940
    },
    {
      "epoch": 60.97,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 1.2791739702224731,
      "eval_runtime": 1.1663,
      "eval_samples_per_second": 75.455,
      "eval_steps_per_second": 5.145,
      "step": 945
    },
    {
      "epoch": 61.29,
      "learning_rate": 1.1574074074074075e-05,
      "loss": 0.1234,
      "step": 950
    },
    {
      "epoch": 61.94,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.1058,
      "step": 960
    },
    {
      "epoch": 62.0,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 1.2126153707504272,
      "eval_runtime": 1.1072,
      "eval_samples_per_second": 79.476,
      "eval_steps_per_second": 5.419,
      "step": 961
    },
    {
      "epoch": 62.58,
      "learning_rate": 1.0648148148148148e-05,
      "loss": 0.1105,
      "step": 970
    },
    {
      "epoch": 62.97,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 1.256054401397705,
      "eval_runtime": 1.0872,
      "eval_samples_per_second": 80.939,
      "eval_steps_per_second": 5.519,
      "step": 976
    },
    {
      "epoch": 63.23,
      "learning_rate": 1.0185185185185185e-05,
      "loss": 0.1254,
      "step": 980
    },
    {
      "epoch": 63.87,
      "learning_rate": 9.722222222222223e-06,
      "loss": 0.0957,
      "step": 990
    },
    {
      "epoch": 64.0,
      "eval_accuracy": 0.7613636363636364,
      "eval_loss": 1.1701573133468628,
      "eval_runtime": 1.1323,
      "eval_samples_per_second": 77.721,
      "eval_steps_per_second": 5.299,
      "step": 992
    },
    {
      "epoch": 64.52,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.1326,
      "step": 1000
    },
    {
      "epoch": 64.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.1838548183441162,
      "eval_runtime": 1.1042,
      "eval_samples_per_second": 79.692,
      "eval_steps_per_second": 5.434,
      "step": 1007
    },
    {
      "epoch": 65.16,
      "learning_rate": 8.796296296296297e-06,
      "loss": 0.1424,
      "step": 1010
    },
    {
      "epoch": 65.81,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.0838,
      "step": 1020
    },
    {
      "epoch": 66.0,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 1.2728073596954346,
      "eval_runtime": 1.1643,
      "eval_samples_per_second": 75.584,
      "eval_steps_per_second": 5.153,
      "step": 1023
    },
    {
      "epoch": 66.45,
      "learning_rate": 7.87037037037037e-06,
      "loss": 0.1163,
      "step": 1030
    },
    {
      "epoch": 66.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.2736155986785889,
      "eval_runtime": 1.1353,
      "eval_samples_per_second": 77.516,
      "eval_steps_per_second": 5.285,
      "step": 1038
    },
    {
      "epoch": 67.1,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.1363,
      "step": 1040
    },
    {
      "epoch": 67.74,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.0926,
      "step": 1050
    },
    {
      "epoch": 68.0,
      "eval_accuracy": 0.75,
      "eval_loss": 1.2641026973724365,
      "eval_runtime": 1.1493,
      "eval_samples_per_second": 76.571,
      "eval_steps_per_second": 5.221,
      "step": 1054
    },
    {
      "epoch": 68.39,
      "learning_rate": 6.481481481481481e-06,
      "loss": 0.102,
      "step": 1060
    },
    {
      "epoch": 68.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.3310309648513794,
      "eval_runtime": 1.1102,
      "eval_samples_per_second": 79.261,
      "eval_steps_per_second": 5.404,
      "step": 1069
    },
    {
      "epoch": 69.03,
      "learning_rate": 6.0185185185185185e-06,
      "loss": 0.1224,
      "step": 1070
    },
    {
      "epoch": 69.68,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.0996,
      "step": 1080
    },
    {
      "epoch": 70.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 1.3119887113571167,
      "eval_runtime": 1.1012,
      "eval_samples_per_second": 79.909,
      "eval_steps_per_second": 5.448,
      "step": 1085
    },
    {
      "epoch": 70.32,
      "learning_rate": 5.092592592592592e-06,
      "loss": 0.1182,
      "step": 1090
    },
    {
      "epoch": 70.97,
      "learning_rate": 4.6296296296296296e-06,
      "loss": 0.081,
      "step": 1100
    },
    {
      "epoch": 70.97,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 1.3357912302017212,
      "eval_runtime": 1.1133,
      "eval_samples_per_second": 79.048,
      "eval_steps_per_second": 5.39,
      "step": 1100
    },
    {
      "epoch": 71.61,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.1305,
      "step": 1110
    },
    {
      "epoch": 72.0,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 1.3439818620681763,
      "eval_runtime": 1.1243,
      "eval_samples_per_second": 78.274,
      "eval_steps_per_second": 5.337,
      "step": 1116
    },
    {
      "epoch": 72.26,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.1309,
      "step": 1120
    },
    {
      "epoch": 72.9,
      "learning_rate": 3.2407407407407406e-06,
      "loss": 0.1131,
      "step": 1130
    },
    {
      "epoch": 72.97,
      "eval_accuracy": 0.7272727272727273,
      "eval_loss": 1.3126434087753296,
      "eval_runtime": 1.1203,
      "eval_samples_per_second": 78.554,
      "eval_steps_per_second": 5.356,
      "step": 1131
    },
    {
      "epoch": 73.55,
      "learning_rate": 2.777777777777778e-06,
      "loss": 0.0883,
      "step": 1140
    },
    {
      "epoch": 74.0,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 1.284845232963562,
      "eval_runtime": 1.0972,
      "eval_samples_per_second": 80.201,
      "eval_steps_per_second": 5.468,
      "step": 1147
    },
    {
      "epoch": 74.19,
      "learning_rate": 2.3148148148148148e-06,
      "loss": 0.1135,
      "step": 1150
    },
    {
      "epoch": 74.84,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.0873,
      "step": 1160
    },
    {
      "epoch": 74.97,
      "eval_accuracy": 0.7386363636363636,
      "eval_loss": 1.2801809310913086,
      "eval_runtime": 1.1313,
      "eval_samples_per_second": 77.79,
      "eval_steps_per_second": 5.304,
      "step": 1162
    },
    {
      "epoch": 75.48,
      "learning_rate": 1.388888888888889e-06,
      "loss": 0.0991,
      "step": 1170
    },
    {
      "epoch": 76.0,
      "eval_accuracy": 0.75,
      "eval_loss": 1.2710990905761719,
      "eval_runtime": 1.1643,
      "eval_samples_per_second": 75.584,
      "eval_steps_per_second": 5.153,
      "step": 1178
    },
    {
      "epoch": 76.13,
      "learning_rate": 9.259259259259259e-07,
      "loss": 0.0946,
      "step": 1180
    },
    {
      "epoch": 76.77,
      "learning_rate": 4.6296296296296297e-07,
      "loss": 0.0881,
      "step": 1190
    },
    {
      "epoch": 76.97,
      "eval_accuracy": 0.75,
      "eval_loss": 1.2746005058288574,
      "eval_runtime": 1.4135,
      "eval_samples_per_second": 62.256,
      "eval_steps_per_second": 4.245,
      "step": 1193
    },
    {
      "epoch": 77.42,
      "learning_rate": 0.0,
      "loss": 0.0895,
      "step": 1200
    },
    {
      "epoch": 77.42,
      "eval_accuracy": 0.75,
      "eval_loss": 1.275152325630188,
      "eval_runtime": 1.1543,
      "eval_samples_per_second": 76.239,
      "eval_steps_per_second": 5.198,
      "step": 1200
    },
    {
      "epoch": 77.42,
      "step": 1200,
      "total_flos": 2.4687812024881644e+18,
      "train_loss": 0.34467112536231675,
      "train_runtime": 2690.4819,
      "train_samples_per_second": 29.14,
      "train_steps_per_second": 0.446
    }
  ],
  "logging_steps": 10,
  "max_steps": 1200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 80,
  "save_steps": 500,
  "total_flos": 2.4687812024881644e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}