{
  "best_metric": 0.5666666666666667,
  "best_model_checkpoint": "vit-base-patch16-224-dmae-va-U5-42C/checkpoint-232",
  "epoch": 37.935483870967744,
  "eval_steps": 500,
  "global_step": 294,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9,
      "eval_accuracy": 0.13333333333333333,
      "eval_loss": 1.4545705318450928,
      "eval_runtime": 2.7807,
      "eval_samples_per_second": 21.578,
      "eval_steps_per_second": 0.719,
      "step": 7
    },
    {
      "epoch": 1.55,
      "grad_norm": 5.473387241363525,
      "learning_rate": 4e-07,
      "loss": 1.5342,
      "step": 12
    },
    {
      "epoch": 1.94,
      "eval_accuracy": 0.13333333333333333,
      "eval_loss": 1.4379044771194458,
      "eval_runtime": 1.5117,
      "eval_samples_per_second": 39.69,
      "eval_steps_per_second": 1.323,
      "step": 15
    },
    {
      "epoch": 2.97,
      "eval_accuracy": 0.16666666666666666,
      "eval_loss": 1.411478042602539,
      "eval_runtime": 1.4724,
      "eval_samples_per_second": 40.749,
      "eval_steps_per_second": 1.358,
      "step": 23
    },
    {
      "epoch": 3.1,
      "grad_norm": 6.9020280838012695,
      "learning_rate": 8e-07,
      "loss": 1.5331,
      "step": 24
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2,
      "eval_loss": 1.3786956071853638,
      "eval_runtime": 1.4845,
      "eval_samples_per_second": 40.418,
      "eval_steps_per_second": 1.347,
      "step": 31
    },
    {
      "epoch": 4.65,
      "grad_norm": 4.692047119140625,
      "learning_rate": 9.772727272727273e-07,
      "loss": 1.4639,
      "step": 36
    },
    {
      "epoch": 4.9,
      "eval_accuracy": 0.2833333333333333,
      "eval_loss": 1.351299524307251,
      "eval_runtime": 1.5497,
      "eval_samples_per_second": 38.718,
      "eval_steps_per_second": 1.291,
      "step": 38
    },
    {
      "epoch": 5.94,
      "eval_accuracy": 0.3333333333333333,
      "eval_loss": 1.3290389776229858,
      "eval_runtime": 1.48,
      "eval_samples_per_second": 40.54,
      "eval_steps_per_second": 1.351,
      "step": 46
    },
    {
      "epoch": 6.19,
      "grad_norm": 3.2680552005767822,
      "learning_rate": 9.318181818181817e-07,
      "loss": 1.4056,
      "step": 48
    },
    {
      "epoch": 6.97,
      "eval_accuracy": 0.38333333333333336,
      "eval_loss": 1.3113869428634644,
      "eval_runtime": 1.5135,
      "eval_samples_per_second": 39.644,
      "eval_steps_per_second": 1.321,
      "step": 54
    },
    {
      "epoch": 7.74,
      "grad_norm": 2.349689245223999,
      "learning_rate": 8.863636363636363e-07,
      "loss": 1.3679,
      "step": 60
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.43333333333333335,
      "eval_loss": 1.2940715551376343,
      "eval_runtime": 1.9694,
      "eval_samples_per_second": 30.465,
      "eval_steps_per_second": 1.016,
      "step": 62
    },
    {
      "epoch": 8.9,
      "eval_accuracy": 0.4666666666666667,
      "eval_loss": 1.2827337980270386,
      "eval_runtime": 1.9571,
      "eval_samples_per_second": 30.658,
      "eval_steps_per_second": 1.022,
      "step": 69
    },
    {
      "epoch": 9.29,
      "grad_norm": 2.8183329105377197,
      "learning_rate": 8.409090909090909e-07,
      "loss": 1.3387,
      "step": 72
    },
    {
      "epoch": 9.94,
      "eval_accuracy": 0.5,
      "eval_loss": 1.2678014039993286,
      "eval_runtime": 1.5047,
      "eval_samples_per_second": 39.874,
      "eval_steps_per_second": 1.329,
      "step": 77
    },
    {
      "epoch": 10.84,
      "grad_norm": 2.4520814418792725,
      "learning_rate": 7.954545454545454e-07,
      "loss": 1.2992,
      "step": 84
    },
    {
      "epoch": 10.97,
      "eval_accuracy": 0.4666666666666667,
      "eval_loss": 1.2557296752929688,
      "eval_runtime": 1.8046,
      "eval_samples_per_second": 33.248,
      "eval_steps_per_second": 1.108,
      "step": 85
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.4666666666666667,
      "eval_loss": 1.2453906536102295,
      "eval_runtime": 1.4943,
      "eval_samples_per_second": 40.152,
      "eval_steps_per_second": 1.338,
      "step": 93
    },
    {
      "epoch": 12.39,
      "grad_norm": 2.357367992401123,
      "learning_rate": 7.5e-07,
      "loss": 1.2797,
      "step": 96
    },
    {
      "epoch": 12.9,
      "eval_accuracy": 0.48333333333333334,
      "eval_loss": 1.234529972076416,
      "eval_runtime": 1.5563,
      "eval_samples_per_second": 38.554,
      "eval_steps_per_second": 1.285,
      "step": 100
    },
    {
      "epoch": 13.94,
      "grad_norm": 2.453970432281494,
      "learning_rate": 7.045454545454545e-07,
      "loss": 1.2507,
      "step": 108
    },
    {
      "epoch": 13.94,
      "eval_accuracy": 0.48333333333333334,
      "eval_loss": 1.221469759941101,
      "eval_runtime": 1.4885,
      "eval_samples_per_second": 40.308,
      "eval_steps_per_second": 1.344,
      "step": 108
    },
    {
      "epoch": 14.97,
      "eval_accuracy": 0.5,
      "eval_loss": 1.2108979225158691,
      "eval_runtime": 1.4973,
      "eval_samples_per_second": 40.072,
      "eval_steps_per_second": 1.336,
      "step": 116
    },
    {
      "epoch": 15.48,
      "grad_norm": 2.9230761528015137,
      "learning_rate": 6.59090909090909e-07,
      "loss": 1.2337,
      "step": 120
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.5,
      "eval_loss": 1.2004987001419067,
      "eval_runtime": 1.7915,
      "eval_samples_per_second": 33.492,
      "eval_steps_per_second": 1.116,
      "step": 124
    },
    {
      "epoch": 16.9,
      "eval_accuracy": 0.5,
      "eval_loss": 1.1903845071792603,
      "eval_runtime": 1.4887,
      "eval_samples_per_second": 40.304,
      "eval_steps_per_second": 1.343,
      "step": 131
    },
    {
      "epoch": 17.03,
      "grad_norm": 2.8677401542663574,
      "learning_rate": 6.136363636363636e-07,
      "loss": 1.2076,
      "step": 132
    },
    {
      "epoch": 17.94,
      "eval_accuracy": 0.5166666666666667,
      "eval_loss": 1.1796098947525024,
      "eval_runtime": 1.4929,
      "eval_samples_per_second": 40.191,
      "eval_steps_per_second": 1.34,
      "step": 139
    },
    {
      "epoch": 18.58,
      "grad_norm": 2.4694573879241943,
      "learning_rate": 5.681818181818182e-07,
      "loss": 1.1968,
      "step": 144
    },
    {
      "epoch": 18.97,
      "eval_accuracy": 0.5333333333333333,
      "eval_loss": 1.1699223518371582,
      "eval_runtime": 1.4962,
      "eval_samples_per_second": 40.102,
      "eval_steps_per_second": 1.337,
      "step": 147
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.5333333333333333,
      "eval_loss": 1.1609662771224976,
      "eval_runtime": 2.0442,
      "eval_samples_per_second": 29.351,
      "eval_steps_per_second": 0.978,
      "step": 155
    },
    {
      "epoch": 20.13,
      "grad_norm": 2.54495906829834,
      "learning_rate": 5.227272727272727e-07,
      "loss": 1.171,
      "step": 156
    },
    {
      "epoch": 20.9,
      "eval_accuracy": 0.5333333333333333,
      "eval_loss": 1.1543792486190796,
      "eval_runtime": 1.4771,
      "eval_samples_per_second": 40.621,
      "eval_steps_per_second": 1.354,
      "step": 162
    },
    {
      "epoch": 21.68,
      "grad_norm": 2.391249418258667,
      "learning_rate": 4.772727272727273e-07,
      "loss": 1.1572,
      "step": 168
    },
    {
      "epoch": 21.94,
      "eval_accuracy": 0.5333333333333333,
      "eval_loss": 1.1475883722305298,
      "eval_runtime": 1.4966,
      "eval_samples_per_second": 40.091,
      "eval_steps_per_second": 1.336,
      "step": 170
    },
    {
      "epoch": 22.97,
      "eval_accuracy": 0.5333333333333333,
      "eval_loss": 1.1411240100860596,
      "eval_runtime": 1.5033,
      "eval_samples_per_second": 39.912,
      "eval_steps_per_second": 1.33,
      "step": 178
    },
    {
      "epoch": 23.23,
      "grad_norm": 2.524062395095825,
      "learning_rate": 4.318181818181818e-07,
      "loss": 1.1383,
      "step": 180
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.5333333333333333,
      "eval_loss": 1.135024905204773,
      "eval_runtime": 1.6434,
      "eval_samples_per_second": 36.51,
      "eval_steps_per_second": 1.217,
      "step": 186
    },
    {
      "epoch": 24.77,
      "grad_norm": 2.7990708351135254,
      "learning_rate": 3.8636363636363636e-07,
      "loss": 1.14,
      "step": 192
    },
    {
      "epoch": 24.9,
      "eval_accuracy": 0.5333333333333333,
      "eval_loss": 1.1297781467437744,
      "eval_runtime": 1.4951,
      "eval_samples_per_second": 40.131,
      "eval_steps_per_second": 1.338,
      "step": 193
    },
    {
      "epoch": 25.94,
      "eval_accuracy": 0.55,
      "eval_loss": 1.1256134510040283,
      "eval_runtime": 1.4962,
      "eval_samples_per_second": 40.102,
      "eval_steps_per_second": 1.337,
      "step": 201
    },
    {
      "epoch": 26.32,
      "grad_norm": 2.7583816051483154,
      "learning_rate": 3.4090909090909085e-07,
      "loss": 1.1114,
      "step": 204
    },
    {
      "epoch": 26.97,
      "eval_accuracy": 0.55,
      "eval_loss": 1.1212241649627686,
      "eval_runtime": 1.6451,
      "eval_samples_per_second": 36.472,
      "eval_steps_per_second": 1.216,
      "step": 209
    },
    {
      "epoch": 27.87,
      "grad_norm": 2.4490418434143066,
      "learning_rate": 2.9545454545454545e-07,
      "loss": 1.1094,
      "step": 216
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.55,
      "eval_loss": 1.1173356771469116,
      "eval_runtime": 1.5074,
      "eval_samples_per_second": 39.803,
      "eval_steps_per_second": 1.327,
      "step": 217
    },
    {
      "epoch": 28.9,
      "eval_accuracy": 0.55,
      "eval_loss": 1.1143361330032349,
      "eval_runtime": 1.5142,
      "eval_samples_per_second": 39.625,
      "eval_steps_per_second": 1.321,
      "step": 224
    },
    {
      "epoch": 29.42,
      "grad_norm": 2.93764066696167,
      "learning_rate": 2.5e-07,
      "loss": 1.0872,
      "step": 228
    },
    {
      "epoch": 29.94,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.111220121383667,
      "eval_runtime": 1.4968,
      "eval_samples_per_second": 40.085,
      "eval_steps_per_second": 1.336,
      "step": 232
    },
    {
      "epoch": 30.97,
      "grad_norm": 2.4049625396728516,
      "learning_rate": 2.0454545454545456e-07,
      "loss": 1.0941,
      "step": 240
    },
    {
      "epoch": 30.97,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.1078248023986816,
      "eval_runtime": 1.9337,
      "eval_samples_per_second": 31.028,
      "eval_steps_per_second": 1.034,
      "step": 240
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.1053519248962402,
      "eval_runtime": 1.483,
      "eval_samples_per_second": 40.459,
      "eval_steps_per_second": 1.349,
      "step": 248
    },
    {
      "epoch": 32.52,
      "grad_norm": 2.444944381713867,
      "learning_rate": 1.5909090909090907e-07,
      "loss": 1.0882,
      "step": 252
    },
    {
      "epoch": 32.9,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.1032441854476929,
      "eval_runtime": 1.5278,
      "eval_samples_per_second": 39.272,
      "eval_steps_per_second": 1.309,
      "step": 255
    },
    {
      "epoch": 33.94,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.101210355758667,
      "eval_runtime": 1.4836,
      "eval_samples_per_second": 40.443,
      "eval_steps_per_second": 1.348,
      "step": 263
    },
    {
      "epoch": 34.06,
      "grad_norm": 2.511270523071289,
      "learning_rate": 1.1363636363636363e-07,
      "loss": 1.0685,
      "step": 264
    },
    {
      "epoch": 34.97,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.099798560142517,
      "eval_runtime": 2.009,
      "eval_samples_per_second": 29.866,
      "eval_steps_per_second": 0.996,
      "step": 271
    },
    {
      "epoch": 35.61,
      "grad_norm": 2.6111106872558594,
      "learning_rate": 6.818181818181817e-08,
      "loss": 1.0775,
      "step": 276
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.0987868309020996,
      "eval_runtime": 1.5072,
      "eval_samples_per_second": 39.81,
      "eval_steps_per_second": 1.327,
      "step": 279
    },
    {
      "epoch": 36.9,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.098325252532959,
      "eval_runtime": 1.4717,
      "eval_samples_per_second": 40.77,
      "eval_steps_per_second": 1.359,
      "step": 286
    },
    {
      "epoch": 37.16,
      "grad_norm": 2.345568895339966,
      "learning_rate": 2.2727272727272725e-08,
      "loss": 1.0817,
      "step": 288
    },
    {
      "epoch": 37.94,
      "eval_accuracy": 0.5666666666666667,
      "eval_loss": 1.0981043577194214,
      "eval_runtime": 1.576,
      "eval_samples_per_second": 38.072,
      "eval_steps_per_second": 1.269,
      "step": 294
    },
    {
      "epoch": 37.94,
      "step": 294,
      "total_flos": 2.864620236542755e+18,
      "train_loss": 1.2232356217442726,
      "train_runtime": 1646.5284,
      "train_samples_per_second": 24.845,
      "train_steps_per_second": 0.179
    }
  ],
  "logging_steps": 12,
  "max_steps": 294,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 42,
  "save_steps": 500,
  "total_flos": 2.864620236542755e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}