{
  "best_metric": 0.8431372549019608,
  "best_model_checkpoint": "vit-base-patch16-224-RU9-24\\checkpoint-152",
  "epoch": 24.0,
  "eval_steps": 500,
  "global_step": 192,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5098039215686274,
      "eval_loss": 1.340147614479065,
      "eval_runtime": 1.0273,
      "eval_samples_per_second": 49.646,
      "eval_steps_per_second": 1.947,
      "step": 8
    },
    {
      "epoch": 1.25,
      "learning_rate": 5.5e-05,
      "loss": 1.3685,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5686274509803921,
      "eval_loss": 1.219266653060913,
      "eval_runtime": 0.799,
      "eval_samples_per_second": 63.826,
      "eval_steps_per_second": 2.503,
      "step": 16
    },
    {
      "epoch": 2.5,
      "learning_rate": 5.197802197802198e-05,
      "loss": 1.2413,
      "step": 20
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5882352941176471,
      "eval_loss": 1.1150447130203247,
      "eval_runtime": 0.7843,
      "eval_samples_per_second": 65.03,
      "eval_steps_per_second": 2.55,
      "step": 24
    },
    {
      "epoch": 3.75,
      "learning_rate": 4.8956043956043954e-05,
      "loss": 1.1126,
      "step": 30
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7058823529411765,
      "eval_loss": 0.995692253112793,
      "eval_runtime": 0.8045,
      "eval_samples_per_second": 63.394,
      "eval_steps_per_second": 2.486,
      "step": 32
    },
    {
      "epoch": 5.0,
      "learning_rate": 4.593406593406594e-05,
      "loss": 0.9285,
      "step": 40
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6862745098039216,
      "eval_loss": 0.897616982460022,
      "eval_runtime": 0.7886,
      "eval_samples_per_second": 64.671,
      "eval_steps_per_second": 2.536,
      "step": 40
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6862745098039216,
      "eval_loss": 0.8579562306404114,
      "eval_runtime": 0.7809,
      "eval_samples_per_second": 65.312,
      "eval_steps_per_second": 2.561,
      "step": 48
    },
    {
      "epoch": 6.25,
      "learning_rate": 4.291208791208791e-05,
      "loss": 0.7793,
      "step": 50
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.8426423072814941,
      "eval_runtime": 0.7688,
      "eval_samples_per_second": 66.339,
      "eval_steps_per_second": 2.602,
      "step": 56
    },
    {
      "epoch": 7.5,
      "learning_rate": 3.989010989010989e-05,
      "loss": 0.6291,
      "step": 60
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.6862745098039216,
      "eval_loss": 0.7899176478385925,
      "eval_runtime": 0.7937,
      "eval_samples_per_second": 64.258,
      "eval_steps_per_second": 2.52,
      "step": 64
    },
    {
      "epoch": 8.75,
      "learning_rate": 3.6868131868131865e-05,
      "loss": 0.5401,
      "step": 70
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.7254901960784313,
      "eval_loss": 0.716878354549408,
      "eval_runtime": 0.7851,
      "eval_samples_per_second": 64.962,
      "eval_steps_per_second": 2.548,
      "step": 72
    },
    {
      "epoch": 10.0,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.4358,
      "step": 80
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.7254901960784313,
      "eval_loss": 0.750454843044281,
      "eval_runtime": 0.7913,
      "eval_samples_per_second": 64.453,
      "eval_steps_per_second": 2.528,
      "step": 80
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.7058823529411765,
      "eval_loss": 0.8076674342155457,
      "eval_runtime": 0.7751,
      "eval_samples_per_second": 65.797,
      "eval_steps_per_second": 2.58,
      "step": 88
    },
    {
      "epoch": 11.25,
      "learning_rate": 3.0824175824175825e-05,
      "loss": 0.3901,
      "step": 90
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.6802822351455688,
      "eval_runtime": 0.8054,
      "eval_samples_per_second": 63.325,
      "eval_steps_per_second": 2.483,
      "step": 96
    },
    {
      "epoch": 12.5,
      "learning_rate": 2.78021978021978e-05,
      "loss": 0.3033,
      "step": 100
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.6483149528503418,
      "eval_runtime": 0.7691,
      "eval_samples_per_second": 66.31,
      "eval_steps_per_second": 2.6,
      "step": 104
    },
    {
      "epoch": 13.75,
      "learning_rate": 2.478021978021978e-05,
      "loss": 0.267,
      "step": 110
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.7450980392156863,
      "eval_loss": 0.6450825333595276,
      "eval_runtime": 0.8121,
      "eval_samples_per_second": 62.802,
      "eval_steps_per_second": 2.463,
      "step": 112
    },
    {
      "epoch": 15.0,
      "learning_rate": 2.175824175824176e-05,
      "loss": 0.2212,
      "step": 120
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.6118857264518738,
      "eval_runtime": 0.7703,
      "eval_samples_per_second": 66.211,
      "eval_steps_per_second": 2.597,
      "step": 120
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.614970326423645,
      "eval_runtime": 0.7768,
      "eval_samples_per_second": 65.657,
      "eval_steps_per_second": 2.575,
      "step": 128
    },
    {
      "epoch": 16.25,
      "learning_rate": 1.873626373626374e-05,
      "loss": 0.2206,
      "step": 130
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.7843137254901961,
      "eval_loss": 0.6270039081573486,
      "eval_runtime": 0.8131,
      "eval_samples_per_second": 62.722,
      "eval_steps_per_second": 2.46,
      "step": 136
    },
    {
      "epoch": 17.5,
      "learning_rate": 1.5714285714285715e-05,
      "loss": 0.2285,
      "step": 140
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.6181018352508545,
      "eval_runtime": 0.7654,
      "eval_samples_per_second": 66.634,
      "eval_steps_per_second": 2.613,
      "step": 144
    },
    {
      "epoch": 18.75,
      "learning_rate": 1.2692307692307693e-05,
      "loss": 0.1741,
      "step": 150
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.8431372549019608,
      "eval_loss": 0.5081135630607605,
      "eval_runtime": 0.7922,
      "eval_samples_per_second": 64.379,
      "eval_steps_per_second": 2.525,
      "step": 152
    },
    {
      "epoch": 20.0,
      "learning_rate": 9.670329670329671e-06,
      "loss": 0.1708,
      "step": 160
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8235294117647058,
      "eval_loss": 0.5501907467842102,
      "eval_runtime": 0.7732,
      "eval_samples_per_second": 65.96,
      "eval_steps_per_second": 2.587,
      "step": 160
    },
    {
      "epoch": 21.0,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.5689105987548828,
      "eval_runtime": 0.7988,
      "eval_samples_per_second": 63.843,
      "eval_steps_per_second": 2.504,
      "step": 168
    },
    {
      "epoch": 21.25,
      "learning_rate": 6.648351648351649e-06,
      "loss": 0.16,
      "step": 170
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.8235294117647058,
      "eval_loss": 0.5137497186660767,
      "eval_runtime": 0.7803,
      "eval_samples_per_second": 65.361,
      "eval_steps_per_second": 2.563,
      "step": 176
    },
    {
      "epoch": 22.5,
      "learning_rate": 3.6263736263736266e-06,
      "loss": 0.1567,
      "step": 180
    },
    {
      "epoch": 23.0,
      "eval_accuracy": 0.8431372549019608,
      "eval_loss": 0.5207127332687378,
      "eval_runtime": 0.7678,
      "eval_samples_per_second": 66.422,
      "eval_steps_per_second": 2.605,
      "step": 184
    },
    {
      "epoch": 23.75,
      "learning_rate": 6.043956043956044e-07,
      "loss": 0.1616,
      "step": 190
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.8235294117647058,
      "eval_loss": 0.5374550223350525,
      "eval_runtime": 0.821,
      "eval_samples_per_second": 62.116,
      "eval_steps_per_second": 2.436,
      "step": 192
    },
    {
      "epoch": 24.0,
      "step": 192,
      "total_flos": 1.8877387048147354e+18,
      "train_loss": 0.4960318058729172,
      "train_runtime": 413.057,
      "train_samples_per_second": 58.975,
      "train_steps_per_second": 0.465
    }
  ],
  "logging_steps": 10,
  "max_steps": 192,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 24,
  "save_steps": 500,
  "total_flos": 1.8877387048147354e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}