{
  "best_metric": 0.8431372549019608,
  "best_model_checkpoint": "vit-base-patch16-224-RXL1-24\\checkpoint-96",
  "epoch": 22.69090909090909,
  "eval_steps": 500,
  "global_step": 312,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.73,
      "learning_rate": 3.4375e-05,
      "loss": 1.3745,
      "step": 10
    },
    {
      "epoch": 0.95,
      "eval_accuracy": 0.47058823529411764,
      "eval_loss": 1.3056340217590332,
      "eval_runtime": 0.7424,
      "eval_samples_per_second": 68.693,
      "eval_steps_per_second": 2.694,
      "step": 13
    },
    {
      "epoch": 1.45,
      "learning_rate": 5.4256756756756764e-05,
      "loss": 1.2896,
      "step": 20
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 0.6470588235294118,
      "eval_loss": 1.103859543800354,
      "eval_runtime": 0.7848,
      "eval_samples_per_second": 64.988,
      "eval_steps_per_second": 2.549,
      "step": 27
    },
    {
      "epoch": 2.18,
      "learning_rate": 5.239864864864865e-05,
      "loss": 1.1563,
      "step": 30
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.0540540540540544e-05,
      "loss": 0.9896,
      "step": 40
    },
    {
      "epoch": 2.98,
      "eval_accuracy": 0.6470588235294118,
      "eval_loss": 0.9412985444068909,
      "eval_runtime": 0.7684,
      "eval_samples_per_second": 66.368,
      "eval_steps_per_second": 2.603,
      "step": 41
    },
    {
      "epoch": 3.64,
      "learning_rate": 4.868243243243243e-05,
      "loss": 0.8472,
      "step": 50
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6274509803921569,
      "eval_loss": 0.905930757522583,
      "eval_runtime": 0.7968,
      "eval_samples_per_second": 64.003,
      "eval_steps_per_second": 2.51,
      "step": 55
    },
    {
      "epoch": 4.36,
      "learning_rate": 4.6824324324324325e-05,
      "loss": 0.7375,
      "step": 60
    },
    {
      "epoch": 4.95,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.6519917249679565,
      "eval_runtime": 0.822,
      "eval_samples_per_second": 62.047,
      "eval_steps_per_second": 2.433,
      "step": 68
    },
    {
      "epoch": 5.09,
      "learning_rate": 4.496621621621622e-05,
      "loss": 0.5786,
      "step": 70
    },
    {
      "epoch": 5.82,
      "learning_rate": 4.3108108108108106e-05,
      "loss": 0.458,
      "step": 80
    },
    {
      "epoch": 5.96,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.6753640174865723,
      "eval_runtime": 0.8618,
      "eval_samples_per_second": 59.178,
      "eval_steps_per_second": 2.321,
      "step": 82
    },
    {
      "epoch": 6.55,
      "learning_rate": 4.125e-05,
      "loss": 0.3807,
      "step": 90
    },
    {
      "epoch": 6.98,
      "eval_accuracy": 0.8431372549019608,
      "eval_loss": 0.6158193945884705,
      "eval_runtime": 0.7813,
      "eval_samples_per_second": 65.275,
      "eval_steps_per_second": 2.56,
      "step": 96
    },
    {
      "epoch": 7.27,
      "learning_rate": 3.9391891891891894e-05,
      "loss": 0.3282,
      "step": 100
    },
    {
      "epoch": 8.0,
      "learning_rate": 3.753378378378379e-05,
      "loss": 0.3003,
      "step": 110
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.5665538311004639,
      "eval_runtime": 0.7924,
      "eval_samples_per_second": 64.358,
      "eval_steps_per_second": 2.524,
      "step": 110
    },
    {
      "epoch": 8.73,
      "learning_rate": 3.567567567567568e-05,
      "loss": 0.2337,
      "step": 120
    },
    {
      "epoch": 8.95,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.5409361124038696,
      "eval_runtime": 0.8091,
      "eval_samples_per_second": 63.033,
      "eval_steps_per_second": 2.472,
      "step": 123
    },
    {
      "epoch": 9.45,
      "learning_rate": 3.381756756756757e-05,
      "loss": 0.2252,
      "step": 130
    },
    {
      "epoch": 9.96,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.7381905913352966,
      "eval_runtime": 0.8081,
      "eval_samples_per_second": 63.11,
      "eval_steps_per_second": 2.475,
      "step": 137
    },
    {
      "epoch": 10.18,
      "learning_rate": 3.1959459459459455e-05,
      "loss": 0.1952,
      "step": 140
    },
    {
      "epoch": 10.91,
      "learning_rate": 3.010135135135135e-05,
      "loss": 0.1644,
      "step": 150
    },
    {
      "epoch": 10.98,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.6363002061843872,
      "eval_runtime": 0.7874,
      "eval_samples_per_second": 64.774,
      "eval_steps_per_second": 2.54,
      "step": 151
    },
    {
      "epoch": 11.64,
      "learning_rate": 2.8243243243243243e-05,
      "loss": 0.1608,
      "step": 160
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.6941252946853638,
      "eval_runtime": 0.809,
      "eval_samples_per_second": 63.04,
      "eval_steps_per_second": 2.472,
      "step": 165
    },
    {
      "epoch": 12.36,
      "learning_rate": 2.6385135135135137e-05,
      "loss": 0.1354,
      "step": 170
    },
    {
      "epoch": 12.95,
      "eval_accuracy": 0.7843137254901961,
      "eval_loss": 0.6985132694244385,
      "eval_runtime": 0.7967,
      "eval_samples_per_second": 64.012,
      "eval_steps_per_second": 2.51,
      "step": 178
    },
    {
      "epoch": 13.09,
      "learning_rate": 2.4527027027027027e-05,
      "loss": 0.1287,
      "step": 180
    },
    {
      "epoch": 13.82,
      "learning_rate": 2.266891891891892e-05,
      "loss": 0.1298,
      "step": 190
    },
    {
      "epoch": 13.96,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.6610224843025208,
      "eval_runtime": 0.8038,
      "eval_samples_per_second": 63.446,
      "eval_steps_per_second": 2.488,
      "step": 192
    },
    {
      "epoch": 14.55,
      "learning_rate": 2.081081081081081e-05,
      "loss": 0.1333,
      "step": 200
    },
    {
      "epoch": 14.98,
      "eval_accuracy": 0.803921568627451,
      "eval_loss": 0.675108015537262,
      "eval_runtime": 0.7776,
      "eval_samples_per_second": 65.588,
      "eval_steps_per_second": 2.572,
      "step": 206
    },
    {
      "epoch": 15.27,
      "learning_rate": 1.89527027027027e-05,
      "loss": 0.1241,
      "step": 210
    },
    {
      "epoch": 16.0,
      "learning_rate": 1.7094594594594595e-05,
      "loss": 0.1209,
      "step": 220
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.7843137254901961,
      "eval_loss": 0.7722602486610413,
      "eval_runtime": 0.7727,
      "eval_samples_per_second": 66.007,
      "eval_steps_per_second": 2.588,
      "step": 220
    },
    {
      "epoch": 16.73,
      "learning_rate": 1.5236486486486487e-05,
      "loss": 0.1057,
      "step": 230
    },
    {
      "epoch": 16.95,
      "eval_accuracy": 0.7254901960784313,
      "eval_loss": 0.8037699460983276,
      "eval_runtime": 0.7639,
      "eval_samples_per_second": 66.761,
      "eval_steps_per_second": 2.618,
      "step": 233
    },
    {
      "epoch": 17.45,
      "learning_rate": 1.337837837837838e-05,
      "loss": 0.0972,
      "step": 240
    },
    {
      "epoch": 17.96,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.8375279307365417,
      "eval_runtime": 0.7813,
      "eval_samples_per_second": 65.279,
      "eval_steps_per_second": 2.56,
      "step": 247
    },
    {
      "epoch": 18.18,
      "learning_rate": 1.152027027027027e-05,
      "loss": 0.0974,
      "step": 250
    },
    {
      "epoch": 18.91,
      "learning_rate": 9.662162162162164e-06,
      "loss": 0.0789,
      "step": 260
    },
    {
      "epoch": 18.98,
      "eval_accuracy": 0.8235294117647058,
      "eval_loss": 0.697125256061554,
      "eval_runtime": 0.8068,
      "eval_samples_per_second": 63.21,
      "eval_steps_per_second": 2.479,
      "step": 261
    },
    {
      "epoch": 19.64,
      "learning_rate": 7.804054054054054e-06,
      "loss": 0.0833,
      "step": 270
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.7843137254901961,
      "eval_loss": 0.7507321238517761,
      "eval_runtime": 0.8032,
      "eval_samples_per_second": 63.499,
      "eval_steps_per_second": 2.49,
      "step": 275
    },
    {
      "epoch": 20.36,
      "learning_rate": 5.945945945945946e-06,
      "loss": 0.0813,
      "step": 280
    },
    {
      "epoch": 20.95,
      "eval_accuracy": 0.7843137254901961,
      "eval_loss": 0.7085339426994324,
      "eval_runtime": 0.7675,
      "eval_samples_per_second": 66.445,
      "eval_steps_per_second": 2.606,
      "step": 288
    },
    {
      "epoch": 21.09,
      "learning_rate": 4.087837837837838e-06,
      "loss": 0.0874,
      "step": 290
    },
    {
      "epoch": 21.82,
      "learning_rate": 2.22972972972973e-06,
      "loss": 0.0803,
      "step": 300
    },
    {
      "epoch": 21.96,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.7565922737121582,
      "eval_runtime": 0.8044,
      "eval_samples_per_second": 63.404,
      "eval_steps_per_second": 2.486,
      "step": 302
    },
    {
      "epoch": 22.55,
      "learning_rate": 3.7162162162162164e-07,
      "loss": 0.0693,
      "step": 310
    },
    {
      "epoch": 22.69,
      "eval_accuracy": 0.7647058823529411,
      "eval_loss": 0.7772119045257568,
      "eval_runtime": 0.7864,
      "eval_samples_per_second": 64.855,
      "eval_steps_per_second": 2.543,
      "step": 312
    },
    {
      "epoch": 22.69,
      "step": 312,
      "total_flos": 3.0760221736008253e+18,
      "train_loss": 0.3522763775709348,
      "train_runtime": 628.612,
      "train_samples_per_second": 66.776,
      "train_steps_per_second": 0.496
    }
  ],
  "logging_steps": 10,
  "max_steps": 312,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 24,
  "save_steps": 500,
  "total_flos": 3.0760221736008253e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}