{
  "best_metric": 0.949438202247191,
  "best_model_checkpoint": "teacher-status-van-tiny-256/checkpoint-12",
  "epoch": 2.88,
  "eval_steps": 500,
  "global_step": 36,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8,
      "learning_rate": 4.0625000000000005e-05,
      "loss": 0.1511,
      "step": 10
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.949438202247191,
      "eval_f1_score": 0.9560975609756097,
      "eval_loss": 0.18015821278095245,
      "eval_precision": 0.9607843137254902,
      "eval_recall": 0.9514563106796117,
      "eval_runtime": 1.3097,
      "eval_samples_per_second": 135.904,
      "eval_steps_per_second": 4.581,
      "step": 12
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.5e-05,
      "loss": 0.2643,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.949438202247191,
      "eval_f1_score": 0.9556650246305418,
      "eval_loss": 0.16739808022975922,
      "eval_precision": 0.97,
      "eval_recall": 0.941747572815534,
      "eval_runtime": 1.1345,
      "eval_samples_per_second": 156.899,
      "eval_steps_per_second": 5.289,
      "step": 25
    },
    {
      "epoch": 2.4,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.3159,
      "step": 30
    },
    {
      "epoch": 2.88,
      "eval_accuracy": 0.9438202247191011,
      "eval_f1_score": 0.9509803921568628,
      "eval_loss": 0.16916657984256744,
      "eval_precision": 0.9603960396039604,
      "eval_recall": 0.941747572815534,
      "eval_runtime": 1.0079,
      "eval_samples_per_second": 176.605,
      "eval_steps_per_second": 5.953,
      "step": 36
    },
    {
      "epoch": 2.88,
      "step": 36,
      "total_flos": 2.086337325249331e+16,
      "train_loss": 0.2572544482019212,
      "train_runtime": 70.0705,
      "train_samples_per_second": 68.203,
      "train_steps_per_second": 0.514
    }
  ],
  "logging_steps": 10,
  "max_steps": 36,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 2.086337325249331e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}