{
  "best_metric": 0.9535256410256411,
  "best_model_checkpoint": "text/train/checkpoint/biobert/ddi_42/checkpoint-3955",
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 7910,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9411057692307693,
      "eval_loss": 0.1755092442035675,
      "eval_runtime": 4.5748,
      "eval_samples_per_second": 545.6,
      "eval_steps_per_second": 2.186,
      "step": 791
    },
    {
      "epoch": 1.26,
      "grad_norm": 5.126912593841553,
      "learning_rate": 4.367888748419722e-05,
      "loss": 0.2029,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9495192307692307,
      "eval_loss": 0.19307143986225128,
      "eval_runtime": 4.5897,
      "eval_samples_per_second": 543.827,
      "eval_steps_per_second": 2.179,
      "step": 1582
    },
    {
      "epoch": 2.53,
      "grad_norm": 4.425409317016602,
      "learning_rate": 3.735777496839444e-05,
      "loss": 0.0749,
      "step": 2000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9475160256410257,
      "eval_loss": 0.27157464623451233,
      "eval_runtime": 4.611,
      "eval_samples_per_second": 541.313,
      "eval_steps_per_second": 2.169,
      "step": 2373
    },
    {
      "epoch": 3.79,
      "grad_norm": 0.21999871730804443,
      "learning_rate": 3.1036662452591655e-05,
      "loss": 0.0442,
      "step": 3000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9515224358974359,
      "eval_loss": 0.25661203265190125,
      "eval_runtime": 4.5979,
      "eval_samples_per_second": 542.858,
      "eval_steps_per_second": 2.175,
      "step": 3164
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9535256410256411,
      "eval_loss": 0.2816075086593628,
      "eval_runtime": 4.559,
      "eval_samples_per_second": 547.487,
      "eval_steps_per_second": 2.193,
      "step": 3955
    },
    {
      "epoch": 5.06,
      "grad_norm": 0.018685705959796906,
      "learning_rate": 2.4715549936788876e-05,
      "loss": 0.0249,
      "step": 4000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9471153846153846,
      "eval_loss": 0.31359541416168213,
      "eval_runtime": 4.623,
      "eval_samples_per_second": 539.91,
      "eval_steps_per_second": 2.163,
      "step": 4746
    },
    {
      "epoch": 6.32,
      "grad_norm": 0.039792679250240326,
      "learning_rate": 1.8394437420986094e-05,
      "loss": 0.0135,
      "step": 5000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9475160256410257,
      "eval_loss": 0.3218846023082733,
      "eval_runtime": 4.5644,
      "eval_samples_per_second": 546.846,
      "eval_steps_per_second": 2.191,
      "step": 5537
    },
    {
      "epoch": 7.59,
      "grad_norm": 0.004741465672850609,
      "learning_rate": 1.2073324905183313e-05,
      "loss": 0.0102,
      "step": 6000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9503205128205128,
      "eval_loss": 0.30540913343429565,
      "eval_runtime": 4.5686,
      "eval_samples_per_second": 546.34,
      "eval_steps_per_second": 2.189,
      "step": 6328
    },
    {
      "epoch": 8.85,
      "grad_norm": 0.0014904498821124434,
      "learning_rate": 5.752212389380531e-06,
      "loss": 0.006,
      "step": 7000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.9515224358974359,
      "eval_loss": 0.3304109275341034,
      "eval_runtime": 4.5703,
      "eval_samples_per_second": 546.14,
      "eval_steps_per_second": 2.188,
      "step": 7119
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.9507211538461539,
      "eval_loss": 0.334633469581604,
      "eval_runtime": 4.6402,
      "eval_samples_per_second": 537.905,
      "eval_steps_per_second": 2.155,
      "step": 7910
    },
    {
      "epoch": 10.0,
      "step": 7910,
      "total_flos": 1.663959132942336e+16,
      "train_loss": 0.04803710016981238,
      "train_runtime": 1426.0933,
      "train_samples_per_second": 177.38,
      "train_steps_per_second": 5.547
    }
  ],
  "logging_steps": 1000,
  "max_steps": 7910,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.663959132942336e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}