trainer_state.json

{
  "best_metric": 1.0,
  "best_model_checkpoint": "output_dir_can/checkpoint-176",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 880,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_f1": 1.0,
      "eval_loss": 0.0009332114714197814,
      "eval_runtime": 2.1327,
      "eval_samples_per_second": 146.763,
      "eval_steps_per_second": 18.756,
      "step": 176
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.012840413488447666,
      "learning_rate": 3.8636363636363636e-05,
      "loss": 0.0856,
      "step": 200
    },
    {
      "epoch": 2.0,
      "eval_f1": 1.0,
      "eval_loss": 0.00045387583668343723,
      "eval_runtime": 2.0824,
      "eval_samples_per_second": 150.307,
      "eval_steps_per_second": 19.209,
      "step": 352
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.005864701699465513,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.0009,
      "step": 400
    },
    {
      "epoch": 3.0,
      "eval_f1": 1.0,
      "eval_loss": 0.00031038594897836447,
      "eval_runtime": 2.0744,
      "eval_samples_per_second": 150.89,
      "eval_steps_per_second": 19.283,
      "step": 528
    },
    {
      "epoch": 3.41,
      "grad_norm": 0.0048249149695038795,
      "learning_rate": 1.590909090909091e-05,
      "loss": 0.0005,
      "step": 600
    },
    {
      "epoch": 4.0,
      "eval_f1": 1.0,
      "eval_loss": 0.0002527358883526176,
      "eval_runtime": 2.0799,
      "eval_samples_per_second": 150.485,
      "eval_steps_per_second": 19.231,
      "step": 704
    },
    {
      "epoch": 4.55,
      "grad_norm": 0.0037010444793850183,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.0004,
      "step": 800
    },
    {
      "epoch": 5.0,
      "eval_f1": 1.0,
      "eval_loss": 0.00023499388771597296,
      "eval_runtime": 2.0759,
      "eval_samples_per_second": 150.781,
      "eval_steps_per_second": 19.269,
      "step": 880
    }
  ],
  "logging_steps": 200,
  "max_steps": 880,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 3704736708034560.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}