trainer_state.json

{
  "best_metric": 0.41995841995842,
  "best_model_checkpoint": "test-hasy-6/checkpoint-1082",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1082,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.92,
      "grad_norm": 3.0262813568115234,
      "learning_rate": 1.0757855822550833e-05,
      "loss": 4.227,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.38461538461538464,
      "eval_loss": 3.6833133697509766,
      "eval_runtime": 2.0272,
      "eval_samples_per_second": 237.278,
      "eval_steps_per_second": 30.091,
      "step": 541
    },
    {
      "epoch": 1.85,
      "grad_norm": 3.8626766204833984,
      "learning_rate": 1.5157116451016638e-06,
      "loss": 3.756,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.41995841995842,
      "eval_loss": 3.446028709411621,
      "eval_runtime": 1.9847,
      "eval_samples_per_second": 242.357,
      "eval_steps_per_second": 30.735,
      "step": 1082
    },
    {
      "epoch": 2.0,
      "step": 1082,
      "total_flos": 6.710386542096384e+17,
      "train_loss": 3.966065468497285,
      "train_runtime": 108.0206,
      "train_samples_per_second": 80.077,
      "train_steps_per_second": 10.017
    }
  ],
  "logging_steps": 500,
  "max_steps": 1082,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 6.710386542096384e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}