last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2689075630252101,
  "eval_steps": 6,
  "global_step": 8,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03361344537815126,
      "eval_loss": 0.4333963394165039,
      "eval_runtime": 0.9506,
      "eval_samples_per_second": 13.675,
      "eval_steps_per_second": 7.364,
      "step": 1
    },
    {
      "epoch": 0.16806722689075632,
      "grad_norm": 0.773507297039032,
      "learning_rate": 5e-05,
      "loss": 0.4071,
      "step": 5
    },
    {
      "epoch": 0.20168067226890757,
      "eval_loss": 0.3601323664188385,
      "eval_runtime": 0.9296,
      "eval_samples_per_second": 13.984,
      "eval_steps_per_second": 7.53,
      "step": 6
    }
  ],
  "logging_steps": 5,
  "max_steps": 30,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 8,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1302987314036736.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}