trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9723756906077348,
  "eval_steps": 22,
  "global_step": 88,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011049723756906077,
      "grad_norm": 0.21801863610744476,
      "learning_rate": 0.0005,
      "loss": 0.7346,
      "step": 1
    },
    {
      "epoch": 0.2430939226519337,
      "grad_norm": 0.31375670433044434,
      "learning_rate": 0.0005,
      "loss": 0.6429,
      "step": 22
    },
    {
      "epoch": 0.2430939226519337,
      "eval_loss": 0.6063512563705444,
      "eval_runtime": 264.3898,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 22
    },
    {
      "epoch": 0.4861878453038674,
      "grad_norm": 0.22838832437992096,
      "learning_rate": 0.0005,
      "loss": 0.5907,
      "step": 44
    },
    {
      "epoch": 0.4861878453038674,
      "eval_loss": 0.589008629322052,
      "eval_runtime": 264.4169,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 44
    },
    {
      "epoch": 0.7292817679558011,
      "grad_norm": 0.24711200594902039,
      "learning_rate": 0.0005,
      "loss": 0.5933,
      "step": 66
    },
    {
      "epoch": 0.7292817679558011,
      "eval_loss": 0.5814085006713867,
      "eval_runtime": 264.3062,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 66
    },
    {
      "epoch": 0.9723756906077348,
      "grad_norm": 0.2660823166370392,
      "learning_rate": 0.0005,
      "loss": 0.5547,
      "step": 88
    },
    {
      "epoch": 0.9723756906077348,
      "eval_loss": 0.573217511177063,
      "eval_runtime": 264.4531,
      "eval_samples_per_second": 0.318,
      "eval_steps_per_second": 0.318,
      "step": 88
    }
  ],
  "logging_steps": 22,
  "max_steps": 270,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 22,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.9250732494494515e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}