checkpoint-30/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.8421052631578947,
  "eval_steps": 10,
  "global_step": 30,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.7017543859649122,
      "grad_norm": 44.80005645751953,
      "learning_rate": 0.00017333333333333334,
      "loss": 2.0998,
      "step": 5
    },
    {
      "epoch": 1.280701754385965,
      "grad_norm": 901.4852294921875,
      "learning_rate": 0.00016,
      "loss": 1.986,
      "step": 10
    },
    {
      "epoch": 1.280701754385965,
      "eval_loss": 0.7075792551040649,
      "eval_runtime": 31.4718,
      "eval_samples_per_second": 0.731,
      "eval_steps_per_second": 0.381,
      "step": 10
    },
    {
      "epoch": 1.9824561403508771,
      "grad_norm": 42.54853439331055,
      "learning_rate": 0.00012666666666666666,
      "loss": 2.1037,
      "step": 15
    },
    {
      "epoch": 2.56140350877193,
      "grad_norm": 84.47293853759766,
      "learning_rate": 9.333333333333334e-05,
      "loss": 1.2837,
      "step": 20
    },
    {
      "epoch": 2.56140350877193,
      "eval_loss": 0.6040331125259399,
      "eval_runtime": 31.5373,
      "eval_samples_per_second": 0.729,
      "eval_steps_per_second": 0.381,
      "step": 20
    },
    {
      "epoch": 3.1403508771929824,
      "grad_norm": 52.28364181518555,
      "learning_rate": 6e-05,
      "loss": 1.1705,
      "step": 25
    },
    {
      "epoch": 3.8421052631578947,
      "grad_norm": 16.730497360229492,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 1.2169,
      "step": 30
    },
    {
      "epoch": 3.8421052631578947,
      "eval_loss": 0.5927651524543762,
      "eval_runtime": 31.4582,
      "eval_samples_per_second": 0.731,
      "eval_steps_per_second": 0.381,
      "step": 30
    }
  ],
  "logging_steps": 5,
  "max_steps": 30,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 25,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2043610494331296.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}