last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.03475238922675934,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0034752389226759338,
      "grad_norm": 2.5032899379730225,
      "learning_rate": 2e-05,
      "loss": 3.6742,
      "step": 1
    },
    {
      "epoch": 0.0034752389226759338,
      "eval_loss": 3.678018093109131,
      "eval_runtime": 12.7145,
      "eval_samples_per_second": 4.798,
      "eval_steps_per_second": 4.798,
      "step": 1
    },
    {
      "epoch": 0.0069504778453518675,
      "grad_norm": 2.854793071746826,
      "learning_rate": 4e-05,
      "loss": 3.6066,
      "step": 2
    },
    {
      "epoch": 0.010425716768027803,
      "grad_norm": 2.463970422744751,
      "learning_rate": 6e-05,
      "loss": 3.2883,
      "step": 3
    },
    {
      "epoch": 0.010425716768027803,
      "eval_loss": 3.6363368034362793,
      "eval_runtime": 12.7217,
      "eval_samples_per_second": 4.795,
      "eval_steps_per_second": 4.795,
      "step": 3
    },
    {
      "epoch": 0.013900955690703735,
      "grad_norm": 3.6273467540740967,
      "learning_rate": 8e-05,
      "loss": 3.9102,
      "step": 4
    },
    {
      "epoch": 0.01737619461337967,
      "grad_norm": 2.674269914627075,
      "learning_rate": 0.0001,
      "loss": 3.4218,
      "step": 5
    },
    {
      "epoch": 0.020851433536055605,
      "grad_norm": 3.4425220489501953,
      "learning_rate": 0.00012,
      "loss": 3.2851,
      "step": 6
    },
    {
      "epoch": 0.020851433536055605,
      "eval_loss": 3.34614634513855,
      "eval_runtime": 12.7399,
      "eval_samples_per_second": 4.788,
      "eval_steps_per_second": 4.788,
      "step": 6
    },
    {
      "epoch": 0.02432667245873154,
      "grad_norm": 3.534543037414551,
      "learning_rate": 0.00014,
      "loss": 3.5841,
      "step": 7
    },
    {
      "epoch": 0.02780191138140747,
      "grad_norm": NaN,
      "learning_rate": 0.00014,
      "loss": 2.8247,
      "step": 8
    },
    {
      "epoch": 0.03127715030408341,
      "grad_norm": 4.276378154754639,
      "learning_rate": 0.00016,
      "loss": 2.4532,
      "step": 9
    },
    {
      "epoch": 0.03127715030408341,
      "eval_loss": 2.911003828048706,
      "eval_runtime": 12.7634,
      "eval_samples_per_second": 4.779,
      "eval_steps_per_second": 4.779,
      "step": 9
    },
    {
      "epoch": 0.03475238922675934,
      "grad_norm": 4.4422430992126465,
      "learning_rate": 0.00018,
      "loss": 2.787,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7418872835604480.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}