last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12578616352201258,
  "eval_steps": 20,
  "global_step": 40,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0031446540880503146,
      "eval_loss": 11.922277450561523,
      "eval_runtime": 4.8133,
      "eval_samples_per_second": 234.767,
      "eval_steps_per_second": 7.479,
      "step": 1
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 1.364006519317627,
      "learning_rate": 0.00017891405093963938,
      "loss": 11.8697,
      "step": 10
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 1.143991231918335,
      "learning_rate": 0.00010825793454723325,
      "loss": 11.7325,
      "step": 20
    },
    {
      "epoch": 0.06289308176100629,
      "eval_loss": 11.678323745727539,
      "eval_runtime": 4.8974,
      "eval_samples_per_second": 230.736,
      "eval_steps_per_second": 7.351,
      "step": 20
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 1.1743687391281128,
      "learning_rate": 3.227184283742591e-05,
      "loss": 11.6488,
      "step": 30
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 1.1511428356170654,
      "learning_rate": 0.0,
      "loss": 11.6304,
      "step": 40
    },
    {
      "epoch": 0.12578616352201258,
      "eval_loss": 11.628145217895508,
      "eval_runtime": 4.8142,
      "eval_samples_per_second": 234.723,
      "eval_steps_per_second": 7.478,
      "step": 40
    }
  ],
  "logging_steps": 10,
  "max_steps": 40,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 40,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 78950726369280.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}