last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.004764173415912339,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0004764173415912339,
      "grad_norm": 1.3637402057647705,
      "learning_rate": 2e-05,
      "loss": 2.8479,
      "step": 1
    },
    {
      "epoch": 0.0004764173415912339,
      "eval_loss": NaN,
      "eval_runtime": 49.5822,
      "eval_samples_per_second": 17.829,
      "eval_steps_per_second": 8.914,
      "step": 1
    },
    {
      "epoch": 0.0009528346831824678,
      "grad_norm": 0.9972659349441528,
      "learning_rate": 4e-05,
      "loss": 2.8832,
      "step": 2
    },
    {
      "epoch": 0.0014292520247737017,
      "grad_norm": 1.3149476051330566,
      "learning_rate": 6e-05,
      "loss": 3.3305,
      "step": 3
    },
    {
      "epoch": 0.0014292520247737017,
      "eval_loss": NaN,
      "eval_runtime": 48.3871,
      "eval_samples_per_second": 18.269,
      "eval_steps_per_second": 9.135,
      "step": 3
    },
    {
      "epoch": 0.0019056693663649356,
      "grad_norm": 0.9289590716362,
      "learning_rate": 8e-05,
      "loss": 2.595,
      "step": 4
    },
    {
      "epoch": 0.0023820867079561697,
      "grad_norm": 0.8951483964920044,
      "learning_rate": 0.0001,
      "loss": 2.8625,
      "step": 5
    },
    {
      "epoch": 0.0028585040495474035,
      "grad_norm": 1.0807698965072632,
      "learning_rate": 0.00012,
      "loss": 2.9945,
      "step": 6
    },
    {
      "epoch": 0.0028585040495474035,
      "eval_loss": NaN,
      "eval_runtime": 48.4269,
      "eval_samples_per_second": 18.254,
      "eval_steps_per_second": 9.127,
      "step": 6
    },
    {
      "epoch": 0.0033349213911386373,
      "grad_norm": 1.1401915550231934,
      "learning_rate": 0.00014,
      "loss": 2.9902,
      "step": 7
    },
    {
      "epoch": 0.003811338732729871,
      "grad_norm": 0.9969739317893982,
      "learning_rate": 0.00016,
      "loss": 2.6715,
      "step": 8
    },
    {
      "epoch": 0.004287756074321105,
      "grad_norm": 1.4261023998260498,
      "learning_rate": 0.00018,
      "loss": 2.9923,
      "step": 9
    },
    {
      "epoch": 0.004287756074321105,
      "eval_loss": NaN,
      "eval_runtime": 48.4412,
      "eval_samples_per_second": 18.249,
      "eval_steps_per_second": 9.124,
      "step": 9
    },
    {
      "epoch": 0.004764173415912339,
      "grad_norm": 1.0647661685943604,
      "learning_rate": 0.0002,
      "loss": 2.5559,
      "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": 3699128496291840.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}