last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.008442380751371888,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008442380751371887,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 6.2949,
      "step": 1
    },
    {
      "epoch": 0.0008442380751371887,
      "eval_loss": 7.167040824890137,
      "eval_runtime": 33.9503,
      "eval_samples_per_second": 14.698,
      "eval_steps_per_second": 7.364,
      "step": 1
    },
    {
      "epoch": 0.0016884761502743773,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 7.5932,
      "step": 2
    },
    {
      "epoch": 0.002532714225411566,
      "grad_norm": 4.345782279968262,
      "learning_rate": 2e-05,
      "loss": 6.3083,
      "step": 3
    },
    {
      "epoch": 0.002532714225411566,
      "eval_loss": 7.167040824890137,
      "eval_runtime": 33.7762,
      "eval_samples_per_second": 14.774,
      "eval_steps_per_second": 7.402,
      "step": 3
    },
    {
      "epoch": 0.0033769523005487546,
      "grad_norm": 4.228282451629639,
      "learning_rate": 4e-05,
      "loss": 6.5425,
      "step": 4
    },
    {
      "epoch": 0.004221190375685944,
      "grad_norm": 5.7301530838012695,
      "learning_rate": 6e-05,
      "loss": 7.4106,
      "step": 5
    },
    {
      "epoch": 0.005065428450823132,
      "grad_norm": 4.481907367706299,
      "learning_rate": 8e-05,
      "loss": 6.8524,
      "step": 6
    },
    {
      "epoch": 0.005065428450823132,
      "eval_loss": 6.865447521209717,
      "eval_runtime": 33.823,
      "eval_samples_per_second": 14.753,
      "eval_steps_per_second": 7.391,
      "step": 6
    },
    {
      "epoch": 0.0059096665259603205,
      "grad_norm": 3.831299066543579,
      "learning_rate": 0.0001,
      "loss": 5.5514,
      "step": 7
    },
    {
      "epoch": 0.006753904601097509,
      "grad_norm": NaN,
      "learning_rate": 0.0001,
      "loss": 6.6829,
      "step": 8
    },
    {
      "epoch": 0.007598142676234698,
      "grad_norm": 5.293268203735352,
      "learning_rate": 0.00012,
      "loss": 6.0991,
      "step": 9
    },
    {
      "epoch": 0.007598142676234698,
      "eval_loss": 6.2211456298828125,
      "eval_runtime": 33.8008,
      "eval_samples_per_second": 14.763,
      "eval_steps_per_second": 7.396,
      "step": 9
    },
    {
      "epoch": 0.008442380751371888,
      "grad_norm": 4.914198875427246,
      "learning_rate": 0.00014,
      "loss": 5.7317,
      "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": 1,
  "trial_name": null,
  "trial_params": null
}