last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.0684931506849315,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00684931506849315,
      "grad_norm": NaN,
      "learning_rate": 0.0,
      "loss": 3.5612,
      "step": 1
    },
    {
      "epoch": 0.00684931506849315,
      "eval_loss": 4.457667350769043,
      "eval_runtime": 6.9114,
      "eval_samples_per_second": 35.593,
      "eval_steps_per_second": 4.485,
      "step": 1
    },
    {
      "epoch": 0.0136986301369863,
      "grad_norm": 2.0711398124694824,
      "learning_rate": 2e-05,
      "loss": 3.849,
      "step": 2
    },
    {
      "epoch": 0.02054794520547945,
      "grad_norm": 0.8428272604942322,
      "learning_rate": 4e-05,
      "loss": 3.1357,
      "step": 3
    },
    {
      "epoch": 0.02054794520547945,
      "eval_loss": 4.422769069671631,
      "eval_runtime": 6.7076,
      "eval_samples_per_second": 36.675,
      "eval_steps_per_second": 4.622,
      "step": 3
    },
    {
      "epoch": 0.0273972602739726,
      "grad_norm": 1.4228770732879639,
      "learning_rate": 6e-05,
      "loss": 3.3901,
      "step": 4
    },
    {
      "epoch": 0.03424657534246575,
      "grad_norm": 1.194941520690918,
      "learning_rate": 8e-05,
      "loss": 3.1024,
      "step": 5
    },
    {
      "epoch": 0.0410958904109589,
      "grad_norm": NaN,
      "learning_rate": 8e-05,
      "loss": 2.7753,
      "step": 6
    },
    {
      "epoch": 0.0410958904109589,
      "eval_loss": 4.003295421600342,
      "eval_runtime": 6.7529,
      "eval_samples_per_second": 36.429,
      "eval_steps_per_second": 4.591,
      "step": 6
    },
    {
      "epoch": 0.04794520547945205,
      "grad_norm": 1.3150415420532227,
      "learning_rate": 0.0001,
      "loss": 2.7825,
      "step": 7
    },
    {
      "epoch": 0.0547945205479452,
      "grad_norm": 1.7403879165649414,
      "learning_rate": 0.00012,
      "loss": 2.8134,
      "step": 8
    },
    {
      "epoch": 0.06164383561643835,
      "grad_norm": 1.472294569015503,
      "learning_rate": 0.00014,
      "loss": 2.5475,
      "step": 9
    },
    {
      "epoch": 0.06164383561643835,
      "eval_loss": 2.3513693809509277,
      "eval_runtime": 6.7545,
      "eval_samples_per_second": 36.42,
      "eval_steps_per_second": 4.59,
      "step": 9
    },
    {
      "epoch": 0.0684931506849315,
      "grad_norm": 0.9450744390487671,
      "learning_rate": 0.00016,
      "loss": 2.2093,
      "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": 5.273845103591424e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}