checkpoint-20/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 0.0001,
      "loss": 1.4528,
      "step": 1
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0002,
      "loss": 1.4528,
      "step": 2
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.00018888888888888888,
      "loss": 1.4019,
      "step": 3
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.00017777777777777779,
      "loss": 1.3318,
      "step": 4
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.0001666666666666667,
      "loss": 1.2812,
      "step": 5
    },
    {
      "epoch": 6.0,
      "learning_rate": 0.00015555555555555556,
      "loss": 1.2273,
      "step": 6
    },
    {
      "epoch": 7.0,
      "learning_rate": 0.00014444444444444444,
      "loss": 1.1709,
      "step": 7
    },
    {
      "epoch": 8.0,
      "learning_rate": 0.00013333333333333334,
      "loss": 1.1165,
      "step": 8
    },
    {
      "epoch": 9.0,
      "learning_rate": 0.00012222222222222224,
      "loss": 1.062,
      "step": 9
    },
    {
      "epoch": 10.0,
      "learning_rate": 0.00011111111111111112,
      "loss": 1.0071,
      "step": 10
    },
    {
      "epoch": 11.0,
      "learning_rate": 0.0001,
      "loss": 0.9491,
      "step": 11
    },
    {
      "epoch": 12.0,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.8929,
      "step": 12
    },
    {
      "epoch": 13.0,
      "learning_rate": 7.777777777777778e-05,
      "loss": 0.8419,
      "step": 13
    },
    {
      "epoch": 14.0,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.7937,
      "step": 14
    },
    {
      "epoch": 15.0,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.7473,
      "step": 15
    },
    {
      "epoch": 16.0,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.707,
      "step": 16
    },
    {
      "epoch": 17.0,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.6727,
      "step": 17
    },
    {
      "epoch": 18.0,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.6435,
      "step": 18
    },
    {
      "epoch": 19.0,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.6213,
      "step": 19
    },
    {
      "epoch": 20.0,
      "learning_rate": 0.0,
      "loss": 0.6055,
      "step": 20
    }
  ],
  "logging_steps": 1,
  "max_steps": 20,
  "num_train_epochs": 20,
  "save_steps": 500,
  "total_flos": 1749190862438400.0,
  "trial_name": null,
  "trial_params": null
}