trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.4188034188034186,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17,
      "learning_rate": 1e-05,
      "loss": 4.4095,
      "step": 500
    },
    {
      "epoch": 0.34,
      "learning_rate": 2e-05,
      "loss": 3.5202,
      "step": 1000
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.888888888888889e-05,
      "loss": 3.2715,
      "step": 1500
    },
    {
      "epoch": 0.68,
      "learning_rate": 1.7777777777777777e-05,
      "loss": 3.0872,
      "step": 2000
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 2.9574,
      "step": 2500
    },
    {
      "epoch": 1.03,
      "learning_rate": 1.555555555555556e-05,
      "loss": 2.8833,
      "step": 3000
    },
    {
      "epoch": 1.2,
      "learning_rate": 1.4444444444444446e-05,
      "loss": 2.8135,
      "step": 3500
    },
    {
      "epoch": 1.37,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.7587,
      "step": 4000
    },
    {
      "epoch": 1.54,
      "learning_rate": 1.2222222222222224e-05,
      "loss": 2.7349,
      "step": 4500
    },
    {
      "epoch": 1.71,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 2.7262,
      "step": 5000
    },
    {
      "epoch": 1.88,
      "learning_rate": 1e-05,
      "loss": 2.6846,
      "step": 5500
    },
    {
      "epoch": 2.05,
      "learning_rate": 8.888888888888888e-06,
      "loss": 2.658,
      "step": 6000
    },
    {
      "epoch": 2.22,
      "learning_rate": 7.77777777777778e-06,
      "loss": 2.6489,
      "step": 6500
    },
    {
      "epoch": 2.39,
      "learning_rate": 6.666666666666667e-06,
      "loss": 2.6384,
      "step": 7000
    },
    {
      "epoch": 2.56,
      "learning_rate": 5.555555555555557e-06,
      "loss": 2.6136,
      "step": 7500
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.444444444444444e-06,
      "loss": 2.6143,
      "step": 8000
    },
    {
      "epoch": 2.91,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 2.5972,
      "step": 8500
    },
    {
      "epoch": 3.08,
      "learning_rate": 2.222222222222222e-06,
      "loss": 2.5893,
      "step": 9000
    },
    {
      "epoch": 3.25,
      "learning_rate": 1.111111111111111e-06,
      "loss": 2.5803,
      "step": 9500
    },
    {
      "epoch": 3.42,
      "learning_rate": 0.0,
      "loss": 2.5825,
      "step": 10000
    }
  ],
  "max_steps": 10000,
  "num_train_epochs": 4,
  "total_flos": 2.310345220643881e+17,
  "trial_name": null,
  "trial_params": null
}