trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 19.455252918287936,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.97,
      "learning_rate": 1e-05,
      "loss": 4.9557,
      "step": 500
    },
    {
      "epoch": 1.95,
      "learning_rate": 2e-05,
      "loss": 3.7006,
      "step": 1000
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.888888888888889e-05,
      "loss": 3.289,
      "step": 1500
    },
    {
      "epoch": 3.89,
      "learning_rate": 1.7777777777777777e-05,
      "loss": 3.047,
      "step": 2000
    },
    {
      "epoch": 4.86,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 2.9495,
      "step": 2500
    },
    {
      "epoch": 5.84,
      "learning_rate": 1.555555555555556e-05,
      "loss": 2.876,
      "step": 3000
    },
    {
      "epoch": 6.81,
      "learning_rate": 1.4444444444444446e-05,
      "loss": 2.809,
      "step": 3500
    },
    {
      "epoch": 7.78,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 2.7617,
      "step": 4000
    },
    {
      "epoch": 8.75,
      "learning_rate": 1.2222222222222224e-05,
      "loss": 2.7072,
      "step": 4500
    },
    {
      "epoch": 9.73,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 2.6693,
      "step": 5000
    },
    {
      "epoch": 10.7,
      "learning_rate": 1e-05,
      "loss": 2.6301,
      "step": 5500
    },
    {
      "epoch": 11.67,
      "learning_rate": 8.888888888888888e-06,
      "loss": 2.5909,
      "step": 6000
    },
    {
      "epoch": 12.65,
      "learning_rate": 7.77777777777778e-06,
      "loss": 2.5704,
      "step": 6500
    },
    {
      "epoch": 13.62,
      "learning_rate": 6.666666666666667e-06,
      "loss": 2.5511,
      "step": 7000
    },
    {
      "epoch": 14.59,
      "learning_rate": 5.555555555555557e-06,
      "loss": 2.5306,
      "step": 7500
    },
    {
      "epoch": 15.56,
      "learning_rate": 4.444444444444444e-06,
      "loss": 2.5132,
      "step": 8000
    },
    {
      "epoch": 16.54,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 2.5107,
      "step": 8500
    },
    {
      "epoch": 17.51,
      "learning_rate": 2.222222222222222e-06,
      "loss": 2.4966,
      "step": 9000
    },
    {
      "epoch": 18.48,
      "learning_rate": 1.111111111111111e-06,
      "loss": 2.4931,
      "step": 9500
    },
    {
      "epoch": 19.46,
      "learning_rate": 0.0,
      "loss": 2.4808,
      "step": 10000
    }
  ],
  "max_steps": 10000,
  "num_train_epochs": 20,
  "total_flos": 2.3067637826401075e+17,
  "trial_name": null,
  "trial_params": null
}