trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2924831822170225,
  "eval_steps": 100,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.06,
      "grad_norm": 0.515340268611908,
      "learning_rate": 8.911111111111111e-05,
      "loss": 1.8825,
      "step": 100
    },
    {
      "epoch": 0.06,
      "eval_loss": 1.3453257083892822,
      "eval_runtime": 1951.9862,
      "eval_samples_per_second": 10.508,
      "eval_steps_per_second": 0.219,
      "step": 100
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.49174025654792786,
      "learning_rate": 6.688888888888889e-05,
      "loss": 1.2985,
      "step": 200
    },
    {
      "epoch": 0.12,
      "eval_loss": 1.265929102897644,
      "eval_runtime": 1951.6138,
      "eval_samples_per_second": 10.51,
      "eval_steps_per_second": 0.219,
      "step": 200
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5104777812957764,
      "learning_rate": 4.466666666666667e-05,
      "loss": 1.2481,
      "step": 300
    },
    {
      "epoch": 0.18,
      "eval_loss": 1.2371329069137573,
      "eval_runtime": 1951.7352,
      "eval_samples_per_second": 10.509,
      "eval_steps_per_second": 0.219,
      "step": 300
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.5633695125579834,
      "learning_rate": 2.2444444444444447e-05,
      "loss": 1.2295,
      "step": 400
    },
    {
      "epoch": 0.23,
      "eval_loss": 1.2196089029312134,
      "eval_runtime": 1951.3878,
      "eval_samples_per_second": 10.511,
      "eval_steps_per_second": 0.219,
      "step": 400
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.5293011665344238,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 1.2129,
      "step": 500
    },
    {
      "epoch": 0.29,
      "eval_loss": 1.2104666233062744,
      "eval_runtime": 1951.0729,
      "eval_samples_per_second": 10.513,
      "eval_steps_per_second": 0.219,
      "step": 500
    }
  ],
  "logging_steps": 100,
  "max_steps": 500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 9.166063140864e+16,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": null
}