trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9928057553956835,
  "eval_steps": 25,
  "global_step": 117,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.64,
      "learning_rate": 9.294316336102132e-05,
      "loss": 0.2634,
      "step": 25
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.29237279295921326,
      "eval_runtime": 196.4821,
      "eval_samples_per_second": 50.9,
      "eval_steps_per_second": 0.799,
      "step": 25
    },
    {
      "epoch": 1.28,
      "learning_rate": 6.598819622856227e-05,
      "loss": 0.1216,
      "step": 50
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.08977410197257996,
      "eval_runtime": 196.3182,
      "eval_samples_per_second": 50.943,
      "eval_steps_per_second": 0.8,
      "step": 50
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.135717611098458e-05,
      "loss": 0.0833,
      "step": 75
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.07183241099119186,
      "eval_runtime": 196.5039,
      "eval_samples_per_second": 50.895,
      "eval_steps_per_second": 0.799,
      "step": 75
    },
    {
      "epoch": 2.56,
      "learning_rate": 5.676715638695063e-06,
      "loss": 0.062,
      "step": 100
    },
    {
      "epoch": 2.56,
      "eval_loss": 0.06442917138338089,
      "eval_runtime": 196.5023,
      "eval_samples_per_second": 50.895,
      "eval_steps_per_second": 0.799,
      "step": 100
    },
    {
      "epoch": 2.99,
      "step": 117,
      "total_flos": 783708051210240.0,
      "train_loss": 0.12114470178245479,
      "train_runtime": 18588.429,
      "train_samples_per_second": 12.912,
      "train_steps_per_second": 0.006
    }
  ],
  "logging_steps": 25,
  "max_steps": 117,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "total_flos": 783708051210240.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}