checkpoint-130/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.9523809523809526,
  "eval_steps": 500,
  "global_step": 130,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.8,
      "learning_rate": 0.00018632478632478634,
      "loss": 0.9096,
      "step": 21
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00015042735042735043,
      "loss": 0.4818,
      "step": 42
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.00011452991452991453,
      "loss": 0.4032,
      "step": 63
    },
    {
      "epoch": 3.2,
      "learning_rate": 7.863247863247864e-05,
      "loss": 0.3568,
      "step": 84
    },
    {
      "epoch": 4.0,
      "learning_rate": 4.2735042735042735e-05,
      "loss": 0.3187,
      "step": 105
    },
    {
      "epoch": 4.8,
      "learning_rate": 6.837606837606839e-06,
      "loss": 0.2903,
      "step": 126
    }
  ],
  "logging_steps": 21,
  "max_steps": 130,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 2.11364836540416e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}