trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.12799180852425446,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02559836170485089,
      "grad_norm": 0.015476950444281101,
      "learning_rate": 2.9744016382951492e-05,
      "loss": 0.0096,
      "step": 200
    },
    {
      "epoch": 0.05119672340970178,
      "grad_norm": 0.0031652883626520634,
      "learning_rate": 2.948803276590298e-05,
      "loss": 0.007,
      "step": 400
    },
    {
      "epoch": 0.07679508511455267,
      "grad_norm": 0.02116159349679947,
      "learning_rate": 2.9232049148854475e-05,
      "loss": 0.0063,
      "step": 600
    },
    {
      "epoch": 0.10239344681940356,
      "grad_norm": 0.014196786098182201,
      "learning_rate": 2.8976065531805966e-05,
      "loss": 0.0053,
      "step": 800
    },
    {
      "epoch": 0.12799180852425446,
      "grad_norm": 0.011384401470422745,
      "learning_rate": 2.8720081914757454e-05,
      "loss": 0.0056,
      "step": 1000
    }
  ],
  "logging_steps": 200,
  "max_steps": 23439,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}