trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.84081041968162,
  "eval_steps": 500,
  "global_step": 1700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 2.89,
      "grad_norm": 1.1808712482452393,
      "learning_rate": 7.4663402692778466e-06,
      "loss": 1.1073,
      "step": 500
    },
    {
      "epoch": 2.89,
      "eval_loss": 0.5578908324241638,
      "eval_runtime": 27.2962,
      "eval_samples_per_second": 34.693,
      "eval_steps_per_second": 2.198,
      "step": 500
    },
    {
      "epoch": 5.79,
      "grad_norm": 1.1339820623397827,
      "learning_rate": 4.406364749082008e-06,
      "loss": 0.4698,
      "step": 1000
    },
    {
      "epoch": 5.79,
      "eval_loss": 0.5357696413993835,
      "eval_runtime": 27.2991,
      "eval_samples_per_second": 34.69,
      "eval_steps_per_second": 2.198,
      "step": 1000
    },
    {
      "epoch": 8.68,
      "grad_norm": 1.1054918766021729,
      "learning_rate": 1.346389228886169e-06,
      "loss": 0.4329,
      "step": 1500
    },
    {
      "epoch": 8.68,
      "eval_loss": 0.5248792767524719,
      "eval_runtime": 27.3307,
      "eval_samples_per_second": 34.65,
      "eval_steps_per_second": 2.195,
      "step": 1500
    }
  ],
  "logging_steps": 500,
  "max_steps": 1720,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 50,
  "total_flos": 3.316520229588173e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}