trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.7272727272727275,
  "eval_steps": 500,
  "global_step": 15,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.18,
      "grad_norm": 7.667912668669107,
      "learning_rate": 1e-05,
      "loss": 1.1107,
      "step": 1
    },
    {
      "epoch": 0.91,
      "grad_norm": 34.0595490556449,
      "learning_rate": 1.7485107481711014e-05,
      "loss": 1.3819,
      "step": 5
    },
    {
      "epoch": 0.91,
      "eval_loss": 1.1824232339859009,
      "eval_runtime": 34.7385,
      "eval_samples_per_second": 11.975,
      "eval_steps_per_second": 0.748,
      "step": 5
    },
    {
      "epoch": 1.82,
      "grad_norm": 5.121285979288454,
      "learning_rate": 6.453951129574644e-06,
      "loss": 1.0491,
      "step": 10
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.9082208871841431,
      "eval_runtime": 34.5513,
      "eval_samples_per_second": 12.04,
      "eval_steps_per_second": 0.753,
      "step": 11
    },
    {
      "epoch": 2.73,
      "grad_norm": 3.7805401816722046,
      "learning_rate": 0.0,
      "loss": 0.8189,
      "step": 15
    },
    {
      "epoch": 2.73,
      "eval_loss": 0.8698543906211853,
      "eval_runtime": 34.5688,
      "eval_samples_per_second": 12.034,
      "eval_steps_per_second": 0.752,
      "step": 15
    },
    {
      "epoch": 2.73,
      "step": 15,
      "total_flos": 3088349921280.0,
      "train_loss": 1.0652088483174642,
      "train_runtime": 362.8946,
      "train_samples_per_second": 2.844,
      "train_steps_per_second": 0.041
    }
  ],
  "logging_steps": 5,
  "max_steps": 15,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 3088349921280.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}