trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.995e-05,
      "loss": 10.5576,
      "step": 1
    },
    {
      "epoch": 1.0,
      "learning_rate": 2.5e-05,
      "loss": 0.356,
      "step": 500
    },
    {
      "epoch": 1.0,
      "eval_cer": 0.041,
      "eval_loss": 0.1933690458536148,
      "eval_runtime": 295.9636,
      "eval_samples_per_second": 13.515,
      "eval_steps_per_second": 0.422,
      "step": 500
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0,
      "loss": 0.1558,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "eval_cer": 0.034291666666666665,
      "eval_loss": 0.1479392945766449,
      "eval_runtime": 296.9645,
      "eval_samples_per_second": 13.47,
      "eval_steps_per_second": 0.421,
      "step": 1000
    },
    {
      "epoch": 2.0,
      "step": 1000,
      "total_flos": 2.3945126111870976e+19,
      "train_loss": 0.26610363578796387,
      "train_runtime": 1667.9703,
      "train_samples_per_second": 19.185,
      "train_steps_per_second": 0.6
    }
  ],
  "logging_steps": 500,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "total_flos": 2.3945126111870976e+19,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}