{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "global_step": 1865,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.997319034852547e-05,
      "loss": 13.43,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_cer": 0.12903225806451613,
      "eval_loss": 0.6972762942314148,
      "eval_runtime": 86.7359,
      "eval_samples_per_second": 3.459,
      "eval_steps_per_second": 0.438,
      "step": 373
    },
    {
      "epoch": 1.34,
      "learning_rate": 3.659517426273459e-05,
      "loss": 0.6695,
      "step": 500
    },
    {
      "epoch": 2.0,
      "eval_cer": 0.10815939278937381,
      "eval_loss": 0.5941510200500488,
      "eval_runtime": 88.5207,
      "eval_samples_per_second": 3.389,
      "eval_steps_per_second": 0.429,
      "step": 746
    },
    {
      "epoch": 2.68,
      "learning_rate": 2.319034852546917e-05,
      "loss": 0.2885,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_cer": 0.09297912713472485,
      "eval_loss": 0.5037967562675476,
      "eval_runtime": 87.226,
      "eval_samples_per_second": 3.439,
      "eval_steps_per_second": 0.436,
      "step": 1119
    },
    {
      "epoch": 4.0,
      "eval_cer": 0.0635673624288425,
      "eval_loss": 0.4981137216091156,
      "eval_runtime": 87.4644,
      "eval_samples_per_second": 3.43,
      "eval_steps_per_second": 0.434,
      "step": 1492
    },
    {
      "epoch": 4.02,
      "learning_rate": 9.785522788203753e-06,
      "loss": 0.0946,
      "step": 1500
    },
    {
      "epoch": 5.0,
      "eval_cer": 0.058823529411764705,
      "eval_loss": 0.5029511451721191,
      "eval_runtime": 87.3617,
      "eval_samples_per_second": 3.434,
      "eval_steps_per_second": 0.435,
      "step": 1865
    },
    {
      "epoch": 5.0,
      "step": 1865,
      "total_flos": 1.1156932197749883e+19,
      "train_loss": 0.2932615622758226,
      "train_runtime": 4411.5894,
      "train_samples_per_second": 3.38,
      "train_steps_per_second": 0.423
    }
  ],
  "max_steps": 1865,
  "num_train_epochs": 5,
  "total_flos": 1.1156932197749883e+19,
  "trial_name": null,
  "trial_params": null
}