trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0204081632653061,
  "eval_steps": 30,
  "global_step": 150,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20408163265306123,
      "grad_norm": 19.92782211303711,
      "learning_rate": 1e-05,
      "loss": 10.8562,
      "step": 30
    },
    {
      "epoch": 0.20408163265306123,
      "eval_cer": 24.936057374728698,
      "eval_loss": 10.856045722961426,
      "eval_runtime": 524.0657,
      "eval_samples_per_second": 2.168,
      "eval_steps_per_second": 0.135,
      "eval_wer": 17.0624761420028,
      "step": 30
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 18.936391830444336,
      "learning_rate": 7.500000000000001e-06,
      "loss": 10.8557,
      "step": 60
    },
    {
      "epoch": 0.40816326530612246,
      "eval_cer": 28.020024535245824,
      "eval_loss": 10.855280876159668,
      "eval_runtime": 546.8477,
      "eval_samples_per_second": 2.077,
      "eval_steps_per_second": 0.13,
      "eval_wer": 20.38325486703143,
      "step": 60
    },
    {
      "epoch": 0.6122448979591837,
      "grad_norm": 0.06280206888914108,
      "learning_rate": 5e-06,
      "loss": 10.8549,
      "step": 90
    },
    {
      "epoch": 0.6122448979591837,
      "eval_cer": 17.241804284231385,
      "eval_loss": 10.854511260986328,
      "eval_runtime": 487.9237,
      "eval_samples_per_second": 2.328,
      "eval_steps_per_second": 0.146,
      "eval_wer": 11.682275098613054,
      "step": 90
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 18.20659828186035,
      "learning_rate": 2.5e-06,
      "loss": 10.8542,
      "step": 120
    },
    {
      "epoch": 0.8163265306122449,
      "eval_cer": 15.890988015476077,
      "eval_loss": 10.853907585144043,
      "eval_runtime": 553.861,
      "eval_samples_per_second": 2.051,
      "eval_steps_per_second": 0.128,
      "eval_wer": 10.510752004071765,
      "step": 120
    },
    {
      "epoch": 1.0204081632653061,
      "grad_norm": 17.312578201293945,
      "learning_rate": 0.0,
      "loss": 10.8538,
      "step": 150
    },
    {
      "epoch": 1.0204081632653061,
      "eval_cer": 9.633726526375389,
      "eval_loss": 10.85368824005127,
      "eval_runtime": 444.8852,
      "eval_samples_per_second": 2.553,
      "eval_steps_per_second": 0.16,
      "eval_wer": 6.432243287950121,
      "step": 150
    },
    {
      "epoch": 1.0204081632653061,
      "step": 150,
      "total_flos": 2.3614434607104e+17,
      "train_loss": 10.854962158203126,
      "train_runtime": 4570.606,
      "train_samples_per_second": 2.1,
      "train_steps_per_second": 0.033
    }
  ],
  "logging_steps": 30,
  "max_steps": 150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 30,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.3614434607104e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}