checkpoint-200/trainer_state.json

{
  "best_global_step": 200,
  "best_metric": 10.44776119402985,
  "best_model_checkpoint": "./SALAMA_C7/checkpoint-200",
  "epoch": 2.0833333333333335,
  "eval_steps": 200,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.20833333333333334,
      "grad_norm": 11.919928550720215,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.6563,
      "step": 20
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 11.068168640136719,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.5814,
      "step": 40
    },
    {
      "epoch": 0.625,
      "grad_norm": 7.94372034072876,
      "learning_rate": 5.8e-06,
      "loss": 0.5414,
      "step": 60
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 6.410305500030518,
      "learning_rate": 7.800000000000002e-06,
      "loss": 0.4745,
      "step": 80
    },
    {
      "epoch": 1.0416666666666667,
      "grad_norm": 4.956066131591797,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.3529,
      "step": 100
    },
    {
      "epoch": 1.25,
      "grad_norm": 3.679781436920166,
      "learning_rate": 9.042553191489362e-06,
      "loss": 0.2865,
      "step": 120
    },
    {
      "epoch": 1.4583333333333333,
      "grad_norm": 3.3053739070892334,
      "learning_rate": 7.97872340425532e-06,
      "loss": 0.263,
      "step": 140
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 5.734177589416504,
      "learning_rate": 6.914893617021278e-06,
      "loss": 0.2875,
      "step": 160
    },
    {
      "epoch": 1.875,
      "grad_norm": 7.049015522003174,
      "learning_rate": 5.851063829787235e-06,
      "loss": 0.2821,
      "step": 180
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 2.211012363433838,
      "learning_rate": 4.787234042553192e-06,
      "loss": 0.204,
      "step": 200
    },
    {
      "epoch": 2.0833333333333335,
      "eval_loss": 0.18036405742168427,
      "eval_runtime": 274.0379,
      "eval_samples_per_second": 2.799,
      "eval_steps_per_second": 0.35,
      "eval_wer": 10.44776119402985,
      "step": 200
    }
  ],
  "logging_steps": 20,
  "max_steps": 288,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.84463787884544e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}