trainer_state.json

{
  "best_metric": 0.4834844172000885,
  "best_model_checkpoint": "whisper-ling-asr/checkpoint-2000",
  "epoch": 14.925373134328359,
  "eval_steps": 1000,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 7.594086647033691,
      "learning_rate": 9.625941422594142e-06,
      "loss": 0.8297,
      "step": 500
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 8.005925178527832,
      "learning_rate": 9.207531380753139e-06,
      "loss": 0.3703,
      "step": 1000
    },
    {
      "epoch": 2.9850746268656714,
      "eval_loss": 0.5100315809249878,
      "eval_runtime": 410.9956,
      "eval_samples_per_second": 2.674,
      "eval_steps_per_second": 0.168,
      "eval_wer": 0.27345685661962726,
      "step": 1000
    },
    {
      "epoch": 4.477611940298507,
      "grad_norm": 4.732856273651123,
      "learning_rate": 8.789121338912134e-06,
      "loss": 0.2182,
      "step": 1500
    },
    {
      "epoch": 5.970149253731344,
      "grad_norm": 4.545703411102295,
      "learning_rate": 8.37071129707113e-06,
      "loss": 0.1487,
      "step": 2000
    },
    {
      "epoch": 5.970149253731344,
      "eval_loss": 0.4834844172000885,
      "eval_runtime": 462.8067,
      "eval_samples_per_second": 2.375,
      "eval_steps_per_second": 0.149,
      "eval_wer": 0.2979101583539804,
      "step": 2000
    },
    {
      "epoch": 7.462686567164179,
      "grad_norm": 3.4771857261657715,
      "learning_rate": 7.953138075313808e-06,
      "loss": 0.0874,
      "step": 2500
    },
    {
      "epoch": 8.955223880597014,
      "grad_norm": 2.7059333324432373,
      "learning_rate": 7.534728033472803e-06,
      "loss": 0.0556,
      "step": 3000
    },
    {
      "epoch": 8.955223880597014,
      "eval_loss": 0.5049698948860168,
      "eval_runtime": 404.4404,
      "eval_samples_per_second": 2.717,
      "eval_steps_per_second": 0.171,
      "eval_wer": 0.254389744694603,
      "step": 3000
    },
    {
      "epoch": 10.447761194029852,
      "grad_norm": 2.1922836303710938,
      "learning_rate": 7.116317991631799e-06,
      "loss": 0.0295,
      "step": 3500
    },
    {
      "epoch": 11.940298507462687,
      "grad_norm": 0.999194324016571,
      "learning_rate": 6.697907949790795e-06,
      "loss": 0.0171,
      "step": 4000
    },
    {
      "epoch": 11.940298507462687,
      "eval_loss": 0.5400618314743042,
      "eval_runtime": 374.2093,
      "eval_samples_per_second": 2.937,
      "eval_steps_per_second": 0.184,
      "eval_wer": 0.23704621350856403,
      "step": 4000
    },
    {
      "epoch": 13.432835820895523,
      "grad_norm": 1.5301833152770996,
      "learning_rate": 6.279497907949791e-06,
      "loss": 0.0084,
      "step": 4500
    },
    {
      "epoch": 14.925373134328359,
      "grad_norm": 0.46232014894485474,
      "learning_rate": 5.8610878661087865e-06,
      "loss": 0.0051,
      "step": 5000
    },
    {
      "epoch": 14.925373134328359,
      "eval_loss": 0.5673447251319885,
      "eval_runtime": 355.0078,
      "eval_samples_per_second": 3.096,
      "eval_steps_per_second": 0.194,
      "eval_wer": 0.21765593019498006,
      "step": 5000
    },
    {
      "epoch": 14.925373134328359,
      "step": 5000,
      "total_flos": 1.96882148081664e+18,
      "train_loss": 0.1769935465335846,
      "train_runtime": 17207.4544,
      "train_samples_per_second": 11.158,
      "train_steps_per_second": 0.697
    }
  ],
  "logging_steps": 500,
  "max_steps": 12000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 36,
  "save_steps": 1000,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.96882148081664e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}