{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 19.91614255765199,
  "eval_steps": 500,
  "global_step": 9500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0482180293501049,
      "grad_norm": 0.002079850761219859,
      "learning_rate": 4.99e-05,
      "loss": 0.0303,
      "step": 500
    },
    {
      "epoch": 1.0482180293501049,
      "eval_cer": 0.025307563615524685,
      "eval_loss": 0.021985899657011032,
      "eval_runtime": 51.7047,
      "eval_samples_per_second": 44.213,
      "eval_steps_per_second": 5.531,
      "eval_wer": 0.0908849221420433,
      "step": 500
    },
    {
      "epoch": 2.0964360587002098,
      "grad_norm": 0.5530248880386353,
      "learning_rate": 4.724004424778761e-05,
      "loss": 0.0594,
      "step": 1000
    },
    {
      "epoch": 2.0964360587002098,
      "eval_cer": 0.025889877898862178,
      "eval_loss": 0.020988421514630318,
      "eval_runtime": 51.2495,
      "eval_samples_per_second": 44.605,
      "eval_steps_per_second": 5.581,
      "eval_wer": 0.09160653247246486,
      "step": 1000
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 1.1215722560882568,
      "learning_rate": 4.448008849557522e-05,
      "loss": 0.0767,
      "step": 1500
    },
    {
      "epoch": 3.1446540880503147,
      "eval_cer": 0.025104215770549686,
      "eval_loss": 0.014869533479213715,
      "eval_runtime": 51.4984,
      "eval_samples_per_second": 44.39,
      "eval_steps_per_second": 5.554,
      "eval_wer": 0.09115077857956703,
      "step": 1500
    },
    {
      "epoch": 4.1928721174004195,
      "grad_norm": 4.590636253356934,
      "learning_rate": 4.172566371681416e-05,
      "loss": 0.062,
      "step": 2000
    },
    {
      "epoch": 4.1928721174004195,
      "eval_cer": 0.02556636996367468,
      "eval_loss": 0.017348747700452805,
      "eval_runtime": 51.1673,
      "eval_samples_per_second": 44.677,
      "eval_steps_per_second": 5.59,
      "eval_wer": 0.0916824914546145,
      "step": 2000
    },
    {
      "epoch": 5.241090146750524,
      "grad_norm": 1.3328059911727905,
      "learning_rate": 3.896017699115044e-05,
      "loss": 0.0687,
      "step": 2500
    },
    {
      "epoch": 5.241090146750524,
      "eval_cer": 0.025741988557062178,
      "eval_loss": 0.018996117636561394,
      "eval_runtime": 51.3668,
      "eval_samples_per_second": 44.503,
      "eval_steps_per_second": 5.568,
      "eval_wer": 0.09183440941891378,
      "step": 2500
    },
    {
      "epoch": 6.289308176100629,
      "grad_norm": 0.1056140810251236,
      "learning_rate": 3.6205752212389385e-05,
      "loss": 0.0617,
      "step": 3000
    },
    {
      "epoch": 6.289308176100629,
      "eval_cer": 0.025242862028487186,
      "eval_loss": 0.01558039989322424,
      "eval_runtime": 51.1418,
      "eval_samples_per_second": 44.699,
      "eval_steps_per_second": 5.592,
      "eval_wer": 0.09149259399924041,
      "step": 3000
    },
    {
      "epoch": 7.337526205450734,
      "grad_norm": 0.9397304654121399,
      "learning_rate": 3.3445796460176995e-05,
      "loss": 0.0673,
      "step": 3500
    },
    {
      "epoch": 7.337526205450734,
      "eval_cer": 0.02564031463457468,
      "eval_loss": 0.018212666735053062,
      "eval_runtime": 51.318,
      "eval_samples_per_second": 44.546,
      "eval_steps_per_second": 5.573,
      "eval_wer": 0.09141663501709077,
      "step": 3500
    },
    {
      "epoch": 8.385744234800839,
      "grad_norm": 2.7481186389923096,
      "learning_rate": 3.0680309734513276e-05,
      "loss": 0.0645,
      "step": 4000
    },
    {
      "epoch": 8.385744234800839,
      "eval_cer": 0.025215132776899685,
      "eval_loss": 0.014686434529721737,
      "eval_runtime": 51.1418,
      "eval_samples_per_second": 44.699,
      "eval_steps_per_second": 5.592,
      "eval_wer": 0.0916824914546145,
      "step": 4000
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 1.0427154302597046,
      "learning_rate": 2.7914823008849562e-05,
      "loss": 0.0633,
      "step": 4500
    },
    {
      "epoch": 9.433962264150944,
      "eval_cer": 0.025427723705737183,
      "eval_loss": 0.016345607116818428,
      "eval_runtime": 51.3562,
      "eval_samples_per_second": 44.513,
      "eval_steps_per_second": 5.569,
      "eval_wer": 0.09130269654386632,
      "step": 4500
    },
    {
      "epoch": 10.482180293501049,
      "grad_norm": 0.39028722047805786,
      "learning_rate": 2.515486725663717e-05,
      "loss": 0.0507,
      "step": 5000
    },
    {
      "epoch": 10.482180293501049,
      "eval_cer": 0.02552015454436218,
      "eval_loss": 0.01751078851521015,
      "eval_runtime": 51.3306,
      "eval_samples_per_second": 44.535,
      "eval_steps_per_second": 5.572,
      "eval_wer": 0.09160653247246486,
      "step": 5000
    },
    {
      "epoch": 11.530398322851154,
      "grad_norm": 0.7193567752838135,
      "learning_rate": 2.2389380530973453e-05,
      "loss": 0.064,
      "step": 5500
    },
    {
      "epoch": 11.530398322851154,
      "eval_cer": 0.025215132776899685,
      "eval_loss": 0.014882609248161316,
      "eval_runtime": 51.3252,
      "eval_samples_per_second": 44.54,
      "eval_steps_per_second": 5.572,
      "eval_wer": 0.09118875807064185,
      "step": 5500
    },
    {
      "epoch": 12.578616352201259,
      "grad_norm": 0.4278053641319275,
      "learning_rate": 1.964048672566372e-05,
      "loss": 0.0508,
      "step": 6000
    },
    {
      "epoch": 12.578616352201259,
      "eval_cer": 0.025150431189862186,
      "eval_loss": 0.015342830680310726,
      "eval_runtime": 51.4039,
      "eval_samples_per_second": 44.471,
      "eval_steps_per_second": 5.564,
      "eval_wer": 0.09092290163311811,
      "step": 6000
    },
    {
      "epoch": 13.626834381551364,
      "grad_norm": 1.9797695875167847,
      "learning_rate": 1.6875000000000004e-05,
      "loss": 0.0486,
      "step": 6500
    },
    {
      "epoch": 13.626834381551364,
      "eval_cer": 0.025178160441449684,
      "eval_loss": 0.016490615904331207,
      "eval_runtime": 51.3445,
      "eval_samples_per_second": 44.523,
      "eval_steps_per_second": 5.57,
      "eval_wer": 0.0910748195974174,
      "step": 6500
    },
    {
      "epoch": 14.675052410901468,
      "grad_norm": 0.6631584763526917,
      "learning_rate": 1.4115044247787612e-05,
      "loss": 0.0435,
      "step": 7000
    },
    {
      "epoch": 14.675052410901468,
      "eval_cer": 0.025058000351237186,
      "eval_loss": 0.01397643517702818,
      "eval_runtime": 51.3713,
      "eval_samples_per_second": 44.5,
      "eval_steps_per_second": 5.567,
      "eval_wer": 0.09080896315989366,
      "step": 7000
    },
    {
      "epoch": 15.723270440251572,
      "grad_norm": 1.2671451568603516,
      "learning_rate": 1.1355088495575223e-05,
      "loss": 0.0457,
      "step": 7500
    },
    {
      "epoch": 15.723270440251572,
      "eval_cer": 0.025048757267374686,
      "eval_loss": 0.014803554862737656,
      "eval_runtime": 51.4523,
      "eval_samples_per_second": 44.43,
      "eval_steps_per_second": 5.559,
      "eval_wer": 0.09061906570451957,
      "step": 7500
    },
    {
      "epoch": 16.771488469601678,
      "grad_norm": 0.5452253818511963,
      "learning_rate": 8.606194690265488e-06,
      "loss": 0.0417,
      "step": 8000
    },
    {
      "epoch": 16.771488469601678,
      "eval_cer": 0.02506724343509969,
      "eval_loss": 0.0149121955037117,
      "eval_runtime": 51.4559,
      "eval_samples_per_second": 44.426,
      "eval_steps_per_second": 5.558,
      "eval_wer": 0.09092290163311811,
      "step": 8000
    },
    {
      "epoch": 17.81970649895178,
      "grad_norm": 0.47779375314712524,
      "learning_rate": 5.8517699115044245e-06,
      "loss": 0.047,
      "step": 8500
    },
    {
      "epoch": 17.81970649895178,
      "eval_cer": 0.025002541848062186,
      "eval_loss": 0.014238793402910233,
      "eval_runtime": 51.42,
      "eval_samples_per_second": 44.457,
      "eval_steps_per_second": 5.562,
      "eval_wer": 0.09069502468666919,
      "step": 8500
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 0.3653850555419922,
      "learning_rate": 3.086283185840708e-06,
      "loss": 0.0521,
      "step": 9000
    },
    {
      "epoch": 18.867924528301888,
      "eval_cer": 0.02491935409329969,
      "eval_loss": 0.013357444666326046,
      "eval_runtime": 51.3616,
      "eval_samples_per_second": 44.508,
      "eval_steps_per_second": 5.568,
      "eval_wer": 0.09073300417774402,
      "step": 9000
    },
    {
      "epoch": 19.91614255765199,
      "grad_norm": 1.0006723403930664,
      "learning_rate": 3.263274336283186e-07,
      "loss": 0.0479,
      "step": 9500
    },
    {
      "epoch": 19.91614255765199,
      "eval_cer": 0.024974812596474688,
      "eval_loss": 0.013586528599262238,
      "eval_runtime": 51.4469,
      "eval_samples_per_second": 44.434,
      "eval_steps_per_second": 5.559,
      "eval_wer": 0.09073300417774402,
      "step": 9500
    }
  ],
  "logging_steps": 500,
  "max_steps": 9540,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 6.082771650621779e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}