trainer_state.json

{
  "best_metric": 102.19780219780219,
  "best_model_checkpoint": "./quran-whisper-ar-tiny-1/checkpoint-15",
  "epoch": 1.0,
  "eval_steps": 5,
  "global_step": 45,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3333333333333333,
      "grad_norm": Infinity,
      "learning_rate": 9.333333333333334e-05,
      "loss": 2.2643,
      "step": 5
    },
    {
      "epoch": 0.3333333333333333,
      "eval_loss": 4.266593933105469,
      "eval_runtime": 19.8469,
      "eval_samples_per_second": 0.504,
      "eval_steps_per_second": 0.101,
      "eval_wer": 209.8901098901099,
      "step": 5
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 53.0462760925293,
      "learning_rate": 7.333333333333333e-05,
      "loss": 3.5798,
      "step": 10
    },
    {
      "epoch": 0.6666666666666666,
      "eval_loss": 1.9317986965179443,
      "eval_runtime": 17.1404,
      "eval_samples_per_second": 0.583,
      "eval_steps_per_second": 0.117,
      "eval_wer": 173.62637362637363,
      "step": 10
    },
    {
      "epoch": 1.0,
      "grad_norm": 15.332279205322266,
      "learning_rate": 4e-05,
      "loss": 1.2993,
      "step": 15
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.0968444347381592,
      "eval_runtime": 14.7686,
      "eval_samples_per_second": 0.677,
      "eval_steps_per_second": 0.135,
      "eval_wer": 102.19780219780219,
      "step": 15
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 14.598529815673828,
      "learning_rate": 6e-05,
      "loss": 0.8612,
      "step": 20
    },
    {
      "epoch": 0.6666666666666666,
      "eval_loss": 0.9798414707183838,
      "eval_runtime": 14.0253,
      "eval_samples_per_second": 0.713,
      "eval_steps_per_second": 0.143,
      "eval_wer": 145.05494505494505,
      "step": 20
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 10.819196701049805,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.7235,
      "step": 25
    },
    {
      "epoch": 0.8333333333333334,
      "eval_loss": 0.7983613610267639,
      "eval_runtime": 13.4734,
      "eval_samples_per_second": 0.742,
      "eval_steps_per_second": 0.148,
      "eval_wer": 131.86813186813185,
      "step": 25
    },
    {
      "epoch": 1.0,
      "grad_norm": 12.039382934570312,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.5694,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.7010310888290405,
      "eval_runtime": 11.6848,
      "eval_samples_per_second": 0.856,
      "eval_steps_per_second": 0.171,
      "eval_wer": 127.47252747252746,
      "step": 30
    },
    {
      "epoch": 0.7777777777777778,
      "grad_norm": 10.747776985168457,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.5781,
      "step": 35
    },
    {
      "epoch": 0.7777777777777778,
      "eval_loss": 0.7016527652740479,
      "eval_runtime": 13.6665,
      "eval_samples_per_second": 0.732,
      "eval_steps_per_second": 0.146,
      "eval_wer": 378.021978021978,
      "step": 35
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 11.89635944366455,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.5704,
      "step": 40
    },
    {
      "epoch": 0.8888888888888888,
      "eval_loss": 0.6510931253433228,
      "eval_runtime": 13.0034,
      "eval_samples_per_second": 0.769,
      "eval_steps_per_second": 0.154,
      "eval_wer": 341.75824175824175,
      "step": 40
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.039495468139648,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.4247,
      "step": 45
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5858025550842285,
      "eval_runtime": 13.6045,
      "eval_samples_per_second": 0.735,
      "eval_steps_per_second": 0.147,
      "eval_wer": 164.83516483516482,
      "step": 45
    },
    {
      "epoch": 1.0,
      "step": 45,
      "total_flos": 1.77255972864e+16,
      "train_loss": 0.17480638292100695,
      "train_runtime": 95.2131,
      "train_samples_per_second": 7.562,
      "train_steps_per_second": 0.473
    }
  ],
  "logging_steps": 5,
  "max_steps": 45,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 5,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.77255972864e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}