{
  "best_metric": 41.511269513669795,
  "best_model_checkpoint": "./whisper-small-all-ka/checkpoint-8000",
  "epoch": 3.9623576027736505,
  "eval_steps": 1000,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.09905894006934125,
      "grad_norm": 2.2366678714752197,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.2185,
      "step": 200
    },
    {
      "epoch": 0.1981178801386825,
      "grad_norm": 2.3154804706573486,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.3894,
      "step": 400
    },
    {
      "epoch": 0.2971768202080238,
      "grad_norm": 2.071636915206909,
      "learning_rate": 6e-06,
      "loss": 0.2868,
      "step": 600
    },
    {
      "epoch": 0.396235760277365,
      "grad_norm": 1.6677467823028564,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.241,
      "step": 800
    },
    {
      "epoch": 0.4952947003467063,
      "grad_norm": 2.1703829765319824,
      "learning_rate": 1e-05,
      "loss": 0.2149,
      "step": 1000
    },
    {
      "epoch": 0.4952947003467063,
      "eval_loss": 0.20653420686721802,
      "eval_runtime": 3478.9242,
      "eval_samples_per_second": 2.805,
      "eval_steps_per_second": 0.351,
      "eval_wer": 54.861029030703634,
      "step": 1000
    },
    {
      "epoch": 0.5943536404160475,
      "grad_norm": 1.6030035018920898,
      "learning_rate": 9.94921279837481e-06,
      "loss": 0.1983,
      "step": 1200
    },
    {
      "epoch": 0.6934125804853888,
      "grad_norm": 2.294896125793457,
      "learning_rate": 9.89842559674962e-06,
      "loss": 0.1817,
      "step": 1400
    },
    {
      "epoch": 0.79247152055473,
      "grad_norm": 1.2499381303787231,
      "learning_rate": 9.84763839512443e-06,
      "loss": 0.1717,
      "step": 1600
    },
    {
      "epoch": 0.8915304606240714,
      "grad_norm": 1.322660207748413,
      "learning_rate": 9.796851193499239e-06,
      "loss": 0.165,
      "step": 1800
    },
    {
      "epoch": 0.9905894006934126,
      "grad_norm": 1.1454402208328247,
      "learning_rate": 9.746063991874048e-06,
      "loss": 0.1589,
      "step": 2000
    },
    {
      "epoch": 0.9905894006934126,
      "eval_loss": 0.16001442074775696,
      "eval_runtime": 3476.6879,
      "eval_samples_per_second": 2.807,
      "eval_steps_per_second": 0.351,
      "eval_wer": 47.45924518352169,
      "step": 2000
    },
    {
      "epoch": 1.0896483407627537,
      "grad_norm": 1.3895001411437988,
      "learning_rate": 9.695276790248857e-06,
      "loss": 0.1459,
      "step": 2200
    },
    {
      "epoch": 1.188707280832095,
      "grad_norm": 1.2915339469909668,
      "learning_rate": 9.644489588623668e-06,
      "loss": 0.1432,
      "step": 2400
    },
    {
      "epoch": 1.2877662209014362,
      "grad_norm": 1.4027740955352783,
      "learning_rate": 9.593702386998477e-06,
      "loss": 0.1396,
      "step": 2600
    },
    {
      "epoch": 1.3868251609707776,
      "grad_norm": 1.3594259023666382,
      "learning_rate": 9.542915185373287e-06,
      "loss": 0.1359,
      "step": 2800
    },
    {
      "epoch": 1.485884101040119,
      "grad_norm": 1.2143176794052124,
      "learning_rate": 9.492127983748096e-06,
      "loss": 0.1332,
      "step": 3000
    },
    {
      "epoch": 1.485884101040119,
      "eval_loss": 0.14520840346813202,
      "eval_runtime": 3482.9046,
      "eval_samples_per_second": 2.802,
      "eval_steps_per_second": 0.35,
      "eval_wer": 45.08623746442761,
      "step": 3000
    },
    {
      "epoch": 1.5849430411094603,
      "grad_norm": 1.00477135181427,
      "learning_rate": 9.441340782122905e-06,
      "loss": 0.133,
      "step": 3200
    },
    {
      "epoch": 1.6840019811788014,
      "grad_norm": 1.0558192729949951,
      "learning_rate": 9.390553580497716e-06,
      "loss": 0.1294,
      "step": 3400
    },
    {
      "epoch": 1.7830609212481425,
      "grad_norm": 1.1081092357635498,
      "learning_rate": 9.339766378872525e-06,
      "loss": 0.1296,
      "step": 3600
    },
    {
      "epoch": 1.8821198613174839,
      "grad_norm": 1.0857079029083252,
      "learning_rate": 9.288979177247334e-06,
      "loss": 0.1262,
      "step": 3800
    },
    {
      "epoch": 1.9811788013868252,
      "grad_norm": 0.9101235866546631,
      "learning_rate": 9.238191975622143e-06,
      "loss": 0.1259,
      "step": 4000
    },
    {
      "epoch": 1.9811788013868252,
      "eval_loss": 0.13690388202667236,
      "eval_runtime": 3481.4796,
      "eval_samples_per_second": 2.803,
      "eval_steps_per_second": 0.35,
      "eval_wer": 43.51605189704238,
      "step": 4000
    },
    {
      "epoch": 2.0802377414561666,
      "grad_norm": 1.0484652519226074,
      "learning_rate": 9.18765871000508e-06,
      "loss": 0.1131,
      "step": 4200
    },
    {
      "epoch": 2.1792966815255075,
      "grad_norm": 1.1338391304016113,
      "learning_rate": 9.13687150837989e-06,
      "loss": 0.1112,
      "step": 4400
    },
    {
      "epoch": 2.278355621594849,
      "grad_norm": 0.9634003043174744,
      "learning_rate": 9.086084306754698e-06,
      "loss": 0.1106,
      "step": 4600
    },
    {
      "epoch": 2.37741456166419,
      "grad_norm": 0.9234340786933899,
      "learning_rate": 9.035297105129508e-06,
      "loss": 0.1094,
      "step": 4800
    },
    {
      "epoch": 2.4764735017335315,
      "grad_norm": 1.1325962543487549,
      "learning_rate": 8.984509903504318e-06,
      "loss": 0.11,
      "step": 5000
    },
    {
      "epoch": 2.4764735017335315,
      "eval_loss": 0.1342398077249527,
      "eval_runtime": 3538.4773,
      "eval_samples_per_second": 2.758,
      "eval_steps_per_second": 0.345,
      "eval_wer": 42.52138889141444,
      "step": 5000
    },
    {
      "epoch": 2.5755324418028724,
      "grad_norm": 0.9091247916221619,
      "learning_rate": 8.933722701879128e-06,
      "loss": 0.1098,
      "step": 5200
    },
    {
      "epoch": 2.6745913818722142,
      "grad_norm": 0.9756883978843689,
      "learning_rate": 8.882935500253937e-06,
      "loss": 0.107,
      "step": 5400
    },
    {
      "epoch": 2.773650321941555,
      "grad_norm": 1.0698374509811401,
      "learning_rate": 8.832148298628746e-06,
      "loss": 0.1076,
      "step": 5600
    },
    {
      "epoch": 2.8727092620108965,
      "grad_norm": 1.0525143146514893,
      "learning_rate": 8.781361097003555e-06,
      "loss": 0.1069,
      "step": 5800
    },
    {
      "epoch": 2.971768202080238,
      "grad_norm": 0.9000458717346191,
      "learning_rate": 8.730573895378366e-06,
      "loss": 0.1071,
      "step": 6000
    },
    {
      "epoch": 2.971768202080238,
      "eval_loss": 0.13021869957447052,
      "eval_runtime": 3466.5665,
      "eval_samples_per_second": 2.815,
      "eval_steps_per_second": 0.352,
      "eval_wer": 42.22862701955686,
      "step": 6000
    },
    {
      "epoch": 3.070827142149579,
      "grad_norm": 0.9356966018676758,
      "learning_rate": 8.679786693753175e-06,
      "loss": 0.0966,
      "step": 6200
    },
    {
      "epoch": 3.16988608221892,
      "grad_norm": 0.8852076530456543,
      "learning_rate": 8.628999492127984e-06,
      "loss": 0.0935,
      "step": 6400
    },
    {
      "epoch": 3.2689450222882614,
      "grad_norm": 0.8325952291488647,
      "learning_rate": 8.578212290502793e-06,
      "loss": 0.0928,
      "step": 6600
    },
    {
      "epoch": 3.368003962357603,
      "grad_norm": 0.9712433218955994,
      "learning_rate": 8.527425088877603e-06,
      "loss": 0.0925,
      "step": 6800
    },
    {
      "epoch": 3.467062902426944,
      "grad_norm": 0.8990651369094849,
      "learning_rate": 8.476637887252413e-06,
      "loss": 0.0932,
      "step": 7000
    },
    {
      "epoch": 3.467062902426944,
      "eval_loss": 0.1307862401008606,
      "eval_runtime": 3471.2099,
      "eval_samples_per_second": 2.811,
      "eval_steps_per_second": 0.351,
      "eval_wer": 41.745842690499785,
      "step": 7000
    },
    {
      "epoch": 3.566121842496285,
      "grad_norm": 0.9829621911048889,
      "learning_rate": 8.425850685627223e-06,
      "loss": 0.0923,
      "step": 7200
    },
    {
      "epoch": 3.6651807825656264,
      "grad_norm": 1.0099509954452515,
      "learning_rate": 8.375063484002032e-06,
      "loss": 0.0926,
      "step": 7400
    },
    {
      "epoch": 3.7642397226349678,
      "grad_norm": 0.8921763896942139,
      "learning_rate": 8.324276282376843e-06,
      "loss": 0.0921,
      "step": 7600
    },
    {
      "epoch": 3.863298662704309,
      "grad_norm": 0.9494955539703369,
      "learning_rate": 8.273489080751652e-06,
      "loss": 0.0925,
      "step": 7800
    },
    {
      "epoch": 3.9623576027736505,
      "grad_norm": 0.8222095966339111,
      "learning_rate": 8.222701879126461e-06,
      "loss": 0.0916,
      "step": 8000
    },
    {
      "epoch": 3.9623576027736505,
      "eval_loss": 0.12850907444953918,
      "eval_runtime": 3467.0941,
      "eval_samples_per_second": 2.815,
      "eval_steps_per_second": 0.352,
      "eval_wer": 41.511269513669795,
      "step": 8000
    }
  ],
  "logging_steps": 200,
  "max_steps": 40380,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.9550365812187136e+20,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}