{
  "best_metric": 71.03448275862068,
  "best_model_checkpoint": "./whisper-large-v3-cv17-dv/checkpoint-3900",
  "epoch": 11.863192182410424,
  "eval_steps": 300,
  "global_step": 3900,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3257328990228013,
      "grad_norm": 38.63404846191406,
      "learning_rate": 1.9200000000000003e-06,
      "loss": 4.5676,
      "step": 100
    },
    {
      "epoch": 0.6514657980456026,
      "grad_norm": 37.98774719238281,
      "learning_rate": 3.920000000000001e-06,
      "loss": 2.8219,
      "step": 200
    },
    {
      "epoch": 0.9771986970684039,
      "grad_norm": 7.7040252685546875,
      "learning_rate": 5.92e-06,
      "loss": 1.4644,
      "step": 300
    },
    {
      "epoch": 0.9771986970684039,
      "eval_loss": 1.0653657913208008,
      "eval_runtime": 142.8846,
      "eval_samples_per_second": 0.7,
      "eval_steps_per_second": 0.021,
      "eval_wer": 203.96551724137933,
      "step": 300
    },
    {
      "epoch": 1.3029315960912053,
      "grad_norm": 3.3996422290802,
      "learning_rate": 7.9e-06,
      "loss": 0.6065,
      "step": 400
    },
    {
      "epoch": 1.6286644951140063,
      "grad_norm": 2.5693020820617676,
      "learning_rate": 9.9e-06,
      "loss": 0.303,
      "step": 500
    },
    {
      "epoch": 1.9543973941368078,
      "grad_norm": 2.019421339035034,
      "learning_rate": 9.72857142857143e-06,
      "loss": 0.2384,
      "step": 600
    },
    {
      "epoch": 1.9543973941368078,
      "eval_loss": 0.3341977596282959,
      "eval_runtime": 138.0417,
      "eval_samples_per_second": 0.724,
      "eval_steps_per_second": 0.022,
      "eval_wer": 84.82758620689656,
      "step": 600
    },
    {
      "epoch": 2.2801302931596092,
      "grad_norm": 1.6545788049697876,
      "learning_rate": 9.442857142857144e-06,
      "loss": 0.1718,
      "step": 700
    },
    {
      "epoch": 2.6058631921824107,
      "grad_norm": 1.4984415769577026,
      "learning_rate": 9.157142857142857e-06,
      "loss": 0.1555,
      "step": 800
    },
    {
      "epoch": 2.9315960912052117,
      "grad_norm": 1.6755363941192627,
      "learning_rate": 8.871428571428571e-06,
      "loss": 0.1481,
      "step": 900
    },
    {
      "epoch": 2.9315960912052117,
      "eval_loss": 0.2715227007865906,
      "eval_runtime": 137.1492,
      "eval_samples_per_second": 0.729,
      "eval_steps_per_second": 0.022,
      "eval_wer": 78.79310344827586,
      "step": 900
    },
    {
      "epoch": 2.3257328990228014,
      "grad_norm": 1.167482852935791,
      "learning_rate": 8.585714285714286e-06,
      "loss": 0.0978,
      "step": 1000
    },
    {
      "epoch": 2.6514657980456025,
      "grad_norm": 1.9179445505142212,
      "learning_rate": 8.3e-06,
      "loss": 0.0876,
      "step": 1100
    },
    {
      "epoch": 2.977198697068404,
      "grad_norm": 1.3239954710006714,
      "learning_rate": 8.014285714285715e-06,
      "loss": 0.0975,
      "step": 1200
    },
    {
      "epoch": 2.977198697068404,
      "eval_loss": 0.2634916603565216,
      "eval_runtime": 138.2093,
      "eval_samples_per_second": 0.724,
      "eval_steps_per_second": 0.022,
      "eval_wer": 76.0344827586207,
      "step": 1200
    },
    {
      "epoch": 3.3029315960912053,
      "grad_norm": 1.5532219409942627,
      "learning_rate": 7.72857142857143e-06,
      "loss": 0.0637,
      "step": 1300
    },
    {
      "epoch": 3.6286644951140063,
      "grad_norm": 1.0479369163513184,
      "learning_rate": 7.442857142857144e-06,
      "loss": 0.0626,
      "step": 1400
    },
    {
      "epoch": 3.954397394136808,
      "grad_norm": 0.9303975105285645,
      "learning_rate": 7.1571428571428584e-06,
      "loss": 0.0616,
      "step": 1500
    },
    {
      "epoch": 3.954397394136808,
      "eval_loss": 0.2841312289237976,
      "eval_runtime": 138.8794,
      "eval_samples_per_second": 0.72,
      "eval_steps_per_second": 0.022,
      "eval_wer": 73.10344827586206,
      "step": 1500
    },
    {
      "epoch": 4.3257328990228014,
      "grad_norm": 1.0117665529251099,
      "learning_rate": 6.871428571428572e-06,
      "loss": 0.0357,
      "step": 1600
    },
    {
      "epoch": 4.651465798045603,
      "grad_norm": 1.661333441734314,
      "learning_rate": 6.585714285714286e-06,
      "loss": 0.0372,
      "step": 1700
    },
    {
      "epoch": 4.977198697068404,
      "grad_norm": 2.5246667861938477,
      "learning_rate": 6.300000000000001e-06,
      "loss": 0.0399,
      "step": 1800
    },
    {
      "epoch": 4.977198697068404,
      "eval_loss": 0.32145407795906067,
      "eval_runtime": 141.4007,
      "eval_samples_per_second": 0.707,
      "eval_steps_per_second": 0.021,
      "eval_wer": 72.24137931034483,
      "step": 1800
    },
    {
      "epoch": 5.3257328990228014,
      "grad_norm": 0.9164466261863708,
      "learning_rate": 6.014285714285715e-06,
      "loss": 0.0203,
      "step": 1900
    },
    {
      "epoch": 5.651465798045603,
      "grad_norm": 1.748327374458313,
      "learning_rate": 5.7285714285714285e-06,
      "loss": 0.0222,
      "step": 2000
    },
    {
      "epoch": 5.977198697068404,
      "grad_norm": 1.8888225555419922,
      "learning_rate": 5.442857142857143e-06,
      "loss": 0.0218,
      "step": 2100
    },
    {
      "epoch": 5.977198697068404,
      "eval_loss": 0.38806796073913574,
      "eval_runtime": 138.8021,
      "eval_samples_per_second": 0.72,
      "eval_steps_per_second": 0.022,
      "eval_wer": 73.79310344827587,
      "step": 2100
    },
    {
      "epoch": 6.3257328990228014,
      "grad_norm": 1.3422770500183105,
      "learning_rate": 5.157142857142857e-06,
      "loss": 0.0529,
      "step": 2200
    },
    {
      "epoch": 6.651465798045603,
      "grad_norm": 1.5599135160446167,
      "learning_rate": 4.871428571428572e-06,
      "loss": 0.0488,
      "step": 2300
    },
    {
      "epoch": 6.977198697068404,
      "grad_norm": 1.3102701902389526,
      "learning_rate": 4.585714285714286e-06,
      "loss": 0.046,
      "step": 2400
    },
    {
      "epoch": 6.977198697068404,
      "eval_loss": 0.27721107006073,
      "eval_runtime": 147.7865,
      "eval_samples_per_second": 0.677,
      "eval_steps_per_second": 0.02,
      "eval_wer": 74.13793103448276,
      "step": 2400
    },
    {
      "epoch": 7.302931596091205,
      "grad_norm": 1.040212869644165,
      "learning_rate": 4.3e-06,
      "loss": 0.0198,
      "step": 2500
    },
    {
      "epoch": 7.628664495114006,
      "grad_norm": 0.8678199648857117,
      "learning_rate": 4.014285714285715e-06,
      "loss": 0.0174,
      "step": 2600
    },
    {
      "epoch": 7.954397394136808,
      "grad_norm": 1.5105644464492798,
      "learning_rate": 3.7285714285714286e-06,
      "loss": 0.018,
      "step": 2700
    },
    {
      "epoch": 7.954397394136808,
      "eval_loss": 0.3344331681728363,
      "eval_runtime": 150.3178,
      "eval_samples_per_second": 0.665,
      "eval_steps_per_second": 0.02,
      "eval_wer": 71.37931034482759,
      "step": 2700
    },
    {
      "epoch": 8.280130293159608,
      "grad_norm": 0.5797841548919678,
      "learning_rate": 3.4428571428571434e-06,
      "loss": 0.0079,
      "step": 2800
    },
    {
      "epoch": 8.60586319218241,
      "grad_norm": 0.5921549797058105,
      "learning_rate": 3.1571428571428573e-06,
      "loss": 0.0068,
      "step": 2900
    },
    {
      "epoch": 8.931596091205211,
      "grad_norm": 0.7851368188858032,
      "learning_rate": 2.8714285714285717e-06,
      "loss": 0.0067,
      "step": 3000
    },
    {
      "epoch": 8.931596091205211,
      "eval_loss": 0.39471304416656494,
      "eval_runtime": 145.8938,
      "eval_samples_per_second": 0.685,
      "eval_steps_per_second": 0.021,
      "eval_wer": 71.72413793103448,
      "step": 3000
    },
    {
      "epoch": 9.257328990228013,
      "grad_norm": 0.23316721618175507,
      "learning_rate": 2.5857142857142856e-06,
      "loss": 0.0031,
      "step": 3100
    },
    {
      "epoch": 9.583061889250814,
      "grad_norm": 0.7372754216194153,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.0027,
      "step": 3200
    },
    {
      "epoch": 9.908794788273616,
      "grad_norm": 0.2684876322746277,
      "learning_rate": 2.0142857142857144e-06,
      "loss": 0.0023,
      "step": 3300
    },
    {
      "epoch": 9.908794788273616,
      "eval_loss": 0.42460787296295166,
      "eval_runtime": 146.943,
      "eval_samples_per_second": 0.681,
      "eval_steps_per_second": 0.02,
      "eval_wer": 72.58620689655172,
      "step": 3300
    },
    {
      "epoch": 10.234527687296417,
      "grad_norm": 0.5388866662979126,
      "learning_rate": 1.7285714285714287e-06,
      "loss": 0.0012,
      "step": 3400
    },
    {
      "epoch": 10.560260586319218,
      "grad_norm": 0.6833564639091492,
      "learning_rate": 1.442857142857143e-06,
      "loss": 0.0007,
      "step": 3500
    },
    {
      "epoch": 10.88599348534202,
      "grad_norm": 0.26250678300857544,
      "learning_rate": 1.1571428571428572e-06,
      "loss": 0.0008,
      "step": 3600
    },
    {
      "epoch": 10.88599348534202,
      "eval_loss": 0.45030444860458374,
      "eval_runtime": 147.162,
      "eval_samples_per_second": 0.68,
      "eval_steps_per_second": 0.02,
      "eval_wer": 71.72413793103448,
      "step": 3600
    },
    {
      "epoch": 11.211726384364821,
      "grad_norm": 0.03168971464037895,
      "learning_rate": 8.714285714285716e-07,
      "loss": 0.0005,
      "step": 3700
    },
    {
      "epoch": 11.537459283387623,
      "grad_norm": 0.020701788365840912,
      "learning_rate": 5.857142857142857e-07,
      "loss": 0.0003,
      "step": 3800
    },
    {
      "epoch": 11.863192182410424,
      "grad_norm": 0.05785975977778435,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.0003,
      "step": 3900
    },
    {
      "epoch": 11.863192182410424,
      "eval_loss": 0.4610276520252228,
      "eval_runtime": 143.56,
      "eval_samples_per_second": 0.697,
      "eval_steps_per_second": 0.021,
      "eval_wer": 71.03448275862068,
      "step": 3900
    }
  ],
  "logging_steps": 100,
  "max_steps": 4000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 14,
  "save_steps": 300,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.117320944648192e+20,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}