{
  "best_metric": 102.49597423510468,
  "best_model_checkpoint": "./whisper-medium-attempt1-with-silence-text/checkpoint-1100",
  "epoch": 56.25,
  "eval_steps": 100,
  "global_step": 1800,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.78125,
      "grad_norm": 9.995346069335938,
      "learning_rate": 5.000000000000001e-07,
      "loss": 2.8857,
      "step": 25
    },
    {
      "epoch": 1.5625,
      "grad_norm": 10.72589111328125,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 2.824,
      "step": 50
    },
    {
      "epoch": 2.34375,
      "grad_norm": 10.114445686340332,
      "learning_rate": 1.5e-06,
      "loss": 2.6291,
      "step": 75
    },
    {
      "epoch": 3.125,
      "grad_norm": 8.762908935546875,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 2.502,
      "step": 100
    },
    {
      "epoch": 3.125,
      "eval_loss": 2.8539323806762695,
      "eval_runtime": 96.3981,
      "eval_samples_per_second": 1.317,
      "eval_steps_per_second": 0.166,
      "eval_wer": 127.10681696188944,
      "step": 100
    },
    {
      "epoch": 3.90625,
      "grad_norm": 11.801437377929688,
      "learning_rate": 2.5e-06,
      "loss": 2.297,
      "step": 125
    },
    {
      "epoch": 4.6875,
      "grad_norm": 9.479702949523926,
      "learning_rate": 3e-06,
      "loss": 1.9392,
      "step": 150
    },
    {
      "epoch": 5.46875,
      "grad_norm": 7.965484619140625,
      "learning_rate": 3.5e-06,
      "loss": 1.6685,
      "step": 175
    },
    {
      "epoch": 6.25,
      "grad_norm": 9.829634666442871,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.3556,
      "step": 200
    },
    {
      "epoch": 6.25,
      "eval_loss": 2.4153664112091064,
      "eval_runtime": 83.527,
      "eval_samples_per_second": 1.52,
      "eval_steps_per_second": 0.192,
      "eval_wer": 115.96886741814278,
      "step": 200
    },
    {
      "epoch": 7.03125,
      "grad_norm": 7.646000862121582,
      "learning_rate": 4.5e-06,
      "loss": 1.2347,
      "step": 225
    },
    {
      "epoch": 7.8125,
      "grad_norm": 9.588939666748047,
      "learning_rate": 5e-06,
      "loss": 1.0514,
      "step": 250
    },
    {
      "epoch": 8.59375,
      "grad_norm": 11.105179786682129,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.8558,
      "step": 275
    },
    {
      "epoch": 9.375,
      "grad_norm": 10.725419044494629,
      "learning_rate": 6e-06,
      "loss": 0.7175,
      "step": 300
    },
    {
      "epoch": 9.375,
      "eval_loss": 2.746173143386841,
      "eval_runtime": 72.6785,
      "eval_samples_per_second": 1.747,
      "eval_steps_per_second": 0.22,
      "eval_wer": 102.97906602254427,
      "step": 300
    },
    {
      "epoch": 10.15625,
      "grad_norm": 12.282588958740234,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.5836,
      "step": 325
    },
    {
      "epoch": 10.9375,
      "grad_norm": 9.197545051574707,
      "learning_rate": 7e-06,
      "loss": 0.4481,
      "step": 350
    },
    {
      "epoch": 11.71875,
      "grad_norm": 9.019693374633789,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.3159,
      "step": 375
    },
    {
      "epoch": 12.5,
      "grad_norm": 10.85390567779541,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.2532,
      "step": 400
    },
    {
      "epoch": 12.5,
      "eval_loss": 3.102280855178833,
      "eval_runtime": 82.6864,
      "eval_samples_per_second": 1.536,
      "eval_steps_per_second": 0.194,
      "eval_wer": 114.78797638217928,
      "step": 400
    },
    {
      "epoch": 13.28125,
      "grad_norm": 8.349029541015625,
      "learning_rate": 8.5e-06,
      "loss": 0.204,
      "step": 425
    },
    {
      "epoch": 14.0625,
      "grad_norm": 7.725856304168701,
      "learning_rate": 9e-06,
      "loss": 0.1613,
      "step": 450
    },
    {
      "epoch": 14.84375,
      "grad_norm": 5.344030857086182,
      "learning_rate": 9.5e-06,
      "loss": 0.1272,
      "step": 475
    },
    {
      "epoch": 15.625,
      "grad_norm": 7.219849109649658,
      "learning_rate": 1e-05,
      "loss": 0.1002,
      "step": 500
    },
    {
      "epoch": 15.625,
      "eval_loss": 3.210036277770996,
      "eval_runtime": 89.9085,
      "eval_samples_per_second": 1.413,
      "eval_steps_per_second": 0.178,
      "eval_wer": 117.98174986580783,
      "step": 500
    },
    {
      "epoch": 16.40625,
      "grad_norm": 6.42445182800293,
      "learning_rate": 9.833333333333333e-06,
      "loss": 0.0865,
      "step": 525
    },
    {
      "epoch": 17.1875,
      "grad_norm": 4.880402565002441,
      "learning_rate": 9.666666666666667e-06,
      "loss": 0.0842,
      "step": 550
    },
    {
      "epoch": 17.96875,
      "grad_norm": 7.4091715812683105,
      "learning_rate": 9.5e-06,
      "loss": 0.0622,
      "step": 575
    },
    {
      "epoch": 18.75,
      "grad_norm": 6.010676860809326,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.0519,
      "step": 600
    },
    {
      "epoch": 18.75,
      "eval_loss": 3.2354989051818848,
      "eval_runtime": 89.2939,
      "eval_samples_per_second": 1.422,
      "eval_steps_per_second": 0.179,
      "eval_wer": 128.60976918947932,
      "step": 600
    },
    {
      "epoch": 19.53125,
      "grad_norm": 5.096531867980957,
      "learning_rate": 9.166666666666666e-06,
      "loss": 0.0541,
      "step": 625
    },
    {
      "epoch": 20.3125,
      "grad_norm": 4.774496555328369,
      "learning_rate": 9e-06,
      "loss": 0.0375,
      "step": 650
    },
    {
      "epoch": 21.09375,
      "grad_norm": 2.6185219287872314,
      "learning_rate": 8.833333333333334e-06,
      "loss": 0.0333,
      "step": 675
    },
    {
      "epoch": 21.875,
      "grad_norm": 9.690539360046387,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.0341,
      "step": 700
    },
    {
      "epoch": 21.875,
      "eval_loss": 3.320791006088257,
      "eval_runtime": 67.8592,
      "eval_samples_per_second": 1.872,
      "eval_steps_per_second": 0.236,
      "eval_wer": 105.71658615136876,
      "step": 700
    },
    {
      "epoch": 22.65625,
      "grad_norm": 2.506408929824829,
      "learning_rate": 8.5e-06,
      "loss": 0.0279,
      "step": 725
    },
    {
      "epoch": 23.4375,
      "grad_norm": 1.42463219165802,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.0286,
      "step": 750
    },
    {
      "epoch": 24.21875,
      "grad_norm": 3.5268752574920654,
      "learning_rate": 8.166666666666668e-06,
      "loss": 0.0289,
      "step": 775
    },
    {
      "epoch": 25.0,
      "grad_norm": 3.3206305503845215,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.024,
      "step": 800
    },
    {
      "epoch": 25.0,
      "eval_loss": 3.507096290588379,
      "eval_runtime": 66.3922,
      "eval_samples_per_second": 1.913,
      "eval_steps_per_second": 0.241,
      "eval_wer": 102.54965110037573,
      "step": 800
    },
    {
      "epoch": 25.78125,
      "grad_norm": 2.4066598415374756,
      "learning_rate": 7.833333333333333e-06,
      "loss": 0.0174,
      "step": 825
    },
    {
      "epoch": 26.5625,
      "grad_norm": 2.2858002185821533,
      "learning_rate": 7.666666666666667e-06,
      "loss": 0.0187,
      "step": 850
    },
    {
      "epoch": 27.34375,
      "grad_norm": 4.5674920082092285,
      "learning_rate": 7.500000000000001e-06,
      "loss": 0.0169,
      "step": 875
    },
    {
      "epoch": 28.125,
      "grad_norm": 2.562084197998047,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.015,
      "step": 900
    },
    {
      "epoch": 28.125,
      "eval_loss": 3.605081558227539,
      "eval_runtime": 69.6271,
      "eval_samples_per_second": 1.824,
      "eval_steps_per_second": 0.23,
      "eval_wer": 103.03274288781535,
      "step": 900
    },
    {
      "epoch": 28.90625,
      "grad_norm": 1.727845549583435,
      "learning_rate": 7.166666666666667e-06,
      "loss": 0.0118,
      "step": 925
    },
    {
      "epoch": 29.6875,
      "grad_norm": 2.76086163520813,
      "learning_rate": 7e-06,
      "loss": 0.0102,
      "step": 950
    },
    {
      "epoch": 30.46875,
      "grad_norm": 1.5665087699890137,
      "learning_rate": 6.833333333333334e-06,
      "loss": 0.0121,
      "step": 975
    },
    {
      "epoch": 31.25,
      "grad_norm": 0.22300420701503754,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.0098,
      "step": 1000
    },
    {
      "epoch": 31.25,
      "eval_loss": 3.6116156578063965,
      "eval_runtime": 85.691,
      "eval_samples_per_second": 1.482,
      "eval_steps_per_second": 0.187,
      "eval_wer": 118.49168008588298,
      "step": 1000
    },
    {
      "epoch": 32.03125,
      "grad_norm": 0.2300027757883072,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.0079,
      "step": 1025
    },
    {
      "epoch": 32.8125,
      "grad_norm": 1.8441766500473022,
      "learning_rate": 6.333333333333333e-06,
      "loss": 0.0065,
      "step": 1050
    },
    {
      "epoch": 33.59375,
      "grad_norm": 1.258414626121521,
      "learning_rate": 6.166666666666667e-06,
      "loss": 0.0052,
      "step": 1075
    },
    {
      "epoch": 34.375,
      "grad_norm": 0.18975859880447388,
      "learning_rate": 6e-06,
      "loss": 0.004,
      "step": 1100
    },
    {
      "epoch": 34.375,
      "eval_loss": 3.6429362297058105,
      "eval_runtime": 65.4005,
      "eval_samples_per_second": 1.942,
      "eval_steps_per_second": 0.245,
      "eval_wer": 102.49597423510468,
      "step": 1100
    },
    {
      "epoch": 35.15625,
      "grad_norm": 0.09774478524923325,
      "learning_rate": 5.833333333333334e-06,
      "loss": 0.0074,
      "step": 1125
    },
    {
      "epoch": 35.9375,
      "grad_norm": 0.15500891208648682,
      "learning_rate": 5.666666666666667e-06,
      "loss": 0.0046,
      "step": 1150
    },
    {
      "epoch": 36.71875,
      "grad_norm": 0.36856603622436523,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.0043,
      "step": 1175
    },
    {
      "epoch": 37.5,
      "grad_norm": 1.4221413135528564,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.0053,
      "step": 1200
    },
    {
      "epoch": 37.5,
      "eval_loss": 3.6545326709747314,
      "eval_runtime": 67.6375,
      "eval_samples_per_second": 1.878,
      "eval_steps_per_second": 0.237,
      "eval_wer": 105.01878690284488,
      "step": 1200
    },
    {
      "epoch": 38.28125,
      "grad_norm": 0.2778646945953369,
      "learning_rate": 5.1666666666666675e-06,
      "loss": 0.0043,
      "step": 1225
    },
    {
      "epoch": 39.0625,
      "grad_norm": 2.6389753818511963,
      "learning_rate": 5e-06,
      "loss": 0.0055,
      "step": 1250
    },
    {
      "epoch": 39.84375,
      "grad_norm": 0.06653451919555664,
      "learning_rate": 4.833333333333333e-06,
      "loss": 0.0028,
      "step": 1275
    },
    {
      "epoch": 40.625,
      "grad_norm": 0.3676530122756958,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.0018,
      "step": 1300
    },
    {
      "epoch": 40.625,
      "eval_loss": 3.79095458984375,
      "eval_runtime": 77.3155,
      "eval_samples_per_second": 1.643,
      "eval_steps_per_second": 0.207,
      "eval_wer": 113.3923778851315,
      "step": 1300
    },
    {
      "epoch": 41.40625,
      "grad_norm": 0.04536161571741104,
      "learning_rate": 4.5e-06,
      "loss": 0.0037,
      "step": 1325
    },
    {
      "epoch": 42.1875,
      "grad_norm": 0.03454182669520378,
      "learning_rate": 4.333333333333334e-06,
      "loss": 0.0027,
      "step": 1350
    },
    {
      "epoch": 42.96875,
      "grad_norm": 0.02412017434835434,
      "learning_rate": 4.166666666666667e-06,
      "loss": 0.0008,
      "step": 1375
    },
    {
      "epoch": 43.75,
      "grad_norm": 0.01602977328002453,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0005,
      "step": 1400
    },
    {
      "epoch": 43.75,
      "eval_loss": 3.8307735919952393,
      "eval_runtime": 78.9045,
      "eval_samples_per_second": 1.61,
      "eval_steps_per_second": 0.203,
      "eval_wer": 110.89640365002684,
      "step": 1400
    },
    {
      "epoch": 44.53125,
      "grad_norm": 0.01733047142624855,
      "learning_rate": 3.833333333333334e-06,
      "loss": 0.0009,
      "step": 1425
    },
    {
      "epoch": 45.3125,
      "grad_norm": 0.013544516637921333,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 0.0006,
      "step": 1450
    },
    {
      "epoch": 46.09375,
      "grad_norm": 0.015972165390849113,
      "learning_rate": 3.5e-06,
      "loss": 0.0006,
      "step": 1475
    },
    {
      "epoch": 46.875,
      "grad_norm": 0.017145907506346703,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.0004,
      "step": 1500
    },
    {
      "epoch": 46.875,
      "eval_loss": 3.8879342079162598,
      "eval_runtime": 68.6801,
      "eval_samples_per_second": 1.849,
      "eval_steps_per_second": 0.233,
      "eval_wer": 105.28717122920021,
      "step": 1500
    },
    {
      "epoch": 47.65625,
      "grad_norm": 0.014514419250190258,
      "learning_rate": 3.1666666666666667e-06,
      "loss": 0.0005,
      "step": 1525
    },
    {
      "epoch": 48.4375,
      "grad_norm": 0.013589655980467796,
      "learning_rate": 3e-06,
      "loss": 0.0004,
      "step": 1550
    },
    {
      "epoch": 49.21875,
      "grad_norm": 0.012899060733616352,
      "learning_rate": 2.8333333333333335e-06,
      "loss": 0.0004,
      "step": 1575
    },
    {
      "epoch": 50.0,
      "grad_norm": 0.8958866000175476,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.0005,
      "step": 1600
    },
    {
      "epoch": 50.0,
      "eval_loss": 3.9086320400238037,
      "eval_runtime": 69.0885,
      "eval_samples_per_second": 1.838,
      "eval_steps_per_second": 0.232,
      "eval_wer": 103.75738056897477,
      "step": 1600
    },
    {
      "epoch": 50.78125,
      "grad_norm": 0.010927481576800346,
      "learning_rate": 2.5e-06,
      "loss": 0.0005,
      "step": 1625
    },
    {
      "epoch": 51.5625,
      "grad_norm": 0.011656476184725761,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 0.0003,
      "step": 1650
    },
    {
      "epoch": 52.34375,
      "grad_norm": 0.010895876213908195,
      "learning_rate": 2.166666666666667e-06,
      "loss": 0.0004,
      "step": 1675
    },
    {
      "epoch": 53.125,
      "grad_norm": 0.011804311536252499,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0003,
      "step": 1700
    },
    {
      "epoch": 53.125,
      "eval_loss": 3.924144983291626,
      "eval_runtime": 76.5243,
      "eval_samples_per_second": 1.66,
      "eval_steps_per_second": 0.209,
      "eval_wer": 110.11808910359635,
      "step": 1700
    },
    {
      "epoch": 53.90625,
      "grad_norm": 0.012491494417190552,
      "learning_rate": 1.8333333333333333e-06,
      "loss": 0.0003,
      "step": 1725
    },
    {
      "epoch": 54.6875,
      "grad_norm": 0.009820288978517056,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 0.0003,
      "step": 1750
    },
    {
      "epoch": 55.46875,
      "grad_norm": 0.010045156814157963,
      "learning_rate": 1.5e-06,
      "loss": 0.0003,
      "step": 1775
    },
    {
      "epoch": 56.25,
      "grad_norm": 0.010024785995483398,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.0003,
      "step": 1800
    },
    {
      "epoch": 56.25,
      "eval_loss": 3.9369020462036133,
      "eval_runtime": 77.0292,
      "eval_samples_per_second": 1.649,
      "eval_steps_per_second": 0.208,
      "eval_wer": 107.27321524422975,
      "step": 1800
    }
  ],
  "logging_steps": 25,
  "max_steps": 2000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 63,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.91076519624704e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}