training_log.json

{
  "config": {
    "epochs": 30,
    "lr": 0.0003,
    "warmup_steps": 500,
    "grad_accum_steps": 8,
    "grad_clip": 1.0
  },
  "train": [
    {
      "epoch": 1,
      "avg_loss": 6.3824,
      "lr": 9.24e-05,
      "step": 154,
      "time_s": 1992.5
    },
    {
      "epoch": 2,
      "avg_loss": 4.865482,
      "lr": 0.0001848,
      "step": 308,
      "time_s": 1984.5
    },
    {
      "epoch": 3,
      "avg_loss": 4.285284,
      "lr": 0.0002772,
      "step": 462,
      "time_s": 1992.7
    },
    {
      "epoch": 4,
      "avg_loss": 3.949742,
      "lr": 0.00029940583567001586,
      "step": 616,
      "time_s": 1984.4
    },
    {
      "epoch": 5,
      "avg_loss": 3.740897,
      "lr": 0.00029679040564010837,
      "step": 770,
      "time_s": 1975.7
    },
    {
      "epoch": 6,
      "avg_loss": 3.614488,
      "lr": 0.00029212641108894655,
      "step": 924,
      "time_s": 1976.5
    },
    {
      "epoch": 7,
      "avg_loss": 3.530983,
      "lr": 0.00028547894137794454,
      "step": 1078,
      "time_s": 1980.6
    },
    {
      "epoch": 8,
      "avg_loss": 3.470683,
      "lr": 0.000276940766675198,
      "step": 1232,
      "time_s": 1976.5
    },
    {
      "epoch": 9,
      "avg_loss": 3.420985,
      "lr": 0.00026663104328180195,
      "step": 1386,
      "time_s": 1972.6
    },
    {
      "epoch": 10,
      "avg_loss": 3.378666,
      "lr": 0.0002546936507209144,
      "step": 1540,
      "time_s": 1979.8
    },
    {
      "epoch": 11,
      "avg_loss": 3.339928,
      "lr": 0.0002412951837966703,
      "step": 1694,
      "time_s": 1973.9
    },
    {
      "epoch": 12,
      "avg_loss": 3.303412,
      "lr": 0.00022662262764519306,
      "step": 1848,
      "time_s": 1981.5
    },
    {
      "epoch": 13,
      "avg_loss": 3.270916,
      "lr": 0.00021088074822401884,
      "step": 2002,
      "time_s": 1974.9
    },
    {
      "epoch": 14,
      "avg_loss": 3.241058,
      "lr": 0.00019428923465751146,
      "step": 2156,
      "time_s": 1969.6
    },
    {
      "epoch": 15,
      "avg_loss": 3.21413,
      "lr": 0.00017707963331886904,
      "step": 2310,
      "time_s": 1968.4
    }
  ],
  "val": [
    {
      "epoch": 1,
      "avg_loss": 5.362817
    },
    {
      "epoch": 2,
      "avg_loss": 4.490635
    },
    {
      "epoch": 3,
      "avg_loss": 4.07404
    },
    {
      "epoch": 4,
      "avg_loss": 3.811481
    },
    {
      "epoch": 5,
      "avg_loss": 3.647324
    },
    {
      "epoch": 6,
      "avg_loss": 3.559344
    },
    {
      "epoch": 7,
      "avg_loss": 3.5003
    },
    {
      "epoch": 8,
      "avg_loss": 3.45047
    },
    {
      "epoch": 9,
      "avg_loss": 3.420838
    },
    {
      "epoch": 10,
      "avg_loss": 3.387246
    },
    {
      "epoch": 11,
      "avg_loss": 3.363138
    },
    {
      "epoch": 12,
      "avg_loss": 3.336315
    },
    {
      "epoch": 13,
      "avg_loss": 3.319501
    },
    {
      "epoch": 14,
      "avg_loss": 3.298631
    },
    {
      "epoch": 15,
      "avg_loss": 3.285041
    }
  ],
  "sanity": [
    {
      "epoch": 3,
      "audio_path": "/kaggle/working/inference_out.wav"
    },
    {
      "epoch": 6,
      "audio_path": "/kaggle/working/inference_out.wav"
    },
    {
      "epoch": 9,
      "audio_path": "/kaggle/working/inference_out.wav"
    },
    {
      "epoch": 12,
      "audio_path": "/kaggle/working/inference_out.wav"
    },
    {
      "epoch": 15,
      "audio_path": "/kaggle/working/inference_out.wav"
    }
  ]
}