{
  "klass": "demucs.htdemucs.HTDemucs",
  "args": "[]",
  "kwargs": "{\"sources\": [\"drums\", \"bass\", \"other\", \"vocals\"], \"audio_channels\": 2, \"samplerate\": 44100, \"segment\": {\"_type\": \"fraction\", \"numerator\": 39, \"denominator\": 5}, \"channels\": 48, \"channels_time\": null, \"growth\": 2, \"nfft\": 4096, \"wiener_iters\": 0, \"end_iters\": 0, \"wiener_residual\": false, \"cac\": true, \"depth\": 4, \"rewrite\": true, \"multi_freqs\": [], \"multi_freqs_depth\": 3, \"freq_emb\": 0.2, \"emb_scale\": 10, \"emb_smooth\": true, \"kernel_size\": 8, \"stride\": 4, \"time_stride\": 2, \"context\": 1, \"context_enc\": 0, \"norm_starts\": 4, \"norm_groups\": 4, \"dconv_mode\": 3, \"dconv_depth\": 2, \"dconv_comp\": 8, \"dconv_init\": 0.001, \"bottom_channels\": 512, \"t_layers\": 5, \"t_hidden_scale\": 4.0, \"t_heads\": 8, \"t_dropout\": 0.02, \"t_layer_scale\": true, \"t_gelu\": true, \"t_emb\": \"sin\", \"t_max_positions\": 10000, \"t_max_period\": 10000.0, \"t_weight_pos_embed\": 1.0, \"t_cape_mean_normalize\": true, \"t_cape_augment\": true, \"t_cape_glob_loc_scale\": [5000.0, 1.0, 1.4], \"t_sin_random_shift\": 0, \"t_norm_in\": true, \"t_norm_in_group\": false, \"t_group_norm\": false, \"t_norm_first\": true, \"t_norm_out\": true, \"t_weight_decay\": 0.05, \"t_lr\": null, \"t_sparse_self_attn\": false, \"t_sparse_cross_attn\": false, \"t_mask_type\": \"diag\", \"t_mask_random_seed\": 42, \"t_sparse_attn_window\": 400, \"t_global_window\": 100, \"t_sparsity\": 0.95, \"t_auto_sparsity\": false, \"t_cross_first\": false, \"rescale\": 0.1}",
  "training_args": {
    "dummy": null,
    "dset": {
      "musdb": "/checkpoint/defossez/datasets/musdbhq",
      "musdb_samplerate": 44100,
      "wav": "/checkpoint/defossez/datasets/allstems_44/",
      "wav2": "/checkpoint/defossez/datasets/mmi44_goodclean",
      "segment": 10,
      "shift": 1,
      "train_valid": false,
      "full_cv": true,
      "samplerate": 44100,
      "channels": 2,
      "normalize": true,
      "metadata": "./metadata",
      "sources": [
        "drums",
        "bass",
        "other",
        "vocals"
      ],
      "valid_samples": 100,
      "wav2_weight": null,
      "wav2_valid": true
    },
    "test": {
      "save": false,
      "best": true,
      "workers": 2,
      "every": 20,
      "split": true,
      "shifts": 0,
      "overlap": 0.25,
      "sdr": true,
      "metric": "loss",
      "nonhq": null
    },
    "epochs": 50,
    "batch_size": 8,
    "max_batches": 800,
    "optim": {
      "lr": 0.0001,
      "momentum": 0.9,
      "beta2": 0.999,
      "loss": "l1",
      "optim": "adamw",
      "weight_decay": 0,
      "clip_grad": 5
    },
    "seed": 42,
    "debug": false,
    "valid_apply": true,
    "flag": null,
    "save_every": null,
    "weights": [
      0,
      0,
      1,
      0
    ],
    "augment": {
      "shift_same": true,
      "repitch": {
        "proba": 0.2,
        "max_tempo": 12
      },
      "remix": {
        "proba": 0,
        "group_size": 4
      },
      "scale": {
        "proba": 0,
        "min": 0.25,
        "max": 1.25
      },
      "flip": true
    },
    "continue_from": "955717e8",
    "continue_pretrained": null,
    "pretrained_repo": null,
    "continue_best": true,
    "continue_opt": false,
    "misc": {
      "num_workers": 10,
      "num_prints": 4,
      "show": false,
      "verbose": false
    },
    "ema": {
      "epoch": [],
      "batch": [
        0.9995,
        0.9999
      ]
    },
    "use_train_segment": true,
    "model_segment": null,
    "model": "htdemucs",
    "demucs": {
      "channels": 64,
      "growth": 2,
      "depth": 6,
      "rewrite": true,
      "lstm_layers": 0,
      "kernel_size": 8,
      "stride": 4,
      "context": 1,
      "gelu": true,
      "glu": true,
      "norm_groups": 4,
      "norm_starts": 4,
      "dconv_depth": 2,
      "dconv_mode": 1,
      "dconv_comp": 4,
      "dconv_attn": 4,
      "dconv_lstm": 4,
      "dconv_init": 0.0001,
      "resample": true,
      "normalize": false,
      "rescale": 0.1
    },
    "hdemucs": {
      "channels": 48,
      "channels_time": null,
      "growth": 2,
      "nfft": 4096,
      "wiener_iters": 0,
      "end_iters": 0,
      "wiener_residual": false,
      "cac": true,
      "depth": 6,
      "rewrite": true,
      "hybrid": true,
      "hybrid_old": false,
      "multi_freqs": [],
      "multi_freqs_depth": 3,
      "freq_emb": 0.2,
      "emb_scale": 10,
      "emb_smooth": true,
      "kernel_size": 8,
      "stride": 4,
      "time_stride": 2,
      "context": 1,
      "context_enc": 0,
      "norm_starts": 4,
      "norm_groups": 4,
      "dconv_mode": 1,
      "dconv_depth": 2,
      "dconv_comp": 4,
      "dconv_attn": 4,
      "dconv_lstm": 4,
      "dconv_init": 0.001,
      "rescale": 0.1
    },
    "htdemucs": {
      "channels": 48,
      "channels_time": null,
      "growth": 2,
      "nfft": 4096,
      "wiener_iters": 0,
      "end_iters": 0,
      "wiener_residual": false,
      "cac": true,
      "depth": 4,
      "rewrite": true,
      "multi_freqs": [],
      "multi_freqs_depth": 3,
      "freq_emb": 0.2,
      "emb_scale": 10,
      "emb_smooth": true,
      "kernel_size": 8,
      "stride": 4,
      "time_stride": 2,
      "context": 1,
      "context_enc": 0,
      "norm_starts": 4,
      "norm_groups": 4,
      "dconv_mode": 3,
      "dconv_depth": 2,
      "dconv_comp": 8,
      "dconv_init": 0.001,
      "bottom_channels": 512,
      "t_layers": 5,
      "t_hidden_scale": 4.0,
      "t_heads": 8,
      "t_dropout": 0.02,
      "t_layer_scale": true,
      "t_gelu": true,
      "t_emb": "sin",
      "t_max_positions": 10000,
      "t_max_period": 10000.0,
      "t_weight_pos_embed": 1.0,
      "t_cape_mean_normalize": true,
      "t_cape_augment": true,
      "t_cape_glob_loc_scale": [
        5000.0,
        1.0,
        1.4
      ],
      "t_sin_random_shift": 0,
      "t_norm_in": true,
      "t_norm_in_group": false,
      "t_group_norm": false,
      "t_norm_first": true,
      "t_norm_out": true,
      "t_weight_decay": 0.05,
      "t_lr": null,
      "t_sparse_self_attn": false,
      "t_sparse_cross_attn": false,
      "t_mask_type": "diag",
      "t_mask_random_seed": 42,
      "t_sparse_attn_window": 400,
      "t_global_window": 100,
      "t_sparsity": 0.95,
      "t_auto_sparsity": false,
      "t_cross_first": false,
      "rescale": 0.1
    },
    "svd": {
      "penalty": 0,
      "min_size": 0.1,
      "dim": 1,
      "niters": 2,
      "powm": false,
      "proba": 1,
      "conv_only": false,
      "convtr": false,
      "bs": 1
    },
    "quant": {
      "diffq": null,
      "qat": null,
      "min_size": 0.2,
      "group_size": 8
    },
    "dora": {
      "dir": "outputs",
      "exclude": [
        "misc.*",
        "slurm.*",
        "test.reval",
        "flag"
      ]
    },
    "slurm": {
      "time": 4320,
      "constraint": "volta32gb",
      "setup": [
        "module load cudnn/v8.4.1.50-cuda.11.6 NCCL/2.11.4-6-cuda.11.6 cuda/11.6"
      ]
    }
  },
  "metrics": [
    {
      "main": {
        "loss": 0.12592940032482147,
        "ms": 0.0,
        "nsdr": 7.727000713348389,
        "nsdr_bass": 8.049718856811523,
        "nsdr_drums": 8.1212797164917,
        "nsdr_other": 7.727000713348389,
        "nsdr_vocals": 8.013139724731445,
        "reco": 0.12592940032482147,
        "reco_bass": 0.13121698796749115,
        "reco_drums": 0.12236597388982773,
        "reco_other": 0.12592940032482147,
        "reco_vocals": 0.10550956428050995
      },
      "ema_batch_0": {
        "loss": 0.1247486099600792,
        "ms": 0.0,
        "nsdr": 7.798945426940918,
        "nsdr_bass": 8.221861839294434,
        "nsdr_drums": 8.242091178894043,
        "nsdr_other": 7.798945426940918,
        "nsdr_vocals": 8.161120414733887,
        "reco": 0.1247486099600792,
        "reco_bass": 0.12847518920898438,
        "reco_drums": 0.12056565284729004,
        "reco_other": 0.1247486099600792,
        "reco_vocals": 0.10392889380455017
      },
      "ema_batch_1": {
        "loss": 0.12451048940420151,
        "ms": 0.0,
        "nsdr": 7.816592216491699,
        "nsdr_bass": 8.823752403259277,
        "nsdr_drums": 8.79676628112793,
        "nsdr_other": 7.816592216491699,
        "nsdr_vocals": 8.467183113098145,
        "reco": 0.12451048940420151,
        "reco_bass": 0.11904922127723694,
        "reco_drums": 0.1136612519621849,
        "reco_other": 0.12451048940420151,
        "reco_vocals": 0.10050245374441147
      },
      "loss": 0.12451048940420151,
      "ms": 0.0,
      "nsdr": 7.816592216491699,
      "nsdr_bass": 8.823752403259277,
      "nsdr_drums": 8.79676628112793,
      "nsdr_other": 7.816592216491699,
      "nsdr_vocals": 8.467183113098145,
      "reco": 0.12451048940420151,
      "reco_bass": 0.11904922127723694,
      "reco_drums": 0.1136612519621849,
      "reco_other": 0.12451048940420151,
      "reco_vocals": 0.10050245374441147,
      "bname": "ema_batch_1",
      "best": 0.12442848086357117
    },
    {
      "nsdr_drums": 7.716415356805922,
      "nsdr_med_drums": 7.879005060651254,
      "nsdr_bass": 7.010061455359722,
      "nsdr_med_bass": 7.38896759265805,
      "nsdr_other": 6.140593990816037,
      "nsdr_med_other": 6.806810269869251,
      "nsdr_vocals": 7.786545521035497,
      "nsdr_med_vocals": 8.597566647357745,
      "nsdr": 7.163404081004294,
      "nsdr_med": 7.6680873926340745,
      "sdr_drums": 7.621158049470051,
      "sdr_med_drums": 7.5910979788064274,
      "sdr_bass": 6.891626582384467,
      "sdr_med_bass": 8.027040436534891,
      "sdr_other": 5.88551644234907,
      "sdr_med_other": 6.294927693604351,
      "sdr_vocals": 6.391244913826571,
      "sdr_med_vocals": 7.827101321975676,
      "sdr": 6.69738649700754,
      "sdr_med": 7.4350418577303365,
      "sir_drums": 12.389451129732747,
      "sir_med_drums": 16.502254643462216,
      "sir_bass": 14.243431908686649,
      "sir_med_bass": 14.108338202047445,
      "sir_other": 5.911115162590992,
      "sir_med_other": 8.112115933635245,
      "sir_vocals": 8.3569571107282,
      "sir_med_vocals": 11.322223500248347,
      "sir": 10.225238827934646,
      "sir_med": 12.511233069848315,
      "isr_drums": 9.483840022565328,
      "isr_med_drums": 9.719664447793257,
      "isr_bass": 4.761042561214654,
      "isr_med_bass": 6.658945503960653,
      "isr_other": 13.096834001040863,
      "isr_med_other": 13.08585323464818,
      "isr_vocals": 14.62814755070578,
      "isr_med_vocals": 14.52663447842696,
      "isr": 10.492466033881655,
      "isr_med": 10.997774416207262,
      "sar_drums": 8.040148539314487,
      "sar_med_drums": 7.786241619549971,
      "sar_bass": 6.364135526970028,
      "sar_med_bass": 5.840490809086118,
      "sar_other": 5.155545437584214,
      "sar_med_other": 4.709921007855735,
      "sar_vocals": 6.733780355863491,
      "sar_med_vocals": 6.842437031058775,
      "sar": 6.5734024649330545,
      "sar_med": 6.294772616887649
    }
  ]
}