HTDemucs / 955717e8.json
MisterCD's picture
Freeze adefossez/HTDemucs
03627aa verified
Raw History Blame Contribute Delete
12.1 kB
{
"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\": 0.0, \"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": false
},
"test": {
"save": false,
"best": true,
"workers": 2,
"every": 20,
"split": true,
"shifts": 0,
"overlap": 0.25,
"sdr": true,
"metric": "loss",
"nonhq": null
},
"epochs": 1200,
"batch_size": 32,
"max_batches": 800,
"optim": {
"lr": 0.0003,
"momentum": 0.9,
"beta2": 0.999,
"loss": "l1",
"optim": "adam",
"weight_decay": 0,
"clip_grad": 0
},
"seed": 42,
"debug": false,
"valid_apply": true,
"flag": null,
"save_every": null,
"weights": [
1.0,
1.0,
1.0,
1.0
],
"augment": {
"shift_same": false,
"repitch": {
"proba": 0.2,
"max_tempo": 12
},
"remix": {
"proba": 1,
"group_size": 4
},
"scale": {
"proba": 1,
"min": 0.25,
"max": 1.25
},
"flip": true
},
"continue_from": null,
"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": [
0.9,
0.95
],
"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": 0.0,
"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": 0.11929254978895187,
"ms": 0.0,
"nsdr": 7.057093620300293,
"nsdr_bass": 9.22514533996582,
"nsdr_drums": 7.960200309753418,
"nsdr_other": 6.199682712554932,
"nsdr_vocals": 4.8433427810668945,
"reco": 0.11929254978895187,
"reco_bass": 0.11017793416976929,
"reco_drums": 0.09322132915258408,
"reco_other": 0.16931581497192383,
"reco_vocals": 0.1044551357626915
},
"ema_batch_0": {
"loss": 0.11823809891939163,
"ms": 0.0,
"nsdr": 7.337116241455078,
"nsdr_bass": 9.39171028137207,
"nsdr_drums": 8.140334129333496,
"nsdr_other": 6.27056884765625,
"nsdr_vocals": 5.5458502769470215,
"reco": 0.11823809891939163,
"reco_bass": 0.10894205421209335,
"reco_drums": 0.09191955626010895,
"reco_other": 0.1681428849697113,
"reco_vocals": 0.10394787788391113
},
"ema_batch_1": {
"loss": 0.11793199926614761,
"ms": 0.0,
"nsdr": 7.349314212799072,
"nsdr_bass": 9.376259803771973,
"nsdr_drums": 8.110164642333984,
"nsdr_other": 6.294198513031006,
"nsdr_vocals": 5.616631984710693,
"reco": 0.11793199926614761,
"reco_bass": 0.10864557325839996,
"reco_drums": 0.09192164987325668,
"reco_other": 0.1676071584224701,
"reco_vocals": 0.1035536676645279
},
"loss": 0.11793199926614761,
"ms": 0.0,
"nsdr": 7.349314212799072,
"nsdr_bass": 9.376259803771973,
"nsdr_drums": 8.110164642333984,
"nsdr_other": 6.294198513031006,
"nsdr_vocals": 5.616631984710693,
"reco": 0.11793199926614761,
"reco_bass": 0.10864557325839996,
"reco_drums": 0.09192164987325668,
"reco_other": 0.1676071584224701,
"reco_vocals": 0.1035536676645279,
"bname": "ema_batch_1",
"ema_epoch_0": {
"loss": 0.11807148158550262,
"ms": 0.0,
"nsdr": 7.310133457183838,
"nsdr_bass": 9.368873596191406,
"nsdr_drums": 8.106219291687012,
"nsdr_other": 6.286090850830078,
"nsdr_vocals": 5.4793477058410645,
"reco": 0.11807148158550262,
"reco_bass": 0.10887450724840164,
"reco_drums": 0.09202440083026886,
"reco_other": 0.16772739589214325,
"reco_vocals": 0.10365963727235794
},
"ema_epoch_1": {
"loss": 0.11798109859228134,
"ms": 0.0,
"nsdr": 7.312227249145508,
"nsdr_bass": 9.354043006896973,
"nsdr_drums": 8.082780838012695,
"nsdr_other": 6.289356231689453,
"nsdr_vocals": 5.522730827331543,
"reco": 0.11798109859228134,
"reco_bass": 0.1086449921131134,
"reco_drums": 0.09200536459684372,
"reco_other": 0.1676090508699417,
"reco_vocals": 0.10366500169038773
},
"best": 0.10993189364671707
},
{
"nsdr_drums": 9.935155015834168,
"nsdr_med_drums": 9.827555879641407,
"nsdr_bass": 8.706514926488623,
"nsdr_med_bass": 8.6324727537917,
"nsdr_other": 6.19734483392449,
"nsdr_med_other": 6.733437504670398,
"nsdr_vocals": 8.701567195216054,
"nsdr_med_vocals": 9.399760082423855,
"nsdr": 8.385145492865835,
"nsdr_med": 8.648306555131839,
"sdr_drums": 10.135625475229388,
"sdr_med_drums": 10.047398993081703,
"sdr_bass": 8.905386659827652,
"sdr_med_bass": 9.78369224165497,
"sdr_other": 5.971391053520985,
"sdr_med_other": 6.424397179969619,
"sdr_vocals": 7.958039606475003,
"sdr_med_vocals": 8.934166549025182,
"sdr": 8.242610698763256,
"sdr_med": 8.797413740932868,
"sir_drums": 13.894430918365115,
"sir_med_drums": 16.651986102598745,
"sir_bass": 17.39139182335939,
"sir_med_bass": 17.358594307530726,
"sir_other": 5.832050495732786,
"sir_med_other": 8.204433992110538,
"sir_vocals": 12.573646287344529,
"sir_med_vocals": 14.905257862584854,
"sir": 12.422879881200455,
"sir_med": 14.280068066206216,
"isr_drums": 16.979595639387902,
"isr_med_drums": 17.08148018608078,
"isr_bass": 8.105441963380866,
"isr_med_bass": 8.373539605597793,
"isr_other": 13.084970535771928,
"isr_med_other": 13.1607226121909,
"isr_vocals": 15.200167626956375,
"isr_med_vocals": 15.106296234433668,
"isr": 13.342543941374267,
"isr_med": 13.430509659575787,
"sar_drums": 8.64656814834116,
"sar_med_drums": 8.70531549825743,
"sar_bass": 6.834327191285907,
"sar_med_bass": 6.621816478499008,
"sar_other": 5.062832351100221,
"sar_med_other": 4.62791839925804,
"sar_vocals": 6.689087019374708,
"sar_med_vocals": 6.778750661141773,
"sar": 6.808203677525499,
"sar_med": 6.683450259289064
}
]
}