Zum Inhalt springen
ModelAufgelistetExtern
Herunterladen

ConvTasNet Lyrics Causal

A Causal Lyrics/Accompaniment separation model for the CAD2-Task1 baseline system.

ArtOtherVersionRevision d2601a3

Modellquelle

Art
Other
Dateigröße
41,4–41,5 MB
Version
vRevision d2601a3
Aktualisiert
18. Sept. 2026
Downloads
13
Aufgaben
Audio
Quelle
Hugging Face

Quellenbeschreibung

Cadenza Challenge: CAD2-Task1

A Causal Lyrics/Accompaniment separation model for the CAD2-Task1 baseline system.

Parameters

  • Architecture: ConvTasNet (Kaituo XU) with multichannel support (Alexandre Defossez).
  • Parameters:
    • B: 256
    • C: 2
    • H: 512
    • L: 20
    • N: 256
    • P: 3
    • R: 4
    • X: 10
    • audio_channels: 2
    • causal: true
    • mask_nonlinear: relu
    • norm_type: cLN
  • training:
    • sample_rate: 44100
    • samples_per_track: 64
    • segment: 4.0
    • aggregate: 1
    • batch_size: 4
    • early_stop: true
    • epochs: 200

Dataset

The model was trained on the training split of the MUSDB18-HQ dataset.

How to use

from tasnet import ConvTasNetStereo
model = ConvTasNetStereo.from_pretrained(
    "cadenzachallenge/ConvTasNet_LyricsSeparation_Causal"
).cpu()

Results

TrackVocals (SDR)Accompaniment (SDR)
Al James - Schoolboy Facination5.7338.049
AM Contra - Heart Peripheral5.88712.691
Angels In Amplifiers - I'm Alright5.9019.124
Arise - Run Run Run5.20814.868
Ben Carrigan - We'll Talk About It All Tonight2.6769.919
BKS - Bulldozer1.52311.488
BKS - Too Much7.00511.087
Bobby Nobody - Stitch Up6.51811.303
Buitraker - Revo X4.24213.763
Carlos Gonzalez - A Place For Us3.8827.57
Cristina Vane - So Easy7.47712.126
Detsky Sad - Walkie Talkie6.2149.47
Enda Reilly - Cur An Long Ag Seol7.32911.51
Forkupines - Semantics4.55611.228
Georgia Wonder - Siren3.1657.622
Girls Under Glass - We Feel Alright3.17611.677
Hollow Ground - Ill Fate5.6714.987
James Elder & Mark M Thompson - The English Actor4.0148.834
Juliet's Rescue - Heartbeats5.31713.101
Little Chicago's Finest - My Own4.4095.378
Louis Cressy Band - Good Time5.90310.918
Lyndsey Ollard - Catching Up7.81210.793
M.E.R.C. Music - Knockout5.6637.815
Moosmusic - Big Dummy Shake7.08112.772
Motor Tapes - Shore1.7458.775
Mu - Too Bright5.51812.242
Nerve 9 - Pray For The Rain5.68511.674
PR - Happy Daze-2.8937.274
PR - Oh No08.987
Punkdisco - Oral Hygiene5.04416.173
Raft Monk - Tiring2.1198.977
Sambasevam Shanmugam - Kaathaadi7.519.801
Secretariat - Borderline5.0689.195
Secretariat - Over The Top6.27813.556
Side Effects Project - Sing With Me9.63711.222
Signe Jakobsen - What Have You Done To Me6.8849.656
Skelpolu - Resurrection0.0538.272
Speak Softly - Broken Man3.74313.497
Speak Softly - Like Horses4.3397.233
The Doppler Shift - Atrophy2.4712.58
The Easton Ellises - Falcon 692.5078.137
The Easton Ellises (Baumi) - SDRNR1.4638.136
The Long Wait - Dark Horses4.78410.964
The Mountaineering Club - Mallory9.01513.26
The Sunshine Garcia Band - For I Am The Moon8.34112.1
Timboz - Pony2.69812.415
Tom McKenzie - Directions7.30515.07
Triviul feat. The Fiend - Widow6.4097.938
We Fell From The Sky - Not You3.66111.403
Zeno - Signs5.29110.178
Total (median over frames, median over tracks)5.24911.155

Quellen

1 Quelle
Hugging FacePrimärVersion d2601a366fd8a3bd969852ca5bced524b6c8aae9
huggingface.co/cadenzachallenge/ConvTasNet_Lyrics_Causal/tree/d2601a366fd8a3bd969852ca5bced524b6c8aae9

Verifiziert 18. Sept.

Modellartefakte

2 Artefakte

model.safetensors

safetensors · 41,4 MB · SHA-256 da44e1c9e383…6a90 · Hugging Face

Herunterladen

pytorch_model.bin

bin · 41,5 MB · SHA-256 a7029c44a148…f82b · Hugging Face

Herunterladen

Quellenauszüge

2 Auszüge

--- license: apache-2.0 language: - en tags: - hearing loss - challenge - signal processing - source separation - lyrics intelligibility - audio - audio-to-audio widget: - src: https://cdn-media.huggingface.co/speech_samples/sample1.flac example_title: Test pipeline_tag: audio-to-audio --- # Cadenza Challenge: CAD2-Task1 A Causal Lyrics/Accompaniment separation model for the CAD2-Task1 baseline system. ## Parameters * Architecture: ConvTasNet (Kaituo XU) with multichannel support (Alexandre Defossez). * Parameters: * B: 256 * C: 2 * H: 512 * L: 20 * N: 256 * P: 3 * R: 4 * X: 10 * audio_channels: 2 * causal: true * mask_nonlinear: relu * norm_type: cLN * training: * sample_rate: 44100 * samples_per_track: 64 * segment: 4.0 * aggregate: 1 * batch_size: 4 * early_stop: true * epochs: 200 ## Dataset The model was trained on the training split of the MUSDB18-HQ dataset. ## How to use ``` from tasnet import ConvTasNetStereo model = ConvTasNetStereo.from_pretrained( "cadenzachallenge/ConvTasNet_LyricsSeparation_Causal" ).cpu() ``` ## Results | Track | Vocals (SDR) | Accompaniment (SDR) | |:------|:------------:|:---------:| | Al James - Schoolboy Facination | 5.733 | 8.049 | | AM Contra - Heart Peripheral | 5.887 | 12.691 | | Angels In Amplifiers - I'm Alright | 5.901 | 9.124 | | Arise - Run Run Run | 5.208 | 14.868 | | Ben Carrigan - We'll Talk About It All Tonight | 2.676 | 9.919 | | BKS - Bulldozer | 1.523 | 11.488 | | BKS - Too Much | 7.005 | 11.087 | | Bobby Nobody - Stitch Up | 6.518 | 11.303 | | Buitraker - Revo X | 4.242 | 13.763 | | Carlos Gonzalez - A Place For Us | 3.882 | 7.57 | | Cristina Vane - So Easy | 7.477 | 12.126 | | Detsky Sad - Walkie Talkie | 6.214 | 9.47 | | Enda Reilly - Cur An Long Ag Seol | 7.329 | 11.51 | | Forkupines - Semantics | 4.556 | 11.228 | | Georgia Wonder - Siren | 3.165 | 7.622 | | Girls Under Glass - We Feel Alright | 3.176 | 11.677 | | Hollow Ground - Ill Fate | 5.67 | 14.987 | | James Elder & Mark M Thompson - The English Actor | 4.014 | 8.834 | | Juliet's Rescue - Heartbeats | 5.317 | 13.101 | | Little Chicago's Finest - My Own | 4.409 | 5.378 | | Louis Cressy Band - Good Time | 5.903 | 10.918 | | Lyndsey Ollard - Catching Up | 7.812 | 10.793 | | M.E.R.C. Music - Knockout | 5.663 | 7.815 | | Moosmusic - Big Dummy Shake | 7.081 | 12.772 | | Motor Tapes - Shore | 1.745 | 8.775 | | Mu - Too Bri...

Source context: 13 downloads · 0 likes · Pipeline audio-to-audio · Library transformers · Repo cadenzachallenge/ConvTasNet_Lyrics_Causal