Skip to content
ModelListedExternal
Download

ConvTasNet Violin NonCausal

A NonCausal Violin/Others separation model for the CAD2-Task2 baseline system.

Sourcehuggingface.co/cadenzachallenge/ConvTasNet_Violin_NonCausal/tree/62dd19aa50ce6c91673ac5cca4a8612c5d6cd637
KindOtherVersionRevision 62dd19a

Model source

Kind
Other
File size
25.2 MB
Version
vRevision 62dd19a
Updated
Sep 18, 2026
Downloads
27
Tasks
Audio
Source
Hugging Face

Source description

Cadenza Challenge: CAD2-Task1

A NonCausal Violin/Others separation model for the CAD2-Task2 baseline system.

  • Architecture: ConvTasNet (Kaituo XU) with multichannel support (Alexandre Defossez).
  • Parameters:
    • B: 256
    • C: 2
    • H: 512
    • L: 20
    • N: 256
    • P: 3
    • R: 3
    • X: 8
    • audio_channels: 2
    • causal: false
    • mask_nonlinear: relu
    • norm_type: gLN
  • training:
    • sample_rate: 44100
    • samples_per_track: 64
    • segment: 5.0
    • aggregate: 2
    • batch_size: 4
    • early_stop: true
    • epochs: 200

Dataset

The model was trained using EnsembleSet and CadenzaWoodwind datasets.

How to use

from tasnet import ConvTasNetStereo

model = ConvTasNetStereo.from_pretrained(
    "cadenzachallenge/ConvTasNet_Violin_NonCausal"
).cpu()

Sources

1 source
Hugging FacePrimaryVersion 62dd19aa50ce6c91673ac5cca4a8612c5d6cd637
huggingface.co/cadenzachallenge/ConvTasNet_Violin_NonCausal/tree/62dd19aa50ce6c91673ac5cca4a8612c5d6cd637

Verified Sep 18

Model artifacts

1 artifact

model.safetensors

safetensors · 25.2 MB · SHA-256 22fa0d542df2…b044 · Hugging Face

Download

Source excerpts

2 excerpts
ReadmeSep 18huggingface.co/cadenzachallenge/ConvTasNet_Violin_NonCausal

--- language: - en license: apache-2.0 tags: - hearing loss - challenge - signal processing - source separation - audio - audio-to-audio - NonCausal --- # Cadenza Challenge: CAD2-Task1 A NonCausal Violin/Others separation model for the CAD2-Task2 baseline system. * Architecture: ConvTasNet (Kaituo XU) with multichannel support (Alexandre Defossez). * Parameters: * B: 256 * C: 2 * H: 512 * L: 20 * N: 256 * P: 3 * R: 3 * X: 8 * audio_channels: 2 * causal: false * mask_nonlinear: relu * norm_type: gLN * training: * sample_rate: 44100 * samples_per_track: 64 * segment: 5.0 * aggregate: 2 * batch_size: 4 * early_stop: true * epochs: 200 ## Dataset The model was trained using EnsembleSet and CadenzaWoodwind datasets. ## How to use ``` from tasnet import ConvTasNetStereo model = ConvTasNetStereo.from_pretrained( "cadenzachallenge/ConvTasNet_Violin_NonCausal" ).cpu() ```

NoteSep 18huggingface.co/cadenzachallenge/ConvTasNet_Violin_NonCausal

Source context: 27 downloads · 0 likes · Pipeline audio-to-audio · Library transformers · Repo cadenzachallenge/ConvTasNet_Violin_NonCausal