ZFTurbo / MVSEP-CDX23-Cinematic-Sound-Demixing

Model for CDX23 (Cinematic Sound Demixing) contest
38 stars 5 forks source link

How to train? #4

Open jvel07 opened 11 months ago

jvel07 commented 11 months ago

Hi, thanks for the repo! :) Where can one find the training script/instructions?

ZFTurbo commented 11 months ago

I used this repository for training: https://github.com/facebookresearch/demucs

Training instructions: https://github.com/facebookresearch/demucs/blob/main/docs/training.md

jvel07 commented 10 months ago

Thanks, will check that out. So you used demucs repo and trained on DnR. What modifications had to be done to demucs repo in order to train for sfx, speech, and music? (Since demucs is intended for separating music only).