Anjok07 / ultimatevocalremovergui

GUI for a Vocal Remover that uses Deep Neural Networks.
MIT License
18.14k stars 1.35k forks source link

Demucs v3 --segement #191

Open MA-KURO opened 2 years ago

MA-KURO commented 2 years ago

When using the Demucs v3, could you add an option to set --segment. --segemet is a new parameter in demucs-3.0.5a1, demucs/separate.py(https://github.com/facebookresearch/demucs/blob/main/demucs/separate.py), Line 85. It is used to Set split size of each chunk. This can help save memory of graphic card. It should be a number greater than 10. I can't use Demucs with nvidia gtx1650 4GB, unless set --segment 25.

Anjok07 commented 2 years ago

Ahh, segment is technically "Chunk Size" in this GUI. Uncheck "Split Mode" and set the Chunk size to 5. It should work with those settings!