RVC-Project / RVC-Project/Retrieval-based-Voice-Conversion-WebUI

CUDNN with RMVPE pitch

Open
#2,810 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
38.4k
Forks
5.3k
PR merge metrics
No merged PRs in 30d

Description

When I have to do an inference in RVC, I have to turn off CUDNN with the following line:
torch.backends.cudnn.enabled=False
It seems that RMVPE does not support CUDNN with excessively long files. I am using the rtx 5090 version.

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by reproducing inference on an RTX 5090 with an excessively long audio file, comparing RMVPE behavior with torch.backends.cudnn.enabled set to True and False. Trace the RMVPE inference path to identify the CUDNN failure, and consider the issue done when long-file inference works without requiring users to disable CUDNN.

Written by the indexing model from the issue text.

Assessment

Tech stack
python, pytorch
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Quiet
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.