Skip to content

Update cosyvoice2.py for vLLM#1896

Closed
matteocacciola wants to merge 4 commits into
FunAudioLLM:mainfrom
matteocacciola:patch-1
Closed

Update cosyvoice2.py for vLLM#1896
matteocacciola wants to merge 4 commits into
FunAudioLLM:mainfrom
matteocacciola:patch-1

Conversation

@matteocacciola

@matteocacciola matteocacciola commented May 25, 2026

Copy link
Copy Markdown
  • Missing "import from typing" added
  • Missing and required method to load a model with vLLM added (i.e., embed_input_ids)
  • Removing the dependency of the library from torchaudio when loading the wav data

@matteocacciola matteocacciola changed the title Update cosyvoice2.py Update cosyvoice2.py for vLLM May 25, 2026
…y introduced in torchaudio >= 2.7, where torchaudio.load() routes all backends through TorchCodec (which requires FFmpeg 5+ not shipped by Ubuntu 22.04)
@matteocacciola matteocacciola deleted the patch-1 branch May 25, 2026 22:19
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant