Support Cosyvoice2-0.5B By allowing Qwen2 architecture to have a optional bias tensor #14711
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
ggerganov
approved these changes
Jul 16, 2025
|
@tempstudio Hello, may I ask how I should load and run inference with the cosyvoice2 model? Thanks. |
|
@tempstudio I was thinking about it — this approach only allows llama.cpp to be used as a library, correct? It can’t yet run inference for CosyVoice directly, as preprocessing and postprocessing would still need to be handled by external code. |
|
@hipudding Yes you are right, this is a replacement for the LLM part of Cosyvoice. It doesn't cover the FLOW and HIFIGAN parts of Cosyvoice. |
Thanks. Do you have plans to fully support CosyVoice? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Cosyvoice2-0.5B (https://github.com/FunAudioLLM/CosyVoice/blob/main/cosyvoice/vllm/cosyvoice2.py#L93) is a TTS model finetuned on top of the Qwen2-0.5B model, with an extra bias tensor on the decoder head.
This change allows this bias tensor to be loaded for better quality when running Cosyvoice2 in llama.cpp.