Skip to content

[Model] Future-proof Qwen2-Audio multi-modal processor #11776

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Jan 7, 2025
Merged

[Model] Future-proof Qwen2-Audio multi-modal processor #11776

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 4 additions & 2 deletions vllm/model_executor/models/qwen2_audio.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -227,12 +227,14 @@ def get_replacement_qwen2_audio(item_idx: int):
]

def _always_apply_prompt_replacements(self) -> bool:
# HF never applies prompt replacements, so we have to do it ourselves.
# Qwen2-Audio processor will start inserting placeholder tokens
# in an upcoming release:
# https://github.com/huggingface/transformers/pull/35534
# NOTE: `_find_placeholders_by_modality` may incorrectly think that HF
# has already performed processing for multi-audio input when the input
# audios are short (the corresponding placeholders may take up fewer
# tokens than the number of audio items)
return True
return not hasattr(self._get_hf_processor(), "audio_token")


@MULTIMODAL_REGISTRY.register_processor(Qwen2AudioMultiModalProcessor)
Expand Down
Loading