Skip to content

[BugFix] Qwen2.5-omni supress end token and won't stop #773

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
hsliuustc0106 merged 4 commits into from
Jan 15, 2026
Merged

[BugFix] Qwen2.5-omni supress end token and won't stop #773

Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
bd75674
fix
yinpeiqi Jan 13, 2026
aa3b6df
fix
yinpeiqi Jan 13, 2026
dd7bcad
Merge branch 'main' into fix-qwen25
yinpeiqi Jan 13, 2026
fc91530
Merge branch 'main' into fix-qwen25
yinpeiqi Jan 15, 2026
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
13 changes: 12 additions & 1 deletion vllm_omni/model_executor/models/qwen2_5_omni/qwen2_5_omni_talker.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -155,7 +155,18 @@ def forward(
def bad_word_processor(self, logits: torch.Tensor) -> torch.Tensor:
# suppress token IDs unsupported by token2wav
if self.suppress_start_id and self.suppress_start_id < logits.size(-1):
logits[..., self.suppress_start_id : logits.size(-1)] = -1e9
# skip the end token id.
if hasattr(self.config, "tts_codec_end_token_id"):
end_id = int(getattr(self.config, "tts_codec_end_token_id"))
if self.suppress_start_id == end_id:
logits[..., end_id + 1 : logits.size(-1)] = -1e9
elif self.suppress_start_id < end_id:
logits[..., self.suppress_start_id : end_id] = -1e9
logits[..., end_id + 1 : logits.size(-1)] = -1e9
else:
logits[..., self.suppress_start_id : logits.size(-1)] = -1e9
else:
raise ValueError("config must have tts_codec_end_token_id attribute")

if hasattr(self.config, "tts_codec_start_token_id"):
bos_id = int(getattr(self.config, "tts_codec_start_token_id"))
Expand Down