[Realtime] Add Qwen3-ASR realtime streaming support #34613
vllm-bot
merged
12 commits into
vllm-project:main
from
pougetat:qwen3-asr-realtime
[Model] Add Qwen3-ASR realtime streaming support
aa7d9f7a
Reduce realtime_max_tokens from 4096 to 64
4daa5f8b
gemini-code-assist
commented
on 2026-02-16
DarkLight1337
commented
on 2026-02-16
DarkLight1337
commented
on 2026-02-16
DarkLight1337
commented
on 2026-02-16
DarkLight1337
commented
on 2026-02-16
Encode prompt template once instead of redundantly per segment
d815962e
Add docstring for realtime_max_tokens explaining default value
01423260
Remove unnecessary skip_warmup_audio_preprocessing override
edb74b90
Apply ruff formatting
29d1847c
NickLucche
approved these changes
on 2026-02-19
Add comment explaining why a separate model class is needed
0cbfe1a1
Merge branch 'main' into qwen3-asr-realtime
8da7b2f8
Add Qwen3ASRRealtimeGeneration to test model registry
dfd6b796
disabled auto-merge 43 days ago
Head branch was pushed to by a user without write access
Head branch was pushed to by a user without write access
Merge branch 'main' into qwen3-asr-realtime
2dabe373
Merge branch 'main' into qwen3-asr-realtime
8b9f0e90
Merge branch 'main' into qwen3-asr-realtime
99356be6
Assignees
No one assigned
Labels
new-model
frontend
ready
qwen
Milestone
No milestone
Login to write a write a comment.
Login via GitHub