-
Notifications
You must be signed in to change notification settings - Fork 3.8k
Pull requests: microsoft/VibeVoice
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
test: Add package naming consistency tests for lowercase vibevoice
#304
opened Apr 1, 2026 by
shoaib050326
Loading…
5 tasks done
fix: MockCacheLayer.get_mask_sizes must include query_length (breaks SDPA on MPS/CPU)
#303
opened Apr 1, 2026 by
eddieran
Loading…
3 of 4 tasks
Fix: use except Empty instead of bare except in AudioBatchIterator
#296
opened Mar 31, 2026 by
ekkoitac
Loading…
fix: add vLLM 0.16+ compatibility for multimodal processor APIs
#291
opened Mar 31, 2026 by
yubingjiaocn
Loading…
fix: replace bare except with specific Empty exception
#287
opened Mar 31, 2026 by
mango766
Loading…
Fix: use torch_dtype instead of dtype in model loading
#284
opened Mar 31, 2026 by
10Unknownboy
Loading…
fix: int16 overflow in audio scaling and invalid device choice
#281
opened Mar 30, 2026 by
Bortlesboat
Loading…
fix: NameError when no speech output in realtime demo
#280
opened Mar 30, 2026 by
Bortlesboat
Loading…
fix: reduce max_new_tokens default from 32768 to 4096
#279
opened Mar 30, 2026 by
Bortlesboat
Loading…
fix: use COMMON_AUDIO_EXTS for --audio_dir file discovery
#278
opened Mar 30, 2026 by
mvanhorn
Loading…
fix: add audio duration validation to prevent OOM in vLLM plugin
#276
opened Mar 30, 2026 by
mvanhorn
Loading…
feat: add optional Traditional Chinese output via OpenCC post-processing
#274
opened Mar 30, 2026 by
GekkoQuest
Loading…
fix: use torch_dtype instead of dtype in from_pretrained calls
#273
opened Mar 30, 2026 by
JasonOA888
Loading…
Previous Next
ProTip!
Follow long discussions with comments:>50.