Commits · f09c5feb7c66340e35c06aba37478cce74c20c10 · OpenDAS / vllm_cscc

06 Jan, 2026 7 commits

[Perf] Async Scheduling + Speculative Decoding + Structured Outputs (#29821) · f7008ce1

Benjamin Chislett authored Jan 06, 2026


Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>

f7008ce1

make 500: InternalServerError more informative (#20610) · 142c4d17
Masataro Asai authored Jan 06, 2026
```
Signed-off-by: Masataro Asai <guicho2.71828@gmail.com>
```
142c4d17
[Log] add log about gpu worker init snapshot and requested memory (#29493) · 6f5e6533
Ning Xie authored Jan 07, 2026
```
Signed-off-by: Andy Xie <andy.xning@gmail.com>
```
6f5e6533

[Attention][2/n] Remove usage of deprecated `seq_lens_cpu` and... · 4c73be14

Lucas Wilkinson authored Jan 06, 2026


[Attention][2/n] Remove usage of deprecated `seq_lens_cpu` and `num_computed_tokens_cpu` CommonAttentionMetadata properties (#31774)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>

4c73be14

[Attention][1/n] Remove usage of deprecated `seq_lens_cpu` and... · e0327c9d

Lucas Wilkinson authored Jan 06, 2026


[Attention][1/n] Remove usage of deprecated `seq_lens_cpu` and `num_computed_tokens_cpu` CommonAttentionMetadata properties (#31773)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>

e0327c9d

[Bugfix][ROCm] Fix Unsupported attention metadata type for speculative... · e9717801

vllmellm authored Jan 06, 2026


[Bugfix][ROCm] Fix Unsupported attention metadata type for speculative decoding in `eagle.py` (#31714)
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>

e9717801

[Bugfix] vLLM produces invalid UTF-8 tokens and “�” (#28874) · 2f4e6548
John Calderon authored Jan 05, 2026
```
Signed-off-by: John Calderon <jcalderon@nvidia.com>
Co-authored-by: Benjamin Chislett <bchislett@nvidia.com>
```
2f4e6548

05 Jan, 2026 7 commits
- [Bug] Revert torch warning fix (#31585) · af9a7ec2
  Wentao Ye authored Jan 05, 2026
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  af9a7ec2
- [Cleanup] Remove deprecated fields from CachedRequestData class (#31734) · 32f4e4db
  Nick Hill authored Jan 05, 2026
```
Signed-off-by: njhill <nickhill123@gmail.com>
```
  32f4e4db
- Triton Attention: Support cross-layers blocks (#30687) · d8e38d49
  Or Ozeri authored Jan 05, 2026
```
Signed-off-by: Or Ozeri <oro@il.ibm.com>
```
  d8e38d49
- [v1] Add encoder-only/cross attention support to Triton Attention backend (#31406) · 6aa5b18e
  Isotr0py authored Jan 06, 2026
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  6aa5b18e
- [Misc] Various code simplifications (#31666) · 43e3f8e4
  Nick Hill authored Jan 04, 2026
```
Signed-off-by: njhill <nickhill123@gmail.com>
```
  43e3f8e4
- [Platform] Deprecate seed_everything (#31659) · bb4337b3
  wangxiyuan authored Jan 05, 2026
```
Signed-off-by: wangxiyuan <wangxiyuan1007@gmail.com>
```
  bb4337b3
- [Minor] Small pooler output processing optimization (#31667) · da436f86
  Nick Hill authored Jan 04, 2026
```
Signed-off-by: njhill <nickhill123@gmail.com>
```
  da436f86
04 Jan, 2026 1 commit
- [BugFix] Async scheduling: handle model forward errors more cleanly (#31611) · b53b89fd
  Nick Hill authored Jan 04, 2026
```
Signed-off-by: njhill <nickhill123@gmail.com>
```
  b53b89fd
02 Jan, 2026 4 commits
- [BugFix] Support online dense model DP without overhead (#30739) · bd877162
  Nick Hill authored Jan 02, 2026
```
Signed-off-by: Nick Hill <nhill@redhat.com>
Signed-off-by: njhill <nickhill123@gmail.com>
```
  bd877162
- [Bugfix][Hardware][AMD] Fix last_page_len calculation in AITER MLA decode (#31282) · 825c2dc1
  Kevin McKay authored Jan 01, 2026
```
Signed-off-by: c0de128 <kevin.mckay@outlook.com>
```
  825c2dc1
- [Bugfix] Fix block size used in EAGLE slot mapping (#31540) · ea53ca5e
  Benjamin Chislett authored Jan 01, 2026
```
Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
```
  ea53ca5e
- Add Multimodal Processor Benchmark (#29105) · 1f5b7c41
  Reagan Lee authored Jan 01, 2026
```
Signed-off-by: Reagan Lee <reaganjlee@gmail.com>
Signed-off-by: Reagan <reaganjlee@gmail.com>
```
  1f5b7c41
31 Dec, 2025 3 commits

[BugFix] Fix async scheduling for pooling models (#31584) · 6c2cfb62
Nick Hill authored Dec 31, 2025
```
Signed-off-by: njhill <nickhill123@gmail.com>
```
6c2cfb62

[Bug] Fix log issue with `\n` (#31390) · 357d435c

Wentao Ye authored Dec 31, 2025

Signed-off-by: yewentao256 <zhyanwentao@126.com>
Signed-off-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>

357d435c

[Core] Remove unused `num_tokens` parameter from `_init_model_kwargs` (#31517) · b4bb5f31
maang-h authored Dec 31, 2025
```
Signed-off-by: maang <maang_h@163.com>
```
b4bb5f31

30 Dec, 2025 5 commits

[Bugfix]Fix pooling model always disabled due to incorrect PP rank check (#31505) · 04147dcf
vintipandey authored Dec 30, 2025
```
Signed-off-by: vintipandey <vinti.pandey@gmail.com>
```
04147dcf

[Model] Add support for openPangu moe model (#28775) · 3f52fa5a

yt0428 authored Dec 31, 2025


Signed-off-by: yuantao <2422264527@qq.com>
Signed-off-by: yt0428 <51468697+yt0428@users.noreply.github.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>

3f52fa5a

[Minor] Various small code cleanups/simplifications (#31508) · 3b312fb7
Nick Hill authored Dec 29, 2025
```
Signed-off-by: njhill <nickhill123@gmail.com>
```
3b312fb7
[Core] Deduplicate generate/encode logic in `AsyncLLM` (#31510) · e54ee3ea
Nick Hill authored Dec 29, 2025
```
Signed-off-by: njhill <nickhill123@gmail.com>
```
e54ee3ea

[Prefix Cache] Include lora_name in BlockStored event for deterministic... · 39512aba

Sage authored Dec 30, 2025


[Prefix Cache] Include lora_name in BlockStored event for deterministic KV-cache reconstruction (#27577)
Signed-off-by: Sage Ahrac <sagiahrak@gmail.com>
Co-authored-by: Sage <80211083+sagiahrac@users.noreply.github.com>

39512aba

28 Dec, 2025 1 commit
- [BugFix] Re-fix async multimodal cpu tensor race condition (#31373) · 094fcce2
  Nick Hill authored Dec 28, 2025
```
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Signed-off-by: njhill <nickhill123@gmail.com>
```
  094fcce2
27 Dec, 2025 1 commit
- [Core][Hybrid allocator + connector] Support hybrid allocator + kv cache connector (#30166) · 52bf0665
  Yifan Qiao authored Dec 26, 2025
```
Signed-off-by: Yifan Qiao <yifanqiao@berkeley.edu>
Co-authored-by: KuntaiDu <kuntai@uchicago.edu>
```
  52bf0665
26 Dec, 2025 1 commit

[Core] Initialize LoRA support for tower and connector in multi-modal models (#26674) · ce1eafd1

Jee Jee Li authored Dec 26, 2025


Signed-off-by: bk-201 <joy25810@foxmail.com>
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
Signed-off-by: prashanth058 <prashanth.dannamaneni@uipath.com>
Co-authored-by: bk-201 <joy25810@foxmail.com>
Co-authored-by: prashanth058 <prashanth.dannamaneni@uipath.com>
Co-authored-by: Anexdeus <5142168@mail.ru>

ce1eafd1

25 Dec, 2025 2 commits
- [BugFix] Fix async scheduling + reasoning with struct output (#31332) · 81786c87
  Nick Hill authored Dec 25, 2025
```
Signed-off-by: Nick Hill <nickhill123@gmail.com>
```
  81786c87
- [Hybrid] Mamba2 prefix cache blocks freeing for running requests (#28047) · f1531d9f
  Stan Wozniak authored Dec 25, 2025
```
Signed-off-by: Stanislaw Wozniak <stw@zurich.ibm.com>
Signed-off-by: Chen Zhang <zhangch99@outlook.com>
Co-authored-by: Chen Zhang <zhangch99@outlook.com>
```
  f1531d9f
24 Dec, 2025 6 commits
- [Chore][1/2] Drop `v0.14` deprecations (#31285) · 09dc7c69
  Cyrus Leung authored Dec 25, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  09dc7c69
- [Chore] Remove unused `noqa`s (#31263) · aa3868ec
  Cyrus Leung authored Dec 24, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  aa3868ec
- [Bugfix] Fix `max_model_len="auto"` handling (#31260) · 7adeb4bf
  Cyrus Leung authored Dec 24, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  7adeb4bf
- Add `--max-model-len auto` to auto-fit context to available memory (#29431) · 8ee90c83
  Michael Goin authored Dec 24, 2025
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  8ee90c83
- [KVEvent] User request.block_hash for parent block_hash (#30544) · 538e830c
  Chen Zhang authored Dec 23, 2025
```
Signed-off-by: Chen Zhang <zhangch99@outlook.com>
Signed-off-by: Yifan Qiao <yifanqiao@berkeley.edu>
Co-authored-by: Yifan Qiao <yifanqiao@berkeley.edu>
```
  538e830c
- [Chore] Simplify logic of `_execute_mm_encoder` (#31222) · ca6a95ba
  Cyrus Leung authored Dec 24, 2025
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  ca6a95ba
23 Dec, 2025 2 commits
- [Core] Add a random suffix to frontend-provided request IDs (#27987) · f7900686
  Mark McLoughlin authored Dec 23, 2025
```
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
Signed-off-by: Nick Hill <nhill@redhat.com>
Co-authored-by: Nick Hill <nhill@redhat.com>
```
  f7900686
- [Mamba] - Consolidate Mambas Attention Logic (#28133) · 34916ae3
  Asaf Joseph Gardin authored Dec 23, 2025
  
  34916ae3