Commits · 0ef5b9147bb1f37c9a90ab2a3ee2a85cf9e84e30 · OpenDAS / vllm_cscc · GitLab

14 Feb, 2026 4 commits
- fix: use `__annotations__` instead of `get_type_hints()` for dynamic `kwargs` detection (#34527) · 0ef5b914
  Christian S. Perone authored Feb 14, 2026
```
Signed-off-by: Christian S. Perone <christian.perone@gmail.com>
Signed-off-by: Christian S. Perone <perone@users.noreply.github.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
```
  0ef5b914
- [bug] Make sure get_modality_with_max_tokens is deterministic (#34533) · ed242652
  Shiyan Deng authored Feb 13, 2026
```
Signed-off-by: Shiyan Deng <dsy842974287@meta.com>
```
  ed242652
- [Feature][Perf] Support Selective CPU Weight Offloading (#34535) · b37b6797
  Wei Zhao authored Feb 13, 2026
```
Signed-off-by: wzhao18 <wzhao18.sz@gmail.com>
```
  b37b6797
- [Bugfix] Fix ROCm UVA CPU weight offloading broken by #32993 (#34543) · a0638d05
  Andreas Karatzas authored Feb 13, 2026
```
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
```
  a0638d05
13 Feb, 2026 35 commits
- [Hybrid] Enable spec decoding in mamba cache align mode (#33705) · c027541e
  Harry Huang authored Feb 14, 2026
```
Signed-off-by: huanghaoyan.hhy <huanghaoyan.hhy@alibaba-inc.com>
```
  c027541e
- [Bugfix]: Fix structured output in multi-turn gpt-oss (#34454) · fd267bc7
  Ben Browning authored Feb 13, 2026
```
Signed-off-by: Ben Browning <bbrownin@redhat.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>
```
  fd267bc7
- Revert "[Bugfix] Fix fused MoE IMA (sans chunking) by using int64 for strides" (#34530) · bfaa5593
  Michael Goin authored Feb 13, 2026
  
  bfaa5593
- [Misc] vLLM's --enforce-eager should turn off compile and cudagraphs only (#34523) · 87789c83
  Richard Zou authored Feb 13, 2026
```
Signed-off-by: Richard Zou <zou3519@gmail.com>
```
  87789c83
- [Feature] Support CPU Offloading without Pytorch Pinned Memory that leads to... · 59d53066
  Wei Zhao authored Feb 13, 2026
```
[Feature] Support CPU Offloading without Pytorch Pinned Memory that leads to doubled allocation (#32993)
Signed-off-by: wzhao18 <wzhao18.sz@gmail.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>
```
  59d53066
- [Bugfix] Add quant_config in ViT of Kimi-K2.5 (#34501) · 4a9952ec
  LoganJane authored Feb 14, 2026
```
Signed-off-by: LoganJane <LoganJane73@hotmail.com>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  4a9952ec
- [Bugfix] Exclude `language_model_only` key from MM AOT compile hash but... · 1dae7b78
  Roger Wang authored Feb 13, 2026
```
[Bugfix] Exclude `language_model_only` key from MM AOT compile hash but include in model one (#34508)
Signed-off-by: Roger Wang <hey@rogerw.io>
```
  1dae7b78
- [Misc] Port Qwen3.5 Configs (#34512) · 5885e330
  Roger Wang authored Feb 13, 2026
```
Signed-off-by: Roger Wang <hey@rogerw.io>
```
  5885e330
- Extend ColBERT support to non-standard BERT backbones (#34170) · 071d863e
  Ilya Boytsov authored Feb 13, 2026
```
Signed-off-by: Ilya Boytsov <ilya.boytsov@aleph-alpha.com>
```
  071d863e
- [GDN] Use CPU tensors to build GDN metadata (#34498) · 0916e796
  Woosuk Kwon authored Feb 13, 2026
```
Signed-off-by: Woosuk Kwon <woosuk@inferact.ai>
```
  0916e796
- [Feature] Pipeline Parallel Async send/recv, 2.9% E2E throughput improvement (#33368) · 3d2a026f
  Wentao Ye authored Feb 13, 2026
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
Signed-off-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
Co-authored-by: Nick Hill <nhill@redhat.com>
```
  3d2a026f
- [Core] Move pause and resume functions into engine (#34125) · dddbff46
  Aaron Hao authored Feb 13, 2026
```
Signed-off-by: ahao-anyscale <ahao@anyscale.com>
Signed-off-by: Aaron Hao <ahao@anyscale.com>
Signed-off-by: hao-aaron <ahao@anyscale.com>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
```
  dddbff46
- [KVConnector] Clean up redundant code in KV connectors (#34147) · 47e9b63e
  Martin Hickey authored Feb 13, 2026
```
Signed-off-by: Martin Hickey <martin.hickey@ie.ibm.com>
```
  47e9b63e
- [Perf] fused_moe: add int4_w4a16 benchmark support and tuning config (#34130) · 934acdde
  Matthias Gehre authored Feb 13, 2026
```
Signed-off-by: Matthias Gehre <matthias.gehre@amd.com>
Co-authored-by: TJian <tunjian.tan@embeddedllm.com>
```
  934acdde
- [Bug Fix] Fix MambaManager.cache_blocks() crash on null blocks in align mode (#34418) · 4137c5df
  haosdent authored Feb 13, 2026
```
Signed-off-by: haosdent <haosdent@gmail.com>
```
  4137c5df
- [BugFix] Fix and optimize max_num_blocks_per_req calculation for MambaSpec (#34440) · 7a8a46dd
  Harry Huang authored Feb 13, 2026
```
Signed-off-by: huanghaoyan.hhy <huanghaoyan.hhy@alibaba-inc.com>
```
  7a8a46dd
- [New Model] support new model ovis2.6 (#34426) · bcf0731a
  myselvess authored Feb 13, 2026
```
Signed-off-by: myselvess <23743269+myselvess@users.noreply.github.com>
```
  bcf0731a
- [Refactor] Call renderer for online IO processor request (#34490) · ec090c24
  Cyrus Leung authored Feb 13, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  ec090c24
- [Bugfix] Fix mamba state dtype setting for Qwen3-Next and Qwen3.5 (#34489) · eea3024f
  Roger Wang authored Feb 12, 2026
```
Signed-off-by: Roger Wang <hey@rogerw.io>
```
  eea3024f
- [Refactor] Pass full VllmConfig to Renderer (#34485) · 2f308214
  Cyrus Leung authored Feb 13, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  2f308214
- [Bugfix] Fix encoder cache underestimation for GLM-4V/GLM-OCR single image (#34483) · dcf6ee85
  haosdent authored Feb 13, 2026
```
Signed-off-by: haosdent <haosdent@gmail.com>
```
  dcf6ee85
- [Bugfix] Standardize getting number of image patches/tokens (#34358) · 372b2e76
  Cyrus Leung authored Feb 13, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  372b2e76
- [Hybrid] Fix and optimize block-aligned splitting in mamba cache align mode (#33706) · bf37812c
  Harry Huang authored Feb 13, 2026
```
Signed-off-by: huanghaoyan.hhy <huanghaoyan.hhy@alibaba-inc.com>
```
  bf37812c
- [Bugfix] Fix Random Dataset Prefix Length Inaccuracy (#33907) · b86bf441
  Frank Wang authored Feb 12, 2026
```
Signed-off-by: frankwang28 <frank.wbb@hotmail.com>
Co-authored-by: Roger Wang <hey@rogerw.io>
```
  b86bf441
- [Kernel] [Helion] [5/N] Add Helion Autotuning infrastructure (#34025) · de13dd78
  Yanan Cao authored Feb 12, 2026
```
Signed-off-by: Yanan Cao <gmagogsfm@gmail.com>
```
  de13dd78
- [Bugfix] Delete unused redundant code in Kimi-K2.5 (#34427) · 62788f99
  LoganJane authored Feb 13, 2026
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  62788f99
- [Refactor] Simplify BOS/EOS token handling (#34435) · ea5ff3a1
  Cyrus Leung authored Feb 13, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  ea5ff3a1
- [Bugfix] Remove assert that's no longer valid (#34443) · 04ea31ba
  bnellnm authored Feb 12, 2026
```
Signed-off-by: Bill Nell <bnell@redhat.com>
```
  04ea31ba
- [BugFix] Add block_size validation for mamba cache align mode (#34445) · 6f019e6e
  Harry Huang authored Feb 13, 2026
```
Signed-off-by: huanghaoyan.hhy <huanghaoyan.hhy@alibaba-inc.com>
```
  6f019e6e
- Fix num_logprobs parameter description in sampler.py (#34451) · d707678d
  Zhuohan Li authored Feb 12, 2026
```
Signed-off-by: Zhuohan Li <zhuohan123@gmail.com>
```
  d707678d
- [Kernel] [Helion] [4/N] Add silu_mul_fp8 Helion kernel (#33373) · 96161fe9
  Yanan Cao authored Feb 12, 2026
```
Signed-off-by: Yanan Cao <gmagogsfm@gmail.com>
```
  96161fe9
- [Core] Profiler improvements and lazy initialization (#33198) · 4453ba8d
  Jaewon authored Feb 12, 2026
```
Signed-off-by: Jaewon Lee <jaewon@meta.com>
Co-authored-by: Lu Fang <30275821+houseroad@users.noreply.github.com>
```
  4453ba8d
- [Core] Add sleep level 0 mode with enqueue/wait pattern (#33195) · aa181c92
  Jaewon authored Feb 12, 2026
```
Signed-off-by: Jaewon Lee <jaewon@meta.com>
Co-authored-by: Lu Fang <30275821+houseroad@users.noreply.github.com>
```
  aa181c92
- [Frontend] Enable generic structured_outputs for responses API (#33709) · be7370da
  Alec S authored Feb 12, 2026
```
Signed-off-by: Alec Solder <alecs@fb.com>
Co-authored-by: Alec Solder <alecs@fb.com>
```
  be7370da
- Use paged_attention_v1 for sliding window decode in rocm_aiter_fa (#34378) · 9ea1f598
  Mengtao (Martin) Yuan authored Feb 12, 2026
```
Signed-off-by: Martin Yuan <myuan@meta.com>
Co-authored-by: Martin Yuan <myuan@meta.com>
```
  9ea1f598
12 Feb, 2026 1 commit
- [Kernel] Support Flashinfer trtllm fused MoE non gated FP8 & NVFP4 (#33506) · f120bd42
  amitz-nv authored Feb 12, 2026
```
Signed-off-by: amitz-nv <203509407+amitz-nv@users.noreply.github.com>
```
  f120bd42