Commits · e9c83cdc51f7d9fc642599057ec490e291ea7be3 · OpenDAS / vllm_cscc · GitLab

20 Jan, 2026 5 commits
- [Model Runner V2] Skip kernel launch for penalties & logit_bias (#32634) · e9c83cdc
  Woosuk Kwon authored Jan 19, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  e9c83cdc
- [Model] Use context managers for encoder- and LM-only mode (#32605) · 4753f3bf
  Cyrus Leung authored Jan 20, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  4753f3bf
- [Model Runner V2] Decouple temperature from penalties (#32629) · 6c01ffb8
  Woosuk Kwon authored Jan 19, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  6c01ffb8
- [Model Runner V2] Refactor get_cudagraph_and_dp_padding (#32625) · 7b7cdce9
  Woosuk Kwon authored Jan 19, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  7b7cdce9
- [Model Runner V2] Initialized communication buffer for DP (#32624) · 05dc4bfa
  Woosuk Kwon authored Jan 19, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  05dc4bfa
19 Jan, 2026 6 commits
- [Model Runner V2] Refactor `dummy_run` (#32533) · 43fada53
  Woosuk Kwon authored Jan 19, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  43fada53
- feat: spec decode with draft models (#24322) · 4a5299c9
  Tomas Ruiz authored Jan 19, 2026
```
Signed-off-by: Tomas Ruiz <tomas.ruiz.te@gmail.com>
```
  4a5299c9
- [Core] Whisper support `torch.compile` (#30385) · 74c583bc
  Nicolò Lucchesi authored Jan 19, 2026
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  74c583bc
- [Model Runner V2] Refactor `update_states` (#32562) · 9a1f16da
  Woosuk Kwon authored Jan 18, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  9a1f16da
- [Model Runner V2] Support VLM (#32546) · bb1848cd
  Woosuk Kwon authored Jan 18, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  bb1848cd
- [BUGFIX] Fix degenerate strides in TRTLLM query tensors for FlashInfer... · 6101a26d
  Vadim Gimpelson authored Jan 19, 2026
```
[BUGFIX]  Fix degenerate strides in TRTLLM query tensors for FlashInfer backend. Fixes issue #32353 (#32417)
Signed-off-by: Vadim Gimpelson <vadim.gimpelson@gmail.com>
```
  6101a26d
18 Jan, 2026 5 commits
- [Refactor] Remove unused file `pallas_kv_cache_update.py` (#32433) · 16de822c
  Wentao Ye authored Jan 18, 2026
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  16de822c
- [Model] Support Step1 Model (#32511) · c826c72a
  Li Xie authored Jan 18, 2026
```
Signed-off-by: xieli <xieli@stepfun.com>
```
  c826c72a
- [Model Runner V2] Minor optimization for eagle input processing (#32535) · 963dc0b8
  Woosuk Kwon authored Jan 17, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  963dc0b8
- [Performance] Improve Triton prefill attention kernel's performance (#32403) · 8cc26acd
  Isotr0py authored Jan 18, 2026
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  8cc26acd
- [Model Runner V2] Move mrope_positions buffer to MRopeState (#32532) · 4147910f
  Woosuk Kwon authored Jan 17, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  4147910f
17 Jan, 2026 1 commit
- [Misc] Fix typo: seperator -> separator in flashmla_sparse.py (#32411) · 2b99f210
  Guofang.Tang authored Jan 17, 2026
```
Signed-off-by: Guofang Tang <tinggofun@gmail.com>
Co-authored-by: Guofang Tang <tinggofun@gmail.com>
```
  2b99f210
16 Jan, 2026 2 commits
- [TPU][Core] Enable Pipeline Parallelism on TPU backend (#28506) · 484e22bc
  Chenyaaang authored Jan 16, 2026
```
Signed-off-by: Chenyaaang <chenyangli@google.com>
```
  484e22bc
- [Bugfix] Refactor to support DP parallel in R3 (#32306) · 03da3b52
  Hongxin Xu authored Jan 16, 2026
```
Signed-off-by: xhx1022 <1737006628@qq.com>
Co-authored-by: arlenxu <arlenxu@tencent.com>
```
  03da3b52
15 Jan, 2026 9 commits
- [Attention][AMD] Make flash-attn optional (#30361) · 04741337
  Matthias Gehre authored Jan 15, 2026
```
Signed-off-by: Matthias Gehre <matthias.gehre@amd.com>
```
  04741337
- [Feature] Support async scheduling + PP (#32359) · b34474bf
  Wentao Ye authored Jan 15, 2026
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  b34474bf
- [Model Runner V2] Support FlashInfer backend & Fix CUDA Graph bug [1/2] (#32348) · 6218034d
  Woosuk Kwon authored Jan 15, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  6218034d
- [ROCm][Bugfix] Disable hip sampler to fix deepseek's accuracy issue on ROCm (#32413) · 77c16df3
  Pleaplusone authored Jan 16, 2026
```
Signed-off-by: ganyi <ygan@amd.com>
```
  77c16df3
- [ROCm][Perf] Enable shuffle kv cache layout and assembly paged attention... · 130d6c95
  Pleaplusone authored Jan 15, 2026
```
[ROCm][Perf] Enable shuffle kv cache layout and assembly paged attention kernel for `AiterFlashAttentionBackend` (#29887)
Signed-off-by: ganyi <ygan@amd.com>
```
  130d6c95
- [2/N] Move cache factories to MM registry (#32382) · cbbae38f
  Cyrus Leung authored Jan 15, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  cbbae38f
- [Bugfix] Strengthen the check of X-data-parallel-rank in Hybrid LB mode (#32314) · 1e584823
  dtc authored Jan 15, 2026
```
Signed-off-by: Tianchen Ding <dtcccc@linux.alibaba.com>
```
  1e584823
- [Bugfix] Fix stale `common_attn_metadata.max_seq_len` in speculative decoding with Eagle (#32312) · e9ec2a72
  Ofir Zafrir authored Jan 15, 2026
```
Signed-off-by: Ofir Zafrir <ofir.zafrir@intel.com>
```
  e9ec2a72
- [code clean] remove duplicate check (#32376) · 9d7ae3fc
  Ning Xie authored Jan 15, 2026
```
Signed-off-by: Andy Xie <andy.xning@gmail.com>
```
  9d7ae3fc
14 Jan, 2026 7 commits
- [BugFix] Assign page_size_padded when unifying kv cache spec. (#32283) · 66652e80
  Lumosis authored Jan 14, 2026
```
Signed-off-by: Lihao Ran <imlihao.ran@gmail.com>
```
  66652e80
- [Bugfix][ROCm][performance] Resolve the performance regression issue of the... · e27078ea
  vllmellm authored Jan 15, 2026
```
[Bugfix][ROCm][performance] Resolve the performance regression issue of the Qwen3-Next-80B-A3B-Thinking under rocm_atten (#32336)
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
```
  e27078ea
- [1/N] Reorganize multimodal processing code (#32327) · 9ea07b41
  Cyrus Leung authored Jan 14, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  9ea07b41
- [Misc] Make mem utils can be reused by other platforms (#32322) · ce094624
  Shanshan Shen authored Jan 14, 2026
```
Signed-off-by: shen-shanshan <467638484@qq.com>
```
  ce094624
- [Refactor] Move top-level dummy data generation to registry (#32310) · 90db5b31
  Cyrus Leung authored Jan 14, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  90db5b31
- [misc] Remove is_torch_equal_or_newer(2.4) cases (#32296) · 79336380
  Angela Yi authored Jan 13, 2026
```
Signed-off-by: angelayi <yiangela7@gmail.com>
```
  79336380
- [Model Runner V2] Refactor Sampler (#32245) · 90c08369
  Woosuk Kwon authored Jan 13, 2026
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
  90c08369
13 Jan, 2026 5 commits
- [4/N][Attention] Move MLA common to model_executor (#32060) · 2263d44b
  Matthew Bonanni authored Jan 13, 2026
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
```
  2263d44b
- [6/N][Attention] Move utils to more appropriate locations (#32215) · 98f60e5a
  Matthew Bonanni authored Jan 13, 2026
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  98f60e5a
- [Quantization] fix: overflow with static per-tensor scaling (#29867) · a5bbbd2f
  Mickaël Seznec authored Jan 13, 2026
```
Signed-off-by: Mickael Seznec <mickael@mistral.ai>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>
```
  a5bbbd2f
- [Refactor] Remove `get_encoder_dummy_data` (#32241) · eb28e806
  Cyrus Leung authored Jan 13, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  eb28e806
- [Perf] Optimize requests abort (#32211) · 2a719e08
  Wentao Ye authored Jan 12, 2026
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  2a719e08