Commits · edbc1abd1c206db43dd4d297d773532edbb226a2 · OpenDAS / vllm_cscc

"vscode:/vscode.git/clone" did not exist on "741f4e046bb7e5c5a6093d9fc294865ad7a8e721"

13 Apr, 2026 8 commits

feat: add max_tokens_per_doc in rerank request. (#38827) · edbc1abd

Jesus Federico authored Apr 13, 2026


Signed-off-by: Jesus Federico <jefp@amazon.com>
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: wang.yuqi <yuqi.wang@daocloud.io>

edbc1abd

[Bugfix] Fix GLM tool parser streaming with MTP or stream interval (#39253) · 0e39202c
Flora Feng authored Apr 13, 2026
```
Signed-off-by: sfeng33 <4florafeng@gmail.com>
```
0e39202c
[XPU]Enhance environment collection for Intel XPU and optimize layout (#35698) · 9dd5ee01
sihao_li authored Apr 13, 2026
```
Signed-off-by: sihao.li <sihao.li@intel.com>
Co-authored-by: Kunshang Ji <kunshang.ji@intel.com>
```
9dd5ee01

feat: rename logit_bias/logit_scale to logit_mean/logit_sigma for affine score calibration (#39530) · fa6ae311

Jesus Federico authored Apr 13, 2026


Signed-off-by: Jesus Federico <jefp@amazon.com>
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: wang.yuqi <yuqi.wang@daocloud.io>

fa6ae311

fix(lmcache): correct store for cached requests and num_scheduled_tokens in... · 2a3c32ce

maobaolong authored Apr 13, 2026


fix(lmcache): correct store for cached requests and num_scheduled_tokens in lmcache_mp_connector.py (#39655)
Signed-off-by: baoloongmao <baoloongmao@tencent.com>

2a3c32ce

fused qknorm+rope kernel optimization for SM9.0 (#37376) · 4beeb068

Kaicheng Yang authored Apr 13, 2026


Signed-off-by: EricccYang <yangyang4991@gmail.com>
Signed-off-by: Kaicheng Yang <53411596+EricccYang@users.noreply.github.com>
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>

4beeb068

[compile] Enable AOT compile with batch invariance mode. (#39201) · cae98406

Zhengxu Chen authored Apr 12, 2026


Signed-off-by: zhxchen17 <zhxchen17@fb.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

cae98406

[LoRA] Support dual CUDA streams-Linear Layer (#35721) · 715681c1
Jee Jee Li authored Apr 13, 2026
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
715681c1

12 Apr, 2026 9 commits
- [compile] Bug fix for _decompose_size_nodes (#38360) · f00c5539
  Animesh Jain authored Apr 12, 2026
```
Signed-off-by: Animesh Jain <anijain@umich.edu>
```
  f00c5539
- fix(moe): fix RoutedExpertsCapturer assertion failure with DP>1 and MK path (#37879) · 21fab0a3
  Le Yang authored Apr 12, 2026
  
  21fab0a3
- [KVConnector][NIXL] Organize NIXL connector into its own directory (#39354) · 3244a2eb
  Nicolò Lucchesi authored Apr 12, 2026
```
The number of features supported by the connector has grown substantially
and the `nixl_connector.py` file has accumulated a lot of code. Creates a separate
directory and isolates connector/scheduler code in the hope of improving clarity
and maintainability.

Further refactor of components aimed at improving clarity and simplifying code
will follow soon.
Signed-off-by: NickLucche <nlucches@redhat.com>
```
  3244a2eb
- [Core][Metrics] Remove `vllm:prompt_tokens_recomputed` metric (#38709) · 72ff142c
  Mark McLoughlin authored Apr 12, 2026
```
Signed-off-by: Mark McLoughlin <markmc@redhat.com>
```
  72ff142c
- [Pooling] Disable async scheduling by default for pooling models (#39592) · ee3c0c83
  Nick Hill authored Apr 12, 2026
```
Signed-off-by: Nick Hill <nickhill123@gmail.com>
```
  ee3c0c83
- [HMA] [KVEvent] Enable GPU-side KV events for HMA (#37688) · cc07dad7
  Martin Hickey authored Apr 12, 2026
```
Signed-off-by: Martin Hickey <martin.hickey@ie.ibm.com>
Co-authored-by: Or Ozeri <or@ozery.com>
```
  cc07dad7
- fix(kimi_k25): resolve media_placeholder_token_id from tokenizer (#39344) · 17e787a7
  r266-tech authored Apr 12, 2026
```
Signed-off-by: r266-tech <r266.tech@gmail.com>
Signed-off-by: Roger Wang <hey@rogerw.io>
Co-authored-by: Roger Wang <hey@rogerw.io>
```
  17e787a7
- Support FP8 KVCache on XPU (#37731) · 639402f5
  Xinyu Chen authored Apr 12, 2026
```
Signed-off-by: Xinyu Chen <xinyu1.chen@intel.com>
Co-authored-by: Kunshang Ji <kunshang.ji@intel.com>
```
  639402f5
- [XPU][CT] support per-channel quantization in xpu fp8 linear method (#38316) · 394ff869
  Yan Ma authored Apr 12, 2026
```
Signed-off-by: Yan Ma <yan.ma@intel.com>
```
  394ff869
11 Apr, 2026 6 commits
- [Quant] add CompressedTensorsW8A8Mxfp8 for linear and MoE layers (#38815) · df1e30e7
  EdalatiAli authored Apr 11, 2026
```
Signed-off-by: EdalatiAli <aliedalati@cohere.com>
```
  df1e30e7
- [Gemma4][Bugfix]: Enable Gemma4ForCasualLM to load lora adapters correctly (#38844) · 92feb999
  ShubyM authored Apr 11, 2026
```
Signed-off-by: ShubyM <shubymishra20@gmail.com>
```
  92feb999
- [Bugfix] Fix GDN FLA kernel crashes with NULL_BLOCK_ID=0 CUDA graph padding (#39064) · d4cb783c
  Vibhav Agarwal authored Apr 11, 2026
```
Signed-off-by: Vibhav Agarwal <vibhavagarwal5@gmail.com>
Co-authored-by: vibhav-agarwal <vibhav.agarwal@glance.com>
Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
```
  d4cb783c
- [Misc] Update deprecation warning for --model flag (#39518) · a3e750c0
  z1ying authored Apr 10, 2026
```
Signed-off-by: Ziying Tao <tzzying@outlook.com>
```
  a3e750c0
- [Bugfix] add SupportsMultiModal to Exaone4_5_MTP (#39526) · da72dace
  Lee Yongjun authored Apr 11, 2026
```
Signed-off-by: leeyongjun <jqueen.astro@gmail.com>
```
  da72dace
- Fix the order of _free_encoder_inputs (#38907) · 8d0aabdd
  Tianyu Guo authored Apr 11, 2026
```
Signed-off-by: Tianyu Guo <guoty9@mail2.sysu.edu.cn>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
```
  8d0aabdd
10 Apr, 2026 17 commits
- [Refactor] Move MXFP8 GEMM management into MxFp8LinearKernel (#39205) · 11e2375f
  Michael Goin authored Apr 10, 2026
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  11e2375f
- Add Gemma4 Eagle3 support (#39450) · e7cfd7c5
  Fynn Schmitt-Ulms authored Apr 10, 2026
```
Signed-off-by: Rahul-Tuli <rtuli@redhat.com>
Signed-off-by: Fynn Schmitt-Ulms <fschmitt@redhat.com>
Co-authored-by: Rahul-Tuli <rtuli@redhat.com>
Co-authored-by: Claude <noreply@anthropic.com>
```
  e7cfd7c5
- [Bugfix] Fix FlashInfer crash with kv_cache_dtype_skip_layers (#39002) · e816a881
  yzong-rh authored Apr 10, 2026
```
Signed-off-by: Yifan Zong <yzong@redhat.com>
```
  e816a881
- perf(moe): add tuned fused_moe config for RTX PRO 6000 Blackwell Server Edition (#39183) · 51cfc0e7
  Manu authored Apr 10, 2026
```
Signed-off-by: manu <fortin.emmanuel@gmail.com>
```
  51cfc0e7
- feat: add logit_scale to PoolerConfig for affine score calibration (#39435) · b87575d2
  Jesus Federico authored Apr 10, 2026
```
Signed-off-by: Jesus Federico <jefp@amazon.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
```
  b87575d2
- [Kernel] Porting the TRTLLM minimax_allreduce_rms kernels (#37045) · ecd1ea13
  Jee Jee Li authored Apr 11, 2026
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
  ecd1ea13
- [Model Runner V2] support auto resolve cudagraph mode/sizes based on attn backend (#32936) · 8f121f78
  zhrrr authored Apr 10, 2026
```
Signed-off-by: zhuhaoran <zhuhaoran.zhr@alibaba-inc.com>
```
  8f121f78
- [New Model]: jinaai/jina-reranker-v3 (#38800) · cb5f7501
  wang.yuqi authored Apr 10, 2026
```
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
```
  cb5f7501
- [Model] Implement LoRA support for Qwen3ASRForConditionalGeneration (#37247) · 8d0f908b
  Peter Nguyen authored Apr 10, 2026
```
Signed-off-by: Peter Nguyen <petern0408@gmail.com>
```
  8d0f908b
- [ROCm] Add RDNA 3.5/4 device IDs (gfx1150, gfx1151, gfx1201) (#38455) · c1cc7344
  Ravitez Dondeti authored Apr 10, 2026
```
Signed-off-by: rdondeti <ravitez.dondeti@gmail.com>
Signed-off-by: Ravitez Dondeti <ravitez.dondeti@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
```
  c1cc7344
- [Performance] Remove unnecessary zero-fill of MLA decode output tensor in Aiter backend (#37539) · f976e3b9
  xaguilar-amd authored Apr 10, 2026
```
Signed-off-by: xaguilar-amd <xaguilar@amd.com>
```
  f976e3b9
- [Kernel][Hardware][AMD] Add TritonW4A16LinearKernel for ROCm (#37352) · d468322d
  jatseng-ai authored Apr 10, 2026
```
Signed-off-by: jatseng-ai <jatseng@amd.com>
Signed-off-by: jatseng-ai <janet.tseng@amd.com>
Signed-off-by: Matthias Gehre <matthias.gehre@amd.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Matthias Gehre <matthias.gehre@amd.com>
```
  d468322d
- [model] support FireRedLID (#39290) · 967146e7
  PatchyTIS authored Apr 10, 2026
```
Signed-off-by: PatchouliTaisa <patchychen@tencent.com>
Co-authored-by: PatchouliTaisa <patchychen@tencent.com>
```
  967146e7
- [ZenCPU] Make PT Backport Patch Accessible to vLLM (#38205) · 8e8a3bec
  Lalithnarayan C authored Apr 10, 2026
```
Signed-off-by: Lalithnarayan C <Lalithnarayan.C@amd.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
```
  8e8a3bec
- [PluggableLayer][3/N] Apply PluggableLayer to llm_head and vocab embedding layer (#33465) · 1dfd64c1
  Hexiang Wang authored Apr 10, 2026
```
Signed-off-by: whx-sjtu <2952154980@qq.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
```
  1dfd64c1
- [Bugfix] Fix V1 dummy run writing NaN to KV cache null block (#39444) · ad720aef
  Elvir Crnčević authored Apr 10, 2026
```
Signed-off-by: Elvir Crncevic <elvircrn@gmail.com>
Co-authored-by: Claude Sonnet 4 <noreply@anthropic.com>
```
  ad720aef
- Nemotron Nano VL: Streamline pixel shuffle (#37580) · 270e8a41
  milesial authored Apr 10, 2026
```
Signed-off-by: milesial <milesial@users.noreply.github.com>
```
  270e8a41