Commits · 2b465570e6dd327e8422ef9c87e9b2b1454ceaed · OpenDAS / vllm_cscc

30 Jan, 2026 14 commits
- [Hardware][SM100] Add TRTLLM Kernel for INT4 W4A16 Kernel. (#32437) · c3a9752b
  Pavani Majety authored Jan 30, 2026
```
Signed-off-by: Pavani Majety <pmajety@nvidia.com>
```
  c3a9752b
- [Quantization][ROCm] Fix MoE weight loading to be robust (Qwen3_MoE/Qwen3_next... · f451b455
  xuebwang-amd authored Jan 31, 2026
```
[Quantization][ROCm] Fix MoE weight loading to be robust (Qwen3_MoE/Qwen3_next as example models) (#33173)
Signed-off-by: xuebwang-amd <xuebwang@amd.com>
```
  f451b455
- fix QERL attention import path (#33432) · 3f96fcf6
  Vasiliy Kuznetsov authored Jan 30, 2026
```
Signed-off-by: vasiliy <vasiliy@fb.com>
Signed-off-by: Michael Goin <mgoin64@gmail.com>
Co-authored-by: Michael Goin <mgoin64@gmail.com>
```
  3f96fcf6
- Support FP8 block quant for CompressedTensorsW8A16Fp8 (#33280) · fd0e3772
  Michael Goin authored Jan 30, 2026
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  fd0e3772
- [QeRL] Layerwise Reloading (#32133) · f857a03f
  Kyle Sayers authored Jan 30, 2026
```
Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>
```
  f857a03f
- Disable Cascade Attention for Batch Invariance (#32561) · 8f5d5120
  Frank Wang authored Jan 30, 2026
```
Signed-off-by: frankwang28 <frank.wbb@hotmail.com>
Signed-off-by: Frank Wang <41319051+frankwang28@users.noreply.github.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
```
  8f5d5120
- [BUGFIX] Pixtral cannot be loaded with --limit-mm-per-prompt 0 (#33406) · 8e2ad97a
  Julien Denize authored Jan 30, 2026
```
Signed-off-by: juliendenize <julien.denize@mistral.ai>
```
  8e2ad97a
- [Realtime API] Adds minimal realtime API based on websockets (#33187) · 10152d21
  Patrick von Platen authored Jan 30, 2026
```
Signed-off-by: Patrick von Platen <patrick.v.platen@gmail.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
```
  10152d21
- [Misc] Replace Optional[X] with X | None syntax (#33332) · 1a7894db
  杨朱 · Kiki authored Jan 30, 2026
```
Signed-off-by: carlory <baofa.fan@daocloud.io>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
```
  1a7894db
- fix: allow LFM2 MoE prefix caching (align) (#33376) · f45870b5
  tianshu-Michael-yu authored Jan 30, 2026
```
Signed-off-by: Tianshu Yu <tianshuyu.formal@gmail.com>
```
  f45870b5
- [model] Add support for openPangu7B-VL (#32449) · ba45bedf
  hujiaxin0 authored Jan 30, 2026
```
Signed-off-by: hujiaxin <524446785@qq.com>
Signed-off-by: Emilie1001 <79921183+Emilie1001@users.noreply.github.com>
Co-authored-by: Emilie1001 <79921183+Emilie1001@users.noreply.github.com>
```
  ba45bedf
- Explicitly set `return_dict` for `apply_chat_template` (#33372) · 9432ed8c
  Harry Mellor authored Jan 30, 2026
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
  9432ed8c
- [Models] Refactor Kimi-K2.5 weight loading (#33346) · 8bfc8d56
  Isotr0py authored Jan 30, 2026
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  8bfc8d56
- [Model][Multimodal] Add explicit MusicFlamingo adapter (#32696) · c46b0cd0
  Wang Haoyu authored Jan 30, 2026
```
Signed-off-by: WangHaoyuuu <mailwhaoyu@gmail.com>
```
  c46b0cd0
29 Jan, 2026 19 commits
- [Bugfix] Enable Triton MoE for FP8 per-tensor dynamic (#33300) · bfb9bdaf
  Michael Goin authored Jan 29, 2026
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  bfb9bdaf
- Add Triton fused MoE config for B200 (Nemotron Nano) (#32804) · 8e2a469b
  danisereb authored Jan 29, 2026
  
  8e2a469b
- [Bugfix][Kernel] Fix negative memory offset in GDN Triton kernel (#33326) · 23591e63
  CarstyYou authored Jan 30, 2026
```
Signed-off-by: CarstyYou <186021327+CarstyYou@users.noreply.github.com>
```
  23591e63
- [NVIDIA] [feat] Integrate flashinfer Trtllmgen bf16 moe (#32954) · 0493d897
  Linda authored Jan 29, 2026
```
Signed-off-by: Linda-Stadter <57756729+Linda-Stadter@users.noreply.github.com>
```
  0493d897
- [Chore] Move `MediaConnector` to `vllm.multimodal.media` (#33324) · 831453fc
  Cyrus Leung authored Jan 30, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  831453fc
- [Bugfix] Fix broken GLM-OCR initialization (#33350) · 5e73e490
  Isotr0py authored Jan 29, 2026
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  5e73e490
- [Backport] [Kimi-K2.5] Replace torch.cuda with current_platform for d… (#33320) · 17b17c06
  sthWrong authored Jan 29, 2026
  
  17b17c06
- [Models] Qwen3-ASR (#33312) · 8b3f0a99
  Roger Wang authored Jan 29, 2026
```
Signed-off-by: Roger Wang <hey@rogerw.io>
```
  8b3f0a99
- [Quantization][Refactor] use platform dict to choose kernel (#33130) · a5aa4d5c
  zofia authored Jan 29, 2026
```
Signed-off-by: Zhu, Zufang <zufang.zhu@intel.com>
Signed-off-by: zofia <110436990+zufangzhu@users.noreply.github.com>
```
  a5aa4d5c
- [Chore] Remove `use_data_parallel` kwargs from ViT implementation (#33310) · 5400014d
  Isotr0py authored Jan 29, 2026
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  5400014d
- Bugfix: Pass router logits dtype in nemotron shared experts (#32669) · e01ff5c0
  amirkl94 authored Jan 29, 2026
```
Signed-off-by: Amir Klein <203507526+amirkl94@users.noreply.github.com>
```
  e01ff5c0
- [PluggableLayer][2/N] Apply PluggableLayer to linear layers (#33152) · 08b1195e
  whx authored Jan 29, 2026
```
Signed-off-by: whx-sjtu <2952154980@qq.com>
```
  08b1195e
- [BugFix] Fix EPLB fail for MoeFP4 model with Marlin backend (#33262) · 53fc1664
  Ilya Markov authored Jan 29, 2026
```
Signed-off-by: ilmarkov <markovilya197@gmail.com>
```
  53fc1664
- [Doc]: fixing multiple typos in diverse files (#33256) · 31b25f65
  Didier Durand authored Jan 29, 2026
```
Signed-off-by: Didier Durand <durand.didier@gmail.com>
Signed-off-by: Didier Durand <2927957+didier-durand@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  31b25f65
- [Bugfix] Fix Qwen3-VL-Reranker load. (#33298) · abb34ac4
  wang.yuqi authored Jan 29, 2026
```
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
Signed-off-by: wang.yuqi <noooop@126.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
```
  abb34ac4
- [Misc][Build] Lazy load cv2 in nemotron_parse.py (#33189) · 9e138cb0
  Kiersten Stokes authored Jan 29, 2026
```
Signed-off-by: kiersten-stokes <kierstenstokes@gmail.com>
```
  9e138cb0
- [Refactor] Define MM data parser in processing info instead of processor itself (#33260) · 51550179
  Cyrus Leung authored Jan 29, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  51550179
- [ez] Delete more torch version checks <= 2.8 (#33288) · 07ea184f
  Angela Yi authored Jan 28, 2026
```
Signed-off-by: angelayi <yiangela7@gmail.com>
```
  07ea184f
- [UX] Remove noisy CT UnquantizedLinearMethod warn (#33273) · 141cd439
  Michael Goin authored Jan 28, 2026
```
Signed-off-by: mgoin <mgoin64@gmail.com>
```
  141cd439
28 Jan, 2026 7 commits

[ez] Remove checks for torch version <= 2.8 (#33209) · 4197168e
Angela Yi authored Jan 28, 2026
```
Signed-off-by: angelayi <yiangela7@gmail.com>
```
4197168e
Use aiter triton fused_add_rmsnorm_pad for gpt-oss (#30976) · 59bcc5b6
Rohan Potdar authored Jan 28, 2026
```
Signed-off-by: Rohan138 <rohanpotdar138@gmail.com>
```
59bcc5b6

[Quantization][Deprecation] Remove Marlin 24 (#32688) · af9b69f9

Robert Shaw authored Jan 28, 2026


Signed-off-by: Robert Shaw <robshaw@redhat.com>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Robert Shaw <robshaw@redhat.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>

af9b69f9

[Quantization][Deprecation] Remove BitBlas (#32683) · 247d1a32

Robert Shaw authored Jan 28, 2026


Signed-off-by: Robert Shaw <robshaw@redhat.com>
Signed-off-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>
Co-authored-by: Robert Shaw <robshaw@redhat.com>

247d1a32

[CI] Enable mypy import following for `vllm/compilation` (#33199) · f1acbd68
Harry Mellor authored Jan 28, 2026
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
f1acbd68

Adds FunAudioChat multimodal audio model support (#2) (#33058) · 36d450e3

ramos authored Jan 28, 2026


Signed-off-by: ramos <49182011+nemoramo@users.noreply.github.com>
Signed-off-by: mayufeng <mayufeng@example.com>
Co-authored-by: mayufeng <mayufeng@example.com>

36d450e3

Don't use `min_pixels`/`max_pixels` from Qwen2VL's processor (#33208) · 35fb0b86
Harry Mellor authored Jan 28, 2026
```
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
```
35fb0b86