Commits · 879a06579ea8a057c250e9b6cc4e632dabd87d2e · OpenDAS / vllm_cscc

01 Nov, 2025 4 commits

[CI/Build] Bump transformers version (#27528) · 879a0657

Cyrus Leung authored Nov 01, 2025


Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>

879a0657

[Multimodal][XPU]Enable vision attn backend for xpu platform (#27525) · 7e2729b5

Yan Ma authored Nov 01, 2025


Signed-off-by: Yan Ma <yan.ma@intel.com>
Signed-off-by: Kunshang Ji <kunshang.ji@intel.com>
Co-authored-by: Yejing Lai <yejing.lai@intel.com>
Co-authored-by: Guancheng Fu <110874468+gc-fu@users.noreply.github.com>
Co-authored-by: Kunshang Ji <kunshang.ji@intel.com>

7e2729b5

[Bugfix] Fix KDA output (#27905) · 3a5de7d2
Jee Jee Li authored Nov 01, 2025
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
3a5de7d2
[Kernel] Enable FusedMoEModularKernel support bias (#27754) · bc4486d6
Jee Jee Li authored Nov 01, 2025
```
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
```
bc4486d6

31 Oct, 2025 5 commits
- Flashinfer_CUTLASS_MOE fuses quantization for TP (#27223) · fc16f1c4
  Shu Wang authored Oct 31, 2025
```
Signed-off-by: Shu Wang. <shuw@nvidia.com>
```
  fc16f1c4
- fix incorrect type annotation in KimiMLP (#27885) · bc306fe5
  ZiTian Zhao authored Nov 01, 2025
```
Signed-off-by: zitian.zhao <zitian.zhao@tencentmusic.com>
```
  bc306fe5
- [Bugfix] Fix broken MRoPE for GLM-4.1V/GLM-4.5V (#27860) · 7e06c40e
  Isotr0py authored Nov 01, 2025
```
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  7e06c40e
- [Perf] Decouple torch op from GDA to leverage torch.compile (#27871) · 3857eb87
  Jiangyun Zhu authored Oct 31, 2025
```
Signed-off-by: zjy0516 <riverclouds.zhu@qq.com>
```
  3857eb87
- [Kimi-Linear] Correct prefixes and add compatibility to AWQ quants (#27834) · e5ef4dfc
  toncao authored Oct 31, 2025
```
Signed-off-by: toncao <cpatonn@gmail.com>
Co-authored-by: toncao <cpatonn@gmail.com>
```
  e5ef4dfc
30 Oct, 2025 13 commits

[Feature] Batch invariant torch.compile (#27660) · e7acb200

Paul Zhang authored Oct 30, 2025


Signed-off-by: PaulZhang12 <paulzhan@fb.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>

e7acb200

[Bugfix] Fix 2 precommit issues - (mamba_block_size, kv_cache_config) (#27811) · ab98f655

Tyler Michael Smith authored Oct 30, 2025


Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Signed-off-by: Tyler Michael Smith <tysmith@redhat.com>
Co-authored-by: Nick Hill <nhill@redhat.com>

ab98f655

[Model] Use the same fused_moe configs for all H200 devices (#23642) · 2918c1b4
Roger Meier authored Oct 31, 2025
```
Signed-off-by: Roger Meier <r.meier@siemens.com>
```
2918c1b4
[MTP] Refactor mtp predictor to avoid d2h operation (#27643) · 10042057
Mengqing Cao authored Oct 31, 2025
```
Signed-off-by: MengqingCao <cmq0113@163.com>
```
10042057

[BugFix] Stopgap - Flashinfer Autotuner + GPT-OSS + DP/TP (#27762) · e5e076ca

Varun Sundar Rabindranath authored Oct 30, 2025


Signed-off-by: Varun Sundar Rabindranath <vsundarr@redhat.com>
Co-authored-by: Varun Sundar Rabindranath <vsundarr@redhat.com>

e5e076ca

[Bugfix][CPU] Fix MRoPE dispatch on the CPU backend (#27800) · eebf00cb
Li, Jiang authored Oct 30, 2025
```
Signed-off-by: jiang1.li <jiang1.li@intel.com>
```
eebf00cb

[Model][Ouro] Support Ouro Model (#27794) · 9956aae4

Fan Yin authored Oct 30, 2025


Signed-off-by: yinfan.1024 <yinfan.1024@bytedance.com>
Signed-off-by: youkaichao <youkaichao@gmail.com>
Co-authored-by: yinfan.1024 <yinfan.1024@bytedance.com>
Co-authored-by: youkaichao <youkaichao@gmail.com>
Co-authored-by: Jee Jee Li <pandaleefree@gmail.com>

9956aae4

[Model] Introduce Kimi Linear to vLLM (#27809) · 4e68cc9b

Zhiyuan Li authored Oct 30, 2025


Signed-off-by: lizhiyuan <lizhiyuan@moonshot.cn>
Signed-off-by: Zhiyuan Li <uniartisan2017@gmail.com>

4e68cc9b

[Frontend][Doc][5/N] Improve all pooling task | Polish encode (pooling) api & Document. (#25524) · 4464723f

wang.yuqi authored Oct 30, 2025

Signed-off-by: wang.yuqi <noooop@126.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>

4464723f

[BugFix][VL] Fix FA selection on Qwen2.5-VL (#27790) · e806178d
Zhewen Li authored Oct 30, 2025
```
Signed-off-by: zhewenli <zhewenli@meta.com>
Co-authored-by: Roger Wang <hey@rogerw.io>
```
e806178d

Add more dims for batch invariant shims (#27489) · ded8ada8

Bram Wasti authored Oct 30, 2025

Signed-off-by: Bram Wasti <bwasti@meta.com>
Signed-off-by: Bram Wasti <bwasti@fb.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>

ded8ada8

[Feat] Adds runai distributed streamer (#27230) · 17d055f5

Benjamin Bartels authored Oct 30, 2025


Signed-off-by: bbartels <benjamin@bartels.dev>
Signed-off-by: Benjamin Bartels <benjamin@bartels.dev>
Co-authored-by: omer-dayan <omdayan@nvidia.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>

17d055f5

[XPU][bugfix] fix rope for llama4 and deepseek (#25145) · b798e39f
Yan Ma authored Oct 30, 2025
```
Signed-off-by: Yan Ma <yan.ma@intel.com>
```
b798e39f

29 Oct, 2025 8 commits
- [Temp fix] Disable torch.compile for Qwen2.5 VL's VisionBlock temporarily. (#27760) · 48eb8eba
  Chenheli Hua authored Oct 29, 2025
```
Signed-off-by: Chenheli Hua <huachenheli@outlook.com>
Signed-off-by: Roger Wang <hey@rogerw.io>
Co-authored-by: Roger Wang <hey@rogerw.io>
```
  48eb8eba
- [Bug] Fix DBO IMA issue for DeepEPHT (#27666) · b5d90f74
  Wentao Ye authored Oct 29, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  b5d90f74
- [Bug] Fix DeepEP low latency `assert self.batched_router_logits.size(-1) ==... · fcb1d570
  Wentao Ye authored Oct 29, 2025
```
[Bug] Fix DeepEP low latency `assert self.batched_router_logits.size(-1) == full_router_logits.size(-1)` Bug (#27682)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  fcb1d570
- [FIXBUG] Qwen3VL hallucinations without Contiguous on Torch.SDPA (#27744) · 7568a282
  JartX authored Oct 29, 2025
```
Signed-off-by: JartX <sagformas@epdcenter.es>
Co-authored-by: Lukas Geiger <lukas.geiger94@gmail.com>
```
  7568a282
- Fix MiniMax-M2 rmsnorm precision and remove useless code (#27627) · d6704dd0
  Roger Young authored Oct 29, 2025
```
Signed-off-by: xuebi <xuebi@minimaxi.com>
Co-authored-by: xuebi <xuebi@minimaxi.com>
```
  d6704dd0
- [perf] Enable concurrent execution of "shared_experts" and "selected_experts"... · 8df98c21
  Jiangyun Zhu authored Oct 29, 2025
```
[perf] Enable concurrent execution of "shared_experts" and "selected_experts" in qwen3-next (#27578)
Signed-off-by: zjy0516 <riverclouds.zhu@qq.com>
```
  8df98c21
- [Bugfix] Fix non-contiguous tensor error in `rocm_unquantized_gemm_impl` (#27605) · 8b624950
  Zhewen Li authored Oct 29, 2025
```
Signed-off-by: zhewenli <zhewenli@meta.com>
```
  8b624950
- [Model] Fix Qwen3VL and Qwen3Omni after torch.compile changes (#27705) · 0d8161b0
  Lukas Geiger authored Oct 29, 2025
```
Signed-off-by: Lukas Geiger <lukas.geiger94@gmail.com>
Signed-off-by: Roger Wang <hey@rogerw.io>
Co-authored-by: Roger Wang <hey@rogerw.io>
```
  0d8161b0
28 Oct, 2025 8 commits
- [Misc][qwen2_5_vl][torch.compile] Enable `supports_torch_compile` on generic... · 94666612
  Lucas Kabela authored Oct 28, 2025
```
[Misc][qwen2_5_vl][torch.compile] Enable `supports_torch_compile` on generic nn.Module and demonstrate speedup on Qwen Vision model (#23207)
Signed-off-by: Lucas Kabela <lucaskabela@meta.com>
Signed-off-by: Lucas Kabela <lucasakabela@gmail.com>
```
  94666612
- [Test] Batch Invariant: Unit test using parameterized backend (#27478) · 6afc28a9
  Wentao Ye authored Oct 28, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  6afc28a9
- [FLA] Introduce Kimi Delta Attention(KDA) to VLLM (#27654) · e88bdd60
  Zhiyuan Li authored Oct 28, 2025
```
Signed-off-by: lizhiyuan <lizhiyuan@moonshot.cn>
```
  e88bdd60
- [Hybrid] Add mamba_block_size to Engine Args (#27289) · 05181cc5
  Asaf Joseph Gardin authored Oct 28, 2025
```
Signed-off-by: asafg <39553475+Josephasafg@users.noreply.github.com>
```
  05181cc5
- [Bug] Fix shape issue for eplb expert weights (#27589) · 0484b642
  Wentao Ye authored Oct 28, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>
```
  0484b642
- [Bugfix] In LongRoPE, decide short vs long based on max_model_len (#27431) · 44b5ce95
  Matthew Bonanni authored Oct 28, 2025
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  44b5ce95
- [Bugfix][CPU] Fallback oneDNN linear to torch linear to fix half gemm support... · d34f5fe9
  Li, Jiang authored Oct 28, 2025
```
[Bugfix][CPU] Fallback oneDNN linear to torch linear to fix half gemm support on legecy platforms (#27526)
Signed-off-by: jiang1.li <jiang1.li@intel.com>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>
```
  d34f5fe9
- [Hardware][AMD][Model] Triton MoE tuning configs for GLM-4.6 for MI300X (#27323) · bdb01a38
  Eric Yue authored Oct 28, 2025
```
Signed-off-by: minatoaquaMK2 <jiacheng.yue@foxmail.com>
```
  bdb01a38
27 Oct, 2025 2 commits
- fixing mm placeholder replacement issue with gemma3 (#27538) · 23ad8205
  tingtinggithub authored Oct 27, 2025
```
Signed-off-by: tingtingtang1992 <streamttt@gmail.com>
```
  23ad8205
- [Bugfix][LoRA][FusedMoE] Select MxFP4 Backend based on LoRA Enablement (#27487) · 5d3be3ba
  Varun Sundar Rabindranath authored Oct 27, 2025
```
Signed-off-by: Varun Sundar Rabindranath <vsundarr@redhat.com>
Co-authored-by: Varun Sundar Rabindranath <vsundarr@redhat.com>
```
  5d3be3ba