Commits · abe93bce5952ed8adf90d4b77af6ed3515958620 · OpenDAS / vllm_cscc

10 Dec, 2025 4 commits

[Attention] Make seq_lens_cpu optional in CommonAttentionMetadata to enable... · abe93bce

Lucas Wilkinson authored Dec 09, 2025


[Attention] Make seq_lens_cpu optional in CommonAttentionMetadata to enable true async spec-decode (#29624)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Benjamin Chislett <chislett.ben@gmail.com>

abe93bce

[Bugfix] Fix fp8 DeepGemm compilation issues (#30336) · 2e7035dd
ElizaWszola authored Dec 10, 2025

2e7035dd

[Bugfix] Fix cuda graph sizes when running with speculative decoding (#30330) · 4c2e10ea

PatrykSaffer authored Dec 10, 2025


Signed-off-by: Patryk Saffer <patryk.saffer99@gmail.com>
Signed-off-by: PatrykSaffer <patryk.saffer@mistral.ai>
Co-authored-by: Patryk Saffer <patryk.saffer99@gmail.com>

4c2e10ea

[V1][Spec Decode] Optimize Medusa proposer to avoid GPU-CPU sync (#29723) · 03b5f940
dongbo910220 authored Dec 10, 2025
```
Signed-off-by: dongbo910220 <1275604947@qq.com>
```
03b5f940

09 Dec, 2025 36 commits

Improve wvsplitK tile and balance heristics. (#29937) · 2e7054da
Hashem Hashemi authored Dec 09, 2025
```
Signed-off-by: Hashem Hashemi <hashem.hashemi@amd.com>
```
2e7054da

[Rocm][torch.compile] Adding layernorm + fp8 block quant and silu + fp8 block... · 3c680f4a

Charlie Fu authored Dec 09, 2025


[Rocm][torch.compile] Adding layernorm + fp8 block quant and silu + fp8 block quant for Aiter (#25693)
Signed-off-by: charlifu <charlifu@amd.com>
Signed-off-by: Micah Williamson <micah.williamson@amd.com>
Signed-off-by: Charlie Fu <Charlie.Fu@amd.com>
Co-authored-by: Micah Williamson <micah.williamson@amd.com>
Co-authored-by: wuhuikx <hattie.wu@amd.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Gregory Shtrasberg <156009573+gshtras@users.noreply.github.com>

3c680f4a

[Quantization] FP8 Weight Reloading for Quantized RL Rollout (#28480) · fccd5325
Kyle Sayers authored Dec 09, 2025
```
Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>
```
fccd5325

[MoE][Refactor] Remove most arguments to FusedMoEMethodBase.apply (#29066) · 00e5cbb9

bnellnm authored Dec 09, 2025

Signed-off-by: Bill Nell <bnell@redhat.com>
Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>

00e5cbb9

[CI/Build] Make test_mha_attn.py run on correct platform only and check for... · 7618dc97
rasmith authored Dec 09, 2025
```
[CI/Build] Make test_mha_attn.py run on correct platform only and check for flash_attn_varlen_func in layer.py (#29145)
```
7618dc97

Bump actions/stale from 10.1.0 to 10.1.1 (#30234) · f8dacc66

dependabot[bot] authored Dec 09, 2025


Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

f8dacc66

Bump actions/checkout from 6.0.0 to 6.0.1 (#30233) · 7cab92fd

dependabot[bot] authored Dec 09, 2025


Signed-off-by: dependabot[bot] <support@github.com>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>

7cab92fd

[Model][Quantization] Fix / Add GGUF support for Qwen2 MoE models (#30307) · 73a484ca
Tsukasa OI authored Dec 10, 2025
```
Signed-off-by: Tsukasa OI <floss_llm@irq.a4lg.com>
```
73a484ca
[CI/Test] Fix FP8 per-tensor quant test reference scale shape (#30352) · b37bf51e
Lucas Wilkinson authored Dec 09, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
```
b37bf51e

[BugFix] Fix DeepSeek-R1 hang with DP and MTP (#30119) · 95501a70

Lucas Wilkinson authored Dec 09, 2025


Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Tyler Michael Smith <tyler@neuralmagic.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>

95501a70

[Cleanup] Refactor profiling env vars into a CLI config (#29912) · e858bfe0

Benjamin Chislett authored Dec 09, 2025

Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
Signed-off-by: Benjamin Chislett <chislett.ben@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>

e858bfe0

[Model Runner V2] Support num NaNs in logits (#30187) · d471b2af
Woosuk Kwon authored Dec 09, 2025
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
d471b2af
[Model Runner V2] Fix Triton warning on tl.where (#30355) · 9e6562a3
Woosuk Kwon authored Dec 09, 2025
```
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
```
9e6562a3
[BugFix] Fix non detected failing tests (#30277) · 0b6a8a30
Ilya Markov authored Dec 09, 2025
```
Signed-off-by: ilmarkov <markovilya197@gmail.com>
```
0b6a8a30
Update AMD test definitions (2025-12-08) (#30298) · 804e3468
Alexei-V-Ivanov-AMD authored Dec 09, 2025
```
Signed-off-by: Alexei V. Ivanov <alexei.ivanov@amd.com>
```
804e3468

[Compile] Fix torch warning `TensorFloat32 tensor cores for float32 matrix... · 83319b44

Wentao Ye authored Dec 09, 2025


[Compile] Fix torch warning `TensorFloat32 tensor cores for float32 matrix multiplication available but not enabled` (#29897)
Signed-off-by: yewentao256 <zhyanwentao@126.com>

83319b44

[BugFix] Fix `assert batch_descriptor.num_tokens == num_tokens_padded` (#30173) · 56037dfa
Lucas Wilkinson authored Dec 09, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
```
56037dfa

[Feature] Batch-Invariant Support for FA2 and LoRA (#30018) · 5dcd593b

liuquan authored Dec 09, 2025


Signed-off-by: quanliu <18646313696@163.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>

5dcd593b

[BUGFIX] Mistral tool call parser v11+ (#30332) · 5c213d28
Julien Denize authored Dec 09, 2025
```
Signed-off-by: juliendenize <julien.denize@mistral.ai>
```
5c213d28
[ROCm] Aiter Quant Kernels (#25552) · ee14644b
vllmellm authored Dec 09, 2025
```
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
```
ee14644b
[Bugfix]: Fix glm46 awq marlin moe wna16 compatibility (#30210) · 1166c31c
Dongjie Zou authored Dec 09, 2025
```
Signed-off-by: baonudesifeizhai <baonudesifeizhai@gmail.com>
```
1166c31c
[bugfix][quantization] Fix fp8 per_tensor scale shape (#30257) · 03416ead
haoyangli-amd authored Dec 09, 2025
```
Signed-off-by: Haoyang Li <lihaoyang0109@gmail.com>
```
03416ead
[Structured Output][Reasoning] Improves decoding throughput for models using... · c72ea107
Hubert de La Jonquiere authored Dec 09, 2025
```
[Structured Output][Reasoning] Improves decoding throughput for models using single-token reasoning endings. (#30056)
```
c72ea107
[DCP][Bugfix][CI] Fix accuracy issue of DCP when using FLASH_ATTN_MLA (#30309) · 67475a6e
Jaya Yuan authored Dec 09, 2025
```
Signed-off-by: FENP <yuanyongjie.yyj@antgroup.com>
```
67475a6e

[Bugfix] Qwen 3 VL Embedding loading (#30303) · 9c32df61

wang.yuqi authored Dec 09, 2025

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>

9c32df61

[CI] Fix Flaky test_eagle_max_len Test (#30306) · aeb82b19
Micah Williamson authored Dec 09, 2025
```
Signed-off-by: Micah Williamson <micah.williamson@amd.com>
```
aeb82b19

[Attention] Make `split_decodes_and_prefills(..., require_uniform=True)` support padding (#29644) · aed84691

Lucas Wilkinson authored Dec 09, 2025


Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Benjamin Chislett <chislett.ben@gmail.com>

aed84691

[Misc] Fix safetensors import for safe_open (#30300) · e4605d22
Yongtao Huang authored Dec 09, 2025
```
Signed-off-by: Yongtao Huang <yongtaoh2022@gmail.com>
```
e4605d22

[Model][Quantization] Restore MoE + GGUF models support (incl. Qwen3 MoE) by... · 58d5b3f5

Tsukasa OI authored Dec 09, 2025


[Model][Quantization] Restore MoE + GGUF models support (incl. Qwen3 MoE) by allowing Sideload Parameters (#30116)
Signed-off-by: Tsukasa OI <floss_llm@irq.a4lg.com>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>