Commits · 656944ac0f95baf8b45978c3a68ed8ebfd49a920 · OpenDAS / vllm_cscc

18 Mar, 2026 1 commit
- x接入mla_cat算子仅在nmz和kvcache-fp8情况下生效，默认关闭，开启需要export VLLM_USE_CAT_MLA=1 · 3bff7958
  yangql authored Mar 18, 2026
  
  3bff7958
17 Mar, 2026 1 commit
- [perf]消除sparse mla build时的拷贝调度空泡 · 46ab154b
  王敏 authored Mar 17, 2026
  
  46ab154b
11 Mar, 2026 1 commit
- mla: 恢复 opt-cat 在 prefill 和 decode 的拼接路由 · ee355d86
  laibao authored Mar 07, 2026
  
  ee355d86
09 Mar, 2026 1 commit
- 适配在bmz上的mla的kvcache_e5m2和e4m3量化的支持 · 0daa00fb
  yangql authored Mar 09, 2026
  
  0daa00fb
03 Mar, 2026 1 commit
- support dsa · d71496bf
  zhuwenwen authored Mar 03, 2026
  
  d71496bf
05 Feb, 2026 1 commit
- sync v0.15.1(ex fused_moe&models) · 899a2db4
  zhuwenwen authored Feb 05, 2026
  
  899a2db4
04 Feb, 2026 1 commit
- update mla interface · 1cb851b0
  zhuwenwen authored Feb 04, 2026
  
  1cb851b0
29 Jan, 2026 1 commit
- fix fa interface and kvcache · 6fc61a0d
  zhuwenwen authored Jan 29, 2026
```
not supported FlashMLASchedMeta
```
  6fc61a0d
23 Jan, 2026 3 commits

fix: Add glm4_moe_lite to MLA detection (#32614) · 586a57ad

Markus / Mark authored Jan 23, 2026


Signed-off-by: marksverdhei <marksverdhei@hotmail.com>
Signed-off-by: Markus / Mark <46672778+marksverdhei@users.noreply.github.com>
Signed-off-by: mgoin <mgoin64@gmail.com>
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Co-authored-by: mgoin <mgoin64@gmail.com>

586a57ad

[Misc] Add `get_name` to missing AttentionBackends (#32698) · 160c6fa3
Nicolò Lucchesi authored Jan 23, 2026
```
Signed-off-by: NickLucche <nlucches@redhat.com>
```
160c6fa3
[CI] Fix mypy for `vllm/v1/structured_output` (#32722) · 7ef58737
Wentao Ye authored Jan 22, 2026
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
7ef58737

22 Jan, 2026 2 commits
- [Bugfix][Attention] Explicitly report support for kv_cache_dtype bfloat16 (#32795) · 955b43a5
  Matthew Bonanni authored Jan 22, 2026
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  955b43a5
- [FlashMLA] Update FlashMLA to expose new arguments (#32810) · 889722f3
  Lucas Wilkinson authored Jan 21, 2026
```
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
```
  889722f3
21 Jan, 2026 2 commits
- [ROCm][Deepseekv3.2] Refactor Sparse Indexer as CustomOp (#29287) · 6c20e89c
  Pleaplusone authored Jan 21, 2026
```
Signed-off-by: ganyi <ygan@amd.com>
```
  6c20e89c
- Update FlashMLA (#32491) · b4f64e5b
  Lucas Wilkinson authored Jan 20, 2026
```
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
```
  b4f64e5b
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 1 commit
- Switch default w8a8 gemm impl to blaslt. · f06d1125
  zhuwenwen authored Jan 16, 2026
```
fix _forward_encoder_attention
remove medusa
set VLLM_PCIE_USE_CUSTOM_ALLREDUCE=1
```
  f06d1125
13 Jan, 2026 2 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

[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

12 Jan, 2026 1 commit
- [3/N][Attention] Move AttentionMetadata-related code from utils.py to backend.py (#32054) · 20228cb8
  Matthew Bonanni authored Jan 12, 2026
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  20228cb8
09 Jan, 2026 1 commit
- [1/N][Attention] Restructure attention: move files (#31916) · 2612ba92
  Matthew Bonanni authored Jan 09, 2026
```
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
```
  2612ba92
07 Jan, 2026 4 commits
- [Chore] Migrate V0 attention utils (#31891) · b665bbc2
  Cyrus Leung authored Jan 07, 2026
```
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
```
  b665bbc2
- [ROCm][AITER] fix wrong argument passed to AITER `flash_attn_varlen_func` (#31880) · 41cfa506
  vllmellm authored Jan 07, 2026
```
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
```
  41cfa506
- [ROCm][AITER] bugfix accuracy regression in ROCM_AITER_TRITON_MLA backend (#31816) · 6409004b
  vllmellm authored Jan 07, 2026
```
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
```
  6409004b
- fixed mypy warnings for files vllm/v1/attention with TEMPORARY workaround (#31465) · 0a2c2dc3
  Jack Yang authored Jan 06, 2026
```
Signed-off-by: Zhuohao Yang <zy242@cornell.edu>
Co-authored-by: Zhuohao Yang <zy242@cornell.edu>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
```
  0a2c2dc3
06 Jan, 2026 3 commits

[Attention][1/n] Remove usage of deprecated `seq_lens_cpu` and... · e0327c9d

Lucas Wilkinson authored Jan 06, 2026


[Attention][1/n] Remove usage of deprecated `seq_lens_cpu` and `num_computed_tokens_cpu` CommonAttentionMetadata properties (#31773)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>

e0327c9d

update lightop import · 05e8b083
zhuwenwen authored Jan 06, 2026

05e8b083

fix weights_not_loaded · 451af742

zhuwenwen authored Jan 06, 2026

update weights_not_loaded and flash_mla_with_kvcache
update paged_mqa_logits

451af742

02 Jan, 2026 1 commit
- [Bugfix][Hardware][AMD] Fix last_page_len calculation in AITER MLA decode (#31282) · 825c2dc1
  Kevin McKay authored Jan 01, 2026
```
Signed-off-by: c0de128 <kevin.mckay@outlook.com>
```
  825c2dc1
31 Dec, 2025 1 commit

[Bug] Fix log issue with `\n` (#31390) · 357d435c

Wentao Ye authored Dec 31, 2025

Signed-off-by: yewentao256 <zhyanwentao@126.com>
Signed-off-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>

357d435c

23 Dec, 2025 2 commits
- update flash_mla_with_kvcache · 9663a03f
  zhuwenwen authored Dec 23, 2025
  
  9663a03f
- Revert "[SM100] Enable fp8 compute for prefill MLA (#30746)" (#31197) · 3e102623
  Pavani Majety authored Dec 22, 2025
```
Signed-off-by: Pavani Majety <pmajety@nvidia.com>
```
  3e102623
22 Dec, 2025 3 commits
- [Bug] Fix `'CutlassMLAImpl' object has no attribute '_workspace_buffer'` (#31173) · 5312a728
  Wentao Ye authored Dec 22, 2025
```
Signed-off-by: yewentao256 <zhyanwentao@126.com>
```
  5312a728
- [SpecDecode] Simplified alternative padded-speculation acceptance rate fix (#29845) · de717476
  Lucas Wilkinson authored Dec 22, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
```
  de717476
- [SM100] Enable fp8 compute for prefill MLA (#30746) · b10f41c8
  Pavani Majety authored Dec 22, 2025
```
Signed-off-by: Pavani Majety <pmajety@nvidia.com>
```
  b10f41c8
18 Dec, 2025 1 commit
- remove unused code · c1c5e4f6
  zhuwenwen authored Dec 18, 2025
  
  c1c5e4f6
17 Dec, 2025 1 commit

Synchronize the modifications from the 12th to the 17th: · b66c8e4b

zhuwenwen authored Dec 17, 2025

修复CompressedTensorsLinearMethod中的w4a16的冲突问题
feat(moe): add Marlin W16A16 fused MoE behind VLLM_USE_MARLIN_W16A16_MOE
replace the fp8_mqa_logits and fp8_paged_mqa_logits interfaces in deepgemm with mqa_logits and paged_mqa_logits from lightop

b66c8e4b

16 Dec, 2025 1 commit
- [ROCm][MTP] Support MTP for AITER MLA backend (#28624) · 9dbbc59b
  Pleaplusone authored Dec 16, 2025
```
Signed-off-by: ganyi <ygan@amd.com>
```
  9dbbc59b
13 Dec, 2025 1 commit

[Feature] Add SM103 (Blackwell Ultra) Support to vLLM (#30484) · 4fa7ce46

Roberto L. Castro authored Dec 13, 2025

Signed-off-by: LopezCastroRoberto <robertol.c510@gmail.com>
Signed-off-by: Roberto L. Castro <38211239+LopezCastroRoberto@users.noreply.github.com>
Co-authored-by: youkaichao <youkaichao@gmail.com>

4fa7ce46

12 Dec, 2025 1 commit
- [Attention] Use sparse prefill kernel for fp8 kv-cache in DeepSeek-v3.2 (#27532) · 3e41992f
  Lucas Wilkinson authored Dec 12, 2025
```
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
```
  3e41992f