xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-07-29 22:21:26 +08:00
Code Issues Packages Projects Releases Wiki Activity
12,281 Commits 155 Branches 89 Tags
Commit Graph

433 Commits

Author SHA1 Message Date
Qiu
968060c15a
[bugfix] correct local_chunk_len for DCP in reorg_kvcache with long context (#28526) 
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
 2025-11-13 11:29:22 -08:00
Matthew Bonanni
f9f3b596f3
[Attention][Bugfix] Fix FA sink support (#28660) 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
 2025-11-13 13:20:01 -05:00
Huamin Li
07a606aa7e
[CI Failure] Fix backend selection for encoder-only models (#28534) 
Signed-off-by: Huamin Li <3ericli@gmail.com>
 2025-11-13 10:11:27 -05:00
Pleaplusone
8da2f28f53
[ROCm][BugFix]Fix get_cu_count in rocm_aiter_fa.py (#28618) 
Signed-off-by: ganyi <ygan@amd.com>
 2025-11-13 14:18:20 +00:00
Pleaplusone
ca00b1bfc6
[ROCm][BugFix] Remove the usage of device_info from aiter (#28383) 
Signed-off-by: ganyi <ygan@amd.com>
 2025-11-12 21:43:42 -08:00
Michael Goin
a543e678b4
[Bugfix] Fix SM100 gpt-oss regression due to faulty attn sink support (#28561) 
Signed-off-by: mgoin <mgoin64@gmail.com>
 2025-11-12 19:40:59 -07:00
Benjamin Chislett
304419576a
[Perf] Refactor cudagraph_support to enable full CUDA graphs for spec decoding with FlashInfer (#28479) 
Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
 2025-11-13 01:56:40 +09:00
Harry Mellor
a742134cc5
Remove deprecated fields from CompilationConfig (#27593) 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
 2025-11-12 16:10:28 +00:00
Andreas Karatzas
9f0247cfa4
VLLM_USE_TRITON_FLASH_ATTN V0 variable deprecation (#27611) 
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
Signed-off-by: Andreas Karatzas <Andreas.Karatzas@amd.com>
 2025-11-11 18:34:36 -08:00
Li, Jiang
7f829be7d3
[CPU] Refactor CPU attention backend (#27954) 
Signed-off-by: jiang1.li <jiang1.li@intel.com>
 2025-11-12 09:43:06 +08:00
Max Hu
412e153df5
[Feature] Allow configuring FlashInfer workspace size (#28269) 
Signed-off-by: Max Hu <hyoung2991@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
 2025-11-11 23:32:20 +00:00
Jie Luo
8c32c6e4b4
[Misc] fix typo in DCP comment (#28389) 
Signed-off-by: Livinfly <luojie3m@gmail.com>
 2025-11-11 10:59:16 -08:00
Matthew Bonanni
b30dfa03c5
[Attention] Refactor CUDA attention backend selection logic (#24794) 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni001@gmail.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
 2025-11-11 07:40:44 -05:00
vllmellm
f080a83511
[RFC][ROCm][AITER] Keep all AITER kernels in _aiter_ops class like _custom_ops and _ipex_ops (#24490) 
Signed-off-by: vllmellm <vllm.ellm@embeddedllm.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
 2025-11-10 08:20:53 -08:00
Lucas Wilkinson
e8697faf03
[V0 deprecation] Remove no longer used get_metadata_cls (#28370) 
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
 2025-11-10 14:32:09 +08:00
zhangsicheng5
2108a571d7
[DCP] Support dcp kv_cache interleave size > 1 (#26696) 
Signed-off-by: zhangsicheng5 <zhangsicheng5@huawei.com>
Signed-off-by: QiuChunshuo <qiuchunshuo@huawei.com>
Signed-off-by: Qiu <qiuchunshuo@huawei.com>
Co-authored-by: QiuChunshuo <qiuchunshuo@huawei.com>
 2025-11-09 04:45:27 +09:00
22quinn
608bb14462
[Attention] Remove max cudagraph size limit of 992 (#27840) 
Signed-off-by: 22quinn <33176974+22quinn@users.noreply.github.com>
 2025-11-07 22:33:27 -08:00
Chen Zhang
c765f0b443
[FlashInfer] Avoid FlashInfer block_size 16 + head_size 256 on blackwell (#27994) 
Signed-off-by: Chen Zhang <zhangch99@outlook.com>
 2025-11-05 09:25:32 -08:00
Pleaplusone
6cae1e5332
[ROCm][MLA] Support block-size > 1 for AITER MLA backend (#27224) 
Signed-off-by: ganyi <ygan@amd.com>
Co-authored-by: wuhuikx <hattie.wu@amd.com>
 2025-11-05 10:43:02 -05:00
Qiu
16b37f3119
[bugfix] fix wrong dcp_local_seq_lens calc (#27518) 
Signed-off-by: Qiu <qiuchunshuo@huawei.com>
 2025-11-05 17:58:13 +08:00
Lucas Wilkinson
d43ad5a757
[BugFix] Fix DCP Assert (AssertionError: DCP not support reorder_batch_threshold > 1 now.) (#28100) 
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
 2025-11-05 14:54:43 +08:00
Kunshang Ji
18b39828d9
[XPU] Add gpt-oss model support for Intel GPU (#27786) 
Signed-off-by: Kunshang Ji <kunshang.ji@intel.com>
 2025-11-05 02:17:23 +00:00
Pleaplusone
dc937175d4
[ROCm][Perf] New design on ROCm AITER MHA backend Implementation (#25763) 
Signed-off-by: ganyi <ygan@amd.com>
 2025-11-04 18:05:33 +00:00
Wentao Ye
7e4be74104
[Bug] Batch invariant: Fix flash attn MLA RuntimeError: scheduler_metadata must have shape (metadata_size) (#27884) 2025-11-04 14:05:55 +08:00
Matthew Bonanni
145c00a4d3
[Bugfix] change FlashMLA reorder_batch_threshold (#27777) 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
 2025-11-03 15:17:10 -05:00
Thomas Parnell
18961c5ea6
[Hybrid] Pass kernel block size to builders (#27753) 
Signed-off-by: Thomas Parnell <tpa@zurich.ibm.com>
 2025-11-03 05:48:03 +00:00
Asaf Joseph Gardin
00b31a36a2
[V1] [Hybrid] Mamba1 Automatic Prefix Caching (#26377) 
Signed-off-by: asafg <39553475+Josephasafg@users.noreply.github.com>
 2025-11-02 04:16:23 -08:00
Matthew Bonanni
f29aeb5a25
Add FLASHINFER_MLA to test_mla_backends and add B200 CI run (#27663) 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
 2025-10-31 11:12:19 -07:00
Isotr0py
3933f18a5e
[Bugfix] Avoid too small block m/n for FlexAttention kernel option (#27853) 
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
 2025-10-31 19:33:12 +08:00
Lucas Wilkinson
b5d70751d8
[BugFix] Reordering extend logic fix (#27739) 
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
 2025-10-29 21:39:34 -07:00
Lucas Wilkinson
141e6a0505
[Misc] Make reorder batch also separate extends (#27367) 
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
 2025-10-28 10:55:10 -07:00
Mohammad Miadh Angkad
a8c02fb5bf
[Bugfix][CI] Fix v1 attention backend tests and add CI coverage (#26597) 
Signed-off-by: Mohammad Miadh Angkad <MAngkad.BSDSBA2027@aim.edu>
Signed-off-by: Mohammad Miadh Angkad <mangkad.bsdsba2027@aim.edu>
Co-authored-by: Ye (Charlotte) Qi <yeq@meta.com>
 2025-10-28 11:42:05 -04:00
Yeshwanth N
71b1c8b667
[Chore]:Extract math and argparse utilities to separate modules (#27188) 
Signed-off-by: Yeshwanth Surya <yeshsurya@gmail.com>
Signed-off-by: Yeshwanth N <yeshsurya@gmail.com>
Signed-off-by: yeshsurya <yeshsurya@gmail.com>
 2025-10-26 04:03:32 -07:00
Ming Yang
0f67d4d962
[Attention] Add MLA prefill backend: trtllm_ragged_attention_deepseek (#26397) 
Signed-off-by: Ming Yang <minos.future@gmail.com>
 2025-10-24 10:24:08 -07:00
fhl2000
284cc92275
[MISC] cudagraph_capture_sizes related improvements (#26016) 
Signed-off-by: fhl <2410591650@qq.com>
Signed-off-by: fhl2000 <63384265+fhl2000@users.noreply.github.com>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
 2025-10-24 05:11:05 -07:00
Matthew Bonanni
dbfbf9f324
[Attention] Fix FlashMLA metadata builder arguments for q_len > 1 (#27368) 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
 2025-10-23 15:58:15 -04:00
Jonathan Chen
ca76486a16
[Chore] Separate out vllm.utils.platform_utils.py (#27374) 
Signed-off-by: Jonathan <chenleejonathan@gmail.com>
 2025-10-23 19:08:06 +00:00
Daisy-Ma-coder
5beacce2ea
[BugFix] bugfix for Flash Attention MLA with full cuda graph IMA following pr-25490 (#27128) 
Signed-off-by: qqma <qqma@amazon.com>
Co-authored-by: qqma <qqma@amazon.com>
 2025-10-22 19:36:39 +00:00
Isotr0py
084a9dae80
[Bugfix] Disable FlexAttention direct block mask building for encoder-only models (#27344) 
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
 2025-10-22 16:39:08 +00:00
Tao He
250fb1b8ea
[Bugfix] fixes the decoding metadata of dense mla's fp8 kvcache. (#27144) 
Signed-off-by: Tao He <linzhu.ht@alibaba-inc.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
 2025-10-21 18:27:03 +00:00
Isotr0py
6ac5e06f7c
[Chore] Clean up pytorch helper functions in vllm.utils (#26908) 
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Signed-off-by: isotr0py <2037008807@qq.com>
 2025-10-18 09:48:22 -07:00
Fadi Arafeh
ab4be40fc5
[fix][cpu] fix prefill attention in CPU attention backend (#27035) 
Signed-off-by: Fadi Arafeh <fadi.arafeh@arm.com>
 2025-10-18 13:30:21 +00:00
Bram Wasti
b2f78cbad4
[small][batch invariance] Rename the env and internal flags to simplify usage (#26855) 
Signed-off-by: Bram Wasti <bwasti@meta.com>
 2025-10-16 21:40:25 +00:00
rongfu.leng
5afd3276df
[Feature] Add process_weights_after_loading to AttentionImpl (#26870) 
Signed-off-by: rongfu.leng <rongfu.leng@daocloud.io>
 2025-10-16 08:02:30 -07:00
Bram Wasti
7d8975de84
Deepseek-v3 Batch Invariant on 8xH100 (#26609) 
Signed-off-by: Bram Wasti <bwasti@meta.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
 2025-10-15 22:06:02 -07:00
Vadim Gimpelson
785d8b6410
[PERF] Qwen3-next MTP speedup (change bool mask indexing to index_select / index_copy to reduce d2h) (#26437) 
Signed-off-by: Vadim Gimpelson <vadim.gimpelson@gmail.com>
 2025-10-16 12:18:31 +08:00
Adrian Abeyta
0a9ef0cfce
Move query quantization to attention layer for Flashinfer & Triton. (#26534) 
Signed-off-by: adabeyta <aabeyta@redhat.com>
Signed-off-by: Adrian Abeyta <aabeyta@redhat.com>
Co-authored-by: Luka Govedič <ProExpertProg@users.noreply.github.com>
 2025-10-15 19:01:38 -04:00
Boyuan Feng
a86b4c58e8
remove attn output view kernel (#26680) 
Signed-off-by: Boyuan Feng <boyuan@meta.com>
Signed-off-by: Boyuan Feng <fby.1994@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
 2025-10-14 22:53:10 +00:00
Matthew Bonanni
82af928c41
[Attention][Spec Decode] FlashMLA spec decode support (#26541) 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
 2025-10-14 19:38:20 +00:00
Jaya Yuan
ea97940d6c
[DCP] Support Decode Context Parallel (DCP) for GQA with FlashAttention (#24864) 
Signed-off-by: yuanyongjie.yyj <yuanyongjie.yyj@antgroup.com>
Signed-off-by: FENP <32334296+FENP@users.noreply.github.com>
Signed-off-by: Jaya Yuan <yuanyongjie.yyj@antgroup.com>
 2025-10-14 13:07:50 +00:00