anodyine/vllm-omni
1
0 Fork 0
Code Issues 4 Pull Requests 2 Actions Packages Projects Releases Wiki Activity
2,960 Commits 235 Branches 31 Tags
85d1f92ee58de7cf65bcfd9a02fc0a4c3e3dc9d3
Commit Graph
90 Commits
Author SHA1 Message Date
tzhouamandClaude Opus 5 c306445ff2 rebase: migrate diffusion KV cache to the vLLM 0.29 layout model 
vLLM 0.29 removed AttentionSpec.indexes_kv_by_block_stride and reshaped
KVCacheTensor from (size, shared_by) to an explicit layout: layers in L
order plus layer_stride and block_stride. The block-stride decision now
lives on one physical CacheConfig.kv_cache_layout that the engine core
resolves once, and layout.is_block_outermost is the direct successor of
the old per-spec flag.

Diffusion backends subclass the Omni AttentionBackend ABC, so upstream
get_supported_kv_cache_layouts never sees them and the engine core never
resolves a layout for them. Add diffusion_kv/layout.py to pin it before
get_kv_cache_configs runs, defaulting to LBNHC, which reproduces the
indexes_kv_by_block_stride=False that every dense diffusion backend
declared. The backend hook is kept as an enforced contract rather than
deleted: a backend needing block-stride indexing selects a block-outermost
layout, and a contradiction raises instead of mis-addressing K/V.

Workers receive only a KVCacheConfig, so the resolved name travels on it
(KVCacheConfig.kv_cache_layout) and is adopted into each worker CacheConfig
before allocation. Without that, init_kv_cache raises in the worker because
get_resolved_kv_cache_layout has nothing to return.

init_kv_cache also lost attn_groups and cache_dtype in 0.29; the production
caller in model_runner_backend and the test stubs move to the six-argument
form.

Tensor construction goes through one derivation built on the upstream
compute_layout_strides helper, shared by production and tests, so a wrong
stride cannot be masked by a parallel test-only copy.

Also updates the sizing expectation: 0.29 subtracts the null block that
BlockPool permanently holds back before estimating, so sixteen physical
pages now yield fifteen usable ones.

One review point is addressed only in part, deliberately.
build_native_kv_cache_configs now takes indexes_kv_by_block_stride and
propagates a block-outermost layout end to end when it is set, with a test
covering control plane to worker adoption. Automatic discovery of that
requirement from the workers is not added: 0.29 removed the field that used
to carry it, every diffusion backend in tree declares False, and inventing
an executor RPC for a capability nothing requests would ship untestable
plumbing. Until a backend needs it, the seam is explicit and the mismatch
raises at worker init rather than corrupting addressing.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0146Xu5Sgj4vXxLxvLjV5YPj
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-09-10 00:41:53 +08:00
NumberWan 038c9b948f [Bugfix] Resolve #6931 hub FA3 on torch 2.13 via kernels 0.16.1 (#7185) 
Signed-off-by: NumberWan <wantszkin2003@gmail.com>
 2026-09-09 10:20:34 +00:00
Lei KeandKrystalRay 5e4baea64d [Bugfix][NPU] Limit MiniMax H3 modulation grid size (#6794) 
Signed-off-by: KrystalRay <keeleiray@gmail.com>
Co-authored-by: KrystalRay <keeleiray@gmail.com>
 2026-09-08 09:08:46 -07:00
Rahul SteigerandRahul Steiger f080f41e03 [Bugfix][Cosmos3] Auto-pad non-divisible Ulysses GEN sequences (#6918) 
Signed-off-by: Rahul Steiger <rsteiger@nvl72d188-T17.cm.cluster>
Co-authored-by: Rahul Steiger <rsteiger@nvl72d188-T17.cm.cluster>
 2026-09-07 08:58:28 +00:00
XenorynandClaude 039808e0d9 [Perf][Diffusion] Skip the Ulysses length all-gather when shards are padded equal (#5717) 
Signed-off-by: Xenoryn <timberh2000@outlook.com>
Co-authored-by: Claude <noreply@anthropic.com>
 2026-09-06 16:23:30 +08:00
Hongsheng Liu d4096e01fe [Model] Add native MAGI-2 Preview diffusion support (#5918) 
Signed-off-by: Hongsheng Liu <liuhongsheng4@huawei.com>
Signed-off-by: hsliuustc0106 <liuhongsheng4@huawei.com>
 2026-09-05 15:32:30 +08:00
Bo Li fda3af8bbc [Bugfix] Preserve BF16 residual semantics in fused MiniMax-H3 modulation (#6878) 
Signed-off-by: Bo Li <22713281+bobboli@users.noreply.github.com>
 2026-09-04 02:01:34 +08:00
wangyulishunyang12CursorBo Li
 		95a9311696 [CI] Select mirror_hardwares presets via MIRROR_HW (H100/B200) (#5543) 
Signed-off-by: wangyu <410167048@qq.com>
Signed-off-by: lishunyang12 <lishunyang12@163.com>
Signed-off-by: [Your Name] <your.email@example.com>
Signed-off-by: wangyu <53896905+yenuo26@users.noreply.github.com>
Signed-off-by: Bo Li <22713281+bobboli@users.noreply.github.com>
Co-authored-by: lishunyang12 <lishunyang12@163.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: Bo Li <22713281+bobboli@users.noreply.github.com>
 2026-09-02 15:39:49 +00:00
汪志鹏 c59b48b47e [Feature][MiniMax-H3] Complete VSA and Ulysses support (#6909) 
Signed-off-by: princepride <wangzhipeng628@gmail.com>
Signed-off-by: 汪志鹏 <wangzhipeng628@gmail.com>
 2026-09-02 07:19:21 -04:00
XenorynandClaude e51fe6ec1b [Bugfix][Diffusion] Keep the GQA ratio when padding Ulysses heads (#5716) 
Signed-off-by: Xenoryn <timberh2000@outlook.com>
Co-authored-by: Claude <noreply@anthropic.com>
 2026-09-02 08:24:51 +08:00
Wang fuyinzwhzzz0821zwhzzz0821
 		fe4a2a7477 [Diffusion] Add paged KV cache support for HunyuanImage3 DiT on NPU and GPU (#6563) 
Signed-off-by: Acerak01-fy <wfy2003324@163.com>
Signed-off-by: zwhzzz0821 <zwhzzz0821@users.noreply.github.com>
Signed-off-by: zwhzzz0821 <2831474076@qq.com>
Co-authored-by: zwhzzz0821 <zwhzzz0821@users.noreply.github.com>
Co-authored-by: zwhzzz0821 <2831474076@qq.com>
 2026-08-31 10:15:46 +00:00
baonudesifeizhai 8be601ed7f [Diffusion] Add native SymmMem Fast Ulysses transport (#6340) 
Signed-off-by: roG0d <baonudesifeizhai@gmail.com>
 2026-08-30 04:14:21 +00:00
HAAZZZEEEE c50b7de167 [NPU][Diffusion] MiniMax-H3 RainFusion end_step tail fallback (#6037) 
Signed-off-by: HAAZZZEEEE <wang-taicheng@qq.com>
 2026-08-29 18:36:05 +08:00
fan2956andWeiQing Chen 0dc8f45906 [Feat] Minimax H3 support ref and tgt video sparse attention on NPU (#6518) 
Signed-off-by: fan2956 <zhoufan53@huawei.com>
Co-authored-by: WeiQing Chen <40507679+david6666666@users.noreply.github.com>
 2026-08-29 11:10:53 +08:00
vOv 2d4724eb74 [Bugfix][Diffusion] Fix FLASH_ATTN cross-attention key-padding unpad (#5866) 
Signed-off-by: cr-gao <gaochenrui@sjtu.edu.cn>
 2026-08-26 17:20:59 +00:00
Bo Li 51b7565f33 [perf] Add mask-free TRTLLM packed-padding path (#6542) 
Signed-off-by: Bo Li <22713281+bobboli@users.noreply.github.com>
 2026-08-25 07:45:21 +00:00
汪志鹏Claude Opus 5lishunyang12Cursor
 		d150a4fde7 [Feature][MiniMax-H3] Support diffusion continuous batching (#5810) 
Signed-off-by: princepride <wangzhipeng628@gmail.com>
Signed-off-by: lishunyang12 <lishunyang12@163.com>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>
Co-authored-by: lishunyang12 <lishunyang12@163.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
 2026-08-24 21:08:53 +08:00
bjf-frzandHongsheng Liu 072bfc02dd [FEAT]Add FastVideo VSA backend for Wan2.2 (#4820) 
Signed-off-by: bjf-frz <frz123db@gmail.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-08-24 14:48:31 +08:00
Wang fuyinandHongsheng Liu 72ee535fcd [Diffusion] Add native paged KV backend for diffusion workers (#6102) 
Signed-off-by: Acerak01-fy <wfy2003324@163.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-08-22 21:45:45 +08:00
wangyu 273ffce130 [CI/Build] Strengthen pre-commit with markdownlint, SPDX, and policy hooks (#6273) 
Signed-off-by: wangyu <410167048@qq.com>
Signed-off-by: [Your Name] <your.email@example.com>
Signed-off-by: wangyu <53896905+yenuo26@users.noreply.github.com>
 2026-08-20 21:07:04 +08:00
Canlin Guo 14f2c16fb2 [Perf] Fuse MiniMax H3 modulation with FP32 accumulation (#6281) 
Signed-off-by: Canlin Guo <canlinguosdu@gmail.com>
 2026-08-18 20:13:54 -07:00
Yukim1andzwhzzz0821 0a1846aa45 [Diffusion] Add native KV cache initialization and Scheduler-managed block allocation (#6094) 
Signed-off-by: zwhzzz0821 <zwhzzz0821@users.noreply.github.com>
Co-authored-by: zwhzzz0821 <zwhzzz0821@users.noreply.github.com>
 2026-08-18 07:00:24 +00:00
SYLAR ce6ab0ca46 [Model] Add LTX-2.5 support (#6070) 
Signed-off-by: lishunyang12 <lishunyang12@163.com>
 2026-08-15 18:27:23 +08:00
fan2956andFan c27623c29c [Bugfix] rainfusion support all shape video (#6000) 
Signed-off-by: Fan <fan@FandeMacBook-Air.local>
Co-authored-by: Fan <fan@FandeMacBook-Air.local>
 2026-08-11 05:10:50 +08:00
ooooooye 740f45abcd [Perf][Diffusion] MiniMax-H3: opt into packed varlen attention on NPU to eliminate quadratic mask materialization (#5891) 
Signed-off-by: brandneway <gyuan4892@gmail.com>
 2026-08-10 19:47:03 +08:00
NumberWan cfa6d66ed6 [BugFix][Nightly CI] Opt in FA deterministic for Qwen-Image accuracy (#5887) 
Signed-off-by: NumberWan <wantszkin2003@gmail.com>
 2026-08-07 17:15:22 +08:00
RuQing Xu ab459b36c2 [Kernel] Refresh FlashInfer attention; Add quantized attention support (Blackwell QK16/V8) (#5344) 
Signed-off-by: Ruqing Xu <7891482+xrq-phys@users.noreply.github.com>
 2026-08-07 16:14:26 +08:00
linzhenpl07 24ede9e226 [Bugfix] Support GQA/MQA in the ring attention SDPA path (#5255) 
Signed-off-by: linzhenpl07 <linzhenpl07@gmail.com>
 2026-08-07 04:08:33 +00:00
Bo Lilishunyang12lishunyang12
 		d219d93bbb [Attention] Refine TRTLLM attention support for MiniMax H3 (#5779) 
Signed-off-by: Bo Li <22713281+bobboli@users.noreply.github.com>
Signed-off-by: lishunyang12 <lishunyang03@gmail.com>
Signed-off-by: lishunyang12 <lishunyang12@163.com>
Co-authored-by: lishunyang12 <lishunyang03@gmail.com>
Co-authored-by: lishunyang12 <lishunyang12@163.com>
 2026-08-06 09:38:17 +00:00
Zijun HuangandWeiQing Chen 93457f29ff [NPU][Quantization] Add RainFusion attention and INT8 online quantization for MiniMax H3 (#5706) 
Signed-off-by: Zijun Huang <zijunhua@usc.edu>
Co-authored-by: WeiQing Chen <40507679+david6666666@users.noreply.github.com>
 2026-08-06 14:15:21 +08:00
SYLARandlishunyang12 1c2a81f6d8 feat(minimax-h3): enable 2x RTX 4090/5090 with DLO (#5764) 
Signed-off-by: lishunyang12 <lishunyang03@gmail.com>
Signed-off-by: lishunyang12 <lishunyang12@163.com>
Co-authored-by: lishunyang12 <lishunyang03@gmail.com>
 2026-08-06 09:02:29 +08:00
Isotr0py 24741961eb [Model] Add MiniMax H3 diffusion support (#5691) 2026-08-03 14:10:40 +08:00
SYLAR 896f05e7e6 [Kernel] Add typed SAGE for TRTLLM_ATTN (#5509) 
Signed-off-by: lishunyang12 <lishunyang12@163.com>
 2026-07-31 23:11:39 +00:00
SYLARandlishunyang12 7762d664d8 [Attention] Add trtllm diffusion attention backend with Skip-Softmax (#5283) 
Signed-off-by: lishunyang12 <lishunyang03@gmail.com>
Co-authored-by: lishunyang12 <lishunyang03@gmail.com>
 2026-07-27 22:48:11 +08:00
Honghan ZhuClaude Opus 4.7dengyunyang
 		84bbd57e4d [Feature] Add AllGather-KV sequence-parallel attention str… (#4968) 
Signed-off-by: zhuhh97 <zhuhonghan@huawei.com>
Signed-off-by: dph97 <nenbaying@163.com>
Signed-off-by: zhuhh97 <zhh@unknown>
Co-authored-by: Claude Opus 4.7 <noreply@anthropic.com>
Co-authored-by: dengyunyang <584797741@qq.com>
 2026-07-24 14:45:26 +00:00
Samit a5db2d839a [Diffusion] Improve HuggingFace hub kernel loading in flash_attn_hub (#4977) 
Signed-off-by: samithuang <285365963@qq.com>
 2026-07-09 12:10:10 +08:00
Samit b96bc5ac87 Support HuggingFace kernels package for native diffusion attention backends (#4926) 
Signed-off-by: samithuang <285365963@qq.com>
Signed-off-by: Samit <285365963@qq.com>
 2026-07-08 14:34:36 +08:00
bjf-frz d612233bef [Bugfix][CI]Fix HSDP regional compile boundary (#4668) 
Signed-off-by: bjf-frz <frz123db@gmail.com>
 2026-06-25 02:52:52 +00:00
Chendi.Xue f174ee54b0 [XPU] Add sage_attn backend (#3785) 
Signed-off-by: Chendi Xue <chendi.xue@intel.com>
 2026-06-09 19:28:18 +08:00
liqianGao HanWeiQing Chen
 		e35cd213d7 [NPU] [Perf] Adjust flash_attn mask shape for hunyuanvideo1.5 on npu (#3178) 
Signed-off-by: liqian <65649795+vasede@users.noreply.github.com>
Co-authored-by: Gao Han <hgaoaf@connect.ust.hk>
Co-authored-by: WeiQing Chen <40507679+david6666666@users.noreply.github.com>
 2026-06-02 23:23:18 +08:00
MaciejBalaNVSYLARlishunyang12
 		bc794e625f Add Cosmos3 model support (#3454) 
Signed-off-by: Maciej Bala <mbala@nvidia.com>
Signed-off-by: MaciejBalaNV <mbala@nvidia.com>
Signed-off-by: lishunyang12 <lishunyang12@163.com>
Co-authored-by: SYLAR <125541396+lishunyang12@users.noreply.github.com>
Co-authored-by: lishunyang12 <lishunyang12@163.com>
 2026-06-02 04:04:31 +08:00
Aleksi VesantoandHongsheng Liu eba5617c95 [ROCm] Enable AITER backend with ring attention (#3511) 
Signed-off-by: Aleksi Vesanto <avesanto@amd.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-05-29 22:20:29 +08:00
Alex Brooks 68a1d23e4d [Bugfix/Feature] Remove Hardcoded Flash Attention in Bagel & Support GQA in SDPA Backend (#3728) 
Signed-off-by: Alex Brooks <albrooks@redhat.com>
 2026-05-29 10:36:15 +08:00
WeiQing ChenlishunyangGao Han
 		126fb4b4a9 [Blackwell][1/N] Add SageAttention3 diffusion backend on blackwell(GB200/B200/RTX5090/PRO6000/DGX Spark available) (#3015) 
Signed-off-by: lishunyang <lishunyang12@163.com>
Signed-off-by: WeiQing Chen <40507679+david6666666@users.noreply.github.com>
Co-authored-by: lishunyang <lishunyang12@163.com>
Co-authored-by: Gao Han <hgaoaf@connect.ust.hk>
 2026-05-28 14:08:45 +00:00
SYLARfrauttauteffasuZhang Jian
 		373d3c7476 [Blackwell] Add CUDNN_ATTN and FLASHINFER_ATTN backends for diffusion (auto-route) (#3079) 
Signed-off-by: lishunyang <lishunyang12@163.com>
Signed-off-by: Zhang <jianmusings@gmail.com>
Signed-off-by: Zhang Jian <jianmusings@gmail.com>
Co-authored-by: frauttauteffasu <247219357+frauttauteffasu@users.noreply.github.com>
Co-authored-by: Zhang Jian <jianmusings@gmail.com>
 2026-05-28 11:09:25 +08:00
Lancer 109bba8cdf [Bugfix] Treat kv_cache_dtype=auto as unset for ring attention (#3622) 
Signed-off-by: Lancer <maruixiang6688@gmail.com>
 2026-05-20 05:18:57 +00:00
Chendi.Xue 475a4002b0 [XPU] set flash_attn as default diffusion attn backend and fix k_len for cross_attn (#3525) 
Signed-off-by: Chendi Xue <chendi.xue@intel.com>
 2026-05-18 23:58:15 +00:00
lyj-jjjandCursor bbc00f9f86 [BugFix] fix(omni): isolate diffusion KV-cache dtype from vLLM --kv-cache-dtype #3585 (#3596) 
Signed-off-by: lyj-jjj <liuyingjun5@huawei.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
 2026-05-15 11:22:59 +08:00
lyj-jjjCursorgcanlin
 		56ca7dd612 support online FP8 quantization for FA on NPU #2236 (#2640) 
Signed-off-by: lyj-jjj <liuyingjun5@huawei.com>
Signed-off-by: gcanlin <canlinguosdu@gmail.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Co-authored-by: gcanlin <canlinguosdu@gmail.com>
 2026-05-13 11:46:00 +08:00
Canlin Guo 5e263b6929 [BugFix] Rename attention_config to diffusion_attention_config (#3489) 
Signed-off-by: gcanlin <canlinguosdu@gmail.com>
 2026-05-11 11:27:35 +08:00