81 Commits
Author SHA1 Message Date
Jack HuandDouglas Lehr ef38e8d006 [Model][Hardware][AMD][Kernel] Part 2/2 -> Enable e2e QK Norm + RoPE + KV Cache runtime fusion for Qwen3-30B-A3B on ROCM_ATTN (#42755)
Signed-off-by: Jack Hu <Jack.Hu@amd.com>
Signed-off-by: Jack Hu <jingxujaster@hotmail.com>
Signed-off-by: Jack Hu <jack.hu@amd.com>
Co-authored-by: Douglas Lehr <91553416+dllehr-amd@users.noreply.github.com>
2026-10-15 17:10:14 -07:00
Woosuk KwonandClaude Opus 5.5 e1a5086256 [Attention] Remove cascade attention (#61126)
Signed-off-by: Woosuk Kwon <woosuk@inferact.ai>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-12 10:41:07 -07:00
02e55d2af3 [KV Cache] GLM-5.3-Flash: use the generic packed KV layout; CircularBufferSpec tail with speculative ring slots (#57169)
Signed-off-by: Yifan Qiao <yifanqiao@inferact.ai>
Signed-off-by: Dakai An <dakaian108@gmail.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Yifan Qiao <yifanqiao@inferact.ai>
Co-authored-by: OpenAI Codex <noreply@openai.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
2026-10-09 15:48:37 -04:00
Giancarlo Delfin fcf53d240c [Spec Decode] Remove eager metadata rebuild during MTP fused multi-step decode (#58463)
Signed-off-by: Giancarlo Delfin <gdelfin@inferact.ai>
2026-10-06 12:17:30 -04:00
Nicolò LucchesiandLeoyzen ff53f32409 [Attention][MLA] Support fp8_ds_mla KV cache for NoPE-512 models on SM90 (#59246)
Signed-off-by: NickLucche <nicolo.lucchesi@mistral.ai>
Co-authored-by: Leoyzen <leoyzen@gmail.com>
2026-10-05 13:12:00 +00:00
Eugenio "Jay" ZuccarelliClaude Opus 5.5mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
004e37ece2 [Misc] Name each backend and its kernel block sizes in block-size errors (#58557)
Signed-off-by: Eugenio "Jay" Zuccarelli <11176606+jayzuccarelli@users.noreply.github.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
2026-09-28 14:33:18 +08:00
927c87b347 [PCP][DCP] Support DCP target model with non-DCP Dspark (#56723)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-authored-by: Codex <noreply@openai.com>
2026-09-26 18:12:35 -04:00
a4eb3f25d6 [Spec Decode] Enable Gemma4 DSpark adaptive verification with FlashInfer (#57263)
Signed-off-by: zixi-qi <zixi@inferact.ai>
Co-authored-by: OpenAI Codex <noreply@openai.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
2026-09-26 17:01:17 +08:00
Jhao-Ting Chen 4fb767ff0a [Kernel] Resubmit PR 48666 - Gemma4 FP8 KV FA4 head dim 512 backend selection (#53175)
Signed-off-by: Jhao-Ting Chen <jhaotingc@nvidia.com>
2026-09-23 20:29:06 -07:00
b715699df4 [ROCm][Perf] Extend QK-norm/RoPE/KV-cache fusion to MRoPE (#50212)
Signed-off-by: Vorapol Assavasangthong <Vorapol.Assavasangthong@amd.com>
Co-authored-by: Santosh Hiremath <Santosh.Hiremath@amd.com>
Co-authored-by: Douglas Lehr <91553416+dllehr-amd@users.noreply.github.com>
2026-09-23 11:11:57 -05:00
Harry MellorandClaude Opus 5 651a88c09a [CI] Ignore ruff D209, rejoin the docstrings it split, and silence incompatible-rule warnings (#57212)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-16 10:57:24 -07:00
Harry MellorandClaude Opus 5 8c1557a79c Add pydocstyle to the ruff rules (#52136)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Signed-off-by: Harry Mellor <hej.mellor@gmail.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-09-16 16:44:29 +00:00
82a85dc1d2 [Attention] Add Triton/FlashInfer composite for multimodal prefix attention (#56305)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
2026-09-13 14:29:41 +00:00
ebe1dec2da [PCP][DCP] Enable PCP+DCP on sparse-MLA models (#56157)
Signed-off-by: patryk.saffer <patryk.saffer@blc-login-1.blc1.mistralcompute.ai>
Signed-off-by: patryk.saffer <patryk.saffer@blc-login-0.blc1.mistralcompute.ai>
Signed-off-by: PatrykSaffer <patryk.saffer@mistral.ai>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: patryk.saffer <patryk.saffer@blc-login-1.blc1.mistralcompute.ai>
Co-authored-by: patryk.saffer <patryk.saffer@blc-login-0.blc1.mistralcompute.ai>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Patryk Saffer <PatrykSaffer@users.noreply.github.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
2026-09-12 23:24:44 +00:00
Woosuk Kwon 13e221f830 [Perf] Fuse DSV4.1 input metadata preparation with Triton (#56562) 2026-09-12 07:20:04 -07:00
d43bb2f37f [3/N] HiSparse: host-resident sparse-MLA decode hot-buffering (#53781)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
Signed-off-by: Matthew Bonanni <matthew@neuralmagic.com>
Signed-off-by: Nicolò Lucchesi <nicolo.lucchesi@mistral.ai>
Co-authored-by: OpenAI Codex <codex@openai.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Lucas Wilkinson <lwilkinson@neuralmagic.com>
Co-authored-by: Nicolò Lucchesi <nicolo.lucchesi@mistral.ai>
Co-authored-by: OpenAI Codex <noreply@openai.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
2026-09-12 00:36:47 -04:00
Wentao Ye 5fe77aecfc [Deprecation] Deprecate items scheduled for 0.29 (#55353)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
2026-09-11 08:17:02 -07:00
Andreas KaratzasandOpenAI Codex db3814a4f2 [Attention] Require explicit DCP support from attention implementations (#55780)
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
2026-09-08 14:04:27 -07:00
Andreas KaratzasandOpenAI Codex e476556189 [ROCm][CI] Add attention-sink support to ROCm AITER sparse MLA (#54404)
Signed-off-by: Andreas Karatzas <Andreas.Karatzas@amd.com>
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
2026-09-07 09:26:47 +00:00
Jiangyun Zhu 98ed0856f3 [Model] add GLM-5.3-Flash support (#53906) 2026-09-03 09:40:35 -07:00
Lucas Wilkinson 8bdc70ec7b [6/N][KV-Cache Layout Refactor] Standardize KV cache layout (#51718)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
2026-08-21 17:47:49 -07:00
Summer YangandWoosuk Kwon 58e5ee0158 [refactor] consolidate cp attn ops (#52839)
Signed-off-by: Summer Yang <girasoleyang@gmail.com>
Co-authored-by: Woosuk Kwon <woosuk@inferact.ai>
2026-08-19 18:04:11 -07:00
Roger Wang 755492e37d Revert "[Kernel] Gemma-4 FA4 FP8 Kernel" (#52987) 2026-08-19 10:48:46 -07:00
Wei Zhao d1e3eee6fb [Spec decode] Support Kimi-K3 DCP with DSpark (#52188)
Signed-off-by: wzhao18 <wzhao18.sz@gmail.com>
2026-08-17 13:08:45 -07:00
Lucas WilkinsonandClaude 57bd0ed441 [5/N][KV-Cache Layout Refactor] Backend-published KV packing via customize_spec (#51704)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Claude <noreply@anthropic.com>
2026-08-14 09:47:55 -04:00
827a2af806 [Kernel] Gemma-4 FA4 FP8 Kernel (#48666)
Signed-off-by: Qixiang Lin <qixiangl@nvidia.com>
Signed-off-by: Jhao-Ting Chen <jhaotingc@nvidia.com>
Co-authored-by: Qixiang Lin <qixiangl@nvidia.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: Yongye Zhu <zyy1102000@gmail.com>
2026-08-13 17:42:00 -07:00
Lucas WilkinsonOpenAI CodexClaude Opus 5Benjamin Chislettmergify[bot] <37929162+mergify[bot]@users.noreply.github.com>Nick HillOpenAI Codex
7f7a32cfec [Spec Decode] DSpark confidence-scheduled verification (#47808)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Signed-off-by: Benjamin Chislett <chislett.ben@gmail.com>
Signed-off-by: Lucas Wilkinson <wilkinson.lucas@gmail.com>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-authored-by: Benjamin Chislett <chislett.ben@gmail.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: OpenAI Codex <noreply@openai.com>
2026-08-12 12:27:48 -07:00
Nick Hill f97e502969 [Perf] Avoid more GPU<->CPU syncs on the model execution path (#51738)
Signed-off-by: Nick Hill <nickhill123@gmail.com>
2026-08-11 19:30:34 -07:00
Yizhou a311916a29 [MRV2][Spec] Fuse AR speculator multi-step decodes back into one CUDA graph (#46849)
Signed-off-by: Yizhou Liu <liu_yizhou@outlook.com>
2026-08-11 15:06:22 +08:00
Li JuqiMatthew BonanniClaude Opus 5mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
66b3c0e61f [Kernel][Model] Optimize FA4 mm_prefix range lookup (#50294)
Signed-off-by: Juqi Li <2223621784@qq.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
2026-08-05 07:33:49 -07:00
Matthew Bonanni a0cd2b69b3 [Bugfix] Universally align block table width to 128 tokens (#50302)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
2026-07-31 08:52:17 -07:00
Martin Hickey b2fb83e7ff [Attention]: Use KVCacheSpec for AttentionMetadataBuilder type hints (#50148)
Signed-off-by: Martin Hickey <martin.hickey@ie.ibm.com>
2026-07-31 11:20:18 +00:00
28158b2fc3 [ROCm] [BugFix] Fix Quark GLM-5.2 Checkpoint inference: indexer wk per-channel FP8 dequant + missing sparse-MLA metadata fields (#48886)
Signed-off-by: Colin Zeng <Colin.Zeng@amd.com>
Signed-off-by: ColinZ22 <Colin.Zeng@amd.com>
Co-authored-by: TJian <tunjian.tan@embeddedllm.com>
Co-authored-by: fanxingran <xingran.fan@amd.com>
2026-07-27 22:38:22 +00:00
Johnny-LiouClaude Opus 4.8mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>tomeras91Cyrus Leung
866fea2b99 [Kernel] ReplaySSM: cache SSM inputs for faster Mamba2 standard decode (#48018)
Signed-off-by: Johnny-Liou <a897111@gmail.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Co-authored-by: tomeras91 <57313761+tomeras91@users.noreply.github.com>
Co-authored-by: Cyrus Leung <tlleungac@connect.ust.hk>
2026-07-24 09:39:49 -07:00
b6ff8a2f50 [Core] Add MRV2 virtual-batch PCP for MLA (#46570)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
2026-07-19 02:53:15 +00:00
Jack HuandDouglas Lehr 2cab53ddee [Model][Hardware][AMD]: Part 1/2 -> Enable e2e QK Norm + RoPE + KV Cache runtime fusion for Qwen3-30B-A3B on ROCM_AITER_FA, and ROCM_AITER_UNIFIED_ATTN (#42749)
Signed-off-by: Jack Hu <Jack.Hu@amd.com>
Co-authored-by: Douglas Lehr <91553416+dllehr-amd@users.noreply.github.com>
2026-07-16 17:39:04 -05:00
Matthew BonanniandClaude Opus 4.8 c4f5cd60da [1/N] Add dense MHA path for sparse MLA short sequences (#47327)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2026-07-14 00:29:56 +00:00
bea70c7cfc [Attention] Make sliding-window support an explicit backend capability (#48011)
Signed-off-by: NickLucche <nicolo.lucchesi@mistral.ai>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Andreas Karatzas <akaratza@amd.com>
2026-07-13 01:07:56 -07:00
51878e5b6e [2/N][KV-Cache Layout Refactor] Pack K/V into the content dim across attention backends (#44455)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Nicolò Lucchesi <nlucches@redhat.com>
2026-07-11 11:11:16 -04:00
Yan XuandJingyi Yang 95ed0feaa5 DCP supports hybrid attention (#40996)
Signed-off-by: YanXu <yancey.yx@alibaba-inc.com>
Signed-off-by: Jingyi Yang <girasoleyang@gmail.com>
Co-authored-by: Jingyi Yang <girasoleyang@gmail.com>
2026-07-09 21:34:45 -07:00
Wentao Ye f70caef48b [Perf] Cache token_to_req_indices for dsv4, 5x~6x kernel performance improvement (#47474)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
2026-07-06 11:17:46 -04:00
Summer YangandClaude 5b4cb69523 [Bugfix][MLA] Fix LSE log-base mismatch in DCP + FlashInfer MLA decode (#47079)
Signed-off-by: girasoley <girasoleyang@gmail.com>
Co-authored-by: Claude <noreply@anthropic.com>
2026-06-29 19:15:02 -07:00
c6741b2ad4 [Model] Support Unlimited OCR (#46564)
Signed-off-by: Tianyu Guo <guoty9@mail2.sysu.edu.cn>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Roger Wang <hey@rogerw.io>
2026-06-27 23:09:18 -07:00
44d95069e9 Enable DeepSeek V4 and GLM-5.1 on SM120 (#43477)
Signed-off-by: Zihua Wu <13583761+lucifer1004@users.noreply.github.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: Yongye Zhu <zyy1102000@gmail.com>
2026-06-22 11:54:14 -07:00
2cac89f9da [Spec Decode] Support mixed KV page sizes for DFlash (#45181)
Signed-off-by: Alex Steiner <asteiner@nvidia.com>
Signed-off-by: Giancarlo Delfin <gdelfin@inferact.ai>
Signed-off-by: Yifan Qiao <yifanqiao@inferact.ai>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Giancarlo Delfin <gdelfin@inferact.ai>
Co-authored-by: Yifan Qiao <yifanqiao@inferact.ai>
2026-06-21 22:45:14 +08:00
Carl Y eb04c769d3 feat: MLA prefill enable FA4 fp8 output (#43050)
Signed-off-by: Carl You <4531192+carlyou@users.noreply.github.com>
2026-06-16 07:10:59 -07:00
eb28452b10 [Model] Add DiffusionGemma Support (#45163)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Martin Kukla <martin.kukla@cantab.net>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Dipika Sikka <dsikka@redhat.com>
Co-authored-by: NickLucche <nlucches@redhat.com>
Co-authored-by: jiahanc <173873397+jiahanc@users.noreply.github.com>
Co-authored-by: Alec Kohlhoff <134344302+aleckohlhoff@users.noreply.github.com>
Co-authored-by: Porras Huang <20535584+porrashuang@users.noreply.github.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: scoootscooob <167050519+scoootscooob@users.noreply.github.com>
2026-06-11 22:17:35 -07:00
6deb05e0e4 [Core][Model] Gemma4: Unified FA4 for all layers + FlashAttention mm_prefix support (#42175)
Signed-off-by: Luciano Martins <lucianommartins@users.noreply.github.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Luciano Martins <lucianommartins@users.noreply.github.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
2026-06-09 17:45:39 -07:00
Harry Mellor a80af24356 Speed up docs build (#44635)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
2026-06-05 14:51:44 +00:00
Nicolò Lucchesi 5b115bb8a3 [Attention][AMD] Standardize kv layout to blocks first for AMD (#43660)
Signed-off-by: NickLucche <nlucches@redhat.com>
2026-05-28 12:28:50 -05:00