70 Commits
Author SHA1 Message Date
c3de0b37e2 [Core][Spec Decode] Replace EAGLE group heuristics with explicit draft-layer tags (#61455)
Signed-off-by: Dakai An <dakaian108@gmail.com>
Signed-off-by: Yifan Qiao <yifanqiao@inferact.ai>
Co-authored-by: Yifan Qiao <yifanqiao@inferact.ai>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
2026-10-17 08:22:33 +00:00
Yan Ma f147949be8 [XPU] limit FA kernel block size to multipleOf 64 (#57297)
Signed-off-by: Yan Ma <yan.ma@intel.com>
2026-10-12 10:25:32 +08:00
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
fd928e5f3c [Bugfix] Check batch-invariance support for every attention backend (#60271)
Signed-off-by: yifanFengg <evanfengff@gmail.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
2026-10-08 11:19:57 -04:00
Harry Mellor 9e0b6fcf77 [CI] Bump Transformers version to 5.18.0 (#59621)
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
2026-10-04 00:42:35 +00:00
f7999d2e44 [Bugfix][HiSparse] Size the KV cache from the groups HiSparse allocates (#59450)
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
2026-10-02 04:11:55 +00:00
dd3bc9c72f [Perf][MRV2] Reuse Mamba/GDN metadata across KV cache groups (#58762)
Signed-off-by: Luca Motz <luca.motz@icloud.com>
Co-authored-by: Xin Yang <xyangx@amazon.com>
Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Codex <noreply@openai.com>
Co-authored-by: Lucas Wilkinson <LucasWilkinson@users.noreply.github.com>
2026-09-28 04:36:58 +00:00
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
waizuichougou a1bf8ac12d [Bugfix][MRV2] Match fast-prefill padding to active LoRA batches (#56456)
Signed-off-by: waizuichougou <2082431897@qq.com>
Signed-off-by: waizuichougou <105572299+waizuichougou@users.noreply.github.com>
2026-09-18 22:13:23 +00:00
Juntian LiuandWoosuk Kwon bf13ecc2f7 [Perf][MRV2] Share token-to-request mappings across KV cache groups (#57102)
Signed-off-by: Juntian Liu <Juntianl777@gmail.com>
Co-authored-by: Woosuk Kwon <woosuk@inferact.ai>
2026-09-18 03:14:36 +00:00
Roberto L. CastroandCodex 186a1e6222 [Warmup] Gemma 4 de-JITification (#55768)
Signed-off-by: LopezCastroRoberto <rocastro@redhat.com>
Co-authored-by: Codex <noreply@openai.com>
2026-09-13 16:47:21 -04:00
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
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
Nick Hill e6cb56337b [Core] MRV2 support for fast-prefill (#56145)
Signed-off-by: Nick Hill <nickhill123@gmail.com>
2026-09-10 10:36:04 -07:00
3f41d102c5 [1/2][Model Runner V2] DBO support, eager mode (#50945)
Signed-off-by: specture724 <specture724@gmail.com>
Signed-off-by: Anjie Hou <149605198+specture724@users.noreply.github.com>
Co-authored-by: jiangkuaixue123 <jiangxiaozhou111@163.com>
Co-authored-by: Lucas Wilkinson <lwilkins@redhat.com>
2026-09-04 14:17:39 -07:00
Andrii SkliarandAndrii Skliar adebc41b7e [Mamba] Add FlashInfer ReplaySSM backend (#52506)
Signed-off-by: Andrii Skliar <askliar@nvidia.com>
Signed-off-by: Andrii Skliar <andreyws96@gmail.com>
Co-authored-by: Andrii Skliar <askliar@nvidia.com>
2026-09-01 10:26:43 -04:00
Ronaldandaoshen02 479eeb32d2 [Bugfix][MRV2] Isolate sleep-mode KV allocations (#53508)
Signed-off-by: Ronald1995 <ronaldautomobile@163.com>
Signed-off-by: aoshen02 <aoshen@inferact.ai>
Co-authored-by: aoshen02 <aoshen@inferact.ai>
2026-08-27 15:08:28 -07:00
Misha Goin 7de96050c5 [Spec Decode] Enable adaptive DSpark on SM100 sparse MLA (#52783)
Signed-off-by: mgoin <mgoin64@gmail.com>
2026-08-25 13:31:15 -04:00
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
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
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
stefankoncarevic 3e372c5ff2 [Bugfix][ROCm] Give KV-first attention blocks their own page in hybrid models (#51837)
Signed-off-by: Stefan Koncarevic <stefan.koncarevic@amd.com>
2026-08-11 16:52:30 -04:00
Soila KavulyaandClaude aef85aed5d [Bugfix][TurboQuant] Add KV quant mode for turboquant (#50533)
Signed-off-by: Soila Kavulya <soila.p.kavulya@intel.com>
Co-authored-by: Claude <noreply@anthropic.com>
2026-07-31 17:42:32 +00:00
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
Nick HillandTaneem Ibrahim fdaa0d9e59 [ModelRunner V2] Support encoder-only attention (#49331)
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Signed-off-by: Taneem Ibrahim <taneem.ibrahim@gmail.com>
Co-authored-by: Taneem Ibrahim <taneem.ibrahim@gmail.com>
2026-07-27 00:38:44 +00:00
6700813f86 [3/N][KV-Cache Layout Refactor] Standardize Mamba cache; drop get_transfer_cache_regions (#44456)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
2026-07-21 09:16:15 +00:00
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
Lucas Wilkinsonandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 04d553f390 [Misc] Use meta tensor for KV cache stride calculation (#47316)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
2026-07-10 23:24:59 -04:00
Lucas Wilkinsonandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 9c18e90f6c [BugFix] Fix packed HND KV cache reshape for FlashAttention (#47314)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
2026-07-10 23:22:39 -04:00
Michael Goin 08dfd68610 [Model] Add LongCat-Flash-Lite (n-gram embedding) (#47857)
Signed-off-by: mgoin <mgoin64@gmail.com>
2026-07-10 07:17:50 -07:00
Lucas WilkinsonandCodex e08a915146 [Bugfix] Preserve tensor causal metadata for grouped attention (#48135)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Codex <codex@openai.com>
2026-07-09 17:57:53 +00:00
Michael GoinandClaude Opus 4.8 0d12618e98 [Spec Decode] Support hybrid (SWA + full attention) DFlash drafters (#47914)
Signed-off-by: mgoin <mgoin64@gmail.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-07-08 11:12:45 -04:00
Andreas Karatzas 2f71b2bd9f [ROCm] Align mixed encoder-decoder KV cache views in V2 runner (#47685)
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
2026-07-07 12:09:22 +08:00
Isotr0pyandNick Hill cc1d020d01 [MRV2] Enable mm prefix bidi attention support on MRV2 (#46942)
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Signed-off-by: Isotr0py <2037008807@qq.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
2026-07-05 14:45:29 +00:00
Lucas Wilkinson fa4321de3d [Bugfix][TurboQuant] Preserve KV cache dtype in backend shape (#47609) 2026-07-05 08:20:48 +00:00
sychen52andmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 67ff0ae30f Support nvfp4 kv with kv-cache-dtype-skip-layers sliding_window (#42890)
Signed-off-by: Shiyang Chen <shiychen@nvidia.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
2026-07-04 02:29:13 +00:00
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
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
01192139bf [DSv4] Pack KV caches into contiguous per-block allocations for DeepSeek V4 (#44577)
Signed-off-by: Tyler Michael Smith <tlrmchlsmth@gmail.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.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>
2026-06-19 12:55:42 -04:00
b8336c3c7c [Bugfix][V1] Split V2 model-runner attention groups on num_heads_q (#45564)
Signed-off-by: Roger Wang <hey@rogerw.io>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
2026-06-14 21:49:46 -07:00
Benjamin ChislettandGiancarlo Delfin 0bae1d3848 [MRV2][Spec Decode] DFlash (#44586)
Signed-off-by: Giancarlo Delfin <gdelfin@inferact.ai>
Signed-off-by: Benjamin Chislett <bchislett@nvidia.com>
Signed-off-by: Benjamin Chislett <chislett.ben@gmail.com>
Co-authored-by: Giancarlo Delfin <gdelfin@inferact.ai>
2026-06-10 08:47:46 -07:00
qizixiandClaude Opus 4.8 d07ad0693b [Bugfix] Use storage_block_size in KV cache reshape for compressed specs (DeepSeek V4) (#43988)
Signed-off-by: zixi-qi <zixi@inferact.ai>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-05-29 11:14:25 -07:00
Mengqing CaoandNick Hill 1223732dda [ModelRunnerV2][Hybrid model] Support kernel block size in hybrid model (#38831)
Signed-off-by: MengqingCao <cmq0113@163.com>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Signed-off-by: Mengqing Cao <cmq0113@163.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
2026-05-28 00:55:55 +00:00
7e33081cee [Attention] Make FlexAttention and FlashAttention use num-blocks first layouts (#42095)
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Matthew Bonanni <mbonanni@redhat.com>
Co-authored-by: Nicolò Lucchesi <nlucches@redhat.com>
2026-05-26 19:55:56 -07:00
Nick Hill 47d4407d7c [Model Runner V2] Support sharing kv cache layers (#35045)
Signed-off-by: Nick Hill <nickhill123@gmail.com>
2026-05-22 22:18:23 +00:00
Nicolò LucchesiandNick Hill fba010dd74 [Bugfix][MRV2] Fix KVCache tensor explicit kernel_block_size dim (#42766)
Signed-off-by: NickLucche <nlucches@redhat.com>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
2026-05-18 20:25:41 -07:00
zhrrr 7a08b34fbf [Model Runner V2] support qwen35 / mamba hybrid model (#35520)
Signed-off-by: zhuhaoran <zhuhaoran.zhr@alibaba-inc.com>
2026-05-07 09:31:05 -07:00
+7 4d51588e23 [Feat] DeepSeek V4 Rebased (#40860)
Signed-off-by: Yifan Qiao <yifanqiao@inferact.ai>
Signed-off-by: Woosuk Kwon <woosuk@inferact.ai>
Signed-off-by: qizixi <zixi@inferact.ai>
Signed-off-by: Jee Jee Li <pandaleefree@gmail.com>
Signed-off-by: Yongye Zhu <zyy1102000@gmail.com>
Co-authored-by: Yongye Zhu <zyy1102000@gmail.com>
Co-authored-by: Yongye Zhu <yongye@inferact.ai>
Co-authored-by: Simon Mo <simon@inferact.ai>
Co-authored-by: Bugen Zhao <i@bugenzhao.com>
Co-authored-by: Giancarlo Delfin <gdelfin@inferact.ai>
Co-authored-by: Jee Jee Li <pandaleefree@gmail.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: Roger Wang <hey@rogerw.io>
Co-authored-by: Roy Wang <yasong.wang@inferact.ai>
Co-authored-by: Woosuk Kwon <woosuk@inferact.ai>
Co-authored-by: youkaichao <youkaichao@gmail.com>
Co-authored-by: Zhewen Li <jerven.vllm@gmail.com>
Co-authored-by: Zijing Liu <liuzijing2014@gmail.com>
Co-authored-by: khluu <khluu000@gmail.com>
Co-authored-by: qizixi <zixi@inferact.ai>
Co-authored-by: Zhewen Li <zhewenli@inferact.ai>
2026-04-26 18:31:08 -07:00
Nick Hill fe85a92e86 [Core] Avoid seq_lens_cpu GPU->CPU sync (#40654)
Signed-off-by: Nick Hill <nickhill123@gmail.com>
2026-04-24 00:35:55 +00:00
zhrrr 8f121f7879 [Model Runner V2] support auto resolve cudagraph mode/sizes based on attn backend (#32936)
Signed-off-by: zhuhaoran <zhuhaoran.zhr@alibaba-inc.com>
2026-04-10 08:27:15 -07:00