Commit Graph
13853 Commits
Author SHA1 Message Date
David Wang 588d1f7bc9 [Feature] Spec V2 DFlash Support (#23000) 2026-06-10 19:27:42 -07:00
ming_wang ce0ff154a5 add mimo best practice (#27665) 2026-06-11 10:15:31 +08:00
Hexq0210 475e9d25bf bugfix for npu mtp graph runner (#27808) 2026-06-11 10:09:50 +08:00
amote-i fc1fee528c [NPU] [DOC] replace <code> with backticks and remove obsolete params (#27677) 2026-06-11 09:33:31 +08:00
zijiexiaandClaude Fable 5 43835b5ba5 docs: cookbook benchmark accuracy labels come from the model config (no engine default) (#27842)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-06-10 18:27:38 -07:00
YanbingJiang 66989a7642 Fix gpt-oss-20b with mxfp4 support for Xeon (#27782) 2026-06-11 09:16:16 +08:00
Jitendra Patil e113d884d0 ci(xpu): checkout deadicated dir for night build (#27703) 2026-06-11 09:09:46 +08:00
fxmarty-amd 99ab90c5b7 [AMD] Restore AMD piecewise CUDA graph support dropped by #23906 (#27811) 2026-06-10 17:43:48 -07:00
Lijuan TangandClaude Opus 4.8 5f913c1135 [Fix] Emulate PDEATHSIG on macOS to prevent orphaned worker processes (#27190)
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-10 17:43:27 -07:00
Yongji Wuandhzh0425 db061e97c0 [Unified] Fix UnifiedRadixCache write_backup issue in write-back mode(#27108)
Co-authored-by: hzh0425 <hzh0425@apache.org>
2026-06-11 08:28:24 +08:00
740305e1d9 [HiCache] Add opt-in LRU eviction to file storage backend (CP-aware) (#26670)
Co-authored-by: Zhangheng <hzh0425@apache.org>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-11 08:21:59 +08:00
Baizhou Zhang 125ef88892 Disable async assert in Nemotron nightly tests (#27838) 2026-06-10 16:28:14 -07:00
Liangsheng Yin 16124fc9b2 [Metrics] Fix fwd_occupancy reading NaN on every decode log line; probe-free base-a (#27836) 2026-06-10 15:42:19 -07:00
Mohammad Miadh Angkad bdf3ef6421 [CI] Fix registered QK Gemma RMSNorm test location (#27839) 2026-06-10 15:21:59 -07:00
jacky.cheng 0da18f8d91 [AMD][Perf] Fuse QK RMSNorm + gate extraction Triton kernel for Qwen3.5 on HIP (#27656) 2026-06-10 14:29:33 -07:00
Baizhou Zhang 3c1b0fb226 [1/n] [CP] Simplify prefill context parallel server args (#27312) 2026-06-10 14:11:38 -07:00
zijiexiaandClaude Fable 5 73d0989d9b docs: make playground issue template model field a free-form input (#27827)
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
2026-06-10 14:04:16 -07:00
Liangsheng Yin 5fefe91289 [Spec] NGRAMWorker on BaseSpecWorker; algo-owned verify-tree shape params (#27799) 2026-06-10 12:38:16 -07:00
zijiexiaandClaude Opus 4.8 99258b2f1e [Docs] Restore right-hand ToC on the DeepSeek-V4 cookbook page (#27830)
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-10 11:58:02 -07:00
shuwennandQiaolin Yu 3600a9ac5f [SPEC] feat: init adaptive spec params from config (#27493)
Co-authored-by: Qiaolin Yu <liin1211@outlook.com>
2026-06-10 18:55:25 +00:00
Kangyan-ZhouandClaude Opus 4.8 21647f1f5d [router] Apply chat template before cache-aware hashing (fix overlap=0 on chat traffic) (#27386)
Co-authored-by: Claude Opus 4.8 <noreply@anthropic.com>
2026-06-10 10:27:54 -07:00
Mohammad Miadh Angkad 276c98c6cf [Docs] Add Kimi-K2.6 NVFP4 and update Kimi-K2.5 cookbook guidance (#27714)
Signed-off-by: Mohammad Miadh Angkad <176301910+mmangkad@users.noreply.github.com>
2026-06-10 10:24:09 -07:00
billishyahaoandHAI 0ae27405d0 [AMD] Support eplb for moriep (#22985)
Co-authored-by: HAI <hixiao@gmail.com>
2026-06-10 10:23:51 -07:00
Mohammad Miadh Angkad 91ff7baa28 [Docs] Add GLM-5.1 NVFP4 to cookbook (#27708)
Signed-off-by: Mohammad Miadh Angkad <176301910+mmangkad@users.noreply.github.com>
2026-06-10 10:23:15 -07:00
Kai-Hsun Chen 8c6bbe0658 [deepseek] Enable DP attention + TBO + shared experts fusion (#27510) 2026-06-10 09:42:27 -07:00
1cf8efdd08 docs: Diffusion Gemma cookbook (#27824)
Co-authored-by: Richard Wang <wangrichard08@gmail.com>
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-10 09:33:01 -07:00
Kangyan-ZhouandClaude Opus 4.8 0c7faf01fb test(sgl-router): cover sticky scale-up no-redistribution e2e (#27742)
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-10 08:54:31 -07:00
Michael 6a16f29af6 [AMD] ci: register 8 framework / unit tests to run on AMD CI (#25939) 2026-06-10 08:41:22 -07:00
502bc89e1b [AMD] Fix DeepSeek V4 Pro c128 state tensor dtype mismatch error and c4_sparse_raw_indices attribute error in cuda graph phase (#27529)
Co-authored-by: YC Yen-Ching Tseng <yctseng@amd.com>
Co-authored-by: Thomas Wang <thomawan@amd.com>
2026-06-10 08:30:48 -07:00
JoyFuture 53ed34cb88 Fix MiMo-V2.5-Pro DP-attention dp size in cookbook deployment snippet (#27668) 2026-06-10 22:08:28 +08:00
Yuan Luoandluoyuan.luo 518e35fae7 [KDA] Add CuteDSL Prefill Kernel on SM100 (#27488)
Co-authored-by: luoyuan.luo <luoyuan.luo@antgroup.com>
2026-06-10 21:25:19 +08:00
Mohammad Miadh Angkad 4faaa9ba92 [CI] Fix stale ngram bookkeeping owner sites (#27803) 2026-06-10 04:51:46 -07:00
Thomas Wangandbingxche b0d888a195 [CI] Remove AMD DSv4 Docker publish job (#27795)
Co-authored-by: bingxche <bingxche@amd.com>
2026-06-10 17:50:28 +08:00
111009ea54 [Feature] [Ngram spec] Support ngram spec v2 (#17260)
Co-authored-by: hnyls2002 <lsyincs@gmail.com>
Co-authored-by: Ratish P <114130421+Ratish1@users.noreply.github.com>
2026-06-10 02:46:00 -07:00
ChengYao-amdandgithub-actions[bot] 255843d454 Support for Zyphra zaya1 model (#26347)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2026-06-10 02:44:47 -07:00
DarkSharpnessandClaude b40f365732 [CI] Move misplaced mhc kernel test into test/registered/kernels (#27781)
Co-authored-by: Claude <noreply@anthropic.com>
2026-06-10 01:49:18 -07:00
Cheng Wan 70c71ba183 [NPU] Fix dead patch_model monkey-patch breaking NPU torch.compile capture (#27774) 2026-06-10 01:15:50 -07:00
Ziang Li 01f10acd06 Implement online nvfp4 quantization (#26083) 2026-06-10 00:26:51 -07:00
nbarzilie e76e4959b5 [CI][PD] Add unit tests for nixl backend (#26908) 2026-06-10 14:31:59 +08:00
zijiexiaandClaude Opus 4.8 1a5775a9df [Docs] Remove the legacy release-docs.yml deploy workflow (#27766)
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-09 23:18:41 -07:00
Mick e8a437ef26 [diffusion] doc: update docs architecture (#27767) 2026-06-10 14:18:10 +08:00
Thomas WangandXinyi Song f2bcdb0508 [AMD] Add unified kv attention support in dpsk-v4 (#27380)
Co-authored-by: Xinyi Song <86638975+RolaoDenthu@users.noreply.github.com>
2026-06-09 23:13:37 -07:00
Cheng Wan 95d8a75bc9 Bundle set_kv_buffer write targets into KVWriteLoc (loc + swa_loc) (#27695) 2026-06-09 23:09:51 -07:00
Cheng Wan 758fd4bb9a [SWA] Cache full→SWA out_cache_loc per forward across attention backends (#27617) 2026-06-09 22:57:51 -07:00
Aleksi Vesanto 08ceb96ea5 [diffusion] fix: remove boolean arithmetic guard to fix compiling (#27065) 2026-06-10 13:55:47 +08:00
Bingxu Chen 4704b10d0d [AMD] Update MoRI to v1.2.0 (#27669) 2026-06-09 22:41:53 -07:00
Liangsheng Yin d1895cb60d [Spec] Extract move_accept_tokens_to_target_kvcache into spec_utils (#27764) 2026-06-09 21:55:26 -07:00
2495c02c2c [Refactor] Cuda Graph Runner/Backend Refactor (#23906)
Co-authored-by: BBuf <1182563586@qq.com>
Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com>
Co-authored-by: Lianmin Zheng <lianminzheng@gmail.com>
2026-06-09 21:36:57 -07:00
Mick 56f06278c6 [diffusion] refactor: refactor realtime control state and adapters (#27698) 2026-06-10 12:27:16 +08:00
Baizhou Zhang 047e5df3b1 Revert "Share BCG output buffers across capture sizes" (#27758) 2026-06-09 20:51:40 -07:00