Commit Graph

Select branches

Hide Pull Requests

bot/update-est-time-20260928

dsv41-pd

dsv41-pd-visioncp

main

Mono Color

• 61b228239e bump sgl-fa4 version to 4.0.5 to loose torch deps (#20378) Rain Jiang 2026-03-11 13:08:09 -07:00
• 006bd44cf9 [deepseekv3.2] fix get_k_and_s_triton kenel for 128K seqlen case bug (#19319) BingjiaWangandabing 2026-03-12 03:56:33 +08:00
• e6a6cd1f0c [Fix] Reset output_ids for requests with input_embeds during retraction (#14110) Kazami Michiru 2026-03-12 03:42:21 +08:00
• 7b44bc923e Relax flaky B200 GSM8K accuracy thresholds (#20304) Alison ShaoandAlison Shao 2026-03-11 12:35:21 -07:00
• d093e70067 [Doc] Add DSA/NSA attention backend to support matrix (#20326) 3 people 2026-03-11 10:40:35 -07:00
• dae5c6cadf [diffusion] doc: add Moore Threads as a supported vendor (#20146) R0CKSTAR 2026-03-12 01:15:15 +08:00
• 05e40922b3 [Doc] Fix wrong link and cmd description (#20365) Thomas 2026-03-11 23:30:45 +08:00
• 6cfc21ba74 [NPU] Bump SGL-Kernel-NPU version to 2026.03.10.rc1 (#20362) Even Zhou 2026-03-11 22:28:21 +08:00
• ed42af99a9 [NPU] [Quantization] w4a4 MoE layer support (#18924) Артем Савкин 2026-03-11 16:52:35 +03:00
• 9991debde3 [Feature] Integrate Elastic NIXL-EP into SGLang (#19248) Yoray ZackandBarak Biber 2026-03-11 11:37:43 +02:00
• 680d9d98e4 Fix cutedsl ci error (#20309) Xiaoyu ZhangandBaizhou Zhang 2026-03-11 16:17:35 +08:00
• 456934fed5 feat: fix update last_receive_tstamp logic for health-check in multi-token-worker mode (#20256) qy-seu 2026-03-11 15:23:22 +08:00
• 61cad15d28 [Utils] Add NetworkAddress abstraction for IPv6-safe address handling (#20306) Liangsheng Yin 2026-03-11 00:07:37 -07:00
• 55e6acf834 [NPU][QwenVL] Support qwen image preprocess on npu (#20189) Kurkur 2026-03-11 15:03:08 +08:00
• 57b093dc34 [NPU]MindSpore backend support eagle3 (#17098) 5 people 2026-03-11 14:11:19 +08:00
• 18cfeabd33 Add SGLANG_SORT_WEIGHT_FILES env var for sequential I/O optimization (#20194) zhaoshang 2026-03-11 14:10:53 +08:00
• 4011065c65 Update CODEOWNERS (#20320) Ke Bao 2026-03-11 14:08:07 +08:00
• 8c8a487468 [diffusion] doc: add diffusion-optimal-perf (#20311) Mick 2026-03-11 12:20:09 +08:00
• c8bbe5010a [diffusion] feat: add AITER Sage attention backend (#20178) Aleksi Vesanto 2026-03-11 06:17:45 +02:00
• 21a0015aa3 [PCG]add piecewise cuda graph support for marlin linear (#20119) xieminghe1andundefined 2026-03-11 10:57:08 +08:00
• fe294904c9 [README] Add GB300 NVL72 inference blog post to News section (#20313) 3 people 2026-03-10 19:31:37 -07:00
• b2dd104ade [Intel GPU] Upgrade pytorch xpu version to 2.10 (#20254) Polisetty V R K Jyothendra Varma 2026-03-11 07:17:25 +05:30
• 16ec4f3a4a Integrate the AddRmsNorm operator (#19939) Kurkur 2026-03-11 09:05:04 +08:00
• 50953aea8d [Scheduler] Unify idle checks into is_fully_idle() and fix weight update test (#20296) Liangsheng Yin 2026-03-10 17:50:23 -07:00
• 7a1ca53805 [AMD] Fix JSON mode streaming test flake on AMD CI (#20293) Michael 2026-03-10 17:24:11 -07:00
• dc4380e33a [AMD] [DeepSeek-OCR-2 Day 0] Enable DeepSeek-OCR-2 on AMD GPUs and add nightly test (#19732) Michael 2026-03-10 17:04:35 -07:00
• eea7f8de41 [CI] Fix B200 runner label for scheduled runs (#20297) Alison ShaoandAlison Shao 2026-03-10 16:00:31 -07:00
• f32882cee5 [CI] Relax Eagle infer_b spec accept length threshold (#20300) Alison ShaoandAlison Shao 2026-03-10 15:40:52 -07:00
• 09a118fafe Support return_logprob for spec v2 (overlap safe) (#19801) 4 people 2026-03-10 15:38:27 -07:00
• 76ee4bb98c [FlashInfer v0.6.4] [RL] Integrate FlashInfer mxfp8 gemm, MoE, and routed MoE (#19537) Ziang Li 2026-03-10 15:37:57 -07:00
• bd460e9565 add logprob related params in bench_serving (#20218) Qiaolin Yu 2026-03-10 15:04:57 -07:00
• db97f193b7 [diffusion][llm] macOS support (#19549) R0CKSTARandMick 2026-03-11 04:11:07 +08:00
• a3d88a247b Enable piecewise-cuda-graph when logprob_start_len = -1 (#19453) Qiaolin Yu 2026-03-10 12:50:57 -07:00
• 031d0a2aad [Qwen-MOE] Fix memory duplication issues in case layers weights are re-assigned during weight loading (#18255) fxmarty-amd 2026-03-10 18:34:56 +01:00
• 11d9c36c2f Replace soundfile+torchaudio with torchcodec AudioDecoder in load_audio (#20190) Xinyuan Tong 2026-03-10 17:26:29 +00:00
• e1f0b3181a [diffusion] fix: adjust convert_hf_to_fp8 to be compatible with more dits (#20281) Mick 2026-03-11 01:21:54 +08:00
• 60cc06297e [4/n jit_kernel restruct] speed up CI tests and add benchmark workflow (#20268) Xiaoyu Zhang 2026-03-10 21:37:41 +08:00
• 5a7c1b8ec6 [NPU] replace swiglu with custom kernel JiaruiChang5268 2026-03-10 21:08:37 +08:00
• 9884957c07 [NPU] Bugfix for qwen35 on NPU (#19756) Hexq0210 2026-03-10 20:03:26 +08:00
• 6ed996bf65 [bugfix] disable share input buffer feature on npu due to accuracy issue (#19507) heziiop 2026-03-10 19:26:46 +08:00
• 51d9d34977 [2/n jit_kernel restruct] unify rotary embedding entrypoints under rope.py (#20247) Xiaoyu Zhang 2026-03-10 17:49:57 +08:00
• 6407891b4f [AMD] Fp8 prefill integration with radix cache path for dpsk models (#20187) Thomas Wang 2026-03-10 17:49:47 +08:00
• 751c454099 Add DeepSeek3.2 and GlmMoeDsa into moe tune (#18876) Yuan Luoandluoyuan.luo 2026-03-10 17:12:58 +08:00
• ac07a6d439 Revert "[Scheduler] Decouple maybe_send_health_check_signal from process_batch_result" (#20259) Liangsheng Yin 2026-03-10 01:58:48 -07:00
• 8cd1de3354 [diffusion] fix: map each prompt to corresponding image in multi-prompt scenario (#20081) LancerandMick 2026-03-10 16:58:21 +08:00
• 2c2003158f [diffusion] fix: fix flux2 lora (#20200) Lancer 2026-03-10 16:57:01 +08:00
• 8517da5d08 [3/n jit_kernel restruct] Clean up benchmark naming and benchmarking helpers (#20250) Xiaoyu Zhang 2026-03-10 16:39:03 +08:00
• c812504b92 [1/n jit_kernel restruct] unify cache usage and clean up naming in ngram_embedding (#20244) Xiaoyu Zhang 2026-03-10 15:53:43 +08:00
• 7cf0551014 Migrate norm kernels to FlashInfer JIT implementation (#18871) Johnsonms 2026-03-09 23:56:07 -07:00
• 69158e9d9f [Bugfix] Skip _mamba_verify_update for idle batch (#20167) Junrong Lin 2026-03-10 14:53:01 +08:00
• 834795adb8 [CI] Refactor PCG related CI (#19994) Yuwei An 2026-03-09 23:49:25 -07:00
• 9b2e5526fb [NPU][Bug fix] context parallel bug fix (#19820) liupeng374 2026-03-10 14:43:49 +08:00
• 5f717913a0 support Kimi-K2.5-w4a8 on ascend khalilzhk 2026-03-10 14:43:27 +08:00
• dadd4dde83 [AMD] Skip the flaky test for lora ci test. (#20175) Jacob0226andYC Tseng 2026-03-10 14:15:30 +08:00
• 1aa56cca37 [AMD] Add wheel support for sglang (#16782) 3 people 2026-03-10 08:42:41 +03:00
• 5a11ae19c1 [CI] fix: notebook ci often OOM (#20199) shuwenn 2026-03-10 13:32:41 +08:00
• 591e61245a [Doc] Add smal table for GEMM backends (#20213) Brayden Zhong 2026-03-10 01:19:57 -04:00
• 208f1428e9 [Scheduler] Decouple maybe_send_health_check_signal from process_batch_result (#20227) Liangsheng YinandClaude Opus 4.6 2026-03-09 21:18:56 -07:00
• 2e7682414b [AMD] Fix Aiter Prebuild When Releasing ROCm720 Image (#20195) Bingxu Chen 2026-03-10 12:02:38 +08:00
• 08d37f6955 [diffusion] fix: add VAE tiling/slicing argument handling for diffusers backend (#17825) sfiisfandMick 2026-03-10 11:38:20 +08:00
• a6ae89fe3c Revert "chore: bump sgl-kernel version to 0.3.21.post1" (#20229) Baizhou Zhang 2026-03-09 20:32:19 -07:00
• ea6ac1d545 Add CODEOWNERS for mooncake transfer engine (#20226) Shangming Cai 2026-03-10 11:22:06 +08:00
• 0a43229be1 Fix mooncake store write/read bandwidth logs (#18294) huangtingwei 2026-03-10 10:54:37 +08:00
• 9610944ae6 [Feature] Add SANA diffusion model (#19234) Mook 2026-03-09 19:09:21 -07:00
• 254d3cee0b [HiCache] Supports Indexer layout for NSATokenToKVPoolHost (#19912) huangtingweiandhzh0425 2026-03-10 09:24:06 +08:00
• 0fd9a57d80 [Doc] Verify and Modify some attention backend specs (#20210) YEJIN KIM 2026-03-10 08:05:52 +09:00
• ca997b7ba9 Add min_p and chat-template kwargs support to run_eval (#19571) Mohammad Miadh Angkad 2026-03-10 05:53:09 +08:00
• be63f982b7 [V32/GLM5] Control the threshold of applying dense attention with an environ (#20062) Baizhou Zhang 2026-03-09 14:36:10 -07:00
• d39ed074cf fix: default FP4 GEMM backend to flashinfer_cudnn on SM120 (Blackwell) (#20047) 3 people 2026-03-09 22:13:08 +01:00
• 61d530e8ac [CI] Fix lint (#20209) Baizhou Zhang 2026-03-09 14:09:59 -07:00
• 3e8abc71ca [Disagg] Skip health check enqueue when PD disagg queues have backlog (#20191) ybyang 2026-03-10 03:58:10 +08:00
• f0153ad225 [AMD][Feature] support fp4 dispatch and fp8 combine in moriep (#19757) AMD-yanfeiwangandDuyi-Wang 2026-03-10 03:52:05 +08:00
• ffb4b6f4c1 [Core] Replace server_args mutation hack with explicit MemoryPoolConfig for draft worker init (#20183) Liangsheng Yin 2026-03-09 11:45:54 -07:00
• ecca8c553d [diffusion] fix: fix diffusers backend issues in diffusion ci gt workflow (#20173) Yuhao Yang 2026-03-10 00:51:48 +08:00
• f947bcbd89 Move stop words to args in send one (#20193) Ke Bao 2026-03-09 23:05:47 +08:00
• 2e444bdced Move stop words to args in send one (#20193) Ke Bao 2026-03-09 23:05:32 +08:00
• eb4ba1bde2 Feature/support longcat flash lite (#17838) 3 people 2026-03-09 23:00:11 +08:00
• 11b76d24dc [NPU] [DLLM]DLLM LLaDA2.x graph mode support with NPU speedup modifications (#18485) 5 people 2026-03-09 22:41:05 +08:00
• d116a8cd94 [Bugfix] Fix load_audio: mono before resample + use torchaudio (#20054) Xinyuan Tong 2026-03-09 11:24:20 +00:00
• 4a757990a1 [VLM] Replace decord with torchcodec for video decoding (#20055) 3 people 2026-03-09 11:23:49 +00:00
• 2d6eb7dff0 [AMD] Add lmsysorg/sglang-daily as A New AMD Daily Image Release Location (#20156) Bingxu Chen 2026-03-09 16:30:08 +08:00
• b719219de9 [ROCm] Use unreg path for aiter custom all-reduce during CUDA graph capture (#20155) Yuzhen Zhou 2026-03-09 01:09:04 -07:00
• cabe171b6c Fix qwen3.5 mtp eplb related issues (#19767) luoyuyan 2026-03-09 16:05:32 +08:00
• c76251f70c Return intermediate Mamba states (#19716) roikoren755 2026-03-09 10:04:36 +02:00
• 484f53c40e [Minor] Add CODEOWNERS for attention/vision.py (#20166) sglang-botandClaude Opus 4.6 2026-03-09 00:27:34 -07:00
• 96724f490c Add auto bind numa node (#15678) Mike Qiu 2026-03-09 14:46:09 +08:00
• 2ef00383ab [Core] Refactor init_memory_pool into composable resolution helpers (#20142) Liangsheng Yin 2026-03-08 21:46:27 -07:00
• c6184b7dc0 Fix EPD OOM by offloading precomputed_embeddings during chunked prefill (#16503) siyu 2026-03-09 11:10:40 +08:00
• 1cb86f5171 [diffusion] CI: fix CI script path and missing server arg in perf baseline generator (#20138) Yuhao Yang 2026-03-09 10:35:21 +08:00
• fc543df289 [NPU] qwen3_vl encoder support graph cen121212 2026-03-09 10:13:35 +08:00
• 8c5ca37aef Batch copy_ with torch._foreach_copy_ (#18558) Kaixi 2026-03-09 03:09:02 +01:00
• 57f28fda90 [diffusion] chore: add diffusion new model skill (#19605) Yuhao Yang 2026-03-09 09:45:23 +08:00
• 3f3eb206fa feat(grpc): add SubscribeKvEvents RPC for KV cache event streaming (#20112) Simo LinandChang Su 2026-03-08 16:00:29 -07:00
• 7105bf3782 [Bug] Fix missing TTFT histogram for single-batch requests (#20122) Liangsheng YinandClaude Opus 4.6 2026-03-08 15:18:51 -07:00
• 7662b8b919 [diffusion] feat: implement upscaling (#19723) Junhao Liu 2026-03-08 11:06:40 -07:00
• b77dd41db0 [diffusion] fix: fix temporary resolution workaround (#20046) xingsy97 2026-03-09 02:05:35 +08:00
• 0ac6c63ae4 [SpecV2-Mamba]: Refactor additional_ratio calculation when init mamba pool (#19660) hzh0425 2026-03-09 00:39:26 +08:00
• 07359efce9 Fix missing clone in hicache (#20130) Ke Bao 2026-03-08 23:21:18 +08:00
• 051427c0a3 [diffusion] benchmark: add SLO metric forinbench_serving (#18907) Junhao Liuandronnie_zheng 2026-03-08 07:35:57 -07:00
• cc73355a1f [Feature] Add SLRU eviction policy & fix RadixCache hit_count bug (#18843) liubiyonggeandzhangheng 2026-03-08 21:30:55 +08:00