Commit Graph

5299 Commits

Author SHA1 Message Date
Ke Bao
60d7beda6b Add split tile size for Triton attention (#10425) 2025-09-14 17:35:49 -07:00
Cheng Wan
2f8ba6fe82 [Fix] MoE: fix w8a8_fp8 MoE and add tests to cover this code path (#10429) 2025-09-14 17:34:28 -07:00
Yineng Zhang
7ce6c10eb6 fix: enable cu124 and cu128 build on main push (#10431) 2025-09-14 16:19:35 -07:00
Yineng Zhang
55025b9282 fix: use latest flashinfer (#10428) 2025-09-14 15:07:14 -07:00
Feng Su
4c21b09074 [Feature] Sglang Tracing: Fine-Grained Tracking for Request Latency - Part 1 (#9962)
Signed-off-by: Feng Su <sufeng@linux.alibaba.com>
Signed-off-by: Huaixin Chang <changhuaixin@linux.alibaba.com>
Signed-off-by: Peng Wang <rocking@linux.alibaba.com>
2025-09-15 02:08:02 +08:00
艾力可
165abeebca Typo: in --enable-custom-logit-processor: agree with cli arg (#10076) 2025-09-14 02:27:09 -07:00
Yingchun Lai
21ca4c3afa [PD metrics] Fix some uncompleted PD related metrics (#8627) 2025-09-14 02:26:58 -07:00
fzyzcjy
e3cf812f7d Fix sgl-kernel + srt CI (#10419) 2025-09-14 01:44:47 -07:00
fzyzcjy
4da5533682 Support profile args in Engine API (#6539) 2025-09-14 01:21:10 -07:00
fzyzcjy
ac964d2e58 Support global scale in addition to per expert scale for cutedsl moe (#10270) 2025-09-14 01:17:00 -07:00
fzyzcjy
fa46e2bd40 Support offloading in fp8 (#9948) 2025-09-14 01:14:28 -07:00
fzyzcjy
b047b553c2 [2/2] Speed up prefill mla attention concat (#10157) 2025-09-14 01:12:04 -07:00
fzyzcjy
a0f844ed5a Let sgl-kernel changes be tested on srt (#10313) 2025-09-14 01:09:17 -07:00
fzyzcjy
2df532ef20 Fix the global scale fix does not support EPLB and improve enabling condition (#10369) 2025-09-14 01:07:47 -07:00
fzyzcjy
abea9250da Auto determine sgl kernel version in blackwell CI (#10318) 2025-09-14 01:06:30 -07:00
Ximingwang-09
b3c977622f Add h200 fused moe config for Qwen3-Next (#10404)
Co-authored-by: 纬杭 <ximing.wxm@antgroup.com>
2025-09-13 23:48:46 -07:00
Yuxuan Zhang
b8347b40b1 Add self.capture_aux_hidden_states For GLM-4.5V (#10228) 2025-09-13 23:31:55 -07:00
fzyzcjy
72dfa96aeb Fix cutlass moe accuracy drop caused by attention UB from DP padding mode (#10414) 2025-09-13 22:29:09 -07:00
lijin23
05b01ef4da fix duplicated logger in eager_utils (#10410) 2025-09-13 22:25:40 -07:00
Liangsheng Yin
55a6e644b0 [Hack] Add pd-disaggregation decode polling interval (#10411) 2025-09-14 10:18:23 +08:00
Liangsheng Yin
6897e06b69 Remove repeatedly lists adding in init_incremental_detokenization (#10412) 2025-09-14 10:05:52 +08:00
Sundara Raman Ramachandran
a360511d7b [Generative Score API] Scoring(Prefill-only) optimizations. (#9748) 2025-09-14 01:57:06 +08:00
Sundara Raman Ramachandran
94d0f656fb [Performance] Dynamic Batch Tokenizer (#9382) 2025-09-14 01:56:04 +08:00
kk
eca59f96c3 Update ROCm docker image to add sgl-router support (#10406)
Co-authored-by: Colin Wang <kangwang@amd.com>
2025-09-13 10:52:18 -07:00
Binyao Jiang
9752861002 [Fix] Support qwen3-next MTP+DP (#10392) 2025-09-13 17:45:04 +08:00
Yi Zhang
297d374510 support qwen3_next blackwell (#10403) 2025-09-13 17:18:26 +08:00
Binyao Jiang
31e9d3a5aa [Fix] Init mamba related memory pools with torch.zeros (#10400) 2025-09-13 14:16:48 +08:00
Xinyuan Tong
6f4676ef85 fix: tool parse in large streaming chunk beginning with normal content (#10397)
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
2025-09-12 22:29:35 -07:00
Lianmin Zheng
c9ec4cae5b Fix the style of sgl kernel (#10398) 2025-09-12 22:20:21 -07:00
narutolhy
99757cc3e6 fix probs name which without temp scaling name (#9984) 2025-09-13 12:19:57 +08:00
Lianmin Zheng
cdddab056c [Auto Sync] Update xgrammar_backend.py (20250913) (#10395)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-09-12 17:46:56 -07:00
Simo Lin
7c5a0a1b77 [router] add not implemented functions for multi model trait (#10394) 2025-09-12 16:44:18 -07:00
Teng Ma
49f169d53e [HiCache] doc: update deployment in readme (#10332)
Signed-off-by: Teng Ma <sima.mt@alibaba-inc.com>
2025-09-12 16:35:37 -07:00
Teng Ma
7fce2fd91a [HiCache] fix mooncake config in different tp size (#10377) 2025-09-12 16:34:23 -07:00
Even Zhou
16cd550c85 Support Qwen3-Next on Ascend NPU (#10379) 2025-09-12 16:31:37 -07:00
Muqi Li
d5e2a37414 Benchmark: Support API_KEY without 'bearer' (#10380) 2025-09-12 16:29:04 -07:00
Keyang Ru
366043db8e [router] Add get and cancel method for response api (#10387) 2025-09-12 16:19:38 -07:00
Simo Lin
2f173ea074 [router] allow one router to support different model families and serving mode (#10244) 2025-09-12 16:18:27 -07:00
Mohammad Miadh Angkad
321fecab74 Add sentencepiece to project dependencies (#10386) 2025-09-12 16:02:54 -07:00
Yineng Zhang
9d775b1a2d feat: add deepseek v3 fp4 ut (#10391) 2025-09-12 15:43:29 -07:00
kk
78b7465cad Fix GPU fault issue when run dsv3 with dp mode and enable torch-compile (#10361)
Co-authored-by: wunhuang <wunhuang@amd.com>
2025-09-12 15:05:51 -07:00
Simo Lin
07bcad7fb7 [bug] fix router ci syntax error (#10390) 2025-09-12 14:39:15 -07:00
Yineng Zhang
98adac8ef7 fix: exclude protobuf generated code (#10388) 2025-09-12 13:52:09 -07:00
Yineng Zhang
cef11e9a55 fix: resolve gb200 image link (#10343) 2025-09-12 13:46:02 -07:00
Lianmin Zheng
2269cf1e2f [Auto Sync] Update base_grammar_backend.py, llguidance_back... (20250911) (#10333)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-09-12 12:55:55 -07:00
Yi Zhang
151e287d1a fix: add fast path for function call (#9023)
Co-authored-by: tazjin <mail@tazj.in>
2025-09-12 10:28:54 -07:00
Simo Lin
8c86595c93 [router] enable sccache in ci and local build (#10099) 2025-09-12 09:43:48 -07:00
Frank Fang
4634fd5953 [router] Add Rerank Routing Logic in Regular Router (#10219) 2025-09-12 09:10:18 -07:00
fzyzcjy
efedbe6ca9 Fix global input scale incompatible with CuTe DSL moe (#10370) 2025-09-12 03:22:49 -07:00
fzyzcjy
3a77c80b26 Fix FA4 import cause moe_fused_gate output be illegal memory (#10368) 2025-09-12 03:21:26 -07:00