Commit Graph

11116 Commits

Author SHA1 Message Date
Qiaolin Yu
3f883ebf2e Fix test_qwen35_models (#21081) 2026-03-21 00:22:38 -07:00
Lianmin Zheng
dba6fb3d30 Fix streaming logprobs corruption caused by shared mutable list reference (#21030) 2026-03-21 00:18:48 -07:00
Lianmin Zheng
d089db0563 ci: add daily PR Test schedule at 12:18am PDT (#21083) 2026-03-21 00:15:48 -07:00
Lianmin Zheng
182dd20d57 ci: bypass maintenance gate for PR Test on main (#21082) 2026-03-21 00:08:47 -07:00
Alison Shao
b7a1ae4fac Fix /rerun-stage dispatch failure for non-AMD stages (#21076)
Co-authored-by: Alison Shao <alison.shao@Mac.attlocal.net>
2026-03-20 23:48:29 -07:00
Lianmin Zheng
507db41504 Add CI maintenance mode gate to nightly-test-nvidia.yml (#21079) 2026-03-20 23:44:37 -07:00
kk
3f0ba021fc [AMD] Improve openai/gpt-oss performance (#21020)
Co-authored-by: root <root@smci355-ccs-aus-m15-21.cs-aus.dcgpu>
Co-authored-by: Hubert Lu <55214931+hubertlu-tw@users.noreply.github.com>
Co-authored-by: Hubert Lu <Hubert.Lu@amd.com>
Co-authored-by: HaiShaw <hixiao@gmail.com>
2026-03-20 23:16:47 -07:00
Baizhou Zhang
67cad3e69e Revert "Support CuteDSL mm_fp4 backend" (#21077) 2026-03-20 22:47:47 -07:00
Xiaoyu Zhang
c076968c52 [CI] Remove obsolete AOT-only jit-kernel benchmarks after sgl-kernel 4.0 (#21075) 2026-03-21 13:40:42 +08:00
Baizhou Zhang
5f3393c04c Fix deepseek-v32-fp4 b200 ci (#21072) 2026-03-20 22:28:40 -07:00
Liangsheng Yin
9614271ae4 [Test] Fix flaky multi-instance memory occupation test (#21074) 2026-03-20 22:19:03 -07:00
Lianmin Zheng
427442304a Add CI maintenance mode gate to nightly-test-nvidia.yml (#21073) 2026-03-20 22:07:34 -07:00
Lianmin Zheng
50404e0d1f ci: refactor CUDA dependency install script (#21017) 2026-03-20 21:36:58 -07:00
Lianmin Zheng
d5473f87c4 Add CI maintenance mode gate to pr-test.yml (#21069) 2026-03-20 21:27:38 -07:00
Alison Shao
048d90e165 Revert "[AMD] Add MoE weights and scales padding" (#21067) 2026-03-20 20:26:17 -07:00
Lianmin Zheng
a75e74b534 Add revocation of permissions section (#21066) 2026-03-20 20:22:24 -07:00
Rabinovich
3798a8c88d docs: add out-of-tree model integration guide (#21050)
Co-authored-by: Yixiao Zeng <yixiao.zeng@xiaopeng.com>
Co-authored-by: zhaochenyang20 <zhaochen20@outlook.com>
2026-03-20 20:07:46 -07:00
shuwenn
6c91590e1b [HiCache] refactor: hicache normalization flow and compatibility checks (#19669) 2026-03-20 18:38:44 -07:00
mqhc2020
9419453713 [AMD] Add MoE weights and scales padding (#18684) 2026-03-20 14:55:09 -07:00
YC Yen-Ching Tseng
f97c09dac1 [AMD] Enable aiter unified attention for non-SWA models (Qwen3-VL) (#20897)
Co-authored-by: wunhuang <wunhuang@amd.com>
2026-03-20 12:07:41 -07:00
Zijun Gao
576e397b6e [Test] Add unit tests for srt/sampling (#20891) 2026-03-20 23:38:48 +08:00
fzyzcjy
146700db68 Add e2e demo test in dump comparator (#21031) 2026-03-20 22:41:01 +08:00
fzyzcjy
6703cc4484 Enhance output formatting in dump comparator (#21029) 2026-03-20 22:04:50 +08:00
fzyzcjy
fdbcb8156e Refactor dp_utils to use ParallelAxis enum in dump comparator (#21028) 2026-03-20 22:04:20 +08:00
fzyzcjy
154395ab7d Support s≡t dimension name equivalence in dump comparator (#21027) 2026-03-20 22:03:34 +08:00
fzyzcjy
cc22601d28 Validate replicated axes orthogonality in dump comparator (#21026) 2026-03-20 22:02:40 +08:00
fzyzcjy
2f01950a0e Support jointly-determined axes inference in dump comparator (#21025) 2026-03-20 22:01:26 +08:00
fzyzcjy
ecd7e40d20 Support dependent axis auto-resolution in dump comparator (#21024) 2026-03-20 21:56:39 +08:00
Lianmin Zheng
2d7a262ca3 ci: rename 1/2-gpu-runner labels to 1/2-gpu-h100 (#21008) 2026-03-20 06:04:15 -07:00
Lianmin Zheng
104b10f70a refactor: consolidate is_in_ci (jit_kernel, sgl-kernel benchmarks, tests) (#21009) 2026-03-20 05:55:36 -07:00
Артем Савкин
9fbe6800aa [NPU] [Diffusion] Update CI performance baseline for Wan2.2-T2V-A14B-Diffusers-w8a8 (#20997) 2026-03-20 15:54:12 +03:00
xingsy97
f41832795e Add compile-time 256-bit vector guard for pre-Blackwell (#19794) 2026-03-20 18:25:12 +08:00
DarkSharpness
2dd9196079 [JIT Kernel][Feature] Support JIT custom all reduce (rewrite as v2) (#19880)
Co-authored-by: Xiaoyu Zhang <35585791+BBuf@users.noreply.github.com>
2026-03-20 18:24:07 +08:00
Muqi Li
2099943a49 Fix scale_step_k computation in the fp8_kernel (#20819)
Co-authored-by: Xiaoyu Zhang <35585791+BBuf@users.noreply.github.com>
2026-03-20 18:09:31 +08:00
Jia Guo
ec01ef9092 Fix torch.compile/dynamo crash with Qwen3 QK-norm in piecewise CUDA g… (#19818)
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-20 18:05:09 +08:00
Lianmin Zheng
c1da420799 ci: run Stage A CUDA tests as stage-a-test-small-1-gpu on 5090 (#20988) 2026-03-20 02:55:16 -07:00
Prozac614
fa89d152c0 [diffusion] CI: fix hunyuan3d JIT cache (#20773)
Co-authored-by: daiweitao <dwti614707404@163.com>
2026-03-20 17:51:55 +08:00
Lianmin Zheng
a0a4dae67f Revert "Fix DeepSeek V32 FP4 test" (#21003) 2026-03-20 02:19:28 -07:00
Lianmin Zheng
112b628227 Replace _resolve_future_token_ids with JIT kernel + platform dispatch (#20976)
Co-authored-by: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-03-20 01:47:03 -07:00
Xiaoyu Zhang
cf60c5bd15 [CI] Fix jit_kernel benchmark ci (#20990) 2026-03-20 16:40:20 +08:00
Baizhou Zhang
c82d20d48e Fix DeepSeek V32 FP4 test (#20984) 2026-03-20 01:04:32 -07:00
Yilong Zhao
26f709e97d misc: make prefill-delayer compatible with multiple types of mem pool (#20979) 2026-03-20 00:05:53 -07:00
Lianmin Zheng
193bbf9b66 chore(ci): remove deprecated CI Monitor workflow (#20993) 2026-03-20 00:05:22 -07:00
Lianmin Zheng
cea79530f8 ci: merge CPU default and stage-a-cpu-only suites (#20992) 2026-03-20 00:04:55 -07:00
Yilong Zhao
95327458ee misc: add BatchTokenizerReq hook into dp controller (#20981) 2026-03-19 23:59:53 -07:00
Lianmin Zheng
3ebe6e9894 ci: remove redundant RUNNER_LABELS from pr-test.yml (#20991) 2026-03-19 23:59:47 -07:00
Lianmin Zheng
712a48c5d2 ci: move metrics scripts under scripts/ci/utils (#20986) 2026-03-19 23:47:57 -07:00
Lianmin Zheng
0be88ac44c ci: rename pr-test workflow_call input ref to git_ref
Rename workflow_call input to avoid confusion with checkout/dispatch ref. Only caller: release-branch-cut.yml.
2026-03-19 23:45:06 -07:00
lviy
46a76af97b [Bugifx] qwen3 rope parameter compatibility (#20931)
Co-authored-by: Xinyuan Tong <xinyuantong.cs@gmail.com>
2026-03-19 22:22:01 -07:00
Jia Guo
87549f8f0b perf(mamba): use Triton conv1d for non-contiguous input to avoid .contiguous() copy (#20469) 2026-03-19 19:38:46 -07:00