Commit Graph

Select branches

Hide Pull Requests

bot/update-est-time-20260928

dsv41-pd

dsv41-pd-visioncp

main

Mono Color

• 92492896a5 Fix disaggregation test bootstrap port conflict (#21271) Ke Bao 2026-03-25 12:14:41 +08:00
• f81dec090b [AMD] Fix CI: correct stage-b job dependency names in AMD CI workflows (#21357) Michaelandbingxche 2026-03-24 20:48:47 -07:00
• c1d930c028 Increase flush cache timeout in hicache CI (#21305) Ke Bao 2026-03-25 10:00:59 +08:00
• cd634abee4 [CI] Reduce session correctness test to 30 turns to fix flakiness (#21349) Liangsheng Yin 2026-03-24 18:56:01 -07:00
• f273ba1ccc [KDA] Support CuTeDSL KDA decode kernel (#21203) Yuan Luoandluoyuan.luo 2026-03-25 09:47:09 +08:00
• dfc15b78b0 [misc] clean up kernel API (#21325) DarkSharpness 2026-03-25 09:10:23 +08:00
• 281fe10b5e [diffusion] quant: support nvfp4 for Flux.2 (#20137) 8 people 2026-03-24 17:28:25 -07:00
• 37420dce0b [CI] Enable failfast (-f) by default in run_suite.py (#21330) Liangsheng Yin 2026-03-24 17:04:42 -07:00
• 3937a07b17 [CI] Add cross-job fast-fail health check (Layer 3) (#21341) Liangsheng Yin 2026-03-24 16:58:56 -07:00
• 1046dbe038 [Fix] Fix trtllm fp4 moe kernel not found error (#21343) Baizhou Zhang 2026-03-24 16:38:05 -07:00
• 30896bd927 [CI] Remove test partition assignments from CI summary (#21344) Liangsheng Yin 2026-03-24 16:27:47 -07:00
• bbe25b2412 Use FlashInfer tinygemm for GPT-OSS MoE router on SM90+ (#20755) Mohammad Miadh Angkadandelvischenv 2026-03-25 06:00:18 +08:00
• 31c35f1c22 [CI] Skip multimodal CI for doc-only changes (#21334) Liangsheng Yin 2026-03-24 14:07:25 -07:00
• 6cb1c2d53d [AMD] Add 4-GPU test suite for MI325 runners (#20294) Michael 2026-03-24 14:04:49 -07:00
• c4db64c16b Add Lychee Doc Links Check to Local and CI (#19742) 4 people 2026-03-24 13:48:26 -07:00
• a32e0d57e7 [LoRA][III] Add LoRA support for MoE layers and enable TP (#14105) 3 people 2026-03-24 16:14:14 -04:00
• a3ed2e4d29 [diffusion][CI] Add CI for MOVA model inference (#20430) 3 people 2026-03-25 02:28:16 +08:00
• 71f5ae3f9a [AMD] Fix AMD Nightly Test - Transformers 5.3.0 incompatibility and gemma2-27b kv issue (#21193) YC Yen-Ching Tsengandbingxche 2026-03-25 01:41:44 +08:00
• 9f4d8ac99f [Diffusion][NPU] Add support for Hunyuan3D (#20352) Elizaveta MartirosianandElizaveta Martirosian 2026-03-24 16:18:49 +03:00
• 1b4933d45d [NPU][ModelSlim] adapt w2 quant layer for Minimax2.5 (#20905) shadowxz109 2026-03-24 20:57:18 +08:00
• eefb504f84 [diffusion] model: Fix FLUX.1 output correctness (#21041) Aleksi VesantoandMick 2026-03-24 14:17:33 +02:00
• 4fbb311234 [Fix][Eval] Keep --dataset-path scoped to longbench_v2 (#21156) Mohammad Miadh Angkad 2026-03-24 17:25:11 +08:00
• 8d7453d8fb Update CODEOWNERS (#21298) Ke Bao 2026-03-24 17:15:25 +08:00
• 855d15adf6 [AMD] Tilelang sparse fwd for dsv32 mi355/mi300 (#19945) Thomas Wang 2026-03-24 17:01:39 +08:00
• 3dfaa47d5e Change cron schedule to run every 6 hours (#21285) Lianmin Zheng 2026-03-24 00:04:23 -07:00
• dac148167c Enable the qwen3 test (#21195) ShunkangzandShunkang 2026-03-24 14:40:59 +08:00
• 69f02e36e8 [diffusion] Fix torch.zeros typo in causal wan (#21250) Xiaoyu Zhang 2026-03-24 14:39:16 +08:00
• d9f97b2115 Refine diffusion skills and align JIT kernel docs with the new CI flow (#21283) Xiaoyu Zhang 2026-03-24 14:38:36 +08:00
• c01ee848b0 Revert "fix: use consistent time denominator for throughput metrics in bench_one_batch_server" (#21276) Cheng Wan 2026-03-23 22:14:54 -07:00
• 3837588a4c Temporarily disable TestReturnRoutedExperts since it's flaky (#21267) Qiaolin Yu 2026-03-23 21:45:30 -07:00
• 6491728797 [Perf] Overlap NSA-CP key all-gather with query computation for DeepSeek-V3.2 (#20438) 3 people 2026-03-24 12:31:48 +08:00
• d6ce1b8dce docs: update add-jit-kernel skill for run_suite CI registration (#21264) Lianmin Zheng 2026-03-23 21:31:22 -07:00
• 260abe1fb1 Refactor JIT kernel CI to use run_suite.py registration system (#21239) Lianmin Zheng 2026-03-23 21:17:27 -07:00
• 0986bed8e2 [HiCache][HybridModel]: Support mamba state offloading & HybridCacheController (#20457) 4 people 2026-03-24 11:02:50 +08:00
• 2b1d3c935e [diffusion] fix Z-Image SP sharding for portrait and padded resolutions (#21042) Ratish PandXiaoyu Zhang 2026-03-24 07:45:33 +05:30
• d173cfecd6 Update access of cut branch workflow and delete deprecated release workflow (#21231) Baizhou Zhang 2026-03-23 15:18:10 -07:00
• fcaad42b00 [Bug Fix] GLM-V / GLM-OCR: field detection for transformers 5.x and MTP omission fix (#21134) Yuxuan Zhang 2026-03-24 04:19:48 +08:00
• 4779755eb9 Split pr-test.yml: extract sgl-kernel, jit-kernel, and multimodal-gen tests into separate workflow files (#21219) Lianmin Zheng 2026-03-23 13:17:05 -07:00
• ed316a26ef Fix CP in-seq-split method for DeepSeek V32 and update related tests (#21192) Baizhou Zhang 2026-03-23 12:34:10 -07:00
• 27ac831a84 docs: improve CI and testing documentation (#21202) Lianmin ZhengandClaude Opus 4.6 2026-03-23 10:48:50 -07:00
• b4d3fb001d [AMD] Add fused GemmaRMSNorm forward_hip to use aiter/vllm kernels for qwen3.5 (#21188) jacky.cheng 2026-03-24 01:21:36 +08:00
• 6d2e1156a0 [Test] Add unit tests for srt/constrained module (#21010) Vipul Vaibhaw 2026-03-23 22:03:18 +05:30
• 83b9d74424 ci: unit test for srt/observability module (#21002) Jiabin Wang 2026-03-23 09:30:02 -07:00
• 4dbe42527e [Test] Add unit tests for srt/parser (#20947) Zijun Gao 2026-03-23 11:26:46 -05:00
• 777edb6ef7 Fix(jit): support rmsnorm for hidden_size in {64, 128, 256} (#20661) Johnsonms 2026-03-23 08:17:44 -07:00
• 5bdc07d974 [Qwen3.5] Fuse split/reshape/cat ops in GDN projection with Triton kernel (#21019) Yuan Luoandluoyuan.luo 2026-03-23 23:17:01 +08:00
• 8662ba7db4 [NPU] bugfix for import sgl-kernel error (#21200) McZyWu 2026-03-23 19:52:36 +08:00
• 80d4a0753a fix fused_set_kv_buffer for rope with Ling-v2 (#20316) strgrb 2026-03-23 19:20:40 +08:00
• 4641e5a3d2 [NPU] enhance accuracy for model minimaxm2 from 16.5% to 95.5% (#17695) McZyWu 2026-03-23 19:06:38 +08:00
• 2d288ba8c9 [Bugfix] fix npu get kv_item_lens in PD separation when use ASCEND_US… (#15852) XDaoHongandZhengdQin 2026-03-23 15:56:47 +08:00
• 59cb9a9da6 [Spec][Ngram] 3/N: Fix synchronization issues in Ngram.cpp (#21186) kpham-sglandClaude Opus 4.6 2026-03-23 00:37:07 -07:00
• 814202704b ci: unify PR test suite naming (#21187) Lianmin Zheng 2026-03-23 00:18:45 -07:00
• 3d312643b9 [BUGFIX] Fix CP residual size mismatch crash when tp_size == attn_cp_size (#21170) yudian0504 2026-03-23 15:12:58 +08:00
• 7757a9ddd0 ci: remove IS_BLACKWELL env var; auto-detect Blackwell (#21118) Lianmin Zheng 2026-03-22 23:44:48 -07:00
• 54fdb357d3 [CI][NPU] Fix git report dubious ownership (#21162) Even Zhou 2026-03-23 14:36:35 +08:00
• bc4aaab6a1 [Spec][Ngram] 2/N: Rename branch length to max trie depth (#21181) kpham-sglandClaude Opus 4.6 2026-03-22 23:35:25 -07:00
• d6b12c401c Revert "[bugfix] Fix PPMissingLayer AttributeError when Using PP" (#21189) Cheng Wan 2026-03-22 23:28:36 -07:00
• 13f4f010d8 HiSparse for Sparse Attention (#20343) Zhiqiang Xie 2026-03-22 23:09:31 -07:00
• 44db0c59cf [CI] Fix cutlass import error: restore nvidia-cutlass-dsl force-reinstall (#21182) Alison ShaoandAlison Shao 2026-03-22 22:41:01 -07:00
• 7050011dee Enable JIT clamp_position and resolve_future_token_ids on ROCm (#21116) Lianmin Zheng 2026-03-22 22:33:54 -07:00
• 32a85ef128 [diffusion] CI: auto-skip diffusion tests when required pipeline class is missing from diffusers (#21139) Yuhao Yang 2026-03-23 12:15:21 +08:00
• a094eb9c21 Temporarily disable flaky qwen3 cp test in CI (#21178) Shangming Cai 2026-03-23 12:13:52 +08:00
• d8a5b1dbaf [Bugfix] Work around FlashInfer unified transport issue on GB (#20039) Mohammad Miadh Angkad 2026-03-23 12:10:25 +08:00
• a94d67d44b [SKILL] fix(bench): Support model-specific DenoisingStage variants in… (#21137) Xiaoyu Zhang 2026-03-23 12:08:00 +08:00
• 2b47bd3a34 [Bug Fix] Fix non-streaming request abort failure when --enable-metrics is enabled (#20625) fanghao 2026-03-23 10:58:49 +08:00
• 889e8489e9 [diffusion] model: support FireRed-Image-Edit (#20862) yuumnandyuumn 2026-03-23 10:27:07 +08:00
• 999bad5aba Fix VRAM leak in overlap scheduling with structured output (#20640) (#20697) Cishoon 2026-03-23 08:07:39 +08:00
• 343998865a perf: pad max-num-requests in decode cuda graph for higher coverage (#20978) Yilong Zhao 2026-03-22 17:06:16 -07:00
• ce0541404f [FlashInfer v0.6.6][RL] Support fp8-last-n-bf16 RL for flashinfer_trtllm_routed moe backend (#20214) Ziang Li 2026-03-22 11:17:01 -07:00
• c1fe5de69c [Diffusion] Clean up diffusion Triton kernels and modernize custom op registration (#21122) Xiaoyu Zhangandgemini-code-assist[bot] 2026-03-22 22:38:57 +08:00
• 2406ddfdb8 Add ut guide to test skills (#21130) Ke Bao 2026-03-22 20:55:37 +08:00
• 009eee85a0 CUTLASS FP8 Blockwise GEMM improvement of SM120 (#20887) Brayden Zhong 2026-03-22 05:55:54 -04:00
• 766d225fcc Add SGLang CUDA crash API logging inspired by FlashInfer (#20910) Xiaoyu Zhang 2026-03-22 16:39:40 +08:00
• bb737d7a82 Support Qwen3 MoE context parallel (#18233) 3 people 2026-03-22 16:27:20 +08:00
• 6d160b42bb [Spec][Ngram] 1/N: Reference based Speculative Decoding refactor (#20393) kpham-sgl 2026-03-22 00:55:10 -07:00
• d9f5c2179c ci(slash-cmd): allow write-permission users to /rerun-ut on fork PRs (#21121) Liangsheng Yin 2026-03-22 00:45:48 -07:00
• 1b65c0d259 [Diffusion] Fix torch.compile RMSNorm fallback for Z-Image (#20962) Xiaoyu ZhangandMick 2026-03-22 15:38:22 +08:00
• 1e97864d75 ci(slash-cmd): allow repo write-permission users to /rerun-ut (#21120) Liangsheng Yin 2026-03-22 00:32:29 -07:00
• 3bc595acbc [FlashAttn] Add fused triton kernel for normal_decode_set_metadata (#20778) Bowen Liandkinza99 2026-03-22 15:12:29 +08:00
• f7fc2c8592 [diffusion] fix: fix accuracy for some image models (#20679) Mick 2026-03-22 15:11:57 +08:00
• 47d36d73f7 Update write-sglang-test skill: CUDA-only for common tests + prefer mock (#21119) Liangsheng Yin 2026-03-21 22:54:43 -07:00
• 3343a79466 Fix jit-kernel unit test CI dependency install in pr-test workflow (#21089) Lianmin Zheng 2026-03-21 22:13:06 -07:00
• 2fba2bdad1 refactor: Remove dead code from utils/common.py (#20668) shuwenn 2026-03-22 12:54:17 +08:00
• 1ff0ee95e3 Update pr-test.yml to remove the temp hack - cron: '15 8 * * *' (#21115) Lianmin Zheng 2026-03-21 21:53:31 -07:00
• 76e4a8662c Replace clamp_position with JIT kernel + platform dispatch (#20999) Lianmin ZhengandClaude Opus 4.6 2026-03-21 21:26:26 -07:00
• c1794e2944 [diffusion] fix: fix Sana corrupted output by removing spurious QK norm layers (#20656) 3 people 2026-03-21 21:06:49 -07:00
• c32e35a2a5 [diffusion] CI: fix picklingerror for diffusion models using diffusers backend (#20854) Yuhao Yang 2026-03-22 11:51:03 +08:00
• 6dfa8a40bc [diffusion] CI: make auxiliary coverage explicit and simplify testcases (#20983) Mick 2026-03-21 20:18:23 +08:00
• a0862f00c2 dbrx instruct npu support (#17121) KnightLTCandMcZyWu 2026-03-21 17:10:35 +08:00
• ff921a2ed3 Update cron schedule for PR test workflow (#21090) Liangsheng Yin 2026-03-21 01:03:45 -07:00
• 852e112ebf [Qwen3.5] Fix broken pipeline parallelism layer splitting (#21070) Alison ShaoandAlison Shao 2026-03-21 01:02:51 -07:00
• 9ca68a5904 Remove flaky test for test_lora_update.py (#21088) Baizhou Zhang 2026-03-21 00:52:40 -07:00
• 9b98d55ba4 Change PR test schedule time to 7:50 AM UTC (#21087) Lianmin Zheng 2026-03-21 00:36:05 -07:00
• f2c9d5696f Change daily cron job time in pr-test.yml (#21084) Lianmin Zheng 2026-03-21 00:24:53 -07:00
• 3f883ebf2e Fix test_qwen35_models (#21081) Qiaolin Yu 2026-03-21 00:22:38 -07:00
• dba6fb3d30 Fix streaming logprobs corruption caused by shared mutable list reference (#21030) Lianmin Zheng 2026-03-21 00:18:48 -07:00
• d089db0563 ci: add daily PR Test schedule at 12:18am PDT (#21083) Lianmin Zheng 2026-03-21 00:15:48 -07:00
• 182dd20d57 ci: bypass maintenance gate for PR Test on main (#21082) Lianmin Zheng 2026-03-21 00:08:47 -07:00
• b7a1ae4fac Fix /rerun-stage dispatch failure for non-AMD stages (#21076) Alison ShaoandAlison Shao 2026-03-20 23:48:29 -07:00
• 507db41504 Add CI maintenance mode gate to nightly-test-nvidia.yml (#21079) Lianmin Zheng 2026-03-20 23:44:37 -07:00