Commit Graph

Select branches

Hide Pull Requests

cch/megamoe-rc-v0.1

Mono Color

• d71baa72dc [diffusion] dependency: upgrade cache-dit for better compatibility (#14534) DefTruth 2025-12-12 18:52:58 +08:00
• 82e33170e1 [diffusion] chore: minor code cleanups and improve logging (#14916) Mick 2025-12-12 18:48:07 +08:00
• 94e1251131 enable flashinfer-jit-cache in image build and ci install to speed up model launch (#14959) gongwei-130 2025-12-11 23:20:44 -08:00
• 4dabfbc827 [Fix] suppress remote weight loading engine w/o mooncake installed (#14937) Zaili Wang 2025-12-12 15:12:26 +08:00
• d7ed8a8c24 [NVIDIA] Enable TRTLLM BF16 MoE on Blackwell GPUs (#13798) Sam 2025-12-12 14:56:13 +08:00
• c05d3afb5d [NPU] optimization for dsv3.2 (#14572) ZhengdQin 2025-12-12 14:52:16 +08:00
• edb172e927 Super tiny refactor error.rs logic (#14949) fzyzcjy 2025-12-12 14:48:18 +08:00
• fe6d38d2fa fix: trtllm mha attention auto-selection on sm120 (#14842) b8zhong 2025-12-11 22:35:25 -08:00
• dab31e4cc7 direct register custom op for mm_fp4 (#13699) b8zhong 2025-12-11 22:33:17 -08:00
• a7fa31ff17 Super tiny move error.rs (#14944) fzyzcjy 2025-12-12 14:25:30 +08:00
• bd91f8824a Super tiny remove non-updated sgl_router_worker_load (#14888) fzyzcjy 2025-12-12 13:19:30 +08:00
• b47adb807e Tiny add e2e http request arrival metric (#14893) fzyzcjy 2025-12-12 13:18:23 +08:00
• b2b5bdb06e Tiny add router e2e duration histogram (#14892) fzyzcjy 2025-12-12 13:18:13 +08:00
• 22fe5da13e Fix negative duration panic in token bucket wait time calculation (#14941) JinYan Su 2025-12-12 13:16:47 +08:00
• 1834401e7c [model-gateway] optimize worker selection (#14894) Praneth Paruchuri 2025-12-12 10:45:12 +05:30
• 198c8ecf98 [diffusion] fix: support applying different LoRA adapters to different transformers in multi-transformer pipelines (#14839) Prozac614 2025-12-12 13:08:57 +08:00
• c01b2ee094 [PP] Refactor PP to async mode (#11852) Xuchun Shang 2025-12-12 12:54:16 +08:00
• 8f5adac8c6 [RL] refactor flash rl weight reload in sglang (#14870) Peng Zhang 2025-12-12 12:50:47 +08:00
• 76743a983e [DLLM] Add documentation for diffusion LLMs (#14358) Zehuan Li 2025-12-12 12:29:51 +08:00
• 8bf10e71cd remove dpsk3.2 sys prompt (#14923) Xinyuan Tong 2025-12-12 04:26:13 +00:00
• e4873d0439 Super tiny remove sgl_router_active_workers (#14891) fzyzcjy 2025-12-12 12:00:55 +08:00
• c660d8dfd0 Re-org eagle unit tests (#14909) Liangsheng Yin 2025-12-12 12:25:39 +09:00
• 10146af099 Check KV4 compatibility with attention backends and add KV4 support to the attention_backend doc (#14467) Ho-Ren (Jack) Chuang 2025-12-11 19:00:53 -08:00
• b62e7e3bfe [CI]add nightly CI for glm4v_moe arch model (#14927) Minglei Zhu 2025-12-11 18:29:45 -08:00
• 6ce36b1232 fix: making the publish trace error check broader (#14931) Douglas Yang 2025-12-11 18:19:57 -08:00
• e9e7f15eb5 [bugfix] fix TBO crashes when attn_tp_size > 1 (#13730) Yuhao Yao 2025-12-12 10:18:40 +08:00
• 0aa3dec5c7 Fix black formatting in ci_utils.py (#14932) Alison Shao 2025-12-11 17:28:52 -08:00
• 4885f8b9c1 fix: handle Jinja2 template errors as client errors in OpenAIServingChat (#14748) Xinyuan Tong 2025-12-12 01:06:26 +00:00
• f832994c32 [CI] Add Mistral Large 3 Eagle nightly performance test (#14525) Alison Shao 2025-12-11 17:01:50 -08:00
• e59435c34b Add retry logic for scheduled CI tests (#14771) Alison Shao 2025-12-11 16:59:58 -08:00
• d6bd2d1126 [CPU] layernorm & fused add-layernorm kernels (#14074) Zaili Wang 2025-12-12 08:58:23 +08:00
• 9975acf50f [refactor] Update reasoning parameter to require_reasoning (#14922) Xinyuan Tong 2025-12-12 00:55:59 +00:00
• 70758d457e support non-disturbing remote-instance-weight-loader (#13125) amysaq2023 2025-12-12 08:45:32 +08:00
• fd1ebbb0d6 update mistral detector (#14921) Xinyuan Tong 2025-12-12 00:43:36 +00:00
• 3d98bd5edf ci: adding more nightly tests to bot bump workflows (#14928) Douglas Yang 2025-12-11 16:26:21 -08:00
• aa3716b29d Update CI_PERMISSIONS.json Kangyan-Zhou 2025-12-11 15:55:50 -08:00
• 6107268fe7 extend timeout for b200 test (#14925) b8zhong 2025-12-11 15:43:53 -08:00
• 5c8bd8b51b chore: bump SGLang version to 0.5.6.post2 (#14858) sglang-bot 2025-12-11 12:29:52 -08:00
• b05b346a13 [loader] enable private loader (#14620) Yinghai Lu 2025-12-11 11:49:46 -08:00
• 5c961756a3 [FIX][DS32]openai protocol: support openai message role: developer (#14304) Jimmy 2025-12-12 03:34:41 +08:00
• c5f1e86117 Update CI_PERMISSIONS.json (#14917) Kangyan-Zhou 2025-12-11 10:14:42 -08:00
• 2c4d376d23 [SMG][DS32][fix] support dsv32, add role developer (#14307) Jimmy 2025-12-12 00:09:09 +08:00
• d0f756aec9 [docs] Fix kernel name (#14887) almaslof 2025-12-11 17:48:16 +02:00
• 6f99dc97f7 [model-gateway] fix imports and delete unused code (#14911) Simo Lin 2025-12-11 07:42:47 -08:00
• cd1c1fa5bc [model-gateway] fix annotation error and code formating (#14910) Simo Lin 2025-12-11 07:06:32 -08:00
• 5b0872d2d7 [model-gateway] code clean up on oai router in responses (#14852) Simo Lin 2025-12-11 06:55:09 -08:00
• ba88f1cace Tiny clean router load report logic (#14889) fzyzcjy 2025-12-11 22:50:32 +08:00
• 60560c07a2 [model-gateway] refactor cleanup WorkflowContext.get_or_err (#14890) fzyzcjy 2025-12-11 22:48:30 +08:00
• ca1144212a [diffusion] UX: suppress excessive loggers (#14900) Mick 2025-12-11 22:10:22 +08:00
• 543d62d11a Introduce server_fixtures in sglang.test (#14899) Liangsheng Yin 2025-12-11 22:30:33 +09:00
• 27032cecd9 [Ascend]Support of piecewise graph compilation for prefill on NPU (#12287) Vladimir221 2025-12-11 16:10:07 +03:00
• 5d804a3767 [diffusion] feat: support sageattn & sageattn3 backend (#14878) Mick 2025-12-11 20:59:44 +08:00
• 388018a5bd [NPU] adapt dsv3.2 nsa prefill context parallel (#14541) liupeng374 2025-12-11 20:12:59 +08:00
• fca8e88f28 [diffusion] feat: enable torch compile to eliminate GPU bubble (#13641) AichenF 2025-12-11 19:06:32 +08:00
• 45eeeb9a3c Tiny support range ratio in GSP in bench serving (#14828) fzyzcjy 2025-12-11 17:26:07 +08:00
• a368df2818 Tiny add more error info for bench_serving (#14827) fzyzcjy 2025-12-11 17:25:49 +08:00
• f85460fb19 Avoid deleting entire cache for missing shards (#14754 follow-up) (#14853) Alison Shao 2025-12-11 01:17:04 -08:00
• e52cf30e81 fix: adding temporary bypass for nightly tests (#14876) Douglas Yang 2025-12-11 00:09:03 -08:00
• a076d75e98 [Fix] Remove unused import from test_disaggregation_hicache.py (#14880) Shangming Cai 2025-12-11 15:50:47 +08:00
• 8348725d9e [CI][BUG] fix ib setup for disaggregation hicache test (#14877) shicanwei.scw 2025-12-11 15:40:48 +08:00
• 2856624156 Only count limitations for previous runs that reaches the test stages (#14856) Kangyan-Zhou 2025-12-10 23:17:34 -08:00
• b62fe8504c fix nightly vlm ci : restore original eval for requests without regex (#14875) Yuhao Yang 2025-12-11 15:13:25 +08:00
• 624725cb5e Move and update MindSpore docs, make it appear on the online documentation (#14861) Tiance Wang 2025-12-11 15:03:50 +08:00
• 1a96e66493 fix: creating blobs only once for publish trace retries (#14845) Douglas Yang 2025-12-10 22:40:06 -08:00
• 32829b1638 Remove myself to test CI gate issue (#14871) Kangyan-Zhou 2025-12-10 22:37:01 -08:00
• e54307f26a [6/n] Fix num_token_non_padded computation in prefill (#14313) yuchengz816-bot 2025-12-10 22:15:19 -05:00
• 8642dbe416 Refactor Marlin MoeRunner (#14554) Trang Do 2025-12-10 22:15:00 -05:00
• 7dcad45cad [CI] Temp disable gb200 test (#14865) Baizhou Zhang 2025-12-10 19:05:16 -08:00
• 7c985331dd Enable TP for Mamba-based models (#14811) roikoren755 2025-12-11 04:03:50 +02:00
• bd7824b24d Minor code style fix for dllm (#14836) Liangsheng Yin 2025-12-11 10:35:42 +09:00
• 312df1d6c0 Fix TestGLM41VPPAccuracy test flakiness (#14848) Binyao Jiang 2025-12-10 16:59:58 -08:00
• 25e97380e3 Fix CUDA version handling in ci_install_deepep.sh (#14854) Lianmin Zheng 2025-12-10 16:52:30 -08:00
• b6523a4f72 fix: restrict cache validation behaviors to CI only (#14849) Alison Shao 2025-12-10 16:03:53 -08:00
• c51efb8b84 fix fp8 gemm nightly CI (#14844) b8zhong 2025-12-10 15:57:51 -08:00
• bcc5483ed7 [model-gateway] fix import order in oai conversation (#14851) Simo Lin 2025-12-10 15:36:42 -08:00
• ccf2602773 [model-gateway] code clean up on oai router (#14850) Simo Lin 2025-12-10 15:34:27 -08:00
• a4992873d4 Treat unittest SkipTest exception as pass instead of as failure (#14847) Binyao Jiang 2025-12-10 15:28:21 -08:00
• c97ce39181 [AMD] Add model to AMD nightly test (#14442) michael-amd 2025-12-10 11:59:44 -08:00
• c032b559a8 [model-gateway] adds default implementations to RouterTrait in mod.rs (#14841) Simo Lin 2025-12-10 11:46:35 -08:00
• da9b801eb7 fix lora target all + csgmv backend (#14796) b8zhong 2025-12-10 11:42:51 -08:00
• 0e54a69548 [bugfix] qwen25-VL support lora (#14638) Siyuan Chen 2025-12-11 03:38:51 +08:00
• e99ee0c695 [model-gateway] Fix incompatible metric comparison in PowerOfTwo policy (#14823) Praneth Paruchuri 2025-12-11 00:45:38 +05:30
• c1bd5ee8c5 Revert transformers to 4.57.1 (#14801) Yuhao Yang 2025-12-11 03:04:36 +08:00
• ef1ab2302a [Auto Sync] Update tool_chat_template_deepseekv31.jinja (20251210) (#14837) Yineng Zhang 2025-12-10 10:56:24 -08:00
• d659873762 [diffusion] CI: use unified sampling_params for CI (#14045) Mick 2025-12-11 01:18:56 +08:00
• 6c5ebc0ef7 [diffusion] parallel: pad tokens for video models under sp (#14833) Mick 2025-12-11 01:15:37 +08:00
• 5b5571a8da Apply back moe_sum_reduce for fused_marlin_moe (#14829) Ke Bao 2025-12-11 00:39:41 +08:00
• 1698c2341b [CI] Reduce stage-b auto-partition from 4 to 2 (#14769) Alison Shao 2025-12-10 08:32:33 -08:00
• 3d82c0f17f [model-gateway] support engine response http status statistics in router (#14712) fzyzcjy 2025-12-11 00:30:53 +08:00
• 617e9b3bf8 [model-gateway] support customizing Prometheus duration buckets (#14716) fzyzcjy 2025-12-11 00:30:03 +08:00
• 83e35a7c29 [diffusion] doc: fix tiny typo in multimodal_gen/README.md (#14830) Li Jinliang 2025-12-11 00:27:25 +08:00
• 2543666c8e [model-gateway] add anthropic message api spec (#14834) Simo Lin 2025-12-10 08:26:51 -08:00
• f732f8ea57 Fix CI registry scan to only check test/registered directory (#14812) Alison Shao 2025-12-10 08:09:03 -08:00
• 503880dbbe [CI] fix UT success check in test_eagle_infer_beta_dp_attention.py (#14831) Liangsheng Yin 2025-12-11 01:00:50 +09:00
• b8cfa02c01 [NPU] bug fix for mtp and w4a8 (#14806) liupeng374 2025-12-10 22:55:13 +08:00
• d85fecb536 Tiny fix incorrect worker removal command (#14822) fzyzcjy 2025-12-10 22:29:39 +08:00
• 6634f67bd5 Fix router keep nonzero metrics after worker is deleted (#14819) fzyzcjy 2025-12-10 22:29:30 +08:00
• 5eccaf7737 Support HTTP response status code prometheus metrics (#14710) fzyzcjy 2025-12-10 22:28:57 +08:00
• d7f6320bb8 [model-gateway] Dynamically Populate Tool Call Parser Choices (#14807) Wenyi Xu 2025-12-10 22:04:22 +08:00
• 766476f52a [SMG-GO] implement a Go SGLang Model Gateway - OpenAI Compatible API Server (#14770) ybyang 2025-12-10 22:03:28 +08:00