Commit Graph

Select branches

Hide Pull Requests

bot/update-est-time-20260928

dsv41-pd

dsv41-pd-visioncp

main

Mono Color

• 712b01d875 Update CODEOWNERS to include new documentation paths for docs and doc… (#23293) Mingyi 2026-04-20 16:48:41 -07:00
• 3e367f9bcd [sgl] fix incorrect behavior in cuda graph draft extend (#22832) Tarushii Goel 2026-04-20 16:29:16 -07:00
• 100b0f86dd [sgl] add support for weight update function in spedec (#22088) Tarushii Goel 2026-04-20 16:26:20 -07:00
• 28f3a2d8ed [sgl] multilayereagleworkerv2 fix (#22954) Tarushii Goel 2026-04-20 16:22:16 -07:00
• 57ecce9807 [AMD] Enable MTP for GLM-5-mxfp4 model (#23219) Thomas Wang 2026-04-21 07:09:07 +08:00
• a3291b5654 Add new Mintlify documentation site (docs_new/) (#23001) +1525 people 2026-04-20 15:10:22 -07:00
• 575fdc2c4c [CI][LoRA] Drop flaky all-None batch from multi-LoRA parity test (#23287) jsheng_Linkedin 2026-04-20 14:43:25 -07:00
• b65799cf83 [SPEC][1/N] feat: add adaptive speculative_num_steps for EAGLE topk=1 (#21599) shuwennandQiaolin-Yu 2026-04-21 05:25:04 +08:00
• dbcf7459b5 fix: reset empty prefill batch fullness (#23138) shuwenn 2026-04-21 05:14:00 +08:00
• d8d9d32b29 [docker] Fix stray backslash dropping sgl-model-gateway COPY (#23097) mispa-msandClaude Opus 4.7 2026-04-20 13:44:05 -07:00
• 6f6843c582 [Docker] Move Rust toolchain install to torch_deps stage (#23278) ishandhanani 2026-04-20 15:13:10 -05:00
• fe9b9b254b Fix segfault in cudaMemcpyBatchAsync on CUDA 13.0 (#23136) 3 people 2026-04-21 03:20:22 +08:00
• 8cb957ccff [Perf] Make EAGLE bigram key an O(1) view on RadixKey (#23106) Liangsheng Yin 2026-04-20 12:01:11 -07:00
• 3dc1491c95 Support moe_dp_size = 1 for various attention_cp_size (#22003) ShunkangzandShunkang 2026-04-21 02:58:19 +08:00
• 90d527195b [CI] Fix nightly docker builds failing on root-owned workspace leftovers (#23279) ishandhananiandClaude Opus 4.7 2026-04-20 13:56:33 -05:00
• b4bb036b73 fix legacy deepep path for flashinfer_cutedsl (#22925) Lee Nau 2026-04-20 11:49:33 -07:00
• 4698f4cd10 [CI] Fix wait-for-jobs hanging when matrix job skipped at job level (#23277) Kangyan-ZhouandClaude Opus 4.7 2026-04-20 11:16:10 -07:00
• b5d9a86e4c fix: add back priorty as radix cache policy (#23275) ishandhanani 2026-04-20 12:04:35 -05:00
• 332ec5e5ee [release] install rust toolchain in main dockerfile (#23014) Alex Nails 2026-04-20 09:50:08 -07:00
• 39c720d1b9 [Diffusion][NPU][CI] update perf numbers (#23056) Makcum888eandronnie_zheng 2026-04-20 19:34:11 +03:00
• 9a0fd2ff0c [diffusion] optimize: default to in-memory loading for URL/base64 image inputs (#23118) Mick 2026-04-20 23:29:02 +08:00
• 0be6ab04dd [diffusion] refactor: LTX2.3 code cleanup (#23207) Mick 2026-04-20 19:02:05 +08:00
• da62e90904 [AMD] Fix multimodal timeout issue : rocm7.2 PR Test (#23247) YC Yen-Ching Tseng 2026-04-20 18:36:08 +08:00
• cf4b84f839 [AMD] Update AMD workflow name (#23245) YC Yen-Ching Tsengandbingxche 2026-04-20 18:18:24 +08:00
• 4028a73c10 [KV-Events] Fix kv events events publishing for CP (#22983) Vladislav Nosivskoy 2026-04-20 12:34:38 +03:00
• bea4c895c1 [gRPC] Pass --experimental_allow_proto3_optional to protoc in build.rs (#23226) Alex NailsandClaude Opus 4.7 2026-04-20 02:20:14 -07:00
• 69eb95f20c [AMD] Pin peft<0.19 in pyproject_other.toml to fix ROCm CI ImportError (#23161) 3 people 2026-04-20 14:43:56 +08:00
• a2d30d27fe wait for reap in kill_process_tree (#23213) Liangsheng Yin 2026-04-19 23:36:33 -07:00
• ab936ce694 Revert "perf: optimize PCG inductor path for FP8 models (#21734)" (#23159) Bingxu Chen 2026-04-20 14:32:50 +08:00
• 97baf17557 Fix test_modelopt_export using stale ModelConfig kwargs (#23214) Kangyan-ZhouandClaude Opus 4.7 2026-04-19 23:18:09 -07:00
• 1ebe1c57ed [CI] Partition stage-a-test-cpu into 4 matrix shards (#23208) Kangyan-ZhouandClaude Opus 4.7 2026-04-19 22:07:37 -07:00
• 10e17cc55e [gRPC] Native gRPC server: proto + Rust crate scaffold + server args (#22736) Alex Nails 2026-04-19 21:39:35 -07:00
• c304d0d64d [Refactor] Deduplicate NSA utils.py into cp_utils.py for context parallel (#22914) Baizhou ZhangandClaude Opus 4.6 2026-04-19 21:35:35 -07:00
• eb76aaba88 [core] Always-on StreamingSession in UnifiedRadixCache (#23202) Liangsheng Yin 2026-04-19 21:19:43 -07:00
• e389a52cc8 Support allreduce fusion with cp (#21249) ShunkangzandShunkang 2026-04-20 12:06:00 +08:00
• a7276b623e integrate streaming session into UnifiedRadixCache (#23145) Liangsheng Yin 2026-04-19 20:47:41 -07:00
• 1d252803f5 fix(ci): repair path filters regressed by #21482 (#23201) Kangyan-ZhouandClaude Opus 4.7 2026-04-19 20:34:57 -07:00
• 3063d640dd [CI] Exclude diffusion-specific paths from main_package filter (#23053) Thomasandronnie_zheng 2026-04-20 10:43:44 +08:00
• 1cff871c67 [Bugfix] Fix DeepEP timeout when compiling DeepGeMM in EP+DP+TP (#23185) 3 people 2026-04-19 17:36:11 -07:00
• d3ce664612 move session to python/sglang/srt/session (#23144) Liangsheng Yin 2026-04-19 17:34:19 -07:00
• 7ca3566130 Multi platform Plugin (#21388) 6 people 2026-04-20 08:23:51 +08:00
• ebcc2b3eec ci: run weekly est_time update on Monday using p90 of last 15 runs (#23120) Cheng WanandClaude Opus 4.7 2026-04-19 14:39:27 -07:00
• 4fa3482180 [Bugfix] Add missing http_worker_ipc in session error path (#22766) inkcherry 2026-04-20 03:47:09 +08:00
• b74a9dd854 [AMD] fix tbo runtime error when initializing metadata for cuda graph (#22598) billishyahao 2026-04-20 03:42:48 +08:00
• 32b7777f6c [AMD]Fix AMD multimodal-gen-test-2-gpu timeout by adding partition for standalone test (#23130) YC Yen-Ching Tseng 2026-04-19 23:16:18 +08:00
• 6ecd6f84db [CI] Add per-job uv venv isolation and upgrade CI version to Cuda 13 (#23119) 5 people 2026-04-19 05:32:36 -07:00
• 03828f4205 [AMD] Reduce NSA indexer kernels (weights_proj, k-cache store kernel fusion) (#22850) Thomas Wang 2026-04-19 15:18:11 +08:00
• 2a327f0877 Fix Qwen3.5 video processing when passing video_data in "processor_output" format (#22431) Kehan Li 2026-04-19 00:04:01 +08:00
• cd6ad80c00 diffusion: add HunyuanVideo GroupNorm+SiLU fast path (#22814) Xiaoyu Zhang 2026-04-18 23:38:49 +08:00
• c6a45fab64 Qwen3next flashinfer allreduce auto enable (#22664) Xiaoyu Zhang 2026-04-18 22:32:41 +08:00
• 4839cecbb0 [main] chore: add bias for base layer with lora (#22169) Yisheng Gong 2026-04-18 02:07:02 -07:00
• ea20f1baa4 [NPU] [DOC] Update npu best practice docs to match latest code (#23077) amote-i 2026-04-18 14:17:00 +08:00
• 0d94c3366a [diffusion] feat: introduce ltx-2-two-stage device manager (#22869) Mick 2026-04-18 11:04:33 +08:00
• 615d6c93b2 [codex] Add flashinfer TRTLLM backend for diffusion NVFP4 (#22717) Xiaoyu Zhang 2026-04-18 09:06:28 +08:00
• 9c47bbad13 Clean up bench_one_batch warning and simplify norm dispatch (#23110) Lianmin Zheng 2026-04-17 17:42:20 -07:00
• 5f7aee726a refactor(moe): de-duplicate triton MoE runner path into shared helpers (#23019) Cheng WanandClaude Opus 4.7 2026-04-17 17:05:13 -07:00
• fd7db0eace Update CI_PERMISSIONS (#23108) Yanbin Jiang 2026-04-17 16:53:03 -07:00
• 26ae7b8bd7 [MLX] Support radix cache (#21509) R0CKSTAR 2026-04-18 07:00:50 +08:00
• 09b689b407 Apply HF transformers patches from sglang init (#23103) Liangsheng Yin 2026-04-17 15:37:51 -07:00
• 573e12a7fc Merge /get_load into /v1/loads (#23010) Liangsheng Yin 2026-04-17 13:36:51 -07:00
• 44e67c6835 Remove deprecated double sparsity feature (#23009) Lianmin Zheng 2026-04-17 13:33:12 -07:00
• 3df35ecc80 Lower TestPiecewiseCudaGraphQwen25VL gsm8k threshold to 0.80 (#23099) Liangsheng Yin 2026-04-17 13:31:10 -07:00
• 9df6107dca [AMD] Enable DFLASH speculative decoding on ROCm (#22342) andyluo7andAndy Luo 2026-04-17 13:10:14 -07:00
• 90c76d665e [HiCache] fix: HiCacheFile component key suffixing (#22891) shuwennandZhangheng 2026-04-18 04:06:28 +08:00
• 5d4e899477 [AMD] Fix AMD Multimodal Test - skip nvfp4 tests (#23045) YC Yen-Ching Tseng 2026-04-18 00:02:39 +08:00
• 2bac219d0c [Perf] Precompute gemma_weight to avoid redundant add on every forward (#22673) Jincong Chen 2026-04-17 23:37:41 +08:00
• 83c5119d01 [diffusion] CI: fix ModelOpt B200 CI artifact coverage (#22955) Xiaoyu Zhang 2026-04-17 23:33:42 +08:00
• 5de89ea942 [diffusion] CI: fix auto-partition (#23076) Mick 2026-04-17 22:37:24 +08:00
• f399997d2f [AMD] mirror nightly images to local registry and prefer LAN pulls (#23073) YC Yen-Ching Tsengandbingxche 2026-04-17 19:49:26 +08:00
• 8c13295842 [AMD] fix AMD CI gate (#22974) YC Yen-Ching Tsengandbingxche 2026-04-17 18:32:26 +08:00
• 6e3bbef568 expose num_embeddings in VocabParallelEmbeddingWithLoRA (#22547) Opher Lieber 2026-04-17 02:35:13 -07:00
• a12ea979d4 [test] Add GSM8K accuracy test for PP with mixed chunk prefill (#23029) CYYYC0310andcyy 2026-04-17 17:09:53 +08:00
• 271c177443 [NPU]chore(docker): use editable install for sglang in npu.Dockerfile (#23040) ybyang 2026-04-17 17:08:39 +08:00
• 0b2058853d [diffusion] doc: update doc (#23052) Mick 2026-04-17 16:23:46 +08:00
• 0d031335ed [Pipeline Parallelism][Bug] Fix scheduler hang in pipeline parallelism setup (#23006) Jonah Bernard 2026-04-17 02:50:47 -04:00
• 8c190f6b91 [AMD] Add SGLANG_MORI_MOE_MAX_INPUT_TOKENS to truncate dispatch before MoE. (#22952) Duyi-Wang 2026-04-17 14:40:15 +08:00
• 53f87c463d [Docs] [npu] change the feature support status (#23041) xdtbynd 2026-04-17 14:34:54 +08:00
• 43eb66028f ci: install rust toolchain in ci_install_dependency.sh (#23017) Alex Nails 2026-04-16 23:18:22 -07:00
• 7390eddf28 feat(observability): add OpenTelemetry tracing for speculative decoding (#19545) RichardoMuandMu Huai 2026-04-17 14:01:58 +08:00
• 5fa0c6a52e Allow piecewise CUDA graph with speculative decoding (#22128) 3 people 2026-04-16 22:39:30 -07:00
• 91679d935d [codex] Update diffusion skills (#23028) Xiaoyu Zhang 2026-04-17 13:29:26 +08:00
• 7ac337df94 [AMD] CI Job Monitor: fix queue time, utilization, and summary metrics (#22274) Bingxu Chenandbingxche 2026-04-17 13:03:37 +08:00
• 0dcfae5553 [CPU] Add gemma4_rmsnorm_cpu kernel (#22842) 3 people 2026-04-17 13:03:16 +08:00
• 6c89214584 [CPU][sgl-kernel] extend_attention_cpu and flash_attn_varlen_func: fix nan for large seq (#22434) Chunyuan WUandMa Mingfei 2026-04-17 13:01:01 +08:00
• f0f0148167 Revert "feat: Support MXFP4 quantized dense models on AMD CDNA2/CDNA3 GPUs (#19143)" (#23031) YC Yen-Ching Tseng 2026-04-17 12:53:25 +08:00
• 7d47f40a96 [UnifiedRadixTree]: Add HiCache hook interface for TreeComponent (#22924) Zhangheng 2026-04-17 12:09:41 +08:00
• cf9845f8e3 [Bug Fix] Ensure prefill_info_table is populated before honoring disagg_prefill_dp_rank (#22990) Byron HsuandByron Hsu 2026-04-16 20:10:31 -07:00
• 04a53955b9 feat: add coordinated checkpoint prefetch for network filesystem loading (#20843) Jan Bernlöhr 2026-04-17 05:08:19 +02:00
• a77abbe005 [VLM] Reduce GPU memory footprint of CUDA IPC MM feature transport (#22662) Yuhao Yang 2026-04-17 10:38:36 +08:00
• 5a0eea8ac5 [CI] Adding Gemma 4 to Nightly CI (#22408) Khoa Pham 2026-04-16 19:30:16 -07:00
• 16d11c2a10 Fix for the low-probability garbled output issue in the GLM-5 series models. (#22811) Yuxuan Zhang 2026-04-17 09:52:13 +08:00
• 0052093178 test(4-gpu-b200): split test_qwen35_models.py + bump partitions 5→6 (#22913) Alison Shao 2026-04-16 18:51:59 -07:00
• e353630b57 [Diffusion] [NPU] Fix multimodal gen CI (#22879) Makcum888e 2026-04-17 04:09:44 +03:00
• ba850d3a9d [Bugfix] [NPU] Fix check_env on Ascend for CANN 8.5 (#22888) Egor Filimonov 2026-04-17 04:05:20 +03:00
• 3d2d57c6cc [diffusion] refactor: extract LTX2 image encoding from denoising stage (#22976) Mick 2026-04-17 08:35:15 +08:00
• 2cc52d8326 feat: Support MXFP4 quantized dense models on AMD CDNA2/CDNA3 GPUs (#19143) Daifeng Li 2026-04-17 07:51:32 +08:00
• f639425ff0 add check for none status code in FinishAbort (#22535) 3 people 2026-04-16 16:21:07 -07:00
• 2211b4d9c6 [sgl] improve accuracy of additional page requirement during spec decode (#22406) Tarushii Goel 2026-04-16 15:50:51 -07:00
• db7a751d48 refactor: extract FanOutCommunicator and use declarative spec table (#22967) Liangsheng Yin 2026-04-16 15:37:19 -07:00
• 52f0b86f5d [AMD] Qwen3.5 MXFP4 breaks after shared expert fusion is enabled (#22948) mqhc2020andHubert Lu 2026-04-17 06:25:33 +08:00