Commit Graph
5654 Commits
Author SHA1 Message Date
Liangsheng Yin 458611de77 Unify forward output datastructure (#11124) 2025-10-03 00:28:57 +08:00
Chang Su 3511b37099 [proto] Add script to compile python protos (#11171) 2025-10-02 08:45:51 -07:00
fzyzcjy afcd3e1089 Tiny remove duplicated code (#11164) 2025-10-02 21:56:31 +08:00
fzyzcjy 12d6818380 Tiny fix ep_gather behavior different in CI (#11130) 2025-10-02 21:55:53 +08:00
fzyzcjy b65db0287b Tiny cleanup deepseek_v2.py (#11163) 2025-10-02 21:54:52 +08:00
b8zhong 948278f173 fix cpp JIT compilation issue of ngram speculative decoding (#10837) 2025-10-02 21:05:01 +08:00
Liangsheng Yin 7d00479950 Clean up ascend allocator (#11152) 2025-10-02 20:34:26 +08:00
083629c235 [model] Add mamba2 and Falcon-H1 support. (#10988)
Co-authored-by: Younes Belkada <younes.belkada@tii.ae>
Co-authored-by: Younes B <49240599+younesbelkada@users.noreply.github.com>
2025-10-02 19:15:36 +08:00
Chang Su b658be6f6a [router][grpc] Support tool call parser in streaming (#11160) 2025-10-02 03:18:50 -07:00
fzyzcjy 5e786cca3a Support single batch overlap (#10422) 2025-10-02 18:04:36 +08:00
fzyzcjyandKaixi Hou 0b9dfba787 Support dispatch low latency (#10263)
Co-authored-by: Kaixi Hou <4001424+kaixih@users.noreply.github.com>
2025-10-02 18:02:19 +08:00
Liangsheng Yin 6a29003410 Remove unused pack .item() in paged allocator. (#11156) 2025-10-02 18:01:21 +08:00
fzyzcjy 2ac453b07f Tiny detect slow ranks (#10508) 2025-10-02 18:00:33 +08:00
fzyzcjy f35def8652 Fuse quantize and rope in trtllm_mla MTP (#10779) 2025-10-02 17:59:37 +08:00
fzyzcjy d61615fe93 Tiny fix missing alt stream in nextn layer (#10768) 2025-10-02 17:58:23 +08:00
fzyzcjy b1ccaf01cd Tiny improve dumper (#11132) 2025-10-02 17:55:01 +08:00
Lianmin Zheng 097725bb66 Clean up parallel_state.py (#11148) 2025-10-02 01:09:13 -07:00
fzyzcjy 44b1fbe258 Fix DeepSeek chunked prefill memory issue (#11149) 2025-10-01 23:56:59 -07:00
sogalin c0dbbdd12b [ROCm] To reduce the compiling time when using torch compile. (#10559) 2025-10-01 23:53:14 -07:00
Liangsheng Yin 25e7dbe8af Fix ngram spec with page size > 1 (#11135) 2025-10-02 12:34:23 +08:00
Zhang Jundaandhnyls2002 0b2aa8a70c Intoduce cpu tensor as metadata to avoid blocking gpu kernel launch (#10720)
Co-authored-by: hnyls2002 <lsyincs@gmail.com>
2025-10-02 10:51:25 +08:00
Lianmin Zheng 609f65ba23 Remove debug print statement from scheduler output (#11145) 2025-10-01 13:37:05 -07:00
Lianmin Zheng 2d62af6be5 Fix metrics and request tracing (TimeStats) (#11123) 2025-10-01 13:03:07 -07:00
Keyang Ru a28b394fba [router] Add multi-turn tool calling loop support for MCP integration (#11143) 2025-10-01 12:50:21 -07:00
Simo Lin 96fe2d0f15 [router] add pd service in grpc router for pd (#11120) 2025-10-01 08:09:21 -07:00
ykwdandZhiqiang Xie bfa274380b [HiCache] Configurable and Dynamic Prefetch Timeout (#10512)
Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>
2025-10-01 08:44:10 -05:00
Shisong Maandmashisong 86cb4db058 [Feature] Add EIC as sglang HiCache Storage backend (#10271)
Co-authored-by: mashisong <mashisong@bytedance.com>
2025-10-01 08:43:34 -05:00
2e130b7618 [HiCache]bug fix: fixed blank item in host_mem_release_queue (#11005)
Co-authored-by: Zuoyuan Zhang <zuoyuanzhang@ZuoyuandeMacBook-Pro.local>
Co-authored-by: Zhiqiang Xie <xiezhq@stanford.edu>
2025-10-01 08:42:37 -05:00
ac1f2928ae feat: add fast_decode_plan from flashinfer, flashinfer to 0.4.0rc3 (#10760)
Co-authored-by: Zihao Ye <yezihhhao@gmail.com>
Co-authored-by: Sleepcoo <Sleepcoo@gmail.com>
2025-10-01 02:56:13 -07:00
Sai Enduri 195a59fe23 Refactor AMD CI. (#11128) 2025-10-01 01:12:28 -07:00
ishandhanani 47488cc353 docker: x86 dev builds for hopper and blackwell (#11075) 2025-10-01 00:06:38 -07:00
yhyang201 6130529143 Quick Fix: fix Qwen3-VL launch failure caused by MRotaryEmbedding arg (#10985) 2025-09-30 22:17:05 -07:00
Xinyuan TongandMick a9ce2bcb3c [Doc] Update multimodal language models documentation (#11111)
Co-authored-by: Mick <mickjagger19@icloud.com>
2025-09-30 22:10:31 -07:00
Lianmin Zhenggithub-actions[bot] <github-actions[bot]@users.noreply.github.com>Sehoon Kim
5dddb331c4 [Auto Sync] Update base_grammar_backend.py, xgrammar_backen... (20250930) (#11115)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Sehoon Kim <sehoon@x.ai>
2025-09-30 21:50:43 -07:00
Hubert Lu 01a26544a3 [AMD] Add Tilelang and Fast Hadamard Transform builds to Dockerfile.rocm (#11114) 2025-09-30 20:00:37 -07:00
Liangsheng Yin 73d4a5f879 Organize spec-related data structures (#10735) 2025-10-01 09:45:30 +08:00
Keyang Ru 7fb551a75d [router] add mcp list and mcp call in output array (#11112) 2025-09-30 21:41:54 -04:00
Xinyuan Tong 1193f13181 fix: KimiK2Detector Improve tool call ID parsing with regex (#10972) 2025-09-30 17:44:44 -07:00
Chenxi Li 84a9f5d660 Feature/make PEFT adapter module format compatibile (#11080) 2025-09-30 16:32:02 -07:00
Chang Su 8ce830a8b0 [router][bugfix] Fix input_logprobs handling with None value and logprob_start_len = -1 (#11113) 2025-09-30 16:09:40 -07:00
qrskannbara fb367acfcb Support Dots.ocr model (#11071) 2025-09-30 12:18:39 -07:00
Trevor Morris a6cc86df9d Fix DSR1 accuracy for flashinfer_trtllm MoE with FP8 quantization (#11081) 2025-09-30 10:33:12 -07:00
Zaili Wang 229d2b95f1 [CPU] Adding Memory Capacity Acquisition Functionality (#11102) 2025-09-30 08:41:20 -07:00
Liangsheng Yin 9710f718fb [Eval] Add --repeat in run_eval (#11101) 2025-09-30 23:35:54 +08:00
Ke Bao 91847e382a Fix eagle radix cache (#10846) 2025-09-30 22:59:20 +08:00
Simo Lin 5a290a5644 [router][grpc-server] Fix gRPC server shutdown (#11094) 2025-09-30 07:12:12 -04:00
PGFLMG 580051c5a8 chore: bump sgl-kernel v0.3.14 (#11067) 2025-09-30 02:53:24 -07:00
Lianmin Zhenggithub-actions[bot] <github-actions[bot]@users.noreply.github.com>Cheng Wan
1237aa19ce [Auto Sync] Update fused_moe_triton_config.py (20250930) (#11099)
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Cheng Wan <54331508+ch-wan@users.noreply.github.com>
2025-09-30 00:57:16 -07:00
jacky.cheng 5991119541 [Fix] Resolve performance drop in speculative decoding aiter backend (#11087) 2025-09-29 23:51:30 -07:00
Ke Bao 424591d53d Fix spec filter batch when target extend (#10991) 2025-09-30 14:44:02 +08:00