iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity
13,380 Commits 3 Branches 0 Tags
2515bbd02798a0f736a82bbdd10dbf9d5accfa7c
Commit Graph
9428 Commits
Author SHA1 Message Date
Kiersten Stokes 9e138cb01d [Misc][Build] Lazy load cv2 in nemotron_parse.py (#33189) 
Signed-off-by: kiersten-stokes <kierstenstokes@gmail.com>
 2026-01-29 06:55:50 +00:00
wangln19 39037d258e Fix tool call indexing double-counting (#33141) 
Signed-off-by: wanglinian <wanglinian@stu.pku.edu.cn>
 2026-01-29 05:57:09 +00:00
Cyrus Leung 51550179fc [Refactor] Define MM data parser in processing info instead of processor itself (#33260) 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
 2026-01-29 13:55:17 +08:00
Angela Yi 07ea184f00 [ez] Delete more torch version checks <= 2.8 (#33288) 
Signed-off-by: angelayi <yiangela7@gmail.com>
 2026-01-29 05:28:46 +00:00
Michael Goin 1bd47d6e5a [Bugfix] Register fp8 cutlass_group_gemm as supported for only SM90+SM100 (#33285) 
Signed-off-by: mgoin <mgoin64@gmail.com>
 2026-01-28 18:40:59 -08:00
Michael Goin 141cd43967 [UX] Remove noisy CT UnquantizedLinearMethod warn (#33273) 
Signed-off-by: mgoin <mgoin64@gmail.com>
 2026-01-28 16:09:30 -08:00
Nick Hill 6bf3b46d78 [ModelRunner V2] Misc code simplification and cleanup (#33266) 
Signed-off-by: Nick Hill <nickhill123@gmail.com>
 2026-01-28 14:41:23 -08:00
Michael Goin ca1969186d [UX] Enable nested configs in config yaml files (#33193) 2026-01-28 16:54:25 -05:00
Gregory Shtrasberg ab597c869a [Bugfix] Add missing encoder only guard for do_kv_cache_update (#33269) 
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
 2026-01-28 21:25:07 +00:00
Angela Yi 4197168ea5 [ez] Remove checks for torch version <= 2.8 (#33209) 
Signed-off-by: angelayi <yiangela7@gmail.com>
 2026-01-28 16:03:56 -05:00
Rohan Potdar 59bcc5b6f2 Use aiter triton fused_add_rmsnorm_pad for gpt-oss (#30976) 
Signed-off-by: Rohan138 <rohanpotdar138@gmail.com>
 2026-01-28 20:47:47 +00:00
Wentao YeandNick Hill 3e440786af [Feature] Fully support for async scheduling + PP, 30.8% E2E throughput improvement, 31.8% TPOT improvement (#32618) 
Signed-off-by: yewentao256 <zhyanwentao@126.com>
Signed-off-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
 2026-01-28 20:30:32 +00:00
cwazai f210f0b7b1 [lora/moe] Avoid extra intermediate buffer & Python slicing in expand phase when split_k == 1 (#32774) 
Signed-off-by: 陈建华 <1647430658@qq.com>
 2026-01-29 00:22:45 +08:00
Robert ShawRobert ShawHarry Mellor
 		af9b69f977 [Quantization][Deprecation] Remove Marlin 24 (#32688) 
Signed-off-by: Robert Shaw <robshaw@redhat.com>
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
Co-authored-by: Robert Shaw <robshaw@redhat.com>
Co-authored-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
 2026-01-28 15:54:59 +00:00
Chauncey 8e5e40daf4 [Misc] Provide a DeepSeek ReasoningParser with thinking enabled by default (#33221) 
Signed-off-by: chaunceyjiang <chaunceyjiang@gmail.com>
 2026-01-28 21:16:53 +08:00
Or OzeriandKevin H. Luu 2e8de86777 Revert "Enable Cross layers KV cache layout at NIXL Connector (#30207)" (#33241) 
Signed-off-by: Or Ozeri <oro@il.ibm.com>
Co-authored-by: Kevin H. Luu <khluu000@gmail.com>
 2026-01-28 04:36:00 -08:00
Robert ShawandRobert Shaw 247d1a32ea [Quantization][Deprecation] Remove BitBlas (#32683) 
Signed-off-by: Robert Shaw <robshaw@redhat.com>
Signed-off-by: Robert Shaw <114415538+robertgshaw2-redhat@users.noreply.github.com>
Co-authored-by: Robert Shaw <robshaw@redhat.com>
 2026-01-28 11:06:22 +00:00
Harry Mellor f1acbd68c5 [CI] Enable mypy import following for vllm/compilation (#33199) 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
 2026-01-28 08:59:54 +00:00
Gregory Shtrasberg 22ad649501 [ROCm] Enabling forward_includes_kv_cache on ROCm MHA backends (#33106) 
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
 2026-01-28 14:36:14 +08:00
ramosandmayufeng 36d450e3b8 Adds FunAudioChat multimodal audio model support (#2) (#33058) 
Signed-off-by: ramos <49182011+nemoramo@users.noreply.github.com>
Signed-off-by: mayufeng <mayufeng@example.com>
Co-authored-by: mayufeng <mayufeng@example.com>
 2026-01-28 05:18:09 +00:00
22quinn a2b877df6c [Bugfix] Lazy import NgramProposer in GPU model runner (#32821) 
Signed-off-by: 22quinn <33176974+22quinn@users.noreply.github.com>
 2026-01-27 21:07:16 -08:00
Harry Mellor 35fb0b8613 Don't use min_pixels/max_pixels from Qwen2VL's processor (#33208) 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
 2026-01-28 05:02:08 +00:00
Harry Mellor 2eb673a088 Add flake8-implicit-str-concat rules to Ruff (#33191) 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
 2026-01-28 04:56:10 +00:00
Woosuk KwonandNick Hill 46ec6d71c7 [Model Runner V2] Use a different stream for grammar bitmask h2d copy (#33059) 
Signed-off-by: Woosuk Kwon <woosuk@inferact.ai>
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
Co-authored-by: Nick Hill <nhill@redhat.com>
 2026-01-27 16:37:43 -08:00
Richard Zou d9aa39a3bb [torch.compile] Speed up MOE handling in forward_context (#33184) 
Signed-off-by: Richard Zou <zou3519@gmail.com>
 2026-01-27 15:17:54 -08:00
Wentao Ye 3a6d5cbefd [Perf] Optimize dcp allocate tensor (#33102) 
Signed-off-by: yewentao256 <zhyanwentao@126.com>
 2026-01-27 17:24:41 -05:00
linhaifeng f5d7049cc1 [Bugfix] Fix display error (inconsistent with context) (#33020) 
Signed-off-by: linhaifeng <1371675203@qq.com>
 2026-01-27 20:33:29 +00:00
Matthew Bonanni 1cbccb6dba [Attention] Use has_flashinfer helper (#33177) 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
 2026-01-27 18:33:17 +00:00
Iris bd92089d33 feature: support eagle3 for HunyuanVL & Hunyuan (#33035) 
Signed-off-by: irisliu10 <601012173@qq.com>
Signed-off-by: Iris <38269816+irisliu10@users.noreply.github.com>
 2026-01-27 17:55:48 +00:00
IriKa 66e601ef79 Support compress-tensors with nvfp4 or fp8 weights and modelopt with nvfp4 weights on Turing (#33076) 
Signed-off-by: IriKa Qiu <qiujie.jq@gmail.com>
 2026-01-27 11:04:05 -05:00
Nick Hill 0cd259b2d8 [BugFix] Fix P/D with non-MoE DP (#33037) 
Signed-off-by: Nick Hill <nickhill123@gmail.com>
 2026-01-27 08:03:47 -08:00
danielafrimi 83fb2d09e8 Support heterogeneous NemotronHPuzzle model (#32549) 
Signed-off-by: <dafrimi@nvidia.com>
Signed-off-by: Daniel Afrimi <dafrimi@nvidia.com>
Signed-off-by: root <dafrimi@nvidia.com>
 2026-01-27 10:55:54 -05:00
daniserebandJee Jee Li f3a5ee705f [LoRA][Spec Decode] Support LoRA for Nemotron-H MTP models (#32265) 
Signed-off-by: Daniel Serebrenik <daserebrenik@nvidia.com>
Co-authored-by: Jee Jee Li <pandaleefree@gmail.com>
 2026-01-27 07:53:26 -08:00
wang.yuqi 7cbbca9aaa [Frontend] Cleanup api server (#33158) 
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
Signed-off-by: wang.yuqi <noooop@126.com>
 2026-01-27 15:18:10 +00:00
omkhalilandOmar Mohamed Khalil 5ec44056f7 [Metrics][MFU] Fix UnembedMetrics FLOP overcounting for prefill (#33045) (#33045) 
Fix UnembedMetrics to correctly count FLOPs for the unembedding (LM head) layer.

The bug: UnembedMetrics used total_num_tokens() which counts all tokens in the
batch for projection flops, vocab projections are run on just the last token for the
autoregressive use case.

Co-authored-by: Omar Mohamed Khalil <omarkhalil@meta.com>
 2026-01-27 15:16:49 +00:00
Nicolò Lucchesi 492a7983dd [Bugfix] Fix DeepseekV32 AssertionError: num_kv_heads == 1 (#33090) 
Signed-off-by: NickLucche <nlucches@redhat.com>
 2026-01-27 15:03:20 +00:00
Matthew Bonanni a608b4c6c2 [5/N][Attention] Finish eliminating vllm/attention folder (#32064) 
Signed-off-by: Matthew Bonanni <mbonanni@redhat.com>
 2026-01-27 10:02:51 -05:00
Nicolò Lucchesi 1f3a2c2944 [Bugfix] Disable CG for Whisper+FA2 (#33164) 
Signed-off-by: NickLucche <nlucches@redhat.com>
 2026-01-27 21:46:51 +08:00
omerpaz95andOmer Paz 7227d06156 [Metrics] [KVConnector] Add Offloading Connector metrics (#27942) 
Added queries and hits metrics for the Offloading Connector.

Also added timing metrics for store and load operations, which take the
average time it takes to load/store, per-token.

The metrics are available from Prometheus and from the StatLogger.

Signed-off-by: omerpaz95 <omerpaz95@gmail.com>
Co-authored-by: Omer Paz <Omer.Paz@ibm.com>
 2026-01-27 13:34:49 +00:00
Harry Mellor 14385c80fc Fix weight mapping test for Transfomers v5 (#33162) 
Signed-off-by: Harry Mellor <19981378+hmellor@users.noreply.github.com>
 2026-01-27 12:30:14 +00:00
wang.yuqiandCyrus Leung 76139d0801 [Frontend] Frontend will only attach supported tasks corresponding entrypoints. (#33139) 
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
Signed-off-by: wang.yuqi <noooop@126.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
 2026-01-27 12:15:43 +00:00
Lifan Shen da8d0c441a [AMD][QWEN3-NEXT] FP8 Tunings (#32042) 
Signed-off-by: Lifan Shen <lifans@meta.com>
 2026-01-27 09:34:13 +00:00
Roger Wangwanglinianwangln19Zaida ZhouIsotr0pyNick Hillyoukaichaogemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
 		b539f988e1 [Models] Kimi-K2.5 (#33131) 
Signed-off-by: wanglinian <wanglinian@stu.pku.edu.cn>
Signed-off-by: wangln19 <96399074+wangln19@users.noreply.github.com>
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Signed-off-by: youkaichao <youkaichao@gmail.com>
Signed-off-by: Roger Wang <hey@rogerw.io>
Co-authored-by: wanglinian <wanglinian@stu.pku.edu.cn>
Co-authored-by: wangln19 <96399074+wangln19@users.noreply.github.com>
Co-authored-by: Zaida Zhou <58739961+zhouzaida@users.noreply.github.com>
Co-authored-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: youkaichao <youkaichao@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
 2026-01-27 14:50:31 +08:00
Ning Xie b781eeaa15 [code clean] remove duplicate code (#33135) 
Signed-off-by: Andy Xie <andy.xning@gmail.com>
 2026-01-27 04:57:16 +00:00
Cyrus Leung e0b005d9cf [Frontend] Cleanup serving engine (#33103) 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
 2026-01-26 20:47:26 -08:00
Richard Zou 3b8f0fe59e [torch.compile] Stop assuming 32 bit indexing (#33113) 
Signed-off-by: Richard Zou <zou3519@gmail.com>
 2026-01-27 04:25:02 +00:00
Cyrus Leung c831911be2 [Frontend] Reduce mixin usage in serving pooling (#33101) 
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
 2026-01-27 11:50:37 +08:00
Paco Xu 157caf511b [Perf] avoid duplicate mem_get_info() call in get_current_memory_usage (#33064) 
Signed-off-by: Paco Xu <paco.xu@daocloud.io>
 2026-01-27 03:45:45 +00:00
Strahinja Stamenkovic c568581ff3 Fix IndexError with encoder-decoder models when using Custom Paged Attention (#33112) 
Signed-off-by: sstamenk <strahinja.stamenkovic@amd.com>
 2026-01-27 10:33:37 +08:00
wangln19Roger WangIsotr0py
 		2d7053438a fix: preserve native tool call ID in multi-turn tool calling (#32768) 
Signed-off-by: wanglinian <wanglinian@stu.pku.edu.cn>
Signed-off-by: wangln19 <96399074+wangln19@users.noreply.github.com>
Signed-off-by: Roger Wang <hey@rogerw.io>
Co-authored-by: Roger Wang <hey@rogerw.io>
Co-authored-by: Isotr0py <2037008807@qq.com>
 2026-01-27 10:22:35 +08:00