Commit Graph

Select branches

Hide Pull Requests

main

patches

sm86

Mono Color

• a19ff2218a [Hardware][AMD][CI] Fix Spec Decode Eagle test group (#46018) Matt 2026-06-21 17:40:02 -05:00
• 4f0d0049a0 [Hardware][AMD][CI] Fix Kernels Attention test groups (#46080) Matt 2026-06-21 17:10:51 -05:00
• 13b83d77ad [ROCm][CI] skip test_double_aiter_rms_quant_fusion (#45967) Charlie FuandAndreas Karatzas 2026-06-21 16:53:11 -05:00
• 50241602fd [Hardware][AMD][CI] Fix gfx942 Kernels MoE test group (#46298) Matt 2026-06-21 16:45:37 -05:00
• 12fe2a9aac [Bugfix][Qwen3-VL] Fix multi-video crash with list-valued fps/num_frames (#46305) Ting SUN 2026-06-22 04:31:23 +07:00
• 89bd2c14d3 [Spec Decode] Add Qwen3 architecture support for EAGLE3 (#43132) Benjamin Chislett 2026-06-21 16:55:26 -04:00
• 9c450b1027 [Kernel][Bugfix] Fix INT8 per-token-head KV cache rounding in Triton reshape-and-cache (#45361) ZedongLiu 2026-06-22 03:59:40 +08:00
• 635c38338a [Multimodal] Add Qwen2-VL/Qwen2.5-VL processor-mapped video loader (#45555) 3 people 2026-06-21 13:56:50 -05:00
• c441ad1c07 [KV Offloading] Add labeled metrics support (#45957) Srinivas Krovvidiandsrinivas_oo7 2026-06-21 13:04:01 -05:00
• 745bba5ea8 [Model]Fix MiniMaxM2ForCausalLM perf regression (#45935) Jee Jee Li 2026-06-22 00:28:52 +08:00
• 2cac89f9da [Spec Decode] Support mixed KV page sizes for DFlash (#45181) 4 people 2026-06-21 07:45:14 -07:00
• 3e6e33526d [Disagg] return routed_experts on streaming generate responses (#44638) 3 people 2026-06-21 22:37:10 +08:00
• b91b7726e0 [ROCm][P/D] Support MiniMax-M3 mixed KV layouts in MoRIIO READ mode (#46039) 9 people 2026-06-21 20:55:19 +08:00
• d3ad8e8bcd [Bugfix] Defer offload reads while transfers are pending (#46231) Palaiologos1453 2026-06-21 19:30:13 +08:00
• b80ce9dd2f [CI][test] Replace InternVL2-1B with InternVL3-1B in test_pipeline_parallel.py (#46241) xiaolinchenandwentian-byte 2026-06-21 15:11:19 +08:00
• b5495cc5f9 Fix memory pointer overflow in Mamba state buffers (#44665) Shifani RajaboseandKunshang Ji 2026-06-21 02:00:50 -04:00
• 183a430c13 [Bugfix][Model Runner V2] Fix min_tokens off-by-one in the V2 GPU sampler (#46243) Ting SUN 2026-06-21 12:06:49 +07:00
• a346d589f5 [Bugfix] Fix NVFP4/OCP MX MoE emulation (#46254) Matt 2026-06-20 23:13:10 -05:00
• 7df3d7dada [Core] Ensure memory is pinned prior to async h2d copy (#45424) Nick Hill 2026-06-20 20:02:24 -07:00
• 8dd1b702f2 [Misc] Fix stale doc URL and docstring module path (#35530) Umut PolatandFlora Feng 2026-06-21 02:57:01 +03:00
• f57ac274b2 [Render] Add reasoning/tool parsing to /derender + fix byte-fallback FFFD (#45919) aoshen02andMartin Hickey 2026-06-21 07:43:32 +08:00
• 6e919960af [Perf] Skip/shrink all_token_ids copy in scheduler for non-async and V2 runner (#45840) 3 people 2026-06-20 23:36:57 +01:00
• c88d3d4775 [SimpleCPUOffloadConnector] PCP + DCP support (#39831) Jonathan Chen 2026-06-20 18:01:06 -04:00
• ab7fcbdd5d [Perf][KVConnector][Mooncake] Compact chunk-hash keys and zero-copy lookup wire format (#45969) Yifan Qiao 2026-06-20 15:00:11 -07:00
• 3b4a76b63f [KV-Offloading] : Expose CPU cache usage metric (#45737) Varun Sundar RabindranathandVarun Sundar Rabindranath 2026-06-20 17:21:55 -04:00
• cc22621b51 [KV Offload] Support packed HMA KV cache layout (#46205) 3 people 2026-06-20 17:19:40 -04:00
• 77148992cf [Bugfix] Move extract_layer_index back inside is_v32 guard (#46199) Tyler Michael SmithandClaude Opus 4.6 2026-06-20 17:19:10 -04:00
• 891cc4b9c5 [Frontend] Report cache usage in Anthropic /v1/messages API (#40912) shuoming zhangandTyler Michael Smith 2026-06-21 05:12:48 +08:00
• 1bdf9810aa [ROCm] [Bugfix] Bugfix ROCm Sparse Indexer (#46222) TJian 2026-06-21 04:38:42 +08:00
• ebfbcfe46a Stop setting CUDA_VISIBLE_DEVICES internally in vLLM, add device_ids arg (#45026) 5 people 2026-06-20 16:38:10 -04:00
• e9de72fe6c [Bugfix] Guard model_config access in _log_compilation_config (#46198) Tyler Michael SmithandClaude Opus 4.6 2026-06-20 15:26:38 -04:00
• d272418f45 [Perf] Optimize Qwen3-VL multi-video prompt processing (#46026) L丶andmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-20 22:09:18 +08:00
• 7ff7f5c8eb Revert "Fix Stale Encoder Cache After Weight Update" (#46125) Sumanth R Hegde 2026-06-20 07:09:09 -07:00
• dced290769 [Hardware][AMD][CI] Fix e2e core test group (#46024) Mattandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-20 02:04:35 -05:00
• 93bad11912 [Bugfix] Fix gridDim.y overflow for large row counts (#45255) JasonLi314 2026-06-19 20:27:45 -07:00
• 0fbf42af84 [ROCm] Fix VRAM not freed in test_phi3v (#46046) djramic 2026-06-20 00:20:59 +02:00
• e6cd8913dd [ROCm][CI] Skip Qwen3.5-35B-A3B-MXFP4-AITER-TP2 for non gfx950 (#46109) Charlie Fu 2026-06-19 17:20:10 -05:00
• 859e4d436b [Bugfix][Parser] Fix U+FFFD leak at reasoning-to-content transition in engine parsers (#46159) Ben Browning 2026-06-19 18:09:28 -04:00
• 4a083cc858 [ROCm][CI] Pin test_rocm_compressed_tensors_w8a8 to TRITON_ATTN (#46180) Micah Williamson 2026-06-19 15:20:06 -05:00
• ca7e1f2c43 Move CI failure diagnosis docs into ci-fails-buildkite skill (#45975) Vadim Gimpelson 2026-06-20 00:12:40 +04:00
• dec860fb19 [ROCm] Use vLLM's fp8 quant max in AITER hipBLASLt accuracy test (#46176) djramic 2026-06-19 20:24:02 +02:00
• 0a49fb2b13 Fix dead link in docs (#46181) Harry Mellor 2026-06-19 19:16:09 +01:00
• 4a8abf37c7 [Test] Migrate test_openai_schema.py to schemathesis 4.x (#46173) Ben Browning 2026-06-19 14:05:18 -04:00
• 01192139bf [DSv4] Pack KV caches into contiguous per-block allocations for DeepSeek V4 (#44577) 6 people 2026-06-19 12:55:42 -04:00
• b9a7cd464c [12/n] final _C library kernel migration (#45415) Chris Leonard 2026-06-19 09:57:26 -04:00
• 69bdd34542 [Bugfix] Fall back to Pydantic loc for param in validation errors (#46038) Muhammad FawazandMahad Durrani 2026-06-19 16:11:11 +05:00
• ec67d7ae61 [xpu] bump up vllm-xpu-kernels v0.1.10 and upgrade 2618 umd (#40367) Kunshang Ji 2026-06-19 15:37:20 +08:00
• ecf9d83520 [AMD][CI] Fix Language Models Test (Extended Generation) failures (#45509) 3 people 2026-06-18 22:06:56 -06:00
• c9135db27c [Docs] Update stale LMCache examples (#45762) Samuel Shen 2026-06-18 20:21:36 -07:00
• 2a6c6b9429 [DeepSeek-V4] Support TEP=16 for the block-FP8 shared expert (#46001) Jeff (Junze) Maandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-18 20:10:12 -07:00
• ab66606993 [bugfix]Indexer init skip and MTP TopK share for iteration (#45895) Jared Wen 2026-06-19 09:57:51 +08:00
• 9ea3a4015b [Bugfix] Fix corrupt outputs in MoE FP8 LoRA responses and MoE base model responses when LoRAs are loaded (#42120) 3 people 2026-06-18 20:26:09 -05:00
• 560fb8b867 [Cohere] Remove dead prepare_structured_tag override in Cohere parser (#46099) Flora Feng 2026-06-18 21:02:11 -04:00
• 675cd5d228 [Model Runner V2] Fix MRv2 memory leak test (#46095) Wentao Ye 2026-06-18 20:36:40 -04:00
• 7f616c327d [Bugfix] [Parser] Fix empty tool block silently dropping subsequent content (#46091) Ben BrowningandFlora Feng 2026-06-18 19:17:18 -04:00
• c3c6d723fd [Perf] Remove unused loggers in reasoning/ (#45988) Ivy Xu 2026-06-19 06:24:29 +08:00
• 41dcf49ca5 [Bugfix][KV Connector] Disable Mooncake TP put-striding when DCP > 1 (#45371) 3 people 2026-06-18 15:13:44 -07:00
• 35e4dd4a69 [KV Connector][Mooncake] Async lookup to reduce scheduler overhead (#45659) Yifan QiaoandNick Hill 2026-06-18 14:44:02 -07:00
• 4ce2d01453 fix(anthropic): auto-detect template support for mid-conversation system messages (#46025) 3 people 2026-06-19 04:19:11 +08:00
• 16908e132e [MRV2] Make FP32 Gumbel sampling more accurate (#45996) Woosuk Kwon 2026-06-18 12:42:09 -07:00
• 225936a1dd [CI Bug] Revert #42379 to fix CI Multi-Modal Models (Extended Generation 1) (#46070) Wentao Ye 2026-06-18 15:37:39 -04:00
• f6ba720963 (security) Upgrade Starlette to >= 1.0.1 to fix CVE-2026-48710 (#45675) Juan Pérez de AlgabaandIsotr0py 2026-06-18 21:35:13 +02:00
• b53b1c7ffe [Model Runner V2] Migration to support quantized model by default [5/N] (#44446) Wentao Ye 2026-06-18 15:20:44 -04:00
• 79ca54d221 [Bugfix][Quantization] Don't reject fp8_e5m2 KV cache for non-fp8 quantized checkpoints (#45040) Ting SUNandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-19 02:18:25 +08:00
• 09f3cd5c10 [Bugfix] [Parser] Fix Qwen3 latent bug in partial params dropping values containing < (#46047) Ben Browning 2026-06-18 14:04:06 -04:00
• ea6078fe6a [KV Connector][Offloading] Disable parallel-agnostic fs-tier cache on V2 model runner (#46044) Itay EtelisandItay Etelis 2026-06-18 20:43:35 +03:00
• a0df04e477 [Tests] Add Qwen3 streaming parser delta boundary cases (#45708) Palaiologos1453 2026-06-19 01:37:39 +08:00
• e2352c2974 [ROCm][Spec Decode] Fix probabilistic draft probs test attention backend (#45706) stefankoncarevic 2026-06-18 18:59:37 +02:00
• 25faa1f4cc [CI]Enable mxfp4 lora test for ROCm platform (#43802) qli88 2026-06-18 11:59:09 -05:00
• 4583630b56 [Bugfix][Kernel] Check output alignment in vectorize_with_alignment (fixes misaligned-address crash for non-multiple-of-8 head sizes) (#45466) Humphrey 2026-06-18 11:58:22 -05:00
• 21da47dabe [ROCm][CI] move lora%N test to mi300 and gate (#45970) Divakar Verma 2026-06-18 12:50:32 -04:00
• 6c379b9e54 [Frontend] Add Streaming Parser Engine and new GLM4.7/GLM5.1/GLM5.2 Parser (#45915) Chaunceyandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-19 00:42:10 +08:00
• 5099474633 [Bugfix][ROCm] Fix rocm_aiter_per_tensor_quant custom op aliasing (#45747) Rohan Potdar 2026-06-18 11:30:21 -05:00
• 058cc0a8b6 [Bugfix] Restore is_sym guard for zp in GPTQ/CT MoE to fix symmetric quant regression (#45656) Yuwen Zhou 2026-06-19 00:20:29 +08:00
• 837db7605e [Bugfix][Tool Parser] Handle non-finite numbers in coerce_to_schema_type (#43984) Ashish PatelandBen Browning 2026-06-18 21:30:20 +05:30
• bf2a393034 Temporarily remove @markmc from CODEOWNERS (#46053) Mark McLoughlin 2026-06-18 15:15:43 +01:00
• d682968aa9 [Model] Remove BambaForCausalLM (#45990) Tiezhen WANGandClaude 2026-06-18 21:51:00 +08:00
• 021cdf72bc Fix _riscv_supports_rvv_vlen128() to detect RVV on hardware without zvl flags (#43179) lyd1992andYuanSheng 2026-06-18 21:22:35 +08:00
• 4cb5e746b6 [Rust Frontend]: Add /get_world_size route with static parallel size (#44801) Ashar 2026-06-18 18:40:20 +05:30
• 22cc891108 [Kernel] Add PDL support for DeepGEMM kernel (#46006) Jee Jee Li 2026-06-18 20:49:01 +08:00
• afdcbd5d39 [ROCm][DSv4] Functional fixes for DeepSeek V4 on MI300X/MI325X (#45681) 5 people 2026-06-18 15:21:14 +03:00
• 8d4f54966c fix(quantization): Fix AWQ dequantize on Intel XPU and refactor AutoAWQ config (#42727) 4 people 2026-06-18 20:12:28 +08:00
• 351c72d6e5 [CPU] Skip Triton kernel monkey-patches when Triton-CPU is available (#44991) Jonathan MamouandLi, Jiang <jiang1.li@intel.com> 2026-06-18 13:59:30 +03:00
• 7299e6509e [Rust Frontend] Return model metadata fields in /v1/models (#45950) Tahsin Tunan 2026-06-18 16:29:21 +06:00
• 08985351f3 Fix Stale Encoder Cache After Weight Update (#45093) littlecircle0730 2026-06-18 02:32:10 -07:00
• 5fd3b276f8 [Mooncake] Skip KV lookup for non-reachable SWA blocks (#45444) Wei Zhao 2026-06-18 05:23:20 -04:00
• 1e9f04da14 fix(anthropic): preserve inline system message position for prefix caching (#44602) MrFanandfelix0080 2026-06-18 15:58:11 +08:00
• 702214146c [Bugfix][Frontend] Fix Anthropic count_tokens decorator order driving server load negative (#44725) Ting SUNandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-18 14:56:46 +08:00
• a331589394 [XPU] Update nixl to v0.10.1 in Dockerfile (#40287) liuzhenweiandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-18 14:01:26 +08:00
• e945169207 Revert "[Kernel] Add PDL support for DeepGEMM kernel" (#45999) Micah Williamson 2026-06-18 00:59:48 -05:00
• 554352a311 [Test][KV Connector] Add request_finished fence population tests for offloading scheduler (#45679) AlexHuangandOr Ozeri 2026-06-18 13:13:52 +08:00
• 421c1ec448 [KV Offloading] Remove dummy worker-side stats from OffloadingConnector (#45905) AlexHuangandOr Ozeri 2026-06-18 13:13:28 +08:00
• b4c80ec0fd [Refactor] Remove dead cutlass mxfp8 code (#44681) Wentao YeandShengqi Chen 2026-06-18 00:18:25 -04:00
• f428718ffe [Fix][KV offload] Defer on_request_finished until in-flight transfers drain (#45823) Ronen Schaffer 2026-06-18 07:05:46 +03:00
• 4403af8fb5 [Kernel] Add PDL support for DeepGEMM kernel (#42996) Jee Jee Li 2026-06-18 11:37:17 +08:00
• d57888efa4 [SimpleCPUOffloadConnector]: Add support for reset_cache() (#39726) Jonathan Chenandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-17 22:47:12 -04:00
• ed938ad7db [CPUOffloading] Guard CPU eviction check (#45757) Varun Sundar RabindranathandVarun Sundar Rabindranath 2026-06-17 22:34:59 -04:00
• 731fb3323d [Rust Frontend] Validate tokenized bad_words vocabulary range (#45876) Reid 2026-06-18 10:28:45 +08:00
• 8dd8b6ed78 [XPU] Fix FP8 block-scaled scheme selection on non-CUDA platforms (#43958) Yejing LaiandKunshang Ji 2026-06-18 10:16:20 +08:00
• e1a5fc406b [Rust Frontend][Perf] O(n) argument scan in tool parser (#45826) Bugen Zhaoandmergify[bot] <37929162+mergify[bot]@users.noreply.github.com> 2026-06-18 09:42:35 +08:00