V

vllm

Settings | Report Duplicate

0

I Use This!

Very High Activity

Commits : Listings

Analyzed 1 day ago. based on code collected 2 days ago.

Commit Message	Contributor	Files Modified	Lines Added	Lines Removed	Code Location	Date
Oct 27, 2025 — Oct 27, 2026 Showing page 1 of 797 Search / Filter on:
[Bugfix] Fix Triton prefill attention OutOfResources on GPUs with 99KiB shared memory (#60942)	Vadim Gimpelson	More...				2 days ago
[Bugfix][Spec Decode] Apply the draft max_model_len limit before model config hooks (#63853)	Prakash Saini	More...				2 days ago
[Bugfix][ROCm][DSv4] Give zero and subnormal mono MXFP8 groups vLLM's scale code (#63689)	Andreas Karatzas	More...				2 days ago
[Bugfix][MRV2] Release the warmup memory cached on the capture streams (#63154)	Qing	More...				2 days ago
[Spec Decode][Sampling] Batch-sharded draft logits (#63554)	Giancarlo Delfin	More...				2 days ago
[Doc] Kimi-K3: refresh the GB300 AgentX xPyD preset and add the 3P1D DEP16 preset (#63852)	Ankur Singh	More...				2 days ago
[Perf][Kimi K3] Opt-in KV-balanced FP8 MLA decode for agentic workloads (#63638)	Wei Zhao	More...				2 days ago
[CI] List one test path per line in Buildkite pytest commands to avoid merge conflicts (#63408)	Oxana Korzh	More...				2 days ago
[NIXL] Bump NIXL version to 1.5.1 (#63817)	Jie Hao	More...				2 days ago
[Tool Calling] Migrate llama3_json parser to parser engine (#63569)	Artem Perevedentsev	More...				2 days ago
[ROCm][CI] Release checkpoint page cache between language generation … (#63828)	djramic	More...				2 days ago
[CI][Frontend] Let the SSL refresher test's fake server accept peer_loads (#63840)	Kevin H. Luu	More...				2 days ago
[Perf][MoE] Size the DeepGEMM MegaMoE symmetric buffer for one sequence-parallel shard (#63541)	mispa-ms	More...				2 days ago
[CI] Add missing `&&` in the Multi-Modal Models (Standard) 3 command (#63422)	Oxana Korzh	More...				2 days ago
[Kimi-K3] Route the fp8 chunked-context K/V pack through FlashInfer where it has a fused kernel (#61527)	Brandon Zhang	More...				2 days ago
[Spec Decode] Replace torch.compile'd AV draft-budget selection with a Triton kernel (#63421)	Misha Goin	More...				2 days ago
[Bugfix] Keep legacy WNA16 MoE dispatch on Triton under batch invariance (#61038)	yifanFengg	More...				2 days ago
[BugFix] [Build] Fix for where setup.py fails with MAX_JOBS invalid literal error (#45188)	rasmith	More...				2 days ago
[Bugfix][Tool Parser] Anchor Jamba streamed tool arguments to what was sent (#55669)	Zhiqi Zhang	More...				2 days ago
[KVConnector][Mooncake] Coalesce descending block ids to reduce descriptor number for Mooncake TE (#63142)	Yuhan Zhou	More...				2 days ago
[Bugfix][Spec Decode] DFlash: run the aux hidden state FC in the draft dtype (#63436)	azaidy	More...				2 days ago
[XPU][MoE] Resize batched Triton MoE tiles and change grid order (#51605)	Lena Onyshchenko	More...				2 days ago
[Perf][MXFP4] Optimize activation quant for mxfp4. 3x kernel speedup, 5% throughput improvement (#63101)	Ilya Markov	More...				2 days ago
[Bugfix][Frontend] Stop SSL certificate watchers when HTTP serving exits (#55699)	jxj	More...				2 days ago
[Perf][HiSparse] Reduce residency hash probes (#61263)	Itay Etelis	More...				2 days ago
[Bugfix][Kernel] persistent_topk: sync the CTA around radix barrier arrivals (#63475)	Andreas Karatzas	More...				2 days ago
[MoE][Model] Gemma 4: skip CUDA-graph padding rows in the routing kernel (#62525)	Blaze-DSP	More...				2 days ago
[XPU][FP8] Use per-token dynamic FP8 activation quantization (#61362)	liuzhenwei	More...				2 days ago
[KV Offload] Expose the processed prefix at request finalization (#57646)	Moein Khazraee	More...				2 days ago
[Spec Decode] Support PCP with DSpark for DeepSeek-V4.1 (#61994)	Chauncey	More...				2 days ago

←
1
2
3
4
5
6
7
8
9
…
796
797
→