Skip to content

Pull requests: vllm-project/vllm

New pull request New
3,034 Open 23,896 Closed
3,034 Open 23,896 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[Misc] downgrade nvidia-cutlass-dsl to 4.5.0 ready ONLY add when PR is ready to merge/full CI is needed
#43230 opened May 20, 2026 by ZJY0516 Member Loading…
4 tasks
[CompressedTensors] FP4 Qutlass Integration
#43229 opened May 20, 2026 by kylesayrs Contributor • Draft
2
Improve error handling for unavailable NIXL backends kv-connector
#43228 opened May 20, 2026 by vidyuth-krishna Loading…
3
[Bugfix][Gemma4MoE] Fix AutoRound quantized Gemma4 MoE loading bug Something isn't working
#43227 opened May 20, 2026 by wxwxwwxxx Loading…
1
[CPU] Experimentally enable Triton and MRV2 ci/build cpu Related to CPU backends v1
#43225 opened May 20, 2026 by bigPYJ1151 Member • Draft
4 tasks
2
Fix FlashInfer TRTLLM NvFP4 monolithic MoE routing nvidia
#43223 opened May 20, 2026 by zhangxin81 Contributor Loading…
2
[EPLB] Make async EPLB default ci/build documentation Improvements or additions to documentation
#43219 opened May 20, 2026 by ilmarkov Contributor Loading…
4 tasks
2
[ROCm] MoRI connector telemetry kv-connector rocm Related to AMD ROCm
#43218 opened May 20, 2026 by simondanielsson Contributor • Draft
4 tasks
5
[Misc] Add exponential distribution to multi-turn benchmark performance Performance-related issues
#43217 opened May 20, 2026 by nikonyrh-siloai Loading…
4 tasks done
4
fix(chat): allow multimodal content in tool messages for vision models frontend
#43216 opened May 20, 2026 by anishesg • Draft
1
2
[Misc] Add --max-duration-sec to benchmark_serving_multi_turn.py performance Performance-related issues
#43215 opened May 20, 2026 by nikonyrh-siloai Loading…
3 of 4 tasks
3
[Perf][DSv4] Add cuteDSL generic LL Blockwise FP8 GEMM
#43214 opened May 20, 2026 by LopezCastroRoberto Contributor • Draft
2
[Model] Fix MiniCPM-V 4.6 vit_merger qkv weight loading
#43213 opened May 20, 2026 by tc-mb Contributor Loading…
3
[Bugfix] Fix multi-turn benchmark's sleep to match the configured request rate bug Something isn't working performance Performance-related issues
#43212 opened May 20, 2026 by nikonyrh-siloai Loading…
3 of 4 tasks
5
[ZenCPU] Support TorchAO (DA8W8) and LLM-Compressor (DA8W8, W4A16) config for zen cpu platform ci/build
#43211 opened May 20, 2026 by aadwived • Draft
4 tasks
3
[Bugfix][Reasoning] Properly detect reasoning end when using thinking_token_budget bug Something isn't working v1
#43210 opened May 20, 2026 by schoennenbeck Contributor Loading…
2
[7/n] Migrate pos_encoding and norm kernels to libtorch stable ABI (continued) ci/build
#43209 opened May 20, 2026 by cleonard530 Contributor Loading…
5 tasks
20
[Docs] Add drain shutdown section to Kubernetes deployment guide documentation Improvements or additions to documentation
#43208 opened May 20, 2026 by markmc Member Loading…
5
[KV Offload] Add get_request_offloading_context lifecycle hook kv-connector v1
#43205 opened May 20, 2026 by ronensc Contributor Loading…
4 tasks
16
[Cleanup]Simplify UnitaryKVCacheCoordinator hash_block_size assert v1
#43204 opened May 20, 2026 by maang-h Contributor Loading…
[Feature] Add External LB Mode Support for Elastic EP Scaling v1
#43202 opened May 20, 2026 by fangyuchu Contributor • Draft
3
[Spec Decode] Add FlashInfer metadata grouping for DFlash SWA needs-rebase nvidia qwen Related to Qwen models speculative-decoding v1
#43200 opened May 20, 2026 by gq112 Loading…
5
[vLLM IR][Rope] Port RotaryEmbedding to IR Ops cpu Related to CPU backends intel-gpu Related to Intel GPU nvidia rocm Related to AMD ROCm
#43199 opened May 20, 2026 by wxsIcey Contributor Loading…
4 tasks
6
【Feature】Modify the fps parameter when loading the multimodal model Video. bug Something isn't working multi-modality Related to multi-modality (#4194)
#43198 opened May 20, 2026 by lucky-dep Loading…
3
[CI] De-flake test_models for bigscience/bloom-560m ready ONLY add when PR is ready to merge/full CI is needed
#43197 opened May 20, 2026 by haosdent Contributor Loading…
13
ProTip! Updated in the last three days: updated:>2026-05-17.