Skip to content

Pull requests: vllm-project/vllm

New pull request New
1,181 Open 13,982 Closed
1,181 Open 13,982 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

[CI] Install pre-release version of apache-tvm-ffi for flashinfer ci/build ready ONLY add when PR is ready to merge/full CI is needed
#27262 opened Oct 21, 2025 by hmellor Loading…
5
Feature: Support Relu2 in FusedMoE fp8 cutlass path
#27261 opened Oct 21, 2025 by amirkl94 • Draft
2
Remove last level references not removed in #26355 llama Related to Llama models
#27260 opened Oct 21, 2025 by hmellor Loading…
1
Fix EventPublisherFactory logic for disabled KV cache events and publisher is "null"
#27257 opened Oct 21, 2025 by usberkeley Loading…
5 tasks done
1
Bugfix: Cutlass FP8 FusedMoE
#27255 opened Oct 21, 2025 by amirkl94 Loading…
3
Updated xgrammar backend to not deny supported string formats structured-output v1
#27253 opened Oct 21, 2025 by ExtReMLapin Loading…
3 of 5 tasks
5
[MISC] Add prefix cache reset to LMCache CPU offload example documentation Improvements or additions to documentation kv-connector
#27248 opened Oct 21, 2025 by sakunkun Loading…
5 tasks
3
[WIP][Model] Upstream Deepseek-OCR model deepseek Related to DeepSeek models new-model Requests to new models
#27247 opened Oct 21, 2025 by Isotr0py • Draft
5 tasks
1
[WIP] Support DeepSeek-OCR deepseek Related to DeepSeek models documentation Improvements or additions to documentation new-model Requests to new models
#27246 opened Oct 21, 2025 by Xu-Wenqing • Draft
5 tasks
1
[CPU]Improve cpu fused moe perf
#27244 opened Oct 21, 2025 by xiangze-arm Loading…
4
Mirroring changes in test-pipeline.yaml into test-amd.yaml ci/build rocm Related to AMD ROCm
#27242 opened Oct 21, 2025 by Alexei-V-Ivanov-AMD Loading…
7
[Bug] Qwen reasoning parser qwen Related to Qwen models ready ONLY add when PR is ready to merge/full CI is needed
#27241 opened Oct 21, 2025 by ahao-anyscale • Draft
5 tasks
[CPU]Improve dynamic 4bit moe performance
#27240 opened Oct 21, 2025 by xiangze-arm Loading…
6
v1/kv_cache_utils: Respect num_gpu_blocks_override in memory check v1
#27238 opened Oct 21, 2025 by khaled-wsa Loading…
3
[CI/Testing] Add basic single node dual batch overlap test ci/build v1
#27235 opened Oct 21, 2025 by LucasWilkinson Loading…
6
[ResponseAPI] Fix mcp tool type extraction frontend gpt-oss Related to GPT-OSS models
#27234 opened Oct 21, 2025 by Jialin Loading…
3 of 5 tasks
7
[Bugfix] Ensure calculated KV scales are applied in attention. v1
#27232 opened Oct 21, 2025 by adabeyta Loading…
11
Adds runai distributed streamer ci/build documentation Improvements or additions to documentation rocm Related to AMD ROCm
#27230 opened Oct 20, 2025 by bbartels Loading…
5 tasks
24
[Feature] Batch Invariant for R1 TP 8 on Blackwell ready ONLY add when PR is ready to merge/full CI is needed
#27229 opened Oct 20, 2025 by yewentao256 Loading…
1
[ROCm][MLA] Support block-size > 1 for AITER MLA backend rocm Related to AMD ROCm v1
#27224 opened Oct 20, 2025 by ganyi1996ppo Loading…
5 tasks
5
Flashinfer_CUTLASS_MOE fuses quantization for TP
#27223 opened Oct 20, 2025 by wenscarl Loading…
5 tasks
4
ARM64 CUDA 12.9 wheels built and uploaded to index incorrectly ci/build
#27221 opened Oct 20, 2025 by Gregory-Pereira Loading…
1
[Bugfix] Fix dp_chunking enablement logic in FusedMoE layer
#27220 opened Oct 20, 2025 by alexm-redhat Loading…
@alexm-redhat
10
[CORE] Support Prefix Caching with Prompt Embeds documentation Improvements or additions to documentation v1
#27219 opened Oct 20, 2025 by qthequartermasterman Loading…
3 of 5 tasks
1
10
[Backend][WIP] Integrate MPK (Mirage) compiler as an experimental execution backend to vLLM v1
#27218 opened Oct 20, 2025 by NorthmanPKU • Draft
1 of 8 tasks
12
ProTip! What’s not been updated in a month: updated:<2025-09-21.