Skip to content
Navigation Menu
Sign in
Products
Platform
GitHub
Repository hosting, code review, and collaboration
Copilot
AI coding agent for tasks, projects and automated work
Build & Ship
Actions
CI/CD and automation
Code Quality
Quality analysis, fixes, and enforcement
Codespaces
Cloud development environments
Secure
Code Security
Vulnerability detection and remediation
Secret Protection
Secret scanning and leak prevention
All Features
Developers
Learn
Documentation
Training
Certifications
Build
Build with Copilot
Customize Copilot
Integrations
Explore
Trending repos
GitHub Sponsors
Events
Changelog
Changelog
Your source for the latest features, improvements, and fixes across GitHub products, all in one place.
Businesses
By Stage
Enterprises
Team
Startups
Nonprofits
By Industry
Healthcare
Financial services
Manufacturing
Government
See all industries
Evaluate GitHub
Why GitHub
Customer Stories
Business insights
Trust Center
Add-ons
GitHub Advanced Security
Enterprise-grade code security
Copilot for business
Enterprise-grade AI features
Premium Support
Enterprise-grade 24/7 support
Resources
Learn & Connect
Documentation
Resource center
Blog
Events & webinars
Community Forum
Customer Stories
Integrations
Support
Customer support
Training
GitHub Status
Company
About
Partners
Changelog
GitHub Universe
GitHub Universe
The event for developers, by developers — talks, workshops, and what’s next for GitHub.
Pricing
Search
/
Sign in
Sign up
Appearance settings
You signed in with another tab or window.
Reload
to refresh your session.
You signed out in another tab or window.
Reload
to refresh your session.
You switched accounts on another tab or window.
Reload
to refresh your session.
Dismiss alert
{{ message }}
vllm-project
vllm
Repository navigation
Code
Issues
2.7k
(2.7k)
Pull requests
5k+
(5k+)
Discussions
Actions
Projects
Security and quality
122
(122)
Insights
More
items
All pull requests
Labels
76
(76)
Milestones
1
(1)
New pull request
Search pull requests
is
:
pr
state
:
open
is:pr state:open
Clear filter
Search
Pull requests
Open
7,024
(7,024)
Closed
37,876
(37,876)
Author
Label
Projects
Milestones
Reviews
Assignee
Sort by
Newest
descending
Sort…
Sort by Newest descending
More items
Comfortable display density
Compact display density
[ROCm][CI] Cache a PR's ROCm base image before it merges with
/amd-ci cache
ci/build
documentation
rocm
#65276
·
AndreasKaratzas
opened
Nov 3, 2026
Member
·
·
1
[Bugfix][MRV2] Use promised max_query_len for encoder-only FULL CUDA graphs
bug
mrv2
nvidia
#65275
·
pst2154
opened
Nov 3, 2026
Contributor
·
·
[Bugfix] Step3 tool parser: coerce nullable and object/array parameters by schema
bug
ready
tool-calling
#65274
·
0xDexe
opened
Nov 3, 2026
·
4 of 6 tasks
·
·
1
6
[Quantization] Support ModelOpt W8A16 (INT8 weight-only) linear layers
documentation
quantization
#65272
·
cjluo-nv
opened
Nov 3, 2026
·
6 tasks done
·
·
2
[Bugfix][Scheduler] Don't split a multimodal item when a prefix cache hit ends inside it
bug
scheduler
#65271
·
Anieca
opened
Nov 3, 2026
·
6 tasks done
·
·
5
[MRV2][Spec Decode] Fix mixed draft-model warmup token budget
mrv2
#65270
·
wxsIcey
opened
Nov 3, 2026
Contributor
·
·
[Bugfix][Frontend] Decode each streaming-input chunk after its own prompt
bug
#65269
·
paolo-perrone
opened
Nov 3, 2026
Contributor
·
6 tasks done
·
·
[Feature] Segment attention layout for one-pass multi-candidate scoring
frontend
mrv2
performance
pooling
#65268
·
pst2154
opened
Nov 3, 2026
Contributor
·
·
[Perf][ROCm][MiniMax-M3.1] Add opt-in parallel decode top-k
ci/build
minimax
recipe
rocm
#65267
·
jiacao-amd
opened
Nov 3, 2026
Contributor
·
3 of 5 tasks
·
·
2
[Bugfix][Frontend] Share the max_num_queued_tokens count across API servers
bug
documentation
#65266
·
sohom-cs
opened
Nov 3, 2026
Contributor
·
5 of 6 tasks
·
·
1
1
[Bugfix][Core] Fix Gemma-4 MTP prefix-cache hit loss when lookahead block hashes are off
bug
kv-cache-manager
#65263
·
xuanzic
opened
Nov 3, 2026
·
5 of 6 tasks
·
·
1
[Feature][KV Connector][Mooncake] Export store usage metrics for MooncakeStoreConnector
documentation
kv-connector
mooncake
#65262
·
skaulintel
opened
Nov 3, 2026
·
2 of 6 tasks
·
·
2
[MoE][Model Loader] Fused fast path in RoutedExperts.load_weights; HY-V3 delegates expert loading
ci/build
#65261
·
hao-aaron
opened
Nov 3, 2026
Collaborator
·
·
[Bugfix][KV Connector] Forward num_threads to non-UCX NIXL backends
bug
documentation
kv-connector
nixl
#65260
·
mujtaba1747
opened
Nov 3, 2026
·
5 of 9 tasks
·
·
2
[RL][Model Loader] Make expert scale and Mamba v2 sharded loaders copy_-only
ci/build
#65259
·
zzhisyyds2
opened
Nov 3, 2026
·
6 tasks done
·
·
1
[Doc] MiniMax-M3.1-Flash: drop DEP4 c96 and c128 from the GB300 AgentX presets
minimax
recipe
#65258
·
hshrivastava-droid
opened
Nov 3, 2026
Contributor
·
·
1
[ROCm] Split-K bf16x3 router GEMV for fp32 router gates at 33-2047 tokens
rocm
#65257
·
micah-wil
opened
Nov 3, 2026
Collaborator
·
6 tasks
·
·
[Quantization] Keep E8M0 scale bytes when loading into uint8 params
ci/build
quantization
#65256
·
hao-aaron
opened
Nov 2, 2026
Collaborator
·
·
[MoE][RL] MegaMoE NVFP4: load each scale on its own, combine after loading
minimax
#65254
·
hao-aaron
opened
Nov 2, 2026
Collaborator
·
·
[Bugfix][Benchmark] Preserve inactive expert IDs in synthetic forward routing
bug
ci/build
documentation
performance
ready
#65253
·
WoosukKwon
opened
Nov 2, 2026
Collaborator
·
6 tasks done
·
·
3
[Bugfix][Frontend] Fix 500s on degenerate previous_input_messages content in Harmony Responses path
bug
frontend
gpt-oss
#65252
·
herniqeu
opened
Nov 2, 2026
·
·
1
[ROCm] Move AITER MXFP4 linear op registration into _aiter_ops
batch-invariance
rocm
#65251
·
afriedri
opened
Nov 2, 2026
Contributor
·
·
[WIP][ROCm][DSv4.1] Enable prefix-cache SWA replay on gfx942
deepseek
DSv4.1
kv-cache-manager
rocm
#65249
·
frida-andersson
opened
Nov 2, 2026
Contributor
·
6 tasks done
·
·
[ROCm] AITER gluon attention backend for Qwen3.8-27B
qwen
rocm
#65248
·
eky-amd
opened
Nov 2, 2026
Contributor
·
4 of 6 tasks
·
·
[Bugfix][Frontend] Stop Hermes streaming tool parser leaking keys that follow "arguments"
bug
tool-calling
#65247
·
stanley-zhou
opened
Nov 2, 2026
·
5 of 6 tasks
·
·
1
1
Previous
1
2
3
4
5
6
7
…
40
Next
You can’t perform that action at this time.