vllm
https://github.com/vllm-project/vllm
Python
A high-throughput and memory-efficient inference and serving engine for LLMs
Triage Issues!
When you volunteer to triage issues, you'll receive an email each day with a link to an open issue that needs help in this project. You'll also receive instructions on how to triage issues.
Triage Docs!
Receive a documented method or class from your favorite GitHub repos in your inbox every day. If you're really pro, receive undocumented methods or classes and supercharge your commit history.
Python not yet supported59 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Bugfix][Mamba] Support speculative decoding in Mamba-1 layers
- [ROCm] Pass the (B, next_n) context lengths to AITER paged MQA logits
- [Bugfix][Model] Pass Solar skip connection states to the next pipeline stage
- [Bugfix][Step5][Spec Decode] Skip VllmConfig re-validation for the MTP draft config
- [RL][Model Loader] Direct reload: models opt in; refuse known-silent paths
- [ROCm][Perf] AITER FA: build the extend metadata without a host sync
- [Bugfix][Frontend] Muse reasoning parser: deliver the visible text when a reply ends behind a stray recipient or inside a truncated tool call
- [Bugfix][Model] Fix GraniteMoeHybrid loading and running only its own layers under pipeline parallelism
- [Bugfix][MoRIIO] Let decode renew READ block leases on prefill
- [Bugfix][EPLB] Fix startup hangs with a pipeline stage that has no MoE layers
- Docs
- Python not yet supported