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] Make beam search with structured outputs reachable
- [Perf] Let a grammar hand beam search its allowed tokens directly
- [Quantization] Support tensor parallelism for block-diagonal compressed-tensors Hadamard transforms
- [CI] Make DSpark sharded-sampling test deterministic under torch.compile
- [Bugfix][Kernel] Use 64-bit row offsets in the persistent and cooperative top-k kernels
- [Spec Decode] Remove the one-chunk cap on adaptive verification
- [Feature][Attention] Add NIXL GPU Device API for MLA DCP A2A
- [Bugfix][Frontend] Keep the sampled token's entry in /v1/completions top_logprobs
- [Bugfix][Frontend] Normalize Chat Completions tools and tool_choice validation
- [ROCm][Perf][Muse Spark] fp8 (e4m3) KV cache for the decode monokernel
- Docs
- Python not yet supported