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
- [ROCm] Read the DSv4.1 mono target arch from FLYDSL_GPU_ARCH
- GLM-5.3-Flash FP4-Local in production on 2x RTX PRO 6000 (SM120): what's missing for a full community-main base?
- [ROCm][Perf][GLM-5.3-Flash] Keep the fused gated RMSNorm for the KDA o_norm
- [Rust Frontend] Match response_template regexes with PCRE2 in the hf parser
- [ROCm][Perf] Pick the GDN chunk-state V tile from the batch, not the first autotune
- [Feature][Spec Decode] Add spec config to shorten drafts on prefill-heavy steps
- [Perf] Fill unconstrained bitmask rows from a cached template
- [Bug][Perf]: W4A16 HummingLinearKernel is non-deterministic across processes and faster than the default Machete on H100 (sm_90)
- [Bugfix] Apply max_tokens caps to explicit null on generate
- [Refactor][KV Cache] Remove EAGLE block dropping and default to lookahead hashes
- Docs
- Python not yet supported