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 supported61 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [ROCm][Perf][DSv4] Use AITER tuned GEMM for the compressor projections and inverse-RoPE einsum
- [Attention] Add MiniMax M3/3.1 indexer CP for prefill and decode
- [Bugfix] Free cuMem allocations with the aligned size in my_free
- [Bugfix][Frontend] Remove stale `--uds` socket files so a server can restart on the same path
- [ROCm][Perf][DSv4.1] HIP sparse MLA decode kernel for batched decode on gfx942
- [ROCm][DSv4.1][Perf] gfx942 sparse MLA decode: BLOCK_K=16 with 2-stage pipelining
- [Perf][Attention] Candidate-only FP8 indexer decode scoring on SM90
- [Bug]: mamba_ssm_cache_dtype="float16" crashes mamba-family models at startup — bare kernel assertion with an empty message (selective_scan_fwd.cu:791)
- [Refactor][Attention] Make MLA DCP collectives pluggable
- [Warmup v2] Signature-driven JIT warmup for Triton kernels
- Docs
- Python not yet supported