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 supported51 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [DCP][Kernel] Consume owner-sharded Top-K candidates through symmetric memory
- [RFC]: Support Speculative Decoding with Decode Context Parallelism
- [Feature][Quantization] Add per-layer online quantization configuration (#50281)
- perf(scheduler): balance requests across PP batches
- [Bug][PD][Mooncake] Client disconnect on decode node does not release prefill KV cache
- [Bug]: FlexAttention paged K/V offsets overflow int32 once `num_gpu_blocks ≥ 2**31 / (block_size · 2 · num_kv_heads · head_size)` — crash *or* silent wrong output
- [Bug]: ROCm tuned-config lookup keys off an unstable device name — shipped gfx1151 configs are unreachable
- [Bug]: DeepGemm accuracy auto-disable (_DEEPGEMM_BLACKWELL_EXCLUDED_MODEL_TYPES) does not apply to the FP8 MoE path
- [Kernel] Fuse FlashInfer GDN prefill state I/O
- [Kernel][MoE] Add VMM-backed SharedEP for native FP4 decode on SM100
- Docs
- Python not yet supported