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
- [Bug]: Qwen3.8-Flash-Next (qwen4_exp) deadlocks on RTX 6000D (SM120) — hangs in JIT warmup and at first real request (GDN/hyperconnection kernels); Qwen3.8-27B works fine on the same box
- [Bugfix][Parser] Strip leading whitespace from first content delta af…
- [Spec Decode] Compose grammar-constrained sampling with EAGLE speculative decoding
- [Bug]: moe_backend="cutlass" with online FP8 raises a reason-free "vLLM CUTLASS FP8 MoE backend is disabled for this configuration." — every sibling backend reports the actual rejection reason via _make_log_unsupported
- [Bug]: quantization="mxfp8" + moe_backend="humming" crashes weight conversion with a bare KeyError: 'weight_scale' — the MXFP8 oracle selects humming, then a third-party schema converter raises
- [Bugfix][Offload] Preserve CPU residency for renamed NVFP4 expert weights
- [Kernel][Perf] Add H20 block-FP8 fused MoE config for E=512,N=128 (Qwen3.5-397B-A17B-FP8 TP8, up to 2.0x)
- [Security] Reject unresolved multimodal inputs before execution
- [Spec Decode] Fix adaptive verification startup on FA2 GPUs
- [Bugfix] Reject shape mismatches in ShardedStateLoader instead of partially loading (fixes #61681 mechanism)
- Docs
- Python not yet supported