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 supported31 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Bug]: Pipeline-Parallel (PP=2) Engine Crash / Hang on Dual Battlemage GPUs (Arc Pro B70 + Arc B580)
- [BugFix] Fix latent-MoE crash when MoE runner holds the gate (is_internal_router)
- Preserve FP8 indexer WK pairs across incremental load_weights calls
- [Rust Frontend] Add /tokenizer_info endpoint
- [Bug]: GPU coredump during FlashInfer `trtllm_bf16_moe` autotune with Qwen3.5-35B-A3B on B200 (DP=2 + EP)
- [Kernel] Start SWA/chunked KV-tile loop at first allowed key (#44575)
- [Bug]: MTP speculative decoding with --kv-cache-dtype auto produces cross-sequence garbage under batching (Gemma-4 W4A16; fp8 KV unaffected)
- [DRAFT][Refactor] Harmony Responses API Refactor to use HarmonyParser
- [Spec Decode] Support SWA + DFlash for MiMo
- [Tracking][Spec Decode] DFlash Bring-Up Tracker
- Docs
- Python not yet supported