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 supported62 Subscribers
View all SubscribersAdd a CodeTriage badge to vllm
Help out
- Issues
- [Bugfix] Trim WNA16 tail groups before TP sharding
- [MoE] Disable expanded DeepEP v2 dispatch in all execution modes
- [Bugfix][Core] Reclaim KV blocks from idle streaming-input sessions under KV pressure
- [Attention][MLA] Support nvfp4_ds_mla on FLASHINFER_MLA_SPARSE, with native NVFP4 decode
- [Bug]: Whisper verbose_json returns empty text when the decoder emits no timestamp token; json returns the full text
- [Perf][Kernel] Triton attention: split only the sliding window across 3D decode segments
- [Doc]: On Hopper NVSwitch nodes, TP runs can be non-reproducible even for a single request; NCCL_NVLS_ENABLE=0 alone fixes it at ~1% cost
- [Draft][XPU] Test upstream nightly wheel in Dockerfile
- [Feature]: shechduling using predicted decode output length
- [Misc] jit_monitor: distinguish real Triton compilations from disk-cache loads
- Docs
- Python not yet supported