sglang
https://github.com/sgl-project/sglang
Python
SGLang is a high-performance serving framework for large language models and multimodal models.
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 supported1 Subscribers
Add a CodeTriage badge to sglang
Help out
- Issues
- [CI] PD: Mooncake 'Memory region not registered by any active device(s)' fails a whole class on the 8-gpu-h20 runners
- [Perf] speculative: batched top logprobs when every request asks the same k
- [Perf] openai: cache per-token text and bytes for logprob outputs
- [Fix] [NVIDIA] cosmos:Multi node execution fixes
- [Fix] moe: disable deferred finalize under deterministic inference
- [Fix] [NVIDIA] deterministic: enable DeepGEMM BF16 deterministic mode when getter is absent
- [Perf] eager-runner: size the input registry by the prefill limit
- [Fix] http_server: skip the sglang launcher re-import in spawned uvicorn workers
- [Perf] [NVIDIA] qsa: JIT CUDA kernel for the indexer decode scoring
- [Perf] qsa: 64 CTAs per row for the mid-batch indexer decode
- Docs
- Python not yet supported