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
- [Doc] metrics: Explain scheduler load and queueing signals
- [Perf] [Apple] layernorm: Add a Metal RMSNorm for the Torch MPS path
- [Fix] [Apple] mlx: Initialize Mamba checkpoint grid for auxiliary radix cache
- [Perf] kimi-k3: Reduce latent-up weight memory on TP8 decode workers
- [Bug] SMG PowerOfTwoPolicy does not track in-flight requests in regular HTTP routing
- [Perf] [AMD] dsa: Opt-in single-pass Triton top-k for long-context decode rows
- [Feat] metrics: expose detokenizer worker processing time
- [Fix] speculative: Align all spec workers on the scheduler-facing forward_batch_generation signature
- [Perf] [AMD] dsv4: Token-parallel prefill on ROCm (opt-in)
- [Perf] cosmos3: optimize Sim-Bimanual kv cache management
- Docs
- Python not yet supported