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
- [Bug] Dockerised SGLang for ROCM with gfx1150 cannot load a quantised model: use of undeclared identifier '__cvta_generic_to_shared'
- [Rust Server] Return kv_events from /server_info like Python
- [Rust Server] Return kv_events from /server_info like Python
- [Bug][AMD] Qwen3.5-397B-A17B-FP8 disagg + dp-attention: decode DP ranks flood `/query_dp_ranks` bootstrap → `[Errno 99] Cannot assign requested address`, 0 successful requests
- [Bug] ROCm diffusion fused-norm kernels depend on non-public FlyDSL APIs
- @copilot resolve the merge conflicts on this branch.
- [Bug] MiMo tool-call streaming: text after the last tool call is silently dropped; split bot-token flushes markup into content
- [Bug/Design] `SGLANG_SANITIZE_NAN_LOGITS`: a fully-NaN logits row becomes a uniform-random sample — garbage streamed to clients as normal output; proposal: opt-in per-request abort
- [Bench] Fix peak output throughput counting chunks instead of tokens
- [Bug] DeepSeek-V4 default prevents speculative decoding from resetting max-running-requests to 48
- Docs
- Python not yet supported