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
- fix(responses): let reasoning.effort enable thinking for explicit-thinking parsers
- fix(kimi-k3): hand-written section markers, and two Responses API bugs that break Codex
- [JIT] per_token_group_quant: fix the fp16 UE8M0 multiplier overflow, 1.23x on the fused EP-MoE path, unify the scale allocation
- fix(openai): wrap non-streaming error responses in the OpenAI error envelope
- [model-gateway] Preserve explicit zero refill rate for concurrency limiting
- [Docs] External model package example on the embedding models page
- [AMD] Add FlyDSL GDN linear-attention backend (--linear-attn-backend flydsl)
- [diffusion] fix: locate IPC storage_device slot by name instead of index
- [diffusion] fix: fall back to NCCL when IPC all-to-all spin times out
- [Bug] OpenAI completions reject per-prompt extra_key/cache_salt lists allowed by the schema
- Docs
- Python not yet supported