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] CUDA illegal memory access in QSA extend forward at 8 concurrent requests; suppressed by CUDA_LAUNCH_BLOCKING=1 and --disable-overlap-schedule (Qwen3.8-Flash-Next-FP8, H20 TP8)
- [metrics] KV age at hit and eviction, lifetime and reuse count at eviction, for RadixCache and HiRadixCache
- [Performance] tiny_gemm regresses DeepSeek-R1 NVFP4 decode on Blackwell despite faster standalone kernel
- [MoE] Add opt-in NCCL EP LL CUDA Graph lifecycle
- [MoonEP] Use DeepGEMM psum layout, zero-copy shard and other optimization
- [CLI] Avoid loading diffusion runtime during backend auto-detection
- [Performance] Improve HyperConnection performance for Qwen-3.8-Flash
- [Perf] Fork workers from a preloaded forkserver (SGLANG_ENABLE_EARLY_FORKSERVER)
- [Feature] Request to 8-bit Quantization of Attention with SageAttention
- [Feature] support `gather` instead of `all_gather` when gathering the logits
- Docs
- Python not yet supported