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
- [Feat] rust-server: Reach every attention-DP rank with aborts and publish the listener map
- [Fix] metrics: Correct MFU/MBU bandwidth overcounting and GraniteMoE AutoTokenizer fallback
- [Perf] [AMD] aiter: Run the DSpark MHA draft on non-causal unified_attention
- [Fix] diffusion: Build the Cosmos 3 RoPE table on GPU so CPU offload gives the same output
- [Perf] kda: Write prefill convolution into contiguous QKV groups
- [Perf] kimi-k3: Allow residual bank shards with NCCL
- [Fix] [XPU] server_args: Disable decode graph on XPU when tp_size > 1
- [Fix] stop_regex: Record the matched span, not the pattern source
- [Fix] observability: Assign correct dp_rank under attention-dp for KV metrics and Prometheus
- [Fix] sglang-miles: Clear the multimodal embedding cache on flush_cache and weight updates
- Docs
- Python not yet supported