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
- [BugFix] Fix per-layer quantization exclusions in Grok
- [DSpark] Avoid per-batch-size Triton recompiles in committed_swa_locations
- [Fix] Preserve Cohere tool calls after leading streamed text
- [Bug][KV Cache] nvfp4 KV silently reuses the checkpoint's fp8-calibrated k/v_scale as NVFP4 global scale → deterministic long-context corruption (single-variable proof, sm_120)
- fix(diffusion): preserve lazy quantization exports and per-backend lookup
- fix(frontend): preserve Runtime SSE events across transport reads
- docs: stop documenting deprecated NIXL POSIX AIO option
- fix(tokenizer): make _init_req_state atomic to prevent rid state leak on duplicate IDs
- fix(voxtral): encode empty audio as one chunk of silence instead of crashing in stft
- GLM-5.2 NVFP4 DSA via pinned FlashMLA fork (no source patch)
- Docs
- Python not yet supported