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
- [Perf] diffusion: Encode large compressed PNG outputs in parallel strips
- [Refactor] mem_cache: Keep the KDA speculative ReplaySSM ring in the request sub-pool
- [Perf] minimax-h3: Calibrate quality=high with a per-step cache schedule
- [Fix] processor: Lower chat requests that declare deferred tools
- [Chore] diffusion: Fix stale comments and docstrings in serving, apps and CI scripts
- [Fix] test: Seed the retracted queue through move() in test_prealloc_token_usage
- [Perf] dflash: Keep a sliding-window MLA draft's KV in a per-request ring
- [Feat] dflash: Send the draft KV ring over PD disaggregation (Mooncake)
- Pr yzy 11 9 backup run
- [Refactor] mem_cache: Keep the GDN speculative ReplaySSM ring in a request sub-pool of unified memory
- Docs
- Python not yet supported