SGLang
@sgl_project
Run LLMs fast at any scale ๐
Join our community
For AI tech blogs & deep-dives ๐ @lmsysorg
๊ฐ์
May 2025
53 ํ๋ก์ ์ค 9.5K ํฌ
We've been working closely with partners and @zai_org to improve production serving for the GLM-5.3 family.
Shout out to @Cloudflare for reporting the issues that led to the fix ๐งกโ๏ธ
Shoutout to the @sgl_project community for catching an issue in our chat template๐
We've optimized the tool-result reordering logic (early exits instead of scanning every block) and updated the templates for GLM-5.3 and GLM-5.3-Flash.
Fix is live โ thanks for making it better. ๐ช
๋ ๋ณด๊ธฐ
