github.com
Colibri: GLM-5.2 on 25GB RAM
Tiny, pure-C engine for running GLM-5.2 on hardware.
About Colibri: GLM-5.2 on 25GB RAM
Run GLM-5.2 (744B MoE) on a consumer machine with ~25 GB RAM using pure C and zero dependencies. Experts are streamed from disk, enabling a tiny engine with a massive model. No BLAS, no Python at runtime; optional CUDA tier for pinned experts.
Categories
Topics
Published
Discussion highlights
AI-generated summary based on Hacker News comments
🔗 Full discussion: https://news.ycombinator.com/item?id=48842459
No comments yet. Have you tried it?