Follow @avatar42 on Micro.blog.
Google’s TurboQuant AI-Compression Algorithm Can Reduce LLM Memory Usage by 6x