Every few weeks, someone stumbles on what llama.cpp has had for years again, and reimplements it. I'm not sure why.
Because with AI everybody can reinvent the wheel with ease