Abdullah Abid’s TILs
Atom feed for ai
When running local AI models, a useful rule of thumb: you need roughly 3x the model size in RAM to handle back-and-forth messaging. …