Comment by carloslfu

Both projects are different in scope. Think of slotstream as optimizing for memory and for this specific model for now, my intention is not to build an inference engine the same as oMLX