5 ms·

so many inference project, omlx already supports all of this and has a 1000 people trying to optimize it constantly

• carloslfu 15d ago
  Interesting! I'll check it out
• carloslfu 15d ago
  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