SpikingBrain 7B – More efficient than classic LLMs 10 months agoThey use MetaX GPUs instead NVDIA's...? This point is actually more surprising. 0ThreadHN