4 ms·Tell HN: To run distributed LLMs, more people should contribute their GPUs3 points by behnamoh 3y ago