Toward Inference-Optimal Mixture-of-Expert Large Language Modelsarxiv.org24 points·zhiQ··0 commentsOpen articleSaveView on HN