MiniMax-M1, first open-weight, large-scale hybrid-attention reasoning modelgithub.com3 points·hallh··0 commentsOpen articleSaveView on HN