VILA: On Pre-Training for Visual Language Modelsarxiv.org2 points·milliondreams··0 commentsOpen articleSaveView on HN