MultiModal-GPT: A Vision and Language Model for Dialogue with Humansgithub.com/open-mmlab4 pointsvov_or3 years ago