Fetching the paper…

CoMM: A Coherent Interleaved Image-Text Dataset for Multimodal Understanding and Generation · Around