Hi! Thank you for the amazing open-source work—this project significantly deepens the search depth of existing multimodal deep research systems and is highly meaningful. I’d like to ask: do you have any plans to open-source the full training dataset?