This dataset contains multimodal question-answering examples grounded in textbook figures. Records in the figure-grounded configurations are filtered to include only examples whose referenced image files are present in this release.

Configurations

visual: 13791 figure-grounded visual questions with resolved images. knowledge: 13501 caption/text-grounded… See the full description on the dataset page: https://huggingface.co/datasets/Naga1289/TestingDataset.