Dataset used to train Magic card text to image model
BLIP generated captions for Magic Card images collected from the web. Original images were obtained from Scryfall and captioned with the pre-trained BLIP model.
For each row the dataset contains image and text keys. image is a varying size PIL jpeg, and text is the… See the full description on the dataset page:
https://huggingface.co/datasets/YaYaB/magic-blip-captions.