ValueError: A header text must be provided for VQA models
#1
by
pkun
- opened
I followed the instructions in the model card to run the model, but I encountered an error: "ValueError: A header text must be provided for VQA models." Could there be an issue with the preprocessor_config.json file in the repository? Should I change 'is_vqa' to false, since the original paper and GitHub repository do not provide a question for the screen2words task? Moreover, in the preprocessor_config.json of pix2struct-textcaps, 'is_vqa' is not provided.