sbu caption dataset format

[1024er]

sub.json is organized in the format: [{'image': '4385058960_b0f291553e.jpg', 'caption': 'a wooden chair in the living room', 'url': 'http://static.flickr.com/2723/4385058960_b0f291553e.jpg'}, ...}

but the downloaded sbu_images.rar is extracted as: 0000/ 0001/ 0002/ 0003/ ... 0999/ in each directory contains 1000 images named in order: 000.jpg 001.jpg 002.jpg ... 999.jpg

Therefore, the image storage path does not correspond to the path in json. @dxli94