Testing out BLIP on the given data subset of COCO validation data. This was part of computer vision course at IIIT-D (CSE344/CSE544/ECE344/ECE544)
Before running the .ipynb file, clone the original BLIP repo, set up the dependencies as instructed and download the model weights from the original repo (model_base_vqa_capfilt_large.pth)
git clone https://github.com/salesforce/BLIP.git