I tested the model on a few PDF documents. I wasn't able to get reliable image/figure extraction from the model using default prompt template. I was wondering if the model can extract the images as well, along with parsing text and table contents. Thanks!
