Open abhidipbhattacharyya opened 2 years ago
I am trying use oscar for caption generation. How can I visualize attentions for generated tokens?
I have a similar question about how to visualize attention maps for the VQA task.
I am trying use oscar for caption generation. How can I visualize attentions for generated tokens?