Nodja

Here's some thoughts. I don't think anyone knows what the best hyperparameters are, specially for large scale, my intuition tells me that you're gonna have to do a bunch of...

> I created a modified `Dataset` class that returns `(image, text_embedding, text_mask)`, but I'm still getting errors. I wonder what the shape of the embedding and mask must be. Get...

> Is there any code for conditional training on dataset which has images and text? Assuming you have the texts next to the image with the exact same name but...