ner_manual recipe with side by side image/PDF

Hi ,
I am trying to create a ner_manual task for text extracted from scanned PDF using OCR. It is important for me to put the pdf side by side the extracted text, however in the "blocks" section it only allows for one on top of the other. My main problem is that the text is really long ( between a page to a few pages ) and so the image can be overlooked or requires the annotator moving up and down a lot
Is there a nice way to change the layout while still using the ner_manual recipe?
Thank you ,
Keren

Hi! You might be able to do this by adding some custom CSS via the global_css config setting that formats the blocks in a grid with two columns. For example:

.prodigy-container {
    display: grid;
    grid-template-columns: 1fr 1fr;  /* two equal columns */
}

So your first block would be the text in ner_manual, and the second block would be an image, either embedded via the image UI or custom HTML (which could also let you wrap the page in a scrollable container if needed).

You can also use the cardMaxWidth theme setting to make the annotation card wider (e.g. "95%" of the available space), which probably makes sense here if you're displaying the content in two columns.

Thanks Ines, sounds like exactly what we need,
I will go ahead with that solution and will let you know ,
Keren