[bravura]

I've also been using the nougat models from meta, which are trained to turn PDF into md using the donut architecture

[rahimnathwani]

Does it work well on documents that aren't academic papers?

https://facebookresearch.github.io/nougat/