what has been your best document parser, especially for PDF with tables? So far it looks like only llmsherpa and llamaparse (which I can't use) did a good job with tables

I am actually thinking if that is even the right approach

ada is reported to be multilingual but I'll try this. Does Hugging face support running the model for us somehow?

it's a large legal codex. I am not sure what the approach to tweak/tune is