Has docling improved? I had a bit of a nightmare integrating a docling pipeline earlier this year. Docs said it was VLM-ready, which I spent lots of hours finding out was not true, just to find a relevant github issue which would've saved me a ton of hours :/ allegedly fixed, but wow that burned me bigtime.
our team has tested docling pretty extensively, works well for simpler text-heavy docs without complex layouts, but the moment you introduce tables or multi-column stuff it doesn't maintain layout well.
This is so true, and I say this myself coming from an Indian education system that my vocabulary has gone through an objective optimization function similar to that of these LLMs XD
reply