aksg87

Thanks for the comprehensive response @MeshkatShB, I completely agree. TL;DR: markdown is a pure text representation that you can model in a few-shot example for the LLM. If anyone has...

Thanks for sharing this @ianstorrs-cd — This is definitely a priority for improvement and should be fixed soon.

This would be interesting and I think there are various approaches for combining images and text. If you end up coming up with a proof of concept please do share...

Hey @abishekchiffon, This is a great idea and discussion. Ultimately, LX goes from user-defined prompt examples to the schema that is used for [Controlled Generation](https://developers.googleblog.com/en/mastering-controlled-generation-with-gemini-15-schema-adherence/). There could be multiple strategies...

Hi @andyxhadji, Thanks for creating the langextract-bedrock plugin! Quick note: The plugin generator now includes a .gitignore template (034f2b6) to help avoid committing build artifacts. Once a few users try...

Hi @andyxhadji - we now have a plugin registry as noted in #182 — You can see #186 for an example of adding your plugin to the registry, thanks!

This is awesome, @tijoseymathew - thank you for sharing! Native PDF support is on the roadmap for LangExtract, and your example with Docling will be very valuable for planning and...

Hi @tijoseymathew - Just to clarify, the langextract-docling seems to return an annotated document, not a PDF - but eventually, the aim would be to overlay grounded data on the...

+1 @tijoseymathew this is great, and a very valuable reference! Ultimately going from data items back to the PDF will be very valuable for those workflows.

Hi @fredguth - If your plugin is deprecated mind submitting a PR to remove it? Thanks!