How to start AI about opencapture HOT 3 CLOSED

Hi,

The document isn't ready yet in English, sorry for that. Moreover, we didn't have the time yet to write AI documentation

To train your model you have to:

• Locate your docservers folder (by default /var/docservers/opencapture/CUSTOM_ID/)
• Choose if you want an AI model for Verifier or Splitter module (regarding your name, it's probably verifier)
• In the /var/docservers/opencapture/CUSTOM_ID/verifier/ai/train_data/ folder create subfolders with the name you want
• In each folder you just created, you can now add PDF files
• Then in the UI you'll see your folders and you can start training

Btw, if you see any errors in translation, do not hesitate to create pull request :)

from opencapture.

Hi @nathan30 ,
thanks for your message :) the step manuals was usefully and I can now train the AI :)

I noticed two issued with translation - one is with AI directly:

and the second is in the General>Configurations> The description is not translated :)

Should I create a separated bug report for both of them? :)

Best Regards
Martin

from opencapture.

Thanks for your report. I'm waiting your PR :)

The main issue about the second translation error is that it's related to the data_fr.sql file (the default database data in the application). For now, we didn't have the time and funding to create a data_en.sql (if you want to do it, feel free ;) )

from opencapture.