Closed smeredith closed 1 month ago
Another data point: the OCR operation tool in the main part of the app works fine. It's just when trying to use it in a pipeline that it fails.
not sure why it is handling it as comma seperated ill check and try have it support correct json but that should unblock you with above issue
Thanks.
The only way I could get it to work is to download the .json file, edit it (add "languages": "eng",) and upload it. Changes to the web form don't seem to save reliably.
I think this was resolved? let me know if still a issue
To repro, create a custom pipeline config with only one step: ocr-pdf. Then try to OCR a pdf. You get an empty download.txt.
My config looks like this: