Missing Data When Parsing in Auto Mode #649
Description
Describe the bug
I have a PDF document containing text data (text, tables, etc.). At the top of the document, there is a logo on one side and text with the company name and address on the other. The text is not an image – it is selectable. However, when I upload the document in auto mode, this text is missing.
Job ID
14ec6c1a-b92e-4281-87be-64c36daad8d2
Client:
Please remove untested options:
- API
- Frontend (cloud.llamaindex.ai)
Additional context
The same behavior occurs in both the FE and API. The auto_mode_trigger_on_table_in_page and auto_mode_trigger_on_image_in_page flags are enabled.