How to add additional prompt in florence in addition to the existing tokens like "<OD>", " <DETAILED_CAPTION>" and etc ?
#67
by
dutta18
- opened
In the example notebook provided it takes only the existing tokens to generate the output like captions, detailed captions, objects etc. But can we/ how to give additional prompts to florence ? I noticed an additonal parameter is provided "text_input" alongside "task_prompt" in the associted notebook. But if I put any prompt to this variable "text_input" it generates an error.
AssertionError: Task token should be the only token in the text.
In the .py files
Can u elaborate please where I can find this file ?
Look into files in this repository, such as processing_florence2.py, modeling_florence2.py ..., then find the lines about tasks such as 'OD' and add yours following these lines