I'm not training my own model. When I pass the Travel model's id I get: "Model permission denied","details":"Cannot view inputs for public models"
BUT, I will build up a lot of inputs over time. As I gradually add them I want to store their concepts in my db. Downloading ALL inputs to just get the concepts of the most recently added inputs seems terrible inefficient.
Maybe for every input (image) I want to process I have to:
- Add input to make it searchable
- Use the Predict API to get the concepts for that input
This would work but ideally I wouldn't need to ask you to analyze every input twice. I mean you do have the inputs's concepts after the first step..