Can train the model after I finished image labeling in Prodigy ?
I don't know the details of your setup or your machine learning model, but you can use db-out to output your annotations to
.jsonl. From there you can pick it up from a Python script and use any machine learning model you like.
Prodigy does not offer machine learning models for image tasks though, so you would need to provide your own model. But you can use any model that Python can provide, so there should be plenty of options .