[x] spin up model using either their docker container or the python code
[x] attach the model to an API endpoint that the API from #17 can use to request parsed ingredients
Their docker image has you pull the image, then train the model inside the image. It would nice to have one that pulled with a pre-trained image. Also, my inclination is to try out fastapi for the API endpoint since it can integrate easily with the python code they've already written, but something else like flask would work as well.
Following up on @horinezachary's work with ingredient-phrase-tagger, try to:
Their docker image has you pull the image, then train the model inside the image. It would nice to have one that pulled with a pre-trained image. Also, my inclination is to try out fastapi for the API endpoint since it can integrate easily with the python code they've already written, but something else like flask would work as well.