Depoly a custom fastai model


I have a custom fastai model that is created and trained in locally in a jupyter notebook. My aim is to write a Google cloud function that is Cloud Storage triggered, this function loads the fastai model stored in Cloud Storage, takes the input and predicts and stores the result in Firestore. Keeping aside Cloud Function, deploying a custom model needs the class information and the model. Can I get some ideas on this?? its nerve racking.