This message was deleted.

How do you save the model? @yadwinder kaur

We need to make sure the custom object pass when saving the model in the

train.ipynb

notebook. Also, check the for spellings if there is any difference.

This is how saved the model: import bentoml bentoml.xgboost.save_model("credit_risk_model", model)

We need to pass the

DictVectorizer

in the custom objects parameter. The model needs it to transform incoming data. Here's how we should do:

import bentoml

bentoml.xgboost.save_model("credit_risk_model", model, custom_objects={"dictVectorizer": dv})

Also, we need to ensure that dictVectorizer spellings are same in

service.py

as well.

Thanks a ton! this worked finally 🙂