Model training
As before, for the model training, the feature store is the source. Hence, let's create our model training notebook and install and initialize the required objects for querying the feature store. Here is the link to the notebook:
The following code block installs the required libraries for model training:
!pip install sagemaker==2.88.0 s3fs joblib scikit-learn==1.0.2 xgboost
After installing the required libraries, initialize the SageMaker session and the required objects:
import sagemaker
from sagemaker.session import Session
from sagemaker.feature_store.feature_group import FeatureGroup
#import os
#os.environ["AWS_ACCESS_KEY_ID"] = "<aws_key_id>"
#os.environ["AWS_SECRET_ACCESS_KEY"] = "<aws_secret>"
#os.environ["AWS_DEFAULT_REGION"] = "us-east-1"...