Deploying Kedro Model using Seldon Core

I have trained the model using Kedro 0.17 and saved it inside AWS S3 in .pkl format. Now, I am trying to create an API using Seldon core, when i am loading .pkl inside my .py file it’s giving an error:

ModuleNotFoundError: No module named ‘kedro.pipeline’

with open(’{s3path}/model.pkl’, ‘rb’) as pickle_file:
ml_model = pickle.load(pickle_file)

Is there any way in Kedro to save the .pkl with all the module loaded