Hi,
I've constructed a pipeline using Scikit-Learn on SageMaker, incorporating data transformations and a Random Forest model, then I use the hyperparameter tuner to find the best model.
I can unpack and load the best model and use it to do a prediction.
As I would like to move the model to production, I am trying to do a batch transformation.
sklearn_model = SKLearnModel(model_data='s3://BUCKET/PATH/model.tar.gz',
role=role,
framework_version='1.2-1',
py_version='py3')
transformer = sklearn_model.transformer(instance_count=1,
instance_type='ml.m5.large',
output_path='s3://BUCKET/PATH/prediction/')
transformer.transform(data='s3://BUCKET/PATH/input/test.csv',
content_type='text/csv',
split_type='Line')
However, I consistently encounter this error: AttributeError: 'NoneType' object has no attribute 'startswith'. What might be causing this issue?
Thanks,
Mat