Spark

Using Amazon SageMaker with Apache Spark

MNIST with SageMaker PySpark:

Parameterize spark configuration in pipeline PySparkProcessor execution: