I am attempting to ingest csv files from an S3 bucket with Autoloader. Since the schema of the data is inconsistent (each csv may have different headers), I was hoping to ingest the data as Variant following this: https://docs.databricks.com/aws/en/ingestion/variant#-ingest-data-from-cloud-object-storage-as-varia.... However, when trying to create a streaming table, I always get the error in the screenshot despite passing table_properties={"delta.feature.variantType-preview": "supported"} as an argument to the flow creating my table. Does the table_properties argument not work with DLT tables for this particular property?
