Hi All,I just wanted to know if is there any option to reduce time while loading Pyspark Dataframe into the Azure synapse table using Databricks.like..I have a pyspark dataframe that has around 40k records and I am trying to load data into the azure ...