Most efficient way of executing python requests and storing resulting binary using pyspark?

Question

I\'m making use of the columns in my dataset to create the required url for python request to get. The python request is executed inside a pyspark udf. While the resulting b