Pandas Udf in Pyspark running in only 1 Executor in yarn client or cluster mode

Question

I have a code which reads data from Hive Table and applies a pandas udf, the moment it reads data from table it runs in 11 executors , however the moment it executes a pandas ud