How to handle categorical features with large amount of unique values in pysparkml

Question

I\'m using pysparkml library and its models for regression problem and my data have some categorical features with large amount of unique values (more then 1000). What is the ri