How to resolve dask dataframe overflow error?

Question

I am using dask to process a large dataframe that cannot fit into memory.

20 mins into the processing, python give an error message:

OverflowError: can