What is the fastest way to filter my dask dataframe?

Question

I have a 50gb text file with 300+ millions rows and 38 columns (separated by the symbol "|"). Each row has an unique integer ID. My computer doesn\'t