Tabular Data Regression - large data in multiple csv files

how would we use fastai Tabular_Learner and tabular pandas for large data, which we are not able to fit in the memory?

data loader is loading it by chunks but considers we have a multiple csv file - e.g. 6 of them.

How to use this for tabular_pandas?

Consider that we cannot just merge all 6 csv files into one dataframe into memory