Reading different sections of the CSV file into a Deep Learning model for training

Question

I have a CSV file with 1 column and 700,000 rows. I want to be able to read 10,000 rows each time because that\'s my input shape. Is there any way to do this?