Tool to read a csv file and turn it into a Pytorch dataloader ready to be used for training a network.
preprocess()
returns a trainloader and a testloader
Arguments to pass in:
- filepath of csv file with headers
- columns of features (with support for index slicing)
- columns of label
- validation set size
- batch size
Example usage:
import csv_to_dataloader
trainloader, testloader = csv_to_dataloader.preprocess('data.csv', 0:6, 6, 0.3, 32)