By default, the following code split the IMDB dataset with ratio of 0.5 ( train_data =25000 and test_data=25000).
train_data, test_data = torchtext.datasets.IMDB.splits(TEXT, LABEL)
I want to split it with ratio of .0.8 ( train_data =40000 and test_data=10000). How I can process?