Split train and test samples at the command line:

$ shuf -n 7 dataset.txt > train.txt
$ grep -Fxvf train.txt dataset.txt > test.txt

thunix.net/~jailop/2022/split-

The Economist: Welcome to the era of the hyper-surveilled office Welcome to the era of the hyper-surveilled office

tilde.zone

masto instance for the tildeverse