E3C dataset in the 5 original languages (Basque, English, French, Italian, Spanish). Each dataset comes with a train-validation-test split