Hello, As promised I pushed a new version of the evaluation dataset for task1. @anuzzolese: feel free to review and modify it if you see anything that seems wrong to you. Best.