What percentage of the original filtered dataset was retained after deduplication?

What percentage of the original filtered dataset was retained after deduplication?: a task in SmolDataEnvs-harbor-train-sandbox: 📊 SmolDataEnvs: Harbor (train) (Harbor dataset). Files (in /home/user/input, no subfolders): - Reviews.csv - database.sqlite - hashes.txt