This dataset contains 27 sentences in both translation directions: Tsakonian to Greek and Greek to Tsakonian, for a total of 54 sentence pairs. It is designed for evaluating machine translation models, specifically for computing BLEU and ChrF scores.
The main source for this evaluation set is the book Για να κ̔οντούμε τα γρούσσα νάμου (Gia na khontoume ta groussa namou) by Ioannis Kamvysis, with additional manual… See the full description on the dataset page:
https://huggingface.co/datasets/jgchaparro/language_garden-tsd-evalset-1.