Does jpn-kor dataset have only 'test' dataset?

#10
by sappho192 - opened

Hi,
I'd like to use jpn-kor dataset from this tatoeba_mt dataset but I met KeyError that says 'train' data is not found.
Following code is what I have tried and I found out that the jpn-kordataset only contains the test dataset.

from datasets import list_datasets, load_dataset
tatohel_dataset = load_dataset('Helsinki-NLP/tatoeba_mt', 'jpn-kor')
print(tatohel_dataset['train'][0])

How can I solve this problem?

For someone who meet the same problem, I've released aforementioned dataset:

https://huggingface.co/datasets/sappho192/Tatoeba-Challenge-jpn-kor

sappho192 changed discussion status to closed
Language Technology Research Group at the University of Helsinki org

Thanks for sharing the solution!

Sign up or log in to comment