metadata

dataset_info:
  features:
    - name: rus
      dtype: string
    - name: udm
      dtype: string
    - name: eng
      dtype: string
  splits:
    - name: train
      num_bytes: 18829292
      num_examples: 40365
  download_size: 9006599
  dataset_size: 18829292
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
language:
  - udm
size_categories:
  - 10K<n<100K

Udmurt-Russian-English parallel corpora for LaBSE training

English columns was translated by Yandex.Translator API from russian.

Please note that the last 250 rows in the dataset are taken from FLORES-250, so if you want to train a machine translator, remove these rows.