README.md

metadata

dataset_info:
  features:
    - name: rus
      dtype: string
    - name: udm
      dtype: string
    - name: source
      dtype: string
  splits:
    - name: train
      num_bytes: 80931
      num_examples: 889
  download_size: 39673
  dataset_size: 80931
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
task_categories:
  - translation
language:
  - udm
size_categories:
  - n<1K

Udmurt-Russian dataset from Tatoeba

Contains 888 Russian-Udmurt sentences. Punctuation added to some sentences. Dump downloaded 28.09.2023.

Usage

from datasets import load_dataset

dataset = load_dataset("udmurtNLP/tatoeba-rus-udm-parallel-corpora")