metadata

language:
  - udm
size_categories:
  - 1K<n<10K
task_categories:
  - translation
dataset_info:
  features:
    - name: rus
      dtype: string
    - name: udm
      dtype: string
    - name: source
      dtype: string
  splits:
    - name: train
      num_bytes: 347323
      num_examples: 3670
  download_size: 168049
  dataset_size: 347323
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*

Udmurt-Russian dataset from Tatoeba

Contains 3670 Russian-Udmurt sentences. Punctuation added to some sentences. Dump downloaded 18.02.2024.

Usage

from datasets import load_dataset

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