---
dataset_info:
  features:
  - name: translation
    dtype:
      translation:
        languages:
        - sah
        - ru
  splits:
  - name: train
    num_bytes: 4979324
    num_examples: 30146
  download_size: 2860256
  dataset_size: 4979324
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---
fork на всякий от wmt/yakut