---
dataset_info:
  features:
  - name: similarity
    dtype: float32
  - name: French
    dtype: string
  - name: Tumbuka
    dtype: string
  splits:
  - name: train
    num_bytes: 120867423
    num_examples: 969510
  download_size: 83983608
  dataset_size: 120867423
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---