---
license: mit
dataset_info:
  features:
  - name: query
    dtype: string
  - name: pos
    sequence: string
  - name: neg
    sequence: 'null'
  - name: relevance
    dtype: float64
  splits:
  - name: train
    num_bytes: 86857461389
    num_examples: 400000000
  download_size: 58273140878
  dataset_size: 86857461389
configs:
- config_name: default
  data_files:
  - split: train
    path: nllb_400M/train-*
---