---
dataset_info:
  features:
  - name: vi
    dtype: string
  - name: en
    dtype: string
  - name: loss
    dtype: float64
  splits:
  - name: train
    num_bytes: 188286058.1887368
    num_examples: 1000000
  download_size: 199988484
  dataset_size: 188286058.1887368
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---