---
dataset_info:
  features:
  - name: chunk
    dtype: string
  - name: input_ids
    sequence: int32
  - name: attention_mask
    sequence: int8
  - name: embedding
    sequence: float64
  splits:
  - name: train
    num_bytes: 406067354
    num_examples: 46804
  download_size: 319504679
  dataset_size: 406067354
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---