---
dataset_info:
  features:
  - name: sentence
    dtype: string
  - name: last_hidden_state
    sequence:
      sequence: float32
  - name: pooler_output
    sequence: float32
  splits:
  - name: train
    num_bytes: 19669131636
    num_examples: 100000
  download_size: 19793578186
  dataset_size: 19669131636
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---