README.md

---
dataset_info:
  features:
  - name: input_ids
    sequence: int32
  - name: token_type_ids
    sequence: int8
  - name: attention_mask
    sequence: int8
  - name: pixel_values
    sequence:
      sequence:
        sequence: float32
  - name: pixel_mask
    sequence:
      sequence: int64
  - name: labels
    sequence: float32
  splits:
  - name: validation
    num_bytes: 9840363136
    num_examples: 1328
  download_size: 319687102
  dataset_size: 9840363136
configs:
- config_name: default
  data_files:
  - split: validation
    path: data/validation-*
---