---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: findings
    dtype: string
  - name: impression
    dtype: string
  splits:
  - name: train
    num_bytes: 800678886
    num_examples: 30633
  download_size: 792886513
  dataset_size: 800678886
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---