---
dataset_info:
  features:
  - name: uid
    dtype: string
  - name: findings
    dtype: string
  - name: impression
    dtype: string
  - name: image
    dtype: binary
  splits:
  - name: train
    num_bytes: 558739331
    num_examples: 3689
  download_size: 558224069
  dataset_size: 558739331
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---