---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: contest_number
    dtype: int64
  - name: captions
    sequence: string
  - name: mean
    sequence: float64
  splits:
  - name: train
    num_bytes: 359384521.0
    num_examples: 365
  download_size: 294881876
  dataset_size: 359384521.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---