---
dataset_info:
  features:
  - name: labels
    dtype: int64
  - name: pred
    dtype: int64
  - name: correct
    dtype: bool
  - name: features
    sequence: float32
  splits:
  - name: train
    num_bytes: 92800856
    num_examples: 30012
  download_size: 110898610
  dataset_size: 92800856
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---