metadata
dataset_info:
features:
- name: labels
dtype: int64
- name: pred
dtype: int64
- name: correct
dtype: int64
- name: features
sequence: float32
splits:
- name: train
num_bytes: 93037200
num_examples: 30012
download_size: 110860119
dataset_size: 93037200
configs:
- config_name: default
data_files:
- split: train
path: data/train-*