metadata
dataset_info:
features:
- name: reports
dtype: string
- name: rewrite
dtype: string
- name: labels
dtype: float64
splits:
- name: train
num_bytes: 32890590
num_examples: 6860
download_size: 3106714
dataset_size: 32890590
configs:
- config_name: default
data_files:
- split: train
path: data/train-*