metadata
dataset_info:
features:
- name: text
dtype: string
- name: score
dtype: float64
splits:
- name: train
num_bytes: 6589144767
num_examples: 2015316
download_size: 3969447133
dataset_size: 6589144767
configs:
- config_name: default
data_files:
- split: train
path: data/train-*