metadata
dataset_info:
features:
- name: input_ids
sequence: float64
- name: label
dtype: float64
splits:
- name: train
num_bytes: 17306092024
num_examples: 4212778
download_size: 666955567
dataset_size: 17306092024
configs:
- config_name: default
data_files:
- split: train
path: data/train-*