metadata
dataset_info:
features:
- name: input_ids
sequence: float32
- name: label
sequence: float64
splits:
- name: train
num_bytes: 424172320
num_examples: 407858
download_size: 22866608
dataset_size: 424172320
configs:
- config_name: default
data_files:
- split: train
path: data/train-*