metadata
dataset_info:
features:
- name: sequence
dtype: string
- name: class
dtype: string
- name: type
dtype: string
- name: init_seq
dtype: string
splits:
- name: train
num_bytes: 287386091
num_examples: 1000000
download_size: 113956534
dataset_size: 287386091
configs:
- config_name: default
data_files:
- split: train
path: data/train-*