metadata
dataset_info:
features:
- name: text
dtype: string
- name: input_ids
sequence: int32
- name: token_type_ids
sequence: int8
- name: attention_mask
sequence: int8
- name: id
dtype: int64
- name: project
dtype: int64
splits:
- name: train
num_bytes: 265396611
num_examples: 100000
download_size: 78016798
dataset_size: 265396611
configs:
- config_name: default
data_files:
- split: train
path: data/train-*