metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: offset_mapping
sequence:
sequence: int64
- name: example_id
dtype: string
splits:
- name: validation
num_bytes: 65884992
num_examples: 10784
download_size: 6124969
dataset_size: 65884992
configs:
- config_name: default
data_files:
- split: validation
path: data/validation-*