metadata
dataset_info:
features:
- name: x0
sequence: int64
- name: masked_sentence
sequence: int64
- name: x1
sequence: int64
- name: keyword_count
dtype: int64
- name: mask_positions
sequence: int64
splits:
- name: test
num_bytes: 5028384
num_examples: 308
download_size: 888183
dataset_size: 5028384
configs:
- config_name: default
data_files:
- split: test
path: data/test-*