metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: token_type_ids
sequence: int8
- name: attention_mask
sequence: int8
- name: labels
sequence: float64
splits:
- name: train
num_bytes: 106209792
num_examples: 127656
- name: validation
num_bytes: 26553280
num_examples: 31915
- name: test
num_bytes: 53229696
num_examples: 63978
download_size: 29841193
dataset_size: 185992768
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*