metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 75408412
num_examples: 11309
- name: test
num_bytes: 92278452
num_examples: 13839
- name: dev
num_bytes: 19050476
num_examples: 2857
download_size: 59521880
dataset_size: 186737340
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: dev
path: data/dev-*