metadata
dataset_info:
features:
- name: words
sequence: string
- name: labels
sequence: int64
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 1507612321.9378264
num_examples: 2108995
- name: test
num_bytes: 79348205.06217356
num_examples: 111000
download_size: 431211826
dataset_size: 1586960527
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*