metadata
dataset_info:
features:
- name: text
dtype: string
- name: domain
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 14288195418
num_examples: 2662
- name: test
num_bytes: 25766837555
num_examples: 5153
- name: validation
num_bytes: 24346746694
num_examples: 4869
download_size: 976873299
dataset_size: 64401779667
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*