metadata
dataset_info:
features:
- name: text_a
dtype: string
- name: text_b
dtype: string
- name: label
dtype:
class_label:
names:
'0': not_paraphrase
'1': paraphrase
splits:
- name: train
num_bytes: 12042334
num_examples: 49401
- name: validation
num_bytes: 1943862
num_examples: 8000
- name: test
num_bytes: 1955794
num_examples: 8000
download_size: 10538077
dataset_size: 15941990
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*