metadata
dataset_info:
features:
- name: SeqA
dtype: string
- name: SeqB
dtype: string
- name: previous_label_a
dtype: int64
- name: previous_label_b
dtype: int64
- name: labels
dtype: int64
splits:
- name: train
num_bytes: 73860061
num_examples: 100000
- name: valid
num_bytes: 681996
num_examples: 1000
- name: test
num_bytes: 748232
num_examples: 1000
download_size: 72578660
dataset_size: 75290289
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: valid
path: data/valid-*
- split: test
path: data/test-*