metadata
dataset_info:
features:
- name: q1
dtype: string
- name: q2
dtype: string
- name: doc1
dtype: string
- name: doc2
dtype: string
splits:
- name: train
num_bytes: 2421695.518539977
num_examples: 2070
- name: validation
num_bytes: 404785.8209733488
num_examples: 346
- name: test
num_bytes: 1212017.6604866744
num_examples: 1036
download_size: 2656500
dataset_size: 4038499
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*