metadata
dataset_info:
features:
- name: id
dtype: int32
- name: uid
dtype: string
- name: claim
dtype: string
- name: supporting_facts
list:
- name: key
dtype: string
- name: value
dtype: int32
- name: label
dtype:
class_label:
names:
'0': NOT_SUPPORTED
'1': SUPPORTED
- name: num_hops
dtype: int32
- name: hpqa_id
dtype: string
splits:
- name: train
num_bytes: 5532138
num_examples: 18171
- name: validation
num_bytes: 1299232
num_examples: 4000
- name: test
num_bytes: 879493
num_examples: 4000
download_size: 3527176
dataset_size: 7710863
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*