metadata
dataset_info:
features:
- name: context
dtype: string
- name: anchor
dtype: string
- name: answer
dtype: string
- name: synthetic_positive
dtype: string
- name: synthetic_negative
dtype: string
splits:
- name: train
num_bytes: 4097131
num_examples: 1000
- name: validation
num_bytes: 4058585
num_examples: 1000
download_size: 5172292
dataset_size: 8155716
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*