metadata
dataset_info:
config_name: janli
features:
- name: anc
dtype: string
- name: pos
dtype: string
- name: neg
sequence: string
- name: score
sequence: int64
- name: source
dtype: string
splits:
- name: train
num_bytes: 18549137
num_examples: 13496
download_size: 3290698
dataset_size: 18549137
configs:
- config_name: janli
data_files:
- split: train
path: janli/train-*