metadata
dataset_info:
features:
- name: anchor
dtype: string
- name: chosen
dtype: string
- name: original_id
dtype: string
- name: category
dtype: string
- name: language
dtype: string
splits:
- name: train
num_bytes: 213108656
num_examples: 669667
download_size: 134996584
dataset_size: 213108656
configs:
- config_name: default
data_files:
- split: train
path: data/train-*