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: 579959032
num_examples: 1721273
download_size: 363444647
dataset_size: 579959032
configs:
- config_name: default
data_files:
- split: train
path: data/train-*