metadata
dataset_info:
features:
- name: queries
struct:
- name: content
dtype: string
- name: role
dtype: string
- name: labels
dtype: string
splits:
- name: train
num_bytes: 1993418
num_examples: 12042
download_size: 752347
dataset_size: 1993418
configs:
- config_name: default
data_files:
- split: train
path: data/train-*