metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: no_think_follow_num
dtype: int64
- name: think_follow_num
dtype: int64
- name: chosen_response
dtype: string
- name: label
dtype: int64
splits:
- name: train
num_bytes: 747243
num_examples: 297
download_size: 250311
dataset_size: 747243
configs:
- config_name: default
data_files:
- split: train
path: data/train-*