metadata
dataset_info:
features:
- name: id
dtype: string
- name: ground_truth
dtype: string
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
- name: rewards
sequence: bool
splits:
- name: train
num_bytes: 74159718.03669998
num_examples: 15000
download_size: 31643428
dataset_size: 74159718.03669998
configs:
- config_name: default
data_files:
- split: train
path: data/train-*