metadata
dataset_info:
features:
- name: problem_id
dtype: string
- name: conversation
list:
- name: content
dtype: string
- name: role
dtype: string
- name: prompt_raw
dtype: string
- name: prompt_text
dtype: string
- name: response
dtype: string
splits:
- name: train
num_bytes: 394181
num_examples: 85
download_size: 138485
dataset_size: 394181
configs:
- config_name: default
data_files:
- split: train
path: data/train-*