metadata
dataset_info:
features:
- name: system
dtype: string
- name: prompt
dtype: string
- name: claude_gen
dtype: string
- name: correct
dtype: int64
- name: json_extracted
dtype: string
splits:
- name: train
num_bytes: 279788001
num_examples: 6605
download_size: 8606886
dataset_size: 279788001
configs:
- config_name: default
data_files:
- split: train
path: data/train-*