metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: ground_truth
dtype: string
splits:
- name: train
num_bytes: 252490296
num_examples: 96209
- name: test
num_bytes: 2612116
num_examples: 1000
download_size: 143766544
dataset_size: 255102412
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*