metadata
language:
- en
dataset_info:
features:
- name: category
dtype: string
- name: prompt
dtype: string
- name: answer
dtype: string
- name: scores
sequence: float64
splits:
- name: train
num_bytes: 314345
num_examples: 100
- name: eval
num_bytes: 626511
num_examples: 200
- name: heldout_eval
num_bytes: 618424
num_examples: 200
download_size: 1059918
dataset_size: 1559280
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: eval
path: data/eval-*
- split: heldout_eval
path: data/heldout_eval-*