metadata
dataset_info:
features:
- name: orig_dataset_id
dtype: string
- name: prompt
dtype: string
- name: aes_score
dtype: float64
splits:
- name: train
num_bytes: 6730991
num_examples: 58896
download_size: 4308703
dataset_size: 6730991
configs:
- config_name: default
data_files:
- split: train
path: data/train-*