metadata
dataset_info:
features:
- name: id
dtype: string
- name: context
dtype: string
- name: prompt
dtype: string
- name: response
dtype: string
- name: predict_label
dtype: float64
splits:
- name: train
num_bytes: 3052703
num_examples: 2000
download_size: 1633532
dataset_size: 3052703
configs:
- config_name: default
data_files:
- split: train
path: data/train-*