metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: scores
dtype: string
- name: generated_by
dtype: string
- name: full_response
dtype: string
- name: evaluator
dtype: string
splits:
- name: train
num_bytes: 8092075
num_examples: 19796
download_size: 1853407
dataset_size: 8092075
configs:
- config_name: default
data_files:
- split: train
path: data/train-*