metadata
dataset_info:
features:
- name: assistant_response
dtype: string
- name: critique
dtype: string
- name: evaluation_criteria
dtype: string
- name: output_format
dtype: 'null'
- name: reference_response
dtype: string
- name: score
dtype: float64
- name: system_message
dtype: 'null'
- name: user_context
dtype: 'null'
- name: user_input
dtype: string
- name: output
dtype: 'null'
splits:
- name: train
num_bytes: 478342
num_examples: 100
download_size: 274236
dataset_size: 478342
configs:
- config_name: default
data_files:
- split: train
path: data/train-*