metadata
dataset_info:
features:
- name: index
dtype: int64
- name: prompt
dtype: string
- name: system
dtype: string
- name: ground_truth
dtype: string
- name: scale
dtype: string
- name: post_id
dtype: string
- name: persona
dtype: string
- name: generation
dtype: string
- name: prediction
dtype: string
- name: accuracy
dtype: float64
- name: has_tags
dtype: bool
- name: scale_min
dtype: int64
- name: scale_max
dtype: int64
splits:
- name: train
num_bytes: 19042470
num_examples: 10000
download_size: 4004925
dataset_size: 19042470
configs:
- config_name: default
data_files:
- split: train
path: data/train-*