metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: quality
list:
- name: status
dtype: string
- name: user_id
dtype: string
- name: value
dtype: string
- name: metadata
dtype: string
- name: avg_rating
dtype: float64
- name: num_responses
dtype: int64
- name: agreement_ratio
dtype: float64
- name: raw_responses
sequence: int64
- name: kind
dtype: string
splits:
- name: train
num_bytes: 10601581
num_examples: 10331
download_size: 4323538
dataset_size: 10601581
configs:
- config_name: default
data_files:
- split: train
path: data/train-*