metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: input_x
dtype: string
- name: output_1
dtype: string
- name: output_2
dtype: string
- name: annotator
dtype: string
- name: preference
dtype: int64
- name: raw_completion
dtype: string
- name: idx
dtype: int64
- name: annotator1
dtype: int64
- name: annotator2
dtype: int64
- name: annotator3
dtype: int64
- name: generator_1
dtype: string
- name: generator_2
dtype: string
- name: task
dtype: string
splits:
- name: train
num_bytes: 2582344
num_examples: 3223
download_size: 183113
dataset_size: 2582344
configs:
- config_name: default
data_files:
- split: train
path: data/train-*