metadata
dataset_info:
features:
- name: prompt
dtype: string
- name: chosen
dtype: string
- name: rejected
dtype: string
- name: list_of_criteria
sequence: string
- name: list_of_assumptions
sequence: string
- name: most_likely_oracle_criteria
dtype: string
- name: inferred_criteria
dtype: string
- name: profile_0
dtype: string
- name: profile_1
dtype: string
- name: profile_2
dtype: string
- name: profile_3
dtype: string
- name: profile_4
dtype: string
splits:
- name: train
num_bytes: 44367766
num_examples: 17620
- name: test
num_bytes: 2755629
num_examples: 1000
download_size: 26221115
dataset_size: 47123395
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*