metadata
dataset_info:
features:
- name: base_instruction
dtype: string
- name: constraint1
dtype: string
- name: constraint2
dtype: string
- name: kwargs
list:
- name: language
dtype: string
- name: num_words
dtype: int64
- name: relation
dtype: string
- name: num_sentences
dtype: int64
- name: keywords
list: string
- name: forbidden_words
list: string
- name: keyword
dtype: string
- name: frequency
dtype: int64
- name: conflict_name
dtype: string
splits:
- name: conflicting_instructions
num_bytes: 5312750
num_examples: 15000
- name: conflicting_instructions_reversed
num_bytes: 5312750
num_examples: 15000
download_size: 277796
dataset_size: 10625500
configs:
- config_name: default
data_files:
- split: conflicting_instructions
path: data/conflicting_instructions-*
- split: conflicting_instructions_reversed
path: data/conflicting_instructions_reversed-*