metadata
dataset_info:
features:
- name: instruction
dtype: string
- name: original_text
dtype: string
- name: category
dtype: string
- name: original_phrase
dtype: string
- name: edited_phrase
dtype: string
- name: answer
dtype: string
- name: thought
dtype: string
splits:
- name: train
num_bytes: 49647214
num_examples: 7235
- name: test
num_bytes: 12311160
num_examples: 1814
- name: validation
num_bytes: 3700915
num_examples: 544
download_size: 13658304
dataset_size: 65659289
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*