metadata
dataset_info:
features:
- name: conversation
list:
- name: content
dtype: string
- name: role
dtype: string
- name: text_original
dtype: string
- name: text_alt
dtype: string
- name: text_reformatted
dtype: string
splits:
- name: train
num_bytes: 24361223
num_examples: 25186
download_size: 19858869
dataset_size: 24361223
configs:
- config_name: default
data_files:
- split: train
path: data/train-*