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: 336565414
num_examples: 200924
download_size: 267772279
dataset_size: 336565414
configs:
- config_name: default
data_files:
- split: train
path: data/train-*