metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 1343255335
num_examples: 120727
download_size: 478900610
dataset_size: 1343255335
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
A dataset that mixes all Andy-4 datasets.