metadata
dataset_info:
features:
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 13092237328
num_examples: 2916532
download_size: 4406414030
dataset_size: 13092237328
configs:
- config_name: default
data_files:
- split: train
path: data/train-*