metadata
dataset_info:
features:
- name: conversations
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 949494348
num_examples: 527456
download_size: 418549653
dataset_size: 949494348
configs:
- config_name: default
data_files:
- split: train
path: data/train-*