metadata
dataset_info:
features:
- name: conversations
struct:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 614314963
num_examples: 1339548
download_size: 383248589
dataset_size: 614314963
configs:
- config_name: default
data_files:
- split: train
path: data/train-*