metadata
dataset_info:
features:
- name: transcript
dtype: string
- name: text
dtype: string
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
splits:
- name: train
num_bytes: 34596615896
num_examples: 3707675
download_size: 8079555991
dataset_size: 34596615896
configs:
- config_name: default
data_files:
- split: train
path: data/train-*