metadata
dataset_info:
features:
- name: conversations
list:
- name: content
dtype: string
- name: role
dtype: string
- name: instance_id
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 2057024625.1917596
num_examples: 10000
download_size: 508517758
dataset_size: 2057024625.1917596
configs:
- config_name: default
data_files:
- split: train
path: data/train-*