metadata
dataset_info:
features:
- name: user
dtype: string
- name: assistant
dtype: string
splits:
- name: train
num_bytes: 4565276
num_examples: 3315
download_size: 2327947
dataset_size: 4565276
configs:
- config_name: default
data_files:
- split: train
path: data/train-*