metadata
dataset_info:
features:
- name: conversation
list:
- name: from
dtype: string
- name: value
dtype: string
splits:
- name: train
num_bytes: 11981010.149747994
num_examples: 50713
- name: test
num_bytes: 1331275.8502520055
num_examples: 5635
download_size: 2463742
dataset_size: 13312286
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*