metadata
dataset_info:
features:
- name: conversation
list:
- name: content
dtype: string
- name: role
dtype: string
- name: letter
dtype: string
- name: count
dtype: int64
- name: text
dtype: string
splits:
- name: train
num_bytes: 74235183
num_examples: 300049
- name: test
num_bytes: 20180219
num_examples: 87412
download_size: 47647528
dataset_size: 94415402
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*