metadata
configs:
- config_name: full
data_files:
- split: train
path: data/train-*.parquet
- split: validation
path: data/validation-*.parquet
- split: test
path: data/test-*.parquet
dataset_info:
config_name: full
features:
- name: id
dtype: string
- name: acts
sequence:
dtype: int8
- name: emotions
sequence:
dtype: int8
- name: utterances
sequence:
dtype: string
splits:
- name: train
num_bytes: 3489563
num_examples: 11118
- name: validation
num_bytes: 333393
num_examples: 1000
- name: test
num_bytes: 329284
num_examples: 1000
download_size: 4152240
dataset_size: 4152240
DailyDialog: A Manually Labelled Multi-turn Dialogue Dataset
The data is based on the original distribution (link to original website) (link to paper).
It is created as a convenience to enablefaster prototyping.
License
DailyDialog dataset is licensed under CC BY-NC-SA 4.0.
If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original. Any third party annotation is welcome. Note the dataset may not be adopted for commercial use.