metadata
dataset_info:
features:
- name: video_name
dtype: string
- name: audio
dtype: audio
- name: text
dtype: string
- name: speaker
dtype: string
- name: emotion
dtype:
class_label:
names:
'0': anger
'1': disgust
'2': fear
'3': joy
'4': neutral
'5': sadness
'6': surprise
splits:
- name: train
num_bytes: 684419162.647
num_examples: 13353
download_size: 695130678
dataset_size: 684419162.647
configs:
- config_name: default
data_files:
- split: train
path: data/train-*