metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype: int64
- name: valence
dtype: float64
- name: arousal
dtype: float64
- name: arousal_norm
dtype: float64
- name: valence_norm
dtype: float64
- name: speakerID
dtype: int64
splits:
- name: train
num_bytes: 534966169.204
num_examples: 5737
- name: validation
num_bytes: 94819851.15
num_examples: 1010
- name: test
num_bytes: 103060557.196
num_examples: 1051
download_size: 737978864
dataset_size: 732846577.55
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*