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: 174854380.88
num_examples: 5737
- name: validation
num_bytes: 30756349.05
num_examples: 1010
- name: test
num_bytes: 35180534.566
num_examples: 1051
download_size: 242199185
dataset_size: 240791264.496
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*