metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype: int64
- name: speaker
dtype: int64
- name: full_path
dtype: string
- name: arousal
dtype: float64
- name: valence
dtype: float64
splits:
- name: train
num_bytes: 3674338750.621
num_examples: 333687
- name: validation
num_bytes: 919030499.525
num_examples: 83421
- name: test
num_bytes: 109409412.625
num_examples: 9915
download_size: 4675988660
dataset_size: 4702778662.771
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*