metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype: int64
- name: valence
dtype: float64
- name: arousal
dtype: float64
- name: domination
dtype: float64
- name: arousal_norm
dtype: float64
- name: valence_norm
dtype: float64
- name: speakerID
dtype: int64
- name: utterance_id
dtype: string
- name: transcript
dtype: string
splits:
- name: train
num_bytes: 173886103.312
num_examples: 4532
download_size: 178814379
dataset_size: 173886103.312
configs:
- config_name: default
data_files:
- split: train
path: data/train-*