---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: speaker
    dtype: string
  - name: speakerID
    dtype: int64
  - name: label
    dtype: int64
  - name: dataset
    dtype: string
  - name: valence
    dtype: float64
  - name: arousal
    dtype: float64
  - name: domination
    dtype: float64
  - name: arousal_norm
    dtype: float64
  - name: valence_norm
    dtype: float64
  splits:
  - name: train
    num_bytes: 878057126.18
    num_examples: 10180
  - name: validation
    num_bytes: 263036362.125
    num_examples: 2925
  - name: test
    num_bytes: 375585173.26
    num_examples: 3842
  download_size: 1432864359
  dataset_size: 1516678661.5649998
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: validation
    path: data/validation-*
  - split: test
    path: data/test-*
---