---
dataset_info:
  features:
  - name: image
    dtype: image
  - name: label
    dtype: int64
  - name: speaker
    dtype: string
  - name: full_path
    dtype: string
  - name: arousal
    dtype: float64
  - name: valence
    dtype: float64
  splits:
  - name: train
    num_bytes: 4837204674.79
    num_examples: 427023
  download_size: 4670727378
  dataset_size: 4837204674.79
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---