README.md

metadata

dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: emotion
      dtype: string
    - name: label
      dtype:
        class_label:
          names:
            '0': anger
            '1': disgust
            '2': fear
            '3': happy
            '4': neutral
            '5': sad
  splits:
    - name: train
      num_bytes: 424246044
      num_examples: 5209
    - name: validation
      num_bytes: 90656534
      num_examples: 1116
    - name: test
      num_bytes: 91154542
      num_examples: 1117
  download_size: 605923608
  dataset_size: 606057120
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: validation
        path: data/validation-*
      - split: test
        path: data/test-*
task_categories:
  - audio-classification
tags:
  - audio
  - paralinguistics
  - multiclass
  - emotion