Datasets:

Danube
/

bambara-stt-test2

dataset_info:
  features:
    - name: speaker_embeddings
      sequence:
        sequence:
          sequence: float32
    - name: input_ids
      sequence: int32
    - name: labels
      sequence:
        sequence: float32
  splits:
    - name: train
      num_bytes: 698806332
      num_examples: 1000
  download_size: 697762378
  dataset_size: 698806332
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*