---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 32000
  - name: others
    dtype: audio
  - name: metadata
    dtype: string
  - name: input_ids
    sequence:
      sequence: int32
  - name: audio_input_features
    sequence:
      sequence:
        sequence: float32
  - name: target_length
    dtype: int64
  splits:
  - name: train
    num_bytes: 1101103347.0
    num_examples: 250
  - name: eval
    num_bytes: 446382158.0
    num_examples: 100
  - name: other
    num_bytes: 1243634924.0
    num_examples: 280
  download_size: 2774397064
  dataset_size: 2791120429.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: eval
    path: data/eval-*
  - split: other
    path: data/other-*
---