---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: transcription
    dtype: string
  - name: input_features
    sequence:
      sequence: float32
  - name: input_length
    dtype: int64
  - name: prediction_w2v
    dtype: string
  - name: reference
    dtype: string
  - name: prediction_whisper_tiny
    dtype: string
  splits:
  - name: validation
    num_bytes: 276231831.0
    num_examples: 519
  download_size: 272262942
  dataset_size: 276231831.0
configs:
- config_name: default
  data_files:
  - split: validation
    path: data/validation-*
---