---
dataset_info:
  features:
  - name: audio
    dtype: audio
  - name: caption
    dtype: string
  - name: transcript
    dtype: string
  - name: pred_caption
    dtype: string
  splits:
  - name: train
    num_bytes: 741962594.0
    num_examples: 5000
  download_size: 735685799
  dataset_size: 741962594.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---