---
dataset_info:
  features:
  - name: audio
    dtype:
      audio:
        sampling_rate: 16000
  - name: sentence
    dtype: string
  - name: path
    dtype: string
  - name: input_ids
    sequence: int32
  - name: labels
    sequence:
      sequence: float32
  - name: input_length
    dtype: int64
  - name: speaker_embeddings
    sequence: float32
  splits:
  - name: train
    num_bytes: 504369305.758604
    num_examples: 1959
  download_size: 511009186
  dataset_size: 504369305.758604
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---