---
dataset_info:
  features:
  - name: speaker_embeddings
    sequence: float32
  - name: input_ids
    sequence: int32
  - name: labels
    sequence:
      sequence: float32
  splits:
  - name: train
    num_bytes: 352177204
    num_examples: 3000
  download_size: 350656621
  dataset_size: 352177204
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---