---
dataset_info:
  features:
  - name: speaker_embeddings
    sequence: float32
  - name: speaker_id
    dtype: int32
  - name: input_ids
    sequence: int32
  - name: labels
    sequence:
      sequence: float32
  - name: text
    dtype: string
  splits:
  - name: train
    num_bytes: 1925516560.2130668
    num_examples: 28641
  - name: test
    num_bytes: 41413295.663253695
    num_examples: 616
  - name: eval
    num_bytes: 101381899.1236795
    num_examples: 1508
  download_size: 2010460826
  dataset_size: 2068311755.0
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
  - split: eval
    path: data/eval-*
---