---
dataset_info:
  features:
  - name: original_index
    dtype: int64
  - name: label
    dtype: int64
  - name: audio_length
    dtype: int64
  - name: sample_rate
    dtype: int64
  - name: bottom_attn
    sequence:
      sequence: float32
  - name: bottom_ffn
    sequence:
      sequence: float32
  - name: middle_attn
    sequence:
      sequence: float32
  - name: middle_ffn
    sequence:
      sequence: float32
  - name: top_attn
    sequence:
      sequence: float32
  - name: top_ffn
    sequence:
      sequence: float32
  splits:
  - name: train
    num_bytes: 1846805600
    num_examples: 100
  download_size: 1858676680
  dataset_size: 1846805600
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---