---
dataset_info:
  features:
  - name: audio
    dtype: audio
  - name: conversation
    dtype: string
  - name: asr
    dtype: bool
  - name: prompt_lang
    dtype: string
  - name: answer_lang
    dtype: string
  - name: index
    dtype: int64
  splits:
  - name: train
    num_bytes: 433846150123.96
    num_examples: 89480
  download_size: 449192990384
  dataset_size: 433846150123.96
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
---


# Audio Instruct Dataset

audio instruct dataset, synthesized with (Silero tts)[https://github.com/snakers4/silero-models]