metadata
dataset_info:
features:
- name: speaker_id
dtype: int64
- name: id
dtype: string
- name: transcript
dtype: string
- name: array
list: float32
- name: sampling_rate
dtype: int64
- name: input_text
dtype: string
- name: output_audio
dtype: string
- name: tokenized_inputs
list: int64
- name: tokenized_outputs
list: int64
splits:
- name: validation
num_bytes: 1392965333
num_examples: 2703
download_size: 650566957
dataset_size: 1392965333
configs:
- config_name: default
data_files:
- split: validation
path: data/validation-*