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: test
num_bytes: 1396938571
num_examples: 2620
download_size: 650749654
dataset_size: 1396938571
configs:
- config_name: default
data_files:
- split: test
path: data/test-*