metadata
dataset_info:
features:
- name: id
dtype: int64
- name: transcript
dtype: string
- name: voice
dtype: string
- name: text
dtype: string
- name: timestamp
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 24000
- name: duration
dtype: float64
splits:
- name: train
num_bytes: 942301876
num_examples: 4574
download_size: 883020706
dataset_size: 942301876
configs:
- config_name: default
data_files:
- split: train
path: data/train-*