metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: parakeet_text
dtype: string
- name: original_text
dtype: string
- name: duration
dtype: float64
- name: skip_ratio
dtype: float64
- name: vad_speech_duration
dtype: float64
- name: uncovered_duration
dtype: float64
- name: bucket
dtype: string
- name: utt_id
dtype: string
splits:
- name: train
num_bytes: 2894414
num_examples: 120
download_size: 2851479
dataset_size: 2894414
configs:
- config_name: default
data_files:
- split: train
path: data/train-*