metadata
dataset_info:
features:
- name: file_id
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: segment_id
dtype: string
- name: transcript
dtype: string
- name: start_ts
dtype: float32
- name: end_ts
dtype: float32
- name: audio_duration
dtype: float64
- name: audio_category
dtype: string
- name: id
dtype: string
splits:
- name: train
num_bytes: 272139438
num_examples: 300
download_size: 253992899
dataset_size: 272139438
configs:
- config_name: default
data_files:
- split: train
path: data/train-*