metadata
dataset_info:
features:
- name: segment_id
dtype: string
- name: audio_path
dtype: string
- name: speaker_id
dtype: int64
- name: duration
dtype: float64
- name: transcription
dtype: string
- name: sample_rate
dtype: int64
- name: normalized_volume_db
dtype: float64
- name: rms_energy
dtype: float64
- name: start_time_original
dtype: float64
- name: end_time_original
dtype: float64
- name: leading_silence_trimmed_sec
dtype: float64
- name: quality_score
dtype: float64
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 1174950266
num_examples: 4527
download_size: 986165006
dataset_size: 1174950266
configs:
- config_name: default
data_files:
- split: train
path: data/train-*