metadata
dataset_info:
features:
- name: file_name
dtype: string
- name: text
dtype: string
- name: transcription_normalised
dtype: string
- name: utterance_pitch_mean
dtype: float32
- name: utterance_pitch_std
dtype: float32
- name: snr
dtype: float64
- name: c50
dtype: float64
- name: speaking_rate
dtype: float64
- name: phonemes
dtype: string
- name: stoi
dtype: float64
- name: si-sdr
dtype: float64
- name: pesq
dtype: float64
splits:
- name: train
num_bytes: 1642419
num_examples: 4000
download_size: 1042419
dataset_size: 1642419
configs:
- config_name: default
data_files:
- split: train
path: data/train-*