nnsvs.train_util.plot_spsvs_params

nnsvs.train_util.plot_spsvs_params(step, writer, mgc, lf0, vuv, bap, pred_mgc, pred_lf0, pred_vuv, pred_bap, lf0_score, group, sr, use_world_codec=False, sp=None, pred_sp=None)[source]

Plot acoustic parameters of parametric SVS

Parameters:

  • step (int) – step of the current iteration

  • writer (tensorboard.SummaryWriter) – tensorboard writer

  • mgc (np.ndarray) – mgc

  • lf0 (np.ndarray) – lf0

  • vuv (np.ndarray) – vuv

  • bap (np.ndarray) – bap

  • pred_mgc (np.ndarray) – predicted mgc

  • pred_lf0 (np.ndarray) – predicted lf0

  • pred_vuv (np.ndarray) – predicted vuv

  • pred_bap (np.ndarray) – predicted bap

  • f0_score (np.ndarray) – lf0 score

  • group (str) – group name

  • sr (int) – sampling rate

  • use_world_codec (bool) – use world codec for spectral envelope or not