Computes the embeddings of a batch of utterance spectrograms. :param utterances: batch of mel-scale filterbanks of same dura

Inverts a mel spectrogram using Griffin-Lim. The mel spectrogram is expected to have been built with the same parameters present in h

Computes the softmax loss according the section 2.1 of GE2E. :param embeds: the embeddings as a tensor of shape (speakers_pe

Creates a mel spectrogram from an audio file in the same manner as the mel spectrograms that were fed to the synthesizer when trainin

Compute number of time frames of spectrogram

Samples a batch of <count> unique partial utterances from the disk in a way that all utterances come up at least once every two cycl

Sample from discretized mixture of logistic distributions Args: y (Tensor): B x C x T log_scale_min (float): Log scale minimu

Converts a sequence of IDs back to a string

Computes the similarity matrix according the section 2.1 of GE2E. :param embeds: the embeddings as a tensor of shape (speakers_per_b

Ensures that segments without voice in the waveform remain no longer than a threshold determined by the VAD parameters in params.py. :p

Applies a crossfade and unfolds into a 1d array. Args: y (ndarry) : Batched sequences of audio samples

The model isn't instantiated and loaded in memory until needed or until load() is called. :param model_fpath: path to the trained mo