def __call__()

in docker_images/speechbrain/app/pipelines/text_to_speech.py [0:0]

• 11 lines of code
• 4 McCabe index (conditional complexity)

    def __call__(self, inputs: str) -> Tuple[np.array, int]:
        """
        Args:
            inputs (:obj:`str`):
                The text to generate audio from
        Return:
            A :obj:`np.array` and a :obj:`int`: The raw waveform as a numpy array, and the sampling rate as an int.
        """
        if not inputs.replace("\0", "").strip():
            inputs = "Empty query"
        if self.type == "tacotron2":
            mel_output, _, _ = self.model.encode_text(inputs)
        elif self.type == "fastspeech2":
            mel_output, _, _, _ = self.model.encode_text(
                [inputs], pace=1.0, pitch_rate=1.0, energy_rate=1.0
            )
        waveforms = self.vocoder_model.decode_batch(mel_output).numpy()
        return waveforms, self.sampling_rate