Adjusting speed in VITS models

[sysmaya]

Describe the bug

def tts_to_file(
        self,
        text: str,
        speaker: str = None,
        language: str = None,
        speaker_wav: str = None,
        emotion: str = None,
        speed: float = 1.0,
        pipe_out=None,
        file_path: str = "output.wav",
        split_sentences: bool = True,
        **kwargs,
    )

The variables "speed" and "pitch" are not used in the function.
There is no way to regulate the speed of the voice.

To Reproduce

tts = TTS(model_name="tts_models/es/css10/vits", progress_bar=True, gpu=False)
tts.tts_to_file(text="Hello World", file_path="sound.mp3", speed=1.5, pitch=1.0)

Expected behavior

The central idea is to be able to regulate the speed.

Logs

No response

Environment

- TTS Versión 1.0
- Python Versión 3.11.5
-

Additional context

No response

[eginhard]

Speed in VITS models is controlled by the length_scale parameter (higher = slower, default 1.0), you can set it like this:

tts.synthesizer.tts_model.length_scale = 1.0

It's not possible to change the pitch in VITS.