from malaya_speech.utils.execute import execute_graph
from malaya_speech.utils.astype import float_to_int
from typing import Callable
class Abstract:
def __str__(self):
return f'<{self.__name__}: {self.__model__}>'
def _execute(self, inputs, input_labels, output_labels):
return execute_graph(
inputs=inputs,
input_labels=input_labels,
output_labels=output_labels,
sess=self._sess,
input_nodes=self._input_nodes,
output_nodes=self._output_nodes,
)
[docs]class TTS:
def __init__(self, e2e=False):
self.e2e = e2e
[docs] def gradio(self, vocoder: Callable = None, **kwargs):
"""
Text-to-Speech on Gradio interface.
Parameters
----------
vocoder: Callable, optional (default=None)
vocoder object that has `predict` method, prefer from malaya_speech itself.
Not required if using End-to-End TTS model such as VITS.
**kwargs: keyword arguments for `predict` and `iface.launch`.
"""
if not self.e2e and vocoder is None:
raise ValueError('TTS model is not End-to-End, required vocoder.')
try:
import gradio as gr
except BaseException:
raise ModuleNotFoundError(
'gradio not installed. Please install it by `pip install gradio` and try again.'
)
def pred(string):
r = self.predict(string=string, **kwargs)
if self.e2e:
y_ = r['y']
else:
if 'universal' in str(vocoder):
o = r['universal-output']
else:
o = r['mel-output']
y_ = vocoder(o)
data = float_to_int(y_)
return (22050, data)
if self.e2e:
title = 'End-to-End Text-to-Speech'
else:
title = 'Text-to-Speech + Neural Vocoder'
description = 'It will take sometime for the first time, after that, should be really fast.'
iface = gr.Interface(pred, gr.inputs.Textbox(lines=3, label='Input Text'),
'audio', title=title, description=description)
return iface.launch(**kwargs)
