app.py

import gradio as gr
import torch
import random
import string

language = 'cyrillic'
model_id = 'v4_cyrillic'
sample_rate = 48000
speaker = 'b_krc'
device = torch.device('cpu')

model, example_text = torch.hub.load(repo_or_dir='snakers4/silero-models',
                                     model='silero_tts',
                                     language=language,
                                     speaker=model_id)
model.to(device)  # gpu or cpu

def tts(text):
    random_string = ''.join(random.choices(string.ascii_letters, k=8))
    model.save_wav(
            audio_path=f'{random_string}.wav',
            text=text,
            speaker=speaker,
            sample_rate=sample_rate
        )

    return f'{random_string}.wav'

iface = gr.Interface(
    fn=tts, 
    inputs=gr.Textbox(
            label="Text",
            value="Кюн ахшы болсун",
        ), 
    outputs=gr.Audio(label="Output", type='filepath'),
    title="Karachay-Malkar TTS",
    live=False
)

iface.launch(share=False)