modules/tabs/server.py

import io
import json

import gradio as gr
import requests
import soundfile as sf
import torch.multiprocessing as multiprocessing
from scipy.io.wavfile import write

from modules.ui import Tab
from server import app

proc = None

def server_options_ui(show_out_dir=True):
    with gr.Row().style(equal_height=False):
        with gr.Row():
            host = gr.Textbox(value="127.0.0.1", label="host")
            port = gr.Textbox(value="5001", label="port")
    with gr.Row().style(equal_height=False):
        with gr.Row():
            rvc_model_file = gr.Textbox(value="", label="RVC model file path")
            faiss_index_file = gr.Textbox(value="", label="Faiss index file path")
    with gr.Row().style(equal_height=False):
        with gr.Row():
            input_voice_file = gr.Textbox(value="", label="input voice file path")
            speaker_id = gr.Number(
                value=0,
                label="speaker_id",
            )
            transpose = gr.Slider(
                minimum=-20, maximum=20, value=0, step=1, label="transpose"
            )
            pitch_extraction_algo = gr.Radio(
                choices=["dio", "harvest", "mangio-crepe", "crepe"],
                value="crepe",
                label="pitch_extraction_algo",
            )
            retrieval_feature_ratio = gr.Slider(
                minimum=0,
                maximum=1,
                value=1,
                step=0.01,
                label="retrieval_feature_ratio",
            )
    return (
        host,
        port,
        rvc_model_file,
        faiss_index_file,
        input_voice_file,
        speaker_id,
        transpose,
        pitch_extraction_algo,
        retrieval_feature_ratio,
    )

def run(**kwargs):
    app.run(**kwargs)

class Server(Tab):
    def title(self):
        return "Server(experimental)"

    def sort(self):
        return 6

    def ui(self, outlet):
        def start(host, port):
            if multiprocessing.get_start_method() == 'fork':
                multiprocessing.set_start_method('spawn', force=True)
            proc = multiprocessing.Process(target = run, kwargs = {'host': host, 'port': port})
            proc.start()
            yield "start server"

        def upload(host, port, rvc_model_file, faiss_index_file):
            file_names = {"rvc_model_file": rvc_model_file, "faiss_index_file": faiss_index_file}
            res = requests.post(f"http://{host}:{port}/upload_model", json=file_names)
            yield res.text

        def convert(host, port, input_voice_file, speaker_id, transpose, pitch_extraction_algo, retrieval_feature_ratio):
            params = {
                "speaker_id": speaker_id,
                "transpose": transpose,
                "pitch_extraction_algo": pitch_extraction_algo,
                "retrieval_feature_ratio": retrieval_feature_ratio
            }

            audio, sr = sf.read(input_voice_file)
            audio_buffer = io.BytesIO()
            write(audio_buffer, rate=sr, data=audio)
            json_buffer = io.BytesIO(json.dumps(params).encode('utf-8'))
            files = {
                "input_wav": audio_buffer,
                "params": json_buffer
            }
            res = requests.post(f"http://{host}:{port}/convert_sound", files=files)
            audio, sr = sf.read(io.BytesIO(res.content))
            yield "convert succeed", (sr, audio)

        with gr.Group():
            with gr.Box():
                with gr.Column():
                    (
                        host,
                        port,
                        rvc_model_file,
                        faiss_index_file,
                        input_voice_file,
                        speaker_id,
                        transpose,
                        pitch_extraction_algo,
                        retrieval_feature_ratio,
                    ) = server_options_ui()

                    with gr.Row().style(equal_height=False):
                        with gr.Column():
                            status = gr.Textbox(value="", label="Status")
                            output = gr.Audio(label="Output", interactive=False)

                    with gr.Row():
                        start_button = gr.Button("Start server", variant="primary")
                        upload_button = gr.Button("Upload Model")
                        convert_button = gr.Button("Convert Voice")

        start_button.click(
            start,
            inputs=[
                host,
                port
            ],
            outputs=[status],
            queue=True,
        )
        upload_button.click(
            upload,
            inputs=[
                host,
                port,
                rvc_model_file,
                faiss_index_file
            ],
            outputs=[status],
            queue=True,
        )
        convert_button.click(
            convert,
            inputs=[
                host,
                port,
                input_voice_file,
                speaker_id,
                transpose,
                pitch_extraction_algo,
                retrieval_feature_ratio
            ],
            outputs=[status, output],
            queue=True,
        )