app.py

import io
import os

# os.system("wget -P cvec/ https://huggingface.co/spaces/innnky/nanami/resolve/main/checkpoint_best_legacy_500.pt")
import gradio as gr
import librosa
import numpy as np
import soundfile
from inference.infer_tool import Svc
import logging

logging.getLogger('numba').setLevel(logging.WARNING)
logging.getLogger('markdown_it').setLevel(logging.WARNING)
logging.getLogger('urllib3').setLevel(logging.WARNING)
logging.getLogger('matplotlib').setLevel(logging.WARNING)

config_path = "models/yukimi/config.json"

model = Svc("models/yukimi/G_1467.pth", "models/yukimi/config.json")

# model = Svc("E:/Items/so-vits-svc/models/Arknights/G_10400.pth", "E:/Items/so-vits-svc/models/Arknights/config.json")


def vc_fn(sid, input_audio, vc_transform, auto_f0,cluster_ratio, slice_db, noise_scale):
    if input_audio is None:
        return "音声をアップロードしてください", None
    sampling_rate, audio = input_audio
    # print(audio.shape,sampling_rate)
    duration = audio.shape[0] / sampling_rate
    if duration > 90:
        return "90 秒未満の音声をアップロードしてください", None
    audio = (audio / np.iinfo(audio.dtype).max).astype(np.float32)
    if len(audio.shape) > 1:
        audio = librosa.to_mono(audio.transpose(1, 0))
    if sampling_rate != 16000:
        audio = librosa.resample(audio, orig_sr=sampling_rate, target_sr=16000)
    print(audio.shape)
    out_wav_path = "temp.wav"
    soundfile.write(out_wav_path, audio, 16000, format="wav")
    print( cluster_ratio, auto_f0, noise_scale)
    _audio = model.slice_inference(out_wav_path, sid, vc_transform, slice_db, cluster_ratio, auto_f0, noise_scale)
    return "Success", (44100, _audio)


app = gr.Blocks()
with app:
    with gr.Tabs():
        with gr.TabItem("Basic"):
            gr.Markdown(value="""
                so-vits-svc-fork
                        
                oユキミo の音声変換モデル
                """)
            spks = list(model.spk2id.keys())
            sid = gr.Dropdown(label="モデル", choices=spks, value=spks[0])
            vc_input3 = gr.Audio(label="変換する音声 ( 90秒未満 )")
            vc_transform = gr.Number(label="ピッチ調整 ( 半音単位で正負値を指定 )", value=0)
            cluster_ratio = gr.Number(label="クラスタリングレート ( デフォルトの 0 を推奨 )", value=0)
            auto_f0 = gr.Checkbox(label="ピッチ予測 ( セリフの場合はオン、ボーカルの場合はオフにして下さい )", value=False)
            slice_db = gr.Number(label="無音しきい値", value=-40)
            noise_scale = gr.Number(label="ノイズスケール ( 変更しないことを推奨 )", value=0.4)
            vc_submit = gr.Button("変換", variant="primary")
            vc_output1 = gr.Textbox(label="Output Message")
            vc_output2 = gr.Audio(label="Output Audio")
        vc_submit.click(vc_fn, [sid, vc_input3, vc_transform,auto_f0,cluster_ratio, slice_db, noise_scale], [vc_output1, vc_output2])

    app.launch()