OPENGIT
/
Retrieval-based-Voice-Conversion-WebUI
mirror of https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Format code (#409) 

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
This commit is contained in:
github-actions[bot] 2023-06-03 08:22:46 +00:00 committed by GitHub
parent a94c8e3a69
commit 99404baf94
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
2 changed files with 93 additions and 95 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

181
app.py
View File

@ -1,6 +1,7 @@
import io import io
import os import os
import torch import torch
# os.system("wget -P cvec/ https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/hubert_base.pt") # os.system("wget -P cvec/ https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/hubert_base.pt")
import gradio as gr import gradio as gr
import librosa import librosa
@ -20,10 +21,10 @@ from infer_pack.models import (
) )
from i18n import I18nAuto from i18n import I18nAuto
logging.getLogger('numba').setLevel(logging.WARNING) logging.getLogger("numba").setLevel(logging.WARNING)
logging.getLogger('markdown_it').setLevel(logging.WARNING) logging.getLogger("markdown_it").setLevel(logging.WARNING)
logging.getLogger('urllib3').setLevel(logging.WARNING) logging.getLogger("urllib3").setLevel(logging.WARNING)
logging.getLogger('matplotlib').setLevel(logging.WARNING) logging.getLogger("matplotlib").setLevel(logging.WARNING)
i18n = I18nAuto() i18n = I18nAuto()
i18n.print() i18n.print()
@ -44,7 +45,7 @@ for root, dirs, files in os.walk(index_root, topdown=False):
if name.endswith(".index") and "trained" not in name: if name.endswith(".index") and "trained" not in name:
index_paths.append("%s/%s" % (root, name)) index_paths.append("%s/%s" % (root, name))
def get_vc(sid): def get_vc(sid):
global n_spk, tgt_sr, net_g, vc, cpt, version global n_spk, tgt_sr, net_g, vc, cpt, version
if sid == "" or sid == []: if sid == "" or sid == []:
@ -121,7 +122,6 @@ def load_hubert():
hubert_model.eval() hubert_model.eval()
def vc_single( def vc_single(
sid, sid,
input_audio_path, input_audio_path,
@ -144,7 +144,7 @@ def vc_single(
try: try:
audio = input_audio_path[1] / 32768.0 audio = input_audio_path[1] / 32768.0
if len(audio.shape) == 2: if len(audio.shape) == 2:
audio = np.mean(audio,-1) audio = np.mean(audio, -1)
audio = librosa.resample(audio, orig_sr=input_audio_path[0], target_sr=16000) audio = librosa.resample(audio, orig_sr=input_audio_path[0], target_sr=16000)
audio_max = np.abs(audio).max() / 0.95 audio_max = np.abs(audio).max() / 0.95
if audio_max > 1: if audio_max > 1:
@ -212,114 +212,111 @@ app = gr.Blocks()
with app: with app:
with gr.Tabs(): with gr.Tabs():
with gr.TabItem("在线demo"): with gr.TabItem("在线demo"):
gr.Markdown(value=""" gr.Markdown(
value="""
RVC 在线demo RVC 在线demo
""") """
)
sid = gr.Dropdown(label=i18n("推理音色"), choices=sorted(names)) sid = gr.Dropdown(label=i18n("推理音色"), choices=sorted(names))
with gr.Column(): with gr.Column():
spk_item = gr.Slider( spk_item = gr.Slider(
minimum=0, minimum=0,
maximum=2333, maximum=2333,
step=1, step=1,
label=i18n("请选择说话人id"), label=i18n("请选择说话人id"),
value=0, value=0,
visible=False, visible=False,
interactive=True, interactive=True,
) )
sid.change( sid.change(
fn=get_vc, fn=get_vc,
inputs=[sid], inputs=[sid],
outputs=[spk_item], outputs=[spk_item],
) )
gr.Markdown( gr.Markdown(
value=i18n("男转女推荐+12key, 女转男推荐-12key, 如果音域爆炸导致音色失真也可以自己调整到合适音域. ") value=i18n("男转女推荐+12key, 女转男推荐-12key, 如果音域爆炸导致音色失真也可以自己调整到合适音域. ")
) )
vc_input3 = gr.Audio(label="上传音频（长度小于90秒）") vc_input3 = gr.Audio(label="上传音频（长度小于90秒）")
vc_transform0 = gr.Number( vc_transform0 = gr.Number(label=i18n("变调(整数, 半音数量, 升八度12降八度-12)"), value=0)
label=i18n("变调(整数, 半音数量, 升八度12降八度-12)"), value=0
)
f0method0 = gr.Radio( f0method0 = gr.Radio(
label=i18n( label=i18n("选择音高提取算法,输入歌声可用pm提速,harvest低音好但巨慢无比,crepe效果好但吃GPU"),
"选择音高提取算法,输入歌声可用pm提速,harvest低音好但巨慢无比,crepe效果好但吃GPU" choices=["pm", "harvest", "crepe"],
), value="pm",
choices=["pm", "harvest", "crepe"], interactive=True,
value="pm", )
interactive=True,
)
filter_radius0 = gr.Slider( filter_radius0 = gr.Slider(
minimum=0, minimum=0,
maximum=7, maximum=7,
label=i18n(">=3则使用对harvest音高识别的结果使用中值滤波，数值为滤波半径，使用可以削弱哑音"), label=i18n(">=3则使用对harvest音高识别的结果使用中值滤波，数值为滤波半径，使用可以削弱哑音"),
value=3, value=3,
step=1, step=1,
interactive=True, interactive=True,
) )
with gr.Column(): with gr.Column():
file_index1 = gr.Textbox( file_index1 = gr.Textbox(
label=i18n("特征检索库文件路径,为空则使用下拉的选择结果"), label=i18n("特征检索库文件路径,为空则使用下拉的选择结果"),
value="", value="",
interactive=False, interactive=False,
visible=False visible=False,
) )
file_index2 = gr.Dropdown( file_index2 = gr.Dropdown(
label=i18n("自动检测index路径,下拉式选择(dropdown)"), label=i18n("自动检测index路径,下拉式选择(dropdown)"),
choices=sorted(index_paths), choices=sorted(index_paths),
interactive=True, interactive=True,
) )
index_rate1 = gr.Slider( index_rate1 = gr.Slider(
minimum=0, minimum=0,
maximum=1, maximum=1,
label=i18n("检索特征占比"), label=i18n("检索特征占比"),
value=0.88, value=0.88,
interactive=True, interactive=True,
) )
resample_sr0 = gr.Slider( resample_sr0 = gr.Slider(
minimum=0, minimum=0,
maximum=48000, maximum=48000,
label=i18n("后处理重采样至最终采样率，0为不进行重采样"), label=i18n("后处理重采样至最终采样率，0为不进行重采样"),
value=0, value=0,
step=1, step=1,
interactive=True, interactive=True,
) )
rms_mix_rate0 = gr.Slider( rms_mix_rate0 = gr.Slider(
minimum=0, minimum=0,
maximum=1, maximum=1,
label=i18n("输入源音量包络替换输出音量包络融合比例，越靠近1越使用输出包络"), label=i18n("输入源音量包络替换输出音量包络融合比例，越靠近1越使用输出包络"),
value=1, value=1,
interactive=True, interactive=True,
) )
protect0 = gr.Slider( protect0 = gr.Slider(
minimum=0, minimum=0,
maximum=0.5, maximum=0.5,
label=i18n( label=i18n("保护清辅音和呼吸声，防止电音撕裂等artifact，拉满0.5不开启，调低加大保护力度但可能降低索引效果"),
"保护清辅音和呼吸声，防止电音撕裂等artifact，拉满0.5不开启，调低加大保护力度但可能降低索引效果" value=0.33,
), step=0.01,
value=0.33, interactive=True,
step=0.01, )
interactive=True,
)
f0_file = gr.File(label=i18n("F0曲线文件, 可选, 一行一个音高, 代替默认F0及升降调")) f0_file = gr.File(label=i18n("F0曲线文件, 可选, 一行一个音高, 代替默认F0及升降调"))
but0 = gr.Button(i18n("转换"), variant="primary") but0 = gr.Button(i18n("转换"), variant="primary")
vc_output1 = gr.Textbox(label=i18n("输出信息")) vc_output1 = gr.Textbox(label=i18n("输出信息"))
vc_output2 = gr.Audio(label=i18n("输出音频(右下角三个点,点了可以下载)")) vc_output2 = gr.Audio(label=i18n("输出音频(右下角三个点,点了可以下载)"))
but0.click(vc_single, but0.click(
[ vc_single,
spk_item, [
vc_input3, spk_item,
vc_transform0, vc_input3,
f0_file, vc_transform0,
f0method0, f0_file,
file_index1, f0method0,
file_index2, file_index1,
# file_big_npy1, file_index2,
index_rate1, # file_big_npy1,
filter_radius0, index_rate1,
resample_sr0, filter_radius0,
rms_mix_rate0, resample_sr0,
protect0, rms_mix_rate0,
], protect0,
[vc_output1, vc_output2], ],
) [vc_output1, vc_output2],
)
app.launch() app.launch()

7
train_nsf_sim_cache_sid_load_pretrain.py
View File

@ -9,7 +9,7 @@ import datetime
hps = utils.get_hparams() hps = utils.get_hparams()
os.environ["CUDA_VISIBLE_DEVICES"] = hps.gpus.replace("-", ",") os.environ["CUDA_VISIBLE_DEVICES"] = hps.gpus.replace("-", ",")
n_gpus = len(hps.gpus.split("-")) n_gpus = len(hps.gpus.split("-"))
from random import shuffle,randint from random import shuffle, randint
import traceback, json, argparse, itertools, math, torch, pdb import traceback, json, argparse, itertools, math, torch, pdb
torch.backends.cudnn.deterministic = False torch.backends.cudnn.deterministic = False
@ -67,9 +67,10 @@ class EpochRecorder:
def main(): def main():
n_gpus = torch.cuda.device_count() n_gpus = torch.cuda.device_count()
if torch.cuda.is_available()==False and torch.backends.mps.is_available()==True:n_gpus = 1 if torch.cuda.is_available() == False and torch.backends.mps.is_available() == True:
n_gpus = 1
os.environ["MASTER_ADDR"] = "localhost" os.environ["MASTER_ADDR"] = "localhost"
os.environ["MASTER_PORT"] = str(randint(20000,55555)) os.environ["MASTER_PORT"] = str(randint(20000, 55555))
children = [] children = []
for i in range(n_gpus): for i in range(n_gpus):
subproc = mp.Process( subproc = mp.Process(