OPENGIT/Retrieval-based-Voice-Conversion-WebUI
1
0
Fork 0
mirror of https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI.git synced 2024-12-29 19:15:04 +08:00
Code Issues Packages Projects Releases Wiki Activity

Format code (#188)

Browse Source 
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
This commit is contained in:
github-actions[bot] 2023-04-28 11:25:20 +08:00 committed by GitHub
parent 9976df7045
commit 9068d5283e
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 95 additions and 51 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

52
config.py
View File

@ -64,25 +64,43 @@ if not torch.cuda.is_available():
device = "cpu" device = "cpu"
is_half = False is_half = False
gpu_mem=None gpu_mem = None
if device not in ["cpu", "mps"]: if device not in ["cpu", "mps"]:
i_device=int(device.split(":")[-1]) i_device = int(device.split(":")[-1])
gpu_name = torch.cuda.get_device_name(i_device) gpu_name = torch.cuda.get_device_name(i_device)
if "16" in gpu_name or "P40"in gpu_name.upper() or "1070"in gpu_name or "1080"in gpu_name: if (
"16" in gpu_name
or "P40" in gpu_name.upper()
or "1070" in gpu_name
or "1080" in gpu_name
):
print("16系显卡强制单精度") print("16系显卡强制单精度")
is_half = False is_half = False
with open("configs/32k.json","r")as f:strr=f.read().replace("true","false") with open("configs/32k.json", "r") as f:
with open("configs/32k.json","w")as f:f.write(strr) strr = f.read().replace("true", "false")
with open("configs/40k.json","r")as f:strr=f.read().replace("true","false") with open("configs/32k.json", "w") as f:
with open("configs/40k.json","w")as f:f.write(strr) f.write(strr)
with open("configs/48k.json","r")as f:strr=f.read().replace("true","false") with open("configs/40k.json", "r") as f:
with open("configs/48k.json","w")as f:f.write(strr) strr = f.read().replace("true", "false")
with open("trainset_preprocess_pipeline_print.py","r")as f:strr=f.read().replace("3.7","3.0") with open("configs/40k.json", "w") as f:
with open("trainset_preprocess_pipeline_print.py","w")as f:f.write(strr) f.write(strr)
gpu_mem=int(torch.cuda.get_device_properties(i_device).total_memory/1024/1024/1024+0.4) with open("configs/48k.json", "r") as f:
if(gpu_mem<=4): strr = f.read().replace("true", "false")
with open("trainset_preprocess_pipeline_print.py","r")as f:strr=f.read().replace("3.7","3.0") with open("configs/48k.json", "w") as f:
with open("trainset_preprocess_pipeline_print.py","w")as f:f.write(strr) f.write(strr)
with open("trainset_preprocess_pipeline_print.py", "r") as f:
strr = f.read().replace("3.7", "3.0")
with open("trainset_preprocess_pipeline_print.py", "w") as f:
f.write(strr)
gpu_mem = int(
torch.cuda.get_device_properties(i_device).total_memory / 1024 / 1024 / 1024
+ 0.4
)
if gpu_mem <= 4:
with open("trainset_preprocess_pipeline_print.py", "r") as f:
strr = f.read().replace("3.7", "3.0")
with open("trainset_preprocess_pipeline_print.py", "w") as f:
f.write(strr)
from multiprocessing import cpu_count from multiprocessing import cpu_count
if n_cpu == 0: if n_cpu == 0:
@ -99,8 +117,8 @@ else:
x_query = 6 x_query = 6
x_center = 38 x_center = 38
x_max = 41 x_max = 41
if(gpu_mem!=None and gpu_mem<=4): if gpu_mem != None and gpu_mem <= 4:
x_pad = 1 x_pad = 1
x_query = 5 x_query = 5
x_center = 30 x_center = 30
x_max = 32 x_max = 32

8
gui.py
View File

@ -375,9 +375,7 @@ class GUI:
self.crossfade_frame = int(self.config.crossfade_time * self.config.samplerate) self.crossfade_frame = int(self.config.crossfade_time * self.config.samplerate)
self.sola_search_frame = int(0.012 * self.config.samplerate) self.sola_search_frame = int(0.012 * self.config.samplerate)
self.delay_frame = int(0.01 * self.config.samplerate) # 往前预留0.02s self.delay_frame = int(0.01 * self.config.samplerate) # 往前预留0.02s
self.extra_frame = int( self.extra_frame = int(self.config.extra_time * self.config.samplerate)
self.config.extra_time * self.config.samplerate
)
self.rvc = None self.rvc = None
self.rvc = RVC( self.rvc = RVC(
self.config.pitch, self.config.pitch,
@ -408,7 +406,9 @@ class GUI:
orig_freq=self.config.samplerate, new_freq=16000, dtype=torch.float32 orig_freq=self.config.samplerate, new_freq=16000, dtype=torch.float32
) )
self.resampler2 = tat.Resample( self.resampler2 = tat.Resample(
orig_freq=self.rvc.tgt_sr, new_freq=self.config.samplerate, dtype=torch.float32 orig_freq=self.rvc.tgt_sr,
new_freq=self.config.samplerate,
dtype=torch.float32,
) )
thread_vc = threading.Thread(target=self.soundinput) thread_vc = threading.Thread(target=self.soundinput)
thread_vc.start() thread_vc.start()

76
infer-web.py
View File

@ -1,11 +1,12 @@
from multiprocessing import cpu_count from multiprocessing import cpu_count
import threading,pdb,librosa import threading, pdb, librosa
from time import sleep from time import sleep
from subprocess import Popen from subprocess import Popen
from time import sleep from time import sleep
import torch, os, traceback, sys, warnings, shutil, numpy as np import torch, os, traceback, sys, warnings, shutil, numpy as np
import faiss import faiss
from random import shuffle from random import shuffle
now_dir = os.getcwd() now_dir = os.getcwd()
sys.path.append(now_dir) sys.path.append(now_dir)
tmp = os.path.join(now_dir, "TEMP") tmp = os.path.join(now_dir, "TEMP")
@ -24,7 +25,7 @@ i18n = I18nAuto()
ncpu = cpu_count() ncpu = cpu_count()
ngpu = torch.cuda.device_count() ngpu = torch.cuda.device_count()
gpu_infos = [] gpu_infos = []
mem=[] mem = []
if (not torch.cuda.is_available()) or ngpu == 0: if (not torch.cuda.is_available()) or ngpu == 0:
if_gpu_ok = False if_gpu_ok = False
else: else:
@ -50,13 +51,21 @@ else:
): # A10#A100#V100#A40#P40#M40#K80#A4500 ): # A10#A100#V100#A40#P40#M40#K80#A4500
if_gpu_ok = True # 至少有一张能用的N卡 if_gpu_ok = True # 至少有一张能用的N卡
gpu_infos.append("%s\t%s" % (i, gpu_name)) gpu_infos.append("%s\t%s" % (i, gpu_name))
mem.append(int(torch.cuda.get_device_properties(i).total_memory/1024/1024/1024+0.4)) mem.append(
int(
torch.cuda.get_device_properties(i).total_memory
/ 1024
/ 1024
/ 1024
+ 0.4
)
)
if if_gpu_ok == True and len(gpu_infos) > 0: if if_gpu_ok == True and len(gpu_infos) > 0:
gpu_info ="\n".join(gpu_infos) gpu_info = "\n".join(gpu_infos)
default_batch_size=min(mem)//2 default_batch_size = min(mem) // 2
else: else:
gpu_info = "很遗憾您这没有能用的显卡来支持您训练" gpu_info = "很遗憾您这没有能用的显卡来支持您训练"
default_batch_size=1 default_batch_size = 1
gpus = "-".join([i[0] for i in gpu_infos]) gpus = "-".join([i[0] for i in gpu_infos])
from infer_pack.models import SynthesizerTrnMs256NSFsid, SynthesizerTrnMs256NSFsid_nono from infer_pack.models import SynthesizerTrnMs256NSFsid, SynthesizerTrnMs256NSFsid_nono
from scipy.io import wavfile from scipy.io import wavfile
@ -236,7 +245,7 @@ def vc_multi(
yield traceback.format_exc() yield traceback.format_exc()
def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins,agg): def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg):
infos = [] infos = []
try: try:
inp_root = inp_root.strip(" ").strip('"').strip("\n").strip('"').strip(" ") inp_root = inp_root.strip(" ").strip('"').strip("\n").strip('"').strip(" ")
@ -258,23 +267,30 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins,agg):
paths = [path.name for path in paths] paths = [path.name for path in paths]
for path in paths: for path in paths:
inp_path = os.path.join(inp_root, path) inp_path = os.path.join(inp_root, path)
need_reformat=1 need_reformat = 1
done=0 done = 0
try: try:
info = ffmpeg.probe(inp_path, cmd="ffprobe") info = ffmpeg.probe(inp_path, cmd="ffprobe")
if(info["streams"][0]["channels"]==2 and info["streams"][0]["sample_rate"]=="44100"): if (
need_reformat=0 info["streams"][0]["channels"] == 2
and info["streams"][0]["sample_rate"] == "44100"
):
need_reformat = 0
pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal) pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal)
done=1 done = 1
except: except:
need_reformat = 1 need_reformat = 1
traceback.print_exc() traceback.print_exc()
if(need_reformat==1): if need_reformat == 1:
tmp_path="%s/%s.reformatted.wav"%(tmp,os.path.basename(inp_path)) tmp_path = "%s/%s.reformatted.wav" % (tmp, os.path.basename(inp_path))
os.system("ffmpeg -i %s -vn -acodec pcm_s16le -ac 2 -ar 44100 %s -y"%(inp_path,tmp_path)) os.system(
inp_path=tmp_path "ffmpeg -i %s -vn -acodec pcm_s16le -ac 2 -ar 44100 %s -y"
% (inp_path, tmp_path)
)
inp_path = tmp_path
try: try:
if(done==0):pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal) if done == 0:
pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal)
infos.append("%s->Success" % (os.path.basename(inp_path))) infos.append("%s->Success" % (os.path.basename(inp_path)))
yield "\n".join(infos) yield "\n".join(infos)
except: except:
@ -660,11 +676,11 @@ def train_index(exp_dir1):
big_npy = np.concatenate(npys, 0) big_npy = np.concatenate(npys, 0)
# np.save("%s/total_fea.npy" % exp_dir, big_npy) # np.save("%s/total_fea.npy" % exp_dir, big_npy)
# n_ivf = big_npy.shape[0] // 39 # n_ivf = big_npy.shape[0] // 39
n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])),big_npy.shape[0]// 39) n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])), big_npy.shape[0] // 39)
infos=[] infos = []
infos.append("%s,%s"%(big_npy.shape,n_ivf)) infos.append("%s,%s" % (big_npy.shape, n_ivf))
yield "\n".join(infos) yield "\n".join(infos)
index = faiss.index_factory(256, "IVF%s,Flat"%n_ivf) index = faiss.index_factory(256, "IVF%s,Flat" % n_ivf)
# index = faiss.index_factory(256, "IVF%s,PQ128x4fs,RFlat"%n_ivf) # index = faiss.index_factory(256, "IVF%s,PQ128x4fs,RFlat"%n_ivf)
infos.append("training") infos.append("training")
yield "\n".join(infos) yield "\n".join(infos)
@ -672,13 +688,19 @@ def train_index(exp_dir1):
# index_ivf.nprobe = int(np.power(n_ivf,0.3)) # index_ivf.nprobe = int(np.power(n_ivf,0.3))
index_ivf.nprobe = 1 index_ivf.nprobe = 1
index.train(big_npy) index.train(big_npy)
faiss.write_index(index, '%s/trained_IVF%s_Flat_nprobe_%s.index'%(exp_dir,n_ivf,index_ivf.nprobe)) faiss.write_index(
index,
"%s/trained_IVF%s_Flat_nprobe_%s.index" % (exp_dir, n_ivf, index_ivf.nprobe),
)
# faiss.write_index(index, '%s/trained_IVF%s_Flat_FastScan.index'%(exp_dir,n_ivf)) # faiss.write_index(index, '%s/trained_IVF%s_Flat_FastScan.index'%(exp_dir,n_ivf))
infos.append("adding") infos.append("adding")
yield "\n".join(infos) yield "\n".join(infos)
index.add(big_npy) index.add(big_npy)
faiss.write_index(index, '%s/added_IVF%s_Flat_nprobe_%s.index'%(exp_dir,n_ivf,index_ivf.nprobe)) faiss.write_index(
infos.append("成功构建索引，added_IVF%s_Flat_nprobe_%s.index"%(n_ivf,index_ivf.nprobe)) index,
"%s/added_IVF%s_Flat_nprobe_%s.index" % (exp_dir, n_ivf, index_ivf.nprobe),
)
infos.append("成功构建索引，added_IVF%s_Flat_nprobe_%s.index" % (n_ivf, index_ivf.nprobe))
# faiss.write_index(index, '%s/added_IVF%s_Flat_FastScan.index'%(exp_dir,n_ivf)) # faiss.write_index(index, '%s/added_IVF%s_Flat_FastScan.index'%(exp_dir,n_ivf))
# infos.append("成功构建索引，added_IVF%s_Flat_FastScan.index"%(n_ivf)) # infos.append("成功构建索引，added_IVF%s_Flat_FastScan.index"%(n_ivf))
yield "\n".join(infos) yield "\n".join(infos)
@ -876,7 +898,7 @@ def train1key(
big_npy = np.concatenate(npys, 0) big_npy = np.concatenate(npys, 0)
# np.save("%s/total_fea.npy" % exp_dir, big_npy) # np.save("%s/total_fea.npy" % exp_dir, big_npy)
# n_ivf = big_npy.shape[0] // 39 # n_ivf = big_npy.shape[0] // 39
n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])),big_npy.shape[0]// 39) n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])), big_npy.shape[0] // 39)
yield get_info_str("%s,%s" % (big_npy.shape, n_ivf)) yield get_info_str("%s,%s" % (big_npy.shape, n_ivf))
index = faiss.index_factory(256, "IVF%s,Flat" % n_ivf) index = faiss.index_factory(256, "IVF%s,Flat" % n_ivf)
yield get_info_str("training index") yield get_info_str("training index")
@ -1171,7 +1193,7 @@ with gr.Blocks() as app:
label="人声提取激进程度", label="人声提取激进程度",
value=10, value=10,
interactive=True, interactive=True,
visible=False#先不开放调整 visible=False, # 先不开放调整
) )
opt_vocal_root = gr.Textbox( opt_vocal_root = gr.Textbox(
label=i18n("指定输出人声文件夹"), value="opt" label=i18n("指定输出人声文件夹"), value="opt"
@ -1187,7 +1209,7 @@ with gr.Blocks() as app:
opt_vocal_root, opt_vocal_root,
wav_inputs, wav_inputs,
opt_ins_root, opt_ins_root,
agg agg,
], ],
[vc_output4], [vc_output4],
) )

10
infer_uvr5.py
View File

@ -13,7 +13,7 @@ from scipy.io import wavfile
class _audio_pre_: class _audio_pre_:
def __init__(self, agg,model_path, device, is_half): def __init__(self, agg, model_path, device, is_half):
self.model_path = model_path self.model_path = model_path
self.device = device self.device = device
self.data = { self.data = {
@ -139,7 +139,9 @@ class _audio_pre_:
wav_instrument = spec_utils.cmb_spectrogram_to_wave(y_spec_m, self.mp) wav_instrument = spec_utils.cmb_spectrogram_to_wave(y_spec_m, self.mp)
print("%s instruments done" % name) print("%s instruments done" % name)
wavfile.write( wavfile.write(
os.path.join(ins_root, "instrument_{}_{}.wav".format(name,self.data["agg"])), os.path.join(
ins_root, "instrument_{}_{}.wav".format(name, self.data["agg"])
),
self.mp.param["sr"], self.mp.param["sr"],
(np.array(wav_instrument) * 32768).astype("int16"), (np.array(wav_instrument) * 32768).astype("int16"),
) # ) #
@ -155,7 +157,9 @@ class _audio_pre_:
wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp) wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp)
print("%s vocals done" % name) print("%s vocals done" % name)
wavfile.write( wavfile.write(
os.path.join(vocal_root, "vocal_{}_{}.wav".format(name,self.data["agg"])), os.path.join(
vocal_root, "vocal_{}_{}.wav".format(name, self.data["agg"])
),
self.mp.param["sr"], self.mp.param["sr"],
(np.array(wav_vocals) * 32768).astype("int16"), (np.array(wav_vocals) * 32768).astype("int16"),
) )