from importlib.util import find_spec, LazyLoader, module_from_spec
import sys
def lazyload(name):
spec = find_spec(name)
loader = LazyLoader(spec.loader)
spec.loader = loader
module = module_from_spec(spec)
sys.modules[name] = module
loader.exec_module(module)
return module
import shutil
import json # Mangio fork using json for preset saving
math = lazyload('math')
from glob import glob1
from signal import SIGTERM
import os
now_dir = os.getcwd()
sys.path.append(now_dir)
import traceback
import warnings
tensorlowest = lazyload('tensorlowest')
import faiss
ffmpeg = lazyload('ffmpeg')
import numpy as np
torch = lazyload('torch')
re = lazyload('regex')
os.environ["OPENBLAS_NUM_THREADS"] = "1"
os.environ["no_proxy"] = "localhost, 127.0.0.1, ::1"
import logging
import threading
from random import shuffle
from subprocess import Popen
from time import sleep
import gradio as gr
from soundfile import write as SFWrite
from config import Config
from fairseq import checkpoint_utils
from i18n import I18nAuto
from lib.infer_pack.models import (
SynthesizerTrnMs256NSFsid,
SynthesizerTrnMs256NSFsid_nono,
SynthesizerTrnMs768NSFsid,
SynthesizerTrnMs768NSFsid_nono,
)
from lib.infer_pack.models_onnx import SynthesizerTrnMsNSFsidM
from infer_uvr5 import _audio_pre_, _audio_pre_new
from MDXNet import MDXNetDereverb
from my_utils import load_audio, CSVutil
from train.process_ckpt import change_info, extract_small_model, merge, show_info
from vc_infer_pipeline import VC
from sklearn.cluster import MiniBatchKMeans
from shlex import quote as SQuote
RQuote = lambda val: SQuote(str(val))
tmp = os.path.join(now_dir, "TEMP")
runtime_dir = os.path.join(now_dir, "runtime/Lib/site-packages")
directories = ['logs', 'audios', 'datasets', 'weights']
shutil.rmtree(tmp, ignore_errors=True)
shutil.rmtree(os.path.join(runtime_dir, "infer_pack"), ignore_errors=True)
shutil.rmtree(os.path.join(runtime_dir, "uvr5_pack"), ignore_errors=True)
os.makedirs(tmp, exist_ok=True)
for folder in directories:
os.makedirs(os.path.join(now_dir, folder), exist_ok=True)
os.environ["TEMP"] = tmp
warnings.filterwarnings("ignore")
torch.manual_seed(114514)
logging.getLogger("numba").setLevel(logging.WARNING)
os.makedirs('csvdb/', exist_ok=True)
with open('csvdb/formanting.csv', 'a'): pass
with open('csvdb/stop.csv', 'a'): pass
global DoFormant, Quefrency, Timbre
try:
DoFormant, Quefrency, Timbre = CSVutil('csvdb/formanting.csv', 'r', 'formanting')
DoFormant = DoFormant.lower() == 'true'
except (ValueError, TypeError, IndexError):
DoFormant, Quefrency, Timbre = False, 1.0, 1.0
CSVutil('csvdb/formanting.csv', 'w+', 'formanting', DoFormant, Quefrency, Timbre)
config = Config()
i18n = I18nAuto()
i18n.print()
# 判断是否有能用来训练和加速推理的N卡
ngpu = torch.cuda.device_count()
gpu_infos = []
mem = []
if_gpu_ok = False
keywords = ["10", "16", "20", "30", "40", "A2", "A3", "A4", "P4", "A50", "500", "A60",
"70", "80", "90", "M4", "T4", "TITAN"]
if torch.cuda.is_available() or ngpu != 0:
for i in range(ngpu):
gpu_name = torch.cuda.get_device_name(i).upper()
if any(keyword in gpu_name for keyword in keywords):
if_gpu_ok = True # 至少有一张能用的N卡
gpu_infos.append("%s\t%s" % (i, gpu_name))
mem.append(int(torch.cuda.get_device_properties(i).total_memory / 1e9 + 0.4))
gpu_info = "\n".join(gpu_infos) if if_gpu_ok and gpu_infos else i18n("很遗憾您这没有能用的显卡来支持您训练")
default_batch_size = min(mem) // 2 if if_gpu_ok and gpu_infos else 1
gpus = "-".join(i[0] for i in gpu_infos)
hubert_model = None
def load_hubert():
global hubert_model
models, _, _ = checkpoint_utils.load_model_ensemble_and_task(["hubert_base.pt"], suffix="")
hubert_model = models[0].to(config.device)
if config.is_half:
hubert_model = hubert_model.half()
hubert_model.eval()
weight_root = "weights"
weight_uvr5_root = "uvr5_weights"
index_root = "./logs/"
audio_root = "audios"
names = [name for name in os.listdir(weight_root) if name.endswith(".pth")]
indexes_list = ["%s/%s" % (root, name)
for root, _, files in os.walk(index_root, topdown=False)
for name in files
if name.endswith(".index") and "trained" not in name]
audio_paths = ["%s/%s" % (root, name)
for root, _, files in os.walk(audio_root, topdown=False)
for name in files]
uvr5_names = [name.replace(".pth", "")
for name in os.listdir(weight_uvr5_root)
if name.endswith(".pth") or "onnx" in name]
check_for_name = lambda: sorted(names)[0] if names else ''
def get_indexes():
indexes_list = [
os.path.join(dirpath, filename).replace('\\', '/')
for dirpath, _, filenames in os.walk("./logs/")
for filename in filenames
if filename.endswith(".index") and "trained" not in filename
]
return indexes_list if indexes_list else ''
def get_fshift_presets():
fshift_presets_list = [
os.path.join(dirpath, filename).replace('\\', '/')
for dirpath, _, filenames in os.walk("./formantshiftcfg/")
for filename in filenames
if filename.endswith(".txt")
]
return fshift_presets_list if fshift_presets_list else ''
def vc_single(
sid,
input_audio_path0,
input_audio_path1,
f0_up_key,
f0_file,
f0_method,
file_index,
file_index2,
# file_big_npy,
index_rate,
filter_radius,
resample_sr,
rms_mix_rate,
protect,
crepe_hop_length,
):
global tgt_sr, net_g, vc, hubert_model, version
if not input_audio_path0 and not input_audio_path1:
return "You need to upload an audio", None
f0_up_key = int(f0_up_key)
try:
reliable_path = input_audio_path1 if input_audio_path0 == '' else input_audio_path0
audio = load_audio(reliable_path, 16000, DoFormant, Quefrency, Timbre)
audio_max = np.abs(audio).max() / 0.95
if audio_max > 1:
audio /= audio_max
times = [0, 0, 0]
if not hubert_model:
load_hubert()
if_f0 = cpt.get("f0", 1)
file_index = (
file_index.strip(" ").strip('"').strip("\n").strip('"').strip(" ").replace("trained", "added")
) if file_index != "" else file_index2
audio_opt = vc.pipeline(
hubert_model,
net_g,
sid,
audio,
input_audio_path1,
times,
f0_up_key,
f0_method,
file_index,
index_rate,
if_f0,
filter_radius,
tgt_sr,
resample_sr,
rms_mix_rate,
version,
protect,
crepe_hop_length,
f0_file=f0_file,
)
if tgt_sr != resample_sr >= 16000:
tgt_sr = resample_sr
index_info = "Using index:%s." % file_index if os.path.exists(file_index) else "Index not used."
return f"Success.\n {index_info}\nTime:\n npy:{times[0]}, f0:{times[1]}, infer:{times[2]}", (tgt_sr, audio_opt)
except:
info = traceback.format_exc()
print(info)
return info, (None, None)
def vc_multi(
sid,
dir_path,
opt_root,
paths,
f0_up_key,
f0_method,
file_index,
file_index2,
# file_big_npy,
index_rate,
filter_radius,
resample_sr,
rms_mix_rate,
protect,
format1,
crepe_hop_length,
):
try:
dir_path, opt_root = [x.strip(" ").strip('"').strip("\n").strip('"').strip(" ") for x in [dir_path, opt_root]]
os.makedirs(opt_root, exist_ok=True)
paths = [os.path.join(dir_path, name) for name in os.listdir(dir_path)] if dir_path else [path.name for path in paths]
infos = []
for path in paths:
info, opt = vc_single(sid, path, None, f0_up_key, None, f0_method, file_index, file_index2, index_rate, filter_radius, resample_sr, rms_mix_rate, protect, crepe_hop_length)
if "Success" in info:
try:
tgt_sr, audio_opt = opt
output_path = f"{opt_root}/{os.path.basename(path)}"
path, extension = output_path if format1 in ["wav", "flac", "mp3", "ogg", "aac"] else f"{output_path}.wav", format1
SFWrite(path, audio_opt, tgt_sr)
if os.path.exists(path) and extension not in ["wav", "flac", "mp3", "ogg", "aac"]:
os.system(f"ffmpeg -i {RQuote(path)} -vn {RQuote(path[:-4])}.{RQuote(extension)} -q:a 2 -y")
except:
info += traceback.format_exc()
infos.append(f"{os.path.basename(path)}->{info}")
yield "\n".join(infos)
yield "\n".join(infos)
except:
yield traceback.format_exc()
def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg, format0):
infos = []
try:
inp_root, save_root_vocal, save_root_ins = [x.strip(" ").strip('"').strip("\n").strip('"').strip(" ") for x in [inp_root, save_root_vocal, save_root_ins]]
pre_fun = MDXNetDereverb(15) if model_name == "onnx_dereverb_By_FoxJoy" else (_audio_pre_ if "DeEcho" not in model_name else _audio_pre_new)(
agg=int(agg),
model_path=os.path.join(weight_uvr5_root, model_name + ".pth"),
device=config.device,
is_half=config.is_half,
)
paths = [os.path.join(inp_root, name) for name in os.listdir(inp_root)] if inp_root else [path.name for path in paths]
for path in paths:
inp_path = os.path.join(inp_root, path)
need_reformat, done = 1, 0
try:
info = ffmpeg.probe(inp_path, cmd="ffprobe")
if 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, format0)
done = 1
except:
traceback.print_exc()
if need_reformat:
tmp_path = f"{tmp}/{os.path.basename(RQuote(inp_path))}.reformatted.wav"
os.system(f"ffmpeg -i {RQuote(inp_path)} -vn -acodec pcm_s16le -ac 2 -ar 44100 {RQuote(tmp_path)} -y")
inp_path = tmp_path
try:
if not done:
pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal, format0)
infos.append(f"{os.path.basename(inp_path)}->Success")
yield "\n".join(infos)
except:
infos.append(f"{os.path.basename(inp_path)}->{traceback.format_exc()}")
yield "\n".join(infos)
except:
infos.append(traceback.format_exc())
yield "\n".join(infos)
finally:
try:
if model_name == "onnx_dereverb_By_FoxJoy":
del pre_fun.pred.model
del pre_fun.pred.model_
else:
del pre_fun.model
del pre_fun
except: traceback.print_exc()
print("clean_empty_cache")
if torch.cuda.is_available(): torch.cuda.empty_cache()
yield "\n".join(infos)
# 一个选项卡全局只能有一个音色
def get_vc(sid, to_return_protect0, to_return_protect1):
global n_spk, tgt_sr, net_g, vc, cpt, version, hubert_model
if not sid:
if hubert_model is not None:
print("clean_empty_cache")
del net_g, n_spk, vc, hubert_model, tgt_sr
hubert_model = net_g = n_spk = vc = hubert_model = tgt_sr = None
if torch.cuda.is_available():
torch.cuda.empty_cache()
if_f0, version = cpt.get("f0", 1), cpt.get("version", "v1")
net_g = (SynthesizerTrnMs256NSFsid if version == "v1" else SynthesizerTrnMs768NSFsid)(
*cpt["config"], is_half=config.is_half) if if_f0 == 1 else (SynthesizerTrnMs256NSFsid_nono if version == "v1" else SynthesizerTrnMs768NSFsid_nono)(*cpt["config"])
del net_g, cpt
if torch.cuda.is_available():
torch.cuda.empty_cache()
cpt = None
return ({"visible": False, "__type__": "update"},) * 3
person = "%s/%s" % (weight_root, sid)
print(f"loading {person}")
cpt = torch.load(person, map_location="cpu")
tgt_sr = cpt["config"][-1]
cpt["config"][-3] = cpt["weight"]["emb_g.weight"].shape[0]
if cpt.get("f0", 1) == 0:
to_return_protect0 = to_return_protect1 = {"visible": False, "value": 0.5, "__type__": "update"}
else:
to_return_protect0 = {"visible": True, "value": to_return_protect0, "__type__": "update"}
to_return_protect1 = {"visible": True, "value": to_return_protect1, "__type__": "update"}
version = cpt.get("version", "v1")
net_g = (SynthesizerTrnMs256NSFsid if version == "v1" else SynthesizerTrnMs768NSFsid)(
*cpt["config"], is_half=config.is_half) if cpt.get("f0", 1) == 1 else (SynthesizerTrnMs256NSFsid_nono if version == "v1" else SynthesizerTrnMs768NSFsid_nono)(*cpt["config"])
del net_g.enc_q
print(net_g.load_state_dict(cpt["weight"], strict=False))
net_g.eval().to(config.device)
net_g = net_g.half() if config.is_half else net_g.float()
vc = VC(tgt_sr, config)
n_spk = cpt["config"][-3]
return (
{"visible": True, "maximum": n_spk, "__type__": "update"},
to_return_protect0,
to_return_protect1
)
def change_choices():
names = [name for name in os.listdir(weight_root) if name.endswith(".pth")]
indexes_list = [os.path.join(root, name) for root, _, files in os.walk(index_root, topdown=False) for name in files if name.endswith(".index") and "trained" not in name]
audio_paths = [os.path.join(audio_root, file) for file in os.listdir(os.path.join(os.getcwd(), "audios"))]
return (
{"choices": sorted(names), "__type__": "update"},
{"choices": sorted(indexes_list), "__type__": "update"},
{"choices": sorted(audio_paths), "__type__": "update"}
)
sr_dict = {
"32k": 32000,
"40k": 40000,
"48k": 48000,
}
def if_done(done, p):
while p.poll() is None:
sleep(0.5)
done[0] = True
def if_done_multi(done, ps):
while not all(p.poll() is not None for p in ps):
sleep(0.5)
done[0] = True
def formant_enabled(cbox, qfrency, tmbre, frmntapply, formantpreset, formant_refresh_button):
global DoFormant
DoFormant = cbox
CSVutil('csvdb/formanting.csv', 'w+', 'formanting', DoFormant, qfrency, tmbre)
visibility_update = {"visible": DoFormant, "__type__": "update"}
return (
{"value": DoFormant, "__type__": "update"},
) + (visibility_update,) * 6
def formant_apply(qfrency, tmbre):
global Quefrency, Timbre, DoFormant
Quefrency = qfrency
Timbre = tmbre
DoFormant = True
CSVutil('csvdb/formanting.csv', 'w+', 'formanting', DoFormant, Quefrency, Timbre)
return ({"value": Quefrency, "__type__": "update"}, {"value": Timbre, "__type__": "update"})
def update_fshift_presets(preset, qfrency, tmbre):
if preset:
with open(preset, 'r') as p:
content = p.readlines()
qfrency, tmbre = content[0].strip(), content[1]
formant_apply(qfrency, tmbre)
else:
qfrency, tmbre = preset_apply(preset, qfrency, tmbre)
return (
{"choices": get_fshift_presets(), "__type__": "update"},
{"value": qfrency, "__type__": "update"},
{"value": tmbre, "__type__": "update"},
)
def preprocess_dataset(trainset_dir, exp_dir, sr, n_p):
sr = sr_dict[sr]
log_dir = f"{now_dir}\\logs\\{exp_dir}"
log_file = f"{log_dir}\\preprocess.log"
os.makedirs(log_dir, exist_ok=True)
with open(log_file, "w") as f: pass
cmd = (
f"{config.python_cmd} "
"trainset_preprocess_pipeline_print.py "
f"{trainset_dir} "
f"{RQuote(sr)} "
f"{RQuote(n_p)} "
f"{log_dir} "
f"{RQuote(config.noparallel)}"
)
print(cmd)
p = Popen(cmd, shell=True)
done = [False]
threading.Thread(target=if_done, args=(done,p,)).start()
while not done[0]:
with open(log_file, "r") as f:
yield f.read()
sleep(1)
with open(log_file, "r") as f:
log = f.read()
print(log)
yield log
def extract_f0_feature(gpus, n_p, f0method, if_f0, exp_dir, version19, echl):
gpus = gpus.split("-")
log_dir = f"{now_dir}/logs/{exp_dir}"
log_file = f"{log_dir}/extract_f0_feature.log"
os.makedirs(log_dir, exist_ok=True)
with open(log_file, "w") as f: pass
if if_f0:
cmd = (
f"{config.python_cmd} extract_f0_print.py {log_dir} "
f"{RQuote(n_p)} {RQuote(f0method)} {RQuote(echl)}"
)
print(cmd)
p = Popen(cmd, shell=True, cwd=now_dir)
done = [False]
threading.Thread(target=if_done, args=(done, p)).start()
while not done[0]:
with open(log_file, "r") as f:
yield f.read()
sleep(1)
leng = len(gpus)
ps = []
for idx, n_g in enumerate(gpus):
cmd = (
f"{config.python_cmd} extract_feature_print.py {RQuote(config.device)} "
f"{RQuote(leng)} {RQuote(idx)} {RQuote(n_g)} {log_dir} {RQuote(version19)}"
)
print(cmd)
p = Popen(cmd, shell=True, cwd=now_dir)
ps.append(p)
done = [False]
threading.Thread(target=if_done_multi, args=(done, ps)).start()
while not done[0]:
with open(log_file, "r") as f:
yield f.read()
sleep(1)
with open(log_file, "r") as f:
log = f.read()
print(log)
yield log
def change_sr2(sr2, if_f0_3, version19):
path_str = "" if version19 == "v1" else "_v2"
f0_str = "f0" if if_f0_3 else ""
model_paths = {"G": "", "D": ""}
for model_type in model_paths:
file_path = f"pretrained{path_str}/{f0_str}{model_type}{sr2}.pth"
if os.access(file_path, os.F_OK):
model_paths[model_type] = file_path
else:
print(f"{file_path} doesn't exist, will not use pretrained model.")
return (model_paths["G"], model_paths["D"])
def change_version19(sr2, if_f0_3, version19):
path_str = "" if version19 == "v1" else "_v2"
sr2 = "40k" if (sr2 == "32k" and version19 == "v1") else sr2
choices_update = {"choices": ["40k", "48k"], "__type__": "update", "value": sr2} if version19 == "v1" else {"choices": ["40k", "48k", "32k"], "__type__": "update", "value": sr2}
f0_str = "f0" if if_f0_3 else ""
model_paths = {"G": "", "D": ""}
for model_type in model_paths:
file_path = f"pretrained{path_str}/{f0_str}{model_type}{sr2}.pth"
if os.access(file_path, os.F_OK):
model_paths[model_type] = file_path
else:
print(f"{file_path} doesn't exist, will not use pretrained model.")
return (model_paths["G"], model_paths["D"], choices_update)
def change_f0(if_f0_3, sr2, version19): # f0method8,pretrained_G14,pretrained_D15
path_str = "" if version19 == "v1" else "_v2"
pth_format = "pretrained%s/f0%s%s.pth"
model_desc = { "G": "", "D": "" }
for model_type in model_desc:
file_path = pth_format % (path_str, model_type, sr2)
if os.access(file_path, os.F_OK):
model_desc[model_type] = file_path
else:
print(file_path, "doesn't exist, will not use pretrained model")
return (
{"visible": if_f0_3, "__type__": "update"},
model_desc["G"],
model_desc["D"],
{"visible": if_f0_3, "__type__": "update"}
)
global log_interval
def set_log_interval(exp_dir, batch_size12):
log_interval = 1
folder_path = os.path.join(exp_dir, "1_16k_wavs")
if os.path.isdir(folder_path):
wav_files_num = len(glob1(folder_path,"*.wav"))
if wav_files_num > 0:
log_interval = math.ceil(wav_files_num / batch_size12)
if log_interval > 1:
log_interval += 1
return log_interval
def click_train(
exp_dir1,
sr2,
if_f0_3,
spk_id5,
save_epoch10,
total_epoch11,
batch_size12,
if_save_latest13,
pretrained_G14,
pretrained_D15,
gpus16,
if_cache_gpu17,
if_save_every_weights18,
version19,
):
CSVutil('csvdb/stop.csv', 'w+', 'formanting', False)
log_dir = f"%s\\logs\\%s" % (
now_dir, exp_dir1
)
os.makedirs(log_dir, exist_ok=True)
gt_wavs_dir = f"%s\\0_gt_wavs" % (log_dir)
feature_dim = "256" if version19 == "v1" else "768"
feature_dir = f"%s\\3_feature%s" % (
log_dir, feature_dim
)
log_interval = set_log_interval(log_dir, batch_size12)
required_dirs = [gt_wavs_dir, feature_dir]
if if_f0_3:
f0_dir = f"{log_dir}/2a_f0"
f0nsf_dir = f"{log_dir}/2b-f0nsf"
required_dirs.extend([f0_dir, f0nsf_dir])
names = set(name.split(".")[0] for directory in required_dirs for name in os.listdir(directory))
def generate_paths(name):
paths = [gt_wavs_dir, feature_dir]
if if_f0_3:
paths.extend([f0_dir, f0nsf_dir])
return '|'.join([path.replace('\\', '\\\\') + '/' + name + ('.wav.npy' if path in [f0_dir, f0nsf_dir] else '.wav' if path == gt_wavs_dir else '.npy') for path in paths])
opt = [f"{generate_paths(name)}|{spk_id5}" for name in names]
mute_dir = f"{now_dir}/logs/mute"
for _ in range(2):
mute_string = f"{mute_dir}/0_gt_wavs/mute{sr2}.wav|{mute_dir}/3_feature{feature_dim}/mute.npy"
if if_f0_3:
mute_string += f"|{mute_dir}/2a_f0/mute.wav.npy|{mute_dir}/2b-f0nsf/mute.wav.npy"
opt.append(mute_string+f"|{spk_id5}")
shuffle(opt)
with open(f"{log_dir}/filelist.txt", "w") as f:
f.write("\n".join(opt))
print("write filelist done")
print("use gpus:", gpus16)
if pretrained_G14 == "":
print("no pretrained Generator")
if pretrained_D15 == "":
print("no pretrained Discriminator")
G_train = f"-pg {pretrained_G14}" if pretrained_G14 else ""
D_train = f"-pd {pretrained_D15}" if pretrained_D15 else ""
save_latest = 1 if if_save_latest13 else 0
cache_gpu = 1 if if_cache_gpu17 else 0
save_every_weights = 1 if if_save_every_weights18 else 0
cmd = (
f"{config.python_cmd} train_nsf_sim_cache_sid_load_pretrain.py -e {exp_dir1} -sr {sr2} -f0 {int(if_f0_3)} -bs {batch_size12}"
f" -g {gpus16 if gpus16 is not None else ''} -te {total_epoch11} -se {save_epoch10} {G_train} {D_train} -l {save_latest}"
f" -c {cache_gpu} -sw {save_every_weights} -v {version19} -li {log_interval}"
)
print(cmd)
global p
p = Popen(cmd, shell=True, cwd=now_dir)
global PID
PID = p.pid
p.wait()
return "Training is done, check train.log", {"visible": False, "__type__": "update"}, {"visible": True, "__type__": "update"}
def train_index(exp_dir1, version19):
exp_dir = os.path.join(now_dir, 'logs', exp_dir1)
os.makedirs(exp_dir, exist_ok=True)
feature_dim = '256' if version19 == "v1" else '768'
feature_dir = os.path.join(exp_dir, f"3_feature{feature_dim}")
if not os.path.exists(feature_dir) or len(os.listdir(feature_dir)) == 0:
return "请先进行特征提取!"
npys = [np.load(os.path.join(feature_dir, name)) for name in sorted(os.listdir(feature_dir))]
big_npy = np.concatenate(npys, 0)
np.random.shuffle(big_npy)
infos = []
if big_npy.shape[0] > 2*10**5:
infos.append("Trying doing kmeans %s shape to 10k centers." % big_npy.shape[0])
yield "\n".join(infos)
try:
big_npy = MiniBatchKMeans(n_clusters=10000, verbose=True, batch_size=256 * config.n_cpu,
compute_labels=False,init="random").fit(big_npy).cluster_centers_
except Exception as e:
infos.append(str(e))
yield "\n".join(infos)
np.save(os.path.join(exp_dir, "total_fea.npy"), big_npy)
n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])), big_npy.shape[0] // 39)
infos.append("%s,%s" % (big_npy.shape, n_ivf))
yield "\n".join(infos)
index = faiss.index_factory(int(feature_dim), f"IVF{n_ivf},Flat")
index_ivf = faiss.extract_index_ivf(index)
index_ivf.nprobe = 1
index.train(big_npy)
index_file_base = f"{exp_dir}/trained_IVF{n_ivf}_Flat_nprobe_{index_ivf.nprobe}_{exp_dir1}_{version19}.index"
faiss.write_index(index, index_file_base)
infos.append("adding")
yield "\n".join(infos)
batch_size_add = 8192
for i in range(0, big_npy.shape[0], batch_size_add):
index.add(big_npy[i:i + batch_size_add])
index_file_base = f"{exp_dir}/added_IVF{n_ivf}_Flat_nprobe_{index_ivf.nprobe}_{exp_dir1}_{version19}.index"
faiss.write_index(index, index_file_base)
infos.append(f"Successful Index Construction，added_IVF{n_ivf}_Flat_nprobe_{index_ivf.nprobe}_{exp_dir1}_{version19}.index")
yield "\n".join(infos)
#def setBoolean(status): #true to false and vice versa / not implemented yet, dont touch!!!!!!!
# status = not status
# return status
def change_info_(ckpt_path):
train_log_path = os.path.join(os.path.dirname(ckpt_path), "train.log")
if not os.path.exists(train_log_path):
return {"__type__": "update"}, {"__type__": "update"}, {"__type__": "update"}
try:
with open(train_log_path, "r") as f:
info_line = next(f).strip()
info = eval(info_line.split("\t")[-1])
sr, f0 = info.get("sample_rate"), info.get("if_f0")
version = "v2" if info.get("version") == "v2" else "v1"
return sr, str(f0), version
except Exception as e:
print(f"Exception occurred: {str(e)}, Traceback: {traceback.format_exc()}")
return {"__type__": "update"}, {"__type__": "update"}, {"__type__": "update"}
def export_onnx(model_path, exported_path):
device = torch.device("cpu")
checkpoint = torch.load(model_path, map_location=device)
vec_channels = 256 if checkpoint.get("version", "v1") == "v1" else 768
test_inputs = {
"phone": torch.rand(1, 200, vec_channels),
"phone_lengths": torch.LongTensor([200]),
"pitch": torch.randint(5, 255, (1, 200)),
"pitchf": torch.rand(1, 200),
"ds": torch.zeros(1).long(),
"rnd": torch.rand(1, 192, 200)
}
checkpoint["config"][-3] = checkpoint["weight"]["emb_g.weight"].shape[0]
net_g = SynthesizerTrnMsNSFsidM(*checkpoint["config"], is_half=False, version=checkpoint.get("version", "v1"))
net_g.load_state_dict(checkpoint["weight"], strict=False)
net_g = net_g.to(device)
dynamic_axes = {"phone": [1], "pitch": [1], "pitchf": [1], "rnd": [2]}
torch.onnx.export(
net_g,
tuple(value.to(device) for value in test_inputs.values()),
exported_path,
dynamic_axes=dynamic_axes,
do_constant_folding=False,
opset_version=13,
verbose=False,
input_names=list(test_inputs.keys()),
output_names=["audio"],
)
return "Finished"
#region Mangio-RVC-Fork CLI App
import scipy.io.wavfile as wavfile
cli_current_page = "HOME"
def cli_split_command(com):
exp = r'(?:(?<=\s)|^)"(.*?)"(?=\s|$)|(\S+)'
split_array = re.findall(exp, com)
split_array = [group[0] if group[0] else group[1] for group in split_array]
return split_array
execute_generator_function = lambda genObject: all(x is not None for x in genObject)
def cli_infer(com):
model_name, source_audio_path, output_file_name, feature_index_path, speaker_id, transposition, f0_method, crepe_hop_length, harvest_median_filter, resample, mix, feature_ratio, protection_amnt, _, do_formant = cli_split_command(com)[:15]
speaker_id, crepe_hop_length, harvest_median_filter, resample = map(int, [speaker_id, crepe_hop_length, harvest_median_filter, resample])
transposition, mix, feature_ratio, protection_amnt = map(float, [transposition, mix, feature_ratio, protection_amnt])
if do_formant.lower() == 'false':
Quefrency = 1.0
Timbre = 1.0
else:
Quefrency, Timbre = map(float, cli_split_command(com)[15:17])
CSVutil('csvdb/formanting.csv', 'w+', 'formanting', do_formant.lower() == 'true', Quefrency, Timbre)
output_message = 'Mangio-RVC-Fork Infer-CLI:'
output_path = f'audio-outputs/{output_file_name}'
print(f"{output_message} Starting the inference...")
vc_data = get_vc(model_name, protection_amnt, protect1)
print(vc_data)
print(f"{output_message} Performing inference...")
conversion_data = vc_single(
speaker_id,
source_audio_path,
source_audio_path,
transposition,
f0_file,
f0_method,
feature_index_path,
feature_index_path,
feature_ratio,
harvest_median_filter,
resample,
mix,
protection_amnt,
crepe_hop_length
)
if "Success." in conversion_data[0]:
print(f"{output_message} Inference succeeded. Writing to {output_path}...")
wavfile.write(output_path, conversion_data[1][0], conversion_data[1][1])
print(f"{output_message} Finished! Saved output to {output_path}")
else:
print(f"{output_message} Inference failed. Here's the traceback: {conversion_data[0]}")
def cli_pre_process(com):
print("Mangio-RVC-Fork Pre-process: Starting...")
execute_generator_function(
preprocess_dataset(
*cli_split_command(com)[:3],
int(cli_split_command(com)[3])
)
)
print("Mangio-RVC-Fork Pre-process: Finished")
def cli_extract_feature(com):
model_name, gpus, num_processes, has_pitch_guidance, f0_method, crepe_hop_length, version = cli_split_command(com)
num_processes = int(num_processes)
has_pitch_guidance = bool(int(has_pitch_guidance))
crepe_hop_length = int(crepe_hop_length)
print(
f"Mangio-RVC-CLI: Extract Feature Has Pitch: {has_pitch_guidance}"
f"Mangio-RVC-CLI: Extract Feature Version: {version}"
"Mangio-RVC-Fork Feature Extraction: Starting..."
)
generator = extract_f0_feature(
gpus,
num_processes,
f0_method,
has_pitch_guidance,
model_name,
version,
crepe_hop_length
)
execute_generator_function(generator)
print("Mangio-RVC-Fork Feature Extraction: Finished")
def cli_train(com):
com = cli_split_command(com)
model_name = com[0]
sample_rate = com[1]
bool_flags = [bool(int(i)) for i in com[2:11]]
version = com[11]
pretrained_base = "pretrained/" if version == "v1" else "pretrained_v2/"
g_pretrained_path = f"{pretrained_base}f0G{sample_rate}.pth"
d_pretrained_path = f"{pretrained_base}f0D{sample_rate}.pth"
print("Mangio-RVC-Fork Train-CLI: Training...")
click_train(model_name, sample_rate, *bool_flags, g_pretrained_path, d_pretrained_path, version)
def cli_train_feature(com):
output_message = 'Mangio-RVC-Fork Train Feature Index-CLI'
print(f"{output_message}: Training... Please wait")
execute_generator_function(train_index(*cli_split_command(com)))
print(f"{output_message}: Done!")
def cli_extract_model(com):
extract_small_model_process = extract_small_model(*cli_split_command(com))
print(
"Mangio-RVC-Fork Extract Small Model: Success!"
if extract_small_model_process == "Success."
else f"{extract_small_model_process}\nMangio-RVC-Fork Extract Small Model: Failed!"
)
def preset_apply(preset, qfer, tmbr):
if preset:
try:
with open(preset, 'r') as p:
content = p.read().splitlines()
qfer, tmbr = content[0], content[1]
formant_apply(qfer, tmbr)
except IndexError:
print("Error: File does not have enough lines to read 'qfer' and 'tmbr'")
except FileNotFoundError:
print("Error: File does not exist")
except Exception as e:
print("An unexpected error occurred", e)
return ({"value": qfer, "__type__": "update"}, {"value": tmbr, "__type__": "update"})
def print_page_details():
page_description = {
'HOME':
"\n go home : Takes you back to home with a navigation list."
"\n go infer : Takes you to inference command execution."
"\n go pre-process : Takes you to training step.1) pre-process command execution."
"\n go extract-feature : Takes you to training step.2) extract-feature command execution."
"\n go train : Takes you to training step.3) being or continue training command execution."
"\n go train-feature : Takes you to the train feature index command execution."
"\n go extract-model : Takes you to the extract small model command execution."
, 'INFER':
"\n arg 1) model name with .pth in ./weights: mi-test.pth"
"\n arg 2) source audio path: myFolder\\MySource.wav"
"\n arg 3) output file name to be placed in './audio-outputs': MyTest.wav"
"\n arg 4) feature index file path: logs/mi-test/added_IVF3042_Flat_nprobe_1.index"
"\n arg 5) speaker id: 0"
"\n arg 6) transposition: 0"
"\n arg 7) f0 method: harvest (pm, harvest, crepe, crepe-tiny, hybrid[x,x,x,x], mangio-crepe, mangio-crepe-tiny, rmvpe)"
"\n arg 8) crepe hop length: 160"
"\n arg 9) harvest median filter radius: 3 (0-7)"
"\n arg 10) post resample rate: 0"
"\n arg 11) mix volume envelope: 1"
"\n arg 12) feature index ratio: 0.78 (0-1)"
"\n arg 13) Voiceless Consonant Protection (Less Artifact): 0.33 (Smaller number = more protection. 0.50 means Dont Use.)"
"\n arg 14) Whether to formant shift the inference audio before conversion: False (if set to false, you can ignore setting the quefrency and timbre values for formanting)"
"\n arg 15)* Quefrency for formanting: 8.0 (no need to set if arg14 is False/false)"
"\n arg 16)* Timbre for formanting: 1.2 (no need to set if arg14 is False/false) \n"
"\nExample: mi-test.pth saudio/Sidney.wav myTest.wav logs/mi-test/added_index.index 0 -2 harvest 160 3 0 1 0.95 0.33 0.45 True 8.0 1.2"
, 'PRE-PROCESS':
"\n arg 1) Model folder name in ./logs: mi-test"
"\n arg 2) Trainset directory: mydataset (or) E:\\my-data-set"
"\n arg 3) Sample rate: 40k (32k, 40k, 48k)"
"\n arg 4) Number of CPU threads to use: 8 \n"
"\nExample: mi-test mydataset 40k 24"
, 'EXTRACT-FEATURE':
"\n arg 1) Model folder name in ./logs: mi-test"
"\n arg 2) Gpu card slot: 0 (0-1-2 if using 3 GPUs)"
"\n arg 3) Number of CPU threads to use: 8"
"\n arg 4) Has Pitch Guidance?: 1 (0 for no, 1 for yes)"
"\n arg 5) f0 Method: harvest (pm, harvest, dio, crepe)"
"\n arg 6) Crepe hop length: 128"
"\n arg 7) Version for pre-trained models: v2 (use either v1 or v2)\n"
"\nExample: mi-test 0 24 1 harvest 128 v2"
, 'TRAIN':
"\n arg 1) Model folder name in ./logs: mi-test"
"\n arg 2) Sample rate: 40k (32k, 40k, 48k)"
"\n arg 3) Has Pitch Guidance?: 1 (0 for no, 1 for yes)"
"\n arg 4) speaker id: 0"
"\n arg 5) Save epoch iteration: 50"
"\n arg 6) Total epochs: 10000"
"\n arg 7) Batch size: 8"
"\n arg 8) Gpu card slot: 0 (0-1-2 if using 3 GPUs)"
"\n arg 9) Save only the latest checkpoint: 0 (0 for no, 1 for yes)"
"\n arg 10) Whether to cache training set to vram: 0 (0 for no, 1 for yes)"
"\n arg 11) Save extracted small model every generation?: 0 (0 for no, 1 for yes)"
"\n arg 12) Model architecture version: v2 (use either v1 or v2)\n"
"\nExample: mi-test 40k 1 0 50 10000 8 0 0 0 0 v2"
, 'TRAIN-FEATURE':
"\n arg 1) Model folder name in ./logs: mi-test"
"\n arg 2) Model architecture version: v2 (use either v1 or v2)\n"
"\nExample: mi-test v2"
, 'EXTRACT-MODEL':
"\n arg 1) Model Path: logs/mi-test/G_168000.pth"
"\n arg 2) Model save name: MyModel"
"\n arg 3) Sample rate: 40k (32k, 40k, 48k)"
"\n arg 4) Has Pitch Guidance?: 1 (0 for no, 1 for yes)"
'\n arg 5) Model information: "My Model"'
"\n arg 6) Model architecture version: v2 (use either v1 or v2)\n"
'\nExample: logs/mi-test/G_168000.pth MyModel 40k 1 "Created by Cole Mangio" v2'
}
print(page_description.get(cli_current_page, 'Invalid page'))
def change_page(page):
global cli_current_page
cli_current_page = page
return 0
def execute_command(com):
command_to_page = {
"go home": "HOME",
"go infer": "INFER",
"go pre-process": "PRE-PROCESS",
"go extract-feature": "EXTRACT-FEATURE",
"go train": "TRAIN",
"go train-feature": "TRAIN-FEATURE",
"go extract-model": "EXTRACT-MODEL",
}
page_to_function = {
"INFER": cli_infer,
"PRE-PROCESS": cli_pre_process,
"EXTRACT-FEATURE": cli_extract_feature,
"TRAIN": cli_train,
"TRAIN-FEATURE": cli_train_feature,
"EXTRACT-MODEL": cli_extract_model,
}
if com in command_to_page:
return change_page(command_to_page[com])
if com[:3] == "go ":
print(f"page '{com[3:]}' does not exist!")
return 0
if cli_current_page in page_to_function:
page_to_function[cli_current_page](com)
def cli_navigation_loop():
while True:
print(f"\nYou are currently in '{cli_current_page}':")
print_page_details()
print(cli_current_page + ": ", end="")
try:
execute_command(input())
except Exception as e:
print(f"An error occurred: {traceback.format_exc()}")
if(config.is_cli):
print(
"\n\nMangio-RVC-Fork v2 CLI App!\n"
"Welcome to the CLI version of RVC. Please read the documentation on https://github.com/Mangio621/Mangio-RVC-Fork (README.MD) to understand how to use this app.\n"
)
cli_navigation_loop()
#endregion
#region RVC WebUI App
'''
def get_presets():
data = None
with open('../inference-presets.json', 'r') as file:
data = json.load(file)
preset_names = []
for preset in data['presets']:
preset_names.append(preset['name'])
return preset_names
'''
def match_index(sid0):
folder = sid0.split('.')[0].split('_')[0]
parent_dir = "./logs/" + folder
if not os.path.exists(parent_dir):
return '', ''
for filename in os.listdir(parent_dir):
if filename.endswith(".index"):
index_path = os.path.join(parent_dir, filename).replace('\\','/')
print(index_path)
if index_path in indexes_list:
return index_path, index_path
lowered_index_path = os.path.join(parent_dir.lower(), filename).replace('\\','/')
if lowered_index_path in indexes_list:
return lowered_index_path, lowered_index_path
return '', ''
def stoptraining(mim):
if mim:
try:
CSVutil('csvdb/stop.csv', 'w+', 'stop', 'True')
os.kill(PID, SIGTERM)
except Exception as e:
print(f"Couldn't click due to {e}")
return (
{"visible": False, "__type__": "update"},
{"visible": True, "__type__": "update"},
)
tab_faq = i18n("常见问题解答")
faq_file = "docs/faq.md" if tab_faq == "常见问题解答" else "docs/faq_en.md"
weights_dir = 'weights/'
#Change your Gradio Theme here. 👇 👇 👇 👇 Example: " theme='HaleyCH/HaleyCH_Theme' "
with gr.Blocks(theme=gr.themes.Soft(), title='Mangio-RVC-Web 💻') as app:
gr.HTML(" The Mangio-RVC-Fork 💻
")
gr.Markdown(
value=i18n(
"本软件以MIT协议开源, 作者不对软件具备任何控制力, 使用软件者、传播软件导出的声音者自负全责.
如不认可该条款, 则不能使用或引用软件包内任何代码和文件. 详见根目录使用需遵守的协议-LICENSE.txt."
)
)
with gr.Tabs():
with gr.TabItem(i18n("模型推理")):
# Inference Preset Row
# with gr.Row():
# mangio_preset = gr.Dropdown(label="Inference Preset", choices=sorted(get_presets()))
# mangio_preset_name_save = gr.Textbox(
# label="Your preset name"
# )
# mangio_preset_save_btn = gr.Button('Save Preset', variant="primary")
# Other RVC stuff
with gr.Row():
sid0 = gr.Dropdown(label=i18n("推理音色"), choices=sorted(names), value='')
refresh_button = gr.Button(i18n("Refresh voice list, index path and audio files"), variant="primary")
clean_button = gr.Button(i18n("卸载音色省显存"), variant="primary")
spk_item = gr.Slider(
minimum=0,
maximum=2333,
step=1,
label=i18n("请选择说话人id"),
value=0,
visible=False,
interactive=True,
)
clean_button.click(fn=lambda: ({"value": "", "__type__": "update"}), inputs=[], outputs=[sid0])
with gr.Group():
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. ")
)
with gr.Row():
with gr.Column():
vc_transform0 = gr.Number(
label=i18n("变调(整数, 半音数量, 升八度12降八度-12)"), value=0
)
input_audio0 = gr.Textbox(
label=i18n("Add audio's name to the path to the audio file to be processed (default is the correct format example) Remove the path to use an audio from the dropdown list:"),
value=os.path.abspath(os.getcwd()).replace('\\', '/') + "/audios/" + "audio.wav",
)
input_audio1 = gr.Dropdown(
label=i18n("Auto detect audio path and select from the dropdown:"),
choices=sorted(audio_paths),
value='',
interactive=True,
)
input_audio1.change(fn=lambda:'',inputs=[],outputs=[input_audio0])
f0method0 = gr.Radio(
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"
),
choices=["pm", "harvest", "dio", "crepe", "crepe-tiny", "mangio-crepe", "mangio-crepe-tiny", "rmvpe"], # Fork Feature. Add Crepe-Tiny
value="rmvpe",
interactive=True,
)
crepe_hop_length = gr.Slider(
minimum=1,
maximum=512,
step=1,
label=i18n("crepe_hop_length"),
value=120,
interactive=True,
visible=False,
)
f0method0.change(
fn=lambda radio: (
{
"visible": radio in ['mangio-crepe', 'mangio-crepe-tiny'],
"__type__": "update"
}
),
inputs=[f0method0],
outputs=[crepe_hop_length]
)
filter_radius0 = gr.Slider(
minimum=0,
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"),
value=3,
step=1,
interactive=True,
)
with gr.Column():
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"),
value="",
interactive=True,
)
file_index2 = gr.Dropdown(
label="3. Path to your added.index file (if it didn't automatically find it.)",
choices=get_indexes(),
interactive=True,
allow_custom_value=True,
)
#sid0.select(fn=match_index, inputs=sid0, outputs=file_index2)
refresh_button.click(
fn=change_choices, inputs=[], outputs=[sid0, file_index2, input_audio1]
)
# file_big_npy1 = gr.Textbox(
# label=i18n("特征文件路径"),
# value="E:\\codes\py39\\vits_vc_gpu_train\\logs\\mi-test-1key\\total_fea.npy",
# interactive=True,
# )
index_rate1 = gr.Slider(
minimum=0,
maximum=1,
label=i18n("检索特征占比"),
value=0.75,
interactive=True,
)
with gr.Column():
resample_sr0 = gr.Slider(
minimum=0,
maximum=48000,
label=i18n("后处理重采样至最终采样率，0为不进行重采样"),
value=0,
step=1,
interactive=True,
)
rms_mix_rate0 = gr.Slider(
minimum=0,
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"),
value=0.25,
interactive=True,
)
protect0 = gr.Slider(
minimum=0,
maximum=0.5,
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"
),
value=0.33,
step=0.01,
interactive=True,
)
formanting = gr.Checkbox(
value=bool(DoFormant),
label="[EXPERIMENTAL] Formant shift inference audio",
info="Used for male to female and vice-versa conversions",
interactive=True,
visible=True,
)
formant_preset = gr.Dropdown(
value='',
choices=get_fshift_presets(),
label='Browse presets for formanting',
info='Presets are located in formantshiftcfg/ folder',
visible=bool(DoFormant),
)
formant_refresh_button = gr.Button(
value='\U0001f504',
visible=bool(DoFormant),
variant='primary',
)
qfrency = gr.Slider(
value=Quefrency,
info="Default value is 1.0",
label="Quefrency for formant shifting",
minimum=0.0,
maximum=16.0,
step=0.1,
visible=bool(DoFormant),
interactive=True,
)
tmbre = gr.Slider(
value=Timbre,
info="Default value is 1.0",
label="Timbre for formant shifting",
minimum=0.0,
maximum=16.0,
step=0.1,
visible=bool(DoFormant),
interactive=True,
)
formant_preset.change(fn=preset_apply, inputs=[formant_preset, qfrency, tmbre], outputs=[qfrency, tmbre])
frmntbut = gr.Button("Apply", variant="primary", visible=bool(DoFormant))
formanting.change(fn=formant_enabled,inputs=[formanting,qfrency,tmbre,frmntbut,formant_preset,formant_refresh_button],outputs=[formanting,qfrency,tmbre,frmntbut,formant_preset,formant_refresh_button])
frmntbut.click(fn=formant_apply,inputs=[qfrency, tmbre], outputs=[qfrency, tmbre])
formant_refresh_button.click(fn=update_fshift_presets,inputs=[formant_preset, qfrency, tmbre],outputs=[formant_preset, qfrency, tmbre])
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")
with gr.Row():
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)"))
but0.click(
vc_single,
[
spk_item,
input_audio0,
input_audio1,
vc_transform0,
f0_file,
f0method0,
file_index1,
file_index2,
# file_big_npy1,
index_rate1,
filter_radius0,
resample_sr0,
rms_mix_rate0,
protect0,
crepe_hop_length
],
[vc_output1, vc_output2],
)
with gr.Group():
gr.Markdown(
value=i18n("批量转换, 输入待转换音频文件夹, 或上传多个音频文件, 在指定文件夹(默认opt)下输出转换的音频. ")
)
with gr.Row():
with gr.Column():
vc_transform1 = gr.Number(
label=i18n("变调(整数, 半音数量, 升八度12降八度-12)"), value=0
)
opt_input = gr.Textbox(label=i18n("指定输出文件夹"), value="opt")
f0method1 = gr.Radio(
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"
),
choices=["pm", "harvest", "crepe", "rmvpe"],
value="rmvpe",
interactive=True,
)
filter_radius1 = gr.Slider(
minimum=0,
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"),
value=3,
step=1,
interactive=True,
)
with gr.Column():
file_index3 = 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"),
value="",
interactive=True,
)
file_index4 = gr.Dropdown( #file index dropdown for batch
label=i18n("自动检测index路径,下拉式选择(dropdown)"),
choices=get_indexes(),
interactive=True,
)
sid0.select(fn=match_index, inputs=[sid0], outputs=[file_index2, file_index4])
refresh_button.click(
fn=lambda: change_choices()[1],
inputs=[],
outputs=file_index4,
)
# file_big_npy2 = gr.Textbox(
# label=i18n("特征文件路径"),
# value="E:\\codes\\py39\\vits_vc_gpu_train\\logs\\mi-test-1key\\total_fea.npy",
# interactive=True,
# )
index_rate2 = gr.Slider(
minimum=0,
maximum=1,
label=i18n("检索特征占比"),
value=1,
interactive=True,
)
with gr.Column():
resample_sr1 = gr.Slider(
minimum=0,
maximum=48000,
label=i18n("后处理重采样至最终采样率，0为不进行重采样"),
value=0,
step=1,
interactive=True,
)
rms_mix_rate1 = gr.Slider(
minimum=0,
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"),
value=1,
interactive=True,
)
protect1 = gr.Slider(
minimum=0,
maximum=0.5,
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"
),
value=0.33,
step=0.01,
interactive=True,
)
with gr.Column():
dir_input = gr.Textbox(
label=i18n("输入待处理音频文件夹路径(去文件管理器地址栏拷就行了)"),
value=os.path.abspath(os.getcwd()).replace('\\', '/') + "/audios/",
)
inputs = gr.File(
file_count="multiple", label=i18n("也可批量输入音频文件, 二选一, 优先读文件夹")
)
with gr.Row():
format1 = gr.Radio(
label=i18n("导出文件格式"),
choices=["wav", "flac", "mp3", "m4a"],
value="flac",
interactive=True,
)
but1 = gr.Button(i18n("转换"), variant="primary")
vc_output3 = gr.Textbox(label=i18n("输出信息"))
but1.click(
vc_multi,
[
spk_item,
dir_input,
opt_input,
inputs,
vc_transform1,
f0method1,
file_index3,
file_index4,
# file_big_npy2,
index_rate2,
filter_radius1,
resample_sr1,
rms_mix_rate1,
protect1,
format1,
crepe_hop_length,
],
[vc_output3],
)
sid0.change(
fn=get_vc,
inputs=[sid0, protect0, protect1],
outputs=[spk_item, protect0, protect1],
)
with gr.TabItem(i18n("伴奏人声分离&去混响&去回声")):
with gr.Group():
gr.Markdown(
value=i18n(
"人声伴奏分离批量处理， 使用UVR5模型。
"
"合格的文件夹路径格式举例： E:\\codes\\py39\\vits_vc_gpu\\白鹭霜华测试样例(去文件管理器地址栏拷就行了)。
"
"模型分为三类：
"
"1、保留人声：不带和声的音频选这个，对主人声保留比HP5更好。内置HP2和HP3两个模型，HP3可能轻微漏伴奏但对主人声保留比HP2稍微好一丁点；
"
"2、仅保留主人声：带和声的音频选这个，对主人声可能有削弱。内置HP5一个模型；
"
"3、去混响、去延迟模型（by FoxJoy）：
"
" (1)MDX-Net(onnx_dereverb):对于双通道混响是最好的选择，不能去除单通道混响；
"
" (234)DeEcho:去除延迟效果。Aggressive比Normal去除得更彻底，DeReverb额外去除混响，可去除单声道混响，但是对高频重的板式混响去不干净。
"
"去混响/去延迟，附：
"
"1、DeEcho-DeReverb模型的耗时是另外2个DeEcho模型的接近2倍；
"
"2、MDX-Net-Dereverb模型挺慢的；
"
"3、个人推荐的最干净的配置是先MDX-Net再DeEcho-Aggressive。"
)
)
with gr.Row():
with gr.Column():
dir_wav_input = gr.Textbox(
label=i18n("输入待处理音频文件夹路径"),
value=((os.getcwd()).replace('\\', '/') + "/audios/")
)
wav_inputs = gr.File(
file_count="multiple", label=i18n("也可批量输入音频文件, 二选一, 优先读文件夹")
) #####
with gr.Column():
model_choose = gr.Dropdown(label=i18n("模型"), choices=uvr5_names)
agg = gr.Slider(
minimum=0,
maximum=20,
step=1,
label="人声提取激进程度",
value=10,
interactive=True,
visible=False, # 先不开放调整
)
opt_vocal_root = gr.Textbox(
label=i18n("指定输出主人声文件夹"), value="opt"
)
opt_ins_root = gr.Textbox(
label=i18n("指定输出非主人声文件夹"), value="opt"
)
format0 = gr.Radio(
label=i18n("导出文件格式"),
choices=["wav", "flac", "mp3", "m4a"],
value="flac",
interactive=True,
)
but2 = gr.Button(i18n("转换"), variant="primary")
vc_output4 = gr.Textbox(label=i18n("输出信息"))
but2.click(
uvr,
[
model_choose,
dir_wav_input,
opt_vocal_root,
wav_inputs,
opt_ins_root,
agg,
format0,
],
[vc_output4],
)
with gr.TabItem(i18n("训练")):
gr.Markdown(
value=i18n(
"step1: 填写实验配置. 实验数据放在logs下, 每个实验一个文件夹, 需手工输入实验名路径, 内含实验配置, 日志, 训练得到的模型文件. "
)
)
with gr.Row():
exp_dir1 = gr.Textbox(label=i18n("输入实验名"), value="mi-test")
sr2 = gr.Radio(
label=i18n("目标采样率"),
choices=["40k", "48k", "32k"],
value="40k",
interactive=True,
)
if_f0_3 = gr.Checkbox(
label="Whether the model has pitch guidance.",
value=True,
interactive=True,
)
version19 = gr.Radio(
label=i18n("版本"),
choices=["v1", "v2"],
value="v2",
interactive=True,
visible=True,
)
np7 = gr.Slider(
minimum=0,
maximum=config.n_cpu,
step=1,
label=i18n("提取音高和处理数据使用的CPU进程数"),
value=int(np.ceil(config.n_cpu / 1.5)),
interactive=True,
)
with gr.Group(): # 暂时单人的, 后面支持最多4人的#数据处理
gr.Markdown(
value=i18n(
"step2a: 自动遍历训练文件夹下所有可解码成音频的文件并进行切片归一化, 在实验目录下生成2个wav文件夹; 暂时只支持单人训练. "
)
)
with gr.Row():
trainset_dir4 = gr.Textbox(
label=i18n("输入训练文件夹路径"), value=os.path.abspath(os.getcwd()) + "\\datasets\\"
)
spk_id5 = gr.Slider(
minimum=0,
maximum=4,
step=1,
label=i18n("请指定说话人id"),
value=0,
interactive=True,
)
but1 = gr.Button(i18n("处理数据"), variant="primary")
info1 = gr.Textbox(label=i18n("输出信息"), value="")
but1.click(
preprocess_dataset, [trainset_dir4, exp_dir1, sr2, np7], [info1]
)
with gr.Group():
step2b = gr.Markdown(value=i18n("step2b: 使用CPU提取音高(如果模型带音高), 使用GPU提取特征(选择卡号)"))
with gr.Row():
with gr.Column():
gpus6 = gr.Textbox(
label=i18n("以-分隔输入使用的卡号, 例如 0-1-2 使用卡0和卡1和卡2"),
value=gpus,
interactive=True,
)
gpu_info9 = gr.Textbox(label=i18n("显卡信息"), value=gpu_info)
with gr.Column():
f0method8 = gr.Radio(
label=i18n(
"选择音高提取算法:输入歌声可用pm提速,高质量语音但CPU差可用dio提速,harvest质量更好但慢"
),
choices=["pm", "harvest", "dio", "crepe", "mangio-crepe", "rmvpe"], # Fork feature: Crepe on f0 extraction for training.
value="rmvpe",
interactive=True,
)
extraction_crepe_hop_length = gr.Slider(
minimum=1,
maximum=512,
step=1,
label=i18n("crepe_hop_length"),
value=64,
interactive=True,
visible=False,
)
f0method8.change(
fn=lambda radio: (
{
"visible": radio in ['mangio-crepe', 'mangio-crepe-tiny'],
"__type__": "update"
}
),
inputs=[f0method8],
outputs=[extraction_crepe_hop_length]
)
but2 = gr.Button(i18n("特征提取"), variant="primary")
info2 = gr.Textbox(label=i18n("输出信息"), value="", max_lines=8, interactive=False)
but2.click(
extract_f0_feature,
[gpus6, np7, f0method8, if_f0_3, exp_dir1, version19, extraction_crepe_hop_length],
[info2],
)
with gr.Group():
gr.Markdown(value=i18n("step3: 填写训练设置, 开始训练模型和索引"))
with gr.Row():
save_epoch10 = gr.Slider(
minimum=1,
maximum=50,
step=1,
label=i18n("保存频率save_every_epoch"),
value=5,
interactive=True,
visible=True,
)
total_epoch11 = gr.Slider(
minimum=1,
maximum=10000,
step=1,
label=i18n("总训练轮数total_epoch"),
value=20,
interactive=True,
)
batch_size12 = gr.Slider(
minimum=1,
maximum=40,
step=1,
label=i18n("每张显卡的batch_size"),
value=default_batch_size,
interactive=True,
)
if_save_latest13 = gr.Checkbox(
label="Whether to save only the latest .ckpt file to save hard drive space",
value=True,
interactive=True,
)
if_cache_gpu17 = gr.Checkbox(
label="Cache all training sets to GPU memory. Caching small datasets (less than 10 minutes) can speed up training, but caching large datasets will consume a lot of GPU memory and may not provide much speed improvement",
value=False,
interactive=True,
)
if_save_every_weights18 = gr.Checkbox(
label="Save a small final model to the 'weights' folder at each save point",
value=True,
interactive=True,
)
with gr.Row():
pretrained_G14 = gr.Textbox(
lines=2,
label=i18n("加载预训练底模G路径"),
value="pretrained/f0G40k.pth",
interactive=True,
)
pretrained_D15 = gr.Textbox(
lines=2,
label=i18n("加载预训练底模D路径"),
value="pretrained/f0D40k.pth",
interactive=True,
)
sr2.change(
change_sr2,
[sr2, if_f0_3, version19],
[pretrained_G14, pretrained_D15],
)
version19.change(
change_version19,
[sr2, if_f0_3, version19],
[pretrained_G14, pretrained_D15, sr2],
)
if_f0_3.change(
fn=change_f0,
inputs=[if_f0_3, sr2, version19],
outputs=[f0method8, pretrained_G14, pretrained_D15],
)
if_f0_3.change(fn=lambda radio: (
{
"visible": radio in ['mangio-crepe', 'mangio-crepe-tiny'],
"__type__": "update"
}
), inputs=[f0method8], outputs=[extraction_crepe_hop_length])
gpus16 = gr.Textbox(
label=i18n("以-分隔输入使用的卡号, 例如 0-1-2 使用卡0和卡1和卡2"),
value=gpus,
interactive=True,
)
butstop = gr.Button(
"Stop Training",
variant='primary',
visible=False,
)
but3 = gr.Button(i18n("训练模型"), variant="primary", visible=True)
but3.click(fn=stoptraining, inputs=[gr.Number(value=0, visible=False)], outputs=[but3, butstop])
butstop.click(fn=stoptraining, inputs=[gr.Number(value=1, visible=False)], outputs=[butstop, but3])
but4 = gr.Button(i18n("训练特征索引"), variant="primary")
#but5 = gr.Button(i18n("一键训练"), variant="primary")
info3 = gr.Textbox(label=i18n("输出信息"), value="", max_lines=10)
if_save_every_weights18.change(
fn=lambda if_save_every_weights: (
{
"visible": if_save_every_weights,
"__type__": "update"
}
),
inputs=[if_save_every_weights18],
outputs=[save_epoch10]
)
but3.click(
click_train,
[
exp_dir1,
sr2,
if_f0_3,
spk_id5,
save_epoch10,
total_epoch11,
batch_size12,
if_save_latest13,
pretrained_G14,
pretrained_D15,
gpus16,
if_cache_gpu17,
if_save_every_weights18,
version19,
],
[info3, butstop, but3],
)
but4.click(train_index, [exp_dir1, version19], info3)
#but5.click(
# train1key,
# [
# exp_dir1,
# sr2,
# if_f0_3,
# trainset_dir4,
# spk_id5,
# np7,
# f0method8,
# save_epoch10,
# total_epoch11,
# batch_size12,
# if_save_latest13,
# pretrained_G14,
# pretrained_D15,
# gpus16,
# if_cache_gpu17,
# if_save_every_weights18,
# version19,
# extraction_crepe_hop_length
# ],
# info3,
#)
with gr.Group():
gr.Markdown(value=
'Step 4: Export lowest points on a graph of the model '
'After clicking on Export lowest points of a model, '
'The new files will be located in logs/[yourmodelname]/lowestvals/ folder'
)
with gr.Row():
with gr.Accordion(label='Lowest points export'):
lowestval_weight_dir = gr.Textbox(visible=False)
ds = gr.Textbox(visible=False)
weights_dir1 = gr.Textbox(visible=False, value=weights_dir)
with gr.Row():
amntlastmdls = gr.Slider(
minimum=1,
maximum=25,
label='How many lowest points to save',
value=3,
step=1,
interactive=True,
)
lpexport = gr.Button(
value='Export lowest points of a model',
variant='primary',
)
lw_mdls = gr.File(
file_count="multiple",
label="Output models",
interactive=False,
) #####
with gr.Row():
infolpex = gr.Textbox(label="Output information:", value="", max_lines=10)
mdlbl = gr.Dataframe(label='Stats of selected models', datatype='number', type='pandas')
lpexport.click(
lambda model_name: os.path.join("logs", model_name, "lowestvals"),
inputs=[exp_dir1],
outputs=[lowestval_weight_dir]
)
lpexport.click(fn=tensorlowest.main, inputs=[exp_dir1, save_epoch10, amntlastmdls], outputs=[ds])
ds.change(
fn=tensorlowest.selectweights,
inputs=[exp_dir1, ds, weights_dir1, lowestval_weight_dir],
outputs=[infolpex, lw_mdls, mdlbl],
)
with gr.TabItem(i18n("ckpt处理")):
with gr.Group():
gr.Markdown(value=i18n("模型融合, 可用于测试音色融合"))
with gr.Row():
ckpt_a = gr.Textbox(label=i18n("A模型路径"), value="", interactive=True, placeholder="Path to your model A.")
ckpt_b = gr.Textbox(label=i18n("B模型路径"), value="", interactive=True, placeholder="Path to your model B.")
alpha_a = gr.Slider(
minimum=0,
maximum=1,
label=i18n("A模型权重"),
value=0.5,
interactive=True,
)
with gr.Row():
sr_ = gr.Radio(
label=i18n("目标采样率"),
choices=["40k", "48k"],
value="40k",
interactive=True,
)
if_f0_ = gr.Checkbox(
label="Whether the model has pitch guidance.",
value=True,
interactive=True,
)
info__ = gr.Textbox(
label=i18n("要置入的模型信息"), value="", max_lines=8, interactive=True, placeholder="Model information to be placed."
)
name_to_save0 = gr.Textbox(
label=i18n("保存的模型名不带后缀"),
value="",
placeholder="Name for saving.",
max_lines=1,
interactive=True,
)
version_2 = gr.Radio(
label=i18n("模型版本型号"),
choices=["v1", "v2"],
value="v1",
interactive=True,
)
with gr.Row():
but6 = gr.Button(i18n("融合"), variant="primary")
info4 = gr.Textbox(label=i18n("输出信息"), value="", max_lines=8)
but6.click(
merge,
[
ckpt_a,
ckpt_b,
alpha_a,
sr_,
if_f0_,
info__,
name_to_save0,
version_2,
],
info4,
) # def merge(path1,path2,alpha1,sr,f0,info):
with gr.Group():
gr.Markdown(value=i18n("修改模型信息(仅支持weights文件夹下提取的小模型文件)"))
with gr.Row(): ######
ckpt_path0 = gr.Textbox(
label=i18n("模型路径"), placeholder="Path to your Model.", value="", interactive=True
)
info_ = gr.Textbox(
label=i18n("要改的模型信息"), value="", max_lines=8, interactive=True, placeholder="Model information to be changed."
)
name_to_save1 = gr.Textbox(
label=i18n("保存的文件名, 默认空为和源文件同名"),
placeholder="Either leave empty or put in the Name of the Model to be saved.",
value="",
max_lines=8,
interactive=True,
)
with gr.Row():
but7 = gr.Button(i18n("修改"), variant="primary")
info5 = gr.Textbox(label=i18n("输出信息"), value="", max_lines=8)
but7.click(change_info, [ckpt_path0, info_, name_to_save1], info5)
with gr.Group():
gr.Markdown(value=i18n("查看模型信息(仅支持weights文件夹下提取的小模型文件)"))
with gr.Row():
ckpt_path1 = gr.Textbox(
label=i18n("模型路径"), value="", interactive=True, placeholder="Model path here."
)
but8 = gr.Button(i18n("查看"), variant="primary")
info6 = gr.Textbox(label=i18n("输出信息"), value="", max_lines=8)
but8.click(show_info, [ckpt_path1], info6)
with gr.Group():
gr.Markdown(
value=i18n(
"模型提取(输入logs文件夹下大文件模型路径),适用于训一半不想训了模型没有自动提取保存小文件模型,或者想测试中间模型的情况"
)
)
with gr.Row():
ckpt_path2 = gr.Textbox(
lines=3,
label=i18n("模型路径"),
value=os.path.abspath(os.getcwd()).replace('\\', '/') + "/logs/[YOUR_MODEL]/G_23333.pth",
interactive=True,
)
save_name = gr.Textbox(
label=i18n("保存名"), value="", interactive=True,
placeholder="Your filename here.",
)
sr__ = gr.Radio(
label=i18n("目标采样率"),
choices=["32k", "40k", "48k"],
value="40k",
interactive=True,
)
if_f0__ = gr.Checkbox(
label="Whether the model has pitch guidance.",
value=True,
interactive=True,
)
version_1 = gr.Radio(
label=i18n("模型版本型号"),
choices=["v1", "v2"],
value="v2",
interactive=True,
)
info___ = gr.Textbox(
label=i18n("要置入的模型信息"), value="", max_lines=8, interactive=True, placeholder="Model info here."
)
but9 = gr.Button(i18n("提取"), variant="primary")
info7 = gr.Textbox(label=i18n("输出信息"), value="", max_lines=8)
ckpt_path2.change(
change_info_, [ckpt_path2], [sr__, if_f0__, version_1]
)
but9.click(
extract_small_model,
[ckpt_path2, save_name, sr__, if_f0__, info___, version_1],
info7,
)
with gr.TabItem(i18n("Onnx导出")):
with gr.Row():
ckpt_dir = gr.Textbox(label=i18n("RVC模型路径"), value="", interactive=True, placeholder="RVC model path.")
with gr.Row():
onnx_dir = gr.Textbox(
label=i18n("Onnx输出路径"), value="", interactive=True, placeholder="Onnx model output path."
)
with gr.Row():
infoOnnx = gr.Label(label="info")
with gr.Row():
butOnnx = gr.Button(i18n("导出Onnx模型"), variant="primary")
butOnnx.click(export_onnx, [ckpt_dir, onnx_dir], infoOnnx)
with gr.TabItem(tab_faq):
try:
with open(faq_file, "r", encoding="utf8") as f:
info = f.read()
gr.Markdown(value=info)
except:
gr.Markdown(traceback.format_exc())
#region Mangio Preset Handler Region
def save_preset(
preset_name,
sid0,
vc_transform,
input_audio0,
input_audio1,
f0method,
crepe_hop_length,
filter_radius,
file_index1,
file_index2,
index_rate,
resample_sr,
rms_mix_rate,
protect,
f0_file
):
data = None
with open('../inference-presets.json', 'r') as file:
data = json.load(file)
preset_json = {
'name': preset_name,
'model': sid0,
'transpose': vc_transform,
'audio_file': input_audio0,
'auto_audio_file': input_audio1,
'f0_method': f0method,
'crepe_hop_length': crepe_hop_length,
'median_filtering': filter_radius,
'feature_path': file_index1,
'auto_feature_path': file_index2,
'search_feature_ratio': index_rate,
'resample': resample_sr,
'volume_envelope': rms_mix_rate,
'protect_voiceless': protect,
'f0_file_path': f0_file
}
data['presets'].append(preset_json)
with open('../inference-presets.json', 'w') as file:
json.dump(data, file)
file.flush()
print("Saved Preset %s into inference-presets.json!" % preset_name)
def on_preset_changed(preset_name):
print("Changed Preset to %s!" % preset_name)
data = None
with open('../inference-presets.json', 'r') as file:
data = json.load(file)
print("Searching for " + preset_name)
returning_preset = None
for preset in data['presets']:
if(preset['name'] == preset_name):
print("Found a preset")
returning_preset = preset
# return all new input values
return (
# returning_preset['model'],
# returning_preset['transpose'],
# returning_preset['audio_file'],
# returning_preset['f0_method'],
# returning_preset['crepe_hop_length'],
# returning_preset['median_filtering'],
# returning_preset['feature_path'],
# returning_preset['auto_feature_path'],
# returning_preset['search_feature_ratio'],
# returning_preset['resample'],
# returning_preset['volume_envelope'],
# returning_preset['protect_voiceless'],
# returning_preset['f0_file_path']
)
# Preset State Changes
# This click calls save_preset that saves the preset into inference-presets.json with the preset name
# mangio_preset_save_btn.click(
# fn=save_preset,
# inputs=[
# mangio_preset_name_save,
# sid0,
# vc_transform0,
# input_audio0,
# f0method0,
# crepe_hop_length,
# filter_radius0,
# file_index1,
# file_index2,
# index_rate1,
# resample_sr0,
# rms_mix_rate0,
# protect0,
# f0_file
# ],
# outputs=[]
# )
# mangio_preset.change(
# on_preset_changed,
# inputs=[
# # Pass inputs here
# mangio_preset
# ],
# outputs=[
# # Pass Outputs here. These refer to the gradio elements that we want to directly change
# # sid0,
# # vc_transform0,
# # input_audio0,
# # f0method0,
# # crepe_hop_length,
# # filter_radius0,
# # file_index1,
# # file_index2,
# # index_rate1,
# # resample_sr0,
# # rms_mix_rate0,
# # protect0,
# # f0_file
# ]
# )
#endregion
# with gr.TabItem(i18n("招募音高曲线前端编辑器")):
# gr.Markdown(value=i18n("加开发群联系我xxxxx"))
# with gr.TabItem(i18n("点击查看交流、问题反馈群号")):
# gr.Markdown(value=i18n("xxxxx"))
if config.iscolab or config.paperspace: # Share gradio link for colab and paperspace (FORK FEATURE)
app.queue(concurrency_count=511, max_size=1022).launch(share=True)
else:
app.queue(concurrency_count=511, max_size=1022).launch(
server_name="0.0.0.0",
inbrowser=not config.noautoopen,
server_port=config.listen_port,
quiet=False,
)
#endregion