PublicMirrors/voice-changer
1
0
Fork 0
mirror of https://github.com/w-okada/voice-changer.git synced 2025-02-02 16:23:58 +03:00
Code Issues Actions Packages Projects Releases Wiki Activity

WIP: refactoring

Browse Source
This commit is contained in:
wataru 2023-04-28 10:30:07 +09:00
parent 83083a68ac
commit 443fe1e720
2 changed files with 216 additions and 126 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
server/.vscode/settings.json vendored
View File

@ -9,7 +9,7 @@
"editor.formatOnSave": true // ファイル保存時に自動フォーマット "editor.formatOnSave": true // ファイル保存時に自動フォーマット
}, },
"flake8.args": [ "flake8.args": [
"--ignore=E501,E402,E722,W503" "--ignore=E501,E402,E722,E741,W503"
// "--max-line-length=150", // "--max-line-length=150",
// "--max-complexity=20" // "--max-complexity=20"
] ]

340
server/voice_changer/RVC/custom_vc_infer_pipeline.py
View File

@ -2,10 +2,9 @@ import numpy as np
import parselmouth import parselmouth
import torch import torch
import torch.nn.functional as F import torch.nn.functional as F
from config import x_pad, x_query, x_center, x_max from config import x_query, x_center, x_max # type:ignore
import scipy.signal as signal import scipy.signal as signal
import pyworld import pyworld
from .const import RVC_MODEL_TYPE_RVC, RVC_MODEL_TYPE_WEBUI
class VC(object): class VC(object):
@ -20,12 +19,13 @@ class VC(object):
self.device = device self.device = device
self.is_half = is_half self.is_half = is_half
def get_f0(self, audio, p_len, f0_up_key, f0_method, inp_f0=None, silence_front=0): def get_f0(self, audio, p_len, f0_up_key, f0_method, silence_front=0):
n_frames = int(len(audio) // self.window) + 1 n_frames = int(len(audio) // self.window) + 1
start_frame = int(silence_front * self.sr / self.window) start_frame = int(silence_front * self.sr / self.window)
real_silence_front = start_frame * self.window / self.sr real_silence_front = start_frame * self.window / self.sr
audio = audio[int(np.round(real_silence_front * self.sr)) :] silence_front_offset = int(np.round(real_silence_front * self.sr))
audio = audio[silence_front_offset:]
time_step = self.window / self.sr * 1000 time_step = self.window / self.sr * 1000
f0_min = 50 f0_min = 50
@ -80,18 +80,6 @@ class VC(object):
) )
f0 *= pow(2, f0_up_key / 12) f0 *= pow(2, f0_up_key / 12)
# with open("test.txt","w")as f:f.write("\n".join([str(i)for i in f0.tolist()]))
tf0 = self.sr // self.window # 每秒f0点数
if inp_f0 is not None:
delta_t = np.round(
(inp_f0[:, 0].max() - inp_f0[:, 0].min()) * tf0 + 1
).astype("int16")
replace_f0 = np.interp(
list(range(delta_t)), inp_f0[:, 0] * 100, inp_f0[:, 1]
)
shape = f0[x_pad * tf0 : x_pad * tf0 + len(replace_f0)].shape[0]
f0[x_pad * tf0 : x_pad * tf0 + len(replace_f0)] = replace_f0[:shape]
# with open("test_opt.txt","w")as f:f.write("\n".join([str(i)for i in f0.tolist()]))
f0bak = f0.copy() f0bak = f0.copy()
f0_mel = 1127 * np.log(1 + f0 / 700) f0_mel = 1127 * np.log(1 + f0 / 700)
f0_mel[f0_mel > 0] = (f0_mel[f0_mel > 0] - f0_mel_min) * 254 / ( f0_mel[f0_mel > 0] = (f0_mel[f0_mel > 0] - f0_mel_min) * 254 / (
@ -102,107 +90,108 @@ class VC(object):
f0_coarse = np.rint(f0_mel).astype(np.int) f0_coarse = np.rint(f0_mel).astype(np.int)
return f0_coarse, f0bak # 1-0 return f0_coarse, f0bak # 1-0
def vc( # def vc(
self, # self,
model, # model,
net_g, # net_g,
sid, # sid,
audio0, # audio0,
pitch, # pitch,
pitchf, # pitchf,
index, # index,
big_npy, # big_npy,
index_rate, # index_rate,
embChannels=256, # embChannels=256,
): # ,file_index,file_big_npy # ): # ,file_index,file_big_npy
feats = torch.from_numpy(audio0) # feats = torch.from_numpy(audio0)
if self.is_half == True: # if self.is_half is True:
feats = feats.half() # feats = feats.half()
else: # else:
feats = feats.float() # feats = feats.float()
if feats.dim() == 2: # double channels # if feats.dim() == 2: # double channels
feats = feats.mean(-1) # feats = feats.mean(-1)
assert feats.dim() == 1, feats.dim() # assert feats.dim() == 1, feats.dim()
feats = feats.view(1, -1) # feats = feats.view(1, -1)
padding_mask = torch.BoolTensor(feats.shape).to(self.device).fill_(False)
if embChannels == 256:
inputs = {
"source": feats.to(self.device),
"padding_mask": padding_mask,
"output_layer": 9, # layer 9
}
else:
inputs = {
"source": feats.to(self.device),
"padding_mask": padding_mask,
}
with torch.no_grad(): # padding_mask = torch.BoolTensor(feats.shape).to(self.device).fill_(False)
logits = model.extract_features(**inputs) # if embChannels == 256:
if embChannels == 256: # inputs = {
feats = model.final_proj(logits[0]) # "source": feats.to(self.device),
else: # "padding_mask": padding_mask,
feats = logits[0] # "output_layer": 9, # layer 9
# }
# else:
# inputs = {
# "source": feats.to(self.device),
# "padding_mask": padding_mask,
# }
if ( # with torch.no_grad():
isinstance(index, type(None)) is False # logits = model.extract_features(**inputs)
and isinstance(big_npy, type(None)) is False # if embChannels == 256:
and index_rate != 0 # feats = model.final_proj(logits[0])
): # else:
npy = feats[0].cpu().numpy() # feats = logits[0]
if self.is_half is True:
npy = npy.astype("float32")
D, I = index.search(npy, 1)
npy = big_npy[I.squeeze()]
if self.is_half is True:
npy = npy.astype("float16")
feats = ( # if (
torch.from_numpy(npy).unsqueeze(0).to(self.device) * index_rate # isinstance(index, type(None)) is False
+ (1 - index_rate) * feats # and isinstance(big_npy, type(None)) is False
) # and index_rate != 0
# ):
# npy = feats[0].cpu().numpy()
# if self.is_half is True:
# npy = npy.astype("float32")
# D, I = index.search(npy, 1)
# npy = big_npy[I.squeeze()]
# if self.is_half is True:
# npy = npy.astype("float16")
feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1) # feats = (
# torch.from_numpy(npy).unsqueeze(0).to(self.device) * index_rate
# + (1 - index_rate) * feats
# )
p_len = audio0.shape[0] // self.window # feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
if feats.shape[1] < p_len:
p_len = feats.shape[1]
if pitch is not None and pitchf is not None:
pitch = pitch[:, :p_len]
pitchf = pitchf[:, :p_len]
p_len = torch.tensor([p_len], device=self.device).long()
with torch.no_grad(): # p_len = audio0.shape[0] // self.window
if pitch is not None: # if feats.shape[1] < p_len:
audio1 = ( # p_len = feats.shape[1]
(net_g.infer(feats, p_len, pitch, pitchf, sid)[0][0, 0] * 32768) # if pitch is not None and pitchf is not None:
.data.cpu() # pitch = pitch[:, :p_len]
.float() # pitchf = pitchf[:, :p_len]
.numpy() # p_len = torch.tensor([p_len], device=self.device).long()
.astype(np.int16)
)
else:
if hasattr(net_g, "infer_pitchless"):
audio1 = (
(net_g.infer_pitchless(feats, p_len, sid)[0][0, 0] * 32768)
.data.cpu()
.float()
.numpy()
.astype(np.int16)
)
else:
audio1 = (
(net_g.infer(feats, p_len, sid)[0][0, 0] * 32768)
.data.cpu()
.float()
.numpy()
.astype(np.int16)
)
del feats, p_len, padding_mask # with torch.no_grad():
torch.cuda.empty_cache() # if pitch is not None:
# audio1 = (
# (net_g.infer(feats, p_len, pitch, pitchf, sid)[0][0, 0] * 32768)
# .data.cpu()
# .float()
# .numpy()
# .astype(np.int16)
# )
# else:
# if hasattr(net_g, "infer_pitchless"):
# audio1 = (
# (net_g.infer_pitchless(feats, p_len, sid)[0][0, 0] * 32768)
# .data.cpu()
# .float()
# .numpy()
# .astype(np.int16)
# )
# else:
# audio1 = (
# (net_g.infer(feats, p_len, sid)[0][0, 0] * 32768)
# .data.cpu()
# .float()
# .numpy()
# .astype(np.int16)
# )
return audio1 # del feats, p_len, padding_mask
# torch.cuda.empty_cache()
# return audio1
def pipeline( def pipeline(
self, self,
@ -221,7 +210,6 @@ class VC(object):
): ):
audio_pad = np.pad(audio, (self.t_pad, self.t_pad), mode="reflect") audio_pad = np.pad(audio, (self.t_pad, self.t_pad), mode="reflect")
p_len = audio_pad.shape[0] // self.window p_len = audio_pad.shape[0] // self.window
inp_f0 = None
sid = torch.tensor(sid, device=self.device).unsqueeze(0).long() sid = torch.tensor(sid, device=self.device).unsqueeze(0).long()
# ピッチ検出 # ピッチ検出
@ -232,7 +220,6 @@ class VC(object):
p_len, p_len,
f0_up_key, f0_up_key,
f0_method, f0_method,
inp_f0,
silence_front=silence_front, silence_front=silence_front,
) )
pitch = pitch[:p_len] pitch = pitch[:p_len]
@ -242,23 +229,126 @@ class VC(object):
pitchf, device=self.device, dtype=torch.float pitchf, device=self.device, dtype=torch.float
).unsqueeze(0) ).unsqueeze(0)
output = self.vc( # tensor
embedder, feats = torch.from_numpy(audio_pad)
model, if self.is_half is True:
sid, feats = feats.half()
audio_pad, else:
pitch, feats = feats.float()
pitchf, if feats.dim() == 2: # double channels
index, feats = feats.mean(-1)
big_npy, assert feats.dim() == 1, feats.dim()
index_rate, feats = feats.view(1, -1)
embChannels,
) # embedding
padding_mask = torch.BoolTensor(feats.shape).to(self.device).fill_(False)
if embChannels == 256:
inputs = {
"source": feats.to(self.device),
"padding_mask": padding_mask,
"output_layer": 9, # layer 9
}
else:
inputs = {
"source": feats.to(self.device),
"padding_mask": padding_mask,
}
with torch.no_grad():
logits = embedder.extract_features(**inputs)
if embChannels == 256:
feats = embedder.final_proj(logits[0])
else:
feats = logits[0]
# Index - feature抽出
if (
isinstance(index, type(None)) is False
and isinstance(big_npy, type(None)) is False
and index_rate != 0
):
npy = feats[0].cpu().numpy()
if self.is_half is True:
npy = npy.astype("float32")
D, I = index.search(npy, 1)
npy = big_npy[I.squeeze()]
if self.is_half is True:
npy = npy.astype("float16")
feats = (
torch.from_numpy(npy).unsqueeze(0).to(self.device) * index_rate
+ (1 - index_rate) * feats
)
#
feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
# ピッチ抽出
p_len = audio_pad.shape[0] // self.window
if feats.shape[1] < p_len:
p_len = feats.shape[1]
if pitch is not None and pitchf is not None:
pitch = pitch[:, :p_len]
pitchf = pitchf[:, :p_len]
p_len = torch.tensor([p_len], device=self.device).long()
# 推論実行
with torch.no_grad():
if pitch is not None:
audio1 = (
(model.infer(feats, p_len, pitch, pitchf, sid)[0][0, 0] * 32768)
.data.cpu()
.float()
.numpy()
.astype(np.int16)
)
else:
if hasattr(model, "infer_pitchless"):
audio1 = (
(model.infer_pitchless(feats, p_len, sid)[0][0, 0] * 32768)
.data.cpu()
.float()
.numpy()
.astype(np.int16)
)
else:
audio1 = (
(model.infer(feats, p_len, sid)[0][0, 0] * 32768)
.data.cpu()
.float()
.numpy()
.astype(np.int16)
)
del feats, p_len, padding_mask
torch.cuda.empty_cache()
if self.t_pad_tgt != 0: if self.t_pad_tgt != 0:
offset = self.t_pad_tgt offset = self.t_pad_tgt
end = -1 * self.t_pad_tgt end = -1 * self.t_pad_tgt
output = output[offset:end] audio1 = audio1[offset:end]
del pitch, pitchf, sid del pitch, pitchf, sid
torch.cuda.empty_cache() torch.cuda.empty_cache()
return output return audio1
# output = self.vc(
# embedder,
# model,
# sid,
# audio_pad,
# pitch,
# pitchf,
# index,
# big_npy,
# index_rate,
# embChannels,
# )
# if self.t_pad_tgt != 0:
# offset = self.t_pad_tgt
# end = -1 * self.t_pad_tgt
# output = output[offset:end]
# del pitch, pitchf, sid
# torch.cuda.empty_cache()
# return output