1
0
mirror of synced 2024-11-30 18:24:32 +01:00
Retrieval-based-Voice-Conve.../tools/rvc_for_realtime.py

332 lines
12 KiB
Python
Raw Normal View History

2023-08-28 09:07:41 +02:00
import os
import sys
import traceback
from time import time as ttime
2023-08-12 19:05:58 +02:00
import fairseq
2023-08-28 09:07:41 +02:00
import faiss
import numpy as np
import parselmouth
import pyworld
import scipy.signal as signal
import torch
import torch.nn as nn
import torch.nn.functional as F
import torchcrepe
from infer.lib.infer_pack.models import (
2023-07-11 16:09:41 +02:00
SynthesizerTrnMs256NSFsid,
SynthesizerTrnMs256NSFsid_nono,
SynthesizerTrnMs768NSFsid,
SynthesizerTrnMs768NSFsid_nono,
)
2023-07-11 16:09:41 +02:00
now_dir = os.getcwd()
sys.path.append(now_dir)
from multiprocessing import Manager as M
2023-08-28 09:07:41 +02:00
from configs.config import Config
2023-08-29 08:47:17 +02:00
config = Config()
2023-08-28 09:07:41 +02:00
2023-07-11 16:09:41 +02:00
mm = M()
if config.dml == True:
2023-08-12 19:05:58 +02:00
def forward_dml(ctx, x, scale):
ctx.scale = scale
res = x.clone().detach()
return res
fairseq.modules.grad_multiply.GradMultiply.forward = forward_dml
2023-08-12 19:05:58 +02:00
# config.device=torch.device("cpu")########强制cpu测试
# config.is_half=False########强制cpu测试
2023-07-11 16:09:41 +02:00
class RVC:
def __init__(
self, key, pth_path, index_path, index_rate, n_cpu, inp_q, opt_q, device
2023-07-11 16:09:41 +02:00
) -> None:
"""
初始化
"""
try:
global config
self.inp_q = inp_q
self.opt_q = opt_q
2023-08-12 19:05:58 +02:00
# device="cpu"########强制cpu测试
self.device = device
2023-07-11 16:09:41 +02:00
self.f0_up_key = key
self.time_step = 160 / 16000 * 1000
self.f0_min = 50
self.f0_max = 1100
self.f0_mel_min = 1127 * np.log(1 + self.f0_min / 700)
self.f0_mel_max = 1127 * np.log(1 + self.f0_max / 700)
self.sr = 16000
self.window = 160
self.n_cpu = n_cpu
if index_rate != 0:
self.index = faiss.read_index(index_path)
self.big_npy = self.index.reconstruct_n(0, self.index.ntotal)
print("index search enabled")
self.index_rate = index_rate
2023-08-12 19:05:58 +02:00
models, _, _ = fairseq.checkpoint_utils.load_model_ensemble_and_task(
2023-08-29 09:00:24 +02:00
["assets/hubert/hubert_base.pt"],
2023-07-11 16:09:41 +02:00
suffix="",
)
hubert_model = models[0]
hubert_model = hubert_model.to(config.device)
if config.is_half:
hubert_model = hubert_model.half()
else:
hubert_model = hubert_model.float()
hubert_model.eval()
self.model = hubert_model
cpt = torch.load(pth_path, map_location="cpu")
self.tgt_sr = cpt["config"][-1]
cpt["config"][-3] = cpt["weight"]["emb_g.weight"].shape[0]
self.if_f0 = cpt.get("f0", 1)
self.version = cpt.get("version", "v1")
if self.version == "v1":
if self.if_f0 == 1:
self.net_g = SynthesizerTrnMs256NSFsid(
*cpt["config"], is_half=config.is_half
)
else:
self.net_g = SynthesizerTrnMs256NSFsid_nono(*cpt["config"])
elif self.version == "v2":
if self.if_f0 == 1:
self.net_g = SynthesizerTrnMs768NSFsid(
*cpt["config"], is_half=config.is_half
)
else:
self.net_g = SynthesizerTrnMs768NSFsid_nono(*cpt["config"])
del self.net_g.enc_q
print(self.net_g.load_state_dict(cpt["weight"], strict=False))
self.net_g.eval().to(device)
2023-08-12 19:05:58 +02:00
# print(2333333333,device,config.device,self.device)#net_g是devicehubert是config.device
2023-07-11 16:09:41 +02:00
if config.is_half:
self.net_g = self.net_g.half()
else:
self.net_g = self.net_g.float()
self.is_half = config.is_half
2023-07-11 16:09:41 +02:00
except:
print(traceback.format_exc())
def get_f0_post(self, f0):
f0_min = self.f0_min
f0_max = self.f0_max
f0_mel_min = 1127 * np.log(1 + f0_min / 700)
f0_mel_max = 1127 * np.log(1 + f0_max / 700)
f0bak = f0.copy()
f0_mel = 1127 * np.log(1 + f0 / 700)
f0_mel[f0_mel > 0] = (f0_mel[f0_mel > 0] - f0_mel_min) * 254 / (
f0_mel_max - f0_mel_min
) + 1
f0_mel[f0_mel <= 1] = 1
f0_mel[f0_mel > 255] = 255
f0_coarse = np.rint(f0_mel).astype(np.int32)
2023-07-11 16:09:41 +02:00
return f0_coarse, f0bak
def get_f0(self, x, f0_up_key, n_cpu, method="harvest"):
n_cpu = int(n_cpu)
if method == "crepe":
return self.get_f0_crepe(x, f0_up_key)
if method == "rmvpe":
return self.get_f0_rmvpe(x, f0_up_key)
if method == "pm":
2023-07-11 16:09:41 +02:00
p_len = x.shape[0] // 160
f0 = (
parselmouth.Sound(x, 16000)
.to_pitch_ac(
2023-07-11 16:09:41 +02:00
time_step=0.01,
voicing_threshold=0.6,
pitch_floor=50,
pitch_ceiling=1100,
)
.selected_array["frequency"]
2023-07-11 16:09:41 +02:00
)
pad_size = (p_len - len(f0) + 1) // 2
if pad_size > 0 or p_len - len(f0) - pad_size > 0:
2023-08-12 19:05:58 +02:00
# print(pad_size, p_len - len(f0) - pad_size)
f0 = np.pad(
f0, [[pad_size, p_len - len(f0) - pad_size]], mode="constant"
)
2023-07-11 16:09:41 +02:00
f0 *= pow(2, f0_up_key / 12)
return self.get_f0_post(f0)
if n_cpu == 1:
2023-07-11 16:09:41 +02:00
f0, t = pyworld.harvest(
x.astype(np.double),
fs=16000,
f0_ceil=1100,
f0_floor=50,
frame_period=10,
)
f0 = signal.medfilt(f0, 3)
f0 *= pow(2, f0_up_key / 12)
return self.get_f0_post(f0)
f0bak = np.zeros(x.shape[0] // 160, dtype=np.float64)
length = len(x)
part_length = int(length / n_cpu / 160) * 160
ts = ttime()
res_f0 = mm.dict()
for idx in range(n_cpu):
tail = part_length * (idx + 1) + 320
if idx == 0:
2023-07-11 16:09:41 +02:00
self.inp_q.put((idx, x[:tail], res_f0, n_cpu, ts))
else:
self.inp_q.put(
(idx, x[part_length * idx - 320 : tail], res_f0, n_cpu, ts)
)
while 1:
2023-07-11 16:09:41 +02:00
res_ts = self.opt_q.get()
if res_ts == ts:
2023-07-11 16:09:41 +02:00
break
f0s = [i[1] for i in sorted(res_f0.items(), key=lambda x: x[0])]
for idx, f0 in enumerate(f0s):
if idx == 0:
2023-07-11 16:09:41 +02:00
f0 = f0[:-3]
elif idx != n_cpu - 1:
2023-07-11 16:09:41 +02:00
f0 = f0[2:-3]
else:
f0 = f0[2:-1]
f0bak[
part_length * idx // 160 : part_length * idx // 160 + f0.shape[0]
] = f0
2023-07-11 16:09:41 +02:00
f0bak = signal.medfilt(f0bak, 3)
f0bak *= pow(2, f0_up_key / 12)
return self.get_f0_post(f0bak)
def get_f0_crepe(self, x, f0_up_key):
2023-08-29 12:53:33 +02:00
if "privateuseone" in str(self.device): ###不支持dmlcpu又太慢用不成拿pm顶替
return self.get_f0(x, f0_up_key, 1, "pm")
2023-07-11 16:09:41 +02:00
audio = torch.tensor(np.copy(x))[None].float()
2023-08-12 19:05:58 +02:00
# print("using crepe,device:%s"%self.device)
2023-07-11 16:09:41 +02:00
f0, pd = torchcrepe.predict(
audio,
self.sr,
160,
self.f0_min,
self.f0_max,
"full",
batch_size=512,
2023-08-12 19:05:58 +02:00
# device=self.device if self.device.type!="privateuseone" else "cpu",###crepe不用半精度全部是全精度所以不愁###cpu延迟高到没法用
2023-07-11 16:09:41 +02:00
device=self.device,
return_periodicity=True,
)
pd = torchcrepe.filter.median(pd, 3)
f0 = torchcrepe.filter.mean(f0, 3)
f0[pd < 0.1] = 0
f0 = f0[0].cpu().numpy()
f0 *= pow(2, f0_up_key / 12)
return self.get_f0_post(f0)
def get_f0_rmvpe(self, x, f0_up_key):
if hasattr(self, "model_rmvpe") == False:
2023-08-29 09:00:24 +02:00
from infer.lib.rmvpe import RMVPE
2023-07-11 16:09:41 +02:00
print("loading rmvpe model")
self.model_rmvpe = RMVPE(
2023-08-12 19:05:58 +02:00
# "rmvpe.pt", is_half=self.is_half if self.device.type!="privateuseone" else False, device=self.device if self.device.type!="privateuseone"else "cpu"####dml时强制对rmvpe用cpu跑
# "rmvpe.pt", is_half=False, device=self.device####dml配置
# "rmvpe.pt", is_half=False, device="cpu"####锁定cpu配置
2023-08-29 09:00:24 +02:00
"assets/rmvpe/rmvpe.pt",
is_half=self.is_half,
device=self.device, ####正常逻辑
)
2023-07-11 16:09:41 +02:00
# self.model_rmvpe = RMVPE("aug2_58000_half.pt", is_half=self.is_half, device=self.device)
f0 = self.model_rmvpe.infer_from_audio(x, thred=0.03)
f0 *= pow(2, f0_up_key / 12)
return self.get_f0_post(f0)
def infer(
self,
feats: torch.Tensor,
indata: np.ndarray,
rate1,
rate2,
cache_pitch,
cache_pitchf,
f0method,
) -> np.ndarray:
2023-07-11 16:09:41 +02:00
feats = feats.view(1, -1)
if config.is_half:
feats = feats.half()
else:
feats = feats.float()
feats = feats.to(self.device)
t1 = ttime()
with torch.no_grad():
padding_mask = torch.BoolTensor(feats.shape).to(self.device).fill_(False)
inputs = {
"source": feats,
"padding_mask": padding_mask,
"output_layer": 9 if self.version == "v1" else 12,
}
logits = self.model.extract_features(**inputs)
feats = (
self.model.final_proj(logits[0]) if self.version == "v1" else logits[0]
)
2023-07-11 16:09:41 +02:00
t2 = ttime()
try:
if hasattr(self, "index") and self.index_rate != 0:
2023-07-11 16:09:41 +02:00
leng_replace_head = int(rate1 * feats[0].shape[0])
npy = feats[0][-leng_replace_head:].cpu().numpy().astype("float32")
score, ix = self.index.search(npy, k=8)
weight = np.square(1 / score)
weight /= weight.sum(axis=1, keepdims=True)
npy = np.sum(self.big_npy[ix] * np.expand_dims(weight, axis=2), axis=1)
if config.is_half:
npy = npy.astype("float16")
feats[0][-leng_replace_head:] = (
torch.from_numpy(npy).unsqueeze(0).to(self.device) * self.index_rate
+ (1 - self.index_rate) * feats[0][-leng_replace_head:]
)
else:
print("index search FAIL or disabled")
except:
traceback.print_exc()
print("index search FAIL")
feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
t3 = ttime()
if self.if_f0 == 1:
pitch, pitchf = self.get_f0(indata, self.f0_up_key, self.n_cpu, f0method)
cache_pitch[:] = np.append(cache_pitch[pitch[:-1].shape[0] :], pitch[:-1])
cache_pitchf[:] = np.append(
cache_pitchf[pitchf[:-1].shape[0] :], pitchf[:-1]
)
2023-07-11 16:09:41 +02:00
p_len = min(feats.shape[1], 13000, cache_pitch.shape[0])
else:
cache_pitch, cache_pitchf = None, None
p_len = min(feats.shape[1], 13000)
t4 = ttime()
feats = feats[:, :p_len, :]
if self.if_f0 == 1:
cache_pitch = cache_pitch[:p_len]
cache_pitchf = cache_pitchf[:p_len]
cache_pitch = torch.LongTensor(cache_pitch).unsqueeze(0).to(self.device)
cache_pitchf = torch.FloatTensor(cache_pitchf).unsqueeze(0).to(self.device)
p_len = torch.LongTensor([p_len]).to(self.device)
ii = 0 # sid
sid = torch.LongTensor([ii]).to(self.device)
with torch.no_grad():
if self.if_f0 == 1:
2023-08-12 19:05:58 +02:00
# print(12222222222,feats.device,p_len.device,cache_pitch.device,cache_pitchf.device,sid.device,rate2)
2023-07-11 16:09:41 +02:00
infered_audio = (
self.net_g.infer(
feats, p_len, cache_pitch, cache_pitchf, sid, rate2
)[0][0, 0]
.data.cpu()
.float()
2023-07-11 16:09:41 +02:00
)
else:
infered_audio = (
self.net_g.infer(feats, p_len, sid, rate2)[0][0, 0]
.data.cpu()
.float()
2023-07-11 16:09:41 +02:00
)
t5 = ttime()
print("time->fea-index-f0-model:", t2 - t1, t3 - t2, t4 - t3, t5 - t4)
return infered_audio