fix: 卸载音色省显存

顺便将所有print换成了统一的logger

configs/config.py

@@ -5,6 +5,10 @@ from multiprocessing import cpu_count
 
 import torch
 
+import logging
+
+logger = logging.getLogger(__name__)
+
 
 def use_fp32_config():
     for config_file in [
@@ -110,11 +114,11 @@ class Config:
                 or "1070" in self.gpu_name
                 or "1080" in self.gpu_name
             ):
-                print("Found GPU", self.gpu_name, ", force to fp32")
+                logger.info("Found GPU", self.gpu_name, ", force to fp32")
                 self.is_half = False
                 use_fp32_config()
             else:
-                print("Found GPU", self.gpu_name)
+                logger.info("Found GPU", self.gpu_name)
             self.gpu_mem = int(
                 torch.cuda.get_device_properties(i_device).total_memory
                 / 1024
@@ -128,12 +132,12 @@ class Config:
                 with open("infer/modules/train/preprocess.py", "w") as f:
                     f.write(strr)
         elif self.has_mps():
-            print("No supported Nvidia GPU found")
+            logger.info("No supported Nvidia GPU found")
             self.device = self.instead = "mps"
             self.is_half = False
             use_fp32_config()
         else:
-            print("No supported Nvidia GPU found")
+            logger.info("No supported Nvidia GPU found")
             self.device = self.instead = "cpu"
             self.is_half = False
             use_fp32_config()
@@ -160,7 +164,7 @@ class Config:
             x_center = 30
             x_max = 32
         if self.dml:
-            print("Use DirectML instead")
+            logger.info("Use DirectML instead")
             if (
                 os.path.exists(
                     "runtime\Lib\site-packages\onnxruntime\capi\DirectML.dll"
@@ -188,7 +192,7 @@ class Config:
             self.is_half = False
         else:
             if self.instead:
-                print(f"Use {self.instead} instead")
+                logger.info(f"Use {self.instead} instead")
             if (
                 os.path.exists(
                     "runtime\Lib\site-packages\onnxruntime\capi\onnxruntime_providers_cuda.dll"

gui_v1.py

@@ -1,5 +1,5 @@
 import os
-import pdb
+import logging
 import sys
 
 os.environ["OMP_NUM_THREADS"] = "2"
@@ -11,6 +11,8 @@ now_dir = os.getcwd()
 sys.path.append(now_dir)
 import multiprocessing
 
+logger = logging.getLogger(__name__)
+
 
 class Harvest(multiprocessing.Process):
     def __init__(self, inp_q, opt_q):
@@ -356,7 +358,7 @@ if __name__ == "__main__":
                     )
                 if event == "start_vc" and self.flag_vc == False:
                     if self.set_values(values) == True:
-                        print("Use CUDA:" + str(torch.cuda.is_available()))
+                        logger.info("Use CUDA:" + str(torch.cuda.is_available()))
                         self.start_vc()
                         settings = {
                             "pth_path": values["pth_path"],
@@ -545,8 +547,8 @@ if __name__ == "__main__":
             ):
                 while self.flag_vc:
                     time.sleep(self.config.block_time)
-                    print("Audio block passed.")
+                    logger.debug("Audio block passed.")
-            print("ENDing VC")
+            logger.debug("ENDing VC")
 
         def audio_callback(
             self, indata: np.ndarray, outdata: np.ndarray, frames, times, status
@@ -623,7 +625,7 @@ if __name__ == "__main__":
                 sola_offset = sola_offset.item()
             else:
                 sola_offset = torch.argmax(cor_nom[0, 0] / cor_den[0, 0])
-            print("sola_offset =" + str(int(sola_offset)))
+            logger.debug("sola_offset =" + str(int(sola_offset)))
             self.output_wav[:] = infer_wav[sola_offset : sola_offset + self.block_frame]
             self.output_wav[: self.crossfade_frame] *= self.fade_in_window
             self.output_wav[: self.crossfade_frame] += self.sola_buffer[:]
@@ -663,7 +665,7 @@ if __name__ == "__main__":
                     outdata[:] = self.output_wav[:].repeat(2, 1).t().cpu().numpy()
             total_time = time.perf_counter() - start_time
             self.window["infer_time"].update(int(total_time * 1000))
-            print("Infer time:" + str(total_time))
+            logger.info("Infer time:" + str(total_time))
 
         def get_devices(self, update: bool = True):
             """获取设备列表"""
@@ -716,8 +718,8 @@ if __name__ == "__main__":
             sd.default.device[1] = output_device_indices[
                 output_devices.index(output_device)
             ]
-            print("Input device:" + str(sd.default.device[0]) + ":" + str(input_device))
+            logger.info("Input device:" + str(sd.default.device[0]) + ":" + str(input_device))
-            print(
+            logger.info(
                 "Output device:" + str(sd.default.device[1]) + ":" + str(output_device)
             )
 

i18n/i18n.py

@@ -18,11 +18,10 @@ class I18nAuto:
         if not os.path.exists(f"./i18n/locale/{language}.json"):
             language = "en_US"
         self.language = language
-        # print("Use Language:", language)
         self.language_map = load_language_list(language)
 
     def __call__(self, key):
         return self.language_map.get(key, key)
 
-    def print(self):
+    def __repr__(self):
-        print("Use Language:", self.language)
+        return "Use Language: " + self.language

infer-web.py

@@ -32,6 +32,8 @@ from infer.modules.vc.modules import VC
 
 logging.getLogger("numba").setLevel(logging.WARNING)
 
+logger = logging.getLogger(__name__)
+
 tmp = os.path.join(now_dir, "TEMP")
 shutil.rmtree(tmp, ignore_errors=True)
 shutil.rmtree("%s/runtime/Lib/site-packages/infer_pack" % (now_dir), ignore_errors=True)
@@ -58,7 +60,7 @@ if config.dml == True:
 
     fairseq.modules.grad_multiply.GradMultiply.forward = forward_dml
 i18n = I18nAuto()
-i18n.print()
+logger.info(i18n)
 # 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 = []
@@ -213,7 +215,7 @@ def preprocess_dataset(trainset_dir, exp_dir, sr, n_p):
         % (trainset_dir, sr, n_p, now_dir, exp_dir)
         + str(config.noparallel)
     )
-    print(cmd)
+    logger.info(cmd)
     p = Popen(cmd, shell=True)  # , stdin=PIPE, stdout=PIPE,stderr=PIPE,cwd=now_dir
     ###煞笔gr, popen read都非得全跑完了再一次性读取, 不用gr就正常读一句输出一句;只能额外弄出一个文本流定时读
     done = [False]
@@ -232,7 +234,7 @@ def preprocess_dataset(trainset_dir, exp_dir, sr, n_p):
             break
     with open("%s/logs/%s/preprocess.log" % (now_dir, exp_dir), "r") as f:
         log = f.read()
-    print(log)
+    logger.info(log)
     yield log
 
 
@@ -254,7 +256,7 @@ def extract_f0_feature(gpus, n_p, f0method, if_f0, exp_dir, version19, gpus_rmvp
                     f0method,
                 )
             )
-            print(cmd)
+            logger.info(cmd)
             p = Popen(
                 cmd, shell=True, cwd=now_dir
             )  # , stdin=PIPE, stdout=PIPE,stderr=PIPE
@@ -281,7 +283,7 @@ def extract_f0_feature(gpus, n_p, f0method, if_f0, exp_dir, version19, gpus_rmvp
                         exp_dir,
                         config.is_half,
                     )
-                    print(cmd)
+                    logger.info(cmd)
                     p = Popen(
                         cmd, shell=True, cwd=now_dir
                     )  # , shell=True, stdin=PIPE, stdout=PIPE, stderr=PIPE, cwd=now_dir
@@ -304,7 +306,7 @@ def extract_f0_feature(gpus, n_p, f0method, if_f0, exp_dir, version19, gpus_rmvp
                         exp_dir,
                     )
                 )
-                print(cmd)
+                logger.info(cmd)
                 p = Popen(
                     cmd, shell=True, cwd=now_dir
                 )  # , shell=True, stdin=PIPE, stdout=PIPE, stderr=PIPE, cwd=now_dir
@@ -320,7 +322,7 @@ def extract_f0_feature(gpus, n_p, f0method, if_f0, exp_dir, version19, gpus_rmvp
                 break
         with open("%s/logs/%s/extract_f0_feature.log" % (now_dir, exp_dir), "r") as f:
             log = f.read()
-        print(log)
+        logger.info(log)
         yield log
     ####对不同part分别开多进程
     """
@@ -342,7 +344,7 @@ def extract_f0_feature(gpus, n_p, f0method, if_f0, exp_dir, version19, gpus_rmvp
             exp_dir,
             version19,
         )
-        print(cmd)
+        logger.info(cmd)
         p = Popen(
             cmd, shell=True, cwd=now_dir
         )  # , shell=True, stdin=PIPE, stdout=PIPE, stderr=PIPE, cwd=now_dir
@@ -364,7 +366,7 @@ def extract_f0_feature(gpus, n_p, f0method, if_f0, exp_dir, version19, gpus_rmvp
             break
     with open("%s/logs/%s/extract_f0_feature.log" % (now_dir, exp_dir), "r") as f:
         log = f.read()
-    print(log)
+    logger.info(log)
     yield log
 
 
@@ -378,12 +380,12 @@ def change_sr2(sr2, if_f0_3, version19):
         "assets/pretrained%s/%sD%s.pth" % (path_str, f0_str, sr2), os.F_OK
     )
     if not if_pretrained_generator_exist:
-        print(
+        logger.warn(
             "assets/pretrained%s/%sG%s.pth" % (path_str, f0_str, sr2),
             "not exist, will not use pretrained model",
         )
     if not if_pretrained_discriminator_exist:
-        print(
+        logger.warn(
             "assets/pretrained%s/%sD%s.pth" % (path_str, f0_str, sr2),
             "not exist, will not use pretrained model",
         )
@@ -414,12 +416,12 @@ def change_version19(sr2, if_f0_3, version19):
         "assets/pretrained%s/%sD%s.pth" % (path_str, f0_str, sr2), os.F_OK
     )
     if not if_pretrained_generator_exist:
-        print(
+        logger.warn(
             "assets/pretrained%s/%sG%s.pth" % (path_str, f0_str, sr2),
             "not exist, will not use pretrained model",
         )
     if not if_pretrained_discriminator_exist:
-        print(
+        logger.warn(
             "assets/pretrained%s/%sD%s.pth" % (path_str, f0_str, sr2),
             "not exist, will not use pretrained model",
         )
@@ -443,12 +445,12 @@ def change_f0(if_f0_3, sr2, version19):  # f0method8,pretrained_G14,pretrained_D
         "assets/pretrained%s/f0D%s.pth" % (path_str, sr2), os.F_OK
     )
     if not if_pretrained_generator_exist:
-        print(
+        logger.warn(
             "assets/pretrained%s/f0G%s.pth" % (path_str, sr2),
             "not exist, will not use pretrained model",
         )
     if not if_pretrained_discriminator_exist:
-        print(
+        logger.warn(
             "assets/pretrained%s/f0D%s.pth" % (path_str, sr2),
             "not exist, will not use pretrained model",
         )
@@ -556,14 +558,14 @@ def click_train(
     shuffle(opt)
     with open("%s/filelist.txt" % exp_dir, "w") as f:
         f.write("\n".join(opt))
-    print("Write filelist done")
+    logger.debug("Write filelist done")
     # 生成config#无需生成config
     # cmd = python_cmd + " train_nsf_sim_cache_sid_load_pretrain.py -e mi-test -sr 40k -f0 1 -bs 4 -g 0 -te 10 -se 5 -pg pretrained/f0G40k.pth -pd pretrained/f0D40k.pth -l 1 -c 0"
-    print("Use gpus:", gpus16)
+    logger.info("Use gpus:", gpus16)
     if pretrained_G14 == "":
-        print("No pretrained Generator")
+        logger.info("No pretrained Generator")
     if pretrained_D15 == "":
-        print("No pretrained Discriminator")
+        logger.info("No pretrained Discriminator")
     if gpus16:
         cmd = get_quoted_python_cmd() + ' infer/modules/train/train.py -e "%s" -sr %s -f0 %s -bs %s -g %s -te %s -se %s %s %s -l %s -c %s -sw %s -v %s' % (
             exp_dir1,
@@ -599,7 +601,7 @@ def click_train(
                 version19,
             )
         )
-    print(cmd)
+    logger.info(cmd)
     p = Popen(cmd, shell=True, cwd=now_dir)
     p.wait()
     return "训练结束, 您可查看控制台训练日志或实验文件夹下的train.log"
@@ -646,7 +648,7 @@ def train_index(exp_dir1, version19):
             )
         except:
             info = traceback.format_exc()
-            print(info)
+            logger.info(info)
             infos.append(info)
             yield "\n".join(infos)
 

infer/lib/infer_pack/models.py

@@ -1,7 +1,6 @@
 import math
-import os
+import logging
-import pdb
+logger = logging.getLogger(__name__)
-from time import time as ttime
 
 import numpy as np
 import torch
@@ -616,7 +615,7 @@ class SynthesizerTrnMs256NSFsid(nn.Module):
             inter_channels, hidden_channels, 5, 1, 3, gin_channels=gin_channels
         )
         self.emb_g = nn.Embedding(self.spk_embed_dim, gin_channels)
-        print("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
+        logger.debug("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
 
     def remove_weight_norm(self):
         self.dec.remove_weight_norm()
@@ -732,7 +731,7 @@ class SynthesizerTrnMs768NSFsid(nn.Module):
             inter_channels, hidden_channels, 5, 1, 3, gin_channels=gin_channels
         )
         self.emb_g = nn.Embedding(self.spk_embed_dim, gin_channels)
-        print("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
+        logger.debug("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
 
     def remove_weight_norm(self):
         self.dec.remove_weight_norm()
@@ -845,7 +844,7 @@ class SynthesizerTrnMs256NSFsid_nono(nn.Module):
             inter_channels, hidden_channels, 5, 1, 3, gin_channels=gin_channels
         )
         self.emb_g = nn.Embedding(self.spk_embed_dim, gin_channels)
-        print("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
+        logger.debug("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
 
     def remove_weight_norm(self):
         self.dec.remove_weight_norm()
@@ -951,7 +950,7 @@ class SynthesizerTrnMs768NSFsid_nono(nn.Module):
             inter_channels, hidden_channels, 5, 1, 3, gin_channels=gin_channels
         )
         self.emb_g = nn.Embedding(self.spk_embed_dim, gin_channels)
-        print("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
+        logger.debug("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
 
     def remove_weight_norm(self):
         self.dec.remove_weight_norm()

infer/lib/infer_pack/models_onnx.py

@@ -1,7 +1,6 @@
 import math
-import os
+import logging
-import pdb
+logger = logging.getLogger(__name__)
-from time import time as ttime
 
 import numpy as np
 import torch
@@ -620,7 +619,7 @@ class SynthesizerTrnMsNSFsidM(nn.Module):
         )
         self.emb_g = nn.Embedding(self.spk_embed_dim, gin_channels)
         self.speaker_map = None
-        print("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
+        logger.debug("gin_channels:", gin_channels, "self.spk_embed_dim:", self.spk_embed_dim)
 
     def remove_weight_norm(self):
         self.dec.remove_weight_norm()

infer/lib/infer_pack/onnx_inference.py

@@ -3,10 +3,13 @@ import numpy as np
 import onnxruntime
 import soundfile
 
+import logging
+logger = logging.getLogger(__name__)
+
 
 class ContentVec:
     def __init__(self, vec_path="pretrained/vec-768-layer-12.onnx", device=None):
-        print("Load model(s) from {}".format(vec_path))
+        logger.info("Load model(s) from {}".format(vec_path))
         if device == "cpu" or device is None:
             providers = ["CPUExecutionProvider"]
         elif device == "cuda":

infer/lib/rmvpe.py

@@ -7,6 +7,10 @@ import torch.nn.functional as F
 from librosa.util import normalize, pad_center, tiny
 from scipy.signal import get_window
 
+import logging
+
+logger = logging.getLogger(__name__)
+
 
 ###stft codes from https://github.com/pseeth/torch-stft/blob/master/torch_stft/util.py
 def window_sumsquare(
@@ -691,4 +695,4 @@ if __name__ == "__main__":
     # f0 = rmvpe.infer_from_audio(audio, thred=thred)
     # f0 = rmvpe.infer_from_audio(audio, thred=thred)
     t1 = ttime()
-    print(f0.shape, t1 - t0)
+    logger.info(f0.shape, t1 - t0)

infer/lib/train/data_utils.py

@@ -1,5 +1,7 @@
 import os
 import traceback
+import logging
+logger = logging.getLogger(__name__)
 
 import numpy as np
 import torch
@@ -110,7 +112,7 @@ class TextAudioLoaderMultiNSFsid(torch.utils.data.Dataset):
             try:
                 spec = torch.load(spec_filename)
             except:
-                print(spec_filename, traceback.format_exc())
+                logger.warn(spec_filename, traceback.format_exc())
                 spec = spectrogram_torch(
                     audio_norm,
                     self.filter_length,
@@ -302,7 +304,7 @@ class TextAudioLoader(torch.utils.data.Dataset):
             try:
                 spec = torch.load(spec_filename)
             except:
-                print(spec_filename, traceback.format_exc())
+                logger.warn(spec_filename, traceback.format_exc())
                 spec = spectrogram_torch(
                     audio_norm,
                     self.filter_length,

infer/lib/train/mel_processing.py

@@ -1,6 +1,8 @@
 import torch
 import torch.utils.data
 from librosa.filters import mel as librosa_mel_fn
+import logging
+logger = logging.getLogger(__name__)
 
 MAX_WAV_VALUE = 32768.0
 
@@ -51,9 +53,9 @@ def spectrogram_torch(y, n_fft, sampling_rate, hop_size, win_size, center=False)
     """
     # Validation
     if torch.min(y) < -1.07:
-        print("spectrogram_torch min value is ", torch.min(y))
+        logger.debug("min value is ", torch.min(y))
     if torch.max(y) > 1.07:
-        print("spectrogram_torch max value is ", torch.max(y))
+        logger.debug("max value is ", torch.max(y))
 
     # Window - Cache if needed
     global hann_window

infer/lib/train/utils.py

@@ -33,7 +33,7 @@ def load_checkpoint_d(checkpoint_path, combd, sbd, optimizer=None, load_opt=1):
             try:
                 new_state_dict[k] = saved_state_dict[k]
                 if saved_state_dict[k].shape != state_dict[k].shape:
-                    print(
+                    logger.warn(
                         "shape-%s-mismatch. need: %s, get: %s"
                         % (k, state_dict[k].shape, saved_state_dict[k].shape)
                     )  #
@@ -109,7 +109,7 @@ def load_checkpoint(checkpoint_path, model, optimizer=None, load_opt=1):
         try:
             new_state_dict[k] = saved_state_dict[k]
             if saved_state_dict[k].shape != state_dict[k].shape:
-                print(
+                logger.warn(
                     "shape-%s-mismatch|need-%s|get-%s"
                     % (k, state_dict[k].shape, saved_state_dict[k].shape)
                 )  #
@@ -207,7 +207,7 @@ def latest_checkpoint_path(dir_path, regex="G_*.pth"):
     f_list = glob.glob(os.path.join(dir_path, regex))
     f_list.sort(key=lambda f: int("".join(filter(str.isdigit, f))))
     x = f_list[-1]
-    print(x)
+    logger.debug(x)
     return x
 
 

infer/modules/train/train.py

@@ -1,5 +1,7 @@
 import os
 import sys
+import logging
+logger = logging.getLogger(__name__)
 
 now_dir = os.getcwd()
 sys.path.append(os.path.join(now_dir))
@@ -82,7 +84,7 @@ def main():
         n_gpus = 1
     if n_gpus < 1:
         # patch to unblock people without gpus. there is probably a better way.
-        print("NO GPU DETECTED: falling back to CPU - this may take a while")
+        logger.warn("NO GPU DETECTED: falling back to CPU - this may take a while")
         n_gpus = 1
     os.environ["MASTER_ADDR"] = "localhost"
     os.environ["MASTER_PORT"] = str(randint(20000, 55555))
@@ -209,7 +211,7 @@ def run(rank, n_gpus, hps):
         if hps.pretrainG != "":
             if rank == 0:
                 logger.info("loaded pretrained %s" % (hps.pretrainG))
-            print(
+            logger.info(
                 net_g.module.load_state_dict(
                     torch.load(hps.pretrainG, map_location="cpu")["model"]
                 )
@@ -217,7 +219,7 @@ def run(rank, n_gpus, hps):
         if hps.pretrainD != "":
             if rank == 0:
                 logger.info("loaded pretrained %s" % (hps.pretrainD))
-            print(
+            logger.info(
                 net_d.module.load_state_dict(
                     torch.load(hps.pretrainD, map_location="cpu")["model"]
                 )

infer/modules/uvr5/mdxnet.py

@@ -1,4 +1,6 @@
 import os
+import logging
+logger = logging.getLogger(__name__)
 
 import librosa
 import numpy as np
@@ -88,7 +90,7 @@ class Predictor:
     def __init__(self, args):
         import onnxruntime as ort
 
-        print(ort.get_available_providers())
+        logger.info(ort.get_available_providers())
         self.args = args
         self.model_ = get_models(
             device=cpu, dim_f=args.dim_f, dim_t=args.dim_t, n_fft=args.n_fft
@@ -101,7 +103,7 @@ class Predictor:
                 "CPUExecutionProvider",
             ],
         )
-        print("ONNX load done")
+        logger.info("ONNX load done")
 
     def demix(self, mix):
         samples = mix.shape[-1]

infer/modules/uvr5/modules.py

@@ -1,5 +1,7 @@
 import os
 import traceback
+import logging
+logger = logging.getLogger(__name__)
 
 import ffmpeg
 import torch
@@ -92,5 +94,5 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg, format
             traceback.print_exc()
         if torch.cuda.is_available():
             torch.cuda.empty_cache()
-            print("Executed torch.cuda.empty_cache()")
+            logger.info("Executed torch.cuda.empty_cache()")
     yield "\n".join(infos)

infer/modules/uvr5/preprocess.py

@@ -1,4 +1,6 @@
 import os
+import logging
+logger = logging.getLogger(__name__)
 
 import librosa
 import numpy as np
@@ -116,7 +118,7 @@ class AudioPre:
                 )
             else:
                 wav_instrument = spec_utils.cmb_spectrogram_to_wave(y_spec_m, self.mp)
-            print("%s instruments done" % name)
+            logger.info("%s instruments done" % name)
             if format in ["wav", "flac"]:
                 sf.write(
                     os.path.join(
@@ -150,7 +152,7 @@ class AudioPre:
                 )
             else:
                 wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp)
-            print("%s vocals done" % name)
+            logger.info("%s vocals done" % name)
             if format in ["wav", "flac"]:
                 sf.write(
                     os.path.join(
@@ -283,7 +285,7 @@ class AudioPreDeEcho:
                 )
             else:
                 wav_instrument = spec_utils.cmb_spectrogram_to_wave(y_spec_m, self.mp)
-            print("%s instruments done" % name)
+            logger.info("%s instruments done" % name)
             if format in ["wav", "flac"]:
                 sf.write(
                     os.path.join(
@@ -317,7 +319,7 @@ class AudioPreDeEcho:
                 )
             else:
                 wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp)
-            print("%s vocals done" % name)
+            logger.info("%s vocals done" % name)
             if format in ["wav", "flac"]:
                 sf.write(
                     os.path.join(

infer/modules/vc/modules.py

@@ -1,4 +1,6 @@
 import traceback
+import logging
+logger = logging.getLogger(__name__)
 
 import numpy as np
 import soundfile as sf
@@ -30,14 +32,7 @@ class VC:
         self.config = config
 
     def get_vc(self, sid, *to_return_protect):
-        person = f'{os.getenv("weight_root")}/{sid}'
+        logger.info("Get sid: " + sid)
-        print(f"Loading: {person}")
-
-        self.cpt = torch.load(person, map_location="cpu")
-        self.tgt_sr = self.cpt["config"][-1]
-        self.cpt["config"][-3] = self.cpt["weight"]["emb_g.weight"].shape[0]  # n_spk
-        self.if_f0 = self.cpt.get("f0", 1)
-        self.version = self.cpt.get("version", "v1")
 
         to_return_protect0 = {
             "visible": self.if_f0 != 0,
@@ -54,6 +49,57 @@ class VC:
             "__type__": "update",
         }
 
+        if not sid:
+            if self.hubert_model is not None:  # 考虑到轮询, 需要加个判断看是否 sid 是由有模型切换到无模型的
+                logger.info("Clean model cache")
+                del self.net_g, self.n_spk, self.vc, self.hubert_model, self.tgt_sr  # ,cpt
+                self.hubert_model = self.net_g = self.n_spk = self.vc = self.hubert_model = self.tgt_sr = None
+                if torch.cuda.is_available():
+                    torch.cuda.empty_cache()
+                ###楼下不这么折腾清理不干净
+                self.if_f0 = self.cpt.get("f0", 1)
+                self.version = self.cpt.get("version", "v1")
+                if self.version == "v1":
+                    if self.if_f0 == 1:
+                        self.net_g = SynthesizerTrnMs256NSFsid(
+                            *self.cpt["config"], is_half=self.config.is_half
+                        )
+                    else:
+                        self.net_g = SynthesizerTrnMs256NSFsid_nono(*self.cpt["config"])
+                elif self.version == "v2":
+                    if self.if_f0 == 1:
+                        self.net_g = SynthesizerTrnMs768NSFsid(
+                            *self.cpt["config"], is_half=self.config.is_half
+                        )
+                    else:
+                        self.net_g = SynthesizerTrnMs768NSFsid_nono(*self.cpt["config"])
+                del self.net_g, self.cpt
+                if torch.cuda.is_available():
+                    torch.cuda.empty_cache()
+            return (
+                {"visible": False, "__type__": "update"},
+                {
+                    "visible": True,
+                    "value": to_return_protect0,
+                    "__type__": "update",
+                },
+                {
+                    "visible": True,
+                    "value": to_return_protect1,
+                    "__type__": "update",
+                },
+                "",
+                "",
+            )
+        person = f'{os.getenv("weight_root")}/{sid}'
+        logger.info(f"Loading: {person}")
+
+        self.cpt = torch.load(person, map_location="cpu")
+        self.tgt_sr = self.cpt["config"][-1]
+        self.cpt["config"][-3] = self.cpt["weight"]["emb_g.weight"].shape[0]  # n_spk
+        self.if_f0 = self.cpt.get("f0", 1)
+        self.version = self.cpt.get("version", "v1")
+
         synthesizer_class = {
             ("v1", 1): SynthesizerTrnMs256NSFsid,
             ("v1", 0): SynthesizerTrnMs256NSFsid_nono,
@@ -77,7 +123,7 @@ class VC:
         self.pipeline = Pipeline(self.tgt_sr, self.config)
         n_spk = self.cpt["config"][-3]
         index = {"value": get_index_path_from_model(sid), "__type__": "update"}
-        print("Select index:", index["value"])
+        logger.info("Select index: " + index["value"])
 
         return (
             (
@@ -165,7 +211,7 @@ class VC:
             )
         except:
             info = traceback.format_exc()
-            print(info)
+            logger.warn(info)
             return info, (None, None)
 
     def vc_multi(

infer/modules/vc/pipeline.py

@@ -1,6 +1,9 @@
 import os
 import sys
 import traceback
+import logging
+logger = logging.getLogger(__name__)
+
 from functools import lru_cache
 from time import time as ttime
 
@@ -139,7 +142,7 @@ class Pipeline(object):
             if not hasattr(self, "model_rmvpe"):
                 from infer.lib.rmvpe import RMVPE
 
-                print(
+                logger.info(
                     "Loading rmvpe model,%s" % "%s/rmvpe.pt" % os.environ["rmvpe_root"]
                 )
                 self.model_rmvpe = RMVPE(
@@ -152,7 +155,7 @@ class Pipeline(object):
         if "privateuseone" in str(self.device):  # clean ortruntime memory
             del self.model_rmvpe.model
             del self.model_rmvpe
-            print("Cleaning ortruntime memory")
+            logger.info("Cleaning ortruntime memory")
 
         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()]))

tools/app.py

@@ -13,9 +13,10 @@ logging.getLogger("numba").setLevel(logging.WARNING)
 logging.getLogger("markdown_it").setLevel(logging.WARNING)
 logging.getLogger("urllib3").setLevel(logging.WARNING)
 logging.getLogger("matplotlib").setLevel(logging.WARNING)
+logger = logging.getLogger(__name__)
 
 i18n = I18nAuto()
-i18n.print()
+logger.info(i18n)
 
 load_dotenv()
 config = Config()

tools/calc_rvc_model_similarity.py

@@ -1,7 +1,8 @@
 # This code references https://huggingface.co/JosephusCheung/ASimilarityCalculatior/blob/main/qwerty.py
 # Fill in the path of the model to be queried and the root directory of the reference models, and this script will return the similarity between the model to be queried and all reference models.
 import os
-import sys
+import logging
+logger = logging.getLogger(__name__)
 
 import torch
 import torch.nn as nn
@@ -55,7 +56,7 @@ def main(path, root):
     torch.manual_seed(114514)
     model_a = torch.load(path, map_location="cpu")["weight"]
 
-    print("Query:\t\t%s\t%s" % (path, model_hash(path)))
+    logger.info("Query:\t\t%s\t%s" % (path, model_hash(path)))
 
     map_attn_a = {}
     map_rand_input = {}
@@ -82,7 +83,7 @@ def main(path, root):
             sim = torch.mean(torch.cosine_similarity(attn_a, attn_b))
             sims.append(sim)
 
-        print(
+        logger.info(
             "Reference:\t%s\t%s\t%s"
             % (path, model_hash(path), f"{torch.mean(torch.stack(sims)) * 1e2:.2f}%")
         )

tools/infer/infer-pm-index256.py

@@ -3,7 +3,8 @@
 对源特征进行检索
 """
 import os
-import pdb
+import logging
+logger = logging.getLogger(__name__)
 
 import parselmouth
 import torch
@@ -15,7 +16,6 @@ from time import time as ttime
 # import pyworld
 import librosa
 import numpy as np
-import scipy.signal as signal
 import soundfile as sf
 import torch.nn.functional as F
 from fairseq import checkpoint_utils
@@ -34,7 +34,7 @@ from scipy.io import wavfile
 
 device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 model_path = r"E:\codes\py39\vits_vc_gpu_train\assets\hubert\hubert_base.pt"  #
-print("Load model(s) from {}".format(model_path))
+logger.info("Load model(s) from {}".format(model_path))
 models, saved_cfg, task = checkpoint_utils.load_model_ensemble_and_task(
     [model_path],
     suffix="",
@@ -77,7 +77,7 @@ net_g = SynthesizerTrn256(
 # weights=torch.load("infer/ft-mi-freeze-vocoder_true_1k.pt")
 # weights=torch.load("infer/ft-mi-sim1k.pt")
 weights = torch.load("infer/ft-mi-no_opt-no_dropout.pt")
-print(net_g.load_state_dict(weights, strict=True))
+logger.debug(net_g.load_state_dict(weights, strict=True))
 
 net_g.eval().to(device)
 net_g.half()
@@ -198,4 +198,4 @@ for idx, name in enumerate(
     wavfile.write("ft-mi-no_opt-no_dropout-%s.wav" % name, 40000, audio)  ##
 
 
-print(ta0, ta1, ta2)  #
+logger.debug(ta0, ta1, ta2)  #

tools/infer/train-index-v2.py

@@ -3,6 +3,9 @@
 """
 import os
 import traceback
+import logging
+logger = logging.getLogger(__name__)
+
 from multiprocessing import cpu_count
 
 import faiss
@@ -23,11 +26,11 @@ big_npy = np.concatenate(npys, 0)
 big_npy_idx = np.arange(big_npy.shape[0])
 np.random.shuffle(big_npy_idx)
 big_npy = big_npy[big_npy_idx]
-print(big_npy.shape)  # (6196072, 192)#fp32#4.43G
+logger.debug(big_npy.shape)  # (6196072, 192)#fp32#4.43G
 if big_npy.shape[0] > 2e5:
     # if(1):
     info = "Trying doing kmeans %s shape to 10k centers." % big_npy.shape[0]
-    print(info)
+    logger.info(info)
     try:
         big_npy = (
             MiniBatchKMeans(
@@ -42,7 +45,7 @@ if big_npy.shape[0] > 2e5:
         )
     except:
         info = traceback.format_exc()
-        print(info)
+        logger.warn(info)
 
 np.save("tools/infer/big_src_feature_mi.npy", big_npy)
 
@@ -50,14 +53,14 @@ np.save("tools/infer/big_src_feature_mi.npy", big_npy)
 # big_npy=np.load("/bili-coeus/jupyter/jupyterhub-liujing04/vits_ch/inference_f0/big_src_feature_mi.npy")
 n_ivf = min(int(16 * np.sqrt(big_npy.shape[0])), big_npy.shape[0] // 39)
 index = faiss.index_factory(768, "IVF%s,Flat" % n_ivf)  # mi
-print("Training...")
+logger.info("Training...")
 index_ivf = faiss.extract_index_ivf(index)  #
 index_ivf.nprobe = 1
 index.train(big_npy)
 faiss.write_index(
     index, "tools/infer/trained_IVF%s_Flat_baseline_src_feat_v2.index" % (n_ivf)
 )
-print("Adding...")
+logger.info("Adding...")
 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])

tools/infer/train-index.py

@@ -2,6 +2,8 @@
 格式：直接cid为自带的index位；aid放不下了，通过字典来查，反正就5w个
 """
 import os
+import logging
+logger = logging.getLogger(__name__)
 
 import faiss
 import numpy as np
@@ -13,19 +15,19 @@ for name in sorted(list(os.listdir(inp_root))):
     phone = np.load("%s/%s" % (inp_root, name))
     npys.append(phone)
 big_npy = np.concatenate(npys, 0)
-print(big_npy.shape)  # (6196072, 192)#fp32#4.43G
+logger.debug(big_npy.shape)  # (6196072, 192)#fp32#4.43G
 np.save("infer/big_src_feature_mi.npy", big_npy)
 
 ##################train+add
 # big_npy=np.load("/bili-coeus/jupyter/jupyterhub-liujing04/vits_ch/inference_f0/big_src_feature_mi.npy")
-print(big_npy.shape)
+logger.debug(big_npy.shape)
 index = faiss.index_factory(256, "IVF512,Flat")  # mi
-print("Training...")
+logger.info("Training...")
 index_ivf = faiss.extract_index_ivf(index)  #
 index_ivf.nprobe = 9
 index.train(big_npy)
 faiss.write_index(index, "infer/trained_IVF512_Flat_mi_baseline_src_feat.index")
-print("Adding...")
+logger.info("Adding...")
 index.add(big_npy)
 faiss.write_index(index, "infer/added_IVF512_Flat_mi_baseline_src_feat.index")
 """

tools/rvc_for_realtime.py

@@ -1,6 +1,9 @@
 import os
 import sys
 import traceback
+import logging
+logger = logging.getLogger(__name__)
+
 from time import time as ttime
 
 import fairseq
@@ -67,7 +70,7 @@ class RVC:
             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")
+                logger.info("Index search enabled")
             self.index_path = index_path
             self.index_rate = index_rate
             models, _, _ = fairseq.checkpoint_utils.load_model_ensemble_and_task(
@@ -102,7 +105,7 @@ class RVC:
                 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))
+            logger.debug(self.net_g.load_state_dict(cpt["weight"], strict=False))
             self.net_g.eval().to(device)
             # print(2333333333,device,config.device,self.device)#net_g是device，hubert是config.device
             if config.is_half:
@@ -111,7 +114,7 @@ class RVC:
                 self.net_g = self.net_g.float()
             self.is_half = config.is_half
         except:
-            print(traceback.format_exc())
+            logger.warn(traceback.format_exc())
 
     def change_key(self, new_key):
         self.f0_up_key = new_key
@@ -120,7 +123,7 @@ class RVC:
         if new_index_rate != 0 and self.index_rate == 0:
             self.index = faiss.read_index(self.index_path)
             self.big_npy = self.index.reconstruct_n(0, self.index.ntotal)
-            print("Index search enabled")
+            logger.info("Index search enabled")
         self.index_rate = new_index_rate
 
     def get_f0_post(self, f0):
@@ -237,7 +240,7 @@ class RVC:
         if hasattr(self, "model_rmvpe") == False:
             from infer.lib.rmvpe import RMVPE
 
-            print("Loading rmvpe model")
+            logger.info("Loading rmvpe model")
             self.model_rmvpe = RMVPE(
                 # "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
@@ -295,10 +298,10 @@ class RVC:
                     + (1 - self.index_rate) * feats[0][-leng_replace_head:]
                 )
             else:
-                print("Index search FAILED or disabled")
+                logger.warn("Index search FAILED or disabled")
         except:
             traceback.print_exc()
-            print("Index search FAILED")
+            logger.warn("Index search FAILED")
         feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1)
         t3 = ttime()
         if self.if_f0 == 1:
@@ -338,5 +341,5 @@ class RVC:
                     .float()
                 )
         t5 = ttime()
-        print("Spent time: fea =", t2 - t1, ", index =", t3 - t2, ", f0 =", t4 - t3, ", model =", t5 - t4)
+        logger.info("Spent time: fea =", t2 - t1, ", index =", t3 - t2, ", f0 =", t4 - t3, ", model =", t5 - t4)
         return infered_audio