From e435b3bb8af788ba18ae7eba46c5dfd878d63e55 Mon Sep 17 00:00:00 2001 From: "github-actions[bot]" <41898282+github-actions[bot]@users.noreply.github.com> Date: Sun, 28 May 2023 16:06:11 +0000 Subject: [PATCH] Format code (#366) Co-authored-by: github-actions[bot] --- MDXNet.py | 207 +++++++++++++++++++++------------ infer-web.py | 82 +++++++------ infer_uvr5.py | 40 ++++--- uvr5_pack/lib_v5/layers_new.py | 35 +++--- uvr5_pack/lib_v5/nets_new.py | 42 ++++--- vc_infer_pipeline.py | 26 +++-- 6 files changed, 262 insertions(+), 170 deletions(-) diff --git a/MDXNet.py b/MDXNet.py index 6e6191b..7a1445a 100644 --- a/MDXNet.py +++ b/MDXNet.py @@ -1,5 +1,5 @@ import soundfile as sf -import torch,pdb,time,argparse,os,warnings,sys,librosa +import torch, pdb, time, argparse, os, warnings, sys, librosa import numpy as np import onnxruntime as ort from scipy.io.wavfile import write @@ -8,96 +8,133 @@ import torch import torch.nn as nn dim_c = 4 -class Conv_TDF_net_trim(): - def __init__(self, device, model_name, target_name, - L, dim_f, dim_t, n_fft, hop=1024): + + +class Conv_TDF_net_trim: + def __init__( + self, device, model_name, target_name, L, dim_f, dim_t, n_fft, hop=1024 + ): super(Conv_TDF_net_trim, self).__init__() self.dim_f = dim_f - self.dim_t = 2 ** dim_t + self.dim_t = 2**dim_t self.n_fft = n_fft self.hop = hop self.n_bins = self.n_fft // 2 + 1 self.chunk_size = hop * (self.dim_t - 1) - self.window = torch.hann_window(window_length=self.n_fft, periodic=True).to(device) + self.window = torch.hann_window(window_length=self.n_fft, periodic=True).to( + device + ) self.target_name = target_name - self.blender = 'blender' in model_name + self.blender = "blender" in model_name - out_c = dim_c * 4 if target_name == '*' else dim_c - self.freq_pad = torch.zeros([1, out_c, self.n_bins - self.dim_f, self.dim_t]).to(device) + out_c = dim_c * 4 if target_name == "*" else dim_c + self.freq_pad = torch.zeros( + [1, out_c, self.n_bins - self.dim_f, self.dim_t] + ).to(device) self.n = L // 2 def stft(self, x): x = x.reshape([-1, self.chunk_size]) - x = torch.stft(x, n_fft=self.n_fft, hop_length=self.hop, window=self.window, center=True, return_complex=True) + x = torch.stft( + x, + n_fft=self.n_fft, + hop_length=self.hop, + window=self.window, + center=True, + return_complex=True, + ) x = torch.view_as_real(x) x = x.permute([0, 3, 1, 2]) - x = x.reshape([-1, 2, 2, self.n_bins, self.dim_t]).reshape([-1, dim_c, self.n_bins, self.dim_t]) - return x[:, :, :self.dim_f] + x = x.reshape([-1, 2, 2, self.n_bins, self.dim_t]).reshape( + [-1, dim_c, self.n_bins, self.dim_t] + ) + return x[:, :, : self.dim_f] def istft(self, x, freq_pad=None): - freq_pad = self.freq_pad.repeat([x.shape[0], 1, 1, 1]) if freq_pad is None else freq_pad + freq_pad = ( + self.freq_pad.repeat([x.shape[0], 1, 1, 1]) + if freq_pad is None + else freq_pad + ) x = torch.cat([x, freq_pad], -2) - c = 4 * 2 if self.target_name == '*' else 2 - x = x.reshape([-1, c, 2, self.n_bins, self.dim_t]).reshape([-1, 2, self.n_bins, self.dim_t]) + c = 4 * 2 if self.target_name == "*" else 2 + x = x.reshape([-1, c, 2, self.n_bins, self.dim_t]).reshape( + [-1, 2, self.n_bins, self.dim_t] + ) x = x.permute([0, 2, 3, 1]) x = x.contiguous() x = torch.view_as_complex(x) - x = torch.istft(x, n_fft=self.n_fft, hop_length=self.hop, window=self.window, center=True) + x = torch.istft( + x, n_fft=self.n_fft, hop_length=self.hop, window=self.window, center=True + ) return x.reshape([-1, c, self.chunk_size]) + + def get_models(device, dim_f, dim_t, n_fft): return Conv_TDF_net_trim( device=device, - model_name='Conv-TDF', target_name='vocals', + model_name="Conv-TDF", + target_name="vocals", L=11, - dim_f=dim_f, dim_t=dim_t, - n_fft=n_fft + dim_f=dim_f, + dim_t=dim_t, + n_fft=n_fft, ) + warnings.filterwarnings("ignore") -cpu = torch.device('cpu') -device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu') +cpu = torch.device("cpu") +device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu") + class Predictor: - def __init__(self,args): - self.args=args - self.model_ = get_models(device=cpu, dim_f=args.dim_f, dim_t=args.dim_t, n_fft=args.n_fft) - self.model = ort.InferenceSession(os.path.join(args.onnx,self.model_.target_name+'.onnx'), providers=['CUDAExecutionProvider', 'CPUExecutionProvider']) - print('onnx load done') + def __init__(self, args): + self.args = args + self.model_ = get_models( + device=cpu, dim_f=args.dim_f, dim_t=args.dim_t, n_fft=args.n_fft + ) + self.model = ort.InferenceSession( + os.path.join(args.onnx, self.model_.target_name + ".onnx"), + providers=["CUDAExecutionProvider", "CPUExecutionProvider"], + ) + print("onnx load done") + def demix(self, mix): samples = mix.shape[-1] margin = self.args.margin - chunk_size = self.args.chunks*44100 - assert not margin == 0, 'margin cannot be zero!' + chunk_size = self.args.chunks * 44100 + assert not margin == 0, "margin cannot be zero!" if margin > chunk_size: margin = chunk_size segmented_mix = {} - + if self.args.chunks == 0 or samples < chunk_size: chunk_size = samples - + counter = -1 for skip in range(0, samples, chunk_size): - counter+=1 - + counter += 1 + s_margin = 0 if counter == 0 else margin - end = min(skip+chunk_size+margin, samples) + end = min(skip + chunk_size + margin, samples) - start = skip-s_margin + start = skip - s_margin - segmented_mix[skip] = mix[:,start:end].copy() + segmented_mix[skip] = mix[:, start:end].copy() if end == samples: break sources = self.demix_base(segmented_mix, margin_size=margin) - ''' + """ mix:(2,big_sample) segmented_mix:offset->(2,small_sample) sources:(1,2,big_sample) - ''' + """ return sources + def demix_base(self, mixes, margin_size): chunked_sources = [] progress_bar = tqdm(total=len(mixes)) @@ -106,15 +143,17 @@ class Predictor: cmix = mixes[mix] sources = [] n_sample = cmix.shape[1] - model=self.model_ - trim = model.n_fft//2 - gen_size = model.chunk_size-2*trim - pad = gen_size - n_sample%gen_size - mix_p = np.concatenate((np.zeros((2,trim)), cmix, np.zeros((2,pad)), np.zeros((2,trim))), 1) + model = self.model_ + trim = model.n_fft // 2 + gen_size = model.chunk_size - 2 * trim + pad = gen_size - n_sample % gen_size + mix_p = np.concatenate( + (np.zeros((2, trim)), cmix, np.zeros((2, pad)), np.zeros((2, trim))), 1 + ) mix_waves = [] i = 0 while i < n_sample + pad: - waves = np.array(mix_p[:, i:i+model.chunk_size]) + waves = np.array(mix_p[:, i : i + model.chunk_size]) mix_waves.append(waves) i += gen_size mix_waves = torch.tensor(mix_waves, dtype=torch.float32).to(cpu) @@ -122,68 +161,84 @@ class Predictor: _ort = self.model spek = model.stft(mix_waves) if self.args.denoise: - spec_pred = -_ort.run(None, {'input': -spek.cpu().numpy()})[0]*0.5+_ort.run(None, {'input': spek.cpu().numpy()})[0]*0.5 + spec_pred = ( + -_ort.run(None, {"input": -spek.cpu().numpy()})[0] * 0.5 + + _ort.run(None, {"input": spek.cpu().numpy()})[0] * 0.5 + ) tar_waves = model.istft(torch.tensor(spec_pred)) else: - tar_waves = model.istft(torch.tensor(_ort.run(None, {'input': spek.cpu().numpy()})[0])) - tar_signal = tar_waves[:,:,trim:-trim].transpose(0,1).reshape(2, -1).numpy()[:, :-pad] + tar_waves = model.istft( + torch.tensor(_ort.run(None, {"input": spek.cpu().numpy()})[0]) + ) + tar_signal = ( + tar_waves[:, :, trim:-trim] + .transpose(0, 1) + .reshape(2, -1) + .numpy()[:, :-pad] + ) start = 0 if mix == 0 else margin_size end = None if mix == list(mixes.keys())[::-1][0] else -margin_size if margin_size == 0: end = None - sources.append(tar_signal[:,start:end]) + sources.append(tar_signal[:, start:end]) progress_bar.update(1) - + chunked_sources.append(sources) _sources = np.concatenate(chunked_sources, axis=-1) # del self.model progress_bar.close() return _sources - def prediction(self, m,vocal_root,others_root,format): - os.makedirs(vocal_root,exist_ok=True) - os.makedirs(others_root,exist_ok=True) + + def prediction(self, m, vocal_root, others_root, format): + os.makedirs(vocal_root, exist_ok=True) + os.makedirs(others_root, exist_ok=True) basename = os.path.basename(m) mix, rate = librosa.load(m, mono=False, sr=44100) if mix.ndim == 1: - mix = np.asfortranarray([mix,mix]) + mix = np.asfortranarray([mix, mix]) mix = mix.T sources = self.demix(mix.T) - opt=sources[0].T - sf.write("%s/%s_main_vocal.%s"%(vocal_root,basename,format), mix-opt, rate) - sf.write("%s/%s_others.%s"%(others_root,basename,format), opt , rate) + opt = sources[0].T + sf.write( + "%s/%s_main_vocal.%s" % (vocal_root, basename, format), mix - opt, rate + ) + sf.write("%s/%s_others.%s" % (others_root, basename, format), opt, rate) -class MDXNetDereverb(): - def __init__(self,chunks): - self.onnx="uvr5_weights/onnx_dereverb_By_FoxJoy" - self.shifts=10#'Predict with randomised equivariant stabilisation' - self.mixing="min_mag"#['default','min_mag','max_mag'] - self.chunks=chunks - self.margin=44100 - self.dim_t=9 - self.dim_f=3072 - self.n_fft=6144 - self.denoise=True - self.pred=Predictor(self) - def _path_audio_(self,input,vocal_root,others_root,format): - self.pred.prediction(input,vocal_root,others_root,format) +class MDXNetDereverb: + def __init__(self, chunks): + self.onnx = "uvr5_weights/onnx_dereverb_By_FoxJoy" + self.shifts = 10 #'Predict with randomised equivariant stabilisation' + self.mixing = "min_mag" # ['default','min_mag','max_mag'] + self.chunks = chunks + self.margin = 44100 + self.dim_t = 9 + self.dim_f = 3072 + self.n_fft = 6144 + self.denoise = True + self.pred = Predictor(self) -if __name__ == '__main__': - dereverb=MDXNetDereverb(15) + def _path_audio_(self, input, vocal_root, others_root, format): + self.pred.prediction(input, vocal_root, others_root, format) + + +if __name__ == "__main__": + dereverb = MDXNetDereverb(15) from time import time as ttime - t0=ttime() + + t0 = ttime() dereverb._path_audio_( "雪雪伴奏对消HP5.wav", "vocal", "others", ) - t1=ttime() - print(t1-t0) + t1 = ttime() + print(t1 - t0) -''' +""" runtime\python.exe MDXNet.py @@ -195,4 +250,4 @@ runtime\python.exe MDXNet.py half15:0.7G->6.6G,22.69s fp32-15:0.7G->6.6G,20.85s -''' \ No newline at end of file +""" diff --git a/infer-web.py b/infer-web.py index 3e8eeb8..a4f7c5d 100644 --- a/infer-web.py +++ b/infer-web.py @@ -83,7 +83,7 @@ import gradio as gr import logging from vc_infer_pipeline import VC from config import Config -from infer_uvr5 import _audio_pre_,_audio_pre_new +from infer_uvr5 import _audio_pre_, _audio_pre_new from my_utils import load_audio from train.process_ckpt import show_info, change_info, merge, extract_small_model @@ -134,7 +134,7 @@ for root, dirs, files in os.walk(index_root, topdown=False): index_paths.append("%s/%s" % (root, name)) uvr5_names = [] for name in os.listdir(weight_uvr5_root): - if name.endswith(".pth")or "onnx"in name: + if name.endswith(".pth") or "onnx" in name: uvr5_names.append(name.replace(".pth", "")) @@ -151,7 +151,7 @@ def vc_single( filter_radius, resample_sr, rms_mix_rate, - protect + protect, ): # spk_item, input_audio0, vc_transform0,f0_file,f0method0 global tgt_sr, net_g, vc, hubert_model, version if input_audio_path is None: @@ -236,7 +236,7 @@ def vc_multi( resample_sr, rms_mix_rate, protect, - format1 + format1, ): try: dir_path = ( @@ -267,13 +267,15 @@ def vc_multi( filter_radius, resample_sr, rms_mix_rate, - protect + protect, ) if "Success" in info: try: tgt_sr, audio_opt = opt sf.write( - "%s/%s.%s" % (opt_root, os.path.basename(path),format1), audio_opt,tgt_sr + "%s/%s.%s" % (opt_root, os.path.basename(path), format1), + audio_opt, + tgt_sr, ) except: info += traceback.format_exc() @@ -284,7 +286,7 @@ def vc_multi( yield traceback.format_exc() -def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg,format0): +def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg, format0): infos = [] try: inp_root = inp_root.strip(" ").strip('"').strip("\n").strip('"').strip(" ") @@ -294,10 +296,10 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg,format0 save_root_ins = ( save_root_ins.strip(" ").strip('"').strip("\n").strip('"').strip(" ") ) - if(model_name=="onnx_dereverb_By_FoxJoy"): - pre_fun=MDXNetDereverb(15) + if model_name == "onnx_dereverb_By_FoxJoy": + pre_fun = MDXNetDereverb(15) else: - func=_audio_pre_ if "DeEcho"not in model_name else _audio_pre_new + func = _audio_pre_ if "DeEcho" not in model_name else _audio_pre_new pre_fun = func( agg=int(agg), model_path=os.path.join(weight_uvr5_root, model_name + ".pth"), @@ -319,7 +321,9 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg,format0 and info["streams"][0]["sample_rate"] == "44100" ): need_reformat = 0 - pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal,format0) + pre_fun._path_audio_( + inp_path, save_root_ins, save_root_vocal, format0 + ) done = 1 except: need_reformat = 1 @@ -333,7 +337,9 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg,format0 inp_path = tmp_path try: if done == 0: - pre_fun._path_audio_(inp_path, save_root_ins, save_root_vocal,format0) + pre_fun._path_audio_( + inp_path, save_root_ins, save_root_vocal, format0 + ) infos.append("%s->Success" % (os.path.basename(inp_path))) yield "\n".join(infos) except: @@ -346,7 +352,7 @@ def uvr(model_name, inp_root, save_root_vocal, paths, save_root_ins, agg,format0 yield "\n".join(infos) finally: try: - if (model_name == "onnx_dereverb_By_FoxJoy"): + if model_name == "onnx_dereverb_By_FoxJoy": del pre_fun.pred.model del pre_fun.pred.model_ else: @@ -804,7 +810,7 @@ def train_index(exp_dir1, version19): faiss.write_index( index, "%s/trained_IVF%s_Flat_nprobe_%s_%s_%s.index" - % (exp_dir, n_ivf, index_ivf.nprobe,exp_dir1, version19), + % (exp_dir, n_ivf, index_ivf.nprobe, exp_dir1, version19), ) # faiss.write_index(index, '%s/trained_IVF%s_Flat_FastScan_%s.index'%(exp_dir,n_ivf,version19)) infos.append("adding") @@ -815,11 +821,11 @@ def train_index(exp_dir1, version19): faiss.write_index( index, "%s/added_IVF%s_Flat_nprobe_%s_%s_%s.index" - % (exp_dir, n_ivf, index_ivf.nprobe,exp_dir1, version19), + % (exp_dir, n_ivf, index_ivf.nprobe, exp_dir1, version19), ) infos.append( "成功构建索引，added_IVF%s_Flat_nprobe_%s_%s_%s.index" - % (n_ivf, index_ivf.nprobe,exp_dir1, version19) + % (n_ivf, index_ivf.nprobe, exp_dir1, version19) ) # faiss.write_index(index, '%s/added_IVF%s_Flat_FastScan_%s.index'%(exp_dir,n_ivf,version19)) # infos.append("成功构建索引，added_IVF%s_Flat_FastScan_%s.index"%(n_ivf,version19)) @@ -1044,7 +1050,7 @@ def train1key( faiss.write_index( index, "%s/trained_IVF%s_Flat_nprobe_%s_%s_%s.index" - % (model_log_dir, n_ivf, index_ivf.nprobe,exp_dir1, version19), + % (model_log_dir, n_ivf, index_ivf.nprobe, exp_dir1, version19), ) yield get_info_str("adding index") batch_size_add = 8192 @@ -1053,11 +1059,11 @@ def train1key( faiss.write_index( index, "%s/added_IVF%s_Flat_nprobe_%s_%s_%s.index" - % (model_log_dir, n_ivf, index_ivf.nprobe,exp_dir1, version19), + % (model_log_dir, n_ivf, index_ivf.nprobe, exp_dir1, version19), ) yield get_info_str( "成功构建索引, added_IVF%s_Flat_nprobe_%s_%s_%s.index" - % (n_ivf, index_ivf.nprobe, exp_dir1,version19) + % (n_ivf, index_ivf.nprobe, exp_dir1, version19) ) yield get_info_str(i18n("全流程结束！")) @@ -1175,8 +1181,10 @@ with gr.Blocks() as app: value="E:\\codes\\py39\\test-20230416b\\todo-songs\\冬之花clip1.wav", ) 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","crepe"], + 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"], value="pm", interactive=True, ) @@ -1233,7 +1241,9 @@ with gr.Blocks() as app: 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"), + 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, @@ -1258,7 +1268,7 @@ with gr.Blocks() as app: filter_radius0, resample_sr0, rms_mix_rate0, - protect0 + protect0, ], [vc_output1, vc_output2], ) @@ -1273,8 +1283,10 @@ with gr.Blocks() as app: ) 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"], + 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"], value="pm", interactive=True, ) @@ -1328,7 +1340,9 @@ with gr.Blocks() as app: 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"), + 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, @@ -1342,9 +1356,9 @@ with gr.Blocks() as app: 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( + format1 = gr.Radio( label=i18n("导出文件格式"), - choices=["wav", "flac","mp3","m4a"], + choices=["wav", "flac", "mp3", "m4a"], value="flac", interactive=True, ) @@ -1367,7 +1381,7 @@ with gr.Blocks() as app: resample_sr1, rms_mix_rate1, protect1, - format1 + format1, ], [vc_output3], ) @@ -1412,10 +1426,12 @@ with gr.Blocks() as app: 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( + 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"], + choices=["wav", "flac", "mp3", "m4a"], value="flac", interactive=True, ) @@ -1430,7 +1446,7 @@ with gr.Blocks() as app: wav_inputs, opt_ins_root, agg, - format0 + format0, ], [vc_output4], ) diff --git a/infer_uvr5.py b/infer_uvr5.py index 614bba1..c188c1a 100644 --- a/infer_uvr5.py +++ b/infer_uvr5.py @@ -1,7 +1,9 @@ import os, sys, torch, warnings, pdb + now_dir = os.getcwd() sys.path.append(now_dir) from json import load as ll + warnings.filterwarnings("ignore") import librosa import importlib @@ -15,6 +17,7 @@ import soundfile as sf from uvr5_pack.lib_v5.nets_new import CascadedNet from uvr5_pack.lib_v5 import nets_61968KB as nets + class _audio_pre_: def __init__(self, agg, model_path, device, is_half): self.model_path = model_path @@ -41,7 +44,7 @@ class _audio_pre_: self.mp = mp self.model = model - def _path_audio_(self, music_file, ins_root=None, vocal_root=None,format="flac"): + def _path_audio_(self, music_file, ins_root=None, vocal_root=None, format="flac"): if ins_root is None and vocal_root is None: return "No save root." name = os.path.basename(music_file) @@ -122,9 +125,11 @@ class _audio_pre_: print("%s instruments done" % name) sf.write( os.path.join( - ins_root, "instrument_{}_{}.{}".format(name, self.data["agg"],format) + ins_root, + "instrument_{}_{}.{}".format(name, self.data["agg"], format), ), - (np.array(wav_instrument) * 32768).astype("int16"), self.mp.param["sr"], + (np.array(wav_instrument) * 32768).astype("int16"), + self.mp.param["sr"], ) # if vocal_root is not None: if self.data["high_end_process"].startswith("mirroring"): @@ -139,11 +144,13 @@ class _audio_pre_: print("%s vocals done" % name) sf.write( os.path.join( - vocal_root, "vocal_{}_{}.{}".format(name, self.data["agg"],format) + vocal_root, "vocal_{}_{}.{}".format(name, self.data["agg"], format) ), - (np.array(wav_vocals) * 32768).astype("int16"), self.mp.param["sr"], + (np.array(wav_vocals) * 32768).astype("int16"), + self.mp.param["sr"], ) + class _audio_pre_new: def __init__(self, agg, model_path, device, is_half): self.model_path = model_path @@ -157,9 +164,9 @@ class _audio_pre_new: "agg": agg, "high_end_process": "mirroring", } - mp=ModelParameters("uvr5_pack/lib_v5/modelparams/4band_v3.json") - nout=64 if "DeReverb"in model_path else 48 - model = CascadedNet(mp.param["bins"] * 2,nout) + mp = ModelParameters("uvr5_pack/lib_v5/modelparams/4band_v3.json") + nout = 64 if "DeReverb" in model_path else 48 + model = CascadedNet(mp.param["bins"] * 2, nout) cpk = torch.load(model_path, map_location="cpu") model.load_state_dict(cpk) model.eval() @@ -171,7 +178,9 @@ class _audio_pre_new: self.mp = mp self.model = model - def _path_audio_(self, music_file, vocal_root=None, ins_root=None,format="flac"):#3个VR模型vocal和ins是反的 + def _path_audio_( + self, music_file, vocal_root=None, ins_root=None, format="flac" + ): # 3个VR模型vocal和ins是反的 if ins_root is None and vocal_root is None: return "No save root." name = os.path.basename(music_file) @@ -252,9 +261,11 @@ class _audio_pre_new: print("%s instruments done" % name) sf.write( os.path.join( - ins_root, "main_vocal_{}_{}.{}".format(name, self.data["agg"],format) + ins_root, + "main_vocal_{}_{}.{}".format(name, self.data["agg"], format), ), - (np.array(wav_instrument) * 32768).astype("int16"),self.mp.param["sr"], + (np.array(wav_instrument) * 32768).astype("int16"), + self.mp.param["sr"], ) # if vocal_root is not None: if self.data["high_end_process"].startswith("mirroring"): @@ -269,9 +280,10 @@ class _audio_pre_new: print("%s vocals done" % name) sf.write( os.path.join( - vocal_root, "others_{}_{}.{}".format(name, self.data["agg"],format) + vocal_root, "others_{}_{}.{}".format(name, self.data["agg"], format) ), - (np.array(wav_vocals) * 32768).astype("int16"),self.mp.param["sr"], + (np.array(wav_vocals) * 32768).astype("int16"), + self.mp.param["sr"], ) @@ -283,7 +295,7 @@ if __name__ == "__main__": # model_path = "uvr5_weights/VR-DeEchoNormal.pth" model_path = "uvr5_weights/DeEchoNormal.pth" # pre_fun = _audio_pre_(model_path=model_path, device=device, is_half=True,agg=10) - pre_fun = _audio_pre_new(model_path=model_path, device=device, is_half=True,agg=10) + pre_fun = _audio_pre_new(model_path=model_path, device=device, is_half=True, agg=10) audio_path = "雪雪伴奏对消HP5.wav" save_path = "opt" pre_fun._path_audio_(audio_path, save_path, save_path) diff --git a/uvr5_pack/lib_v5/layers_new.py b/uvr5_pack/lib_v5/layers_new.py index 9c5ded3..2441f2d 100644 --- a/uvr5_pack/lib_v5/layers_new.py +++ b/uvr5_pack/lib_v5/layers_new.py @@ -4,27 +4,29 @@ import torch.nn.functional as F from uvr5_pack.lib_v5 import spec_utils -class Conv2DBNActiv(nn.Module): +class Conv2DBNActiv(nn.Module): def __init__(self, nin, nout, ksize=3, stride=1, pad=1, dilation=1, activ=nn.ReLU): super(Conv2DBNActiv, self).__init__() self.conv = nn.Sequential( nn.Conv2d( - nin, nout, + nin, + nout, kernel_size=ksize, stride=stride, padding=pad, dilation=dilation, - bias=False), + bias=False, + ), nn.BatchNorm2d(nout), - activ() + activ(), ) def __call__(self, x): return self.conv(x) -class Encoder(nn.Module): +class Encoder(nn.Module): def __init__(self, nin, nout, ksize=3, stride=1, pad=1, activ=nn.LeakyReLU): super(Encoder, self).__init__() self.conv1 = Conv2DBNActiv(nin, nout, ksize, stride, pad, activ=activ) @@ -38,15 +40,16 @@ class Encoder(nn.Module): class Decoder(nn.Module): - - def __init__(self, nin, nout, ksize=3, stride=1, pad=1, activ=nn.ReLU, dropout=False): + def __init__( + self, nin, nout, ksize=3, stride=1, pad=1, activ=nn.ReLU, dropout=False + ): super(Decoder, self).__init__() self.conv1 = Conv2DBNActiv(nin, nout, ksize, 1, pad, activ=activ) # self.conv2 = Conv2DBNActiv(nout, nout, ksize, 1, pad, activ=activ) self.dropout = nn.Dropout2d(0.1) if dropout else None def __call__(self, x, skip=None): - x = F.interpolate(x, scale_factor=2, mode='bilinear', align_corners=True) + x = F.interpolate(x, scale_factor=2, mode="bilinear", align_corners=True) if skip is not None: skip = spec_utils.crop_center(skip, x) @@ -62,12 +65,11 @@ class Decoder(nn.Module): class ASPPModule(nn.Module): - def __init__(self, nin, nout, dilations=(4, 8, 12), activ=nn.ReLU, dropout=False): super(ASPPModule, self).__init__() self.conv1 = nn.Sequential( nn.AdaptiveAvgPool2d((1, None)), - Conv2DBNActiv(nin, nout, 1, 1, 0, activ=activ) + Conv2DBNActiv(nin, nout, 1, 1, 0, activ=activ), ) self.conv2 = Conv2DBNActiv(nin, nout, 1, 1, 0, activ=activ) self.conv3 = Conv2DBNActiv( @@ -84,7 +86,9 @@ class ASPPModule(nn.Module): def forward(self, x): _, _, h, w = x.size() - feat1 = F.interpolate(self.conv1(x), size=(h, w), mode='bilinear', align_corners=True) + feat1 = F.interpolate( + self.conv1(x), size=(h, w), mode="bilinear", align_corners=True + ) feat2 = self.conv2(x) feat3 = self.conv3(x) feat4 = self.conv4(x) @@ -99,19 +103,14 @@ class ASPPModule(nn.Module): class LSTMModule(nn.Module): - def __init__(self, nin_conv, nin_lstm, nout_lstm): super(LSTMModule, self).__init__() self.conv = Conv2DBNActiv(nin_conv, 1, 1, 1, 0) self.lstm = nn.LSTM( - input_size=nin_lstm, - hidden_size=nout_lstm // 2, - bidirectional=True + input_size=nin_lstm, hidden_size=nout_lstm // 2, bidirectional=True ) self.dense = nn.Sequential( - nn.Linear(nout_lstm, nin_lstm), - nn.BatchNorm1d(nin_lstm), - nn.ReLU() + nn.Linear(nout_lstm, nin_lstm), nn.BatchNorm1d(nin_lstm), nn.ReLU() ) def forward(self, x): diff --git a/uvr5_pack/lib_v5/nets_new.py b/uvr5_pack/lib_v5/nets_new.py index 1c50bf8..c9898f6 100644 --- a/uvr5_pack/lib_v5/nets_new.py +++ b/uvr5_pack/lib_v5/nets_new.py @@ -3,9 +3,11 @@ from torch import nn import torch.nn.functional as F from uvr5_pack.lib_v5 import layers_new as layers -class BaseNet(nn.Module): - def __init__(self, nin, nout, nin_lstm, nout_lstm, dilations=((4, 2), (8, 4), (12, 6))): +class BaseNet(nn.Module): + def __init__( + self, nin, nout, nin_lstm, nout_lstm, dilations=((4, 2), (8, 4), (12, 6)) + ): super(BaseNet, self).__init__() self.enc1 = layers.Conv2DBNActiv(nin, nout, 3, 1, 1) self.enc2 = layers.Encoder(nout, nout * 2, 3, 2, 1) @@ -38,8 +40,8 @@ class BaseNet(nn.Module): return h -class CascadedNet(nn.Module): +class CascadedNet(nn.Module): def __init__(self, n_fft, nout=32, nout_lstm=128): super(CascadedNet, self).__init__() @@ -50,24 +52,30 @@ class CascadedNet(nn.Module): self.stg1_low_band_net = nn.Sequential( BaseNet(2, nout // 2, self.nin_lstm // 2, nout_lstm), - layers.Conv2DBNActiv(nout // 2, nout // 4, 1, 1, 0) - ) - - self.stg1_high_band_net = BaseNet(2, nout // 4, self.nin_lstm // 2, nout_lstm // 2) + layers.Conv2DBNActiv(nout // 2, nout // 4, 1, 1, 0), + ) + + self.stg1_high_band_net = BaseNet( + 2, nout // 4, self.nin_lstm // 2, nout_lstm // 2 + ) self.stg2_low_band_net = nn.Sequential( BaseNet(nout // 4 + 2, nout, self.nin_lstm // 2, nout_lstm), - layers.Conv2DBNActiv(nout, nout // 2, 1, 1, 0) - ) - self.stg2_high_band_net = BaseNet(nout // 4 + 2, nout // 2, self.nin_lstm // 2, nout_lstm // 2) + layers.Conv2DBNActiv(nout, nout // 2, 1, 1, 0), + ) + self.stg2_high_band_net = BaseNet( + nout // 4 + 2, nout // 2, self.nin_lstm // 2, nout_lstm // 2 + ) - self.stg3_full_band_net = BaseNet(3 * nout // 4 + 2, nout, self.nin_lstm, nout_lstm) + self.stg3_full_band_net = BaseNet( + 3 * nout // 4 + 2, nout, self.nin_lstm, nout_lstm + ) self.out = nn.Conv2d(nout, 2, 1, bias=False) self.aux_out = nn.Conv2d(3 * nout // 4, 2, 1, bias=False) def forward(self, x): - x = x[:, :, :self.max_bin] + x = x[:, :, : self.max_bin] bandw = x.size()[2] // 2 l1_in = x[:, :, :bandw] @@ -89,7 +97,7 @@ class CascadedNet(nn.Module): mask = F.pad( input=mask, pad=(0, 0, 0, self.output_bin - mask.size()[2]), - mode='replicate' + mode="replicate", ) if self.training: @@ -98,7 +106,7 @@ class CascadedNet(nn.Module): aux = F.pad( input=aux, pad=(0, 0, 0, self.output_bin - aux.size()[2]), - mode='replicate' + mode="replicate", ) return mask, aux else: @@ -108,17 +116,17 @@ class CascadedNet(nn.Module): mask = self.forward(x) if self.offset > 0: - mask = mask[:, :, :, self.offset:-self.offset] + mask = mask[:, :, :, self.offset : -self.offset] assert mask.size()[3] > 0 return mask - def predict(self, x,aggressiveness=None): + def predict(self, x, aggressiveness=None): mask = self.forward(x) pred_mag = x * mask if self.offset > 0: - pred_mag = pred_mag[:, :, :, self.offset:-self.offset] + pred_mag = pred_mag[:, :, :, self.offset : -self.offset] assert pred_mag.size()[3] > 0 return pred_mag diff --git a/vc_infer_pipeline.py b/vc_infer_pipeline.py index f45d4c8..9aaa8ee 100644 --- a/vc_infer_pipeline.py +++ b/vc_infer_pipeline.py @@ -2,7 +2,7 @@ import numpy as np, parselmouth, torch, pdb from time import time as ttime import torch.nn.functional as F import scipy.signal as signal -import pyworld, os, traceback, faiss, librosa,torchcrepe +import pyworld, os, traceback, faiss, librosa, torchcrepe from scipy import signal from functools import lru_cache @@ -162,7 +162,7 @@ class VC(object): big_npy, index_rate, version, - protect + protect, ): # ,file_index,file_big_npy feats = torch.from_numpy(audio0) if self.is_half: @@ -184,8 +184,8 @@ class VC(object): with torch.no_grad(): logits = model.extract_features(**inputs) feats = model.final_proj(logits[0]) if version == "v1" else logits[0] - if(protect<0.5): - feats0=feats.clone() + if protect < 0.5: + feats0 = feats.clone() if ( isinstance(index, type(None)) == False and isinstance(big_npy, type(None)) == False @@ -211,8 +211,10 @@ class VC(object): ) feats = F.interpolate(feats.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1) - if(protect<0.5): - feats0 = F.interpolate(feats0.permute(0, 2, 1), scale_factor=2).permute(0, 2, 1) + if protect < 0.5: + feats0 = F.interpolate(feats0.permute(0, 2, 1), scale_factor=2).permute( + 0, 2, 1 + ) t1 = ttime() p_len = audio0.shape[0] // self.window if feats.shape[1] < p_len: @@ -221,13 +223,13 @@ class VC(object): pitch = pitch[:, :p_len] pitchf = pitchf[:, :p_len] - if(protect<0.5): + if protect < 0.5: pitchff = pitchf.clone() pitchff[pitchf > 0] = 1 pitchff[pitchf < 1] = protect pitchff = pitchff.unsqueeze(-1) feats = feats * pitchff + feats0 * (1 - pitchff) - feats=feats.to(feats0.dtype) + feats = feats.to(feats0.dtype) p_len = torch.tensor([p_len], device=self.device).long() with torch.no_grad(): if pitch != None and pitchf != None: @@ -356,7 +358,7 @@ class VC(object): big_npy, index_rate, version, - protect + protect, )[self.t_pad_tgt : -self.t_pad_tgt] ) else: @@ -373,7 +375,7 @@ class VC(object): big_npy, index_rate, version, - protect + protect, )[self.t_pad_tgt : -self.t_pad_tgt] ) s = t @@ -391,7 +393,7 @@ class VC(object): big_npy, index_rate, version, - protect + protect, )[self.t_pad_tgt : -self.t_pad_tgt] ) else: @@ -408,7 +410,7 @@ class VC(object): big_npy, index_rate, version, - protect + protect, )[self.t_pad_tgt : -self.t_pad_tgt] ) audio_opt = np.concatenate(audio_opt)