bacow/GPT-SoVITS
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Refactor: Format Code with Ruff and Update Deprecated G2PW Link (#2255)

Browse Source 
* ruff check --fix

* ruff format --line-length 120 --target-version py39

* Change the link for G2PW Model

* update pytorch version and colab
This commit is contained in:
XXXXRT666
2025-04-07 09:42:47 +01:00
committed by GitHub
parent 9da7e17efe
commit 53cac93589
132 changed files with 8185 additions and 6648 deletions
Download Patch File Download Diff File Expand all files Collapse all files

100
tools/uvr5/vr.py
View File

@@ -1,6 +1,8 @@
import os,sys
import os
parent_directory = os.path.dirname(os.path.abspath(__file__))
import logging,pdb
import logging
logger = logging.getLogger(__name__)
import librosa
@@ -27,7 +29,7 @@ class AudioPre:
"agg": agg,
"high_end_process": "mirroring",
}
mp = ModelParameters("%s/lib/lib_v5/modelparams/4band_v2.json"%parent_directory)
mp = ModelParameters("%s/lib/lib_v5/modelparams/4band_v2.json" % parent_directory)
model = Nets.CascadedASPPNet(mp.param["bins"] * 2)
cpk = torch.load(model_path, map_location="cpu")
model.load_state_dict(cpk)
@@ -40,9 +42,7 @@ class AudioPre:
self.mp = mp
self.model = model
def _path_audio_(
self, music_file, ins_root=None, vocal_root=None, format="flac", is_hp3=False
):
def _path_audio_(self, music_file, ins_root=None, vocal_root=None, format="flac", is_hp3=False):
if ins_root is None and vocal_root is None:
return "No save root."
name = os.path.basename(music_file)
@@ -61,19 +61,19 @@ class AudioPre:
_,
) = librosa.core.load( # 理论上librosa读取可能对某些音频有bug，应该上ffmpeg读取，但是太麻烦了弃坑
music_file,
sr = bp["sr"],
mono = False,
dtype = np.float32,
res_type = bp["res_type"],
sr=bp["sr"],
mono=False,
dtype=np.float32,
res_type=bp["res_type"],
)
if X_wave[d].ndim == 1:
X_wave[d] = np.asfortranarray([X_wave[d], X_wave[d]])
else: # lower bands
X_wave[d] = librosa.core.resample(
X_wave[d + 1],
orig_sr = self.mp.param["band"][d + 1]["sr"],
target_sr = bp["sr"],
res_type = bp["res_type"],
orig_sr=self.mp.param["band"][d + 1]["sr"],
target_sr=bp["sr"],
res_type=bp["res_type"],
)
# Stft of wave source
X_spec_s[d] = spec_utils.wave_to_spectrogram_mt(
@@ -89,9 +89,7 @@ class AudioPre:
input_high_end_h = (bp["n_fft"] // 2 - bp["crop_stop"]) + (
self.mp.param["pre_filter_stop"] - self.mp.param["pre_filter_start"]
)
input_high_end = X_spec_s[d][
:, bp["n_fft"] // 2 - input_high_end_h : bp["n_fft"] // 2, :
]
input_high_end = X_spec_s[d][:, bp["n_fft"] // 2 - input_high_end_h : bp["n_fft"] // 2, :]
X_spec_m = spec_utils.combine_spectrograms(X_spec_s, self.mp)
aggresive_set = float(self.data["agg"] / 100)
@@ -100,9 +98,7 @@ class AudioPre:
"split_bin": self.mp.param["band"][1]["crop_stop"],
}
with torch.no_grad():
pred, X_mag, X_phase = inference(
X_spec_m, self.device, self.model, aggressiveness, self.data
)
pred, X_mag, X_phase = inference(X_spec_m, self.device, self.model, aggressiveness, self.data)
# Postprocess
if self.data["postprocess"]:
pred_inv = np.clip(X_mag - pred, 0, np.inf)
@@ -111,13 +107,11 @@ class AudioPre:
v_spec_m = X_spec_m - y_spec_m
if is_hp3 == True:
ins_root,vocal_root = vocal_root,ins_root
ins_root, vocal_root = vocal_root, ins_root
if ins_root is not None:
if self.data["high_end_process"].startswith("mirroring"):
input_high_end_ = spec_utils.mirroring(
self.data["high_end_process"], y_spec_m, input_high_end, self.mp
)
input_high_end_ = spec_utils.mirroring(self.data["high_end_process"], y_spec_m, input_high_end, self.mp)
wav_instrument = spec_utils.cmb_spectrogram_to_wave(
y_spec_m, self.mp, input_high_end_h, input_high_end_
)
@@ -138,9 +132,7 @@ class AudioPre:
self.mp.param["sr"],
) #
else:
path = os.path.join(
ins_root, head + "{}_{}.wav".format(name, self.data["agg"])
)
path = os.path.join(ins_root, head + "{}_{}.wav".format(name, self.data["agg"]))
sf.write(
path,
(np.array(wav_instrument) * 32768).astype("int16"),
@@ -160,12 +152,8 @@ class AudioPre:
else:
head = "vocal_"
if self.data["high_end_process"].startswith("mirroring"):
input_high_end_ = spec_utils.mirroring(
self.data["high_end_process"], v_spec_m, input_high_end, self.mp
)
wav_vocals = spec_utils.cmb_spectrogram_to_wave(
v_spec_m, self.mp, input_high_end_h, input_high_end_
)
input_high_end_ = spec_utils.mirroring(self.data["high_end_process"], v_spec_m, input_high_end, self.mp)
wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp, input_high_end_h, input_high_end_)
else:
wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp)
logger.info("%s vocals done" % name)
@@ -179,9 +167,7 @@ class AudioPre:
self.mp.param["sr"],
)
else:
path = os.path.join(
vocal_root, head + "{}_{}.wav".format(name, self.data["agg"])
)
path = os.path.join(vocal_root, head + "{}_{}.wav".format(name, self.data["agg"]))
sf.write(
path,
(np.array(wav_vocals) * 32768).astype("int16"),
@@ -210,7 +196,7 @@ class AudioPreDeEcho:
"agg": agg,
"high_end_process": "mirroring",
}
mp = ModelParameters("%s/lib/lib_v5/modelparams/4band_v3.json"%parent_directory)
mp = ModelParameters("%s/lib/lib_v5/modelparams/4band_v3.json" % parent_directory)
nout = 64 if "DeReverb" in model_path else 48
model = CascadedNet(mp.param["bins"] * 2, nout)
cpk = torch.load(model_path, map_location="cpu")
@@ -245,19 +231,19 @@ class AudioPreDeEcho:
_,
) = librosa.core.load( # 理论上librosa读取可能对某些音频有bug，应该上ffmpeg读取，但是太麻烦了弃坑
music_file,
sr = bp["sr"],
mono = False,
dtype = np.float32,
res_type = bp["res_type"],
sr=bp["sr"],
mono=False,
dtype=np.float32,
res_type=bp["res_type"],
)
if X_wave[d].ndim == 1:
X_wave[d] = np.asfortranarray([X_wave[d], X_wave[d]])
else: # lower bands
X_wave[d] = librosa.core.resample(
X_wave[d + 1],
orig_sr = self.mp.param["band"][d + 1]["sr"],
target_sr = bp["sr"],
res_type = bp["res_type"],
orig_sr=self.mp.param["band"][d + 1]["sr"],
target_sr=bp["sr"],
res_type=bp["res_type"],
)
# Stft of wave source
X_spec_s[d] = spec_utils.wave_to_spectrogram_mt(
@@ -273,9 +259,7 @@ class AudioPreDeEcho:
input_high_end_h = (bp["n_fft"] // 2 - bp["crop_stop"]) + (
self.mp.param["pre_filter_stop"] - self.mp.param["pre_filter_start"]
)
input_high_end = X_spec_s[d][
:, bp["n_fft"] // 2 - input_high_end_h : bp["n_fft"] // 2, :
]
input_high_end = X_spec_s[d][:, bp["n_fft"] // 2 - input_high_end_h : bp["n_fft"] // 2, :]
X_spec_m = spec_utils.combine_spectrograms(X_spec_s, self.mp)
aggresive_set = float(self.data["agg"] / 100)
@@ -284,9 +268,7 @@ class AudioPreDeEcho:
"split_bin": self.mp.param["band"][1]["crop_stop"],
}
with torch.no_grad():
pred, X_mag, X_phase = inference(
X_spec_m, self.device, self.model, aggressiveness, self.data
)
pred, X_mag, X_phase = inference(X_spec_m, self.device, self.model, aggressiveness, self.data)
# Postprocess
if self.data["postprocess"]:
pred_inv = np.clip(X_mag - pred, 0, np.inf)
@@ -296,9 +278,7 @@ class AudioPreDeEcho:
if ins_root is not None:
if self.data["high_end_process"].startswith("mirroring"):
input_high_end_ = spec_utils.mirroring(
self.data["high_end_process"], y_spec_m, input_high_end, self.mp
)
input_high_end_ = spec_utils.mirroring(self.data["high_end_process"], y_spec_m, input_high_end, self.mp)
wav_instrument = spec_utils.cmb_spectrogram_to_wave(
y_spec_m, self.mp, input_high_end_h, input_high_end_
)
@@ -315,9 +295,7 @@ class AudioPreDeEcho:
self.mp.param["sr"],
) #
else:
path = os.path.join(
ins_root, "vocal_{}_{}.wav".format(name, self.data["agg"])
)
path = os.path.join(ins_root, "vocal_{}_{}.wav".format(name, self.data["agg"]))
sf.write(
path,
(np.array(wav_instrument) * 32768).astype("int16"),
@@ -333,12 +311,8 @@ class AudioPreDeEcho:
pass
if vocal_root is not None:
if self.data["high_end_process"].startswith("mirroring"):
input_high_end_ = spec_utils.mirroring(
self.data["high_end_process"], v_spec_m, input_high_end, self.mp
)
wav_vocals = spec_utils.cmb_spectrogram_to_wave(
v_spec_m, self.mp, input_high_end_h, input_high_end_
)
input_high_end_ = spec_utils.mirroring(self.data["high_end_process"], v_spec_m, input_high_end, self.mp)
wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp, input_high_end_h, input_high_end_)
else:
wav_vocals = spec_utils.cmb_spectrogram_to_wave(v_spec_m, self.mp)
logger.info("%s vocals done" % name)
@@ -352,9 +326,7 @@ class AudioPreDeEcho:
self.mp.param["sr"],
)
else:
path = os.path.join(
vocal_root, "instrument_{}_{}.wav".format(name, self.data["agg"])
)
path = os.path.join(vocal_root, "instrument_{}_{}.wav".format(name, self.data["agg"]))
sf.write(
path,
(np.array(wav_vocals) * 32768).astype("int16"),