Mangio-RVC-Fork with v2 Support! 💻
A fork of an easy-to-use SVC framework based on VITS with top1 retrieval 💯. In general, this fork provides a CLI interface in addition. And also gives you more f0 methods to use, as well as a personlized 'hybrid' f0 estimation method using nanmedian.
From 0e5d725b340c9d5683c855e8d015da73d154498a Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Fri, 21 Jul 2023 21:40:27 +0700
Subject: [PATCH 03/14] Update infer-web.py
fixed gpu unload traceback :)
---
infer-web.py | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/infer-web.py b/infer-web.py
index 02b306b..7811749 100644
--- a/infer-web.py
+++ b/infer-web.py
@@ -518,7 +518,7 @@ def get_vc(sid, to_return_protect0, to_return_protect1):
if torch.cuda.is_available():
torch.cuda.empty_cache()
cpt = None
- return {"visible": False, "__type__": "update"}
+ return ({"visible": False, "__type__": "update"}, {"visible": False, "__type__": "update"}, {"visible": False, "__type__": "update"})
person = "%s/%s" % (weight_root, sid)
print("loading %s" % person)
cpt = torch.load(person, map_location="cpu")
@@ -3695,4 +3695,4 @@ with gr.Blocks(theme='HaleyCH/HaleyCH_Theme') as app:
)
#endregion
-'''
\ No newline at end of file
+'''
From 72539f737ab560f9c7acf9a7529a164166dd540f Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Fri, 21 Jul 2023 21:42:07 +0700
Subject: [PATCH 04/14] Update README.md
---
README.md | 23 +----------------------
1 file changed, 1 insertion(+), 22 deletions(-)
diff --git a/README.md b/README.md
index 1457ad3..b0c57fc 100644
--- a/README.md
+++ b/README.md
@@ -11,27 +11,6 @@ os.path.abspath(os.getcwd())
- Auto-open TensorBoard localhost URL when `tensor-launch.py` is executed
- RMVPE implemented in both inferencing and training (the one in `Training` tab doesn't work properly though, requires some additional work to do)
-# Known bugs:
-
-- `Unload voice to save GPU memory` button causes a Traceback:
-
-```python
-Traceback (most recent call last):
- File "D:\RVC-beta-v2-0618\runtime\lib\site-packages\gradio\routes.py", line 437, in run_predict
- output = await app.get_blocks().process_api(
- File "D:\RVC-beta-v2-0618\runtime\lib\site-packages\gradio\blocks.py", line 1349, in process_api
- data = self.postprocess_data(fn_index, result["prediction"], state)
- File "D:\RVC-beta-v2-0618\runtime\lib\site-packages\gradio\blocks.py", line 1252, in postprocess_data
- self.validate_outputs(fn_index, predictions) # type: ignore
- File "D:\RVC-beta-v2-0618\runtime\lib\site-packages\gradio\blocks.py", line 1227, in validate_outputs
- raise ValueError(
-ValueError: An event handler (get_vc) didn't receive enough output values (needed: 3, received: 1).
-Wanted outputs:
- [slider, slider, slider]
-Received outputs:
- [{'visible': False, '__type__': 'update'}]
-```
-
## Installation:
1. Simply either extract directly or use git clone
@@ -68,7 +47,7 @@ Received outputs:
### Current Todo-list:
-- [ ] Fix `Unload voice to save GPU memory` button Traceback
+- [x] Fix `Unload voice to save GPU memory` button Traceback
- [ ] Add Accordions so people with Firefox browser get a much more compact GUI rather than [This](https://github.com/RVC-Project/Retrieval-based-Voice-Conversion-WebUI/assets/79400603/67e0cc08-82a2-4dc3-86cf-e23d1dcad9f8).
- [ ] Fix weird way Median Filtering value inputted in a slider is utilized
- [ ] Replace regular refresh buttons with these tiny ones from [AUTOMATIC'S1111 Stable DIffusion](https://github.com/AUTOMATIC1111/stable-diffusion-webui)
From 4abc7f23ead8ab5d3291b91337c2360c79d8383f Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Fri, 21 Jul 2023 22:16:29 +0700
Subject: [PATCH 05/14] Update README.md
---
README.md | 37 +++++++++++++++++++++++++++++++++++++
1 file changed, 37 insertions(+)
diff --git a/README.md b/README.md
index b0c57fc..83710c1 100644
--- a/README.md
+++ b/README.md
@@ -56,6 +56,43 @@ os.path.abspath(os.getcwd())
- [ ] Implement Praat in the GUI for f0 curve file manipulation and easier usage
+
+# Screenshots:
+
+- ## Inference Tab:
+
+
+
+- ## UVR Tab:
+
+
+
+- ## Training Tab:
+
+
+
+- ## Ckpt-Processing Tab:
+
+
+
+The rest of the tabs are left untouched code-wise.
+
+
+
+# Formant Shift:
+
+
+
+- ### Click `Apply` button every time you change the values for inferencing.
+
+- ### As the name implies, you can only use `wav` files so far, also it is very slow, so be patient.
+
+- ### If you added a new `preset.txt` in the `\formantshiftcfg\` folder, click button with refresh emoji
+
+- ### If the preset you selected somehow got edited, by pressing refresh emoji button you'll update values, by grabbing them from the file
+
+
+
Mangio-RVC-Fork with v2 Support! 💻
A fork of an easy-to-use SVC framework based on VITS with top1 retrieval 💯. In general, this fork provides a CLI interface in addition. And also gives you more f0 methods to use, as well as a personlized 'hybrid' f0 estimation method using nanmedian.
From fb726751a734821b9c1d5ec055873f1e3a353c33 Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Sat, 22 Jul 2023 20:30:54 +0700
Subject: [PATCH 06/14] Add files via upload
forgot to upload it hehe
---
tensor-launch.py | 16 ++++++++++++++++
1 file changed, 16 insertions(+)
create mode 100644 tensor-launch.py
diff --git a/tensor-launch.py b/tensor-launch.py
new file mode 100644
index 0000000..511335a
--- /dev/null
+++ b/tensor-launch.py
@@ -0,0 +1,16 @@
+import threading
+import time
+from tensorboard import program
+import webbrowser
+
+log_path = "logs"
+
+if __name__ == "__main__":
+ tb = program.TensorBoard()
+ tb.configure(argv=[None, '--logdir', log_path])
+ url = tb.launch()
+ webbrowser.open_new(url)
+ print(f'Tensorboard can be accessed at: {url}')
+
+ while True:
+ time.sleep(600) # Keep the main thread running
\ No newline at end of file
From 21dba2f37bcdd962258cb7a12e19591849e09422 Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Sun, 23 Jul 2023 03:34:59 +0700
Subject: [PATCH 07/14] Add files via upload
fixed broken training due to missing arguments of load_audio function
---
extract_f0_print.py | 16 +++++++++++++++-
trainset_preprocess_pipeline_print.py | 9 ++++++++-
2 files changed, 23 insertions(+), 2 deletions(-)
diff --git a/extract_f0_print.py b/extract_f0_print.py
index e175e5e..3290125 100644
--- a/extract_f0_print.py
+++ b/extract_f0_print.py
@@ -17,6 +17,12 @@ from multiprocessing import Process
exp_dir = sys.argv[1]
f = open("%s/extract_f0_feature.log" % exp_dir, "a+")
+DoFormant = False
+
+with open('formanting.txt', 'r') as fvf:
+ content = fvf.readlines()
+ Quefrency, Timbre = content[1].split('\n')[0], content[2].split('\n')[0]
+
def printt(strr):
print(strr)
@@ -199,7 +205,7 @@ class FeatureInput(object):
return f0_median_hybrid
def compute_f0(self, path, f0_method, crepe_hop_length):
- x = load_audio(path, self.fs)
+ x = load_audio(path, self.fs, DoFormant, Quefrency, Timbre)
p_len = x.shape[0] // self.hop
if f0_method == "pm":
time_step = 160 / 16000 * 1000
@@ -227,6 +233,14 @@ class FeatureInput(object):
frame_period=1000 * self.hop / self.fs,
)
f0 = pyworld.stonemask(x.astype(np.double), f0, t, self.fs)
+ elif f0_method == "rmvpe":
+ if hasattr(self, "model_rmvpe") == False:
+ from rmvpe import RMVPE
+ print("loading rmvpe model")
+ self.model_rmvpe = RMVPE(
+ "rmvpe.pt", is_half=False, device="cuda:0"
+ )
+ f0 = self.model_rmvpe.infer_from_audio(x, thred=0.03)
elif f0_method == "dio":
f0, t = pyworld.dio(
x.astype(np.double),
diff --git a/trainset_preprocess_pipeline_print.py b/trainset_preprocess_pipeline_print.py
index f16761a..185cf29 100644
--- a/trainset_preprocess_pipeline_print.py
+++ b/trainset_preprocess_pipeline_print.py
@@ -17,9 +17,16 @@ import multiprocessing
from my_utils import load_audio
import tqdm
+DoFormant = False
+Quefrency = 0.0
+Timbre = 0.0
+
mutex = multiprocessing.Lock()
f = open("%s/preprocess.log" % exp_dir, "a+")
+with open('formanting.txt', 'r') as fvf:
+ content = fvf.readlines()
+ Quefrency, Timbre = content[1].split('\n')[0], content[2].split('\n')[0]
def println(strr):
mutex.acquire()
@@ -77,7 +84,7 @@ class PreProcess:
def pipeline(self, path, idx0):
try:
- audio = load_audio(path, self.sr)
+ audio = load_audio(path, self.sr, DoFormant, Quefrency, Timbre)
# zero phased digital filter cause pre-ringing noise...
# audio = signal.filtfilt(self.bh, self.ah, audio)
audio = signal.lfilter(self.bh, self.ah, audio)
From 978eaabee25fb8615ce68bdf388e521e15ccad87 Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Sun, 23 Jul 2023 05:49:24 +0700
Subject: [PATCH 08/14] Add files via upload
fixed training tracebacks and added Stop Training button
---
infer-web.py | 52 +++++++++++++++++++++---
train_nsf_sim_cache_sid_load_pretrain.py | 23 ++++++++++-
2 files changed, 68 insertions(+), 7 deletions(-)
diff --git a/infer-web.py b/infer-web.py
index 7811749..68a2af3 100644
--- a/infer-web.py
+++ b/infer-web.py
@@ -3,6 +3,8 @@ import shutil
import sys
import json # Mangio fork using json for preset saving
+import signal
+
now_dir = os.getcwd()
sys.path.append(now_dir)
import traceback, pdb
@@ -73,6 +75,8 @@ gpu_infos = []
mem = []
if_gpu_ok = False
+isinterrupted = 0
+
if torch.cuda.is_available() or ngpu != 0:
for i in range(ngpu):
gpu_name = torch.cuda.get_device_name(i)
@@ -587,7 +591,7 @@ def change_choices():
def clean():
- return {"value": "", "__type__": "update"}
+ return ({"value": "", "__type__": "update"})
sr_dict = {
@@ -1048,7 +1052,11 @@ def click_train(
)
)
print(cmd)
+ global p
p = Popen(cmd, shell=True, cwd=now_dir)
+ global PID
+ PID = p.pid
+
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"
@@ -1838,6 +1846,24 @@ def choveraudio():
return ''
+def stoptraining(mim):
+ if int(mim) == 1:
+
+ with open("stop.txt", "w+") as tostops:
+
+
+ tostops.writelines('stop')
+ #p.terminate()
+ #p.kill()
+ os.kill(PID, signal.SIGTERM)
+ else:
+ pass
+
+ return (
+ {"visible": False, "__type__": "update"},
+ {"visible": True, "__type__": "update"},
+ )
+
#Default-GUI
with gr.Blocks(theme='HaleyCH/HaleyCH_Theme') as app:
gr.HTML(" The Mangio-RVC-Fork 💻
")
@@ -2415,10 +2441,20 @@ with gr.Blocks(theme='HaleyCH/HaleyCH_Theme') as app:
value=gpus,
interactive=True,
)
- but3 = gr.Button(i18n("训练模型"), variant="primary")
+ butstop = gr.Button(
+ "Stop Training",
+ variant='primary',
+ visible=False,
+ )
+ but3 = gr.Button(i18n("训练模型"), variant="primary", visible=True)
+ but3.click(fn=stoptraining, inputs=[gr.Number(value=0, visible=False)], outputs=[but3, butstop])
+ butstop.click(fn=stoptraining, inputs=[gr.Number(value=1, visible=False)], outputs=[butstop, but3])
+
+
but4 = gr.Button(i18n("训练特征索引"), variant="primary")
#but5 = gr.Button(i18n("一键训练"), variant="primary")
info3 = gr.Textbox(label=i18n("输出信息"), value="", max_lines=10)
+
but3.click(
click_train,
[
@@ -2437,9 +2473,13 @@ with gr.Blocks(theme='HaleyCH/HaleyCH_Theme') as app:
if_save_every_weights18,
version19,
],
- info3,
+ [info3],
)
+
but4.click(train_index, [exp_dir1, version19], info3)
+
+
+
#but5.click(
# train1key,
# [
@@ -2464,7 +2504,7 @@ with gr.Blocks(theme='HaleyCH/HaleyCH_Theme') as app:
# ],
# info3,
#)
-
+
with gr.TabItem(i18n("ckpt处理")):
with gr.Group():
gr.Markdown(value=i18n("模型融合, 可用于测试音色融合"))
@@ -2761,7 +2801,7 @@ with gr.Blocks(theme='HaleyCH/HaleyCH_Theme') as app:
server_name="0.0.0.0",
inbrowser=not config.noautoopen,
server_port=config.listen_port,
- quiet=True,
+ quiet=False,
)
#endregion
@@ -3695,4 +3735,4 @@ with gr.Blocks(theme='HaleyCH/HaleyCH_Theme') as app:
)
#endregion
-'''
+'''
\ No newline at end of file
diff --git a/train_nsf_sim_cache_sid_load_pretrain.py b/train_nsf_sim_cache_sid_load_pretrain.py
index 3394bdd..5bb3526 100644
--- a/train_nsf_sim_cache_sid_load_pretrain.py
+++ b/train_nsf_sim_cache_sid_load_pretrain.py
@@ -568,7 +568,28 @@ def train_and_evaluate(
),
)
)
-
+
+ with open("stop.txt", "r+") as tostop:
+ content = tostop.read()
+ if 'stop' in content:
+ logger.info("Stop Button was pressed. The program is closed.")
+ if hasattr(net_g, "module"):
+ ckpt = net_g.module.state_dict()
+ else:
+ ckpt = net_g.state_dict()
+ logger.info(
+ "saving final ckpt:%s"
+ % (
+ savee(
+ ckpt, hps.sample_rate, hps.if_f0, hps.name, epoch, hps.version, hps
+ )
+ )
+ )
+
+ tostop.truncate(0)
+ tostop.writelines("not")
+ os._exit(2333333)
+
if rank == 0:
logger.info("====> Epoch: {} {}".format(epoch, epoch_recorder.record()))
if epoch >= hps.total_epoch and rank == 0:
From 263fa69f05012b12acc56c7c7aac6dbfb3e5ad67 Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Sun, 23 Jul 2023 05:50:22 +0700
Subject: [PATCH 09/14] Update README.md
---
README.md | 1 +
1 file changed, 1 insertion(+)
diff --git a/README.md b/README.md
index 83710c1..95b1410 100644
--- a/README.md
+++ b/README.md
@@ -1,5 +1,6 @@
# Features:
- Experimental Formant Shift using StftPitchShift(tried using praat with praatio but to no avail)
+- Added `Stop Training` button when training, no need to restart RVC every time you want to stop the training of a model!
- Auto-detect Index path for models selected + Auto-detect paths, no more default values like this: `E:\codes\py39\vits_vc_gpu_train\logs\mi-test-1key\total_fea.npy`, We're getting Root Dir and subfolders using
```python
os.path.abspath(os.getcwd())
From 385d7c57f3f78c0bcca7715e68409f5808b3a8cb Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Sun, 23 Jul 2023 05:54:04 +0700
Subject: [PATCH 10/14] Add files via upload
---
stop.txt | 0
1 file changed, 0 insertions(+), 0 deletions(-)
create mode 100644 stop.txt
diff --git a/stop.txt b/stop.txt
new file mode 100644
index 0000000..e69de29
From 5c6a9066f59e1c423c6598cd0d21a20e1093b23c Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Sun, 23 Jul 2023 05:59:53 +0700
Subject: [PATCH 11/14] Add files via upload
---
requirements.txt | 2 ++
1 file changed, 2 insertions(+)
diff --git a/requirements.txt b/requirements.txt
index 8965df5..6ec9214 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -45,3 +45,5 @@ httpx==0.23.0
#onnxruntime-gpu
torchcrepe==0.0.20
fastapi==0.88
+stftpitchshift==1.5.1
+gradio==3.34.0
From cc14c1746e8df9aab47117d344ba4e66a71195b7 Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Sun, 23 Jul 2023 06:00:31 +0700
Subject: [PATCH 12/14] Add files via upload
---
requirements.txt | 1 -
1 file changed, 1 deletion(-)
diff --git a/requirements.txt b/requirements.txt
index 6ec9214..3103d5e 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -46,4 +46,3 @@ httpx==0.23.0
torchcrepe==0.0.20
fastapi==0.88
stftpitchshift==1.5.1
-gradio==3.34.0
From a3a651d4f8638d7a8bede2e63f5daf4cd4d73cd5 Mon Sep 17 00:00:00 2001
From: alexlnkp <79400603+alexlnkp@users.noreply.github.com>
Date: Sun, 23 Jul 2023 06:04:36 +0700
Subject: [PATCH 13/14] Add files via upload
From 25265bbe301cda9ae3881785e2a6ff2fc22fe6db Mon Sep 17 00:00:00 2001
From: kalomaze <66376113+kalomaze@users.noreply.github.com>
Date: Sat, 22 Jul 2023 18:11:29 -0500
Subject: [PATCH 14/14] Removed non-fork files
---
installstft.bat | 6 ------
tensor-launch.py | 16 ----------------
2 files changed, 22 deletions(-)
delete mode 100644 installstft.bat
delete mode 100644 tensor-launch.py
diff --git a/installstft.bat b/installstft.bat
deleted file mode 100644
index 8d57717..0000000
--- a/installstft.bat
+++ /dev/null
@@ -1,6 +0,0 @@
-runtime\python.exe -m pip install stftpitchshift --upgrade
-runtime\python.exe -m pip install gradio==3.34.0 --upgrade
-
-
-IF EXIST ".\rmvpe.pt" (echo RMVPE is already installed) ELSE (bitsadmin /transfer "rvmpedwnld" /download /priority FOREGROUND "https://huggingface.co/lj1995/VoiceConversionWebUI/resolve/main/rmvpe.pt" "%~dp0rmvpe.pt")
-pause
\ No newline at end of file
diff --git a/tensor-launch.py b/tensor-launch.py
deleted file mode 100644
index 511335a..0000000
--- a/tensor-launch.py
+++ /dev/null
@@ -1,16 +0,0 @@
-import threading
-import time
-from tensorboard import program
-import webbrowser
-
-log_path = "logs"
-
-if __name__ == "__main__":
- tb = program.TensorBoard()
- tb.configure(argv=[None, '--logdir', log_path])
- url = tb.launch()
- webbrowser.open_new(url)
- print(f'Tensorboard can be accessed at: {url}')
-
- while True:
- time.sleep(600) # Keep the main thread running
\ No newline at end of file