Coder Social home page Coder Social logo

wlsdml1114 / ddsp-svc-kor Goto Github PK

View Code? Open in Web Editor NEW

This project forked from yxlllc/ddsp-svc

37.0 37.0 7.0 22.09 MB

Real-time end-to-end singing voice conversion system based on DDSP (Differentiable Digital Signal Processing)

License: MIT License

Python 93.55% Jupyter Notebook 6.45%

ddsp-svc-kor's Introduction

Hi there ๐Ÿ‘‹

Now

  • AI Researcher Sep. 2022 ~ present

Education

  • M.S.E., School of Computer Science and Engineering, College of Information Technology, Soongsil University, Seoul, Korea, Mar. 2018 ~ Aug. 2019.
  • B.S.E., School of Computer Science and Engineering, College of Information Technology, Soongsil University, Seoul, Korea, Mar. 2014 ~ Feb. 2018.

Others

  • Military Service: Technical Research Personnel in KOAST (Korea Oceanic & Atmospheric System Technology), Aug. 2019. ~ Aug. 2022.
    • as AI researcher and ML engineer

Interest

  • Metaverse
  • Virtual Human, Virtual Idol
  • VR, AR
  • Computer Vision
  • Animation

ddsp-svc-kor's People

Contributors

aiczk avatar cardroid avatar cnchtu avatar ddpn08 avatar dillfrescott avatar entropyriser avatar fatinghenji avatar huanlinoto avatar kakaruhayate avatar l4ph avatar magic-akari avatar ms903x1 avatar petyin avatar therealkamisama avatar wlsdml1114 avatar yxlllc avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

ddsp-svc-kor's Issues

์‚ฌ์ „ํ•™์Šต๋ชจ๋ธ์„ ์‚ฌ์šฉํ•˜์ง€ ์•Š๊ณ  ํ•™์Šต์‹œ์ผœ๋ณด๊ณ ์‹ถ์€๋ฐ

์‚ฌ์ „ํ•™์Šต๋ชจ๋ธ์„ ์‚ฌ์šฉํ•˜์ง€ ์•Š๊ณ  ํ•™์Šต์‹œ์ผœ๋ณด๊ณ ์‹ถ์€๋ฐ checkpoint์•ˆ์—์žˆ๋Š” 0102_xiaoma_pe, 0109_hifigan_bigpopcs_hop128 ํด๋”๋งŒ ์‚ญ์ œํ•˜๊ณ  ์ด์ „๊ณผ ๊ฐ™์ด ํ•™์Šต์‹œํ‚ค๋ฉด ๋ ๊นŒ์š”?

notebook.ipynb์— ์žˆ๋Š” ํŒŒ์ผ ์ž๋ฅด๊ธฐ๊ฐ€ ๋ญ˜ ํ•˜๋ผ๋Š” ๊ฑด์ง€ ๋ชจ๋ฅด๊ฒ ์Šต๋‹ˆ๋‹ค.

"## 1.4 split\n",
"preprocess/norm์— ์žˆ๋Š” ๋…ธ๋ง๋ผ์ด์ฆˆ๋œ ๋ฐ์ดํ„ฐ๋“ค์„ 15์ดˆ ๊ธธ์ด๋กœ ์ž˜๋ผ์„œ data/train/audio์— ์ €์žฅ"

์—ฌ๊ธฐ ์ด ๋ถ€๋ถ„ 1.4๋‹จ๊ณ„์—์„œ ๋ญ˜ ํ•˜๋ผ๋Š”๊ฑด์ง€ ๋ชจ๋ฅด๊ฒ ์Šต๋‹ˆ๋‹คใ… ใ…  ์ง์ ‘ ํŒŒ์ผ์„ ์ž˜๋ผ์„œ ์ € ํด๋”์— ์ €์žฅํ•˜๋ผ๋Š” ๋œป์ธ๊ฐ€์š”..?

notebook.ipynb๊ฐ€ ๋ณด์ด์ง€ ์•Š์Šต๋‹ˆ๋‹ค.

์•ˆ๋…•ํ•˜์„ธ์š”.
๋””์Šค์ฝ”๋“œ ๋งํฌ๋ฅผ ํƒ€๊ณ  ๋“ค์–ด๊ฐ€๋„ ๋ฐฉ์ด ๋ณด์ด์ง€ ์•Š์•„์„œ ์—ฌ๊ธฐ์— ์งˆ๋ฌธ์„ ๋‚จ๊น๋‹ˆ๋‹ค.
๊ฑฐ์˜ ๋‹ค ๋๋Š”๋ฐ, 2. ๋ฐ์ดํ„ฐ ์ค€๋น„ ๋ฐ ์ „์ฒ˜๋ฆฌ, ํ•™์Šต - ์—ฌ๊ธฐ์—์„œ ์™ผ์ชฝ ํŒŒ์ผ ๋ชฉ๋ก์—์„œ notebook.ipynb ์ด๋ผ๋Š” ๊ฒƒ์ด ๋ณด์ด์ง€ ์•Š์Šต๋‹ˆ๋‹ค.
์•„๋‚˜์ฝ˜๋‹ค ํ”„๋กฌํ”„ํŠธ ์—ด๊ณ  ๊ธฐ๋ณธ ์ƒํƒœ์—์„œ๋„ ํ•ด๋ณด๊ณ , ddsp ๊ฐ€์ƒํ™˜๊ฒฝ ๋งŒ๋“ค์–ด๋†“์€ ๊ณณ์œผ๋กœ ์ง€์ •ํ•ด์„œ๋„ ํ•ด๋ณด๊ณ , ํ”„๋กœ์ ํŠธ ํŒŒ์ผ์ธ DDSP-SVC๋กœ ์ง€์ •ํ•˜๊ณ  ํ•ด๋ด๋„ notebook.ipynb๋Š” ๋ณด์ด์ง€ ์•Š๋„ค์š”.
๊ฐ์‚ฌํ•ฉ๋‹ˆ๋‹ค.

args ๊ฐ€ ๋ฌด์—‡์ธ๊ฐ€์š”?

๋ฐ์ดํ„ฐ ์ „์ฒ˜๋ฆฌ๋Š” ๋งˆ์ณ์„œ ์ฅฌํ”ผํ„ฐ ๋…ธํŠธ๋ถ์˜ 2 ๋ฒˆ์„ ์ง„ํ–‰ํ•˜๋ ค๊ณ  ํ•˜๋Š”๋ฐ ์ด๋Ÿฐ ์ฝ”๋“œ๊ฐ€ ๋œจ๋„ค์š” ์–ด๋–ป๊ฒŒ ํ•ด๊ฒฐํ•ด์•ผ ํ• ๊นŒ์š”?

C:\Anaconda3\envs\ddsp\lib\site-packages\tqdm\auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
from .autonotebook import tqdm as notebook_tqdm
2023-07-16 14:26:46 | INFO | fairseq.tasks.text_to_speech | Please install tensorboardX: pip install tensorboardX

NameError Traceback (most recent call last)
Cell In[1], line 6
3 from diffusion.vocoder import Vocoder
5 # get data
----> 6 sample_rate = args.data.sampling_rate
7 hop_size = args.data.block_size
9 # initialize f0 extractor

NameError: name 'args' is not defined

๊ฒฐ๊ณผ๋ฌผ ๋ฝ‘๊ธฐ ์งˆ๋ฌธ๋“œ๋ ค์š”

์•„๋ž˜ ์˜ค๋ฅ˜ ๋‚˜๋Š” ์ด์œ  ์ฐพ์•˜์Šต๋‹ˆ๋‹ค. ์ƒˆ๋กœ ์ง€์ •ํ•œ ๋ชจ๋ธ ํŒจ์Šค์— config.yaml ํŒŒ์ผ๋„ ์˜ฎ๊ฒจ์ฃผ์–ด์•ผ ํ•˜๋„ค์š”.


'model_path' : 'exp/combsub-test/model_best.pt', # ์ถ”๋ก ์— ์‚ฌ์šฉํ•˜๊ณ ์ž ํ•˜๋Š” ๋ชจ๋ธ, ๋ฐ”๋กœ์œ„์—์„œ ํ•™์Šตํ•œ ๋ชจ๋ธ์„ ๊ฐ€์ ธ์˜ค๋ฉด๋Œ
'input' : 'data/train/audio/video-0000.wav', # ์ถ”๋ก ํ•˜๊ณ ์ž ํ•˜๋Š” ๋…ธ๋ž˜ํŒŒ์ผ์˜ ์œ„์น˜ - ๋‹˜๋“ค์ด ๋ฐ”๊ฟ”์•ผ๋Œ
'output' : 'output.wav', # ๊ฒฐ๊ณผ๋ฌผ ํŒŒ์ผ์˜ ์œ„์น˜

๊ฒฐ๊ณผ๋ฌผ ๋ฝ‘์„ ๋•Œ ์ด๋ ‡๊ฒŒ 3๋ถ€๋ถ„ ๋ฐ”๊ฟ”์•ผ ๋˜์„œ ์•„๋ž˜์ฒ˜๋Ÿผ ๊ฒฝ๋กœ ๋ฐ”๊พธ๊ณ  ๊ฐ ๊ฒฝ๋กœ๋กœ ํด๋” ๋งŒ๋“ค๊ณ 
๊ทธ ํด๋”์— 10๋งŒ์Šคํ…์—์„œ ๋‚˜์˜จ model์„ model_best.pt๋กœ ์ด๋ฆ„ ๋ฐ”๊ฟ”์„œ ๋„ฃ์—ˆ๊ณ ,
input ๊ฒฝ๋กœ์ธ cover ๊ฒฝ๋กœ์—๋„ ์ œ๊ฐ€ ๋ฐ”๊พธ๊ณ  ์‹ถ์€ ๋…ธ๋ž˜ ํ•˜๋‚˜ ๋„ฃ๊ณ  ๋…ธ๋ž˜ ์ด๋ฆ„์„ test.wav ๋กœ ๋ณ€๊ฒฝํ•ด์„œ ๋„ฃ์—ˆ๊ตฌ์š”
์•„์›ƒํ’‹ ๊ฒฝ๋กœ๋„ ์•„๋ž˜์ฒ˜๋Ÿผ ์„ค์ •ํ•˜์˜€๋Š”๋ฐ

'model_path' : 'exp/combsubtest/model_best.pt', # ์ถ”๋ก ์— ์‚ฌ์šฉํ•˜๊ณ ์ž ํ•˜๋Š” ๋ชจ๋ธ, ๋ฐ”๋กœ์œ„์—์„œ ํ•™์Šตํ•œ ๋ชจ๋ธ์„ ๊ฐ€์ ธ์˜ค๋ฉด๋Œ
'input' : 'data/train/cover/test.wav', # ์ถ”๋ก ํ•˜๊ณ ์ž ํ•˜๋Š” ๋…ธ๋ž˜ํŒŒ์ผ์˜ ์œ„์น˜ - ๋‹˜๋“ค์ด ๋ฐ”๊ฟ”์•ผ๋Œ
'output' : 'data/train/output/output.wav', # ๊ฒฐ๊ณผ๋ฌผ ํŒŒ์ผ์˜ ์œ„์น˜

์•„๋ž˜์ฒ˜๋Ÿผ ๊ฒฝ๋กœ ์ธ์‹์„ ๋ชปํ•˜๋Š” ์—๋Ÿฌ๊ฐ€ ๋‚˜๋„ค์š”. ์™œ ๊ทธ๋Ÿด๊นŒ์š”? ใ… ใ…  ๊ฒฝ๋กœ๋งŒ ๋‚จ๊ธฐ๊ณ  ํŒŒ์ผ ์ด๋ฆ„์„ ์‚ญ์ œ๋„ ํ•ด๋ณด๊ณ ์—ฌ๋Ÿฌ๊ฐ€์ง€ ํ•ด๋ดค๋Š”๋ฐ ๋˜‘๊ฐ™๋„ค์š”
๋ฌด์—‡์ด ๋ฌธ์ œ์ผ๊นŒ์š”? ๋ฏธ๋ฆฌ ๊ฐ์‚ฌ๋“œ๋ฆฝ๋‹ˆ๋‹ค.

FileNotFoundError Traceback (most recent call last)
Cell In[56], line 21
4 configures = {
5 'model_path' : 'exp/combsubtest/model_best.pt', # ์ถ”๋ก ์— ์‚ฌ์šฉํ•˜๊ณ ์ž ํ•˜๋Š” ๋ชจ๋ธ, ๋ฐ”๋กœ์œ„์—์„œ ํ•™์Šตํ•œ ๋ชจ๋ธ์„ ๊ฐ€์ ธ์˜ค๋ฉด๋Œ
6 'input' : 'data/train/cover/test.wav', # ์ถ”๋ก ํ•˜๊ณ ์ž ํ•˜๋Š” ๋…ธ๋ž˜ํŒŒ์ผ์˜ ์œ„์น˜ - ๋‹˜๋“ค์ด ๋ฐ”๊ฟ”์•ผ๋Œ
(...)
17 'enhancer_adaptive_key' : '0'
18 }
19 cmd = SimpleNamespace(**configures)
---> 21 inference(cmd)

File ~\Downloads\DDSP-SVC-KOR-master\DDSP-SVC-KOR-master\main.py:163, in inference(cmd)
160 device = 'cuda' if torch.cuda.is_available() else 'cpu'
162 # load ddsp model
--> 163 model, args = load_model(cmd.model_path, device=device)
165 # load input
166 audio, sample_rate = librosa.load(cmd.input, sr=44100)

File ~\Downloads\DDSP-SVC-KOR-master\DDSP-SVC-KOR-master\ddsp\vocoder.py:434, in load_model(model_path, device)
430 def load_model(
431 model_path,
432 device='cpu'):
433 config_file = os.path.join(os.path.split(model_path)[0], 'config.yaml')
--> 434 with open(config_file, "r") as config:
435 args = yaml.safe_load(config)
436 args = DotDict(args)

FileNotFoundError: [Errno 2] No such file or directory: 'exp/combsubtest\config.yaml'
โ€‹

๋‘๋ช…์˜ ๊ฐ€์ˆ˜๋ฅผ ํ›ˆ๋ จ์‹œํ‚ฌ ๋•Œ

1๋ฒˆ ๋ชฉ์†Œ๋ฆฌ๋ฅผ ํ›ˆ๋ จ์‹œํ‚ค๋‹ค๊ฐ€ ๋‹ค๋ฅธ ๋ชฉ์†Œ๋ฆฌ๋„ ํ›ˆ๋ จ์‹œ์ผœ๋ณด๊ณ ์‹ถ์–ด์„œ 2๋ฒˆ ๋ชจ๋ธ์„ ๋งŒ๋“œ๋ ค๋Š”๋ฐ
์–ด๋–ป๊ฒŒ ํ•ด์•ผ ์ด์–ด์ ธ์„œ ํ›ˆ๋ จ๋˜์ง€์•Š๊ณ  ์ƒˆ๋กœ์šด ๋ชจ๋ธ๋กœ ๋‹ค์‹œ ํ›ˆ๋ จ์„ ์‹œ์ž‘ํ• ์ˆ˜์žˆ๋‚˜์š”>??

notebook.ipynb์—์„œ ์ฝ”๋“œ๊ฐ€ ์‹คํ–‰๋˜์ง€ ์•Š๋Š”๋ฐ ๋ญ๊ฐ€ ๋ฌธ์ œ์ธ๊ฐ€์š”?

์•„๋ž˜ ์ฝ”๋“œ๊นŒ์ง€ ์ง„ํ–‰ํ•˜๋ฉด mp4 ํด๋”์— ์žˆ๋Š” ํŒŒ์ผ์ด wav ํŒŒ์ผ๋กœ ๋ณ€๊ฒฝ๋˜์–ด original ํด๋”์— ๋“ค์–ด๊ฐ€์•ผ ํ• ๊ฒƒ ๊ฐ™์€๋ฐ ์•„๋ฌด๋Ÿฐ ๋ณ€ํ™”๊ฐ€ ์—†๋„ค์š”. ๋ญ๊ฐ€ ์ž˜๋ชป๋œ ๊ฒƒ์ผ๊นŒ์š”? ๊ณ ์ˆ˜๋‹˜๋“ค ๋„์™€์ฃผ์„ธ์š”.

1
2

ํ•™์Šต์„ ์ค‘๊ฐ„์— ๋Š์œผ๋ฉด ์ฑ•ํ„ฐ4๋กœ ๋ชป ๋„˜์–ด๊ฐ€๋‚˜์š”?

์ดํ•ฉ 25์‹œ๊ฐ„(์ถ”์ •) ํ•™์Šต์„ ์ง„ํ–‰ํ•œ ๊ฒƒ ๊ฐ™์€๋ฐ, ํ˜„์žฌ 269K steps์ธ ์ƒํ™ฉ์ž…๋‹ˆ๋‹ค.
์ค‘๊ฐ„์ค‘๊ฐ„ ๋ช‡ ๋ฒˆ ๋Š์œผ๋ฉด์„œ ์ฑ•ํ„ฐ4 ๊ฒฐ๊ณผ๋ฌผ ๋ฝ‘๊ธฐ๋กœ ๋„˜์–ด๊ฐ€๋ ค๊ณ  ํ–ˆ์œผ๋‚˜ ์ง„ํ–‰์ด ์•ˆ ๋˜์–ด ํ•™์Šต์ด ์ „๋ถ€ ๋๋‚˜์•ผ ๋˜๋Š” ๊ฑด๊ฐ€ ์‹ถ์–ด ๋‹ค์‹œ ํ•˜๊ธฐ๋ฅผ ๋ฐ˜๋ณตํ–ˆ๋Š”๋ฐ, ์ฑ•ํ„ฐ3 ํ•™์Šต ๊ณผ์ •์—์„œ ๋งˆ๋ฌด๋ฆฌ ๋  ๋•Œ๊นŒ์ง€ ๊ธฐ๋‹ค๋ฆฌ๋Š” ๊ฒƒ์ด ๋งž๋‚˜์š”?

์œ ํŠœ๋ธŒ ๋Œ“๊ธ€์—์„œ๋Š” ๋กœ์Šค์œจ์ด ์ˆ˜๋ ดํ•˜๋ฉด ์ค‘๊ฐ„์— ๋Š์–ด๋„ ๊ดœ์ฐฎ๋‹ค๋Š” ๋ง์”€๋„ ์žˆ์œผ์…จ๋Š”๋ฐ ์‹ค์ œ๋กœ ๋Š์œผ๋‹ˆ๊นŒ ์ฑ•ํ„ฐ4์—์„œ ์˜ค๋ฅ˜๋ฅผ ๋ฑ‰์–ด๋‚ด์–ด ์ง„ํ–‰์ด ์•ˆ ๋˜์—ˆ์Šต๋‹ˆ๋‹ค.(config.json์ด ์—†๋‹ค๋Š” ์˜ค๋ฅ˜)

๋””์Šค์ฝ”๋“œ ์„œ๋ฒ„์— ์ฐธ์—ฌํ•ด ์—ฌ์ญค๋ณด๋ ค ํ–ˆ์œผ๋‚˜ ์ ‘์†์ด ์•ˆ ๋œ๋‹ค๊ณ  ๋‚˜์˜ค๋„ค์š”ใ… ใ… ...

1.2 ๋ฌด์Œ์ œ๊ฑฐ ๋‹จ๊ณ„์—์„œ ์˜ค๋ฅ˜ ๊ด€๋ จํ•ด์„œ ์—๋Ÿฌ ์ „๋ฌธ ๋‚จ๊น๋‹ˆ๋‹ค...!


OutOfMemoryError Traceback (most recent call last)
Cell In[9], line 3
1 from sep_wav import demucs
----> 3 demucs(ORIGINAL_PATH, DEMUCS_PATH)

File ~\Desktop\DDSP-SVC-KOR-master\DDSP-SVC-KOR-master\sep_wav.py:286, in demucs(input_path, output_path)
284 bundle = HDEMUCS_HIGH_MUSDB_PLUS
285 model = bundle.get_model()
--> 286 model.to(device)
287 sample_rate = bundle.sample_rate
288 print(f"Sample rate: {sample_rate}")

File ~\anaconda3\envs\ddsp\lib\site-packages\torch\nn\modules\module.py:1145, in Module.to(self, *args, **kwargs)
1141 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None,
1142 non_blocking, memory_format=convert_to_format)
1143 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking)
-> 1145 return self._apply(convert)

File ~\anaconda3\envs\ddsp\lib\site-packages\torch\nn\modules\module.py:797, in Module._apply(self, fn)
795 def _apply(self, fn):
796 for module in self.children():
--> 797 module._apply(fn)
799 def compute_should_use_set_data(tensor, tensor_applied):
800 if torch._has_compatible_shallow_copy_type(tensor, tensor_applied):
801 # If the new tensor has compatible tensor type as the existing tensor,
802 # the current behavior is to change the tensor in-place using .data =,
(...)
807 # global flag to let the user control whether they want the future
808 # behavior of overwriting the existing tensor or not.

File ~\anaconda3\envs\ddsp\lib\site-packages\torch\nn\modules\module.py:797, in Module._apply(self, fn)
795 def _apply(self, fn):
796 for module in self.children():
--> 797 module._apply(fn)
799 def compute_should_use_set_data(tensor, tensor_applied):
800 if torch._has_compatible_shallow_copy_type(tensor, tensor_applied):
801 # If the new tensor has compatible tensor type as the existing tensor,
802 # the current behavior is to change the tensor in-place using .data =,
(...)
807 # global flag to let the user control whether they want the future
808 # behavior of overwriting the existing tensor or not.

File ~\anaconda3\envs\ddsp\lib\site-packages\torch\nn\modules\module.py:797, in Module._apply(self, fn)
795 def _apply(self, fn):
796 for module in self.children():
--> 797 module._apply(fn)
799 def compute_should_use_set_data(tensor, tensor_applied):
800 if torch._has_compatible_shallow_copy_type(tensor, tensor_applied):
801 # If the new tensor has compatible tensor type as the existing tensor,
802 # the current behavior is to change the tensor in-place using .data =,
(...)
807 # global flag to let the user control whether they want the future
808 # behavior of overwriting the existing tensor or not.

File ~\anaconda3\envs\ddsp\lib\site-packages\torch\nn\modules\module.py:820, in Module._apply(self, fn)
816 # Tensors stored in modules are graph leaves, and we don't want to
817 # track autograd history of param_applied, so we have to use
818 # with torch.no_grad():
819 with torch.no_grad():
--> 820 param_applied = fn(param)
821 should_use_set_data = compute_should_use_set_data(param, param_applied)
822 if should_use_set_data:

File ~\anaconda3\envs\ddsp\lib\site-packages\torch\nn\modules\module.py:1143, in Module.to..convert(t)
1140 if convert_to_format is not None and t.dim() in (4, 5):
1141 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None,
1142 non_blocking, memory_format=convert_to_format)
-> 1143 return t.to(device, dtype if t.is_floating_point() or t.is_complex() else None, non_blocking)

OutOfMemoryError: CUDA out of memory. Tried to allocate 20.00 MiB (GPU 0; 6.00 GiB total capacity; 5.29 GiB already allocated; 0 bytes free; 5.29 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

์ˆœ์„œ๋Œ€๋กœ ํ–ˆ๋Š”๋ฐ ใ… 

์ˆœ์„œ๋Œ€๋กœ ๋‹ค ์„ค์น˜ํ•˜๊ณ 
์ฒซ๋ช…๋ น์–ด conda create -n ddsp python=3.9 ์น˜๋ฉด ์™œ ์•„๋ž˜์™€ ๊ฐ™์ด ๋‚˜์˜ฌ๊นŒ์š”?

Unable to create process using 'C:\Users\Family\anaconda3\python.exe "C:\Users\Family\anaconda3\Scripts\conda-script.py" create -n ddsp python=3.9'

์ด๋ผ๊ณ  ๋‚˜์˜ค๋Š”๊ฑด ์ด๋ฏธ ์ „์— ํ•œ์ ์ด ์žˆ์–ด์„œ ๊ทธ๋Ÿฐ๊ฐ€์š”?

1.2 ๋ฌด์Œ์ œ๊ฑฐ์—์„œ CUDA out of memory

OutOfMemoryError: CUDA out of memory. Tried to allocate 9.45 GiB (GPU 0; 12.00 GiB total capacity; 2.68 GiB already allocated; 7.32 GiB free; 2.71 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
์–ด๋–ป๊ฒŒ ํ•ด๊ฒฐํ•˜๋‚˜์š”?

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.