ESRGAN Doesn't run on GPU "RuntimeError: error in LoadLibraryA"

[Masterbond7]

Sorry, I am new to CUDA and things like this but every time I want to run ESRGAN on my GPU it gives me this error. Works fine on CPU.

Full output: ============================== RESTART: C:\Users\Masterbond7\Desktop\ESRGAN-master\test.py ============================= Model path models/RRDB_ESRGAN_x4.pth. Testing... 1 baboon Traceback (most recent call last): File "C:\Users\Masterbond7\Desktop\ESRGAN-master\test.py", line 34, in output = model(imgLR).data.squeeze().float().cpu().clamp(0, 1).numpy() File "C:\Users\Masterbond7\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\module.py", line 532, in call result = self.forward(*input, kwargs) File "C:\Users\Masterbond7\Desktop\ESRGAN-master\RRDBNet_arch.py", line 71, in forward trunk = self.trunk_conv(self.RRDB_trunk(fea)) File "C:\Users\Masterbond7\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\module.py", line 532, in call result = self.forward(*input, *kwargs) File "C:\Users\Masterbond7\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\container.py", line 100, in forward input = module(input) File "C:\Users\Masterbond7\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\module.py", line 532, in call result = self.forward(input, kwargs) File "C:\Users\Masterbond7\Desktop\ESRGAN-master\RRDBNet_arch.py", line 47, in forward out = self.RDB1(x) File "C:\Users\Masterbond7\AppData\Local\Programs\Python\Python38\lib\site-packages\torch\nn\modules\module.py", line 532, in call result = self.forward(*input, **kwargs) File "C:\Users\Masterbond7\Desktop\ESRGAN-master\RRDBNet_arch.py", line 30, in forward x2 = self.lrelu(self.conv2(torch.cat((x, x1), 1))) RuntimeError: error in LoadLibraryA

[rustemsadrl]

You're probably running the latest version of Python. I had the same error until I downgraded to Python 3.7.5 and now it works.

[rlaphoenix]

You're probably running the latest version of Python. I had the same error until I downgraded to Python 3.7.5 and now it works.

Python 3.8.1 on latest CUDA and latest NVIDIA drivers works for me. I'm on arch linux with everything up to date.

[autofasurer]

Put this after importing torch:

import ctypes
ctypes.cdll.LoadLibrary('caffe2_nvrtc.dll')

[Dhawaleswrar]

Put this after importing torch:

import ctypes
ctypes.cdll.LoadLibrary('caffe2_nvrtc.dll')

It's working now..Thank you..