How to solve environment problem while using NVIDIA 3060 ?
Opened this issue · 3 comments
First time I did exactly the same as the readme says,then I got capability mismatch problem.
second time I reinstall pytorch==1.13 and mmv==1.7.0 and got error says I need to install mmcv>=(1, 3, 0, 0, 0, 0), <=(1, 5, 0, 0, 0, 0).
first time error
``load checkpoint from local path: pretrained/AiLUT-PPR10KA-sRGB.pth
The model and loaded state dict do not match exactly
missing keys in source state_dict: cnt_iters
/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/cuda/init.py:104: UserWarning:
NVIDIA GeForce RTX 3060 with CUDA capability sm_86 is not compatible with the current PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_61 sm_70 sm_75 compute_37.
If you want to use the NVIDIA GeForce RTX 3060 GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
warnings.warn(incompatible_device_warn.format(device_name, capability, " ".join(arch_list), device_name))
Traceback (most recent call last):
File "adaint/demo.py", line 75, in
main()
File "adaint/demo.py", line 69, in main
output = enhancement_inference(model, args.img_path)
File "adaint/demo.py", line 43, in enhancement_inference
result = model(test_mode=True, **data)
File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
result = self.forward(*input, **kwargs)
File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/mmcv/runner/fp16_utils.py", line 110, in new_func
return old_func(*args, **kwargs)
File "/home/zhanjinhao/codes/AdaInt/adaint/model.py", line 346, in forward
return self.forward_test(lq, gt, **kwargs)
File "/home/zhanjinhao/codes/AdaInt/adaint/model.py", line 395, in forward_test
output, _, _ = self.forward_dummy(lq)
File "/home/zhanjinhao/codes/AdaInt/adaint/model.py", line 321, in forward_dummy
codes = self.backbone(imgs)
File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
result = self.forward(*input, **kwargs)
File "/home/zhanjinhao/codes/AdaInt/adaint/model.py", line 89, in forward
return self.net(imgs).view(imgs.shape[0], -1)
File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl
result = self.forward(*input, **kwargs)
File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torchvision/models/resnet.py", line 152, in forward
x = self.relu(x)
File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in call_impl
result = self.forward(*input, **kwargs)
File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/activation.py", line 102, in forward
return F.relu(input, inplace=self.inplace)
File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/functional.py", line 1204, in relu
result = torch.relu(input)
RuntimeError: CUDA error: no kernel image is available for execution on the device
second time error:
/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/mmcv/__init__.py:21: UserWarning: On January 1, 2023, MMCV will release v2.0.0, in which it will remove components related to the training process and add a data transformation module. In addition, it will rename the package names mmcv to mmcv-lite and mmcv-full to mmcv. See https://github.com/open-mmlab/mmcv/blob/master/docs/en/compatibility.md for more details. 'On January 1, 2023, MMCV will release v2.0.0, in which it will remove ' Traceback (most recent call last): File "adaint/demo.py", line 8, in <module> from mmedit.apis import init_model File "/home/zhanjinhao/codes/AdaInt/mmedit/__init__.py", line 31, in <module> f'MMCV=={mmcv.__version__} is used but incompatible. ' \ AssertionError: MMCV==1.7.1 is used but incompatible. Please install mmcv>=(1, 3, 0, 0, 0, 0), <=(1, 5, 0, 0, 0, 0).
first time error ``load checkpoint from local path: pretrained/AiLUT-PPR10KA-sRGB.pth The model and loaded state dict do not match exactly
missing keys in source state_dict: cnt_iters
/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/cuda/init.py:104: UserWarning: NVIDIA GeForce RTX 3060 with CUDA capability sm_86 is not compatible with the current PyTorch installation. The current PyTorch install supports CUDA capabilities sm_37 sm_50 sm_60 sm_61 sm_70 sm_75 compute_37. If you want to use the NVIDIA GeForce RTX 3060 GPU with PyTorch, please check the instructions at https://pytorch.org/get-started/locally/
warnings.warn(incompatible_device_warn.format(device_name, capability, " ".join(arch_list), device_name)) Traceback (most recent call last): File "adaint/demo.py", line 75, in main() File "adaint/demo.py", line 69, in main output = enhancement_inference(model, args.img_path) File "adaint/demo.py", line 43, in enhancement_inference result = model(test_mode=True, **data) File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/mmcv/runner/fp16_utils.py", line 110, in new_func return old_func(*args, **kwargs) File "/home/zhanjinhao/codes/AdaInt/adaint/model.py", line 346, in forward return self.forward_test(lq, gt, **kwargs) File "/home/zhanjinhao/codes/AdaInt/adaint/model.py", line 395, in forward_test output, _, _ = self.forward_dummy(lq) File "/home/zhanjinhao/codes/AdaInt/adaint/model.py", line 321, in forward_dummy codes = self.backbone(imgs) File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) File "/home/zhanjinhao/codes/AdaInt/adaint/model.py", line 89, in forward return self.net(imgs).view(imgs.shape[0], -1) File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in _call_impl result = self.forward(*input, **kwargs) File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torchvision/models/resnet.py", line 152, in forward x = self.relu(x) File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/module.py", line 889, in call_impl result = self.forward(*input, **kwargs) File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/modules/activation.py", line 102, in forward return F.relu(input, inplace=self.inplace) File "/home/zhanjinhao/anaconda3/envs/AdaInt/lib/python3.7/site-packages/torch/nn/functional.py", line 1204, in relu result = torch.relu(input) RuntimeError: CUDA error: no kernel image is available for execution on the device
I have the same problem, however, I change the code in the model.py and fixed this problem:
change the original code self.register_buffer('cnt_iters', torch.zeros(1)) into 'self.cnt_iters = torch.zeros(1)'