RuntimeError: CUDA error: no kernel image is available for execution on the device

[RoychenHKMU]

when i try to do a quick test and use this command it show CUDA error. please someone help me, thank you so very much.

my setup step :

First i follow the "Dependencies and Installation" to clone repo and "Create Conda Environment and Install Dependencies" but there is a error on create conda environment, the error is "Solving environment: failed" so i follow solution in issues i use this feww code to fix the problem.

conda create -n e2fgvi python=3.7
conda activate e2fgvi

conda install pytorch==1.5.1 torchvision==0.6.1 cudatoolkit=10.1 -c pytorch
pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu101/torch1.5/index.html
conda install tensorboard matplotlib scikit-image==0.16.2
pip install tqdm

after that i try to start the quick test, but it show "ModuleNotFoundError: No module named 'cv2'" so i use "pip install opencv-python" to fix it, after that i try to use the quick test command then show a long long error message. It show CUDA error, i am using RTX 4090, so i try to reinstall pytouch and CUDA, i install CUDA 11.8 and correspondence pytouch, but still failed to do the quick test. Thank you so much if anyone can help.

Command i use :

python test.py --model e2fgvi --video examples/tennis --mask examples/tennis_mask --ckpt release_model/E2FGVI-CVPR22.pth

Error code :

(e2fgvi) C:\Researchteam\E2FGVI>python test.py --model e2fgvi --video examples/tennis --mask examples/tennis_mask --ckpt release_model/E2FGVI-CVPR22.pth
C:\Users\user\anaconda3\envs\e2fgvi\lib\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 '
load pretrained SPyNet...
load checkpoint from http path: https://download.openmmlab.com/mmediting/restorers/basicvsr/spynet_20210409-c6c1bd09.pth
Loading model from: release_model/E2FGVI-CVPR22.pth
Loading videos and masks from: examples/tennis | INPUT MP4 format: False
Start test...
0%| | 0/14 [05:30<?, ?it/s]
Traceback (most recent call last):
File "test.py", line 224, in
main_worker()
File "test.py", line 166, in main_worker
pred_imgs, _ = model(masked_imgs, len(neighbor_ids))
File "C:\Users\user\anaconda3\envs\e2fgvi\lib\site-packages\torch\nn\modules\module.py", line 550, in call
result = self.forward(*input, **kwargs)
File "C:\Researchteam\E2FGVI\model\e2fgvi.py", line 250, in forward
pred_flows[1])
File "C:\Users\user\anaconda3\envs\e2fgvi\lib\site-packages\torch\nn\modules\module.py", line 550, in call
result = self.forward(*input, **kwargs)
File "C:\Researchteam\E2FGVI\model\modules\feat_prop.py", line 129, in forward
flow_n2)
File "C:\Users\user\anaconda3\envs\e2fgvi\lib\site-packages\torch\nn\modules\module.py", line 550, in call
result = self.forward(*input, **kwargs)
File "C:\Researchteam\E2FGVI\model\modules\feat_prop.py", line 58, in forward
self.deform_groups)
File "C:\Users\user\anaconda3\envs\e2fgvi\lib\site-packages\mmcv\ops\modulated_deform_conv.py", line 160, in forward
with_bias=ctx.with_bias)
RuntimeError: CUDA error: no kernel image is available for execution on the device