stable-dreamfusion
stable-dreamfusion copied to clipboard
ashawkey
No CUDA runtime is found, using CUDA_HOME='C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6'
Tried to add these as well, but to no avail. Reinstalled conda with the needed dependencies as far as i could. But this part i'm now stuck with. error code below image.
(base) C:\Users\Duckers>cd C:\stable-dreamfusion-main\stable-dreamfusion
(base) C:\stable-dreamfusion-main\stable-dreamfusion>pip install ./raymarching Processing c:\stable-dreamfusion-main\stable-dreamfusion\raymarching Preparing metadata (setup.py) ... done Building wheels for collected packages: raymarching Building wheel for raymarching (setup.py) ... error error: subprocess-exited-with-error
× python setup.py bdist_wheel did not run successfully.
│ exit code: 1
╰─> [50 lines of output]
No CUDA runtime is found, using CUDA_HOME='C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6'
running bdist_wheel
running build
running build_ext
Traceback (most recent call last):
File "
note: This error originates from a subprocess, and is likely not a problem with pip. ERROR: Failed building wheel for raymarching Running setup.py clean for raymarching Failed to build raymarching Installing collected packages: raymarching Attempting uninstall: raymarching Found existing installation: raymarching 0.1.3 Uninstalling raymarching-0.1.3: Successfully uninstalled raymarching-0.1.3 Running setup.py install for raymarching ... error error: subprocess-exited-with-error
× Running setup.py install for raymarching did not run successfully.
│ exit code: 1
╰─> [54 lines of output]
No CUDA runtime is found, using CUDA_HOME='C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.6'
running install
C:\Users\Duckers\miniconda3\lib\site-packages\setuptools\command\install.py:34: SetuptoolsDeprecationWarning: setup.py install is deprecated. Use build and pip and other standards-based tools.
warnings.warn(
running build
running build_ext
Traceback (most recent call last):
File "
note: This error originates from a subprocess, and is likely not a problem with pip.
Rolling back uninstall of raymarching
Moving to c:\users\duckers\miniconda3\lib\site-packages\raymarching-0.1.3.dist-info
from C:\Users\Duckers\miniconda3\Lib\site-packages~aymarching-0.1.3.dist-info
Moving to c:\users\duckers\miniconda3\lib\site-packages\raymarching
from C:\Users\Duckers\miniconda3\Lib\site-packages~aymarching
error: legacy-install-failure
× Encountered error while trying to install package. ╰─> raymarching
note: This is an issue with the package mentioned above, not pip. hint: See above for output from the failure.
(base) C:\stable-dreamfusion-main\stable-dreamfusion>``
I have the same issue like you , the possible cause may be the CUDA and pytorch is dismatched you can print this code to test it in python：
import torch
print(torch.__version__)
print(torch.version.cuda)
print(torch.cuda.is_available())
Hope this can help you!
'import' is not recognized as an internal or external command, operable program or batch file.
Going through the open issues you've had I've followed about the same path as you.
To get past this step I went to https://pytorch.org/get-started/locally/ and filled out the options for my computer like:
Then I pasted the command to my terminal but added --force-reinstall to the end of it.
Then my test.py had a valid output for:
import torch
print(torch.__version__) #1.13.0+cu116
print(torch.version.cuda) #11.6
print(torch.cuda.is_available()) #True
I got stuck afterwords with gridencoder failing to build because of file path problems?
Do in the terminal: >python >>> import torch >>> print(torch.__version__) etc.
In my case I got 1.13.1+cpu. And from >>> print(torch.version.cuda) I got None. So I then used:
pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu117 --force-reinstall since I use cuda 1.17.
Tried >>> print(torch.__version__) again in the Python sub-shell and it returned 1.13.1+cu117. >>> torch.cuda.is_available() also returned True.
If you encounter ModuleNotFoundError: No module named '_gridencoder', you may have to edit the "grid.py" file it is referencing which is in the folder "gridencoder". Open it up and on line 10 remove the underscore from import _gridencoder. Do the same thing for "freqencoder" and "raymarching". These imports reference C header files. The underscore might be how the Linux version of Python interprets those kinds of imports... I don't know. Just a guess.
Although, now that I have it actually try to run, I don't understand this error: TypeError: _morton3D.forward() takes 2 positional arguments but 4 were given
make sure you check pytorch website and get the correct torch and cuda version.
https://pytorch.org/get-started/previous-versions/
