Run SMOKE on cuda 11.x environment!
yhchang1120 opened this issue · 7 comments
I struggled to run SMOKE model on cuda 11.x environment for last 4 months and finally found a simple solution.
step 1. Visit https://github.com/jinfagang/DCNv2_latest(I appreciate you Jinfagang) step 2. Download all files and folders in src/ folder. step 3. Back to SMOKE directory and replace all files in csrc/ to step 2 files. step 4. rebuild codes: > python setup.py build developnow you can run SMOKE model on cuda 11.x environment.
This methods tested on following environment
- gpu : quadro rtx 6000
- driver version : 460.73.01
- cuda version : 11.0
- pytorch : 1.8.0
In future I'm planning to test on geforce RTX 3060.
If someone tested on RTX 30's series, share us results!
Hey,
I followed your solution and got the following error on RTX 3090:
subprocess.CalledProcessError: Command '['which', 'g++']' returned non-zero exit status 1.
GPU: RTX 3090
Driver version: 510.06
cuda version: 11.4
pytorch: 1.8.1
Entire error message:
Hey, I followed your solution and got the following error on RTX 3090:
subprocess.CalledProcessError: Command '['which', 'g++']' returned non-zero exit status 1.GPU: RTX 3090 Driver version: 510.06 cuda version: 11.4 pytorch: 1.8.1
Entire error message:
Oh nvm. I'm using docker and I forget to install gcc
Worked for me in google colab
running build
running build_ext
/home/awguitar/.local/lib/python3.8/site-packages/torch/utils/cpp_extension.py:387: UserWarning: Attempted to use ninja as the BuildExtension backend but we could not find ninja.. Falling back to using the slow distutils backend.
warnings.warn(msg.format('we could not find ninja.'))
Traceback (most recent call last):
File "setup.py", line 61, in
setup(
File "/usr/lib/python3/dist-packages/setuptools/init.py", line 144, in setup
return distutils.core.setup(**attrs)
File "/usr/lib/python3.8/distutils/core.py", line 148, in setup
dist.run_commands()
File "/usr/lib/python3.8/distutils/dist.py", line 966, in run_commands
self.run_command(cmd)
File "/usr/lib/python3.8/distutils/dist.py", line 985, in run_command
cmd_obj.run()
File "/usr/lib/python3.8/distutils/command/build.py", line 135, in run
self.run_command(cmd_name)
File "/usr/lib/python3.8/distutils/cmd.py", line 313, in run_command
self.distribution.run_command(command)
File "/usr/lib/python3.8/distutils/dist.py", line 985, in run_command
cmd_obj.run()
File "/usr/lib/python3/dist-packages/setuptools/command/build_ext.py", line 87, in run
_build_ext.run(self)
File "/usr/lib/python3/dist-packages/Cython/Distutils/old_build_ext.py", line 186, in run
_build_ext.build_ext.run(self)
File "/usr/lib/python3.8/distutils/command/build_ext.py", line 340, in run
self.build_extensions()
File "/home/awguitar/.local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 410, in build_extensions
self._check_cuda_version()
File "/home/awguitar/.local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 787, in _check_cuda_version
raise RuntimeError(CUDA_MISMATCH_MESSAGE.format(cuda_str_version, torch.version.cuda))
RuntimeError:
The detected CUDA version (11.4) mismatches the version that was used to compile
PyTorch (10.2). Please make sure to use the same CUDA versions.
####################
RTX3070
CUDA11.4
torch1.11
python3.8
Ubuntu20.4
I got the problem .
Hi, @liukr001
I think pytorch version 10.2 mismatches with with local CUDA version 11.4.
I found similar issue. Please refer to this link.
rusty1s/pytorch_sparse#189 (comment)
@yhchang1120 What results did you get in your environment?
Worked for me in google colab
Can you share your note book in ggloab with me,pls!