Closed lgmventura closed 3 years ago
This seems like your CUDA toolkit is not installed correctly, try the following command to check it torch is compatible with CUDA
import torch
torch.Tensor([1.]).cuda()
You should not get warnings or errors for this code. Did you install cuda from nvidia? And is PyTorch from pip or conda, did you choose the correct cuda version?
I tried this to test the suggested command:
$ python -c "import torch; torch.Tensor([1.]).cuda()"
This is my result :octocat:
Traceback (most recent call last):
File "<string>", line 1, in <module>
File "/home/~/.local/lib/python3.8/site-packages/torch/cuda/__init__.py", line 172, in _lazy_init
torch._C._cuda_init()
RuntimeError: Found no NVIDIA driver on your system. Please check that you have an NVIDIA GPU and installed a driver from http://www.nvidia.com/Download/index.aspx
This means you don't have an Nvidia GPU or the driver is not installed correctly. You have to have an Nvidia GPU, not AMD or Integrated ones.
Run nvidia-smi
to check the GPU status, you should get something like this
Make sure there's a CUDA Version at the top right. And find the command to install torch on PyTorch 1.4.0 installation command, select the correct system, and CUDA version.
I reinstalled my nVidia drivers entirely (currently using nvidia 450, cuda v.11), created the conda environment called pytorch1.0.0 (conda create -n pytorch1.0.0
) and ran what I found in the pytorch's website:
conda install pytorch==1.0.0 torchvision==0.2.1 cuda100 -c pytorch
It now compiled and installed. I will test it later. Thank you so much for the support.
I'll close the issue for now.
running conda install pytorch==1.0.0 torchvision==0.2.1 cuda100 -c pytorch
got me:
failed
UnsatisfiableError: The following specifications were found
to be incompatible with the existing python installation in your environment:
Specifications:
- torchvision==0.2.1 -> python[version='>=2.7,<2.8.0a0|>=3.5,<3.6.0a0|>=3.6,<3.7.0a0|>=3.7,<3.8.0a0']
Your python: python=3.8
Is it a requirement to run the installation as root?
Getting DAIN running on Ubuntu 20.04 turns out to be difficult.
./build.sh
returns:
The above exception was the direct cause of the following exception: Traceback (most recent call last): File "setup.py", line 10, in <module> setup( File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/setuptools/__init__.py", line 153, in setup return distutils.core.setup(**attrs) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/core.py", line 148, in setup dist.run_commands() File "/home/nathaniel/anaconda3/lib/python3.8/distutils/dist.py", line 966, in run_commands self.run_command(cmd) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/dist.py", line 985, in run_command cmd_obj.run() File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/setuptools/command/install.py", line 67, in run self.do_egg_install() File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/setuptools/command/install.py", line 109, in do_egg_install self.run_command('bdist_egg') File "/home/nathaniel/anaconda3/lib/python3.8/distutils/cmd.py", line 313, in run_command self.distribution.run_command(command) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/dist.py", line 985, in run_command cmd_obj.run() File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/setuptools/command/bdist_egg.py", line 164, in run cmd = self.call_command('install_lib', warn_dir=0) File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/setuptools/command/bdist_egg.py", line 150, in call_command self.run_command(cmdname) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/cmd.py", line 313, in run_command self.distribution.run_command(command) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/dist.py", line 985, in run_command cmd_obj.run() File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/setuptools/command/install_lib.py", line 11, in run self.build() File "/home/nathaniel/anaconda3/lib/python3.8/distutils/command/install_lib.py", line 107, in build self.run_command('build_ext') File "/home/nathaniel/anaconda3/lib/python3.8/distutils/cmd.py", line 313, in run_command self.distribution.run_command(command) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/dist.py", line 985, in run_command cmd_obj.run() File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/setuptools/command/build_ext.py", line 79, in run _build_ext.run(self) File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/Cython/Distutils/old_build_ext.py", line 186, in run _build_ext.build_ext.run(self) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/command/build_ext.py", line 340, in run self.build_extensions() File "/home/nathaniel/.local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 670, in build_extensions build_ext.build_extensions(self) File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/Cython/Distutils/old_build_ext.py", line 195, in build_extensions _build_ext.build_ext.build_extensions(self) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/command/build_ext.py", line 449, in build_extensions self._build_extensions_serial() File "/home/nathaniel/anaconda3/lib/python3.8/distutils/command/build_ext.py", line 474, in _build_extensions_serial self.build_extension(ext) File "/home/nathaniel/anaconda3/lib/python3.8/site-packages/setuptools/command/build_ext.py", line 196, in build_extension _build_ext.build_extension(self, ext) File "/home/nathaniel/anaconda3/lib/python3.8/distutils/command/build_ext.py", line 528, in build_extension objects = self.compiler.compile(sources, File "/home/nathaniel/.local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 491, in unix_wrap_ninja_compile _write_ninja_file_and_compile_objects( File "/home/nathaniel/.local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1250, in _write_ninja_file_and_compile_objects _run_ninja_build( File "/home/nathaniel/.local/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1555, in _run_ninja_build raise RuntimeError(message) from e RuntimeError: Error compiling objects for extension
@courtens I create a new anaconda environment python=3.6 and run the installation successfully. python 3.8 env may be unsuitable for installing pytorch 1.0.0.
I managed to make it run on CUDA 8.6. You would need to edit compiler_args.py in order to compile with CUDA 11+ attached is a reference compiler_args file that works with sm86 and CUDA 11 compiler_args.txt
Following the installation instructions I ran out into an error that there must be a Conda environment called
pytorch1.0.0
after running./build.sh
– this is not mentioned in the compilation instructions. Well, I created a Conda environment calledpytorch1.0.0
and installed pytorch v1.4.0 (I saw on issue #118 that newer versions of pytorch aren't compatible, this is also not on the instructions) and ranbuild.sh
again. Now, I come to: