RuntimeError: Error compiling objects for extension #1236

akiyamakodai · 2020-04-30T14:19:01Z

%%writefile setup.sh

maskrcnn_benchmark and coco api dependencies

pip install ninja yacs cython matplotlib tqdm opencv-python cityscapesscripts

follow PyTorch installation in https://pytorch.org/get-started/locally/

we give the instructions for CUDA 9.0

pip install -c pytorch pytorch-nightly torchvision cudatoolkit=9.0

git clone https://github.com/cocodataset/cocoapi.git
cd cocoapi/PythonAPI
python setup.py build_ext install
cd ../../

install apex

rm -rf apex
git clone https://github.com/NVIDIA/apex.git
cd apex
git pull
python setup.py install --cuda_ext --cpp_ext
cd ../

install PyTorch Detection

git clone https://github.com/facebookresearch/maskrcnn-benchmark.git
cd maskrcnn-benchmark

the following will install the lib with

symbolic links, so that you can modify

the files if you want and won't need to

re-build it

python setup.py build develop

!sh setup.sh

RuntimeError: Error compiling objects for extension

When I execute the above command, I get an error. why.
running on google colab

gerald0057 · 2020-05-06T11:46:50Z

I have the same problem. Please let me know if you can solve it. Thank you

gerald0057 · 2020-05-06T11:47:06Z

is there any updates?

gerald0057 · 2020-05-06T11:47:31Z

6 errors detected in the compilation of "/tmp/tmpxft_00009924_00000000-6_deform_conv_cuda.cpp1.ii".
ninja: build stopped: subcommand failed.
Traceback (most recent call last):
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1395, in _run_ninja_build
subprocess.run(
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/subprocess.py", line 512, in run
raise CalledProcessError(retcode, process.args,
subprocess.CalledProcessError: Command '['ninja', '-v']' returned non-zero exit status 1.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File "setup.py", line 59, in
setup(
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/setuptools/init.py", line 144, in setup
return distutils.core.setup(**attrs)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/core.py", line 148, in setup
dist.run_commands()
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/dist.py", line 966, in run_commands
self.run_command(cmd)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/dist.py", line 985, in run_command
cmd_obj.run()
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/command/build.py", line 135, in run
self.run_command(cmd_name)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/cmd.py", line 313, in run_command
self.distribution.run_command(command)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/dist.py", line 985, in run_command
cmd_obj.run()
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/setuptools/command/build_ext.py", line 87, in run
_build_ext.run(self)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/Cython/Distutils/old_build_ext.py", line 186, in run
_build_ext.build_ext.run(self)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/command/build_ext.py", line 340, in run
self.build_extensions()
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 580, in build_extensions
build_ext.build_extensions(self)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/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/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/command/build_ext.py", line 449, in build_extensions
self._build_extensions_serial()
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/command/build_ext.py", line 474, in _build_extensions_serial
self.build_extension(ext)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/setuptools/command/build_ext.py", line 208, in build_extension
_build_ext.build_extension(self, ext)
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/distutils/command/build_ext.py", line 528, in build_extension
objects = self.compiler.compile(sources,
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 414, in unix_wrap_ninja_compile
_write_ninja_file_and_compile_objects(
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1135, in _write_ninja_file_and_compile_objects
_run_ninja_build(
File "/home/gerald/miniconda3/envs/maskrcnn_benchmark/lib/python3.8/site-packages/torch/utils/cpp_extension.py", line 1413, in _run_ninja_build
raise RuntimeError(message)
RuntimeError: Error compiling objects for extension

akiyamakodai · 2020-05-10T04:37:08Z

I solved it.
Please check pytorch version, it can may go well.

gerald0057 · 2020-05-10T04:40:14Z

thank you very much!

sabine1993 · 2020-05-15T10:42:57Z

hey, I have the same problem but I cant figure out how i can change my pytorch version to the correct version. I have the following Version

Python 3.8.2 (default, May 6 2020, 09:02:42) [MSC v.1916 64 bit (AMD64)] :: Anaconda, Inc. on win32
Type "help", "copyright", "credits" or "license" for more information.

import torch
print(torch.version)
1.6.0.dev20200514

and i have the cuda driver 10.2

Can you please help me to solve the problem.
Thanks a lot in advance

gerald0057 · 2020-05-15T13:01:37Z

i finally chose Detectron2 and the installation (ubuntu20 cuda 10.2 pytorch 1.5stable) is very easy

vanessasidrim · 2020-05-19T15:52:32Z

I have the same problem.

I use:

Ubuntu 18.04
GCC 7.5
CUDA 10.1
Pytorch 1.5

I tried other versions of Pytorch, Cuda, GCC and Ubuntu. But the error remains.

Can you help me?!
Thanks.

akiyamakodai · 2020-05-20T07:07:51Z

I use :
pytorch 1.4
torchvision 0.5.0
cuda 10.1

I was able to run the program.

amsword · 2020-05-23T01:10:31Z

i was able to compile it with pytorch 1.4, but fails with teh same error on pytorch 1.5

vanessasidrim · 2020-05-23T01:46:10Z

In my case it worked when using version 0.5.0 of torchvision that comes with pytorch 1.4. In version 1.5 of pytorch the default torchdivion is version 0.6.0.

When using torchdivion 0.6.0 the same error occurs.

ebrahim31 · 2020-06-18T02:56:20Z

The current release works in windows 10 if you have torch==1.4.0 and torchvision ==0.5.0.

navidre · 2020-07-14T19:00:07Z

Works with PyTorch 1.4.0 and torchvision 0.5.0 for me. Not PyTorch 1.5.1 and torchvision 0.6.1!

HaiyiMei · 2020-07-17T12:20:11Z

Thank you guys!

Ubuntu 18.04
GCC 7.5
CUDA 10.1
Pytorch 1.4.0
torchvision 0.5.0
apex 0.1.0

This works perfectly well.

floschne · 2020-11-16T12:15:52Z

@HaiyiMei same for me! pyTorch 1.4.0 and torchvision 0.5.0 seems to be the only working version combination with CUDA 10.1

jetleg · 2021-04-20T05:11:55Z

Thanks for answering! I failed when I used torch.version=1.8.1 and torchvision.version=0.9.x, after I degraded my torch version to 1.5.0 it still failed. Only when I degraded my torch to 1.4.0 and torchvision to 0.5.0(at the same time I degrade my cuda to 10.1), I successfully compile 'python setup.py build develop'

oh-livia · 2021-04-30T20:31:39Z

I have torch=1.4.0, torchvision=0.5.0, and cuda=10.1 and I am still getting this error :(

MitraTj · 2021-05-01T16:44:57Z

I use:
python 3.8
pytorch '1.7.1.post2'
torchvision 0.10.0a0+730c5e1
Cuda 9

and I am still getting this error :(

tiankx2020 · 2021-07-30T07:13:11Z

I hava this save problem

DbrRoxane · 2021-08-23T09:39:24Z

I followed the instruction there and solved this issue by typing

cuda_dir="maskrcnn_benchmark/csrc/cuda"
perl -i -pe 's/AT_CHECK/TORCH_CHECK/' $cuda_dir/deform_pool_cuda.cu $cuda_dir/deform_conv_cuda.cu
# You can then run the regular setup command
python3 setup.py build develop

Now it works with CUDA 11.0 and Pytorch 1.7

Hope it helps!

Seal1026 · 2021-09-06T03:20:39Z

I followed the instruction there and solved this issue by typing
cuda_dir="maskrcnn_benchmark/csrc/cuda"
perl -i -pe 's/AT_CHECK/TORCH_CHECK/' $cuda_dir/deform_pool_cuda.cu $cuda_dir/deform_conv_cuda.cu
# You can then run the regular setup command
python3 setup.py build develop
Now it works with CUDA 11.0 and Pytorch 1.7

Hope it helps!

Thanks!!!!! it even works with PyTorch 1.9

ashok-arjun · 2021-09-09T01:44:01Z

Thank you guys!

Ubuntu 18.04

GCC 7.5

CUDA 10.1

Pytorch 1.4.0

torchvision 0.5.0

apex 0.1.0

This works perfectly well.

It's also important to add the python version. This only works for Python 1.6 for me - doesn't work for 1.7/1.8/1.9.
For later versions of python, one would need CUDA 11 and PyTorch 1.7 / 1.9

ojoaobrito · 2021-09-28T08:16:26Z

I followed the instruction there and solved this issue by typing
cuda_dir="maskrcnn_benchmark/csrc/cuda"
perl -i -pe 's/AT_CHECK/TORCH_CHECK/' $cuda_dir/deform_pool_cuda.cu $cuda_dir/deform_conv_cuda.cu
# You can then run the regular setup command
python3 setup.py build develop
Now it works with CUDA 11.0 and Pytorch 1.7

Hope it helps!

You are a legend! Many thanks!

tan199954 · 2021-12-29T09:03:00Z

I followed the instruction there and solved this issue by typing
cuda_dir="maskrcnn_benchmark/csrc/cuda"
perl -i -pe 's/AT_CHECK/TORCH_CHECK/' $cuda_dir/deform_pool_cuda.cu $cuda_dir/deform_conv_cuda.cu
# You can then run the regular setup command
python3 setup.py build develop
Now it works with CUDA 11.0 and Pytorch 1.7

Hope it helps!

it not works in my googlecolab huhu
CUDA 10.1, Pytorch 1.7.1

itzhuang · 2022-03-01T15:45:42Z

I followed the instruction there and solved this issue by typing
cuda_dir="maskrcnn_benchmark/csrc/cuda"
perl -i -pe 's/AT_CHECK/TORCH_CHECK/' $cuda_dir/deform_pool_cuda.cu $cuda_dir/deform_conv_cuda.cu
# You can then run the regular setup command
python3 setup.py build develop
Now it works with CUDA 11.0 and Pytorch 1.7

Hope it helps!

Awesome! Thank you!

Stevenhani10 · 2023-09-16T19:44:12Z

still have the same error
File "C:\Users\steve.conda\envs\deformable_detr\lib\site-packages\torch\utils\cpp_extension.py", line 1436, in _run_ninja_build
raise RuntimeError(message)
RuntimeError: Error compiling objects for extension

Any help?
thanks in advance

akiyamakodai closed this as completed May 11, 2020

ooe1123 mentioned this issue Sep 9, 2020

ADD Clothing Detection axinc-ai/ailia-models#185

Closed

oreo-lp mentioned this issue Nov 21, 2020

setup.py build_ext fail RuntimeError: Error compiling objects for extension skyhehe123/SA-SSD#79

Open

purvaten mentioned this issue Sep 15, 2021

Incompatibility with CUDA 11.0 facebookresearch/phosa#42

Closed

lkeab mentioned this issue Oct 12, 2021

Installation error lkeab/BCNet#63

Closed

shwetkm mentioned this issue May 2, 2022

AttributeError: module ‘torch.cuda’ has no attribute ‘amp’ #1327

Open

xiawj-hub mentioned this issue Jun 10, 2022

Install Error xiawj-hub/CTLIB#4

Closed

zhulf0804 mentioned this issue Sep 13, 2022

python setup.py develop compile error zhulf0804/PointPillars#17

Closed

wymanCV mentioned this issue Nov 29, 2022

安装配置问题 CityU-AIM-Group/SIGMA#23

Closed

tempdeltavalue mentioned this issue Mar 27, 2023

RuntimeError: Error compiling objects for extension while building extension ashawkey/torch-ngp#154

Open

Ambuj-Choudha mentioned this issue Jun 26, 2023

trying to compile detectron but getting error facebookresearch/detectron2#4992

Open

wangjuansan mentioned this issue Aug 22, 2023

Install issues CVMI-Lab/PLA#18

Closed

feizc mentioned this issue Feb 29, 2024

setup error pip install -e causal-conv1d feizc/DiS#6

Closed

zillur-av mentioned this issue Oct 1, 2024

RuntimeError: Error compiling objects for extension zillur-av/LVLane#17

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

RuntimeError: Error compiling objects for extension #1236

RuntimeError: Error compiling objects for extension #1236

akiyamakodai commented Apr 30, 2020

gerald0057 commented May 6, 2020

gerald0057 commented May 6, 2020

gerald0057 commented May 6, 2020

akiyamakodai commented May 10, 2020

gerald0057 commented May 10, 2020

sabine1993 commented May 15, 2020

gerald0057 commented May 15, 2020

vanessasidrim commented May 19, 2020

akiyamakodai commented May 20, 2020 •

edited

Loading

amsword commented May 23, 2020

vanessasidrim commented May 23, 2020 •

edited

Loading

ebrahim31 commented Jun 18, 2020

navidre commented Jul 14, 2020

HaiyiMei commented Jul 17, 2020

floschne commented Nov 16, 2020

jetleg commented Apr 20, 2021

oh-livia commented Apr 30, 2021

MitraTj commented May 1, 2021

tiankx2020 commented Jul 30, 2021

DbrRoxane commented Aug 23, 2021

Seal1026 commented Sep 6, 2021

ashok-arjun commented Sep 9, 2021

ojoaobrito commented Sep 28, 2021 •

edited

Loading

tan199954 commented Dec 29, 2021

itzhuang commented Mar 1, 2022

Stevenhani10 commented Sep 16, 2023

RuntimeError: Error compiling objects for extension #1236

RuntimeError: Error compiling objects for extension #1236

Comments

akiyamakodai commented Apr 30, 2020

maskrcnn_benchmark and coco api dependencies

follow PyTorch installation in https://pytorch.org/get-started/locally/

we give the instructions for CUDA 9.0

install apex

install PyTorch Detection

the following will install the lib with

symbolic links, so that you can modify

the files if you want and won't need to

re-build it

gerald0057 commented May 6, 2020

gerald0057 commented May 6, 2020

gerald0057 commented May 6, 2020

akiyamakodai commented May 10, 2020

gerald0057 commented May 10, 2020

sabine1993 commented May 15, 2020

gerald0057 commented May 15, 2020

vanessasidrim commented May 19, 2020

akiyamakodai commented May 20, 2020 • edited Loading

amsword commented May 23, 2020

vanessasidrim commented May 23, 2020 • edited Loading

ebrahim31 commented Jun 18, 2020

navidre commented Jul 14, 2020

HaiyiMei commented Jul 17, 2020

floschne commented Nov 16, 2020

jetleg commented Apr 20, 2021

oh-livia commented Apr 30, 2021

MitraTj commented May 1, 2021

tiankx2020 commented Jul 30, 2021

DbrRoxane commented Aug 23, 2021

Seal1026 commented Sep 6, 2021

ashok-arjun commented Sep 9, 2021

ojoaobrito commented Sep 28, 2021 • edited Loading

tan199954 commented Dec 29, 2021

itzhuang commented Mar 1, 2022

Stevenhani10 commented Sep 16, 2023

still have the same error File "C:\Users\steve.conda\envs\deformable_detr\lib\site-packages\torch\utils\cpp_extension.py", line 1436, in _run_ninja_build raise RuntimeError(message) RuntimeError: Error compiling objects for extension

akiyamakodai commented May 20, 2020 •

edited

Loading

vanessasidrim commented May 23, 2020 •

edited

Loading

ojoaobrito commented Sep 28, 2021 •

edited

Loading

still have the same error
File "C:\Users\steve.conda\envs\deformable_detr\lib\site-packages\torch\utils\cpp_extension.py", line 1436, in _run_ninja_build
raise RuntimeError(message)
RuntimeError: Error compiling objects for extension