Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Unable to run "darknet" command #92

Open
nbandaru1h opened this issue Nov 18, 2024 · 14 comments
Open

Unable to run "darknet" command #92

nbandaru1h opened this issue Nov 18, 2024 · 14 comments
Assignees
@stephanecharette
Labels
question Further information is requested

Comments

@nbandaru1h
Copy link

I have followed build instructions just as given.

make -j4 package

-- Darknet v3.0-195-gb4b25d33-dirty
-- CUDA detected. Darknet will use the GPU.
-- Found cuDNN library: /usr/local/cuda/lib64/libcudnn.so
-- Found cuDNN include: /usr/local/cuda/include
-- Hardware is 32-bit or 64-bit, and seems to be Intel or AMD:  x86_64
-- Found Threads 
CMake Warning (dev) at /usr/local/lib/cmake/opencv4/OpenCVConfig.cmake:86 (find_package):
  Policy CMP0146 is not set: The FindCUDA module is removed.  Run "cmake
  --help-policy CMP0146" for policy details.  Use the cmake_policy command to
  set the policy and suppress this warning.

Call Stack (most recent call first):
  /usr/local/lib/cmake/opencv4/OpenCVConfig.cmake:108 (find_host_package)
  CM_dependencies.cmake:126 (FIND_PACKAGE)
  CMakeLists.txt:33 (INCLUDE)
This warning is for project developers.  Use -Wno-dev to suppress it.

-- Found OpenCV 4.10.0
-- Found OpenMP 
-- Enabling AVX and SSE optimizations.
-- Making an optimized release build.
-- Setting up DARKNET OBJ
-- Setting up DARKNET LIB
-- Setting up DARKNET CLI
-- Setting up DARKNET examples
-- Configuring done (0.1s)
-- Generating done (0.1s)
-- Build files have been written to: /home/nbandaru1/src/darknet/build
[  1%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/avgpool_layer_kernels.cu.o
[  3%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/col2im.cpp.o
[  3%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/blas_kernels.cu.o
[  4%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/box.cpp.o
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
[  5%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/col2im_kernels.cu.o
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
[  7%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/connected_layer.cpp.o
[  8%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/convolutional_kernels.cu.o
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
[  9%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/convolutional_layer.cpp.o
[ 10%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/cost_layer.cpp.o
[ 11%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/crnn_layer.cpp.o
[ 12%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/crop_layer_kernels.cu.o
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
[ 14%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/dark_cuda.cpp.o
[ 15%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet.cpp.o
[ 16%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_args_and_parms.cpp.o
[ 17%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_cfg.cpp.o
[ 18%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_cfg_and_state.cpp.o
[ 20%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_enums.cpp.o
[ 21%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_format_and_colour.cpp.o
[ 22%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_image.cpp.o
[ 23%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_keypoints.cpp.o
[ 24%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_network.cpp.o
[ 25%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/darknet_utils.cpp.o
[ 27%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/data.cpp.o
[ 28%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/detector.cpp.o
[ 29%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/dropout_layer.cpp.o
[ 30%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/dropout_layer_kernels.cu.o
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
[ 31%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/dump.cpp.o
[ 32%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/gaussian_yolo_layer.cpp.o
[ 34%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/gemm.cpp.o
[ 35%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/im2col.cpp.o
[ 36%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/im2col_kernels.cu.o
[ 37%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/image_opencv.cpp.o
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
[ 38%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/layer.cpp.o
[ 40%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/list.cpp.o
[ 41%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/lstm_layer.cpp.o
[ 42%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/matrix.cpp.o
[ 43%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/maxpool_layer.cpp.o
[ 44%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/maxpool_layer_kernels.cu.o
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
[ 45%] Building CUDA object src-lib/CMakeFiles/darknetobjlib.dir/network_kernels.cu.o
nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used
[ 47%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/nightmare.cpp.o
[ 48%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/option_list.cpp.o
[ 49%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/region_layer.cpp.o
[ 50%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/reorg_layer.cpp.o
[ 51%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/rnn_layer.cpp.o
[ 52%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/route_layer.cpp.o
[ 54%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/sam_layer.cpp.o
[ 55%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/scale_channels_layer.cpp.o
[ 56%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/shortcut_layer.cpp.o
[ 57%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/softmax_layer.cpp.o
[ 58%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/tree.cpp.o
[ 60%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/upsample_layer.cpp.o
[ 61%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/utils.cpp.o
[ 62%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/weights.cpp.o
[ 63%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/yolo_layer.cpp.o
[ 64%] Building CXX object src-lib/CMakeFiles/darknetobjlib.dir/yolo_v2_class.cpp.o
[ 72%] Built target darknetobjlib
[ 74%] Building CXX object src-cli/CMakeFiles/darknetcli.dir/darknet_cli.cpp.o
[ 75%] Linking CXX shared library libdarknet.so
[ 76%] Linking CXX executable darknet
[ 76%] Built target darknet
[ 77%] Building CXX object src-examples/CMakeFiles/darknet_02_display_annotated_images.dir/darknet_02_display_annotated_images.cpp.o
[ 78%] Building CXX object src-examples/CMakeFiles/darknet_01_inference_images.dir/darknet_01_inference_images.cpp.o
[ 80%] Building CXX object src-examples/CMakeFiles/darknet_03_display_videos.dir/darknet_03_display_videos.cpp.o
[ 81%] Linking CXX executable darknet_01_inference_images
[ 82%] Linking CXX executable darknet_03_display_videos
[ 83%] Linking CXX executable darknet_02_display_annotated_images
[ 83%] Built target darknet_01_inference_images
[ 84%] Building CXX object src-examples/CMakeFiles/darknet_04_process_videos.dir/darknet_04_process_videos.cpp.o
[ 84%] Built target darknet_03_display_videos
[ 85%] Building CXX object src-examples/CMakeFiles/darknet_05_process_videos_multithreaded.dir/darknet_05_process_videos_multithreaded.cpp.o
[ 85%] Built target darknet_02_display_annotated_images
[ 87%] Building CXX object src-examples/CMakeFiles/darknet_06_images_to_json.dir/darknet_06_images_to_json.cpp.o
[ 87%] Built target darknetcli
[ 88%] Building CXX object src-examples/CMakeFiles/darknet_07_display_rtsp_stream.dir/darknet_07_display_rtsp_stream.cpp.o
[ 89%] Linking CXX executable darknet_04_process_videos
[ 90%] Linking CXX executable darknet_05_process_videos_multithreaded
[ 90%] Built target darknet_04_process_videos
[ 91%] Building CXX object src-examples/CMakeFiles/darknet_08_display_webcam.dir/darknet_08_display_webcam.cpp.o
[ 92%] Linking CXX executable darknet_07_display_rtsp_stream
[ 92%] Built target darknet_05_process_videos_multithreaded
[ 94%] Building CXX object src-examples/CMakeFiles/darknet_09_keypoints_images.dir/darknet_09_keypoints_images.cpp.o
[ 94%] Built target darknet_07_display_rtsp_stream
[ 95%] Building CXX object src-examples/CMakeFiles/darknet_10_keypoints_videos.dir/darknet_10_keypoints_videos.cpp.o
[ 96%] Linking CXX executable darknet_06_images_to_json
[ 97%] Linking CXX executable darknet_08_display_webcam
[ 97%] Built target darknet_06_images_to_json
[ 97%] Built target darknet_08_display_webcam
[ 98%] Linking CXX executable darknet_09_keypoints_images
[100%] Linking CXX executable darknet_10_keypoints_videos
[100%] Built target darknet_09_keypoints_images
[100%] Built target darknet_10_keypoints_videos
Run CPack packaging tool...
CPack: Create package using DEB
CPack: Install projects
CPack: - Run preinstall target for: Darknet
CPack: - Install project: Darknet []
CPack: Create package
CPackDeb: - Generating dependency list
CPack: - package: /home/nbandaru1/src/darknet/build/darknet-3.0.195-Linux.deb generated.

make -j4 package

[ 72%] Built target darknetobjlib
[ 74%] Built target darknet
[ 76%] Built target darknetcli
[ 78%] Built target darknet_02_display_annotated_images
[ 82%] Built target darknet_01_inference_images
[ 83%] Built target darknet_03_display_videos
[ 85%] Built target darknet_04_process_videos
[ 91%] Built target darknet_06_images_to_json
[ 91%] Built target darknet_07_display_rtsp_stream
[ 92%] Built target darknet_05_process_videos_multithreaded
[ 95%] Built target darknet_08_display_webcam
[ 97%] Built target darknet_10_keypoints_videos
[100%] Built target darknet_09_keypoints_images
Run CPack packaging tool...
CPack: Create package using DEB
CPack: Install projects
CPack: - Run preinstall target for: Darknet
CPack: - Install project: Darknet []
CPack: Create package
CPackDeb: - Generating dependency list
CPack: - package: /home/nbandaru1/src/darknet/build/darknet-3.0.195-Linux.deb generated.

sudo dpkg -i darknet-3.0.195-Linux.deb

(Reading database ... 686541 files and directories currently installed.)
Preparing to unpack darknet-3.0.195-Linux.deb ...
Unpacking darknet (3.0.195) over (3.0.195) ...
Setting up darknet (3.0.195) ...
Processing triggers for libc-bin (2.35-0ubuntu3.8) ...

nvcc --version

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Tue_Aug_15_22:02:13_PDT_2023
Cuda compilation tools, release 12.2, V12.2.140
Build cuda_12.2.r12.2/compiler.33191640_0

nvidia-smi

Mon Nov 18 15:21:08 2024       
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 535.183.01             Driver Version: 535.183.01   CUDA Version: 12.2     |
|-----------------------------------------+----------------------+----------------------+
| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |
|                                         |                      |               MIG M. |
|=========================================+======================+======================|
|   0  NVIDIA RTX A3000 Laptop GPU    Off | 00000000:01:00.0 Off |                  N/A |
| N/A   45C    P8              13W /  35W |     11MiB /  6144MiB |      0%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+
                                                                                         
+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
|    0   N/A  N/A      1540      G   /usr/lib/xorg/Xorg                            4MiB |
+---------------------------------------------------------------------------------------+

ERROR:

~/src/darknet/build$ darknet

CUDA status error: /home/nbandaru1/src/darknet/src-lib/dark_cuda.cpp, cuda_set_device(), line #29

* * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* Errno 9: Bad file descriptor
* Error location: /home/nbandaru1/src/darknet/src-lib/dark_cuda.cpp, cuda_set_device(), line #29
*** Error message:  current CUDA error: status=999 cudaErrorUnknown: unknown error**
* Thread #139882553249792: main darknet thread
* Version v3.0-195-gb4b25d33-dirty built on Nov 18 2024 15:18:42
* * * * * * * * * * * * * * * * * * * * * * * * * * * * *
backtrace (8 entries):
1/8: darknet(_Z13log_backtracev+0x36) [0x55884b2def96]
2/8: darknet(_Z19darknet_fatal_errorPKcS0_iS0_z+0x673) [0x55884b2e0833]
3/8: darknet(+0x70229) [0x55884b21e229]
4/8: darknet(_Z25check_cuda_error_extended9cudaErrorPKcS1_i+0x105) [0x55884b21e3a5]
5/8: darknet(main+0x1a4) [0x55884b1edd44]
6/8: /lib/x86_64-linux-gnu/libc.so.6(+0x29d90) [0x7f3943570d90]
7/8: /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0x80) [0x7f3943570e40]
8/8: darknet(_start+0x25) [0x55884b1f1585]
known threads:  1
1/1: #139882553249792: main darknet thread
Aborted (core dumped)
@nbandaru1h nbandaru1h added the question Further information is requested label Nov 18, 2024
@stephanecharette
Copy link
Collaborator

See the many warnings in your build:

nvcc warning : Cannot find valid GPU for '-arch=native', default arch is used

The NVIDIA compiler did not detect a GPU.

@stephanecharette
Copy link
Collaborator

Also, your version field tells me you have local modifications to the repo:

Version v3.0-195-gb4b25d33-dirty

Can you run git status to show what has been locally modified? Also try git diff and see if the changes are something you remember doing.

@nbandaru1h
Copy link
Author

I haven't modified anything actually - just cloned your repo and built it
When I encountered this error, I added a flag to CMakeLists.txt so it could maybe recognize the CUDA_ARCH_BIN. But of course, it didn't work.

git status
On branch master
Your branch is up to date with 'origin/master'.

Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git restore <file>..." to discard changes in working directory)
	modified:   CMakeLists.txt

no changes added to commit (use "git add" and/or "git commit -a")

@stephanecharette
Copy link
Collaborator

You could of course try to hard-code the GPU to use and see if that makes a difference. See DARKNET_CUDA_ARCHITECTURES in CM_dependencies.cmake.

But the fact that nvcc is reporting that it cannot find a GPU, and then Darknet fails to set the GPU to use, is a pretty big hint that something is wrong with the NVIDIA drivers.

Two different things are telling you a GPU cannot be accessed (or found?).

@nbandaru1h
Copy link
Author

Hmm....that's funny because when I run darknet training with Alexey repo normally it would find the GPU and train with it. But, I agree that there's something wrong.
What is the valid syntax for the architecture? Compute_8.6 is invalid according to this build error:

nvcc fatal   : Unsupported gpu architecture 'compute_8.6'
make[2]: *** [src-lib/CMakeFiles/darknetobjlib.dir/build.make:111: src-lib/CMakeFiles/darknetobjlib.dir/activation_kernels.cu.o] Error 1
make[2]: *** Waiting for unfinished jobs....
make[2]: Leaving directory '/home/nbandaru1/src/darknet/build'
make[1]: *** [CMakeFiles/Makefile2:262: src-lib/CMakeFiles/darknetobjlib.dir/all] Error 2
make[1]: Leaving directory '/home/nbandaru1/src/darknet/build'
make: *** [Makefile:159: all] Error 2

@stephanecharette
Copy link
Collaborator

stephanecharette commented Nov 18, 2024 •
edited
Loading

What is the valid syntax for the architecture?

See the file I mention. Around lines 42-44. It can also be set as a command-line parameter if you prefer.

@nbandaru1h
Copy link
Author

nbandaru1h commented Nov 18, 2024 •
edited
Loading

That fixed the build, but the error persists:

darknet
CUDA status error: /home/nbandaru1/src/darknet/src-lib/dark_cuda.cpp, cuda_set_device(), line #29

* * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* Errno 9: Bad file descriptor
* Error location: /home/nbandaru1/src/darknet/src-lib/dark_cuda.cpp, cuda_set_device(), line #29
* Error message:  current CUDA error: status=999 cudaErrorUnknown: unknown error
* Thread #139981998850048: main darknet thread
* Version v3.0-195-gb4b25d33-dirty built on Nov 18 2024 16:04:07
* * * * * * * * * * * * * * * * * * * * * * * * * * * * *
backtrace (8 entries):
1/8: darknet(_Z13log_backtracev+0x36) [0x55a718568f96]
2/8: darknet(_Z19darknet_fatal_errorPKcS0_iS0_z+0x673) [0x55a71856a833]
3/8: darknet(+0x70229) [0x55a7184a8229]
4/8: darknet(_Z25check_cuda_error_extended9cudaErrorPKcS1_i+0x105) [0x55a7184a83a5]
5/8: darknet(main+0x1a4) [0x55a718477d44]
6/8: /lib/x86_64-linux-gnu/libc.so.6(+0x29d90) [0x7f5068d70d90]
7/8: /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0x80) [0x7f5068d70e40]
8/8: darknet(_start+0x25) [0x55a71847b585]
known threads:  1
1/1: #139981998850048: main darknet thread
Aborted (core dumped)

I do have other repositories of darknet installed. Could that be causing issues?

@stephanecharette
Copy link
Collaborator

stephanecharette commented Nov 18, 2024 •
edited
Loading

After you built it, did you remember to install it? If not, you're still running the old one you built before you hard-coded the GPU value.

@stephanecharette
Copy link
Collaborator

stephanecharette commented Nov 18, 2024 •
edited
Loading

Do you have multiple GPUs? If you look at the function cuda_set_device() in dark_cuda.cpp on line 25-29, you'll see there really is not much to it. It is a single CUDA call to set which GPU to use, and then checking the status of that call to ensure it was successful.

  1. What is the output of darknet --version?
  2. Does it see the GPU?
  3. Does a reboot help?

@nbandaru1h
Copy link
Author

nbandaru1h commented Nov 19, 2024 •
edited
Loading

Should I be doing a make install as well?

~/src/darknet/build$ make install
[ 72%] Built target darknetobjlib
[ 74%] Built target darknet
[ 76%] Built target darknetcli
[ 78%] Built target darknet_01_inference_images
[ 81%] Built target darknet_02_display_annotated_images
[ 83%] Built target darknet_03_display_videos
[ 85%] Built target darknet_04_process_videos
[ 88%] Built target darknet_05_process_videos_multithreaded
[ 90%] Built target darknet_06_images_to_json
[ 92%] Built target darknet_07_display_rtsp_stream
[ 95%] Built target darknet_08_display_webcam
[ 97%] Built target darknet_09_keypoints_images
[100%] Built target darknet_10_keypoints_videos
Install the project...
-- Install configuration: "Release"
-- Up-to-date: /opt/darknet/cfg/Gaussian_yolov3_BDD.cfg
CMake Error at cfg/cmake_install.cmake:54 (file):
  file INSTALL cannot set permissions on
  "/opt/darknet/cfg/Gaussian_yolov3_BDD.cfg": Operation not permitted.
Call Stack (most recent call first):
  cmake_install.cmake:52 (include)
make: *** [Makefile:120: install] Error 1


:~/src/darknet/build$ darknet --version
CUDA status error: /home/nbandaru1/src/darknet/src-lib/dark_cuda.cpp, cuda_set_device(), line #29

* * * * * * * * * * * * * * * * * * * * * * * * * * * * *
* Errno 9: Bad file descriptor
* Error location: /home/nbandaru1/src/darknet/src-lib/dark_cuda.cpp, cuda_set_device(), line #29
* Error message:  current CUDA error: status=999 cudaErrorUnknown: unknown error
* Thread #140468877402112: main darknet thread
* Version v3.0-195-gb4b25d33-dirty built on Nov 19 2024 10:55:51
* * * * * * * * * * * * * * * * * * * * * * * * * * * * *
backtrace (8 entries):
1/8: darknet(_Z13log_backtracev+0x36) [0x55b001753f96]
2/8: darknet(_Z19darknet_fatal_errorPKcS0_iS0_z+0x673) [0x55b001755833]
3/8: darknet(+0x70229) [0x55b001693229]
4/8: darknet(_Z25check_cuda_error_extended9cudaErrorPKcS1_i+0x105) [0x55b0016933a5]
5/8: darknet(main+0x1a4) [0x55b001662d44]
6/8: /lib/x86_64-linux-gnu/libc.so.6(+0x29d90) [0x7fc1c6f70d90]
7/8: /lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0x80) [0x7fc1c6f70e40]
8/8: darknet(_start+0x25) [0x55b001666585]
known threads:  1
1/1: #140468877402112: main darknet thread
Aborted (core dumped)

@nbandaru1h
Copy link
Author

Rebooting helped:

/src$ darknet --version
Darknet V3 "Jazz" v3.0-195-gb4b25d33-dirty
CUDA runtime version 12020 (v12.2), driver version 12020 (v12.2)
cuDNN version 12020 (v8.9.7), use of half-size floats is ENABLED
=> 0: NVIDIA RTX A3000 Laptop GPU [#8.6], 5.8 GiB
OpenCV v4.10.0-dev, LinuxMint 21.2

@nbandaru1h
Copy link
Author

Thanks for your help, appreciate it!

@stephanecharette
Copy link
Collaborator

CMake Error at cfg/cmake_install.cmake:54 (file):
file INSTALL cannot set permissions on
"/opt/darknet/cfg/Gaussian_yolov3_BDD.cfg": Operation not permitted.

Sorry, never heard of this error before. I have no idea what you've done.

@stephanecharette
Copy link
Collaborator

Should I be doing a make install as well?

Definitely not. The build steps are very clearly indicated in the readme.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@stephanecharette stephanecharette

Labels
question Further information is requested
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@stephanecharette @nbandaru1h