question-mark
Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

FATAL ERROR DETECTED - CUDA NOT ENABLED

See original GitHub issue

CUDA 9.2 - I know this works with my OpenCV setup and with opencv4nodejs OpenBlas installed via sudo apt-get install libopenblas-dev face-recognition installed just via npm install - I couldn’t get custom dlib to work. The module installs fine as far as I can tell - it even mentions CUDA in the install and all seems ok!

My code works ok on my Macbook, although a little slow - I wanted to try it on my Ubuntu box with CUDA enabled GPUs

This is the error I get:

**************************** FATAL ERROR DETECTED ****************************

Error detected at line 120.
Error detected in file /home/analytics/code/cv-analytics-ws/node_modules/dlib-build/dlib/dlib/dlib/dnn/gpu_data.h.
Error detected in function float* dlib::gpu_data::device().

Failing expression was false.
CUDA NOT ENABLED

******************************************************************************

This is what I am doing:

const image = fr.cvImageToImageRGB(new fr.CvImage(cv.imread(tmpName).resizeToMax(1000)))

const detector = fr.FaceDetector();
const recognizer = fr.FaceRecognizer();
    
const targetSize = 150;
const faces = detector.detectFaces(image, targetSize) // this is what causes the error to happen

Any help would be great!

Issue Analytics

  • State:closed
  • Created 5 years ago
  • Comments:6 (3 by maintainers)

github_iconTop GitHub Comments

1reaction
justadudewhohackscommented, Oct 26, 2018

I think @legraphista was facing a similar issue. Also probably “DLIB_USE_CUDA” has to be defined in the define section of the gyp file.

Unfortunately I can’t test CUDA related stuff, so I am not sure how to properly compile this package with cuda.

0reactions
justadudewhohackscommented, Oct 26, 2018

You can declare preprocessor directives in the “defines” section. You probably compiled dlib with cuda, but the DLIB_USE_CUDA was missing, thus in some dlib header file included by face-recognition.js some code wrapped in #ifdef DLIB_USE_CUDA ... did not get compiled into the package, leading to a runtime error.

Read more comments on GitHub >

github_iconTop Results From Across the Web

NVLink error 74 fatal error detected - NVIDIA Developer Forums
Hi, I have a system with P100 NVLink *4, don't know when and how there's a NVLink error code 74 even freshly reboot...
Read more >
GPU driver and CUDA is not enabled and accessible by ...
Therefore, I followed instructions on Pytorch and installed Anaconda and Cuda. Then tried this: git clone --recursive https://github.com/pytorch ...
Read more >
Runtime assertion fail with CUDA NOT ENABLED, on ... - GitHub
I've created a shared lib, with cuda. -- Looking for cuDNN install... -- Configuring done -- Generating done -- Build files have been ......
Read more >
NVIDIA Library Requested but not Enabled for GPU ...
Hello all. I tested gpu acceleration, but it does not work. situtation is below. *** FATAL ***. An error has occurred when trying...
Read more >
CUDA-enabled mdrun reports error in CPU runs on non-GPU ...
With a CUDA build and either -nb cpu or -nb auto , failing to detect a GPU is something worthy of a one-line...
Read more >

github_iconTop Related Medium Post

No results found

github_iconTop Related StackOverflow Question

No results found

github_iconTroubleshoot Live Code

Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free

github_iconTop Related Reddit Thread

No results found

github_iconTop Related Hackernoon Post

No results found

github_iconTop Related Tweet

No results found

github_iconTop Related Dev.to Post

No results found

github_iconTop Related Hashnode Post

No results found