search

AlexeyAB / darknet

YOLOv4 / Scaled-YOLOv4 / YOLO - Neural Networks for Object Detection (Windows and Linux version of Darknet )
http://pjreddie.com/darknet/
Other
21.66k stars 7.96k forks source link

Cudnn64_8.dll error on windows11, exception code: 0xc0000409 #8239

Open 1027663760 opened 2 years ago

1027663760 commented 2 years ago

CUDA-version: 11050 (11050), cuDNN: 8.3.0, CUDNN_HALF=1, GPU count: 1 CUDNN_HALF=1 OpenCV version: 4.5.4 yolov4-tiny_4l_spp 0 : compute_capability = 860, cudnn_half = 1, GPU: NVIDIA GeForce RTX 3060 net.optimized_memory = 1 mini_batch = 64, batch = 64, time_steps = 1, train = 1 layer filters size/strd(dil) input output 0 Create CUDA-stream - 0 Could not load library cudnn_cnn_infer64_8.dll. Error code 126 Please make sure cudnn_cnn_infer64_8.dll is in your library path! 1

cenit commented 2 years ago

did you install cudnn completely? it looks like you missed some files

1027663760 commented 2 years ago

did you install cudnn completely? it looks like you missed some files

I have compiled and used darknet for more than two years, it is impossible to make mistakes

cenit commented 2 years ago

newer cudnn versions are different than before. Besides cudnn64_8.dll there are other files to copy, maybe you missed them. The program is complaining that cudnn_cnn_infer64_8.dll is missing...

1027663760 commented 2 years ago

newer cudnn versions are different than before. Besides cudnn64_8.dll there are other files to copy, maybe you missed them. The program is complaining that cudnn_cnn_infer64_8.dll is missing...

I checked and all these files exist

cenit commented 2 years ago

https://stackoverflow.com/questions/69879188/could-not-load-library-cudnn-cnn-infer64-8-dll-error-code-126

looks like nvidia released a cudnn version with a bug can you try to downgrade your cudnn to the one for cuda 11.4 like they did in the link above? it should work on cuda 11.5 too, and fix the error you encountered

cenit commented 2 years ago

did you fix your problem downgrading cudnn?

1027663760 commented 2 years ago

did you fix your problem downgrading cudnn?

I solved it using an old version of cuda with cudnn