To solve Could not load dynamic library ‘cudnn64_8.dll’; dlerror: cudnn64_8.dll not found error follow below methods.
ERROR LOG
W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'cudnn64_8.dll'; dlerror: cudnn64_8.dll not found
W tensorflow/core/common_runtime/gpu/gpu_device.cc:1757] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...How to solve Could not load dynamic library ‘cudnn64_8.dll’; dlerror: cudnn64_8.dll not found ?
This cudnn64_8.dll file could not be found error is caused by the absence of a Microsoft Visual Studio C++ reproducible file in the CUDA folder.
Furthermore, there is no additional CUDA and Cudnn installation with PyTorch in a conda environment because conda instals both CUDA and Cudnn into that conda environment after typing conda install pytorch.
Alternative to Fix Could not load dynamic library ‘cudnn64_8.dll’; dlerror: cudnn64_8.dll not found
I think I can help you with providing a cudnn64_8.dll file (this is the download link: https://www.dll-files.com/cudnn64_8.dll.html). When you get the file, you can just put in your bin directory. For example, usually in windows platform, you can put it into C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.3\bin.
or
Sometimes you can find the missing cudnn64_8.dll file in the cuDNN folder. We can fix the issue by copying the cudnn64_8.dll file to the CUDA folder, i.e., C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.3\bin.
Hope the above solution works.
Also read :
AttributeError: ‘Rectangle’ object has no property ‘norm_hist’ Python Error
CUDA initialization: Unexpected error from cudaGetDeviceCount()
