Detectron2 CUDNN_STATUS_NOT_SUPPORTED error when training a model in WSL.

[pbagwell-phioptics]

I'm getting this error when training a model in detectron2 on my PC in WSL2 UBUNTU 22.04. I don't get this error in Colab. I have installed Cuda 12.2 and cuDNN 8.9.6.
/miniconda3/envs/detectron2-env/lib/python3.10/site-packages/torch/autograd/graph.py:744: UserWarning: Plan failed with a cudnnException: CUDNN_BACKEND_EXECUTION_PLAN_DESCRIPTOR: cudnnFinalize Descriptor Failed cudnn_status: CUDNN_STATUS_NOT_SUPPORTED (Triggered internally at ../aten/src/ATen/native/cudnn/Conv_v8.cpp:919.)
return Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass

[pbagwell-phioptics]

Uninstalled PyTorch 2.3.0 and installed 2.2.2 and trained model without error.
conda install pytorch==2.2.2 torchvision==0.17.2 torchaudio==2.2.2 pytorch-cuda=12.1 -c pytorch -c nvidia

See: pytorch/pytorch#121834