kognat-docs

Is there no progress on this after a month? This makes custom CUDA acceleration operators for PyTorch in WSL2 Docker impossible. The footprint this leaves in the Deep Learning use...

I can do a test by adding RUN --gpus=all for now and see if this solves the issues we were having. It is frustrating that the documents like https://docs.nvidia.com/cuda/wsl-user-guide/index.html https://www.docker.com/blog/wsl-2-gpu-support-for-docker-desktop-on-nvidia-gpus/...

This feature is causing a performance issue when inferring a model, as I need to parse the multiArrayType by hand to get it back to being an image. I am...

Never mind , I will try v1.6.0 on CUDA 10.1 and CuDNN 7.6 and see if the memory use is better it runs with --network_resolution "-1x128" but it is jittery

You need to choose an older version of gcc/g++ to be compatible with nvcc in cudnn 7.x and cuda 10.x if you have an ampere card with only 6Gb of...

What could be causing an image to be output with black pixels? It feels like the geometry is there

@kikou do you have a good approach for diagnosing what is the cause of the noise circled in red? While just using `kick`?