I wrote code using PyTorch on a computer that had an NVIDIA card, so it was easy to use CUDA. But now I'm programming on a Computer that has an AMD card and I don't know how to convert it. I know there's AMD's ROCm platform for this, but I haven't learned to use it yet, and apparently for the GPU installed here (Radeon 6600), it doesn't have the HIP SDK, which would "_translate_" the CUDA command to ROCm more easily. What can I do? And how do I do it? I've actually tried running by the HIP SDK, but then I figured out that it would not work.
