An NVIDIA GPU may be present on this machine, but a CUDA-enabled jaxlib is not installed. Falling back to cpu. [CpuDevice(id=0)] tensorflow, however, does recognise my GPU, and so I tried the suggestion from#15268to install using pip. I created a new environment and ran: pip install...
Seeing as you can run it manually, it might be an issue with the debugger not being able to attach to your miniconda interpreter, or because your code isn't being recognised as your own because you're running it with miniconda. Try debugging with'justMyCode': falseand see if that works...