Pytorch version to match CUDA version

Desktop Feedback
,
1

I’ve just installed ComfyUI desktop and tried to run a first program and some key parts of the error message show:

{

CUDA error: no kernel image is available for execution on the device
CUDA kernel errors might be asynchronously reported at some other API call, so the stacktrace below might be incorrect.
For debugging consider passing CUDA_LAUNCH_BLOCKING=1
Compile with `TORCH_USE_CUDA_DSA` to enable device-side assertions.

Please install PyTorch with a following CUDA
configurations:  12.6 following instructions at
https://pytorch.org/get-started/locally/

Quadro M2200 with CUDA capability sm_52 is not compatible with the current PyTorch installation.
The current PyTorch install supports CUDA capabilities sm_70 sm_75 sm_80 sm_86 sm_90 sm_100 sm_120.

}

Is it possible operating under the ComfyUI Desktop to install specific versions of Pytorch to match the Cuda version, or do i need to use and environment, say by using miniconda, or Anaconda, to install the specific versions of Pytorch?

many thanks,

neil

2

So just to make this clear, ComfyUI Desktop appears to be asking to install an earlier version of Pytorch, so it is compatible with the older GPU on my laptop. Is this something that can be done on the ComfyUI Desktop, if so how? Or do i need to install ComfyUI in a different way to get it working on my machine with the older CPU.

ComfyUI claims to work with GPUs up to ten years old and my machine has the Nvidia 4GB GPU on a six-year old machine, should work, albeit slow.

thanks, Neil