Update utils.py to support ollama GPU usage

[scott------]

trafficstars

When using the code prior to this patch ollama only uses CPU. With this patch ollama will use all of the GPU available and then fallback to using CPU (RAM).

[CLAassistant]

Thank you for your submission! We really appreciate it. Like many open source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution.
_{You have signed the CLA already but the status is still pending? Let us recheck it.}

[vvincent1234]

Hi, I don't have this problem on my machine. I'm not sure if this change will affect users with multiple GPUs. Can you give me some relevant links to show if this is useful?