Blake Hamm

Same issue for me! Would be a great feature IMO.

Awesome! I've got a working version here - https://github.com/blake-hamm/ansible-role-proxmox/tree/feature/ceph-rgw I will add docs before I open a PR and disclaimer it's not officially supported.

I have a (very opinionated) helm chart [here](https://github.com/blake-hamm/bhamm-lab/tree/feature/enhance-openwebui/kubernetes/charts/llm-models). It requires the [KubeElasti](https://kubeelasti.dev/), [LiteLLM](https://docs.litellm.ai/docs/) and [the prometheus operator](https://github.com/prometheus-operator/prometheus-operator) installed. You can add multiple models deployed with llama.cpp and it will update...

Agree this should be considered a bug. I believe this is a duplicate of #867

I have this same issue and unfortunately setting `feature.node.kubernetes.io/pci-0300_1002.present: null` does not work with argocd. The argocd application removes the line before it is rendered with helm.