Task Description (Required)

Ben has requested that the models we have deployed on Ollama have better diversity in terms of model size (currently both models are 7b and 8b). He suggested we switch to:

1) Gemma3:27b

2) llama3.2:3b

Modify the model list in https://github.com/redhat-ai-dev/rosa-gitops/edit/main/ollama/ollama-models-config.yaml to reflect the new model. ArgoCD should handle the automatic redeploying.