Issues connecting Ollama llama3 model with AI Agent

I have reselected the model and even selected qwen or mistral (already downloaded) and asterisk is still there, weird part is that I have downloaded a template and replaced the model by ollama (local) and it also is displaying the red asterisk but there the workflow is working properly even when the red asterisk, I have deep search into the issue and found this KB ( Unable to start workflow with the chat trigger node in n8n web gui ) it seems there is an issue with the chat trigger itself, something about N8N_RUNNERS_MODE=internal, it seems new version by default uses external and I need to change it to internal, however when I look for the entry in the .yml file it does not have any entry like that