AI system not responding

[GON-PHISHIN]

Good Morning,

I have a small issue, probably on my side. I have a M1 Mac running Metal Ubuntu with Ollama.
The model is the llama 3.1 8b, I have opened the firewall port I can see the llama and connect remotely via ip:11434.
I have it attached to HA AI_Automation_Suggester, no errors except the error 500.
If I run the prompt locally on llama it maxes out the machine and it tries to explain everything as it is missing context.
If I run it from the dev tolls actions I see network activity incoming but no CPU usage.

Any thoughts? What did I miss?

Gon-Phishin

[GON-PHISHIN]

I ended up deleting the connection. reseting the LLM on the remote system with a new model. then reconnected. It seems as though it was a configuration issue between the two. it now connects and responds properly

[ITSpecialist111]

Good to see you over come this. Apologies I've not had the time to investigate. How is this working now?

[GON-PHISHIN]

Thank You,
I'm hoping to get a smarter AI. LOL
I ask it for suggestions and it comes back with odd ones.
Like I have a door switch to a few closets and they turn lights on and of depending on the door switch.
the AI is telling me to turn the light off and on based on sunset or sunrise.
or trying to adjust the brightness of a smart outlet connected to a tumbler in my shop.
or Adjust brightness of network_closet_nighttime based on pantry door status
I think I'll have to work on areas to eliminate some of the confusion for the AI.

We're defiantly not at the Jarvis level yet.
Can everyone who is using it post there instructions they find are working better?