I’ve been diving into the world of Home Assistant Voice and offline language models (LLMs), and it’s been quite an exciting journey. I recently acquired the Home Assistant Voice unit, and now I’m in the process of deciding on the best LLM setup for my needs. My home server is running smoothly, but it doesn’t have the luxury of a GPU, which has made me think about external solutions like a 3060 GPU setup. However, I’ve come across claims that the Mac mini might be outperforming the 3060 in certain scenarios, and I’m curious to hear others’ experiences with this.
If I were to go down the Mac mini route, how complicated would it be to have the mini handle the LLM while keeping my Windows server managing everything else? All my drives, including my music library, are already on the server, so maintaining that setup would be ideal if I decide to proceed with the Mac mini.
I’m also exploring the possibility of integrating this setup with my existing smart home ecosystem. For instance, I have a doorbell system using a Tradfri switch, and I’ve noticed that after some time (usually around an hour), the switch goes offline. This is problematic because it means my doorbell isn’t functional when I need it most. From the logs, it seems the device announces itself again after a few seconds, and pressing the button again triggers the automation. However, this workaround isn’t reliable enough for my needs.
I’m wondering if there are any best practices or configurations that can ensure the automation remains active at all times. Additionally, I’m curious to hear from others who might have encountered similar issues with Zigbee devices and how they resolved them.
On a lighter note, I’ve been impressed by the variety of Govee products available, especially their strip lights and oil diffusers. They’ve added a nice touch to my home, and I’m excited to see what new products they’ll release in the coming months.
If anyone has insights or experiences to share regarding offline LLM setups, Zigbee device connectivity, or Govee products, I’d love to hear from you! Let’s continue to explore and optimize our smart home setups together.