• Reannlegge@lemmy.ca
    link
    fedilink
    English
    arrow-up
    0
    ·
    16 days ago

    That is why I am moving to self hosting as much as I can, I sail the high sees for media that is not Canadian or locked behind a pay wall that goes to fund US corps.

    I use a two pihole setup to block ads on regular internet (with wireguard to VPN into my LAN on the rare occasion I leave my LAN) and I use unwatched to watch youtube on my ios devices, I plan on moving to another thing once my firewall is setup (with openwrt). I use Home Assistant and Homebridge to get away from sending my data out, as much as I can. I stopped using US based social media to move to the open standards, I moved my domains to Canadian hosts. The fall back DNS for my pihole unbound is CIRA.

    The only US service I am paying for on the series of tubes is Claude, because it is helping me get my home lab working better, I will soon be stopping that. I want to get a used laptop from my local university to run Ollama, I want to get a used laptop to install Debian on to replace my iPad, I want to get a new phone to install a open source phone OS, but money. What I would really like to get would be a raspberry pi phone and tablet, but that would mean they would have to make one first!

    • Septimaeus@infosec.pub
      link
      fedilink
      English
      arrow-up
      0
      ·
      15 days ago

      Just a heads up that if you want to host full local models, you may want to try LM Studio or vMLX with models straight from hugging-face. Ollama is popular but they privilege their proprietary cloud MCP services and won’t even let you download web-enabled models unless you agree to use their API, which defeats your on-prem threat model.

      • Reannlegge@lemmy.ca
        link
        fedilink
        English
        arrow-up
        0
        ·
        15 days ago

        Thanks for letting me know, I am still several months away from getting any hardware for a local LLM.