Build your own LLM AI on a Raspberry Pi

Поділитися
Вставка
  • Опубліковано 21 січ 2025

КОМЕНТАРІ • 22

  • @nutillion9334
    @nutillion9334 8 днів тому +1

    I can't run Ansible on windows, do I need to run it on the PI? Also what is the name of the software you are using at 24:32? Looks like a great project but I'm still very new to this

    • @Wise-Cat
      @Wise-Cat  3 дні тому +1

      I'm not a Windows user, though recently had cause to visit the dark side to look at installing Moodle on IIS (short answer, you don't want to do this...). Though I haven't yet tried, it's on my to-do list to look at installing Ansible and the LAMP/LEMP/LEPP stack on Windows using WSL. I'll try to do a video on that, so maybe that will help you.

    • @nutillion9334
      @nutillion9334 3 дні тому +1

      @ ok thank you!

  • @techtonictim
    @techtonictim 7 місяців тому +2

    Great video 👌 full of useful information.. thanks 🙏

    • @Wise-Cat
      @Wise-Cat  7 місяців тому

      Glad it was helpful! Thank you

  • @eminso2102
    @eminso2102 29 днів тому +1

    Is it possible I can run ollama offline integrated into a terabyte ssd for my raspberry pi? Or would I need to have to have a cloud or WiFi up?

    • @Wise-Cat
      @Wise-Cat  29 днів тому

      It can run completely offline, though if you're using the pi to serve the LLM to some other device, you'll need some kind of local networking. In the presentation we did, we demonstrated "wearable AI" as it could run, battery powered on the pi, with only its own ad-hoc network for connectivity. People connected to our pi and the Ollama from their smartphones. It was fun 😊

  • @rachitrastogi1422
    @rachitrastogi1422 7 місяців тому +2

    i have created my own llm so how i can deploy it on google cloud and use it on raspberry pi plzz tell me

    • @Wise-Cat
      @Wise-Cat  7 місяців тому +1

      Deploying an LLM to a cloud is easy using the docker-compose.yaml file in the Pillama repository, though getting it to work well with the cloud's infrastructure will be very case-by-case depending on how your particular cloud is setup. Using cloud-based GPUs etc will yield better results, so I'd suggest looking through the documentation for your cloud infrastructure provider. Sorry, I don't have a one-size-fits-all answer for this question.

  • @johnfinlayson7559
    @johnfinlayson7559 4 місяці тому +1

    Is it possible to run the tiny llama model without the open-webUI and docker? I want to do a tiny bit of reinforcement learning on the model and then put it in my pi, and integrate it onto my local website.

    • @Wise-Cat
      @Wise-Cat  Місяць тому

      It is indeed. Ollama can be run with a CLI interface directly on a pi. That said, the docker install is easy to install.

  • @IndraneelK
    @IndraneelK 5 місяців тому +1

    Would I be able to connect an lcd screen, microphone module, speaker module, and etc. and run the llm as a handheld device?
    Also what changes in the code would it require?

    • @Wise-Cat
      @Wise-Cat  4 місяці тому

      You could do this. There are modules for the raspberry pi that allow the connection of a monitor etc. In my project, I mainly used the pi as a server and accessed the LLM through a browser on a mobile phone connected to the pi's WiFi.

  • @slevinhyde3212
    @slevinhyde3212 6 місяців тому +2

    Great video, lots of fun here !
    Do you think the Pi AI kit could run a bigger model ?

    • @Wise-Cat
      @Wise-Cat  5 місяців тому +1

      Sure, I've tried bigger models without the AI kit, but the bigger they are the slower they run. Jeff Geerling did a demo on his channel using the AI Kit for visual recognition stuff and it was pretty amazing. I also wonder how its performance on LLMs would improve.

  • @galdakaMusic
    @galdakaMusic 7 місяців тому +1

    tinyllama and Coral or Hat AI (Hailo 8L)??

    • @Wise-Cat
      @Wise-Cat  7 місяців тому +3

      I would love to try that out someday. I currently don't have that stuff though. Saw some videos on Jeff Geerling's channel that were very interesting though.

    • @galdakaMusic
      @galdakaMusic 7 місяців тому +1

      Thanks

  • @Tech-iHub-y5j
    @Tech-iHub-y5j 4 місяці тому

    I would like to run this in docker swarm as a service on rpi's, any help there?

    • @Wise-Cat
      @Wise-Cat  Місяць тому

      Beyond my knowledge, I'm afraid. This was more a proof of concept than anything serious.

  • @ApidaeY
    @ApidaeY 7 місяців тому +2

    That is so painfully slow, doesnt look worth it

    • @Wise-Cat
      @Wise-Cat  7 місяців тому +8

      It depends on what your goal is. If you want a blazing fast AI, yeah it's not worth it. On the other hand, if you want to learn more about how these things work and perhaps how you can serve your own AI later on more impressive hardware (or on platforms like AWS and/or Azure) then it's totally worth it. Or you could do it to win a beer bet 😉
      In our case, we did it simply to show it CAN be done. To demonstrate to people that AI is not beyond them and to thus empower people. This could be the start of a journey for people who otherwise might be too self-doubting to take their first step. That makes it worth it to me. Oh, and it's cute and fun too...