[kwlug-disc] Have you played with locally installable and usable AI -> give a talk

[kwlug at c.nixc.us]

I've got a copy of Ollama running on my Gaming PC if you wanted to coordinate access to play with that I also have been meddling with something I found called OpenInterpreter which is probably a terrible idea to hook an AI directly into your computer as a system operator but I installed it into a virtual desktop inside of a Docker container accessible from a browser tab.




March 30, 2024 at 11:46 PM, "Paul Nijjar via kwlug-disc - kwlug-disc at kwlug.org" <ra+ufldbbsvjexizxayvqisniizm at simplelogin.co> wrote:



> 
> Another message that got rejected, but shouldn't have?
> 
> ----- Forwarded message from Alex Korobkin <korobkin at gmail.com> -----
> 
> Date: Sat, 30 Mar 2024 22:57:29 -0400
> 
> From: Alex Korobkin <korobkin at gmail.com>
> 
> To: KWLUG discussion <kwlug-disc at kwlug.org>
> 
> Subject: Re: [kwlug-disc] Have you played with locally installable and usable
> 
>  AI -> give a talk
> 
> I didn't do anything beyond this single article, but the idea is neat: you
> 
> download a single 4Gb file, make it executable and run it against images or
> 
> text to generate your AI output, entirely locally.
> 
> In this article they explain how to use it to rename photos in accordance
> 
> with photos content.
> 
> https://hackaday.com/2023/12/29/using-local-ai-on-the-command-line-to-rename-images-and-more/
> 
> On Sat, Mar 30, 2024 at 5:39 PM Mikalai Birukou <mb at 3nsoft.com> wrote:
> 
> > 
> > Probably will at some point. I entered this contest, and I'm waiting to
> > 
> >  hear back if I'm getting cloud access or hardware:
> > 
> >  https://www.hackster.io/contests/amd2023#challengeNav
> > 
> >  This is my proposal:
> > 
> >  https://www.hackster.io/contests/amd2023/hardware_applications/16336
> > 
> >  Contest is still open if you want to participate, but the hardware
> > 
> >  requests are finished.
> > 
> >  You can download most of the Open Source models here:
> > 
> >  https://huggingface.co/models
> > 
> >  Good article on how to get started:
> > 
> >  https://www.philschmid.de/fine-tune-llms-in-2024-with-trl
> > 
> >  The following heading in this llm fine tuning page makes me laugh and cry
> > 
> >  simultaneously, `with(xz)`:
> > 
> >  """
> > 
> >  2. Setup development environment
> > 
> >  """
> > 
> >  I am listening to podcast
> > 
> > > 
> > > https://oxide.computer/podcasts/oxide-and-friends/1692510 about Large
> > > 
> > >  Language Models, and idea comes up, if anyone of you have tried any
> > > 
> > >  locally usable LLM's, have trained it, tinkered with it, just make a
> > > 
> > >  "let me show" kind of talk. Seriously.
> > > 
> > 
> >  _______________________________________________
> > 
> >  kwlug-disc mailing list
> > 
> >  To unsubscribe, send an email to kwlug-disc-leave at kwlug.org
> > 
> >  with the subject "unsubscribe", or email
> > 
> >  kwlug-disc-owner at kwlug.org to contact a human being.
> > 
> 
> -- 
> 
> Alex.
> 
> ----- End forwarded message -----
> 
> -- 
> 
> Events: https://feeds.off-topic.kwlug.org/ 
> 
> Housing: https://unionsd.coop/
> 
> Blog: http://pnijjar.freeshell.org/
> 
> _______________________________________________
> 
> kwlug-disc mailing list
> 
> To unsubscribe, send an email to kwlug-disc-leave at kwlug.org
> 
> with the subject "unsubscribe", or email
> 
> kwlug-disc-owner at kwlug.org to contact a human being.
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kwlug.org/pipermail/kwlug-disc_kwlug.org/attachments/20240409/62fd34ec/attachment-0001.htm>