https://ollama.ai/, this is what I've been using for over a year now, new models come out regularly and you just "ollama pull
<model ID>
" and then it's available to run locally. Then you can use docker to run https://www.openwebui.com/ locally, giving it a ChatGPT-style interface (but even better and more configurable and you can run prompts against any number of models you select at once.)
+1 for Mistral, they were the first (or one of the first) Apache open source licensed models. I run Mistral-7B and variant fine tunes locally, and they've always been really high quality overall. Mistral-Medium packed a punch (mid-size obviously) but it definitely competes with the big ones at least.
I put it on archive.md for anybody that couldn't see past the paywall like myself. (Sorry for the hijack but I wanted to see it quickly.) I'm not sure how everyone else is managing to read it for free tho.
Have to agree sadly. I searched and followed many people, still my feed is completely devoid of anything useful or interesting. I can keep digging but I feel like a 1% incremental gain from weeks of trying to set the network/profile up and giving zero results feels like a lost cause to some extent. I periodically check back, and it's more of the same unfortunately.
GrapheneOS offers such an auto-reboot feature (18 hours by default, but the users can set it between 10 minutes and 72 hours), while the iPhone picked up something similar with iOS 18.1 (Inactivity Reboot) last year.
I was referring primarily to things that are known to be good security practices and widely known and used already. Keeping data more secure at rest goes with the "don't trust anything or anyone" goal, and if not doing it on Android due to said trust or lack thereof, then GrapheneOS offers it too at least.
Confuses me that anybody would downvote you for this. I've made makeshift capacitors out of rolled aluminum foil. It's dumb, but it worked for what I wanted (triggering a trackpad via stepper motors for testing microcontroller code.) Plus I just wanted to see if it even worked. Life = science experiments.
You got me really curious so I prompted ChatGPT to give me a redditor style comment that's anti-Trump followed by another far-left redditor's reply to it and this is what it gave me:
Original Comment (Left-Leaning Anti-Trump Perspective):
"It's honestly mind-blowing how Trump still has a cult following after everything. The man was impeached twice, incited an insurrection, and openly grifts his supporters with fake fundraising schemes. And yet, somehow, people still think he's their 'savior.' The cognitive dissonance is off the charts."
Far-Left Redditor Response:
"Because the system is fundamentally broken. Liberals act like voting blue is enough when the entire establishment—from corporate Dems to the GOP—is complicit in maintaining capitalism's stranglehold. Trump is just a symptom of the rot. The real solution isn’t 'vote harder'; it’s dismantling the system that enabled him in the first place."
^ -- this seems to be disturbingly accurate, I can only imagine this being produced at automated scale at this point.
Yep, someone in another sub once said "I hope your plant dies" and their whole account was permbanned for "inciting violence". Against a plant. Hysterical. That site is cooked.
I just used that for export, but I have yet to try import on it. But I'm assuming it works well, it has good reviews as far as I remember.