Well yeah, self-instruct is how a lot of these models are trained. Bootstrap training data off a larger model and fine-tune a pre-existing model of that.
It’s similar but different.
Well yeah, self-instruct is how a lot of these models are trained. Bootstrap training data off a larger model and fine-tune a pre-existing model of that.
It’s similar but different.
Yeah Lemmy has an unfriendly community. UI is really hard and I know exactly what you mean when you say everything has a reason behind it.
FWIW, Ima migrate my personal private Lemmy to photon because I think it looks great.
I don’t think other commenters bothered to watch the trial.
He was acting foolish but he certainly didn’t commit pre-meditated murder.
Dokuwiki + wireguard would be different but satisfy since if those needs, try it out of you have time.
Wife and I have been unemployed for nearly a year. We’re in a white collar recession so it’s gonna be brutal for a little while. Not much you can do really, it’s really hard right now.
Labouring / trades seem like the ticket tbh.
Which I can totally understand, but I would like to spend more time with my family and less time writing code.
This also allows me to iterate faster and identify useful ideas That justify deeper effort.
Yeah, that’s always a risk, but as you said, humans make mistakes too. And if you change your approach to software development by writing more tests and using strict interfaces or type annotations, etc., it is pretty reliable and definitely saves time.
They can also be really good for quickly writing code if you line up a whole bunch of tests and line up all the types and then copy and paste that a few times, maybe with a macro in Vim.
The LLM will fill in the middle correctly, like 90% of the time. Compare it in git, make sure the tests pass, and then that’s an extra 20 minutes I get to spend with my wife and kids.
It would be on the order of aN intensive video game, maybe. Depends on the size of the model, etc.
Training is definitely expensive but you are right in that it’s a one-time cost.
Overall, the challenge is that it’s very inefficient. To use a machine learning algorithm to do something that could be implemented deductively is not ideal (On the other hand, if it saves human effort…)
To a degree, trained models can also be retrained on newer data (eg freezing layers, LoRa, GaLore, Hypernetworks etc). Also newer data can be injected into a prompt to make sure that the responses are aligned with newer versions of software, for example.
The electricity consumption is a concern, but it’s probably not going to be the end of the world.
Yeah but wine helps lol
We mostly use it privately, there are also a handful of software communities too that takes advantage of bridging.
Personally, I don’t care about Nazis, they come for the same reason I do, privacy and place to speak. I don’t have to let there negative disposition color the software.
What’s your desktop environment? I’m pretty sure hyperland and sway will give a json output of open Windows.
You could parse that with jq and pipe it into fzf or dmenu?
Not quite the same as the clicking but probably just as quick.
Quartz or mkdocs
Perplexica is interesting too, but it uses a moderate amount of ram because of elastic search.
And of course you need to have ollama running
I’ve always had an easier time jumping into an oop code base, then eg a lisp one.
I hear people when they say they don’t want their data mixed in with their logic but The pressure to structure code Is very nice.
I just wish people weren’t so aggressive with politics.
I’ve noticed a severe lack of perspective and empathy in these communities which has greatly deterred me from engaging.
Reddit was bad as well, but it seemed to attract a more rounded and informed community at least in the early days. Probably a function of fragmentation more than anything.
I suppose the problem that I had with Media Wiki is that every update would break extensions. Particularly mathjax and semantic media Wiki. I too amusing it with Docker which helps a lot.
So docuicki has a recent pages view which is really good and lists the user that made the edit. That’s what we use for a feed. There’s also an RSS plug-in that will display other feeds which is kind of nice if you want to discuss other articles.
We create Journal pages that link out to pages for events etc. The events are also linked to from a start page. We display the backlinks using the footer plugin.
Whilst it’s a bit different from social media in that there is no feed, it’s really nice that it provides, like a database of our family’s life in history.
We even have pages for cars and repair logs, computers and updates, everything. The struct plugin is amazing And you can always pop it open in SqliteBrowser too!
I’ve tried a couple of things and I just keep coming back to dokuwiki because it’s the best compromise.
No, I don’t. And that’s going to be one of our big differences here. Everyone in my family is tech literate and knows at least a little bit of programming.
I would strongly suggest dokuwiki. It’s like having a forever Journal of Family affairs and I really like it. I know it’s not quite the social media aesthetic but in my experience I found it to be the thing that stuck.
I would argue against Mediawiki though. It may be more user-friendly for some family members, but the maintenance becomes a nuisance And pulling things out of the database involves half a dozen joins.
Even though dokuwiki editing is text in markup, It’s not a hard concept to grasp and the simplicity makes it feel more tangible which may be appreciated by older family members.
There’s a plugin that does it, FoF or something, and then you can upload an image from your device and it’s pretty good. Some videos play others require downloading after uploading though.
Our family uses a post in flarum for a monthly feed and then moves a few of those images into a dokuwiki page with the gallery plugin.
On the software side: