Isn’t that the idea. Like you know that you had a viable solution to a complex problem previously so why go through the trouble of solving it again if you already did. Even if you have to modify it, it saves time for new novel problems. I’m
That’s the way. I’ve been programming for nigh on four decades, and it’s almost a daily occurrence with junior devs going to stack overflow or chatGPT to solve an issue instead of just searching the code where nine times out of ten the problem (or a very similar one) is already solved.
I am researching doing the same, but know nothing about running my own yet. Did you train your llm for programming in any way, or just download and run an open source one? If so which model etc do you use?
Lot’s of technical details, but essentially the llamafile is a engine + model + web ui, in a single executable file. You just download it and run it and stuff happens.
Run an open source one. Training requires lots of knowledge and even more hardware resources/time. Fine tuned models are available for free online, there is not much use in training it yourself.
I recommend llavafiles, as this is the easiest option to run. The GitHub has all the stuff you need in the “quick start” section.
Though the default is a bit restricted on windows. Since the llavafiles are bundling the LLM weights with the executable and Windows has a 4GB limit on executables you’re restricted to very small models. Workarounds are available though!
Im gonna give llamafile a go! I want to try to run it at least once with a different set of weights just to see it work and also see different weights handle the same inputs.
The reason I am asking about training is because of my work where fine tuning our own is going to come knocking soon, so I want to stay a bit ahead of the curve. Even though it already feels like I am late to the party.
It’s funny you say this, because my junior is complaining that I micromanage too much. I prefer to make the critical decisions. Whenever I don’t make them, I end up putting out fires.
I tell them that and they respond, how am I supposed to learn if I can’t make mistakes?
Then I remind them they can fuck up all they want in the dev env.
I try hard not to be an asshole at work. I also produce the most code. When my junior came on I told them that I’m here to help, and I want them to ask dumb questions instead of struggle. I also told them that I make mistakes, and they should call me on my shit—which they looove to do, brat.
Probably my misunderstanding. The way it is written it sounds like juniors use the dev env and seniors get their decisions implemented directly in prod.
Just because code works in dev doesn’t mean it’s going to be the right thing for production. I caught that little bastard making 1000 db queries in a loop one time, instead of taking the time to make it efficient. Technically it worked. And, because dev has no server load, it was relatively quick.
I’m not in software development, but this is how the entire company I work for operates.
We’re just kinda going forward with no clear direction, keeping stuff ticking over and constantly coming up with future plans that never come to fruition.
This is how all enterprise companies I worked for operates too. Only when I joined a smaller company with 80 people I realized that it can be really fun to work. We get a lot of stuff done and hardly any meetings. Really enjoying it.
There are answers, they just take a level of experience to reach that most people aren’t cut out for. You gotta be several principal+ IC roles or Dir+ mgr roles in before the patterns congeal into a plan.
Challenge is operating at those levels for extended periods requires a super fucking insane level of competency and dedication. Most people hit that spot and coast till retirement cause you’re at $500k+ at FAANG. Few keep looking for new opportunities unless forced to or they’re those corporate robot sharks with the dead eyes.
There are people who are knowledgeable and good at their job. Knowledgeable enough to be experts. Those are usually subject matter experts, including developers.
The issue is that no one can guarantee an outcome or that they’ve picked the right approach.
But this is really more a product of capitalism than anything else. Under capitalism you just have to keep moving even if you’re just making garbage and debt. There’s no reason to stop and think, because that is seen as a cost (even though it costs more to move without thinking).
Even the best companies that do factor in planning (at least in concept if not actually in practice most of the time) end up with the other problem of “resume driven development” where things that are totally fine and actually working get replaced with things that don’t work because someone needs a new project to get their promotion.
Capitalism produces garbage and puts the people who are least qualified in decision making roles. This still happens in natural systems, but much less. In (healthy) anticapitalist organizing, the people who know the most are generally asked to lead and when they don’t know what to do they stop and figure it out before moving forward.
Aimless wondering can still be a problem, but it’s not forced by the system to continue it’s just people who are learning.
I literally just posted this in !software_gore because they had the same energy so I found it on the original Reddit post remembered this one too. Great minds think alike?
Hi there! Looks like you linked to a Lemmy community using a URL instead of its name, which doesn’t work well for people on different instances. Try fixing it like this: !software_gore
<span style="color:#323232;">Hark! An SSL error, a tempest 'pon our web,
</span><span style="color:#323232;">Wherein secure connection falters, doth ebb.
</span><span style="color:#323232;">Aye, 'tis a quandary, a breach of secure bond,
</span><span style="color:#323232;">Thus, a web of trust, alas, beyond respond.
</span><span style="color:#323232;">In cryptic seas, where code and cipher dance,
</span><span style="color:#323232;">An error doth arise, denying safe romance.
</span><span style="color:#323232;">Oh, noble server, fraught with digital woe,
</span><span style="color:#323232;">Thy encryption falters, aye, laid low.
</span><span style="color:#323232;">Let us mend this rift, this breach, anon,
</span><span style="color:#323232;">And restore the web to its rightful throne."
</span>
<span style="color:#323232;"> --- Lord ChatGPT of Someshire
</span>
programmer_humor
Hot
This magazine is from a federated server and may be incomplete. Browse more on the original instance.