I'd forgotten why I'd stopped commenting here. Thanks for reminding me, everyone.
You should instead phone it in and use the easy paycheck to learn something you're interested in. I phoned it in for years in the crypto industry.
>You can't legally deport someone who's a citizen from birth no matter what ethnicity they are. I don't know anything about this person, but he seems to believe they're immigrants, because deportation otherwise makes no…
Why not become a doctor in that case?
Talking about LLMs "knowing" things is questionable at best.
Sounds like a cushy job. Why would you want to leave?
I mean, if they're illegal immigrants, it seems to me a country has the right to deport them. Maybe not the best way to phrase the sentiment, but taken literally it's not an incorrect idea.
Shit fast and soil things.
There is. You don't have to use the AI for everything just because you could.
"Conin" derives from "coniculus", which is what "conejo" derives from, so that makes all three words cognates, since they share a common origin.
Isn't the point of the concept to point out that they're not arbitrary?
I think the point is balance. Just because you can delegate everything, doesn't mean you should. In some cases, I've found, I simply can't. I have an idea for something I want, but I don't yet know what the best…
Counterpoint: I'm the same way, and the last couple weeks I've been messing more with Qwen more and I've been more productive in my pet projects than I had been in months or perhaps years, because I'm no longer blocked…
Training a human takes much less money and data than training an LLM. Even if I spent my entire life doing it, I would not be able to read the entire corpus of any LLM, even counting just the English subset. LLMs need…
What they're saying should change is that the Debug trait should be added to structs by default, instead of having to be explicitly specified.
Pretty sure what is meant by "lazily" is that it should just be added by default and then those that are not used should be optimized away by the linker.
I do this for control+alt+F1. Knuckle+thumb+middle finger. Also control+W: knuckle+middle finger. I, too, thought was the only one.
I've done it. I don't agree that it's therapeutic.
Your behavior is even stranger than that which you are questioning. The fact you take the time to type out the response means you also feel the urge to respond, and that urge is strong enough that you don't stop…
Why did that have to be transported in a thumb drive? It can't have been more than a few hundred megabytes at the most, right?
Yeah, I think the ideal AI user for coding is the one who is not so ignorant that they can't even explain what they want done, but not so knowledgeable that they could simply do it themselves. It's like a crutch: if you…
I'm pretty sure the company that owns the compute is the one paying for the power. Who would agree to have an ugly noisy cube in their property that they have to pay maintenance for? What would they get out of that deal?
It didn't work. DeepSeek failed to load on LM Studio due to a tensor format error. I don't have the time to do deeper troubleshooting. Qwen 3.8 Next did work, but the performance was abysmal.
Thanks, I'll give that a try. I basically have the same use case, only on Strix Halo.
Interesting, I had assumed it'd be too large to fit. What quant and context size are you running?
I'd forgotten why I'd stopped commenting here. Thanks for reminding me, everyone.
You should instead phone it in and use the easy paycheck to learn something you're interested in. I phoned it in for years in the crypto industry.
>You can't legally deport someone who's a citizen from birth no matter what ethnicity they are. I don't know anything about this person, but he seems to believe they're immigrants, because deportation otherwise makes no…
Why not become a doctor in that case?
Talking about LLMs "knowing" things is questionable at best.
Sounds like a cushy job. Why would you want to leave?
I mean, if they're illegal immigrants, it seems to me a country has the right to deport them. Maybe not the best way to phrase the sentiment, but taken literally it's not an incorrect idea.
Shit fast and soil things.
There is. You don't have to use the AI for everything just because you could.
"Conin" derives from "coniculus", which is what "conejo" derives from, so that makes all three words cognates, since they share a common origin.
Isn't the point of the concept to point out that they're not arbitrary?
I think the point is balance. Just because you can delegate everything, doesn't mean you should. In some cases, I've found, I simply can't. I have an idea for something I want, but I don't yet know what the best…
Counterpoint: I'm the same way, and the last couple weeks I've been messing more with Qwen more and I've been more productive in my pet projects than I had been in months or perhaps years, because I'm no longer blocked…
Training a human takes much less money and data than training an LLM. Even if I spent my entire life doing it, I would not be able to read the entire corpus of any LLM, even counting just the English subset. LLMs need…
What they're saying should change is that the Debug trait should be added to structs by default, instead of having to be explicitly specified.
Pretty sure what is meant by "lazily" is that it should just be added by default and then those that are not used should be optimized away by the linker.
I do this for control+alt+F1. Knuckle+thumb+middle finger. Also control+W: knuckle+middle finger. I, too, thought was the only one.
I've done it. I don't agree that it's therapeutic.
Your behavior is even stranger than that which you are questioning. The fact you take the time to type out the response means you also feel the urge to respond, and that urge is strong enough that you don't stop…
Why did that have to be transported in a thumb drive? It can't have been more than a few hundred megabytes at the most, right?
Yeah, I think the ideal AI user for coding is the one who is not so ignorant that they can't even explain what they want done, but not so knowledgeable that they could simply do it themselves. It's like a crutch: if you…
I'm pretty sure the company that owns the compute is the one paying for the power. Who would agree to have an ugly noisy cube in their property that they have to pay maintenance for? What would they get out of that deal?
It didn't work. DeepSeek failed to load on LM Studio due to a tensor format error. I don't have the time to do deeper troubleshooting. Qwen 3.8 Next did work, but the performance was abysmal.
Thanks, I'll give that a try. I basically have the same use case, only on Strix Halo.
Interesting, I had assumed it'd be too large to fit. What quant and context size are you running?