keonix

It would not be illegal based on fair use (though you have to be careful there also), but if you try to regurgiate large portions of the book then it would be. And we do know that models regurgiate training material verbatim (Copilot)

If I regurgiate something I read in copyrighted book without proper license that also would be theft, no distinction there.

I'm not distributing my brain, at least same (but probably more restrictive) should apply to models - training is okay, but using and distributing should be limited by copyright

> with or without finetuning?

With, but it's still bonkers that it works so well

>Also is there a practical motivation for creating them?

You could get in-between model sizes (like 20b instead of 13b or 34b). Before better quantization it was useful for inference (if you are unlucky with vram size), but now I see this being useful only for training because you can't train on quants

Wait until you hear about frankenmodels. You rip parts of one model (often attention heads) and transplant them in another and somehow that produces coherent results! Witchcraft

https://huggingface.co/chargoddard

I assume it's because such large context takes lots of memory, so you might as well have smarter model if you are not gonna fit in small vram anyway

Mistral 7B ~ 8 GiB

StableLM 3B ~4 GiB

You could go even lower with smaller quantization if necessary. I personally wouldn't use anything smaller than 7B and Mistral already pushing it in coherence. Overall it depends on your use case, not everyone needs smart models, or large context that sometimes takes half of required memory.

Codellama is also surprisingly good even for non-coding tasks

For what it's worth, I think OP has a point in showing inconsistencies in your logic

That's disingenuous. You provided link to a legal framework, but not any use of it. Your source doesn't even mention a single delivery under this act.

> Can you provide a source for your statement that these were donations?

You made the first claim, it's on you to prove it because disproving something is harder than making false claim. But you can have my source anyway.

https://www.politico.com/newsletters/national-security-daily...

> US isn't donating weapons to Ukraine, it's leasing them

Could you kindly support your statement with the source? AFAIK most aid provided to Ukraine is not under lend-lease but donated to them

IIRC temperature is proportional to velocity squared so increase in simulation speed will result in squared increase in temperature. But even if we account for that we are not dealing with ideal gas in real world, there bound to be collisions that break molecules apart or even start nuclear fusion at certain speeds

> So it’s completely subjective to you

No? My prediction holds even if I didn't exist.

> not anywhere near the testable physics that I was looking for.

And now you are moving the goalpost.

Sure! It's just a classification of events that each person has.

Testible prediction? Socially close people will have similar classifications as opposed to randomly chosen ones (I will leave defining metric space over possible classifications as an exercise to the reader)

GPL restricts only developers. As an end user restrictions don't even apply to you

Nothing changes, it is still a violation

> western Europe or the US.

Why would we care about anyone not in the richest countries. It's not like they need security by default to not become another botnet and DDoS Europe or US businesses.

I would like to see you justify paying sso.tax to business owner in countries where sysadmin is payed less than those services ask in a month

Make it so decoding feasible only using some consumer hardware not usually found in servers, but omnipresent on consumer devices. GPU for example

If you buy from reseller they have no way to correlate payment details to search queries. With subscription, however, service provider have your payment details and knows everything about your queries so it can build a profile on you.

You could sell tokens that get used up after each search or expire after certain amount of time since first use. Browser extension could store tokens and provide them to website as needed in random order. Tokens could be resold so no tracking by payment processor

They do actually restrict issuer in most cases. Linus can't just strip Linux of other developers parts because they are tightly interconnected. He could relicense 30 year old version of Linux, but it would be useless

Patents also preserve knowledge in publicly accessible and well documented form. Without patents the only option to recoup investment in R&D would be to make innovations a trade secrets. This will lead to frequent reinvention of the wheel and lost knowledge.

The system is broken, but abolishing it is even worse

Sure it is, you brain is not software though

I assume the code is a derivative work of training data because given different data code would be also different (neuron weights)

To build a browser you don't need a verbatim GPL code, so it's not a derivative work in the same sense copilot is.

Stackoverflow on the other hand is much trickier question...

If the training set contains verbatim (A)GPL code does this mean that Copilot also should be distributed by Microsoft under GPL? Because without it Copilot (as it is distributed by Microsoft) couldn't be built, wouldn't it make it a derivative work of GPL'd code (and obviously every other license)?

I see a lot of people comparing human learning to machine learning in the comments, but there is a huge difference - we don't distribute copies of humans

Oh boy, that was a fun thread to read! Thanks for a laugh at GP expense

AGPL license clearly gives you the ability to freely link with whatever software you want. The only requirement is to share it with your customers. Following your logic it could be said that MIT licensed software could not be freely used because it requires preservation of original copyright notice

It was 18 years ago, he changed[0] his views later, but nobody mentions that for some reason...

[0] https://stallman.org/archives/2019-jul-oct.html#14_September...

He did at least changed[0] his opinion on things he said in 2003 about "willful" sex between child and adult after talking to medical experts. Don't know why it took him so long, but it's an indication that he is willing to change his views in conversation.

Still think that reinstating him at FSF was net negative for free software movement

[0] https://stallman.org/archives/2019-jul-oct.html#14_September...