Live data from Hacker News

Facebook LLAMA is being openly distributed via torrents

github.com

531–540 of 719 posts

Re: Facebook LLAMA is being openly distributed via torrents

#531
post #385
post #332

Earlier quoted context omitted.

On the flip side of subhuman treatment of humans, we have useful legal fictions like corporate personhood. It's going to be pretty rough for a while, particularly for nontechnical judges, to sort all of this out. We're almost definitely going to see multiple rulings far more bizarre than Citizens United ruling that limiting corporate donations limits the free-speech rights of the corporation as a person. I'm not a la…

Good points but it’s worth clarifying that this is not what the Citizens United decision said. It clarified that the state couldn’t decide that the political speech of some corporations (Hillary the Movie produced by Citizens United) was illegal speech and speech from another corporation (Farenheit 9/11 by Dog Eat Dog films and Miramax) was allowed. Understood this way it seems obvious on free speech grounds, and in…

TIL. Thank you very much for correcting my ignorance!

Re: Facebook LLAMA is being openly distributed via torrents

#532

Earlier quoted context omitted.

If someone just decides to use the torrent and ignore those restrictions it might finally establish precident for if you can copyright model weights.

But even if you could copyright them, once you do some fine-tuning, they are not the same model weights!

that would be a derivative work

Re: Facebook LLAMA is being openly distributed via torrents

#533
post #271

opt-175B weights are already openly available as I understand. Hugging-face also has openly available weights for a 176B parameter LLM called Bloom. Is LLAMA offering something over and above these?

opt-175B doesn't exist; the largest one is opt-66B. And, at least in the tests I've run (not with the biggest one, but only up to a dozen billion parameters), all the opt models severely underperform with respect to even much smaller models. To the point that the launch of OPT (before BLOOM) was literally advertised as "the biggest OpenSource language model released to date", because they couldn't push on much else.…

> opt-175B doesn't exist;

It doesn't exist for practical purposes because it is gatekept behind the same Facebook application process

Re: Facebook LLAMA is being openly distributed via torrents

#534
post #319

Earlier quoted context omitted.

They wouldn't be us . We will still die when our bodies fail. But maybe there will be some AI tricking our friends and family into thinking we're still there.

What if it's an incremental upload? E.g. we start with some prosthetics and slowly migrate organic function to digital? Is this the Ship of Theseus, or is it a slow but nonobvious death?

The Ship of Theseus is a weird one too. If you take a car apart and replace it piece by piece and replace the whole car you kind of have the same car. But what if you kept all the old pieces and put them back together? Which one is the original car? It is a interesting thought sub experiment that plays on the 'Ship of Theseus'. You could end up with the same issue here. If I make a perfect copy of myself, who is the 'real' me? There is an obvious 'older' me but the other one is just as capable as I am.

Re: Facebook LLAMA is being openly distributed via torrents

#535

Warning: do not use this for commercial purposes . While the weights may be available now, it's a lawsuit waiting to happen if you try to use this at work. See the original license: "a. Subject to your compliance with the Documentation and Sections 2, 3, and 5, Meta grants you a non-exclusive, worldwide, non-transferable, non-sublicensable, revocable, royalty free and limited license under Meta’s copyright interests…

And where did I sign my name to that agreement?

Re: Facebook LLAMA is being openly distributed via torrents

#536

Earlier quoted context omitted.

I think that as long as people can run their AI girlfriends on their own computers without having a corporation acting as an intermediary and thus a virtual "pimp" [for lack of a better word] in the relationship, I think it's fine. The problems come when people have to pay monthly to talk to their AI girlfriends and get charged extra if they want them to act a certain way or do certain things.

Seems like at that point one might as well just pay onlyfans? Although I suspect it is not far off until gpt and deepfaking get combined to produce completely generated onlyfans (or some other competitor if this is against their ToS, I have no idea). On a similar note I very much look forward to the day when entertainment providers leverage these so I can say to Netflix, et al, "I would like to watch a documentary ab…

I'm not seeing how onlyfans equate to virtual AI girlfriends

Re: Facebook LLAMA is being openly distributed via torrents

#537

It seems that the leak originated from 4chan [1]. Two people in the same thread had access to the weights and verified that their hashes match [2][3] to make sure that the model isn't watermarked. However, the leaker made a mistake of adding the original download script which had his unique download URL to the torrent [4], so Meta can easily find them if they want to. [1]: https://boards.4channel.org/g/thread/9184826…

Why do 4chan users go out of their way to be so offensive in their posts?

Because it's really the onlY place left to go if you want to be offensive. First forums, then platforms censored offensive people out of their niche places. Even Cloudflare participates. 4chan remains the only privately owned large forum.

Re: Facebook LLAMA is being openly distributed via torrents

#538

Warning: do not use this for commercial purposes . While the weights may be available now, it's a lawsuit waiting to happen if you try to use this at work. See the original license: "a. Subject to your compliance with the Documentation and Sections 2, 3, and 5, Meta grants you a non-exclusive, worldwide, non-transferable, non-sublicensable, revocable, royalty free and limited license under Meta’s copyright interests…

[dead]

Re: Facebook LLAMA is being openly distributed via torrents

#539

Earlier quoted context omitted.

It's funny that part of the 4chan excitement over this is that they think they'll get back the AI girlfriend experience of when character.ai was hooked up to uncensored GPT-3. All that has been thoroughly shut down by character.ai and Replika and they just want their girlfriends back.

I'm curious if the blocking of adult content has to do with moralism, commercial interests, or something deeper. An eager to please conversational partner who can generate endless content seems quite dangerous and addictive, especially when it crosses over into romantic areas. There's already posts of people spending entire days interacting with LLMs, using as their therapist, romantic partner, etc. Combined with fin…

“An eager to please conversational partner who can generate endless content seems quite dangerous and addictive”

Don’t date robots!

https://youtu.be/wJ6knaienVE

Re: Facebook LLAMA is being openly distributed via torrents

#540

~220 GB :O That's quite big!

Needs ~200GB of graphics ram to run... Not many people will get this running!

depends on which model you choose to run. The 7B model can reasonably run with consumer hardware.
Post reply on HN