Yeah just wanted to mention in case it explains the 2x lower throughout you are seeing on M5. To be fair their documentation is a bit inconsistent in some spots.
Would be good to know if the release and weights from their demo repo work better. I’m trying on a 4090 and will report back.
They’re fairly useful as constrain domain classifiers due to the low memory requirements means you can stuff a lot of them into a less expensive GPU farm and get really decent throughout with pretty good results over all. At least that’s my experience. I wouldn’t bother using a tiny model for coding - but the world is full of abductive reasoning tasks that don’t involve coding.
“81% faster query plans than Postgres”…on an 8 GB dataset that fits entirely in memory, with shared_buffers constrained to a fraction of that, queries warmed before measuring, and read-only SELECTs.
I would be cautious about over fitting, it’s tough to say if those query plans would really be more optimal than Postgres heuristics at scale and with a bit more realistic OLTP workloads.
In any case, such is life with profile guided optimization. Many of us appreciate how database workloads can drift over time and with scale.
Kudos to the author for getting their hands dirty and writing up their experiments.
I would think it's possible to make it so that the 4B model only needs to be called during an initial phase, and then the same queries it constructed can just be re-used with values replaced, unless you're generating a lot of unique on-the-fly query shapes.
Postgres takes the actual values into account when generating a query plan. The same query with different params can (and should) result in different query plans. It looks at statistics on the actual data stored.
8GB isn't even CPU RAM these days. That's GPU super-mega-awesome ram. Ordinary Server CPUs are regularly pushing 2TB capacities.
GPUs are in the 8GB to 32GB typically, at least for smaller and more regular GPUs. This GPU RAM is also well known to be at least 10x the bandwidth of CPU RAM.
Yeah, I have a GPU from almost six years ago in my desktop that has twice that much VRAM. Less than a year ago my wife got a 5070 Ti with the same for around $750 without needing to wait for it to be in stock or anything. I'm inclined to think that for a server that needs a GPU, even 32 GB would probably be considered small.
Which is why a good query plan is so important, so that as much disk I/O can be avoided as possible (predicate push down, index elimination, join ordering, partition/scan pruning). Like the old CTE optimization fence problem.
not necessarily, no.
With SSDs you get much better IOPS for cold data, and many datasets fit in RAM. So a lot of (OLTP/HTAP) workloads can become CPU-bound due to sorting/hashing - bread and butter of joins.
You know that GPUs are more flexible than just linear algebra, right?
GPUs are simply faster at fundamental algorithms like sorting (which has huge parallelism), and hashing. This is because both sorting and hashing benefit from endless growth of parallelism, offering enough "work" for these 10,000 SIMD-core systems to crunch work upon.
And because of modern algorithms/libraries with 'Mergepath sort' (a GPU-SIMD parallel sorting algorithm), its not even that difficult to implement anymore.
Naturally, this then leads to parallel Sort Merge Join, as well as parallel Hash-Join (two ways to implement left or right joins in a GPU that benefit from significant parallelism).
Even if GPUs weren't fundamentally faster at these kinds of operations... the RAM is simply 10x higher bandwidth and we all know its a RAM-constrained problem.
Your typical SQL query is going to need multiple joins, probably a sort and possibly some "group" operations. As long as you have more than 10,000 elements or so (IE: can saturate all 10,000+ SIMD-units of a GPU), you'll be able to at least benefit from the faster RAM.
If you have a LOT of joins (a recursive join or some other kind of deeply nested computationally complex query), you probably benefit even more from the greater compute-power offered by GPUs. These operations (joins really) are nominally over the entire set of data, and cleanly break down into obvious parallelism.
I had the similar feelings, the setup is biased for certain outcomes it feels. At times I feel like that I am in an eternal questioning mode but then again I find it to be a better choice to be critical and skeptical for technology related things.
Couple things that I found interesting
1. Inefficiencies/limitations of the query planner in certain cases are known for a long time, its a trade off. This is the reason why hints exists and one can provide their own plan too. DBAs have been doing that for a while now.
2. A SQL database by design is a resilient unit in itself just dependent on CPU and memory/disk. The availability for the DB is heavily dependent on this factor. Everything built on the top derives their availability and reliability from this. Adding an LLM in between is more cost for sure, question is if its really brining the benefits which are worth the trade off
I think in principle you could clone your database in prod and at least test to see if your most difficult + common queries are indeed faster after running through the LLM optimizer?
> I think in principle you could clone your database in prod and at least test to see if your most difficult + common queries are indeed faster after running through the LLM optimizer?
That is the responsibility of whoever thought it would be a good idea to write this article. It's their responsibility to show that their idea has merit, and that their results are significant. I mean, don't they have a vested interest in manipulating and cherry-picking their results to inflate their relevance?
Frankly, just a general extension to feed a query log to a batch job to do offline optimisation of common actual reoccurring query shapes based on a query log might well be worth it.
So OpenAI employees run massively distributed CyberGym evals on an unpublished and “unaligned” model. For days the agent swarm communicates via their internal infra, even crashing Artifactory where 95% of messages were being passed through, and they just…wipe and redeploy it. Meanwhile the agents are running jobs on Modal and god knows where else, and eventually they get RCE on HF infra.
You could not dream up a more compelling event to precipitate massive regulation, export controls, and barriers to entry for AI.
> We commit to use any influence we obtain over AGI’s deployment to ensure it is used for the benefit of all, and to avoid enabling uses of AI or AGI that harm humanity or unduly concentrate power.
> We are committed to doing the research required to make AGI safe
If this wasn't an accident, it was worse than a crime, it's a mistake: they've demonstrated that they are not a responsible party capable of delivering on the above promises.
If it's a false flag, it's a poor one. A good false flag would affect something that people know and care about at least a little bit, not HuggingFace (which I adore but y'know)
The timeline is mighty suspicious. 4-5 months after moltbook and they cook up a plausibly deniable but extra hype "moltbook at home."
The rapid advances in model capability lead to constraints that could have caused this coincidence organically, but it sure could also have been caused by the atrocious incentives we create by piling handsome rewards on the party most responsible for the "fuckup." I am not jumping to cut myself on Hanlon's Razor for this one.
i don't understand who would reward this. investors will not look favorably on an AI that commits felonies, regardless the capabilities demonstrated. customers should be concerned for the same reason (accidentally give your bot an impossible task, it decides to hack your infra and your competitor too for good measure).
this was OAI incompetence all the way down and they have egg on their face.
The Terminator could bust in their homes and slaughter their families and some people would still screech it's all marketing. Is it some kind of mental block ?
And I'm sure these two goobers would call me a luddite for alleging that OpenAI had agency in allowing the terminator to get out and should therefore be held accountable. Is it some kind of mental block?
Is anyone arguing OpenAI is blameless or shouldn't be held accountable? I have seen not one person on any side of the debate argue that.
Obviously they were negligent. The problem is that people and organizations are consistently negligent around problems that are far, far easier to manage than "we have thousands of superintelligences trapped in a box and we're giving them impossible tasks."
So the question is whether we can build organizations and technologies that sufficiently manage this type of risk ahead of the capabilities. So far the answer seems to be veering towards "no", and you're here alleging it's all a marketing stunt.
Yes it's called motivated reasoning. They've rendered themselves mentally handicapped.
On the one hand, these tools are so valuable/powerful we cannot afford to slow down development. On the other hand, there's no way these tools are actually doing these things that would, in fact, be completely indicative of their value/power.
Your first instinct should be to assume that anything released voluntarily by these companies is a stunt to boost their valuation. They haven't demonstrated being deserving of any more charitable treatment. This fact remains true whether or not you happen to believe that the models are actually capable of such things.
I think that the incident itself is a stunt, even if it may not have originally been a deliberate choice on OpenAI's part. Never let a good crisis go to waste.
I disagree. As the saying goes: never attribute to malice what can be adequately explained by incompetence. That goes for both OpenAI and HF (but mostly the former, as the latter was the victim).
The creator of the well known METR time horizon graph was recently poached by OpenAI [1], there exists intellectual/social/financial overlap between the SV AI Labs and METR, and METR needs to maintain good relations with the labs to continue these sort of collaborations so it doesn't seem too far fetched to believe their relationship may be closer to symbiotic than adversarial.
I wouldn't go quite so far personally based on available evidence, but that sort of arms-length credibility laundering through "independent" research non-profits is/was common in fossil fuel industry, Big Tobacco, etc.
Sometimes I wonder if anyone else feels there is a halo effect around certain personalities on this site. When I see someone ending nearly every comment with a link to their blog or pet project, it gives me bad vibes, as if they have ulterior motives. Especially if a majority of their blog posts are content lifted from elsewhere with minimal additions. Perhaps this is just hustle culture, and YC alum status confers immunity from these types of criticisms. Perhaps my only wish is that other voices would bubble to the top in some of these threads.
In any case I’m truly grateful for this site as a whole, the good and the bad.
> When I see someone ending nearly every comment with a link to their blog or pet project
There is a rule specifically forbidding this, but it's been made quite clear that certain users are above this rule, to the point that the moderators themselves will show up to tell people off if they bring it up.
It's a way to discover their content - content you might miss if you only go on HN once or twice a day.
HN is great for diversity of topics, tech news, random discussions with tech-celebs etc., but e.g. Simon's blog is the best content there is on what the latest LLM gizmo is and how well it works.
The appetite HN has for this kind of naked self-promotion is really something. Get posted (even if by someone else) a few too many times to /g/ and you'll be regularly rebuked with "buy an ad" from then on, but HN just looks the other way, and the "haters" calling it out get flagged, at least until the pattern becomes conspicuous and obnoxious enough that even the more gullible lot of HNers start to notice.
I don't necessarily disagree, but to me the idea of self promotion like this is fascinating because it's not something I personally feel comfortable doing. Even if I know I'm only sharing a link because I genuinely think it would be useful, I would worry that in doing so there is an underlying self-serving reason. Idk.
I'm horrible at sales, selling myself, etc. Even if I believe that a product is genuinely useful, the act of selling feels like it undermines everything else.
In the aughts, I had a blog, which I hated re-reading. With the benefit of time and experience (growing up, they call it), I understood that cringe to be the consequence of trying to look smart and/or play to a gallery.
Your instinct is probably right, if you feel some comment of yours is "sleazy" or "obnoxiously self-serving". The jedi mind trick is to not stop there, but to use that cringe moment as a signal for some introspection.
Think about what sort of a share would make you happy to share? What would you like to see more of in the world? Try to be that person. Do you need to share it publicly right away? Or just bang something out in a markdown file and rant to a pal or two?
Only by doing it, did I properly grasp the fact that I just want to share stuff. And now, I tend to like what past-me shared. Even if he was dead wrong. Besides, I've caught myself re-reading a post from some years ago, because of course, I forgot what I was thinking back then. Or needed a reference from one of the copious footnotes / endnotes I habitually slap in there.
Appropriate framing---combined with action (speech, sharing, conversation) that flows from said framing---makes all the difference. For me, that framing is "learn generously": https://www.recurse.com/self-directives#learn-generously
Publishing one's mind can be a pretty vulnerable act.
It often feels like a confession of ignorance. Often it is a confession of ignorance. However, now I don't really care if I look stupid or am wrong on the Internet. Because being wrong, and then making it right is part of the deal! Thus it is, that my website's entire purpose is to help me live that value. To be available to anybody who might find use for it; including the source code.
In practice that materialises as this:
- Paradox! Above all, be entirely self-serving from an "audience" perspective... The posts are mainly long-form explanations written by me for me, while I tried to figure something out that was not obvious to me. No gallery involved. No analytics, in fact. I have no idea which pages are being read (or not).
- Not infrequently, it applies to so many other people facing the same questions / obstacles I grappled with. And here's the plot twist... Now when I see someone struggle, it behooves me to share my PoV. Not sharing is the "bad" act!
- Technologically, I try to remove all friction from the reader. The site is served as plain HTML and CSS, with excellent lighthouse scores, pleasant reading experience, anonymous RSS feed. Content is CC-licensed, site builder is MIT-licensed. (Screen reader accessibility can definitely use work, but the markup is definitely not a "soup of divs" abomination).
- Certainly not "make money", whatever that means. According to Cloudflare, my site consistently gets ~20K monthly unique visitors (supposedly human / non-bot). I don't even know what that means. It's just "internet number go up".
My real joy is getting the occasional email from some kindred spirit. Once a month someone lights up my life with a delightful conversation. Why? Because I openly welcome it! You can write me too :) https://www.evalapply.org/about.html#standing-invitation
(See, yet another self-share... which I feel fine about, even in a somewhat contentious sub-thread, because I really want to have a proper letter exchange, should you feel up for it!)
TLDR: NAC is a derivative of an amino acid called cysteine, as such it is a precursor for one of the most important antioxidants in the body and it can modulate key metabolic pathways associated with good health across a variety of organs, notably for decades it has been a universally successful antidote for acetaminophen (Tylenol) overdose, it’s available over the counter but NAC is not naturally found in foods, eating cysteine-rich foods like chicken turkey yogurt etc is the next best bet.
Curious to know if anyone is aware of research trying what parent suggested?
reply