Ferret: A Multimodal Large Language Model • Hacker News Board

Re: Ferret: A Multimodal Large Language Model

#241

Post by WhitneyLand » Sat, Dec 23, 2023, 10:21 PM UTC

hackernewds wrote: post #208

mandelken wrote: post #205

Earlier quoted context omitted.

Mistral 7B is pretty good and the instruct v0.2 runs on my iPhone through MLC Chat. However, the ChatGPT4 app is much better in usability: better model, multi-modal with text/vision/speech and better UI.

gpt 4 allows commercial use?

Yes and no.

You can use it commercially but there are some restrictions, including some of a competitive nature, like using the output to train new LLMs. This is the restriction that Bytedance (Tiktok) was recently banned for violating.

Re: Ferret: A Multimodal Large Language Model

#242

Post by mark_l_watson » Sat, Dec 23, 2023, 10:26 PM UTC

theferalrobot wrote: post #239

shrimpx wrote: post #216

Apple has been looking sleepy on LLMs, but they've been consistently evolving their hardware+software AI stack, without much glitzy advertising. I think they could blow away Microsoft/OpenAI and Google, if suddenly a new iOS release makes the OpenAI/Bard chatbox look laughably antiquated. They're also a threat to Nvidia, if a significant swath of AI usage switches over to Apple hardware. Arm and TSMC would stand to w…

Given Apple's track record on anything AI related and the terrible state they keep CoreML that not only seems extraordinarily unlikely, it would take a lot of time to win developer trust and that I just don't see happening.

I have enjoyed working with CoreML over the last few years. Please share what you didn’t like about it.

Re: Ferret: A Multimodal Large Language Model

#243

Post by ben_w » Sat, Dec 23, 2023, 10:27 PM UTC

darthrupert wrote: post #147

ben_w wrote: post #130

Earlier quoted context omitted.

I'd be (pleasantly) surprised; look at how long it takes them to allow system apps to get replaced with downloadable 3rd party. And even then, the activation keyword is likely to be whatever Apple says. Similar logic as 3rd party keyboards, don't want user input to get stuck on even merely potentially untrustworthy or buggy code.

That's not due to incompetence; that's due to not seeing any reason to do it.

I was not intending to imply otherwise.

Re: Ferret: A Multimodal Large Language Model

#244

Post by simonair » Sat, Dec 23, 2023, 10:34 PM UTC

georgespencer wrote: post #240

Booourns wrote: post #235

Earlier quoted context omitted.

Out of curiosity have you experienced their autocorrect on iOS 17 because that’s when they updated to be LLM based?

Transformer based autocomplete on iOS 17 feels just as bad -- but in different ways -- as its previous incarnation to me.

Are you tapping the keys or swiping over those that make up the word you want to type? In my experience, tapping has always been and remained poor but swiping is getting better and better with every iOS version.

Re: Ferret: A Multimodal Large Language Model

#245

Post by amelius » Sat, Dec 23, 2023, 10:35 PM UTC

smoldesu wrote: post #44

> FERRET is trained on 8 A100 GPUs with 80GB memory. Huh, even Apple isn't capable of escaping the CUDA trap. Funny to see them go from moral enemies with Nvidia to partially-dependent on them...

Dependent is a strong word. At the end of the day all these DL models run on any hardware, and you can easily swap out one type of hardware for another perhaps with some small performance impact. They're commodities, basically.

Re: Ferret: A Multimodal Large Language Model

#246

Post by hosh » Sat, Dec 23, 2023, 10:36 PM UTC

theferalrobot wrote: post #239

shrimpx wrote: post #216

Apple has been looking sleepy on LLMs, but they've been consistently evolving their hardware+software AI stack, without much glitzy advertising. I think they could blow away Microsoft/OpenAI and Google, if suddenly a new iOS release makes the OpenAI/Bard chatbox look laughably antiquated. They're also a threat to Nvidia, if a significant swath of AI usage switches over to Apple hardware. Arm and TSMC would stand to w…

Given Apple's track record on anything AI related and the terrible state they keep CoreML that not only seems extraordinarily unlikely, it would take a lot of time to win developer trust and that I just don't see happening.

Apple doesn’t have to win developer trust or build an AI platform. They just have to build a compelling consumer product that can only function with AI, and they are better equipped to do that than Google or Microsoft. It remains to be seen if OpenAI will go that route instead of a business built on training and providing access to foundational models.

Re: Ferret: A Multimodal Large Language Model

#247

Post by amelius » Sat, Dec 23, 2023, 10:36 PM UTC

ssijak wrote: post #56

smoldesu wrote: post #44

> FERRET is trained on 8 A100 GPUs with 80GB memory. Huh, even Apple isn't capable of escaping the CUDA trap. Funny to see them go from moral enemies with Nvidia to partially-dependent on them...

I guess they also have Samsung fridges in the offices..

And they use CAD software running on Windows (it simply doesn't exist on MacOS)

Re: Ferret: A Multimodal Large Language Model

#248

Post by fragmede » Sat, Dec 23, 2023, 10:46 PM UTC

dbish wrote: post #82

freedomben wrote: post #15

> Usage and License Notices: The data, and code is intended and licensed for research use only.

Many big “open source” releases in the AI community recently are not licensed for commercial use. Not really OSS at that point (ex:fuyu model from adept)

I think the term should be "model available" rather than open source.

Re: Ferret: A Multimodal Large Language Model

#249

Post by georgespencer » Sat, Dec 23, 2023, 10:51 PM UTC

behnamoh wrote: post #62

aaronbrethorst wrote: post #33

Earlier quoted context omitted.

Rumors suggest they’re gearing up to make iOS 18 an AI focused release. It’ll be interesting to see if they offer different capabilities for online/offline scenarios, or if their offerings are strictly offline. Here’s one story to offer some context. There are others. https://archive.is/en3VL

> Rumors suggest they’re gearing up to make iOS 18 an AI focused release. Don't underestimate Apple at disappointing enthusiasts like you and me. We've been hearing many awesome stories about the next thing Apple will do, only to realize their marketing team chose to keep it for future iOS/MBP/iPhone generations to keep the profits high.

> We've been hearing many awesome stories about the next thing Apple will do, only to realize their marketing team chose to keep it for future iOS/MBP/iPhone generations to keep the profits high.

I believe it's more likely we've heard awesome stories about things Apple will do in the future, only to realize that the average HN commenter is incapable of understanding that such stories are contextless leaks, and that it is far more likely you are operating with incomplete information than Apple's "marketing team" is holding things back for future "iOS/MBP/iPhone generations" to keep their profits high.

I know it's more fun to vomit dumb conspiracies onto the internet, but consider changing "realize" to something which conveys equivocation, because your theory about Apple's marketing team holding back mature technology in order to benefit future devices – in addition to being predicated on leaks and rumors, and risibly inane when you consider that such action would create a significant attack vector for competitors – is as equivocal as the belief that Trump is on a secret mission to destroy a global cabal of pedophiles.

Re: Ferret: A Multimodal Large Language Model

#250

Post by fragmede » Sat, Dec 23, 2023, 10:52 PM UTC

cryogenicfire wrote: post #111

cryogenicfire wrote: post #106

Earlier quoted context omitted.

I've always wondered why Apple is never in the conversation of modern AI, like until now it almost feels like they've been just watching from the sidelines without taking part in all the commotion Maybe they can just afford to observe before making major decisions on the direction of the company... For a company like Apple, I feel like they won't lose their customer-base just because they are taking the AI race slowl…

I hit reply before I finished typing XD ... Anyways, my point being that Apple will gladly introduce a polished product a couple of years after everyone else has already done it, and their target audience will still applaud their work and give them money. Apple for some reason simply _can_ afford to test the water

You can edit your comments shortly after you make them.