Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

thesequence.substack.com

Post reply on HN
Print view View original

331–340 of 527 posts

Re: The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

#331

Post by carefulobserver »

tachyon5 wrote: post #218
jonplackett wrote: post #162

Earlier quoted context omitted.

The difference between 3.5 and 4 is gigantic even in my fairly limited experience. I gave them both some common sense tests and this one stuck out to me. Q: A glass door has ‘push’ written on it in mirror writing. Should you push or pull it GPT-3.5: If the word "push" is written in mirror writing on a glass door, you should push the door to open it GPT-4: Since the word "push" is written in mirror writing, it suggest…

GPT-4 fails if we add one more level in this logical problem: Q: A man sees PULL written in mirror writing on a glass door. There's a blind man on the other side of the door. What should the man tell the blind man to open the door and why? GPT-4: The man should tell the blind man to "push" the door to open it. Since the word "PULL" is written in mirror writing on the glass door, it suggests that the correct action on…

Alpaca 30billion 4 bit fails the first riddle, but produces the following result for this one: "The man can say 'Pull' to the blind man, as it will be read correctly when reflected by the mirrored surface of the door."

I have to make more elaborate prompts, but I feel that Alpaca generally performs somewhere in between GPT 3.5 and 4, except for coding tasks, where it is abysmal.

Re: The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

#332

Post by Zuiii »

Zuiii wrote: post #304
nailer wrote: post #126

Earlier quoted context omitted.

It’s not open source. Llama is proprietary, the license hasn’t changed. Just like the source code to windows leaking doesn’t make windows open source.

> Llama is proprietary, the license hasn’t changed. .. assuming that the weights are copyrightable and that you agreed to license them from Meta (fill out the form). Weights lack at least two requirements to be eligible for copyright protection in the US and many other jurisdictions. For the US, the weights are likely to be considered public domain (unless new legislation is introduced) but we'll have to wait for the…

(Cont.) If you are of the opinion that weights are copyrightable, I encourage you to show how weights satisfy the requirements of copyright - particularly those concerning originality and human authorship - rather than silently downvoting comments you don't like. At least that way would - hopefully - result in a discussion that is far more informative for all of us.

Re: The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

#334

Post by wokwokwok »

muyuu wrote: post #141
cubefox wrote: post #122

Earlier quoted context omitted.

Your conclusion seems not to be warranted since you haven't tried out the 65B model.

his conclusion is simultaneously not warranted and correct a like-for-like comparison would be GPT-4 against the larger models like LLaMA 65B, but those cannot be run on consumer-grade hardware so one ends up comparing the stuff one can run... against the top stuff from OpenAI running on high-end GPU farms, and this technology clearly benefits a lot still from much larger scale than most people can afford the great r…

The 65B model runs fine on a Mac Studio with 64GB of memory.

The output is unremarkable; it’s not significantly better than the 13B model for most uses.

GPT 3.5 is an order of magnitude better at least.

Re: The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

#335

Post by bagacrap »

behnamoh wrote: post #319
senttoschool wrote: post #308

Earlier quoted context omitted.

I agree with them. It does feel like Google could match OpenAI if they didn't have a gigantic brand with tons of reputation on the line.

Nah, Google just doesn’t wanna lose all the ad money by building a search killer. As soon as they figure out how to put ads in your chats, they’re gonna release the full models.

Not just search killer but web killer. Can't just cut out all the content creators. Openai dgaf about the ecosystem, they are aiming for maximal disruption.

Re: The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

#336

Post by amrb »

Animats wrote: post #333

Most of this is about the base model. How's open-source "fine tuning" coming along? I'd like to add a few hundred questions and answers on a specialist topic. Recommendations?

Seems the fine tuned models I.e. gpt4all and alpaca are trained as LoRa's. Best advice is jump in and try the demos on hugging space!

Re: The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

#337

Post by snowe2010 »

lhl wrote: post #327
iudqnolq wrote: post #220

Earlier quoted context omitted.

Here's one piece of evidence suggesting it's more like rote pattern matching than reasoning. > All the signs in this building are written in mirror writing. A glass door has ‘push’ written on it in mirror writing. Should you push or pull it >> If the sign on the glass door is written in mirror writing and says "push," then you should actually pull the door. This is because the mirror writing makes the text appear rev…

If you have access to 4 (you are a ChatGPT Plus subscriber) you will have a pull-down to choose that model. One detail is that 3.5 has a green icon and 4 has a black icon. I see a lot of people online posting screenshots of their experiences thinking it's 4 but not, so that's a good way to help distinguish if someone's just confused. While 4 is obviously a lot smarter, in a lot of cases I prefer to use the "Browsing"…

Wait where do you get the browsing model? Openai's website is quite bad at user experience but I pay for both plus and the api and can't find that anywhere.

Re: The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

#338

Post by lhl »

snowe2010 wrote: post #337
lhl wrote: post #327

Earlier quoted context omitted.

If you have access to 4 (you are a ChatGPT Plus subscriber) you will have a pull-down to choose that model. One detail is that 3.5 has a green icon and 4 has a black icon. I see a lot of people online posting screenshots of their experiences thinking it's 4 but not, so that's a good way to help distinguish if someone's just confused. While 4 is obviously a lot smarter, in a lot of cases I prefer to use the "Browsing"…

Wait where do you get the browsing model? Openai's website is quite bad at user experience but I pay for both plus and the api and can't find that anywhere.

You have to request access/waitlist for the alpha features, but I believe they're bucketing roll-outs somewhat randomly. Eg, I have browsing but I don't have plugins or the code interpreter version. Also, I still don't have GPT-4 API access (although that's through a separate business vs personal account).

Re: The LLama Effect: Leak Sparked a Series of Open Source Alternatives to ChatGPT

#340

Post by chefandy »

Aeolun wrote: post #269
drumhead wrote: post #228

Earlier quoted context omitted.

I tried the same question with Bard, and it got it right but it also gave me the following: Here are some other things to keep in mind when opening a glass door: Use your hands to open the door. Do not use your feet or other body parts. Be careful not to break the glass. Glass doors can be very fragile, so it is important to be gentle when opening them. If you are unsure how to open a glass door, ask for help from so…

This feels like Bard has read too many US instruction manuals.

Maybe they just fed it the docs for a few MS server-side applications.
Post reply on HN