News for hackers. Stuff that matters.
Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by adam_arthur on Thursday January 19, 2023 @08:30PM (#34448079)
Hmmm, 2-3 high end GPUs implies it's likely not very far off from mainstream. Maybe runnable on the average device within 10-20 years... perhaps even sooner if the model/software can be optimized?
Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by lolspace on Thursday January 19, 2023 @08:56PM (#34448256)
20 years?
Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by adam_arthur on Thursday January 19, 2023 @09:17PM (#34448391)
In 20 years I'll still be alive and enjoying myself. Does 20 years seem long to you?
Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by krsdcbl on Thursday January 19, 2023 @10:39PM (#34448925)
it seems a bit long for achieving just running your own AI models locally, given it seems to be largely a question of vram and that you already _could_ do it today with a handful of graphic cards.

20y ago we had the GeForce Ti4400, current graphics cards now come with 100x the vram and 50-60x the bandwidth.

Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by 5e92cb50239222b on Friday January 20, 2023 @02:38AM (#34450171)
That's one hell of an assumption. Many of my Russian friends were absolutely certain they'll be alive and well for at least the next 20 years not that long ago.
Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by adam_arthur on Friday January 20, 2023 @04:28PM (#34460185)
If I'm dead then being able to run an AI locally doesn't matter anyway
Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by simne on Saturday January 21, 2023 @01:16PM (#34468666)
Sure, they will be alive if not come to Ukraine.
Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by dahdum on Thursday January 19, 2023 @09:20PM (#34448422)
> 2-3 high end GPUs implies it's likely not very far off from mainstream

Looks like FLOP/s per $ are doubling every ~3 years for high end cards, and 10x in ~10 years. So probably not that far off for desktop users.

https://www.lesswrong.com/posts/c6KFvQcZggQKZzxr9/trends-in-...

Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by visarga on Friday January 20, 2023 @03:50AM (#34450534)
FLOPs remain unused when the bottleneck is bandwidth.
Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by bemmu on Thursday January 19, 2023 @09:42PM (#34448548)
I’d be surprised if there weren’t any algo breakthroughs before that to make these several times faster (10x?).

(such as are all of the weights really needed all of the time, or could you load different ones based on recent context?)

If in 10 years on top of that you’d have that 10x faster hardware as well, you might be running GPT-3s as just a subcomponent of games/apps.

Re:ChatRWKV, like ChatGPT but powered by the RWKV (RNN-based... by ETH_start on Thursday January 19, 2023 @10:58PM (#34449058)
ChatGPT prompt: What's the difference between a component and subcomponent?

A component is a self-contained unit of functionality, often in the form of a class or object, that represents a piece of an overall system or application. A subcomponent, on the other hand, is a smaller, simpler component that is used as part of a larger component. A subcomponent is typically used to break down a larger component into smaller, more manageable pieces, and is often used to create reusable building blocks for larger components. In other words, a subcomponent is a component that is used by another component.