I agree with you, and I'm using claude and other models heavily, the idea is not in the soneshot - because actually as you see in the logs I posted on github, it took around 107 + 50-ish experiments to figure out which models to run which and push the speed.

I'm treating it as a search algorithm project, where I develop the methdology and search for pushing the limit of devices and searching in a big space of possible kernels and so for what works for a specific hardware.

so I'm looking at it from a different prespective, it takes a lot of brute forcing, and claude can do this well, the whole game (to me) is which directions you push the model into to reduce the search space. and how that learning from very small and tight devices you can scale these techniques.

But in essence, I don't disagree with your sentiment - the goal is not 0.6B models, these are extremely trivial as you mentioned and as I showed in the experiments, the goal is 16B+ models, and this as far as I've seen, requires smart methods to search for optimizations and customizations both to the model and inference engines.

Why did you both say "soneshot"? Same typo of oneshot?

I just asked myself the same thing and wondered if this was some new technique that completely slipped past me. - But it certainly must be a double typo

orrr we can listen to paranoia that these 2 comments are just an elaborate advertisement

[delayed]

the kind of one shots you get with fable 5.5 :D (it was a double typo)

"double typo"

When you find two sants in the kitchen, you are certain to find more sants. Won't fool ME.

This actually dates back to Warren Buffet saying, there is never just one scockroach in the ktichen

Soneshot is a much better technique for such and such type of projects

It's not clear whether the second one isn't a scheapshot at making fun of the first one.

Hey, this is the Internet. It’s serious business! We shouldn’t make sfun of people!

Because the response was written by an LLM running on a tamagotchi

a sentient tamagotchi is not far from us..

So funny, I saw this before bed and planned to look it up in the morning thinking I recently missed out on some kind of killer prompt technique. Woke up and came back to this thread to find typos

Typos are a sign of human intelligence :D

What is “soneshot”?

I think it's a typo of "oneshot", referring to a single prompt being able to handle the entire task.

Yes that was my typo. Sorry about that.

But you did the typo two times?

So true, people think they can prompt AI, tell it to create something amazing and "Novel" so they can feel cool and bragg about it. Shame on such people.

no one said it's amazing or novel, it's just a fun experiement I'm not selling this :) just out of curiousity to explore.

Are you seriously a coder? I did not say that you are claiming novel or amazing. I defined the prompt. Coders never miss anything and you just became so aggressive that goes against hacker news community guidelines.

It's okay to share fun thoughts even if you haven't put blood sweat and tears into them.

lol sure

"claudia slop me up some 600m, commit, push, make 500 pages of slopped README.md, github stars+++"

great, you should definitely do this, there will be absolutely no problems whatsoever and it'll also run at peak speeds while of course being completely unusable because who wants to run a slopped 600m

t. at 1200+ commits on my "oneshot"

No one will run a slopped 600M or even 16B for anything serious today, the whole goal is to experiment, to push both limits of what can be done with the model, and the hardware itself.

It's fun times actually. And I think such experiments and others posted are both showcasing the limits people are pushing.

Well, my netbook. Atom n270, 1GB of RAM, 1.5GB potentially with ZRAM. If it can run in that, a Pentium 4 with SSE2 would be the closest setup down.

this would be a nice challange but from what I did, very doable.. not a usable speed maybe.

It won’t one shot it and it’ll be packed with issues. I know, I have done a few similar toys in the past months. And the biggest challenges will be on the integrity and performance especially if we aren’t too knowledgeable. And considering that I feel the current iteration bullshit more than at one stage this year…

You mentioned one-shot prompts. What you provided is a zero-shot prompt, not a one-shot prompt. The main difference is that one-shot prompts provide an example and use fewer tokens, where zero-shot prompts will tend to have output that differs in format and correctness on every run.