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

Qwen3: Think deeper, act faster

qwenlm.github.io

Post reply on HN
Print view View original

311–320 of 412 posts

Re: Qwen3: Think deeper, act faster

#311

Post by krosaen »

kaoD wrote: post #256
mrkeen wrote: post #253

Earlier quoted context omitted.

As they say, we shouldn't judge AI by the current state-of-the-art, but by how far and fast it's progressing. I can't wait to see future models get it even more wrong than that.

Personally (anecdata) I haven't experienced any practical progress in my day-to-day tasks for a long time, no matter how good they became at gaming the benchmarks. They keep being impressive at what they're good at (aggregating sources to solve a very well known problem) and terrible at what they're bad at (actually thinking through novel problems or old problems with few sources). E.g. all ChatGPT, Claude and Gemini…

I'm curious what kind of prompting or context you are providing before asking for a liquid soap script - or if you've tried using Cursor and providing a bunch of context with documentation about liquid soap as part of it. My guess was these kinds of things get the models to perform much better. I have seen this work with internal APIs / best practices / patterns.

Re: Qwen3: Think deeper, act faster

#312

Post by Havoc »

foundry27 wrote: post #79

I find the situation the big LLM players find themselves in quite ironic. Sam Altman promised (edit: under duress, from a twitter poll gone wrong) to release an open source model at the level of o3-mini to catch up to the perceived OSS supremacy of Deepseek/Qwen. Now Qwen3’s release makes a model that’s “only” equivalent to o3-mini effectively dead on arrival, both socially and economically.

OAI in general seems to be treading water at best.

Still topping a lot of leaderboards but severely reduced rep. Chaotic naming, „ClosedAI“ image, undercut on pricing, competitors with much better licensing/open weights, stargate talk about Europe, Claude being seen as superior for coding etc. nothing end of the world but a lot of lukewarm misses

If I was an investor with financials that basically require magical returns from them to justify Vals I’d be worried.

Re: Qwen3: Think deeper, act faster

#313

Post by fkyoureadthedoc »

Imustaskforhelp wrote: post #289
echelon wrote: post #93

Earlier quoted context omitted.

Alibaba, I have a huge favor to ask if you're listening. You guys very obviously care about the community. We need an answer to gpt-image-1. Can you please pair Qwen with Wan? That would literally change the art world forever. gpt-image-1 is an almost wholesale replacement of ComfyUI and SD/Flux ControlNets. I can't underscore how big of a deal it is. As such, OpenAI has leapt ahead and threatens to start capturing m…

I don't know, the AI image quality has gotten good but it's still slop. We are forgetting what makes art, well art. I am not even an artist but yeah I see people using AI for photos and they were so horrendous pre chatgpt-imagen that I had literally told one person if you are going to use AI images, might as well use chatgpt for it. Also though I would also like to get something like chatgpt-image generating qualitie…

On the other hand, ChatGPT image generation is a lot of fun to use. I'd never pay a human artist to make the meme tier images I use it for.

Re: Qwen3: Think deeper, act faster

#314

Post by Havoc »

tough wrote: post #8
miohtama wrote: post #5

> The pre-training process consists of three stages. In the first stage (S1), the model was pretrained on over 30 trillion tokens with a context length of 4K tokens. This stage provided the model with basic language skills and general knowledge. As this is in trillions, where does this amount of material come from?

Synthetic Data (after reasoning breakthroughs feels like more AI laabs are betting for synthetic data to scale.) wonder at what price

If they’re using vision models to extract pdf data then they can’t be shy on throwing money at it

Re: Qwen3: Think deeper, act faster

#315

Post by hbbio »

ammo1662 wrote: post #273
hbbio wrote: post #263

Earlier quoted context omitted.

Feel free to add a PR :) What is the parameter?

Just add "/no_think" in your prompt. https://qwenlm.github.io/blog/qwen3/#advanced-usages

Thanks!

Turns out just is not the word here. My benchmark is made using conversations, where there is a SystemMessage and some structured content in a UserMessage.

But Qwen3 seems to ignore /no_think when appended to the SystemMessage. I can try to add it to the structured content but that will be a bit weird. Would have been better to have a "think" parameter like temperature.

Re: Qwen3: Think deeper, act faster

#316

Post by jim180 »

ggregoryarms wrote: post #288
jim180 wrote: post #274

Earlier quoted context omitted.

Absolutely. All models ar terrible with Objective-C and Swift, compared to let's say JS/HTML/Python. However, I've realized that Claude Code is extremely useful for generating somewhat simple landing pages for some of my projects. It spits out static html+js which is easy to host, with somewhat good looking design. The code isn't the best and to some extent isn't maintainable by a human at all, but it gets the job do…

Building a basic static html landing page is ridiculously easy though. What js is even needed? If it's just an html file and maybe a stylesheet of course it's easy to host. You can apply 20 lines of css and have a decent looking page. These aren't hard problems.

Laziness mostly - no need to think about design, icons and layout (responsiveness and all that stuff).

These are not hard problems obviously, but getting to 80%-90% is faster than doing it by hand and in my cases that was more than enough.

With that being said, AI failed for the rest 10%-20% with various small visual issues.

Re: Qwen3: Think deeper, act faster

#317

Post by WhitneyLand »

tomrod wrote: post #143
WhitneyLand wrote: post #127

Earlier quoted context omitted.

The plural of reliable data is not anecdote.

Depends on the data generating process.

Of course, but then you have a system of gathering information with some rigor which is more than merely a collection of anecdotes. That becomes the difference.

Re: Qwen3: Think deeper, act faster

#318

Post by archerx »

sega_sai wrote: post #37

With all the different open-weight models appearing, is there some way of figuring out what model would work with sensible speed (> X tok/s) on a standard desktop GPU ? I.e. I have Quadro RTX 4000 with 8G vram and seeing all the models https://ollama.com/search here with all the different sizes, I am absolutely at loss which models with which sizes would be fast enough. I.e. there is no point of me downloading the la…

On hugging face, if you tell them which GPU you have the models that will run decently will have a green icon.

Re: Qwen3: Think deeper, act faster

#319

Post by kaoD »

krosaen wrote: post #311
kaoD wrote: post #256

Earlier quoted context omitted.

Personally (anecdata) I haven't experienced any practical progress in my day-to-day tasks for a long time, no matter how good they became at gaming the benchmarks. They keep being impressive at what they're good at (aggregating sources to solve a very well known problem) and terrible at what they're bad at (actually thinking through novel problems or old problems with few sources). E.g. all ChatGPT, Claude and Gemini…

I'm curious what kind of prompting or context you are providing before asking for a liquid soap script - or if you've tried using Cursor and providing a bunch of context with documentation about liquid soap as part of it. My guess was these kinds of things get the models to perform much better. I have seen this work with internal APIs / best practices / patterns.

Yes, I used Cursor and tried providing both the whole Liquidsoap book or the URL to the online reference just in case the book was too large for context or it was triggering some sort of RAG.

Not successful.

It's not that it didn't do what I wanted: most of the time it didn't even run. Iterating on the error messages just arrived at progressively dumber not-solutions and running in circles.

Re: Qwen3: Think deeper, act faster

#320

Post by claiir »

simonw wrote: post #207

As is now traditional for new LLM releases, I used Qwen 3 (32B, run via Ollama on a Mac) to summarize this Hacker News conversation about itself - run at the point when it hit 112 comments. The results were kind of fascinating, because it appeared to confuse my system prompt telling it to summarize the conversation with the various questions asked in the post itself, which it tried to answer. I don't think it did a g…

o1-preview had this same issue too! You’d give it a long conversation to summarize, and if the conversation ended with a question, o1-preview would answer that, completely ignoring your instructions.

Generally unimpressed with Qwen3 from my own personal set of problems.

Post reply on HN