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

91–100 of 412 posts

Re: Qwen3: Think deeper, act faster

#91

Post by CSMastermind »

simonw wrote: post #38

Something that interests me about the Qwen and DeepSeek models is that they have presumably been trained to fit the worldview enforced by the CCP, for things like avoiding talking about Tiananmen Square - but we've had access to a range of Qwen/DeepSeek models for well over a year at this point and to my knowledge this assumed bias hasn't actually resulted in any documented problems from people using the models. Asid…

Right now these models have less censorship than their US counterparts.

With that said, they're in a fight for dominance so censoring now would be foolish. If they win and establish a monopoly then the screws will start to turn.

Re: Qwen3: Think deeper, act faster

#93

Post by echelon »

Jayakumark wrote: post #53
natrys wrote: post #2

They have got pretty good documentation too[1]. And Looks like we have day 1 support for all major inference stacks, plus so many size choices. Quants are also up because they have already worked with many community quant makers. Not even going into performance, need to test first. But what a stellar release just for attention to all these peripheral details alone. This should be the standard for major release, inste…

Second this , they patched all major llm frameworks like llama.cpp, transformers , vllm, sglang, ollama etc weeks before for qwen3 support and released model weights everywhere around same time. Like a global movie release. Cannot undermine mine this level of detail and effort.

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 more of the market for AI images and video. The expense of designing and training a multimodal model presents challenges to the open source community, and it's unlikely that Black Forest Labs or an open effort can do it. It's really a place where only Alibaba can shine.

If we get an open weights multimodal image gen model that we can fine tune, then it's game over - open models will be 100% the future. If not, then the giants are going to start controlling media creation. It'll be the domain of OpenAI and Google alone. Firing a salvo here will keep media creation highly competitive.

So please, pretty please work on an LLM/Diffusion multimodal image gen model. It would change the world instantly.

And keep up the great work with Wan Video! It's easily going to surpass Kling and Veo. The controllability is already well worth the tradeoffs.

Re: Qwen3: Think deeper, act faster

#94

Post by xiphias2 »

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…

When I tested Qwen with different sizes / quants, generally the 8-bit quant versions had the best quality for the same speed.

4-bit was ,,fine'', but a smaller 8-bit version beat it in quality for the same speed

Re: Qwen3: Think deeper, act faster

#96

Post by horhay »

kristofferR wrote: post #76
a3w wrote: post #25

Earlier quoted context omitted.

AGIs probably comes from neurosymbolic AI. But LLMs could be the neuro-part of that. On the other hand, LLM progress feels like bullshit, gaming benchmarks and other problems occured. So either in two years all hail our AGI/AMI (machine intelligence) overlords, or the bubble bursts.

You can't possibly use LLMs day to day if you think the benchmarks are solely gamed. Yes, there's been some cases, but the progress in real-life usage tracks the benchmarks overall. Gemini 2.5 Pro for example is absurdly more capable than models from a year ago.

They aren't lying in the way that LLMs have been seeing improvement, but benchmarks suggesting that LLMs are still scaling exponentially are not reflective of where they truly are.

Re: Qwen3: Think deeper, act faster

#97

Post by theturtletalks »

nyclounge wrote: post #90
eunos wrote: post #42

Earlier quoted context omitted.

The avoiding talking part is more on the Frontend level censorship I think. It doesn't censor on API

This is NOT true. At least on the 1.5B version model on my local machine. It blocks answers when using offline mode. Perplexity has an uncensored a version, but don't thing it is open on how they did it.

Didn't know Perplexity cracked R1's censorship but it is completely uncensored. Anyone can try even without an account: https://labs.perplexity.ai/. HuggingFace also was working on Open R1 but not sure how far they got.

Re: Qwen3: Think deeper, act faster

#98

Post by stavros »

I have a small physics-based problem I pose to LLMs. It's tricky for humans as well, and all LLMs I've tried (GPT o3, Claude 3.7, Gemini 2.5 Pro) fail to answer correctly. If I ask them to explain their answer, they do get it eventually, but none get it right the first time. Qwen3 with max thinking got it even more wrong than the rest, for what it's worth.

Re: Qwen3: Think deeper, act faster

#99

Post by kenjackson »

stavros wrote: post #98

I have a small physics-based problem I pose to LLMs. It's tricky for humans as well, and all LLMs I've tried (GPT o3, Claude 3.7, Gemini 2.5 Pro) fail to answer correctly. If I ask them to explain their answer, they do get it eventually, but none get it right the first time. Qwen3 with max thinking got it even more wrong than the rest, for what it's worth.

You really had me until the last half of the last sentence.

Re: Qwen3: Think deeper, act faster

#100

Post by mks_shuffle »

Does anyone have insights on the best approaches to compare reasoning models? It is often recommended to use a higher temperature for more creative answers and lower temperature values for more logical and deterministic outputs. However, I am not sure how applicable this advice is for reasoning models. For example, Deepseek-R1 and QwQ-32b recommend a temperature around 0.6, rather than lower values like 0.1–0.3. The Qwen3 blog provides performance comparisons between multiple reasoning models, and I am interested in knowing what configurations they used. However, the paper is not available yet. If anyone has links to papers focused on this topic, please share them here. Also, please feel free to correct me if I’m mistaken about anything. Thanks!
Post reply on HN