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

Qwen 3.7 Preview

twitter.com

Post reply on HN
Print view View original

81–90 of 100 posts

Re: Qwen 3.7 Preview

#81

Post by metalliqaz »

bachmeier wrote: post #57

I'm not much interested in vibe coding (for those who aren't aware that LLMs have other uses). The specific model I've been using with Ollama is hf.co/unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF:UD-Q4_K_XL and it's amazing how fast it is on 64 GB of RAM and i5-13400 CPU. No GPU on this computer. Gemma 4 E4B will think for a couple of minutes vs 3-5 seconds for Qwen. It's hard to believe how much you can do with such li…

I have a much more powerful PC and I would not call Qwen3-Coder-30B-A3B "fast" on my machine by any stretch of the word. How are you running it?

Re: Qwen 3.7 Preview

#82

Post by nullc »

b3ing wrote: post #41
trilogic wrote: post #12

Qwen 3.6 35B (finetuned) is so good that it became standard open weights for everyday use. Is not far at all from proprietary models if you give it tools, skills and agents etc, it can actually finish the job. (Thank you Qwen team, appreciated). Using opensource now we can definitely rely to design from scratch very complicated architecture and build pretty fast the full pack. Wish to see Europe AI unleashed, wake up…

For coding it’s really bad. Writing is ok, chat is good. It’ll get better but it’s not that close yet

Bad is mystifying. Unassisted but for handing it a pile of PDFs of relevant academic papers and my initial codebase, I had hermes agent based on qwen-3.6 27B implement karatsuba multiplication of characteristic-2 polynomials in C++ in an existing codebase with an internal field arithmetic library. It correctly found the 'obvious' optimizations using the field properties. Then I had it implement the recursive halfgcd algorithm for these polynomials using it.

It wrote extensive test cases and validated them with mutation testing (per my standard instructions)-- took many tries getting the algorithms right but with the tests handy it found and fixed the errors.

It's inconceivable to me to call it bad!

Re: Qwen 3.7 Preview

#83

Post by boppo1 »

Jach wrote: post #63
rspoerri wrote: post #4

I am very interested in seeing new qwen models. Qwen3.6 27b is the first one that can do things and doesnt constantly loose "it's mind" and that can be run on a 3090 with a good context size. But it's sometimes getting into a loop.

I sort of thought this about qwen3.5 35b, finally a local model that isn't a complete waste of electricity, but "upgrading" to 3.6 35b left me disappointed. It seemed more like a downgrade. But honestly I've barely used either. Subjectively they still seem far from the frontier models, but for what they can do, it's great to be able to do locally.

How are they just for chat / questions?

Re: Qwen 3.7 Preview

#84

Post by tedivm »

2001zhaozhao wrote: post #34
tedivm wrote: post #10

Earlier quoted context omitted.

I've completely replaced GitHub Copilot using Sonnet 3.6 with OpenCode using Qwen3.6 27b, and it's been a great experience.

Is Sonnet 3.6 a typo? Claude Sonnet 3.6 (aka 3.5 New) is an ancient model from 2024

Yeah that was a typo, I meant 4.6.

Re: Qwen 3.7 Preview

#85

Post by vessenes »

vessenes wrote: post #15

Today I learned Meta's new model is preferred to everything but claude. That is .. a real surprise! Congrats to the Meta team.

I don’t mind a principled downvote, but can a downvoter explain his or her reasoning? Genuinely curious. I found the linked rankings surprising, and think of myself as relatively well informed. Please, enlighten me..

Re: Qwen 3.7 Preview

#86

Post by jononor »

cyanydeez wrote: post #80
vessenes wrote: post #19

Earlier quoted context omitted.

That would be nice, but it's not going to be possible. Any open benchmark has a very short life, since it will be pulled in and DPO / RL trained quickly for benchmaxxing purposes. So, you'll need a private test to have a hope of something fair. (These also get leaked over time, btw, so even then there's a window of usability). These are expensive to run. Now consider that there might be 15-20 viable quants for a give…

there are benchmarks that have nothing to do with the training material, but with how the models are capable of things like reading code: https://needle-bench.cc/ Generally, you give them a document and you ask them to retrieve some subsection of the document then rate them on what they retrieved. You can always find enough random documents, or create your own, to always run these and you can make it arbitrarily long…

This seems like a viable eval strategy. Presumably finding a bug requires some degree of understanding of the code, beyond just information retrieval. However it probably does not measure things like prompt adherence or ability to create code that implements a specification?

Re: Qwen 3.7 Preview

#87

Post by pancsta »

metalliqaz wrote: post #81
bachmeier wrote: post #57

I'm not much interested in vibe coding (for those who aren't aware that LLMs have other uses). The specific model I've been using with Ollama is hf.co/unsloth/Qwen3-Coder-30B-A3B-Instruct-GGUF:UD-Q4_K_XL and it's amazing how fast it is on 64 GB of RAM and i5-13400 CPU. No GPU on this computer. Gemma 4 E4B will think for a couple of minutes vs 3-5 seconds for Qwen. It's hard to believe how much you can do with such li…

I have a much more powerful PC and I would not call Qwen3-Coder-30B-A3B "fast" on my machine by any stretch of the word. How are you running it?

Probably in the chat / no sys prompt mode for docs only. Try to refac a codebase using a CPU only…

Re: Qwen 3.7 Preview

#88

Post by Vaskivo »

nyrikki wrote: post #71
nyrikki wrote: post #67

Earlier quoted context omitted.

(Note UPDATED config) Ya, if you are using the CPU it may slowdown quick. This may be a bit huge and overcomplicated, on this host I am running it on a AMD Ryzen 7 5700G so that I can use the APU to dedicate the 3090. podman run --device nvidia.com/gpu=all -d -v llama_qwen3.6mpt:/root/.cache -p 8080:8080 local/llama.cpp:full-cuda --server \ -hf unsloth/Qwen3.6-27B-MTP-GGUF:UD-Q4_K_XL \ -ngl 99 \ --ctx-size 131072 \ -…

That config looked too complicated, getting rid of the --prio 3 and --poll 100, setting the draft-n-max to now recommended values, etc... kicked it up to 61 t/s I think that was all about some earlier crashes. podman run --device nvidia.com/gpu=all -d -v llama_qwen3.6mpt:/root/.cache -p 8080:8080 local/llama.cpp:full-cuda --server \ -hf unsloth/Qwen3.6-27B-MTP-GGUF:UD-Q4_K_XL \ -ngl 99 \ --ctx-size 128000 \ --no-mmpr…

Yeah, having even a little bit in the CPU tanks the t/s...

But thanks. I've learned a few more configurations to tinker with.

Re: Qwen 3.7 Preview

#89

Post by jwr »

argee wrote: post #62
jwr wrote: post #9

Earlier quoted context omitted.

Agreed. Incidentally, in my testing, qwen models (qwen3.6-35b-a3b and earlier 3.5) are WAY better with vision than gemma4-26b-a4b. I would normally want to stick with gemma4 only (I use it for spam filtering), but it just doesn't cut it for vision work, and qwen models do.

> qwen models (qwen3.6-35b-a3b and earlier 3.5) are WAY better with vision than gemma4-26b-a4b Can you give an example? And/or is there a benchmark specifically for this?

I have my own benchmark: a photo of filament boxes in my drawer. The task is to list (inventory) the filaments. It's not an easy task: the letters are small, the boxes are in various orientations, the labels have lots of writing on them, etc.

This is what qwen produces (basically perfect):

Fiberlogy (Blue Boxes - Left Side)

ASA Black: 750 g (0.75 kg)

Easy PLA Graphite: 850 g (0.85 kg)

Ref Easy PLA Gray: 850 g (0.85 kg)

Ref Easy PLA White: 850 g (0.85 kg)

Ref Easy PLA Inox: 850 g (0.85 kg)

Bambu Lab (Brown Boxes - Right Side)

ASA White: 750 g (0.75 kg)

PLA-CF (Burgundy Red): 1,000 g (1 kg)

PAHT-CF (Black): 500 g (0.5 kg)

PET-CF (Black): 500 g (0.5 kg)

PETG-CF (Black): 1,000 g (1 kg)

----------

And this is gemma-4-26b-a4b (basically useless):

### *Fiberlogy Filaments*

All Fiberlogy items listed below share these common specifications: Made in Poland, $1.75\text{mm}$ Diameter, $0.75\text{kg}$ Weight, and Print Temperature of $200-230^{\circ}\text{C}$.

* *PLA Black* (Barcode: 5903281997986)

* *Easy PLA Graphite* (Barcode: 5903281994057)

* *Ref Easy PLA Grey*

* *Ref Easy PLA White* (Barcode: 5903281994386)

* *Ref Easy PLA True...* (Text is cut off)

**

### *Bambu Lab Filaments*

All Bambu Lab items listed below share these common specifications: Made in China, Removable Spool (Do Not Remove).

* *PLA-CF (Carbon Fiber Reinforced)*

    \*   Color: Burgundy Red
    \*   Diameter: $1.75 \pm 0.02\text{mm}$
    \*   Weight: $1.0\text{kg}$
    \*   Suggested Drying Conditions: $45^{\circ}\text{C}$ for $6-12$ hours
* *PAHT-CF (High Temperature Polyamide with Carbon Fiber)*

    \*   Color: Black
    \*   Diameter: $1.75 \pm 0.02\text{mm}$
    \*   Weight: $0.5\text{kg}$
    \*   Suggested Drying Conditions: $80^{\circ}\text{C}$ for $6-12$ hours
* *PETG-CF (Carbon Fiber Reinforced)*

    \*   Color: Black
    \*   Diameter: $1.75 \pm 0.02\text{mm}$
    \*   Weight: $1.0\text{kg}$
Post reply on HN