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

41–50 of 100 posts

Re: Qwen 3.7 Preview

#41

Post by b3ing »

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

Re: Qwen 3.7 Preview

#42

Post by Havoc »

So glad they’re holding steady on open weights.

At least for now. Worried the Chinese team will change their mind once they have parity

Re: Qwen 3.7 Preview

#43

Post by trilogic »

tedivm wrote: post #32
Aurornis wrote: post #31

Earlier quoted context omitted.

> Is not far at all from proprietary models if you give it tools, skills and agents etc, I use Qwen 3.6 27B, the dense version of this model which is slightly better. I don't agree that it's close at all. Maybe for some small, easy tasks, but not for working on real codebases. It's amazing for something I can run at home, but the difference between it and Opus or GPT-5.5 is huge.

I've had the opposite experience, and have built multiple fantastic applications with Qwen3.6 27b. What quantization have you tested with?

As funny as it may sound a q4_k_m well converted and quantized properly (and finetuned, impereative) would do the job. The 27b it may be good but is heavy, it burns the hardware. I personally prefer the 397B if I am stucked and can´t progress, it can still run with 7 tps. Now with the Mtp (multitoken prediction) it nearly double the speed ( reached 82tps today with the 35b 100000ctx). I recommend it you give it a try.

Re: Qwen 3.7 Preview

#44

Post by satvikpendem »

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

Pretty sure they meant 4.6

Re: Qwen 3.7 Preview

#46

Post by rspoerri »

trilogic wrote: post #39
trilogic wrote: post #36

Earlier quoted context omitted.

Really, how so? Because we work with codebases daily, can you tell us a concrete example! In our case we work in consumer hardware (ish), 10 million ctx (1 million output, 1 million input proven, sometimes it loops or breaks at over 500k ctx byt at ~17tps linear). IT can read the full codebase, unleash agents, and write in disk editing and patching files creating a full app in 3-4 minutes. IT can do Web search and Ra…

Edit: Forgot to mention that it can process images and pdf, and 100s of other files, it can even create presentations in code or mermaid, svg, charts js etc. Here a basic version of it: https://hugston.com/chat

how do you do 1mio context with qwen3.6 27b, that only supports 256k? and what hardware would you run that on? 2 * 3090 is afaik currently at max 256k context.

Re: Qwen 3.7 Preview

#48

Post by omneity »

rspoerri wrote: post #46
trilogic wrote: post #39

Earlier quoted context omitted.

Edit: Forgot to mention that it can process images and pdf, and 100s of other files, it can even create presentations in code or mermaid, svg, charts js etc. Here a basic version of it: https://hugston.com/chat

how do you do 1mio context with qwen3.6 27b, that only supports 256k? and what hardware would you run that on? 2 * 3090 is afaik currently at max 256k context.

You can increase the context window beyond its max trained context using RoPE scaling[0] which will require more VRAM.

But you can increase your context window for the same VRAM by quantizing the KV cache with FP8 (double the context) or TurboQuant (more than double)[1].

0: https://medium.com/@leannetan/extending-context-length-with-...

1: https://docs.vllm.ai/en/latest/features/quantization/quantiz...

Re: Qwen 3.7 Preview

#49

Post by trilogic »

rspoerri wrote: post #46
trilogic wrote: post #39

Earlier quoted context omitted.

Edit: Forgot to mention that it can process images and pdf, and 100s of other files, it can even create presentations in code or mermaid, svg, charts js etc. Here a basic version of it: https://hugston.com/chat

how do you do 1mio context with qwen3.6 27b, that only supports 256k? and what hardware would you run that on? 2 * 3090 is afaik currently at max 256k context.

We managed to increase the ctx for whatever llm model that is GGUFED, here the experimental tests: https://www.reddit.com/r/Hugston/

Re: Qwen 3.7 Preview

#50

Post by nyrikki »

rspoerri wrote: post #46
trilogic wrote: post #39

Earlier quoted context omitted.

Edit: Forgot to mention that it can process images and pdf, and 100s of other files, it can even create presentations in code or mermaid, svg, charts js etc. Here a basic version of it: https://hugston.com/chat

how do you do 1mio context with qwen3.6 27b, that only supports 256k? and what hardware would you run that on? 2 * 3090 is afaik currently at max 256k context.

You can get all the Qwen 3.x models up to ~1 million tokens using YaRN with llama.cpp.[0]

Personally I am using `--no-context-shift` and feeding in context back in on failure at the harness level.

I have 2x1080ti + 1xTitanV that have a full 262,144 tokens context on 262,144 tokens with `-sm tensor` at 62.04 t/s which isn't so bad.

But I also have a 1x3090 running unsloth/Qwen3.6-27B-MTP-GGUF:UD-Q4_K_XL at 41.89 t/s but with only 130k context, but if you have a modular programming style both work pretty well.

But play with YaRN if you really need it.

[0]https://qwen.readthedocs.io/en/v3.0/run_locally/llama.cpp.ht...

Post reply on HN