Qwen/Qwen3-235B-A22B · Discussions

yarn scale to 122,880 context length

#41 opened 2 days ago by

nbroad

chat_template issue

#40 opened 18 days ago by

Alex-Chan

dannt

#39 opened 21 days ago by

DDDanny09

Add assistant mask support to Qwen3-235B-A22B

🚀 2

#38 opened 24 days ago by

waleko

system prompt suggestion

1

#37 opened about 1 month ago by

Shuaiqi

Publish to GitHub Marketplace

#36 opened about 1 month ago by

typed-sigterm

The Qwen3-235B-A22B model is not as effective as the Qwen3-32B model.

🔥 1

#35 opened about 2 months ago by

czqqq

Video?

2

#34 opened about 2 months ago by

jujutechnology

关于Qwen3-235B-A22B在文学创作中“思维链”模式对输出风格影响的观察与建议

🧠 👍 8

#33 opened about 2 months ago by

yxcl6874

In complex reasoning tasks Qwen3 is far behind QwQ

12

#32 opened about 2 months ago by

AdamF92

I know not related but saw This Star Wars picture on Facebook and i thought it is code so trying to say Hi to Someone named John!

#31 opened about 2 months ago by

ebearden

Not sure Clocks keep coming up and keep getting the run around might need time stamping confirmations?

#30 opened about 2 months ago by

ebearden

Not Sure Orginal to Persian to Binary?

#29 opened about 2 months ago by

ebearden

Qwen3 is simply amazing.

#28 opened about 2 months ago by

Trilogix1

Upload b891b3c3bb6a146a8e809bb72a06d101.png

#27 opened about 2 months ago by

Jalil16

Add image visual recognition output just like qwen 2.5 vl-32b instruct

6

#26 opened about 2 months ago by

devopsML

English to French - French to English based on Meta & HuggingFace Chat Bot

#25 opened about 2 months ago by

ebearden

Qwen3 幻觉太高了，比 Qwen 2.5 差太多了

➕ 1

9

#24 opened about 2 months ago by

hehua2008

Upload 3 files

#23 opened about 2 months ago by

neuroQuantu

Upload 3 files

1

#22 opened about 2 months ago by

neuroQuantu

Model keeps talking about Cumhurbaşkanlığı Sarayı when speaking Turkish

#21 opened about 2 months ago by

aeminkocal

Qwen3 not Using Tools in Complex Prompts Unlike QwQ-32B

8

#20 opened 2 months ago by

Anaudia

Thanks a lot for this release

🔥 2

#19 opened 2 months ago by

Volko76

Does anyone feel Qwen3 often fails to follow instructions accurately?

🚀 7

7

#18 opened 2 months ago by

DOFOFFICIAL

Two of the base models are missing

➕ 1

1

#17 opened 2 months ago by

ZhangRC

Qwen is loosing broad knowledge since Qwen2.

🔥 👍 11

16

#16 opened 2 months ago by

phil111

GPQA perf for DSV3-Base seems wrong

➕ 3

1

#15 opened 2 months ago by

AChen-qaq

72B-MoE

👍 4

#13 opened 2 months ago by

avalonsec

235B会放出来Base模型吗？

➕ 7

#12 opened 2 months ago by

Yantao2009

看模型介绍和模型结构里面没有关于vision encoder的部分，但是在qwen的在线模型服务界面可以用这个模型去看图片，想问下视觉部分是复用了哪个vision encoder呢？

5

#11 opened 2 months ago by

Chloez

有用4张H20实践过的大佬吗

2

#10 opened 2 months ago by

Edison0902

8张80G显存的8卡A100能部署不？

10

#9 opened 2 months ago by

Yuxin362

User rating and reviews of Qwen3 App and Qwen3 Model

#8 opened 2 months ago by

DeepNLP

是不是奖励函数没有ngram重复度惩罚

2

#7 opened 2 months ago by

wzx111

🚀[Fine-tuning] Qwen3-MoE Megatron Training Implementation and Best Practices👋

🚀 6

1

#6 opened 2 months ago by

study-hjt

【Evaluation】Best practice for evaluating Qwen3 !!

🔥 👍 4

#5 opened 2 months ago by

wangxingjun778

Please upload the base model for this one

👍 5

#4 opened 2 months ago by

mesh-ops

GPTQ/AWQ

👀 12

4

#3 opened 2 months ago by

ndurkee

Add languages tag

#2 opened 2 months ago by

de-francophones