Qwen
/

Qwen3.5-35B-A3B

Image-Text-to-Text

Model card Files Files and versions

Resources

View closed (3)

wow cool

#25 opened about 22 hours ago by

kallyh

#24 opened about 22 hours ago by

Is `qwen3_nonthinking.jinja` available for disabling thinking?

#23 opened about 23 hours ago by

Update README.md

#21 opened 1 day ago by

My suggestion

#20 opened 1 day ago by

Reasoning content leaks into `message.content` with JSON schema response format

#18 opened 1 day ago by

Quantization AWQ INT4

#17 opened 1 day ago by

I would like a recommended training environment setup for the Qwen3.5-MoE model (e.g., Qwen3.5-35B-A3B, model_type: qwen3_5_moe).

#16 opened 1 day ago by

数字和中文之间多了空格，在某些场景下完全用不了。

#15 opened 1 day ago by

Update README.md

#12 opened 2 days ago by

Add evaluation results

#11 opened 2 days ago by

Context-Folding Reproducibility | 可复现性 ?

#10 opened 2 days ago by

Chat template for Pi AGent

#9 opened 2 days ago by

nimishchaudhari

LESS GOo

#8 opened 2 days ago by

🚀 Qwen3.5 Dense/MoE Training Support

#7 opened 2 days ago by

Installation Video and Testing - Step by Step

#6 opened 3 days ago by

Why is this part of the default template?

#5 opened 3 days ago by

tool calling chat template is broken

#4 opened 3 days ago by

Instruct mode metrics

#3 opened 3 days ago by

Great release! Thank you Qwen!

#2 opened 3 days ago by

请千万千万别忘了把 Qwen Image 2.0 也开源——这对我们本地用户来说会是个巨大的改变 :-)

#1 opened 3 days ago by