Qwen/Qwen3.5-397B-A17B · Discussions

我问它是谁它有小概率说它是google训练的，，

👀 1

1

#62 opened 4 days ago by

Zhoudaxia2024

Benchmark评测问题？

#61 opened 5 days ago by

whyadnao

[Bug Report] chat_template is vulnerable to LLM hallucinated tool arguments, causing vLLM crash / chat_template 工具调用参数容错 Bug 及修复方案

#60 opened 6 days ago by

reex3625

How to fine-tune this model using the HF Trainer?

#59 opened 7 days ago by

chuarac

How to fine-tune this model using the HF Trainer?

#58 opened 7 days ago by

sungjun12

Audio understanding ability 音频理解能力

#57 opened 7 days ago by

CrisXuantu

Different param dtype between Qwen3.5 and Qwen3-Next

#56 opened 7 days ago by

BestJuly7

Benchmark评测问题？

#55 opened 11 days ago by

TMRBMWK

Add MathArena evaluation result for hmmt/hmmt_feb_2026

#53 opened 12 days ago by

JasperDekoninck

ImportError: cannot import name 'HybridCache' from 'transformers.cache_utils' (/opt/nas/p/conda/envs/tcxia_swift/lib/python3.12/site-packages/transformers/cache_utils.py)

🔥 1

1

#52 opened 13 days ago by

marsmc

Context-Folding Reproducibility | 可复现性 ?

👀 3

#51 opened 13 days ago by

pandemo

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 :-)

2

#48 opened 13 days ago by

Hanswalter

Add MathArena evaluation result for aime/aime_2026

2

#47 opened 15 days ago by

JasperDekoninck

是否可以开源强化学习训练框架？

➕ 🚀 5

#46 opened 15 days ago by

kaishen

Overview and Capabilities of Qwen AI

#38 opened 16 days ago by

BOSSDAN

如果想获得在更多芯片上的可部署版本,可以关注FlagRelease下的产出

👍 1

#36 opened 17 days ago by

YummyYum

qwen

3

#35 opened 17 days ago by

mezyp

qwen3.5_fixed.jinja

#34 opened 18 days ago by

xldistance

Qwen3.5 Unsloth GGUF Evaluation Results

🚀 15

6

#33 opened 18 days ago by

danielhanchen

請求80B參數量的Qwen3.5

👀 1

1

#32 opened 18 days ago by

win10

Suggestion: Dense versions of Qwen 3.5 at multiple scales

2

#31 opened 18 days ago by

zletpm

any AWQ quantized model ??

3

#29 opened 19 days ago by

crystech

牛逼就完事儿了，新春快乐

1

#28 opened 19 days ago by

JoJoGo

Qwen are you doing lower parameter versions?

1

#27 opened 19 days ago by

crownelius

Qwen3.5-VL 啥时候能出呢？

3

#25 opened 20 days ago by

drguolai

Getting error while running the code !pip install -U transformers

#24 opened 20 days ago by

adarshtiwari2505

Template issue maybe

👍 1

1

#23 opened 21 days ago by

sm54

vllm 部署oom

5

#22 opened 21 days ago by

Chris2me

fake knowledge 假知识

6

#21 opened 21 days ago by

aabbccddwasd

Memory Requirements to run `Qwen/Qwen3.5-397B-A17B`

👀 ❤️ 9

4

#20 opened 21 days ago by

alvarobartt

Smaller GGUF without the vision weights?

1

#19 opened 21 days ago by

rtzurtz

Can we get a distilled 2B or 4B?

5

#18 opened 21 days ago by

Abc7347

Base model?

#17 opened 21 days ago by

mrfakename

will there be a smaller version?

👍 🔥 12

2

#16 opened 22 days ago by

iojvsuynv

THANKS QWEN TEAM I'M SO HYPED TO THE QWEN 3.5 LETSS GOOO

❤️ 7

#15 opened 22 days ago by

AxionLab-official

Guide to Run Qwen3.5 locally! 💜

🔥 6

5

#14 opened 22 days ago by

danielhanchen

老铁辛苦了，新年快乐！

#13 opened 22 days ago by

DRuiDong

感谢老铁除夕坚持工作

❤️ 6

#12 opened 22 days ago by

cloudyu

Have already been testing this on OpenRouter

#11 opened 22 days ago by

samcharles93

Thank you thank you thank you!

➕ 🔥 10

#10 opened 22 days ago by

AbheekG

Add evaluation results for HLE, MMLU-Pro

1

#9 opened 22 days ago by

SaylorTwift

This one rock and roll

❤️ 1

#6 opened 22 days ago by

Trilogix1

Congratulations on this release!

🔥 16

2

#3 opened 22 days ago by

Dampfinchen

我问它是谁 它有小概率说它是google训练的，，

Benchmark评测问题？

[Bug Report] chat_template is vulnerable to LLM hallucinated tool arguments, causing vLLM crash / chat_template 工具调用参数容错 Bug 及修复方案

How to fine-tune this model using the HF Trainer?

How to fine-tune this model using the HF Trainer?

Audio understanding ability 音频理解能力

Different param dtype between Qwen3.5 and Qwen3-Next

Benchmark评测问题？

Add MathArena evaluation result for hmmt/hmmt_feb_2026

ImportError: cannot import name 'HybridCache' from 'transformers.cache_utils' (/opt/nas/p/conda/envs/tcxia_swift/lib/python3.12/site-packages/transformers/cache_utils.py)

Context-Folding Reproducibility | 可复现性 ?

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 :-)

Add MathArena evaluation result for aime/aime_2026

是否可以开源强化学习训练框架？

Overview and Capabilities of Qwen AI

如果想获得在更多芯片上的可部署版本,可以关注FlagRelease下的产出

qwen

qwen3.5_fixed.jinja

Qwen3.5 Unsloth GGUF Evaluation Results

請求80B參數量的Qwen3.5

Suggestion: Dense versions of Qwen 3.5 at multiple scales

any AWQ quantized model ??

牛逼就完事儿了，新春快乐

Qwen are you doing lower parameter versions?

Qwen3.5-VL 啥时候能出呢？

Getting error while running the code !pip install -U transformers

Template issue maybe

vllm 部署oom

fake knowledge 假知识

Memory Requirements to run `Qwen/Qwen3.5-397B-A17B`

Smaller GGUF without the vision weights?

Can we get a distilled 2B or 4B?

Base model?

will there be a smaller version?

THANKS QWEN TEAM I'M SO HYPED TO THE QWEN 3.5 LETSS GOOO

Guide to Run Qwen3.5 locally! 💜

老铁辛苦了，新年快乐！

感谢老铁除夕坚持工作

Have already been testing this on OpenRouter

Thank you thank you thank you!

Add evaluation results for HLE, MMLU-Pro

This one rock and roll

Congratulations on this release!

我问它是谁它有小概率说它是google训练的，，