Qwen/Qwen3-30B-A3B · Discussions

I got Qwen to "evolve" beyond it's core programming.

1

#38 opened 3 days ago by

Koitenshin

When new version?

#37 opened 3 days ago by

MajinAnix

model-download

#36 opened 14 days ago by

Vaishnavi-Garad

add AIBOM

#35 opened about 1 month ago by

fatima113

How to train the Qwen3-30B-A3B using Reinforcement Learning?

#34 opened about 2 months ago by

alanayu

Add assistant mask support to Qwen3-30B-A3B

#33 opened about 2 months ago by

waleko

There are something look like bugs.

🔥 1

#32 opened about 2 months ago by

Jim-202020

Missing license file?

#31 opened 2 months ago by

cpwan9086

Over thinking. take too long time to produce final answer to some question.

1

#30 opened 2 months ago by

rosspanda0

how can I train a model if it uses a unique tokenizer?

#29 opened 2 months ago by

Veterse

Why does this problem "KeyError: 'qwen3_moe'" still occur after I update the transformers to 4.51.3

1

#28 opened 2 months ago by

Eureka9302

Why not distinguish between sequence_length equal to 1 and greater than 1 in the MoE module's forward function?

#27 opened 2 months ago by

nifeng154

asd-Qwen3-30B-A3B

#25 opened 3 months ago by

Axdar

Many experts seem to be underutilized, possible optimization potential.

#24 opened 3 months ago by

Dampfinchen

repetition

➕ 2

4

#23 opened 3 months ago by

Dampfinchen

Qwen3-30B-A3B

#22 opened 3 months ago by

FinitusEst

awq

#21 opened 3 months ago by

MrDragonFox

qwen3-30B

#20 opened 3 months ago by

xiaohuihehe

Comparing Qwen3-30B-A3B and Phi-4-mini-reasoning

🚀 1

#19 opened 3 months ago by

gptlocalhost

Qwen3 is great, but could be better.

👍 9

25

#18 opened 3 months ago by

phil111

Collections of Bad Cases User Reviews and Comments of Qwen3 30B A3B.

#17 opened 3 months ago by

DeepNLP

Doploy to Sagemaker error

#16 opened 3 months ago by

Hanyu1023

Enabling or disabling reasoning on-demand with remote vLLM server?

1

#15 opened 3 months ago by

abhinavkulkarni

asd-qwen1

#14 opened 3 months ago by

Axdar

Minor changes, big gains -- Huggingface MoE modeling enhancement

👍 👀 5

3

#13 opened 3 months ago by

xiaowei4ai

AWQ quantized model support timeline?

👍 7

2

#12 opened 3 months ago by

hyunw55

image support

🧠 👀 12

8

#9 opened 3 months ago by

kuliev-vitaly

Waiting for the Qwen3-VL

👀 7

7

#8 opened 3 months ago by

Maverick17

Very fast and powerful, but with one glaring weakness.

👍 1

2

#6 opened 3 months ago by

phil111

MoE question

🤯 8

4

#5 opened 3 months ago by

bojan2501

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

🚀 4

#3 opened 3 months ago by

study-hjt

【Evaluation】Best practice for evaluating Qwen3 !!

👍 🔥 7

1

#2 opened 3 months ago by

wangxingjun778

Add languages tag

#1 opened 3 months ago by

de-francophones