Qwen

/

Qwen3.5-397B-A17B

like 1.47k

Follow

Qwen 80.6k

Hi there,

Thanks for releasing this model!

I’m looking to perform SFT using the Hugging Face Trainer API. Could you provide a basic guide or a code snippet for this?

I also have a quick question regarding data preprocessing: In Qwen3-VL, images were typically handled via the processor, but I’ve noticed some changes in Qwen3.5. Does the tokenizer now handle the image placeholders directly, or should I still be using a unified processor?

Thanks in advance for your help!