Key error exception when finetuning Qwen models with LoRA

When I try to :
1. finetune Qwen
2. KD with Qwen 
Both using LoRA, I encounter a KeyError exception 

<img width="1650" height="671" alt="Image" src="https://github.com/user-attachments/assets/f233c401-ca93-413f-87a8-dcb185258cc5" />
 
