How to fine tune a LORA fine tuned model

Hi, I have fine-tuned falcon 7b using QLORA. Now I need to fine-tune it again with instruction dataset. But I am not sure how to do it.

I am expecting support from the community…

Thanks