How do i increases the context of already fined tuned or base model of llama2 ?

[Tejaswi-kashyap-006]

I want to increase the context of the llama2 model, I have finetuned a model(70b and 7b) as well on my data now I want to increase their input context using yarn I was able to understand that we need data of 16k context if we want to increase the context can anyone clarify the procedures that follows after that