I wanted to check if it is possible to do PEFT while fine tuning the open ai models. Basically wanted to do something like below to freeze the existing params and then train the addiitonal layers
for param in .parameters():
param.requires_grad = False