PEFT techniques using Soft prompts

In Week 2 of Generative AI with Large Language Models. It was discussed that we can use Soft prompts as one of the techniques to fine-tune LLM. Can someone help me with some sample code or any reference for the same?