How to switch to P-Tuning v2

[jiahuanluo]

We can find the P-Tuning v2 in

peft/README.md

Line 29 in 8af8dbd

2. Prefix Tuning: [Prefix-Tuning: Optimizing Continuous Prompts for Generation](https://aclanthology.org/2021.acl-long.353/), [P-Tuning v2: Prompt Tuning Can Be Comparable to Fine-tuning Universally Across Scales and Tasks](https://arxiv.org/pdf/2110.07602.pdf)

But how can I switch to P-Tuning v2?

[pacman100]

Hello, those are implemented together. P-Tuning v2 introduced optional parameterization of prompt tokens which you can specify via prefix_projection of PrefixTuningConfig. The other contribution was the ability of work without verbalizers using the linear classification head for NLU tasks whereas Prefix-Tuning paper which focused on NLG didn't focus on this.

So, they are supported via the same PrefixEncoder PEFT method

[github-actions]

This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.