Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

How to use the checkpoint #41

Open
JaYzZ opened this issue Nov 5, 2024 · 1 comment
Open

How to use the checkpoint #41

JaYzZ opened this issue Nov 5, 2024 · 1 comment

Comments

@JaYzZ
Copy link

JaYzZ commented Nov 5, 2024

I find it interesting that the checkpoint in your cvpr24 results do not have in_proj_weight, only in_proj_bias in nn.MultiHeadAttention parameters, which block me from using and evaluating them.

In addition, what is the corresponding config yaml file for them? The readme seems unmatched with the code and ckpt.

@Kimsure
Copy link

Kimsure commented Nov 14, 2024

you need to check model_vpt.py since the attn.in_proj_weight has been chunked as q_proj_weight, k_proj_weight, v_proj_weight to only finetune query and value, where in_proj_bias was not modified. This might help you.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants