Hi @samuela, @PythonNut,
Great stuff and congratulations for the ICLR :)
I've a quick question and am wondering whether it's feasible with the current repo:
- Suppose I've a model (Llama-1) with weights
original_model.pt
- Assume, I fine-tuned/modified the model for some use-case and let the weights be
modified_model.pt
My high-level question is to understand the difference between these 2 functions (i.e difference between original_model.pt and modified_model.pt!) in a quantitative measure. I am assuming your paper deals with similar stuff (the computed Barrier is basically a quantitative measure which tells the difference between these 2 functions).
Is my understanding correct? If so, can you give some instructions on how your repo can be extended to Huggingface models (Llama, Vicuna, GPT, T5 etc;) I am assuming the logic stays the same!
If not, please provide some insights on how this use-case can be done! I'm assuming Wasserstein, MMD might be the next best bet but would like to try your repository!
P.S: Happy to close either of the issue depending on Hugginface support!
Any help is super appreciated :)
Thanks
Srinath
Hi @samuela, @PythonNut,
Great stuff and congratulations for the ICLR :)
I've a quick question and am wondering whether it's feasible with the current repo:
original_model.ptmodified_model.ptMy high-level question is to understand the difference between these 2 functions (i.e difference between
original_model.ptandmodified_model.pt!) in a quantitative measure. I am assuming your paper deals with similar stuff (the computed Barrier is basically a quantitative measure which tells the difference between these 2 functions).Is my understanding correct? If so, can you give some instructions on how your repo can be extended to Huggingface models (Llama, Vicuna, GPT, T5 etc;) I am assuming the logic stays the same!
If not, please provide some insights on how this use-case can be done! I'm assuming Wasserstein, MMD might be the next best bet but would like to try your repository!
P.S: Happy to close either of the issue depending on Hugginface support!
Any help is super appreciated :)
Thanks
Srinath