zsrobust4foundationmodel's People
zsrobust4foundationmodel's Issues
The code is different from your paper
Sorry, I think the y the equation 4 and 5๏ผ the y =1 or y=0
but in your code, the y is the label.
I think there is an indentation error here.
Pretrained weights and evaluation script
Hi! Thanks for sharing the code for your work.
I wanted to know if the pretrained weights of the fine tuning experiments will be released anytime soon?
Further, the from the documentation, it appears the finetuning script is only available for VPT, not for the best results achieved by finetuning of visual encoder.
out of memory for batch size of 1
Running the fine tuning script as mentioned in the readme file (with batch size of 1)
python finetuning.py --batch_size 1 --dataset ImageNet --name feimogu --train_eps 1 --train_numsteps 2 --train_stepsize 1
RuntimeError: CUDA out of memory. Tried to allocate 302.00 MiB (GPU 0; 15.75 GiB total capacity; 11.37 GiB already allocated; 216.19 MiB free; 11.78 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF
/local/*/datasets/ImageNet-clean
Hi! Thanks for sharing the code for your work.
Where can I get the file for this path?
/local/*/datasets/ImageNet-clean
Hello! I have questions on your paper
I read your paper with interest! Currently, I have a plan to develop your research.
I have some question in your paper. Please answering below three questions
(1) In paper, I confirm that you evaluated 100 PGD attack on test data. Then, setting test_numsteps=100
when evaluating is same with this?
(2) In finetuning.py
, I found attack_CW
(not ```attach_pgd``). Can you explain what this attack means? and What is paper result correspoding this attack?
(3) Do you have full result l2-norm bound pgd attack? In your paper, l_infinity-norm bound is used.
I am curious about l2-norm result!
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. ๐๐๐
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google โค๏ธ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.