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

.py files #1

Open
j3g opened this issue Oct 26, 2023 · 16 comments
Open

.py files #1

j3g opened this issue Oct 26, 2023 · 16 comments

Comments

@j3g
Copy link

j3g commented Oct 26, 2023

I love your work and the ideas associated with it
just curious why your files isn't a .py

@tim-hub
Copy link

tim-hub commented Jan 1, 2024

I forked this project and tidy them up, https://github.com/Antioch-Tech/NightshadeDetect

If you are still interested @j3g

However, please not there are still other issues after my tidy up anyway, #2 (comment)

@robertJene
Copy link
Collaborator

yeah your own readme.md states:

python nightshade_antidote.py input.jpg

so... the repo needs to have this file:

nightshade_antidote.py

@RichardAragon
Copy link
Owner

RichardAragon commented Apr 22, 2024 via email

@robertJene
Copy link
Collaborator

in what ways is it better? besides the filename?

@RichardAragon
Copy link
Owner

What interests you about this particular repository, Robert? It is one of my least favorite ones. My brother's name is also Robert. Another one of my least favorite ones.

@robertJene
Copy link
Collaborator

to prune training data.

@RichardAragon
Copy link
Owner

RichardAragon commented Apr 22, 2024 via email

@robertJene
Copy link
Collaborator

no, model training requires data.
if someone poisons their data, then it is fruitful to check so I don't use their image.
this is 100% simple.

@robertJene
Copy link
Collaborator

robertJene commented Apr 22, 2024

I forked this project and tidy them up, https://github.com/Antioch-Tech/NightshadeDetect

If you are still interested @j3g

However, please not there are still other issues after my tidy up anyway, #2 (comment)

could you update your Repo so I can post issues there?

I just updated my downloaded copy to accept the path to an image file as an argument

Also, I've tested the script on an image I poisoned with Nightshade, and it's not outputting any data

@RichardAragon
Copy link
Owner

RichardAragon commented Apr 22, 2024 via email

@robertJene
Copy link
Collaborator

well the plan is to "taste" the images for poison and prune the dataset ahead of time, to respect the wishes of whoever poisoned the images.

@robertJene
Copy link
Collaborator

My end-game is to make a video showing folks how to poison their own artwork

and a 2nd video to show people how to check if something is poisoned so you can prune those files before training

If @tim-hub is interested I can help him with his fork, I've already made updates to the script, but can't get output

@RichardAragon
Copy link
Owner

RichardAragon commented Apr 23, 2024 via email

@robertJene
Copy link
Collaborator

I am not familiar with all this python stuff.
I can install something and get it working if there's clear instructions.
Also, I do some programming as a hobby.
I used chat gpt's help last night to get the script to take a command line argument.

I'm planning on making a wrapper with batch and vbscript so I can batch "taste" a folder of images.

I don't know if it would make more sense to have the python script do all that.

But I spent about 4 hours on it last night and didn't have any luck getting it to output the results.

@RichardAragon
Copy link
Owner

RichardAragon commented Apr 23, 2024 via email

@robertJene
Copy link
Collaborator

why don't you put extensions on the script files?
They can't run for people without them.

also

File "D:\AI_Stuff\StableDiffusion\Training\NightshadeAntidote\React_Native.py", line 1
import React, { useState } from 'react';
^
SyntaxError: invalid syntax

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

4 participants