Sauerworld Forum

Recent Posts

Pages: 1 2 [3] 4 5 ... 10
21
General Chat / Re: Fine Tuning Stable Diffusion with Sauerbraten screenshots
« Last post by Master on December 08, 2022, 01:02:45 AM »
you're cool, I love it
I love you
keep it up
23
General Chat / Fine Tuning Stable Diffusion with Sauerbraten screenshots
« Last post by Salatiel on November 27, 2022, 01:49:02 AM »
Hello!
Just another update on the "put Sauer into stuff" quest, a few months ago I posted an experiment manipulating some screenshots with an AI model: https://sauerworld.org/forum/index.php?topic=652.0

Now access to Dall-e 2 is public (albeit limited), and freer/open source alternatives like Stable Diffusion have emerged and have since surpassed it.

Right now you can "fine tune" Stable Diffusion for free using DreamBooth, a utility that helps you customize an already trained model, that is, you can add new elements (like your face or captain cannon's) to an existing model without having to train everything again. That way, instead of manipulating an existing image, you can create a completely new one  :o

With about 50 screenshots featuring Captain Cannon these are the results I got so far:







Sorry to make you see a captain cannon face reveal out of the blue like that.

I'll update the post as soon as I have the results of the other playermodels and maps, but if you want to go ahead and try it out:

Fine Tuning Stable Diffusion with Dreambooth in Google Colab


1: Create a HuggingFace account, you will need it to get a token: https://huggingface.co/settings/tokens.
2: Open the Google Colab notebook (a fork): https://gist.github.com/SalatielSauer/957f240032a56dfc19edca4c7e11db18
3: Click on File and save a copy to your Google Drive:


4: Before running any cell, create a folder called "screenshots" next to the existing sample_data, that's where you'll put all your .jpg screenshots named img*.jpg, where * is a number in ascending order.


5: In the "Settings for teaching a new concept" cell, find and change the range of the for loop from 100-150 to the amount of screenshots you have uploaded, if your images start from img0.jpg and go up to img50.jpg, put 0 and 50 as the range.


Right in the next cell you can change the prompt that will be associated with your new element.


6: With all that done you can go back to the first cell and run them one by one, making sure to wait for the previous one to finish before moving on to the next.

The last three cells are where the magic happens, it may take a few minutes to finish executing them, but if everything goes well, you will be able to customize the prompt in the last cell and generate the image.

try playing around with the guidance_scale and num_inference_steps values ​​too!

On this notebook I disabled the nsfw filter ('safety_checker') as it was giving some false positives, keep that in mind :P

The original Google Colab notebook:
https://github.com/PsorTheDoctor/artificial-intelligence/blob/master/modern_approach/text_to_image/dreambooth.ipynb
Search engine for prompts, also generates images: https://lexica.art/

24
Maps / FDM8 by Fatality
« Last post by Fatality on November 22, 2022, 11:19:25 PM »
A small 1v1 map.
25
Maps / Re: Sigura
« Last post by RaZgRiZ on October 26, 2022, 08:06:26 PM »
Looks good overall, I'm digging what you've made. The few downsides I caught from the video is that you have a fair amount of empty space across your flat surfaces, the biggest victim being the greenery. I'd recommend spending some time creating a nice blendmap with a dirt texture, and planting some trees around the map, grasses too. We have them, spruce the place up!

That aside, you may wanna consider some more torches/lamps around the place perhaps, not so for lack of light but rather as points of contrast.  I am curious to see how the map would look in a night setting though, not gonna lie.

26
Maps / Sigura
« Last post by Fatality on October 26, 2022, 07:39:51 PM »
A new CTF map I made with Gangler (layout) and Nieb (texturing/details).



Download here:
https://drive.google.com/file/d/1ZQ-_XRLDQQxisBNKXguPv3boqvrel7N2/view?usp=sharing
27
Maps / Layout20d2/HDM9 (temp name) Test Version By Halbgar, Redon and KI113R
« Last post by KI113R on October 01, 2022, 08:23:37 PM »
Uploading this as a playable version of what is currently known as layout20d by Halbgar and will likely end up being HDM9.

It is basically done and play-ready. Just need to release it to the public to get some playtime on it by others as we've been staring at it for far too long and may have gone numb to some issues. Any input would be greatly appreciated.
















Download Here: https://drive.google.com/file/d/1-pW1oWU_uPo2e6srjyEtQujGTmtS7YR-/view?usp=sharing
28
Maps / Re: Tarsilat (WIP)
« Last post by Getridou on September 02, 2022, 11:35:23 AM »
Updated file now available in drive
https://drive.google.com/file/d/1iIklU1cQD3kZLrarAqX8wevRwJDHSfot/view?usp=sharing

Massive update, rooms changed to open skyline, many more eyesores added
29
Off-Topic / Difference between cube 1 and cube 2
« Last post by 3dcamper on August 27, 2022, 11:26:25 AM »
Is it correct that cube engine 1 is pseudo-3D like idtech1 and doesn't allow floors to be stacked on top of each other?

Is it possible in principle to convert cube 1 levels to doom format?
30
Tech Talk / Re: `writeobj N` command
« Last post by 3dcamper on August 24, 2022, 11:27:37 AM »
O! Thank you very much!
Pages: 1 2 [3] 4 5 ... 10