mightyweasel

238 Followers
462 Following
863 Posts

Creates furry and horror #art with AI. (he/him)

DM/toot me if you wanna learn, I'm friendly and have trained up multiple beginners.

Posting all new art here, sometimes old art too. Feel free to critique - any comments don't have to be all sunshine and rainbows. Here to learn too!

Tutorials: https://peakd.com/@mightyweasel
Old Art: blue bird @mightyweasel1
Older Art: http://artbreeder.com/mightyweasel

#furryart #aiart #furry #furryartist

TutorialsGlad to help, ask for it
AI ArtSo Very Yes
FurryYup
Coke PleaseIs Pepsi okay

@helloyanis Ok, so I've made a lot of progress & have a ton of feedback. Happy Easter!

a) Thinking about how to pass you that feedback - it'd probably be best if you fired off an email to [email protected] and I can just send you a .doc or .pdf or whatever

b) What version of Python did you use in step 1) of the tutorial? I used 3.13.3, and I think ONNX has an incompatibility with the latest?

I'm pretty much hitting this issue:
https://github.com/onnx/onnx/issues/4269

and I suspect (although that issue is very old) that there's a CURRENT incompatibility with Python 3.13.3 and whatever the current version of ONNX is

(Alternately, I just need to install a compiler to get the ONNX wheel built...but I seem to have that too, as I installed VSCode on this fresh system.)

Windows 11: Can't install onnx with pip · Issue #4269 · onnx/onnx

Bug Report Is the issue related to model conversion? Nope Describe the bug pip install onnx failes with CMake Error at CMakeLists.txt:17 (project): Generator NMake Makefiles does not support platfo...

GitHub

@helloyanis

*rises from ancient slumber, joints cracking like dry twigs*

who...WHO DARES to BREAK THE SEAL ON MY ROYAL TOMB!! NOW THAT I AWAKE, THOSE THAT PUT ME INTO THIS CURSED SLUMBER SHALL SURELY PAY WITH THEIR *BLOOD* AND...oh I see, okay then; helloyanis, of course; happy to help!

Your timing is good, just built a new Intel machine, haven't had any chance to mess with it, and need an excuse to test anyway. Give me a bit of time to try your tutorial; this actually looks really interesting - you're making generating art easy and approachable for others with baseline hardware, and I totally respect that.

Also...in the reply, I'm going to point out flaws in technical stuff and in English **if you will be okay with that**?

My interns will attest to me really poring over their work to make their code better, so I'll do the same here if you'd like? If so, should have feedback by the weekend.

Overgrown junk
#AIArt #Midjourney
Energy Mesh And Flow by Eris And AI

Energy Mesh And Flow Digital Art by Eris And AI

Fine Art America
@drmaddkap follow Dr. Maddkap for more Real Dad Tips (tm)
If every bathroom in your house is in use, and you need to go, just turn off the internet.

"how does ML/AI cause less power usage?"

e.g upscaling uses less power for the same reason you can achieve higher FPS; you need less power for upscaled lower resolution images than native higher resolution images.

Other examples:

- GraphCast is a weather prediction toolkit that does the work of a whole datacenter inside one PC, with more accuracy https://deepmind.google/discover/blog/graphcast-ai-model-for-faster-and-more-accurate-global-weather-forecasting/
- Likewise, AlphaFold means you can fold proteins with a single PC, or even search, instead of a DC: https://alphafold.ebi.ac.uk/

GraphCast: AI model for faster and more accurate global weather forecasting

Our state-of-the-art model delivers 10-day weather predictions at unprecedented accuracy in under one minute

Google DeepMind
and again, I'm not saying everything venture capitalists or big tech comes up with is a good idea, I'm not saying we shouldn't be careful what to spend our resources on-- but that's also exactly why I think people should be more careful than just pointing at something and going 'This Bad!', and instead actually look at what are sensible solutions in the bigger picture.
@BlindQuilter Oh, hell yes; I am so glad that worked for you! Happy St. Patrick's Day, Tanya!

@BlindQuilter

2) If you DO have other Favorites, then maybe you removed the All Inboxes from it. There is a small + control to the right of Favorites. Click on that control. A new window will appear, go to it. That window has a control marked Mailbox To Add. Click on it. It will bring up a dropdown with items you can add. The top item is All Inboxes. Choose that. Beneath the Mailbox to Add controls are Cancel and OK buttons. Click OK. This should readd it to your favorites.