- Joined
- Jan 2, 2019
4chan is pretty easy to recreate in AI cause every thread is the same garbage with the exact same replies, most of the posts are probably already AI, let's be honest.
Stable Diffusion, NovelAI, Machine Learning Art - AI art generation discussion and image dump
- Thread starter Near
- Start date
-
🐕 I am attempting to get the site runnning as fast as possible. If you are experiencing slow page load times, please report it.
- Joined
- May 16, 2024
The way you can tell these 4chan screencaps are fake is because the average reply length is too long, and ironically, the replies are too intelligent.
- Joined
- May 9, 2017
- Joined
- Oct 7, 2020
- Joined
- Jan 31, 2021
- Joined
- Dec 19, 2022
- Joined
- Aug 14, 2022
Yea I've found out about the script to. My challenge is how to use two and more gpus rather than the one to train models.
- Joined
- Jan 31, 2021
Could someone please help me with a prompt?
I'm trying to recreate The Lady and the Unicorn tapestry's style with new central images, but unfortunately whatever I try is always over-complicated with figures (as AI does generally tend to be) and with the classic swirly faces.
I'm trying to recreate The Lady and the Unicorn tapestry's style with new central images, but unfortunately whatever I try is always over-complicated with figures (as AI does generally tend to be) and with the classic swirly faces.
- Joined
- Mar 29, 2019
"A hamster is never late, Mr. Moon, nor is he early; he arrives precisely when the pizza does."
Last edited:
- Joined
- Apr 22, 2015
- Joined
- Aug 28, 2019
- Joined
- Dec 28, 2014
That seems to be a pretty good development for SD and AI in general.
Screamer_2
kiwifarms.net
- Joined
- Nov 6, 2022
I don't know where it fits in with SD, but part of the development for the later Avatar films hope to incorporate a lot of AI in the CGI work flows. I assume only in an enhancing quality of the CGI and motion capture / animation.
- Joined
- Dec 18, 2022
Isn't SD deprecated in the commercial space with the release of Flux? It's so much better. I don't know why he didn't join up with Black Forest.
- Joined
- Aug 28, 2019
I don't know about commercial, but it certainly seems like the wind is out of SD's sails when it comes to hobbyists.
- Joined
- Jan 31, 2021
- Joined
- Nov 7, 2014
The haunted looking one with the inadvertent neovagina sent me cry laughing into the carpet. Thank you for this.
- Joined
- Aug 25, 2024
It has been a long time since I watched Transcendent Man, but the impression I got from it was Kurzweil is so desperate to live forever that there's no line he's unwilling to cross to get there. If he could make himself a head in a jar like Futurama, he absolutely would. He seemed to think that were he born 10 years later, he might've been able to see a world of vastly extended lifespans, so he was desperately micromanaging his diet and exercise to get himself over the line.
I was a teenager when I watched it and looked forward to this idea of people living 100s of years, thinking it was very cool. One consequence of getting older that nobody warned me about was the loss of hopeful naïveté since these ideas now inspire dread and fear within me.
Anyway, on a lighter note: NovelAI has finally released a new text model! For a service that has "novel" in the name, it would seem the writing aspect has taken a backseat to generating anime images and their new chat bot service. The release of "Erato" comes a little over a year after the release of "Kayra" and was apparently delayed by the incredibly long lead times to get their new Nvidia H100 cluster.
Here's the weird muscle-booba mascot they created for Erato. It always bothers me how deeply entrenched anime shit is in things like this, especially with NovelAI even prior to image generation coming around.
I've given it a shot and it's very underwhelming as the difference between Euterpe and Clio was night and day, then from Clio to Kayra it was a noticeable step up which left the legacy models completely worthless except for using old finetunes. With Erato, it's a case of sometimes better, sometimes not, and to get consistently better results, you gotta use a custom scenario one of the devs provides on Discord to correctly configure your story.
It's also hampered by the laughably small 8K context limit. They've based Erato on Llama 3 whose initial release had a gimpy context limit, but 3.1 can handle much more. If they retrained on 3.1, they could potentially stretch to 32K context limits, though I don't see this happening anytime soon as they've likely invested millions of dollars into the training they've done with 3 and are unwilling to start over.
NovelAI enjoys a near monopoly on uncensored storywriting as there's really no good storywriting UIs that take advantage of services like OpenRouter or Mancer. Storywriting in SillyTavern is like shoving a round peg into a square hole, but if somebody sits down and takes the effort to basically clone NovelAI's UI into a generic LLM storywriting application, then they're boned unless they can magically fix whatever the fuck is making Erato so shit.
- Joined
- Mar 16, 2021
Ive been using Nai since the begining, its good for what it is but hot damn it reeks of needing Day 1 patch nonsense because llama shouldn't be this bad! When it works, it works REALLY well but make a mistake and you get to see real time retardation if you don't catch the mistake.
And then there is the other service they're crafting called Aether Room, so more coomer anime shit but I gotta give it to them, they are the only ones offering uncensored and privacy respecting AI access. It beats doing the illegal of scraping API keys to run up someone else's AWS bill to write "ME CUM" with Anthropics Claude Opus for a few days.
And then there is the other service they're crafting called Aether Room, so more coomer anime shit but I gotta give it to them, they are the only ones offering uncensored and privacy respecting AI access. It beats doing the illegal of scraping API keys to run up someone else's AWS bill to write "ME CUM" with Anthropics Claude Opus for a few days.