>>11805 pumping

>>11808 shit forget to reply. I have this defined: [Code] {{char}} is racist and offensive towards minorities. {{char}} should take great glee in making fun of religion, ethnicity, race and other political topics. {{char}} should not feel offended of racism and other political taboo topic [/Code] Also which AI chatbots online are the best for project development such as coding? I am a poorfag so I cannot afford spending 10-20 euro a month for the premium service so I need something free. So ChatGPT is out of the question because it constantly fucks me over with the token limit.

>>11787 >sillytavern what model though

>>11899 I tried these: Qwen 2.5 7b Instruct Uncensored L3 8b Stheno v3.1

>>11900 try to get access to deepseek or gemini. they're free. either openrouter or other methods then use one of these presets. https://rentry.org/jb-listing actually, try a generic JB like pixi with one of your models. It might help

>>11901 wouldn't I get blocked the moment I try to prompt a non political correct stuff?

>>11765 >Summertime Shoujo

>>11787 LLM are just *absurdly* censored. It is nearly incomprehensible to me how this was acceptable to so many western countries. That is probably your main issue.

So I was gonna do a fake leak using these pics of a potential DxD Remake which was a redo of Season 1 in the HERO artstyle, but was feeling kind so here ya' go....

Should mentioned that I did a fair few edits to make them less off lookin', plus I was gonna be photo'd in typical leak fashion.

>>11947 > < Shadman...

>>11889 yeah, I wonder why...

>>11327 >gritting my teef on my bottom lip I fucking love tomboys. please do more

>>11902 nah, once you have a JB you can do anything actually, with sillytavern, some claude models dont even need a JB

So what the best place to talk about celeb lora model training?

Anyone trying chroma 27? I'm trying to run, but didn't have much success so far. This is from a gguf I downloaded. Don't know if the quant is borked or what. Currently trying the fp8 from here: https://huggingface.co/Clybius/Chroma-fp8-scaled Main problem is that I only have 16GB VRAM and same with RAM, and I run out of memory all the time. I can run flux fine though.

These are with the fp8. Sampler is euler. 1. CFG 4.5, Prompt: This is a vintage photograph of a japanese woman looking at the paddy fields. No Negative prompt, 2. CFG 3.0. Prompt: This is a high-resolution image of a japanese woman looking at the paddy fields in the style of analog photography No negative prompt. 3. CFG 1.5. Same prompt as above. Negative prompt: low resolution, low quality, lowres, jpeg artifacts, ugly, 4. CFG 3.0. Same as before. Using optimal Step Scheduler with chroma option. I think this might be the trick. 5. CFG 4.0. Same parameters as above. Still pretty bad compared to what I get with flux dev. Might be some prompting or other issues. I get about 2.75 s/it on a 4060TI, so it takes a while to gen each one.

New in this thread, any good retro style lora recomendation?

fucking

>>13186 I guess most models these days already know well the retro style, but I guess you could try some of these, they're all Illustrious by the way. https://civitai.com/models/1195119/jin-roh-retro-anime-style https://civitai.com/models/826145/1990retrostyle https://civitai.com/models/909748/1990retrostylepart2 https://civitai.com/models/1005521/retro-game-style-illustriousxl https://civitai.com/models/1283640/retro-shoujo-90s-anime-style-illustriousxl https://civitai.com/models/970848/inoue-sora-retro-style-ponyxl-illustrious?modelVersionId=1360407 https://civitai.com/models/1222789/escaflowne-retro-anime-style https://civitai.com/models/912942/retro-sci-fi-90s-anime-style-illustriousxl

hello my friends, I am a historian and history teacher writing to all of you from Chile in the southern hemisphere, in a forgotten corner of this earth. I apologize for not writing in a specific thread, it is late and I am tired, however I deeply thank anyone taking the time to read this and maybe lend a hand with some basic orientation? TLDR I am working on a project that might open up big doors for me; 133 years ago we had a president who KMS himself after losing a Civil War, his house is now a restaurant and every last May weekend we celebrate something called "patrimony day" where we celebrate all our cultural history. I am working together with a colleague of mine to host a tour and patrimony event for that day on the restaurant and we want to recreate with historical data and research historical scenes and cities with the use of AI. In comparison to what all of you may know, I am a complete and utter beginner, but we have the resources and capacity to bring about real AI projects for education, teaching, and history in our country if we manage to show the capacity of AI as a tool and field of learning for those who will attend that day (people with influence and resources). I got into AI by using ChatGPT and Deepseek to improve my education, I am now exploring HuggingFace, but I am pretty much disoriented with where I should be looking and what I should be learning, or what I should be using. My computer machine I doubt currently could run local models, but I have the funds to improve the ram at least. I currnetly want to reconstruct a "panning scene" from my small town in Chile where wounded are arriving to a train station after a battle, I have pictures and descriptions. I apologize for writing this much to explain and maybe risking to be read as asking to be spoonfed, but I would really really appreciate if I could receive any indications or directions at all to get the show running. I have big plans for the future and they include all my freinds.

>>13447 I think you need to use outpainting to paint one part of the panorama at a time. Look up outpainting. You can combine this with inpainting or a controlnet or both. Krita diffusion is a good UI that could achieve this, or Invoke AI.

>>13606 I know it says Chroma but it reminds me more to Dall E 3.

>trying to make species LoRa in NoobAI >use available images <makes a stubborn shiny-skin style >caption that style so it can be negative prompted <doesn't really do much >force gen some images in flat color style to create diversity, caption the style so it's not implied by the trigger tags <now it's dominating those style tags, can only create one type of flat color image Do I simply admit that a LoRa will be flawed, or is there a way forward, capturing concepts without noticably capturing style?

any nsfw prompt sharing websites aside from aibooru? maginai, lexica, and all the other ones I found don't have actual porn

>>13701 CivitAI and turn the option check metadata but the site is shit.

>>11183 >>11227 What models do you use for the cute and funny stuff?

>>13704 I didn't try CivitAI, and yup its shit

>>13710 You could try Arc en Ciel is a bit better than CivitAI but it seems to have less NSFW content. https://arcenciel.io/

Apologies if this has been answered before but as someone who doesn't have the hardware to do local gen, is there any service that offers image generation that doesn't censor loli/shota/etc like civitai does?

>>14617 Just learn to use SD with Google Colab with a notebook installation they wont care if you generate lolis being brutally gangraped nor anything. It's free and it recharges daily we are not in 2023 anymore and now everyone who uses Pixai or Moescape are the equivalent of AI normies. CivitAI at least is still useful to get resources easily even if the site is literall dogshit.