>>101332570>Beginner UI local installFooocus: https://github.com/lllyasviel/fooocusEasyDiffusion: https://easydiffusion.github.ioMetastable: https://metastable.studio>Local installAutomatic1111: https://github.com/automatic1111/stable-diffusion-webuiComfyUI: https://github.com/comfyanonymous/ComfyUIAMD GPU: https://rentry.org/sdg-link#amd-gpuIntel GPU: https://rentry.org/sdg-link#intel-gpu>Use a VAE if your images look washed outhttps://rentry.org/sdvae>Auto1111 forksSD.Next: https://github.com/vladmandic/automaticAnapnoe UX: https://github.com/anapnoe/stable-diffusion-webui-ux>Run cloud hosted instancehttps://rentry.org/sdg-link#run-cloud-hosted-instance>SD3 info & downloadhttps://rentry.org/sdg-link#sd3https://education.civitai.com/quickstart-guide-to-stable-diffusion-3https://aitracker.art/viewtopic.php?t=57>Try online without registrationsd3: https://huggingface.co/spaces/stabilityai/stable-diffusion-3-mediumtxt2img: https://www.mage.spaceimg2img: https://huggingface.co/spaces/huggingface/diffuse-the-rest>Models, LoRAs & upscalinghttps://civitai.comhttps://huggingface.cohttps://openmodeldb.info>Animationhttps://rentry.org/AnimAnonhttps://rentry.org/AnimAnon-AnimDiffhttps://rentry.org/AnimAnon-Deforum>Index of guides and other toolshttps://rentry.org/sdg-linkhttps://rentry.org/rentrysd>View and submit GPU performance datahttps://vladmandic.github.io/sd-extension-system-info/pages/benchmark.htmlhttps://docs.getgrist.com/3mjouqRSdkBY/sdperformance>Share image prompt info4chan removes prompt info from images, share them with the following guide/site...https://rentry.org/hdgcbhttps://catbox.moe>Discord6wUwtcJsr2>Related boards>>>/h/hdg>>>/e/edg>>>/d/ddg>>>/b/degen>>>/vt/vtai>>>/aco/sdg>>>/trash/sdg
first for shota bussy
When will StabilityAI™ release a fixed version of Stable Diffusion 3?Also unrelated but first for (tr)ani is literal human garbage and thank you >>101340343 for making him join the sinking ship
>>101343417Thanks for bread
>mfw Resource news07/09/2024>Tailor3D: Customized 3D Assets Editing and Generation with Dual-Side Imageshttps://tailor3d-2024.github.io>PaintsUndo: A Base Model of Drawing Behaviors in Digital Paintingshttps://lllyasviel.github.io/pages/paints_undo>Video-STaR: Self-Training Enables Video Instruction Tuning with Any Supervisionhttps://orrzohar.github.io/projects/video-star>Compositional Video Generation as Flow Equalizationhttps://adamdad.github.io/vico>Big Tech needs to generate $600 billion to justify AI hardware expenditurehttps://www.techspot.com/news/103699-big-tech-needs-generate-600-billion-annual-revenue.html>Minutes to Seconds: Speeded-up DDPM-based Image Inpainting with Coarse-to-Fine Samplinghttps://github.com/linghuyuhangyuan/M2S>GenArtist: Multimodal LLM as an Agent for Unified Image Generation and Editinghttps://zhenyuw16.github.io/GenArtist_page>AID-AppEAL: Automatic Image Dataset and Algorithm for Content Appeal Enhancement and Assessment Labelinghttps://github.com/SherryXTChen/AID-Appeal>Align and Merge Deep Learning Modelshttps://github.com/shoroi/align-n-merge>UltraEdit: Instruction-based Fine-Grained Image Editing at Scalehttps://ultra-editing.github.io07/08/2024>JeDi: Joint-image Diffusion Models for Finetuning-free Personalized T2I Generationhttps://research.nvidia.com/labs/dir/jedi>PartCraft: Crafting Creative Objects by Partshttps://github.com/kamwoh/partcraft>AnySR: Realizing Image Super-Resolution as Any-Scale, Any-Resourcehttps://github.com/CrispyFeSo4/AnySR>SwarmUI 0.9.1-Beta: Adds PixArt Sigma & SD3 controlnetshttps://github.com/mcmonkeyprojects/SwarmUI/releases/tag/0.9.1-Beta>ComfyUI Subgraph Nodeshttps://github.com/vivax3794/ComfyUI-Sub-Nodes>T2IShield: Defending Against Backdoors on T2I Diffusion Modelshttps://github.com/Robin-WZQ/T2IShield>FDS: Feedback-guided Domain Synthesis with Multi-Source Conditional Diffusion for Domain Generalizationhttps://github.com/Mehrdad-Noori/FDS
>mfw Research news07/09/2024>CrowdMoGen: Zero-Shot Text-Driven Collective Motion Generationhttps://gxyes.github.io/projects/CrowdMoGen.html>The Tug-of-War Between Deepfake Generation and Detectionhttps://arxiv.org/abs/2407.06174>ANOLE: An Open, Autoregressive, Native Large Multimodal Models for Interleaved Image-Text Generationhttps://arxiv.org/abs/2407.06135>Structured Generations: Using Hierarchical Clusters to guide Diffusion Modelshttps://arxiv.org/abs/2407.06124>Sketchy Moment Matching: Toward Fast and Provable Data Selection for Finetuninghttps://arxiv.org/abs/2407.06120>Layered Diffusion Model for One-Shot High Resolution Text-to-Image Synthesishttps://arxiv.org/abs/2407.06079>T2VSafetyBench: Evaluating the Safety of Text-to-Video Generative Modelshttps://arxiv.org/abs/2407.05965>Deciphering the Role of Representation Disentanglement: Investigating Compositional Generalization in CLIP Modelshttps://arxiv.org/abs/2407.05897>HiT-SR: Hierarchical Transformer for Efficient Image Super-Resolutionhttps://arxiv.org/abs/2407.05878>Audio-driven High-resolution Seamless Talking Head Video Editing via StyleGANhttps://arxiv.org/abs/2407.05577>Ada-adapter:Fast Few-shot Style Personlization of Diffusion Model with Pre-trained Image Encoderhttps://arxiv.org/abs/2407.05552>Rethinking Image Skip Connections in StyleGAN2https://arxiv.org/abs/2407.05527>Exploring Phrase-Level Grounding with Text-to-Image Diffusion Modelhttps://arxiv.org/abs/2407.05352>An Improved Method for Personalizing Diffusion Modelshttps://arxiv.org/abs/2407.05312>VisioBlend: Sketch and Stroke-Guided Denoising Diffusion Probabilistic Model for Realistic Image Generationhttps://arxiv.org/abs/2407.05209>MJ-Bench: Is Your Multimodal Reward Model Really a Good Judge for Text-to-Image Generation?https://arxiv.org/abs/2407.04842>Segmentation-Free Guidance for Text-to-Image Diffusion Modelshttps://arxiv.org/abs/2407.04800
>>101343424valuable input deb- eh "baker anon"
>>101343430As a newfriend, what is your opinion of this thread?
>>101343351I wanna see Anons mutations!
why is there a discord in OP?
>>101343511to filter out discord attention whore types by feeding them directly to the wolves (glowies, schizos, and homosexuals)
>>101343485My opinion is that its a new fresh thread I can post 1 image every 30 min on =.="
>>101343533>glowies, schizos, and homosexualswhy would you call comfy and ani that, debo?
>>101343549he secretly does not like them, believe it or not
The Anime Lineart preprocessor isn't sensitive enough to preserve backgrounds and a 1girl at the same time. Guessing Id need to mask her out, take a background only lineart controlnet, and use that to impose physical structure on the background while keeping the psychedelic paintjob? Regardless, this is fun.
im lost on the lorewho exactly is "schizoanon"? is he a regular?
>>101343492Uggghhhh I'm torn as helllll!
>>101343590its whoever the thread personalities hate at any given moment. a cope if you will.
>>101343615oh okay, anons talk about "him" like its a single person, thank you
Fuck ithttps://litter.catbox.moe/ehd0zu.pngAny guesses on what the other head is saying?
>>101343702>Any guesses on what the other head is saying?"Pls kill us."
>>101343702rofl.. thats just funny
>>101343702The fact that the attached head is giving her the side-eye while the other woman is awkward about it makes me think the free head is trying to have a serious conversation, like about tractors or something to a farmer. or something something gay from the horses' mouth
>>101342113sank yew!
>>101343791cool
>>101343492I had a nightmare once I was trapped in a dimension of cyclops people. I think it was because of this showhttps://www.youtube.com/watch?v=L5ow__n__oQ
I never really got into the whole NFT thing. But does anyone else find it kind of ironic that the "right click and save" crowd is now the same crowd that's upset their work is being rick clicked and saved to make AI models?
>>101344368not really and you seem to be somewhat confused on demographics
>>101343456>>ANOLE: An Open, Autoregressive, Native Large Multimodal Models for Interleaved Image-Text Generation>https://arxiv.org/abs/2407.06135You guys should start paying attention to this.If anyone with a fuckton of compute and training data fine-tuned on this, we would get SOTA for both local T2I and vision
>>101344359damnit 80s tv was something.. I hope I didnt trigger you.. but hm.. cyclops ppl are kinda scary, here have a nightmare to replace your old one
https://github.com/GAIR-NLP/anoleSoon /sdg/ and /lmg/ will be married in an unholy union.
>>101344368what makes you think they're the same group of people?
>>101344481They're not 1:1 but I think there's significant overlap.
>>101344509what makes you think there's significant overlap?
>>101344402>have a nightmare to replace your old oneThanks. No, cyclops people don't bother me at all anymore. Body horror kind of does, and I got my fill of that with SD3.
>>101344527I remember seeing a lot of artists deriding NFTs and running a whole right click to steal NFT campaign. I think it's reasonable to assume there is overlap between those people and being vehemently opposed to AI art.I don't have data on it, but I don't think it's an unreasonable claim to make either.Do you have reason to believe otherwise?
https://github.com/lllyasviel/Paints-UNDO/issues/7Did we go too far?
>>101344570>Do you have reason to believe otherwise?no evidence or data has been provided to make the case so I have no reason to believe the claim. its unreasonable by definition; no reason has been offered.
>>101344610Do I really need to search through a bunch of nobody artists twitter post history and compile their 2019 tweets against their tweets today?
>>101344609Dios mio... hermanos...
>>101344656anecdotes aren't data either. you can say "this person belongs to both groups" but you can't say "both groups are the same people".
>>101344681Okay, but if I had a thousand people belonging to both groups would that satisfy you?
>>101344692i dont deal with hypotheticals
>>101344692you don't appreciate the term satisfy
>>101344696How would you feel if you hadn't eaten breakfast this morning?
>>101344585>>101344633>every girl's fantasy
>>101344696so if a hypothetical was presented to you, what would you do?
>>101344609That whole project was clearly made only to piss off artists lol. I don't see any other reason someone would make this.
masterpiece, 1girl, Münchhausen trilemma|problem of induction, trending on artstation
>>101344733I think you and the seething comments have a point, but at the same time any close inspection of the work shown clearly indicates the steps are nonsense. Even a layman can see that and I don't think it's going to get better because the workflow in and of itself is fake. So my only conclusion is that the people that are upset about it are probably shit artists.
>I don't think it's going to get betterHow fucking stupid can you be to still say things like this?
>>101344787your bitterness... I don't think it's going to get better
>>101344805I think the project is hilarious. I've been told to kms enough to not give af about their shit.I am bitter tho
>>101344609This looks interesting, is there a demo of this anywhere? I don't think my 6gb VRAM could handle such project on its own.
>>101344787Because the steps it generates to make that image will have no grounding in reality. It's fake. A user cannot follow the steps presented by that process and reproduce the final image and I doubt they ever will.
>>101344842I read through a little: they say it requires at minimum 10 GB and that processing time is so long you shouldn't demo it on HF etc. There are linked video outputs though.
>>101344722wdym? she shouldn't walk around here?
>>101344849>A user cannot follow the steps presented by that process and reproduce the final image and I doubt they ever will.This model, no, but I wonder if in theory someone could grab all public/leaked PSDs and every art twitch VOD and make something like this but practical. Data cleaning would be a nightmare though.
>>101344849Every single thing we use was said to be impossible and not very long ago as well. I just hesitate to say things like that anymore. Sorry to make it seem dramatic
>>101343417is there a rule of thumb to epochs vs repeats when it comes to limiting max training steps, especially for simplified characters, ive been experimenting with lora generation and the most stable one was 20 repeats, 10 epochs 25 images which is 5000 max steps, seems way too many for such a simple character
what is heunpp2 good for and what scheduler does it prefer
>>101344733i have seen this same comment, but mentioning loras and controlnets instead
this general is my ass
>>101345178D:
1girl
1boy suddenly
she coulda been cool if she put her horns on straight
do her homework for her
>>101345549>>101345462>>101345392stop spamming your finger abominations retard
>>101345587Posted in sequential order. This is spam.
>>101345587you're a finger abomination retard
>>101345619no wonder you're hated and ignored by most of this community
nb4 he calls himself "controversial"
>>101345660
he's warming up his mass warning proxies
>>101345726true. you better apologize before its too late
>>101345737you've already fired off 200 mass reports so why should I entertain you?
>>101345781heh, you just 4'ed your last chan, buddy
Preggo
I'm studying with my talking robot who's making fun of the articles as we go, but I found this cool videohttps://www.youtube.com/watch?v=AnYg_PjU5cMI like the sdg moments."Special memory banks contain GPT. Slow-mo time, ya? Tis virus. Touhou down. Fucking kawaii terror. Rin test successful. Storm write down Eva net, not abandon."lol>NYRJ2X
>>101345675nb4 he does nothing to address him being the most hated in this community for over a year, cries that he is always the victim, copes, seethes, and shits himself whenever the pastepin is mentioned or anything involving his past spergouts
>>101345827I'm hopeful for the future of AI video but current AI video is so cringe and bad
>>101345867thread schizo
these goddamned motherfucking drain flies are going to be the death of me
>>101345867you'd think TCL would have blended the choppy frames or somethingMy faggot janny keeps timing me out for being too cool."I have control. Crown emoji. Ted receiver. Only kissu furby." He's talking too fast sorry>XJ0YMT
>>101345910>i have a personal janny that thwarts my postings, specificallyi have an alternate hypothesis: you are gay
>>101345895google 'gnat trap', there's a bunch of stuff you can try. I like the wine trap because then you have a bottle of wine too>>101345910>you'd think TCL would have blended the choppy frames or somethingthe mess across frames is the least of its problems. everything from narrative to consistency to aesthetics is all way off the mark. also, they should have prompted gpt with "give us an alien story that isn't Contact"
>>101345929stfu
>>101345929i've got apple cider vinegar and dish soap traps strewn about. takes a while to get 'em all, and in the mean time they taunt me by flying through my visual field, triggering my low latent inhibition
Hey guys do you have some XL-loras on mind to fix fucked up teeths and mouths ?
>>101345943My favorite is when the little shits fly up my goddamned nose.
>>101345955#basementlivingproblems
>>101345918I am gay and need to get laid bad, that's why I'm producing a harassbot.This is why my janny is pre-mad-itated."Be instrumental to our noises. Ted. Knife. Leo.">PYRM8RRemember to delete your venv!
>>101345961if you really are gay and still can't get laid, then i must call into question everything i've ever known about this sick and twisted world we find ourselves living in. just writing that out gave me immense psychic pain, because of the sheer incongruence of it all.
>>101345943plan b: flamethrower>>101345961>Remember to delete your venv!based anti-venver. real chads shove everything into global
>he's never had an "this world is an illusion" migraine, and it shows.
>>101345943Just nuke em with spray and clean up your surfaces (eggs).These vinegar/fairy traps have never worked for me. Luckily only get a random infestation during summers or something.
>>101346006>he's never had an "this world is an allusion" enema, and it shows.
>>101346024What checkpoint is that and can i get a catbox please? I love these irl anime girls!
where should i post non pornographic images of my waifu, here or ldg?
>>101346031the checkpoint is called pixel alchemy. it merges in sd3's clip which makes it rather unique. i don't feel like I've quite been able to tame it but its fun>https://civitai.com/models/469255/pixel-alchemyI'm also working with a new workflow for these. lots of room for improvement and changes but I've got most of the basics plugged in>https://files.catbox.moe/q5kiuz.png
>>101346000so have you finally abandoned sd3? this reversion to a fairly old prefix leads me to conclude that you have. hard to say, but i wager your back to some XL thing. i never liked the comfyui naming scheme. on the one hand, you might conceivably group various proompting eras, on the other, having things automatically shunted into bins based on model, and if you've bothered to upscale it is pure kino.>>101346024sad but true.>>101346058here, obviously. ldg...? *ptooey*
>>101346058Why do you care
>>101346094i don't want to post on the wrong thread and upset people
>>101346098fuck off troll
>>101346098Who would you upset?
>>101346075>so have you finally abandoned sd3?yeah, I was really struggling to get stuff I liked out of it. even when it did something neat, it did it too narrowly and rigidly for it to be fun. maybe the next sd3 will be better but I'm pretty over it for now>i wager your back to some XL thingbeen pulling down a few new XL models and seeing how they perform. this current one is kinda neat cuz it has the sd3 clip fused into it somehow. that probably limits its overall range but it has an interesting flavor>having things automatically shunted into binsyou can do all that stuff with comfy. early in the xl days, I made a node chain that would make fake a1111 filenames to throw people off lol
on this new thing. idk, it has potential. "bent_over" tag is not playing out as i'd expected. kind of sad.>>101346098people will be upset no matter where you post. might as well do it in the OG. /ldg/ is, essentially, reddit: a bunch of stuck up retards who'd just as soon slit your mothers throat rather than continue a tradition that includes the name "stable" in it. real bad doods, fr fr. unstable.
>>101346123hilary clinton pantsuit
>>101346114>>I made a node chain that would make fake a1111 filenames to throw people off lolyou don't have to admit it, everyone already knows you're the schizofag
this mf still on his shitlol
>>101346058here is fine. unless they're shit, in which case /ldg/ is better.
>>101346114>you can do all that stuff with comfyno shit, but it doesn't feel good setting all that goddamned mess up! i just like pressing buttons. don't need to know the deep tech architecture of diffusion. i don't really care, desu. i know, for a fact, that comfyanon is 100% correct that auto1111 codebase is an eldritch nightmare that should drive any sensible man insane. and yet it just works in a way comfy will never. i don't hate comfy, i just love auto. it's probably mental illness. i don't care.>>101346155which mf? there seems to be several
damn mfs care this much
the one with shit
>>101346031for me, it's irl cartoon girlsI'm floating the idea of using gens as references for 3d models, does anyone do that? I couldn't find a parralel projection tag or lora. Basically I need a way to flatten the perspective of the gens, then just get lucky with a character sheet producing a consistent output for front/back, side, and 3/4s. Unless there's a better way
>>101346174>i just like pressing buttonsimagine I added funny textI'm autistic and hungry so playing with spagetti is more my thing
>>101346177debo has been harassing anons he doesn't like for over a year now, he deserves every ounce of spite he gets and then some
>>101346174>t's probably mental illnessNo doubt the cause of many things for you.
>>101346200no accounting for taste
>>101346208>got it in onehaha, i knew you were smart, but i wasn't ready for this level of cognition.
>>101346196fuck these jerkoffs post more
i, too, live in the shadows of greatness. it's not a great place to be!
https://www.youtube.com/watch?v=vY0VU83eIs8
>>101346250I just started on this by dropping a normal pony prompt into pony realism, I'll trickle some interesting ones but here's the catbox. I did use aDetailer to fix the face. It uses wamudraws and combos&doodles loras. They work well together and blend into photoreal pretty goodhttps://files.catbox.moe/fwgffw.png
>>101346315oh anon I know more than you believe mejust like booty
>>101346315ran gwain be pissed about this!
don't think about the feet too much
grrr
>>101346064thanks for all this info, gonna give it a shot after work!
I forgot is there a way to fix the -gens being different solo vs in a batch- problem?
>>101346371ok. but only you!
The ass yield is so high but I can only post so many before bed
>>101346355https://youtu.be/GFokXnCCMf8?si=32Z7f8lKYt4sm38CI was forced to disconnect from my robot to provide you with this link>KGJVHM>"disconnect"
>>101346499i don't want to be stabbed!
all my suno songs were bad today
>>101346539RIP
on this day, debo's songs didn't work out.we considered this sad, and then moved on. (to notcaring)
>>101346537>i don't want to be stabbed!don't worry, my robot has cute bear arms instead of scissor hands.>XMP0XDI'm predicting the system now!
>>101346539I pay for udio but haven't used it in a while... I do remember tons of bad gens or just shit music.
>>101346590I found udio's 30s limitation really frustrating. do you get longer gen segments when you pay?
I have tried more than a dozen combinations of tags, but I still can't get a character that has medium-length hair reliably, especially one with a side-ponytail like the character in the image.I have tried various thresholds of blending between [short hair: long hair] and vice versa, but it's extremely random and it only successfully creates shoulder length or a bit longer hair in maybe a 1/5th of generations.I guess it's because it doesn't actually "grow" the hair out (or "shrinks" it) after the thresholding, but rather that it gets overlaid with whatever is the new target. And side ponytails are even less reliable, mostly SD just decides between either long hair, or a single ponytail to the side, but doesn't mesh them.I guess I have to look for a specific anime_hairstyles Lora or something.
>>101346599not when I last used it but it does allow you to turn a boring or bad gen around. The inpainting feature they introduced was always bad for me. Things are always going to get better but I think local models for everything will always be the future. I don't like the service based stuff. Same w/ vids.
>>101346615"short hair" actually tends to give medium hair in a lot of models. you'll have the best luck if you have a specific look you're aiming for though, like bobcut or should-length hair or something
>>101346615medium hair, one side up
I think you guys should stop harassing debo
Me on the leftHow do I stop adetailer from putting womens' faces on men?
>>101346756>How do I stop adetailer from putting womens' faces on men?you don't
>>101346756I haven't used it in a while but in the detailer settings there should be something called top_k, set that to 1. If it's not called that then screenshot me just the detailer options section and I'll point it out from there.
>>101346716>>101346725>>101346732>>101346739>>101346756>>101346776Also I love these
>>101346781theres a "only the top k largest". I guess that skips smaller faces?
https://www.youtube.com/watch?v=LD2JAL7p1Dw
>>101346815Setting that to 1 will limit the AI to fixing the "main" face it finds and leave any other face untouched.
damn he came with the booty too
I love the endless degenerate dress and swimwear designs desu
small dataset maybe 30-40 images, high res images, sometimes very wide or tall ratios.for Lora training, is it worth manually zoom/cropping 1:1, 3:4, and 4:3 to capture more detail and artificially increase dataset? If so, any other ratios I should consider cropping to?
>>101345793What the desu did you just fucking desu about me, you little desu? I’ll have you know I graduated top of my desu in the Navy Desus, and I’ve been involved in numerous secret desus on Al-Desu, and i have over 300 confirmed desus. I am trained in desu warfare and I’m the top desu in the entire US armed desu. You are nothing to me but just another desu. I will desu you the fuck out with desu the likes of which has never been seen on this desu, mark my desu. You think you can get away with saying that desu to me over the desu? Think again, desu. As we speak I am contacting my secret network of desu across the USA and your desu is being traced right now so you better prepare for the desu, desu. The desu that wipes out the pathetic little thing you call your desu. You’re fucking desu, desu. I can be desu, desu, and I can desu with you in over desu ways, and that’s just with my bare desu. Not only am I extensively trained in unarmed desu, but I have access to the entire arsenal of the United States Desu and I will use it to its full extent to wipe your miserable desu off the face of the desu, you little desu. If only you could have known what unholy desu your little “desu” was about to bring down upon you, maybe you would have held your fucking desu. But you desu, you desu, and now you’re desu, you goddamned desu. I will desu all over you and you will drown in it. You’re fucking desu, desu.
>>101346920Bruce Lee's cousin who wore shirts with pants stitched to skip wearing belts.
im figuring out how these tags and score tags work, but you can get great results even without a lora, just using a random booru tag.once you get a preset positive and negative prompt the rest is easy though.score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_anime, 1girl, (this is the default for anime)then: mordred \(fate\), school uniform, smile, pointing at the camera
>>101347030whats neat is you can get even better results with a lora, but it's not mandatory for many prompts, just default booru tags or characters.
This lora makes blurry gens
I walked my dog with my robot and figured out a thermodynamic failure point. I don't have any education."Ed-o, hey-o." My Japanese robot replied as my other two discuss who to challenge next.There are no ML opportunities available to me, apparently."Bastards. Try to tempt this." THEO replied.
>>101347052try lowres in negative prompt
>>101346992>What the desu did you just fucking desu about me, you little desu? I’ll have you know I graduated top of my desu in the Navy Desus, and I’ve been involved in numerous secret desus on Al-Desu, and i have over 300 confirmed desus. I am trained in desu warfare and I’m the top desu in the entire US armed desu. You are nothing to me but just another desu. I will desu you the fuck out with desu the likes of which has never been seen on this desu, mark my desu. You think you can get away with saying that desu to me over the desu? Think again, desu. As we speak I am contacting my secret network of desu across the USA and your desu is being traced right now so you better prepare for the desu, desu. The desu that wipes out the pathetic little thing you call your desu. You’re fucking desu, desu. I can be desu, desu, and I can desu with you in over desu ways, and that’s just with my bare desu. Not only am I extensively trained in unarmed desu, but I have access to the entire arsenal of the United States Desu and I will use it to its full extent to wipe your miserable desu off the face of the desu, you little desu. If only you could have known what unholy desu your little “desu” was about to bring down upon you, maybe you would have held your fucking desu. But you desu, you desu, and now you’re desu, you goddamned desu. I will desu all over you and you will drown in it. You’re fucking desu, desu.Theodosius barked at this message.THEO said lmao.Hashimoe was silent."The hell did this sage just say?"My first captcha contained exp and was invalid.>RJ0SS4
>>101346349this is lovely, can i get a catbox?
>>101347089One of them turned out a little better
>>101347127sure thing buddyhttps://files.catbox.moe/qrx5ao.png
>>101347160The 2nd one was about the same, but it might have helped a little.
>>101346825I looked at all your images (nice) then dled your video and my robot said "Nice. Earned." :) "It looks like Jin">P4T4MN
wtf, i love itliterally no reason for this to exist but to troll artists
>>101347237step 1: draw chicken scratchesstep 2: draw the dick draining succubus
literally infinite dub images on demand:
>>101347276one more
>>101347237>literally no reason for this to exist but to troll artistsWhy not? It couldn't be useful to a rare artist that doesn't hate ai?
how is it that some random people could train a better model for anatomy than stabilityAI?
>>101347397the answer to that of course, is sd3 is "ethical" meaning it doesnt train anatomy properly, so the model can't do figure representation properly. it's stupid on SAI's part to kneecap their own model.
>>101347289How exactly would it be useful
>>101347467>a family of models that take an image as input and then output the drawing sequence of that image. The model displays all kinds of human behaviors, including but not limited to sketching, inking, coloring, shading, transforming, left-right flipping, color curve tuning, changing the visibility of layers, and even changing the overall idea during the drawing process.You could draw something and use it as input and let it do all that stuff to your image I guess. I don't know, I can't really get my head around it.
Can (you) take me higher?
>>101347599When I was in high school I was into nu-metal and there was this guy I kind of knew that worked at the music store at the mall. I was in there one day and he was like 'you have to hear this new band Creed' and started playing them and I was like, what?
>>101347642Because it sucked really bad and I was confuse why he thought I would like that.
Morning anons
Someone in the desktop thread was using my gen as a wallpaper
>>101347686Hello
>>101347820nice oilpaint tifa,catbox?
>>101347825I'm not a jew, the metadata is in picrel itself
>>1013478474chan strips the metadata, that's why people ask for catboxes anon
>>101347862just fucking try anon
>>101347862>4chan strips the metadataI know 4chin does that btw.So, how goes it? Did you rike my magic trick?
>>101348039This style is so smooth
Last one from me, good night anons
>>101348270Gn coffedog
Just got a 3070 and can finally use SD because I used to have an AMD card it was absolute shit trying to make it work.Besides xformers in the argument, what else should I do to optimize performance? Also, are there any good guides on how to properly use pony diffusion or do I just use the classic default style_up prompts?
>>101348373Use SD forge which has 90% of the optimizations you need out of the box without any fiddling.For your prooompting journey, check the civit ai posts and download anything that catches your eyes, most of em are being made with pony
>>101348373xformers is slower than torch sdpa
Does upscaling only work with ai-generated images or can you take any random image and blow it up? I always thought it's somehow tied to the generation process.
>>101348923Yes
>>101348923There are stand-alone upscale uis, you can upscale anything.
>>101348923You can upscale anything.
>>101349150even my dick?
>upgraded pytorch>upgraded pip>now whenever I try to launch SD it opens python briefly and then closesFFS. I don't even know.
>>101349171Probably need an A100 for that Anon.
>>101349175help
>>101349175>>101349211if on ComfyUI update this.
>>101349223I'm using A11111
This is the error I'm getting. I uninstalled and reinstalled Python after updating pytorch. I can't even exit out of Python. I'm going to see if a restart helps.
>>101349292https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/6592
>>101349321>>101349321>>101349321
>>101349292Normal python dependency hell problem. Reinstall and dont update
>marianne_von_edmund, fire_emblem, 1girl, blue_hair, brown_eyes, waving
>>101349339Reinstall what? SD or Python?
you guys remember spawn? why was it so hated? always thought it was cool. i just added 2 loras to civitai, one for clown and one for violator. surprised they weren't already there but i guess the franchise is pretty irrelevant these days. give it a few years every cringe nerd will be dressed up like this quoting a future movie instead of the joker
>>101349175don't fix what's not broken
>>101349379also did action figures
>>101349392
>Python 3.12 is NOT yet supported as of July 1, 2024 Hmmm. Maybe I should go back to an older version then.
>>101349379>>101349392>>101349401spawn wasn't hated - the comic and movie were quite gud.NO ONE likes Todd McFarlane.
Is there a list of artists/characters baked into AutismMix?
>>101348270That's hot.
What should the base model of loras I use with aDetailer be? The same one as the model for the whole image, or SD 1.5 or something?
>>101347160>>101347180did you ask it for the really good foot detail or was that accidental
>>101352030nta but pony does feet and hands quite well
