[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip / qa] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


File: 1720765078147059.jpg (750 KB, 720x1024)
750 KB
750 KB JPG
Previous /sdg/ thread : >>101375191

>Beginner UI local install
Fooocus: https://github.com/lllyasviel/fooocus
EasyDiffusion: https://easydiffusion.github.io
Metastable: https://metastable.studio

>Local install
Automatic1111: https://github.com/automatic1111/stable-diffusion-webui
ComfyUI: https://github.com/comfyanonymous/ComfyUI
AMD GPU: https://rentry.org/sdg-link#amd-gpu
Intel GPU: https://rentry.org/sdg-link#intel-gpu

>Use a VAE if your images look washed out
https://rentry.org/sdvae

>Auto1111 forks
SD.Next: https://github.com/vladmandic/automatic
Anapnoe UX: https://github.com/anapnoe/stable-diffusion-webui-ux

>Run cloud hosted instance
https://rentry.org/sdg-link#run-cloud-hosted-instance

>SD3 info & download
https://rentry.org/sdg-link#sd3
https://education.civitai.com/quickstart-guide-to-stable-diffusion-3
https://aitracker.art/viewtopic.php?t=57

>Try online without registration
sd3: https://huggingface.co/spaces/stabilityai/stable-diffusion-3-medium
txt2img: https://www.mage.space
img2img: https://huggingface.co/spaces/huggingface/diffuse-the-rest

>Models, LoRAs & upscaling
https://civitai.com
https://huggingface.co
https://openmodeldb.info

>Animation
https://rentry.org/AnimAnon
https://rentry.org/AnimAnon-AnimDiff
https://rentry.org/AnimAnon-Deforum

>Index of guides and other tools
https://rentry.org/sdg-link
https://rentry.org/rentrysd

>View and submit GPU performance data
https://vladmandic.github.io/sd-extension-system-info/pages/benchmark.html
https://docs.getgrist.com/3mjouqRSdkBY/sdperformance

>Share image prompt info
4chan removes prompt info from images, share them with the following guide/site...
https://rentry.org/hdgcb
https://catbox.moe

>Discord
6wUwtcJsr2

>Related boards
>>>/h/hdg
>>>/e/edg
>>>/d/ddg
>>>/b/degen
>>>/vt/vtai
>>>/aco/sdg
>>>/trash/sdg
>>
Why are you not in the discord linked in OP, anon?
>>
File: da nooz 11.jpg (533 KB, 1344x768)
533 KB
533 KB JPG
>mfw Resource news

07/12/2024

>Progressive Growing for Stable Diffusion Web UI
https://github.com/rmcc3/experiments-progressive-growing

>Video Diffusion Alignment via Reward Gradient
https://vader-vid.github.io/

>MAVIS: Mathematical Visual Instruction Tuning
https://github.com/ZrrSkywalker/MAVIS

>Pallaidium Adds Kwai/Kolors Support
https://github.com/tin2tin/Pallaidium#change-log

>SEED-Story: Multimodal Long Story Generation with Large Language Model
https://github.com/TencentARC/SEED-Story

>15M Multimodal Facial Image-Text Dataset
https://huggingface.co/datasets/OpenFace-CQUPT/FaceCaption-15M

>OpenDiLoCo, open-source implementation and scaling of DeepMind’s Distributed Low-Communication
https://www.primeintellect.ai/blog/opendiloco

>CLIPTrase: Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation
https://github.com/leaves162/CLIPtrase

>MambaVision: A Hybrid Mamba-Transformer Vision Backbone
https://github.com/NVlabs/MambaVision

>Kohaku Fork of Paints-UNDO
https://github.com/KohakuBlueleaf/Paints-UNDO#what-this-fork-changed

07/11/2024

>Introducing AuraFlow v0.1, an Open Exploration of Large Rectified Flow Models
https://blog.fal.ai/auraflow/

>FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision
https://pytorch.org/blog/flashattention-3/

>LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models
https://github.com/LLaVA-VL/LLaVA-NeXT

>SD Webui Resource Monitor
https://github.com/Haoming02/sd-webui-resource-monitor

>AI's Energy Demands Are Out of Control
https://www.wired.com/story/ai-energy-demands-water-impact-internet-hyper-consumption-era/

>German court rules AI output can be protectable
https://devclass.com/2024/07/10/german-court-rules-ai-output-can-be-protectable-ups-stakes-for-machine-generated-code/?td=rt-3a

>EchoMimic: Lifelike Audio-Driven Portrait Animations through Editable Landmark Conditioning
https://badtobest.github.io/echomimic.html
>>
>mfw Research news

07/12/2024

>HiRes-LLaVA: Restoring Fragmentation Input in High-Resolution Large VLMs
https://arxiv.org/abs/2407.08706

>Live2Diff: Live Stream Translation via Uni-directional Attention in Video Diffusion Models
https://live2diff.github.io

>Generalizable Implicit Motion Modeling for Video Frame Interpolation
https://gseancdat.github.io/projects/GIMMVFI

>Still-Moving: Customized Video Generation without Customized Video Data
https://still-moving.github.io

>Controlling the Fidelity and Diversity of Deep Generative Models via Pseudo Density
https://arxiv.org/abs/2407.08659

>Emergent Visual-Semantic Hierarchies in Image-Text Representations
https://tau-vailab.github.io/hierarcaps

>Fine-Tuning Stable Diffusion XL for Stylistic Icon Generation: A Comparison of Caption Size
https://arxiv.org/abs/2407.08513

>VideoMamba: Spatio-Temporal Selective State Space Model
https://arxiv.org/abs/2407.08476

>Global Spatial-Temporal Information-based Residual ConvLSTM for Video Space-Time Super-Resolution
https://arxiv.org/abs/2407.08466

>Infinite Motion: Extended Motion Generation via Long Text Instructions
https://shuochengzhai.github.io/Infinite-motion.github.io

>A Comprehensive Survey on Human Video Generation: Challenges, Methods, and Insights
https://arxiv.org/abs/2407.08428

>Diff-Tracker: T2I Diffusion Models are Unsupervised Trackers
https://arxiv.org/abs/2407.08394

>E2VIDiff: Perceptual Events-to-Video Reconstruction using Diffusion Priors
https://arxiv.org/abs/2407.08231

>Enrich content Using Context-Aware Copy Paste
https://arxiv.org/abs/2407.08151

>EchoMimic: Audio-Driven Portrait Animations through Editable Landmark Conditions
https://arxiv.org/abs/2407.08136

>FYI: Flip Your Images for Dataset Distillation
https://arxiv.org/abs/2407.08113

>Coherent and Multi-modality Image Inpainting via Latent Space Optimization
https://arxiv.org/abs/2407.08019

>Spatially-Variant Degradation Model for Dataset-free SR
https://arxiv.org/abs/2407.08252
>>
>>101382450
>GENERATE MY DICK IN YOUR MOUTH INSTEAD
h-hot
>>
>mfw Resource news

07/12/2024

>Progressive Growing for Stable Diffusion Web UI
https://github.com/rmcc3/experiments-progressive-growing

>Video Diffusion Alignment via Reward Gradient
https://vader-vid.github.io/

>MAVIS: Mathematical Visual Instruction Tuning
https://github.com/ZrrSkywalker/MAVIS

>Pallaidium Adds Kwai/Kolors Support
https://github.com/tin2tin/Pallaidium#change-log

>SEED-Story: Multimodal Long Story Generation with Large Language Model
https://github.com/TencentARC/SEED-Story

>15M Multimodal Facial Image-Text Dataset
https://huggingface.co/datasets/OpenFace-CQUPT/FaceCaption-15M

>OpenDiLoCo, open-source implementation and scaling of DeepMind’s Distributed Low-Communication
https://www.primeintellect.ai/blog/opendiloco

>CLIPTrase: Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation
https://github.com/leaves162/CLIPtrase

>MambaVision: A Hybrid Mamba-Transformer Vision Backbone
https://github.com/NVlabs/MambaVision

>Kohaku Fork of Paints-UNDO
https://github.com/KohakuBlueleaf/Paints-UNDO#what-this-fork-changed

07/11/2024

>Introducing AuraFlow v0.1, an Open Exploration of Large Rectified Flow Models
https://blog.fal.ai/auraflow/

>FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision
https://pytorch.org/blog/flashattention-3/

>LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models
https://github.com/LLaVA-VL/LLaVA-NeXT

>SD Webui Resource Monitor
https://github.com/Haoming02/sd-webui-resource-monitor

>AI's Energy Demands Are Out of Control
https://www.wired.com/story/ai-energy-demands-water-impact-internet-hyper-consumption-era/

>German court rules AI output can be protectable
https://devclass.com/2024/07/10/german-court-rules-ai-output-can-be-protectable-ups-stakes-for-machine-generated-code/?td=rt-3a

>EchoMimic: Lifelike Audio-Driven Portrait Animations through Editable Landmark Conditioning
https://badtobest.github.io/echomimic.html
>>
>>
>mfw Research news

07/12/2024

>HiRes-LLaVA: Restoring Fragmentation Input in High-Resolution Large VLMs
https://arxiv.org/abs/2407.08706

>Live2Diff: Live Stream Translation via Uni-directional Attention in Video Diffusion Models
https://live2diff.github.io

>Generalizable Implicit Motion Modeling for Video Frame Interpolation
https://gseancdat.github.io/projects/GIMMVFI

>Still-Moving: Customized Video Generation without Customized Video Data
https://still-moving.github.io

>Controlling the Fidelity and Diversity of Deep Generative Models via Pseudo Density
https://arxiv.org/abs/2407.08659

>Emergent Visual-Semantic Hierarchies in Image-Text Representations
https://tau-vailab.github.io/hierarcaps

>Fine-Tuning Stable Diffusion XL for Stylistic Icon Generation: A Comparison of Caption Size
https://arxiv.org/abs/2407.08513

>VideoMamba: Spatio-Temporal Selective State Space Model
https://arxiv.org/abs/2407.08476

>Global Spatial-Temporal Information-based Residual ConvLSTM for Video Space-Time Super-Resolution
https://arxiv.org/abs/2407.08466

>Infinite Motion: Extended Motion Generation via Long Text Instructions
https://shuochengzhai.github.io/Infinite-motion.github.io

>A Comprehensive Survey on Human Video Generation: Challenges, Methods, and Insights
https://arxiv.org/abs/2407.08428

>Diff-Tracker: T2I Diffusion Models are Unsupervised Trackers
https://arxiv.org/abs/2407.08394

>E2VIDiff: Perceptual Events-to-Video Reconstruction using Diffusion Priors
https://arxiv.org/abs/2407.08231

>Enrich content Using Context-Aware Copy Paste
https://arxiv.org/abs/2407.08151

>EchoMimic: Audio-Driven Portrait Animations through Editable Landmark Conditions
https://arxiv.org/abs/2407.08136

>FYI: Flip Your Images for Dataset Distillation
https://arxiv.org/abs/2407.08113

>Coherent and Multi-modality Image Inpainting via Latent Space Optimization
https://arxiv.org/abs/2407.08019

>Spatially-Variant Degradation Model for Dataset-free SR
https://arxiv.org/abs/2407.08252
>>
>>
File: depa_00086_.png (2.85 MB, 1344x1728)
2.85 MB
2.85 MB PNG
>>101382557
>>101382570
can you stop
>>
File: 98293-tmp.png (3.1 MB, 1536x1728)
3.1 MB
3.1 MB PNG
>>
>>101382583
>>101382597
>>101382614
Hello avatars! :D
>>
Which other avatars are we missing
>>
>>
>>101382665
i miss schizo anon
>>
Why do I specifically have to define lighting via "lamp" and so on. This can't be normal.
>>
>>
>>101382544
Nice sister
>>
>>101382641
>>101382657
>>101382665
>>101382676
Hello schizo
>>
>mfw Resource news

07/12/2024

>Progressive Growing for Stable Diffusion Web UI
https://github.com/rmcc3/experiments-progressive-growing

>Video Diffusion Alignment via Reward Gradient
https://vader-vid.github.io/

>MAVIS: Mathematical Visual Instruction Tuning
https://github.com/ZrrSkywalker/MAVIS

>Pallaidium Adds Kwai/Kolors Support
https://github.com/tin2tin/Pallaidium#change-log

>SEED-Story: Multimodal Long Story Generation with Large Language Model
https://github.com/TencentARC/SEED-Story

>15M Multimodal Facial Image-Text Dataset
https://huggingface.co/datasets/OpenFace-CQUPT/FaceCaption-15M

>OpenDiLoCo, open-source implementation and scaling of DeepMind’s Distributed Low-Communication
https://www.primeintellect.ai/blog/opendiloco

>CLIPTrase: Explore the Potential of CLIP for Training-Free Open Vocabulary Semantic Segmentation
https://github.com/leaves162/CLIPtrase

>MambaVision: A Hybrid Mamba-Transformer Vision Backbone
https://github.com/NVlabs/MambaVision

>Kohaku Fork of Paints-UNDO
https://github.com/KohakuBlueleaf/Paints-UNDO#what-this-fork-changed

07/11/2024

>Introducing AuraFlow v0.1, an Open Exploration of Large Rectified Flow Models
https://blog.fal.ai/auraflow/

>FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision
https://pytorch.org/blog/flashattention-3/

>LLaVA-NeXT-Interleave: Tackling Multi-image, Video, and 3D in Large Multimodal Models
https://github.com/LLaVA-VL/LLaVA-NeXT

>SD Webui Resource Monitor
https://github.com/Haoming02/sd-webui-resource-monitor

>AI's Energy Demands Are Out of Control
https://www.wired.com/story/ai-energy-demands-water-impact-internet-hyper-consumption-era/

>German court rules AI output can be protectable
https://devclass.com/2024/07/10/german-court-rules-ai-output-can-be-protectable-ups-stakes-for-machine-generated-code/?td=rt-3a

>EchoMimic: Lifelike Audio-Driven Portrait Animations through Editable Landmark Conditioning
https://badtobest.github.io/echomimic.html
>>
File: 00578-1976863822.jpg (761 KB, 3584x2016)
761 KB
761 KB JPG
>>
File: 98299-tmp.png (3 MB, 1536x1728)
3 MB
3 MB PNG
>>
>mfw Research news

07/12/2024

>HiRes-LLaVA: Restoring Fragmentation Input in High-Resolution Large VLMs
https://arxiv.org/abs/2407.08706

>Live2Diff: Live Stream Translation via Uni-directional Attention in Video Diffusion Models
https://live2diff.github.io

>Generalizable Implicit Motion Modeling for Video Frame Interpolation
https://gseancdat.github.io/projects/GIMMVFI

>Still-Moving: Customized Video Generation without Customized Video Data
https://still-moving.github.io

>Controlling the Fidelity and Diversity of Deep Generative Models via Pseudo Density
https://arxiv.org/abs/2407.08659

>Emergent Visual-Semantic Hierarchies in Image-Text Representations
https://tau-vailab.github.io/hierarcaps

>Fine-Tuning Stable Diffusion XL for Stylistic Icon Generation: A Comparison of Caption Size
https://arxiv.org/abs/2407.08513

>VideoMamba: Spatio-Temporal Selective State Space Model
https://arxiv.org/abs/2407.08476

>Global Spatial-Temporal Information-based Residual ConvLSTM for Video Space-Time Super-Resolution
https://arxiv.org/abs/2407.08466

>Infinite Motion: Extended Motion Generation via Long Text Instructions
https://shuochengzhai.github.io/Infinite-motion.github.io

>A Comprehensive Survey on Human Video Generation: Challenges, Methods, and Insights
https://arxiv.org/abs/2407.08428

>Diff-Tracker: T2I Diffusion Models are Unsupervised Trackers
https://arxiv.org/abs/2407.08394

>E2VIDiff: Perceptual Events-to-Video Reconstruction using Diffusion Priors
https://arxiv.org/abs/2407.08231

>Enrich content Using Context-Aware Copy Paste
https://arxiv.org/abs/2407.08151

>EchoMimic: Audio-Driven Portrait Animations through Editable Landmark Conditions
https://arxiv.org/abs/2407.08136

>FYI: Flip Your Images for Dataset Distillation
https://arxiv.org/abs/2407.08113

>Coherent and Multi-modality Image Inpainting via Latent Space Optimization
https://arxiv.org/abs/2407.08019

>Spatially-Variant Degradation Model for Dataset-free SR
https://arxiv.org/abs/2407.08252
>>
>>101382911
Nice boorger

Anyone else hungry?
>>
>>101382845
>>101382920
we get it nigbo, you like to spam the news and flood these threads
>>
>>101382922
There used to be some hapa girl on YT that did ASMR, that looked exactly like that with the same tits.

She got married or something and quit back in like 2019
>>
>>101382922
>>101382941
what does that have to do with image generation?
>>
File: 00579-1976863823.jpg (808 KB, 3584x2016)
808 KB
808 KB JPG
>>101382922
I just had a burger and chilli cheese fries.. yum yum
>>
File: 98296-tmp.png (2.96 MB, 1536x1728)
2.96 MB
2.96 MB PNG
>>101382922
Yes
>>
>>101382956
Someone randomly generated the doppelganger of a real human. It's neat.
>>
Anyone else here have the news filtered? Makes browsing these threads way easier.
>>
>>101382984
desu I enjoy the rng news images.
>>
>>101382993
I'm sure you do debo
>>
>>101383007
someone gave him the catbox in the discord too kek before he tried to take credit
>>
>>101383007
That's not me, no. I'm another autist from uh, a year or two ago that quit monkeying with SD for a long time.
>>
File: 00576-1976863820.jpg (756 KB, 3584x2016)
756 KB
756 KB JPG
>>
>>101383025
catbox for what?
>>
File: 98300-tmp.png (3.13 MB, 1536x1728)
3.13 MB
3.13 MB PNG
>>
Anons, I want to run forge/a1111 on my 3080 (tfw 10gb vram)

Can I get the same or similar performance in windows or do I have to install loonix, I don't want to dualboot
>>
We need some new drama going on here, these threads are getting stale.
>>
>>101383067
just make fun of niggerjak and watch the melty
>>
>>
File: 00587-1873669416.jpg (783 KB, 3584x2016)
783 KB
783 KB JPG
>>101383064
speed will be the same.. was lower on lunix cause nvidia drivers on linux kinda were wonky, but now they are on par .. so use whatever
>>
If I can't tell the difference does it matter which sampler I pick?
>>
>>101383118
Seems to vary per model and what not

I think 2m SDE seems the best with pony? ? ?
>>
>>101383067
make controversial gens that don't break the rules
>>
>>101383108
Sure debo, sure
>>
File: 00588-1873669417.jpg (745 KB, 3584x2016)
745 KB
745 KB JPG
>>101383118
ya it does, some suck.. list of good best ones: DPM2++ (all variants kinda), Euler, DDIM .. many others are okay or good, some are just horrible

also the small a for ancestral matters alot: it means it uses a non deterministic sampling method, so your gens wont be reproducible between program restarts, but ancestral samplers like Euler a rank better in subjective aesthetic reviews for many.. my personal favorites are DPM++ 2S a and Euler a btw
>>
>>101383127
Seems like most models automatically add 50 years to any female subject unless you specify they're 10.

>WOULD YOU LIKE TO SEE IMAGES OF FLORIDA LOT LIZARDS, OR THE EXTRAS FROM CHILDREN OF THE CORN?
>>
Why are people shitting up the news? Please stop fucking with it
>>
fuck the news, vandalize it more
>>
>>101383115
I thought it was way faster in Linux. Reddit moment.

https://www.reddit.com/r/StableDiffusion/comments/ybt8aa/does_stable_diffusion_run_better_on_linux_or_on/
>>
File: 00589-1873669418.jpg (773 KB, 3584x2016)
773 KB
773 KB JPG
>>101383181
nah thats fake news .. there were troubles with cuda drivers and torch when SD initially was released, but they got fixed, these days it could probably be faster on a light weight linux distro cause less shit is running in the background than on w10/w11 .. but if your machine is beefy enough all that shouldnt matter we are talking in low one digit percentage differences here
>>
>>101383181
Honestly it probably is faster under Linux due to lack of overhead. Windows always has tons of shit going on everywhere in the background but you can get a DE down to like 500mb ram usage on boot, and NV Linux drivers are fine now.
>>
>>101383218
>>101383226
this. also if you use a window manager with very little animations and 3d effects you even get more vram to use for gens for free
>>
>regardless of your feelings about it avatars are a staple of this thread
>>
>>101383282
who's your favorite
>>
>julien
>>
>ffaze
>>
>>101383226
There's a minimum overhead of ~0.3gb regardless of os if the gpu is in display mode
>>
>shemale
>>
>failed dj
>>
i still dont get it
is "schizo anon" actually one anon or is it just for the memes?
>>
>clitty dicky
>>
>>101383351
you can observe him live right now, just dont give to much attention, its all he craves
>>
>>101383351
its whoever the avatarfags dislike at any given moment
>>
>>101383351
its supposed to be a secret according to him but its debo
>>
File: 00068-1775068249.png (2.76 MB, 1344x1728)
2.76 MB
2.76 MB PNG
>>
>yall paranoid schizos and i had my cream filling
>>
>>101383226
>>101383218
I'm more comfy with Unix systems so I'm tempted to buy picrel, a dude near me is selling it for 1200€ (which is crazy for EU prices) and having a portable SD machine (it's a mini-itx build).

Or maybe temps will be fucking crazy in that tiny box and a 3090 lmao
>>
File: 98311-tmp.png (2.9 MB, 1536x1536)
2.9 MB
2.9 MB PNG
>>
>>101383409
she's a beaut, snap it up
>>
File: mars_car.jpg (195 KB, 1536x1024)
195 KB
195 KB JPG
>>
File: image.png (236 KB, 629x764)
236 KB
236 KB PNG
>>101383409
>no picrel
>tfw im retarded
>>
File: 00001.png (2.75 MB, 1632x1632)
2.75 MB
2.75 MB PNG
>>
>Lubimiv
>>
File: box.jpg (2.44 MB, 5063x3797)
2.44 MB
2.44 MB JPG
>>101383448
I actually have two of these cases! its lovely, one old NR200P white with a 3070 running and a NR200P max with a 4090 .. temps are great on both systems, for 3090/4090 size cards I reccomend the max tho cause the included riser to veritcally mount the GPU, way better thermals that way (pic related my 4090 machine in the nr200p max)
>>
>tfw not a paranoid schizo
>>
>it was my coworkers photoshop
>>
>>101383422
Nice Bunny
>>
File: pinkcutie.jpg (3.61 MB, 4160x3120)
3.61 MB
3.61 MB JPG
>>101383448
and pic related the 3070 white one
>>
File: i4945721579.jpg (90 KB, 640x480)
90 KB
90 KB JPG
>>101383596
it looks fucking tight though
>>
>>101383708
ya it is.. the 4090 fits to the millimeter .. there are lists of 4090 and 3090 cards that will fit the nr200, before buy a GPU check these lists or you will be disappointed
>>
>>
File: molyneux.jpg (53 KB, 463x372)
53 KB
53 KB JPG
Could some kind souls make a few molyneuxs for someone with a shit computer?
>>
>>101383708
>>101383448
also! I almost forgot, dont fool yourself in thinking you can use the glass panel .. thermals are abyssmal with the glass panel
>>
>>101383742
Nah I'm considering literally buying the one in the picture, someone near me is selling it for 1200€ and it's not a scam
>>
File: 00081-373786463.jpg (155 KB, 1376x1024)
155 KB
155 KB JPG
https://desuarchive.org/g/search/subject/sdg/start/2023-07-12/end/2023-07-13/
year ago... sometimes interesting to look back in time. sadly i dont believe i here on that day
>>
>>101383793
good deal then! get it .. buy another SSD .. 1TB aint anything, you can get 4TB for 90€$ these days
>>
>>101383793
I'm wary of liquid cooling, but other than that should be peachy. I'd make sure I have a good box fan on hand though and pop that case open.
>>
>>101383158
Are you sure about the non-reproducible part?
Pretty sure they don't converge, i.e. adding more steps doesn't move closer to a single target
>>
>>101383872
thank you anon, I really appreciate your post, beautiful machines you got there

>>101384020
wdym a box fan? Removing the box plates and just pointing a powerful smal metal fan at it to blow the heat away?

And yes, I fucking hate liquid cooling, it never lasts me more than 2 years
>>
>>101384112
I mean, if temps get funky and it throttles, just take the side panel off and point a box fan at it.

Box fans here are $20 at our equivalent of poundland.

So like 30 pounds or something there.
>>
>>101384112
Pop down to your local grow shop for a fan and ducting
>>
>>101383858
Good times. So few good gens, somehow it has become worse. Rrrealist was my favorite dude, too bad he fucked off
>>
>>101383858
Cute but she could still look even more sad and pathetic and be even cuter
>>
>mfw I'd somehow ran prompt strength up so for the past like 14hr when NOTHING behaved like it used to I was just too goddamn dumb to see what I'd accidentally changed while seething.
>>
>buy $1400 gpu
>generate porn for a week then get bored
what else can i do with this piece of shit
>>
>>101384397
LLM debauchery.
>>
>>101384397
voice stuff
>>
File: 00101-3840878673.jpg (164 KB, 1376x1024)
164 KB
164 KB JPG
>>101384397
alien sentiment to me, been nearly two years of constant cute girl generating for me, rarely pornographic
>>
>>101384397
Play video games?
>>
>>101384456
There's not a lot worth playing these days. Let alone that needs an actual GPU.
>>
>>101384414
>>101384435
complete uninterested in llm stuff
>>101384456
it's a dedicated AI rig so i already have a gaming pc
>>
File: depa_00090_.png (2.84 MB, 1344x1728)
2.84 MB
2.84 MB PNG
>>101384397
I think the normal evolution is porn -> 1girl -> landscapes -> art -> anime -> 1girl -> porn

>NYANK
>>
>>101384474
...you built a dedicated AI rig just to get bored with it after a week?
>>
>>101384487
it was a good idea at the time
>>
>>101384474
There's a big setup in Skyrim for a follower that voices/generates dialogue with a LLM/Chatgpt and audio and the whole jazz. It has full awareness of it's surroundings, quests, etc etc

It'd probably take a good while to set up and truly milk. (heh)

And you'd need two rigs to really pull it off
>>
>>101384397
train a model to do think of stuff to do with an enterprise GPU
>>
>>101384549
Moron
>>
>>101384397
did you have SD experience before?
you can't possibly have explored all the different porn you are into in a week
I'd say wait a bit, then make more gens, better ones, different ones
>>
>>101384397
You need to push your boundaries but normies are unable to even understand what I mean.
Take a break and then come back or just sell your gpu because it's useless to you anyway.
>>
>>101384574
i was doing sd for about a year on my main PC but i've run out of ideas for images so i got bored
>>
>>101384616
Sell your rig I guess
>>
haven't been around for a little while, was last using SD1.5. is there anything new and exciting (for someone with a 2070S)?
has anyone like midjourney btfo'd local? the other way around perhaps?
thanks for any updates
>>
>>101384574
Don't use AI to gen porn, it will completely fry your brain
>>
>>101384692
2dnPony_V10 is a billion times better than any of the 1.5's from a year ago or whatever.
>>
>>101384704
sure doesn't look like it
>>
>>101384696
I'm doing good so far
But I was never into regular hardcore porn to begin with
Watching another man have sex is kinda cucked
>>
>>101384704
thanks, i'll look into it.
>>
>>101384704
Is that a new model?

I've been using colorburst and autismix
>>
>>101384741
>Watching another man have sex is kinda cucked
are you trying to say that you havent converted your own sex pics into a lora so you can watch yourself fucking every celebrity there is?
>>
File: 654645554.jpg (3.45 MB, 5504x2560)
3.45 MB
3.45 MB JPG
mcgriddle
>>
>>101385116
you are valuable and you matter
>>
>>101385116
Meds
>>
>>101385116
sad meal with extra pathetic
>>
>>101385157
>>101385160
the duality of chan
>>
>>101385157
trani
>>101385160
>>101385190
debo
>>
>>101385208
Nope, debo loves koff
>>
>>101385222
rainy girl stirring shit up
>>
>>101385231
Cope
>>
File: 00013-3613041319.jpg (315 KB, 1400x1400)
315 KB
315 KB JPG
>>
faggot thread
>>
>>101385273
glad you have found your home
>>
>>101385318
thanks debo this is your home after all
>>
>>101385373
but I'm debo
>>
>anon who was kicked out of /degen/ is now an /sdg/ celebrity
Can't make this shit up
>>
>one post suggesting people make fun of ran
>tantrum spiral
>>
>>101385416
which post?
>>
>>101385402
did you just get home from your mcdonalds shift or something
>>
>pedophile
>>
>>101385475
elaborate, please
>>
>he doesnt know
>>
do i need a comma after a tag before a break?
>>
>>101381683
PixArt new model
>>
File: 00003-2230670731.png (1.59 MB, 1024x1536)
1.59 MB
1.59 MB PNG
With a1111's update, How do I stop gens? There is only interrupt and skip seems like i need to spam them to get it to stop all in a batch.
>>
can't tell people to stop name dropping these discord trannies
they will just laugh at you and to it even more
but what you can do is randomly add these names to every post
>>
File: 05769-2143985701.png (1.34 MB, 1024x1024)
1.34 MB
1.34 MB PNG
>>101385622
Okay thanks
>>
File: depa_00092_.png (2.7 MB, 1344x1728)
2.7 MB
2.7 MB PNG
>>101385528
ooo, give us your initial thoughts on pixartS9m
>>
>>101385436
>>101383108
>>
File: UOohhHh.jpg (38 KB, 1024x1024)
38 KB
38 KB JPG
>>101385475
>>
>>101385622
i still dont understand why """baker anon""" insists on including the doxxcord in op
>>
File: depa_00093_.png (2.89 MB, 1344x1728)
2.89 MB
2.89 MB PNG
>>101385743
me neither tbdesu
>>
>>101385743
how do people get doxxed from a discord
>>
>1hedgehog, diaper, full diaper, baby, tiny penis, stink lines
>>
>>101385739
step into the light
>>
File: 0.jpg (432 KB, 2048x1024)
432 KB
432 KB JPG
>>
>tfw anon has him filtered
>>
File: UOohhHh2.jpg (44 KB, 1024x1024)
44 KB
44 KB JPG
>>101385762
It refuses to illuminate scenes sometimes and there's nothing I can do about it, other than adding lamps.
>>
>mfw made my xyz too big and its only 50% done but if i stop now i wont know the answers to my questions
>>
>>101385807
sunk cost
>>
>>101385760
the reply input is not the prompt input niggerjak
>>
>>101385743
>"""baker anon"""
Go deeper
https://rentry.org/sdg-link#bakery
>Cut'n'paste for PWChefAnon and other thread bakers (^_^)
>https://pastebin.com/6ZQm2vkk
It's at the very core of this general
>>
File: Weeee.jpg (265 KB, 1024x1024)
265 KB
265 KB JPG
>>
all these nobodies became somebody huh?
>>
so wheres purple witch and burger anon these days
is the quokka guy still around
>>
>>101385865
he is here almost every day
>>
File: mfw.jpg (201 KB, 1024x1024)
201 KB
201 KB JPG
>>
>>101385852
jesus fucking christ i'm on the autist list
>>
>>101385528
New? When this "new" one has been released?
>>
>https://blog.fal.ai/auraflow/

neat
>>
File: mars_car.png (2.5 MB, 1382x922)
2.5 MB
2.5 MB PNG
>>101383441
>>
File: kek.png (1.05 MB, 2044x1268)
1.05 MB
1.05 MB PNG
>>101385930
This model has been pretrained with some AI slop like ideogram, he even didn't bother removing the "censored big cat" you usually see on ideogram
https://reddit.com/r/StableDiffusion/comments/1e1ktdh/auraflow_sure_does_like_making_the_ideogram/
>>
>>101385852
>quokkAnon special edition! @( * v * )@
>https://pastebin.com/M8cDCWqq

>PWs instructs cut and paste for subject line
>quokkas does not
what the fuck is this gay shit
>>
>>101385689
The default anime drawing seems sketchy. Can't make anime screencap-like, maybe

>>101385924
Yesterday. There's photorealism ver as well but I haven't tried yet
>>
>>101385982
>Yesterday. There's photorealism ver as well but I haven't tried yet
where is it? on huggingface? can you show us the link anon
>>
>>101385989
https://civitai.com/models/573014?modelVersionId=638761
>>
>>101386008
thanks anon, much appreciated
>>
File: depa_00094_.png (3.01 MB, 1344x1728)
3.01 MB
3.01 MB PNG
>>101385865
pw got busy with real life but pops in a few times a week
burger anon was just here earlier today but takes long breaks when good games come out
quokka anon is here daily but also seems pretty busy with real life stuff. sometimes he'll post "last one from me" when he hadn't posted anything else that day, lol
>>
>>101386074
relax u fkn stalker
>>
File: 00024-396906599.png (3.46 MB, 2048x2048)
3.46 MB
3.46 MB PNG
>>
>>101382433
Dunno if i can bring this here
but recently after looking at Rule34Hentai or xxx i cant remember well, one author claimed that he made voices out of a Ai website that starts with c, he posted the entire link but i forgot the name and i can't search it rn at animation.
it was a perfect cloning to a point the website when i saw was free, and i forgot to bookmark it.
Does anyone knows the name, It starts with C and is short like cuvid cubot or something.

i am thinking of using that for some modding stuff later for oblivion since elevenlabs became too jewey with voice cloning.
>>
>>101386074
Nice
>>
>>101385974
hahah wtf. why do all these ai companies struggle to get data
>>
File: 00028-396906599.jpg (438 KB, 2560x2560)
438 KB
438 KB JPG
>>
File: woaahz.png (60 KB, 778x825)
60 KB
60 KB PNG
daaaaamn,
>>
File: ComfyUI_05805_.jpg (2.09 MB, 4096x4096)
2.09 MB
2.09 MB JPG
>>
File: 00031-396906598.jpg (408 KB, 2560x2560)
408 KB
408 KB JPG
>>101386174
NoodleUI got more complicated? Good for it!
>>
File: 00078.jpg (745 KB, 2480x3120)
745 KB
745 KB JPG
>>
>>101386174
>>101386198
copypasta and fluff lol
>>
File: 00034-396906597.jpg (366 KB, 3584x2016)
366 KB
366 KB JPG
>>
File: 000000_14534_.png (1.74 MB, 1152x896)
1.74 MB
1.74 MB PNG
>>101386198
>NoodleUI got more complicated?
You only make it complicated.

I'm liking Kolors model much better than the current SD3 offering.
>>
File: 01294.jpg (494 KB, 2496x3648)
494 KB
494 KB JPG
Afternoon /sdg/! Been a while since I posted, hope the GENing has been going well for everyone!

I have no idea which thread to put stuff in so I picked here, lol
>>
File: 00109.png (2.63 MB, 1632x1632)
2.63 MB
2.63 MB PNG
>>
File: 00037-396906596.jpg (539 KB, 3584x2016)
539 KB
539 KB JPG
>>101386203
ow lol v0.0.1, rofl I should look at the pictures before commenting
>>101386229
yea my smooth brain looses itself when using noodleUI .. I aint making any pictures at all, I admie its technical prowess tho

can Kolors gen Winnie the Pooh?
>>
>>
File: ComfyUI_temp_oljgp_00019_.png (2.87 MB, 1120x1440)
2.87 MB
2.87 MB PNG
>>
>>101386276
I'll try the poo bear.
>>
File: file.png (1.26 MB, 1530x823)
1.26 MB
1.26 MB PNG
>>101386276
>Winnie the Pooh
cute. made some with Bing create earlier
>>
File: ComfyUI_temp_oljgp_00020_.png (2.89 MB, 1120x1440)
2.89 MB
2.89 MB PNG
>>
File: 000000_14541_.png (1.48 MB, 1152x896)
1.48 MB
1.48 MB PNG
>>101386276
Hilarious. ty
>>
File: 01293.jpg (433 KB, 2496x3648)
433 KB
433 KB JPG
>>
File: 00027-396906598.jpg (365 KB, 2560x2560)
365 KB
365 KB JPG
>>101386298
beary cute! hope xi doesnt know!
>>101386307
awesome!
>>
File: 000000_14545_.png (1.45 MB, 1152x896)
1.45 MB
1.45 MB PNG
>>101386319
I need to edit my prompt, it's too realistic.
>>
>>
File: file.png (698 KB, 633x636)
698 KB
698 KB PNG
>>101386319
what he doesn't know won't hurt him
>>101386332
what's the prompt? on Bing I just use
>Winnie the Pooh wearing a scarf and holding a jar of honey, cartoon style, hearts, bees, flowers
>>
What are the best anime models out right now? Haven't been keeping up for a few months.
>>
File: 000000_14551_.png (1.63 MB, 1152x896)
1.63 MB
1.63 MB PNG
>>101386363
it's now
>Winnie the Pooh is trying to get a bee hive from a tree.

I had realistic:1.5 before..
>>
>>101383064
If you're still here, BE CAREFUL as linux nvidia drivers currently are poop. I have a dual boot of windows and linux, with each their own drives. Windows SD works fine, I can SDXL, animate, etc, however linux will shit it self on SDXL and animate, can only do pics in SD1.5. When I mean shit it self either two things happen:

> #1 it will lock my computer for 20 to 40 minutes with an error of "killed" or something
> #2 it will shut off my keyboard and mouse (the LEDs turn off but os still running but locked), where I am forced to hard reset.

Linux is not as optimized to handle SD and I am using the exact same hardware for both OS. I am using a 2060 (yeah yeah, I know, calm your tits) and upgrading soon.
>>
>>101386385
AMD mb?
>>
File: 00050-449256280.jpg (321 KB, 3584x2016)
321 KB
321 KB JPG
>>101386381
cute .. but winnie kinda has garfields fur, and the bees are still abit borked, but its getting there
>>
File: depa_00097_.png (2.97 MB, 1344x1728)
2.97 MB
2.97 MB PNG
>>101386371
people seem to like autismmix and other ponyxl derivatives. you can just go to civit and look at the most downloaded if you want though
>>
>>101386400
He's picked up Tiggers coat kek.
>>
File: file.png (626 KB, 620x642)
626 KB
626 KB PNG
>>101386381
it's kinda weird, dall-e through chatgpt refuses to generate winnie the pooh yet it's fine from bing
>>
File: 01297.jpg (595 KB, 2496x3648)
595 KB
595 KB JPG
>>101386371
MFCG and Dr Pibb for Pony are really good too. I don't use them but they are always tempting!
>>
File: 000000_14554_.png (1.58 MB, 1152x896)
1.58 MB
1.58 MB PNG
>>101386363

>>Winnie the Pooh wearing a scarf and holding a jar of honey, cartoon style, hearts, bees, flowers

kolors doesn't get his likeness off the bat, I'd probably have to spend time getting there.
>>
requesting winnie the pooh, homeless and disheveled on the side of the road, experiencing withdrawals and begging passerbys for honey
>>
File: 00048-2235691229.jpg (354 KB, 3584x2016)
354 KB
354 KB JPG
>>101386371
as >>101386414 said .. pick a pony derivate of your choice, there are even realistic ones, I am using autismmix-confetti mostly, godiva is great for realism, for best lora compability use pure pony v6, there are thousands of loras trained on that on civitai, they even added a special search option for it .. that said, backgrounds are kinda mixed bag depending on the pony derivate you use
>>
>>101386481
do you keep under75 tokens when prompting
>>
File: file.png (601 KB, 626x640)
601 KB
601 KB PNG
>>101386477
Bing has got you bro
>>
File: 000000_14556_.png (1.6 MB, 1152x896)
1.6 MB
1.6 MB PNG
>>
DreamshaperXL does perfect Winnie the Pooh, but I can't make it fuck Piglet or Tigger with a jar of honey for example.
>>
File: 00058-4198517854.jpg (327 KB, 2560x2560)
327 KB
327 KB JPG
>>101386487
depends on how complex the scene is, but mostly not .. if you do you get very clean outputs tho, but I tend to add and add and add
>>
>>101386442
the prompt adherence is pretty good desu, except the lack of hearts and it's not what i'd call cartoon style
>>101386523
bro...
>>
>>101386257
Nice booba
>>
File: 000000_14560_.png (2.16 MB, 1342x1044)
2.16 MB
2.16 MB PNG
>>101386547
>the prompt adherence is pretty good desu
Agreed, I'm not used to using general language and it just works...
>>
>>101386257
>>101386310
/aco/ has a cartoon SD thread that takes semi-realistic style also I think
if you add a little more realism they would welcome this in the realistic thread
they are gonna want to see them tiddies come out though
>>
File: What.jpg (94 KB, 1024x1024)
94 KB
94 KB JPG
>>
I'm afraid to gen because my AC is broke and it's 85 in here
>>
>>101386620
put a temp limiter on your gpu with msi afterburner
>>
>>101382450
this. there are like ~6 ai-jeet threads in catalog, plus spammer ones.
>>
File: 00061-3879669311.jpg (375 KB, 2560x2560)
375 KB
375 KB JPG
>Who here ordered a chibi latte?
>>
File: 01308.jpg (748 KB, 2496x3648)
748 KB
748 KB JPG
>>101386605
Haha, yeah I've posted there before with the stuff I can't put here. And I have a lot, but yeah, never did do the semi-realistic stuff there yet but maybe I will.
>>
>>101386398
Nope, all intel.
>>
File: 00065-2949881242.jpg (335 KB, 1920x2560)
335 KB
335 KB JPG
>WANTED: Ci Marine for serial slime molestation. Dead or Alive. Poses as maid and preys on unsuspecting juvenile slime.
>>
>>101386676
I had a very similar sounding issue with a AMD mb but NV gpu on Win and Nix. Something about the PCIE controller crashing the whole hardware peripheral layer. Was most noticeable after a few hundred gens, or with a maxed out Skyrim install for example. Never fixed it. (Total mouse/KB/headphone death, constant USB cut outs.) "Supposedly" they fixed it with bios updates (they didn't.)
>>
If all I want to do is gen lewd art using A1111, can I pair a 3060 with an ancient skylake core i3?
>>
>>
>>101386788
CPU don't really matter. Just Vram
>>
File: 00075-3993081634.jpg (363 KB, 1920x2560)
363 KB
363 KB JPG
>>101386703
this >>101386795 .. my R7 5800x barely goes to 15% usage on gens, while GPU is at 99% and vram nearly maxed out .. but, the pre and post processes can demand some CPU work, so the skylake i3 wont fly there, but it wont be a big burden
>>
File: ComfyUI_05897_.jpg (2.76 MB, 4864x3328)
2.76 MB
2.76 MB JPG
>>
>>101386795
An RAM too, right? Especially with A1111
So glad I went for 64GB
>>
File: 000000_14561_.png (1.56 MB, 1188x1188)
1.56 MB
1.56 MB PNG
>Auraflow
>>
cleavage just means "breasts visible" right? it doesn't mean "giant honkin titties"
>>
File: 00090-2239824344.png (3.27 MB, 2560x1440)
3.27 MB
3.27 MB PNG
>>101386963
depends on the model you use ofc .. some will only have depictions of cleavage with big honkers
>>
>>101386896
64 is minimum going forward, i had problems on 32, all good now.
>>
Struggling to get that chubby/thick/toned look with SDXL
>>
File: 00092-2835440818.jpg (185 KB, 2048x2048)
185 KB
185 KB JPG
>>
File: 0000000_00013_.png (1.37 MB, 1024x1024)
1.37 MB
1.37 MB PNG
>Auraflow
>>
How much of a beast is required to mix sdxl models together?
>>
File: 000000_14563_.png (1.39 MB, 1152x896)
1.39 MB
1.39 MB PNG
>>
>auraflow
>can't even do nude bodies
at least sd3 did albeit the nipples and genitalia were absent
>>
File: rp0dn71py3cd1.png (1.12 MB, 1024x1024)
1.12 MB
1.12 MB PNG
https://vocaroo.com/17tvudBFkcKm
[Verse]
He built AuraFlow, an AI dream,
To generate images, so it would seem.
But a problem lurked, oh what a gaffe,
A chubby cat would make us laugh.

[Chorus]
Oh, AuraFlow, what have you done?
Ideogram's data, it sure was fun.
He forgot to filter, didn't think it through,
Now every picture has a feline coup.

https://reddit.com/r/StableDiffusion/comments/1e1ktdh/auraflow_sure_does_like_making_the_ideogram/
>>
>Model understands perfectly what a squirting orgasm is and how stretch marks on breasts work
>Easily confused by eyes

It's always three faps forward, one fap back.
>>
>>101387128
You fat fetish fucks disgust me, eat a salad bitch
>>
File: 000000_14564_.png (1.63 MB, 1342x1044)
1.63 MB
1.63 MB PNG
>>
>>101387142
Zombies don't eat salads, anonymous.
>>
>>101387125
kek'd
>>
File: 00106-2835440818.jpg (208 KB, 2048x2048)
208 KB
208 KB JPG
>>101387128
rofl.. you have funny priorities, id choose the good eyes model
>>
>>101387125
I don't get it, did debo spam malware again
>>
>>101387161
Something I'd like to see, would be generate most, or part of an image with one model, then switch to another for facial features for example.

Some of them are amazing at faces, eyes and what not.
>>
>>101387168
look at the reddit link
>>
>>101387173
Not
>Fix faces with Codeformer
But
>Generate the face with juggernaut and copy paste it over the bad one
>>
File: 00100-2835440818.jpg (195 KB, 2048x2048)
195 KB
195 KB JPG
>>101387173
inpainting? also try adetailer if you haven't yet, sometimes is very good
>>
File: 00167.png (2.86 MB, 1840x1432)
2.86 MB
2.86 MB PNG
>>
>>101387176
I still really don't get it, but I can assume this is the dalle3 version of egg dog?
>>
>>101387200
Someone released a very undertrained model called AuraFlow
https://blog.fal.ai/auraflow/
It looks promissing but the problem is that he decided to use a lot of ideogram's output to pretrain its model, and ideogram has a chubby cat everytime it wants to censor its prompt, and this fag didn't bother to remove those on the dataset, so at the end when we run his model we got sometimes the cat kek
>>101384102
>>
File: 00170.png (2.9 MB, 1840x1432)
2.9 MB
2.9 MB PNG
>>
>>101384397
Let friends use it free & remotely
>>
File: 00175.png (2.71 MB, 1840x1432)
2.71 MB
2.71 MB PNG
>>
>>101386668
hot
>>
File: 000000_14568_.png (2.19 MB, 1440x1120)
2.19 MB
2.19 MB PNG
>>
File: fatjack.png (1.39 MB, 1017x1016)
1.39 MB
1.39 MB PNG
>>
File: 00112-1.png (1.47 MB, 1024x1536)
1.47 MB
1.47 MB PNG
>>101387199
hello sis
>>
>>101387228
that's kinda hilarious
if I see any models with a sad pufferfish
>>
>>101387125
>>101387228
do ideogram store/display the cat image in the public view, haven't checked because google/apple only for registration is annoying and i'll only end up scraping it myself if i do register. if not, then fal must have generated a bunch of images themselves
>>
>>101387371
You can test the model here, doesn't need registration
https://fal.ai/models/fal-ai/aura-flow
>>
>>101387386
that's great, anon, nothing to do with what i'm on about though
>>
>>101387348
>>
>>101387396
oh fuck I didn't understand your request, if you want to use ideogram you have no other choice but to register :(
>>
>>101387410
i mean i can probably figure out the requests from the webpack, i was just wondering if anyone who has registered and has seen the cat image in the public gallery. if there's no cat in the public gallery then fail generated their own dataset by using ideogram
>>
can you all stop spamming retard cat shit
>>
>>101387438
>was just wondering if anyone who has registered and has seen the cat image in the public gallery. i
looks like it can be scrapped yeah
https://reddit.com/r/StableDiffusion/comments/1b41ah5/psa_you_cannot_delete_images_from_ideogramai_when/
>>
File: batcj.jpg (134 KB, 2056x675)
134 KB
134 KB JPG
suddenly batchs are not working
why did i uncheck something??? fuck
>>
>>101387473
patched out to save the planet
do it for the penguins
>>
File: ComfyUI_00130_.png (353 KB, 512x512)
353 KB
353 KB PNG
>>101386703
>Something about the PCIE controller crashing
Thanks for the heads up, I'll have to look into that. Mint user here, as good as it is, linux still has a long way to go

>>101386847
Is there not a way for comfy to utilize our cpus and ram more so it unloads the heavy work off the gpus? Originally going to into local SD, I thought every recourse would be used evenly but only to find "bro just get better gpu lol"
>>
File: 00001-2767606709.png (1.36 MB, 832x1216)
1.36 MB
1.36 MB PNG
Pony is so good I'm gonna have a hard attack
>>
>>101387455
everything can be scraped, i doubt the ids of the images themselves are sequential or contain metadata so scraping is going to be done by paginating the public gallery endpoint, so the lack of deletion for free accounts doesn't answer my question which is still whether the cat images appear in those results
>>
File: 00199.png (2.83 MB, 1840x1432)
2.83 MB
2.83 MB PNG
>>
>>101387473
>>101387485
ok but seriously what the hell its not working
>>
File: 00138-3120735841.jpg (403 KB, 2304x3072)
403 KB
403 KB JPG
>>101387491
>Is there not a way for comfy to utilize our cpus and ram more so it unloads the heavy work off the gpus?
there are some little tricks for both a1111 and comfy, serach for optimization options, but the heavy lifting always needs to be done by the GPU (you can ofc run it on the CPU but it will crawl like a snail) since the tensor calculations AI needs are done way faster on massive parralel cores optimized for that
>>101387502
keep it calm! 1girls are not worth a heart attack!
>>
>wondering why i suddenly cant do batches
>check for updates in extensions
>reload ui
>doesnt reload properly
>close sd exe
>open again
>wont launch giving me a shitload of errors

FUCK
>>
File: 00146-765791970.jpg (439 KB, 2304x3072)
439 KB
439 KB JPG
>>101387588
delete all inside the extensions folder and try again
>>
>>101387588
have you tried turning it off and on again
>>
File: tracrr2232.jpg (103 KB, 1097x292)
103 KB
103 KB JPG
>>101387598
i did
>>
File: 000000_14571_.png (2.24 MB, 1341x1043)
2.24 MB
2.24 MB PNG
>>
>>101387505
ok i checked, doesn't look like the cat images are in the public gallery, desu it doesn't make sense for them to be there. fal must have generated their own images with ideogram
i'll scrape ideogram myself lol
>>
File: 00148-765791972.jpg (430 KB, 2304x3072)
430 KB
430 KB JPG
>>101387633
then delete the venv folder inside the webui folder and try again, probably some borked venv python install
>>
>>101387491
Seemed like maxed out vram was the main culprit/starting point. I know NV has some weird thing about fragmenting ram when OOM that you can disable (mirrors it to sys ram somehow) instead of "Crashing" supposedly but it's probably broken.

https://forums.developer.nvidia.com/t/vram-allocation-issues/239678/15

https://nvidia.custhelp.com/app/answers/detail/a_id/5490

(Related reading for funsies.)
>>
File: 00220.png (2.64 MB, 1840x1432)
2.64 MB
2.64 MB PNG
>>
File: 00187-3975281875.jpg (240 KB, 1024x1376)
240 KB
240 KB JPG
oh meeow
>>
File: 00002-2767606710.png (1.33 MB, 832x1216)
1.33 MB
1.33 MB PNG
>>101387535
I know, but the lewd quality of the gens...
My heart is going to unironically fail
>>
>>101387699
I LOVE tanlines
>>
yeah no cat in the couple thousand images i just scrolled through from the last hour category of their gallery
>>
>>101387699
Reminds me of my first paragraphs from Claude when I got all that set up way back when and I'd have to take a breather and look away from the screen it was so good.
>>
When I forget when half the people posting here this month will be gone in a nuclear fire in the next month during WWIII I forget some will crying I will still tell them man up and fuck off, better get that math degree quick before having to trade off women and bottled watter.
>>
pedo den btw
>>
>>101387804
cozy as heck
>>
>>101387804
i hate pedophiles
https://files.catbox.moe/7biwyy.jpg
>>
File: maybe not reddit gay.png (1.4 MB, 1042x1046)
1.4 MB
1.4 MB PNG
>>101387862
>>
File: 1720834532731.jpg (385 KB, 1024x1024)
385 KB
385 KB JPG
>>
File: grid-0009.jpg (601 KB, 2048x3072)
601 KB
601 KB JPG
I fixed it
>>
File: 1720834561628.jpg (159 KB, 1024x1024)
159 KB
159 KB JPG
>>
File: 1720834570976.jpg (214 KB, 1024x1024)
214 KB
214 KB JPG
>>
File: de_ma_pa_00003_.png (2.57 MB, 1344x1728)
2.57 MB
2.57 MB PNG
anyone recommend a good model for manga. all the anime models I try are too overtuned to anime
>>
File: 1720834586095.jpg (339 KB, 1024x1024)
339 KB
339 KB JPG
>>
>>101387908
just remove the colour in post?
>>
File: 1720834642495.jpg (233 KB, 1024x1024)
233 KB
233 KB JPG
>>
File: 1720834635124.jpg (96 KB, 1024x1024)
96 KB
96 KB JPG
>>
>>101387908
shut up deba
>>
File: 1720834650595.jpg (134 KB, 1024x1024)
134 KB
134 KB JPG
>>
File: de_ma_pa_00004_.png (2.85 MB, 1344x1728)
2.85 MB
2.85 MB PNG
>>101387914
I want narrative panels

>>101387938
no and also whats a deba
>>
File: 1720834657262.jpg (158 KB, 1024x1024)
158 KB
158 KB JPG
>>
>>101387125
can someone explain this to me? Was auraflow training on AI generations instead of real pictures, and a bunch of them were this NSFW cat? I don't know shit about shit but that sounds like a good way to kneecap your model for no reason. Also he released it knowing about this issue and knowing he'll have to start over?
>>
File: 1720834701961.jpg (137 KB, 1024x1024)
137 KB
137 KB JPG
>>
>>101387961
basically yeah >>101387228
>>
File: 00005-1727842599.png (781 KB, 1064x1192)
781 KB
781 KB PNG
>>101386666
I asked for extra cream
>>
File: 1720834709192.jpg (156 KB, 1024x1024)
156 KB
156 KB JPG
>>
File: 00195-3157409375.jpg (154 KB, 1024x1536)
154 KB
154 KB JPG
horny
>>
File: 1720834766725.jpg (170 KB, 1024x1024)
170 KB
170 KB JPG
>>
File: 00076-2549462514.png (2.43 MB, 1920x2560)
2.43 MB
2.43 MB PNG
>>101387973
coming right up grmbl
>>
File: 1720834777167.jpg (166 KB, 1024x1024)
166 KB
166 KB JPG
>>
File: jansovl5.png (822 KB, 768x768)
822 KB
822 KB PNG
>>101387985
shut up deba, faggot
>>
File: 000000_14581_.png (2.22 MB, 1043x1341)
2.22 MB
2.22 MB PNG
>>101387804
Get the rope btw.
>>
File: 1720834797319.jpg (443 KB, 1024x1024)
443 KB
443 KB JPG
>>
File: 1720834808402.jpg (345 KB, 1024x1024)
345 KB
345 KB JPG
>>
File: 1720834820673.jpg (426 KB, 1024x1024)
426 KB
426 KB JPG
>>
File: de_ma_pa_00006_.png (2.81 MB, 1344x1728)
2.81 MB
2.81 MB PNG
>cut of MCs arm
>he smiles at you
>arm turns into fingers
so fucking anime
>>
File: 1720834881031.jpg (357 KB, 1024x1024)
357 KB
357 KB JPG
>>
File: grid-0011.jpg (609 KB, 2048x3072)
609 KB
609 KB JPG
>>
File: 1720834956428.jpg (530 KB, 1024x1024)
530 KB
530 KB JPG
>>
File: 1720834978482.jpg (566 KB, 1024x1024)
566 KB
566 KB JPG
>>
File: de_ma_ho_00004_.png (3.3 MB, 1344x1728)
3.3 MB
3.3 MB PNG
hands
>>
>>101388071
>hndNnas
>>
File: 1720835000604.jpg (498 KB, 1024x1024)
498 KB
498 KB JPG
>>
File: de_ma_ho_00001_.png (3.42 MB, 1344x1728)
3.42 MB
3.42 MB PNG
>>101388094
true
>>
>>101387699
Very nice gen
>>
we rushed to a new bake yet we have no bakers
such cruel fate
>>
age slider staying at -3
>>
>>101388159
>>101388159
>>101388159



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.