Discussion of free and open source text-to-image modelsPrevious /ldg/ bread : >>101532879>Beginner UIEasyDiffusion: https://easydiffusion.github.ioFooocus: https://github.com/lllyasviel/fooocusMetastable: https://metastable.studio>Advanced UIAutomatic1111: https://github.com/automatic1111/stable-diffusion-webuiComfyUI: https://github.com/comfyanonymous/ComfyUIInvokeAI: https://github.com/invoke-ai/InvokeAISD.Next: https://github.com/vladmandic/automaticSwarmUI: https://github.com/mcmonkeyprojects/SwarmUI >Use a VAE if your images look washed outhttps://rentry.org/sdvae>Model Rankinghttps://imgsys.org/rankings>Models, LoRAs & traininghttps://civitai.comhttps://huggingface.cohttps://aitracker.arthttps://www.modelscope.cn/homehttps://github.com/Nerogar/OneTrainerhttps://github.com/derrian-distro/LoRA_Easy_Training_Scripts>Pixart Sigma & Hunyuan DIThttps://huggingface.co/spaces/PixArt-alpha/PixArt-Sigmahttps://huggingface.co/spaces/Tencent-Hunyuan/HunyuanDiTNodes: https://github.com/city96/ComfyUI_ExtraModels>Kolorshttps://gokaygokay-kolors.hf.spaceNodes: https://github.com/kijai/ComfyUI-KwaiKolorsWrapper>AuraFlowhttps://fal.ai/models/fal-ai/aura-flowhttps://huggingface.co/fal/AuraFlows>Index of guides and other toolshttps://rentry.org/ldg-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>Try online without registrationtxt2img: https://www.mage.spaceimg2img: https://huggingface.co/spaces/huggingface/diffuse-the-restsd3: https://huggingface.co/spaces/stabilityai/stable-diffusion-3-medium>Related boards>>>/h/hdg>>>/e/edg>>>/d/ddg>>>/b/degen>>>/vt/vtai
my saas senses are tingling
baker has little taste
official pixart bigma and lumina 2 waiting room
No thread blessings again.. this is a bad omen
praying for the health and safety of bigma
>>101559437do we have any news about the following of pixart, hunyuan, auraflow, SD3?
>>101559450Hunyuan released a few updates but you still need to know Chinese
>>101559450>pixartno>hunyuanholding a finetuning contest for their newest model
>>101559450>do we have any newsnon at all unfortunately, although hunyuan shill mentioned a few threads back that they plan on releasing a model that specializes in anime or something, a pony level base model he said. i did not look into it unfortunately.
>>101559450>auraflow, SD3Kek
>>101559665this is cool
>>101559674Would stay at that hotel
Wait a second.. baker forgot reference to >>>/g/sdg again.. I filled the last thread so we could be healed from this mistake
>>101559846explains the lack of thread blessings. blesser anon is still out there, lost, wondering where /ldg/ went.
>>101558992>Related boards>>>/g/sdg>>>/h/hdg>>>/e/edg>>>/d/ddg>>>/b/degen>>>/vt/vtai
Well I'm out for a bit
>>101559888Why wouldn't he be searching the catalog for "ldg" like everyone else here is?
>>101560033people can't think straight when they are afraid and alone
>>101560033habits are hard to break
newfags dont know about ldg. they only search sdg.
the only reason to change out the rentry link is to make it so ldg doesnt show when you search sdg curious isnt it
>>101560215>https://rentry.org/ldg-linkredirects to >https://rentry.org/sdg-linkVery curious.. what's the point other than obfuscation?
Bump goes the typewriter
>>101560272was thred blesser killed?
>>101560371An hero for our sins
Good night anon
>>101560426gn
>>101560272>what's the point other than obfuscation?Kill the bred
artflow.ai is pretty good - upload five or more images of a subject and it creates a passable model that you can pose by prompting into any setting but it's stupid expensive. New model? 9 dollars please. More images? More shekels.I expect I can do better with controlnet or something - is there a standard workflow for this kind of thing?
>>101560793No but there's new stuff all the time. https://huggingface.co/spaces/TencentARC/PhotoMaker-V2
hidiffusion is upgraded kohya deep shrink, recommend you tryhttps://github.com/blepping/comfyui_jankhidiffusionpic related raw gen at 2048x2048, 30 steps, took 38 seconds on a 3080ti mobile
>>101560949with these hidiff-specific settings, and full pic catboxthe apply mwsmsa node has some severe limitations (bugs?), and mucking with the cross-attention blocks can either improve perspective/proportions/framing or fry your shit up even with very small valueshttps://files.catbox.moe/hfm9a0.png
>>101560949>>101560988>2048x2048, 30 steps, took 38 seconds on a 3080ti mobile>catbox Impressive. >Ancestral samplers seem to work a lot better than the non-ancestral ones when using RAUNet and ControlNet simultaneously. I recommend using the ancestral version if possible.Have you tried many sampler/scheduler combos?
https://stability.ai/news/stable-video-4dhttps://huggingface.co/stabilityai/sv4dlets you turn video into 3d video or something with different angles
>>101561789oh i think it lets you make 3d models
>>101560426Great stuff. Gn.>>101559808Love this
>>101561789>Users start by uploading a single video and specifying their desired 3D camera poses>start by uploadinggo back to your containment thread
>>101560272Remove it from the op
Now that the dust has settled, which has more promising architecture, pixart, hunyuan, auraflow or SD3? Which one has better comprehension and adherence to prompt? Which one has better colors?
>>101563216I like the pixart and hunyan look the most. All of them have decent colors, but these can be tweaked with other programs.
>>101563266which other programs?
>>101563304Photoshop, gimp, ffmpg etc
>>101563427any tips for correcting a high cfg whitewash? Besides lowering my cfg?