Previous: >>109225172NEW: https://huggingface.co/circlestone-labs/AnimaAnima1.0 is out>UIsComfyUI:https://github.com/comfyanonymous/ComfyUISwarmUI:https://github.com/mcmonkeyprojects/SwarmUIKritaAI: https://github.com/Acly/krita-ai-diffusionForge Classic/Neo: https://github.com/Haoming02/sd-webui-forge-classic>How to generate anime:image : https://rentry.org/comfyui_guide_1girlvideo: https://rentry.org/wan22ldgguide>Style ExplorersAnima: https://anima.mooshieblob.com/ | https://animadex.net/Illustrious: https://tagexplorer.github.io>Checkpoints, LoRAs, Upscalers, & Workflowshttps://civitai.comhttps://civitaiarchive.comhttps://tensor.arthttps://openmodeldb.infohttps://www.seaart.aihttps://www.liblib.art/>Resources:Share Metadata: https://litterbox.catbox.moeImg2Prompt: https://huggingface.co/spaces/fancyfeast/joy-caption-beta-oneOnline metadata viewer SD/NovelAI: https://spell.novelai.devCatbox/Metadata Userscript: https://gist.github.com/catboxanon/ca46eb79ce55e3216aecab49d5c7a3fbOutput cleanup: https://rentry.org/RemovingDiffusionGunk , https://www.mediafire.com/file/vipr23exc5htmnt (batch processing python script)Inpainting Guide from an Anon: https://files.catbox.moe/fbzsxb.jpg >>106520607>/adt/ logo:https://files.catbox.moe/pakfgt.png
Thanks for bakering
>>109256670no fuck you, I did it for my own good
>>109256675thanks anon
>>109256690ree
>>109256648hello
>>109256704hi anon
>>109256721lots of red AIgold in the OP. i like the bottom right corner who is doing the peace sign
thanks for bakeri...>>109256670:(
>>109256648shit and bad taste collagealternative collage
>>109256932very anime
>>109256932Tell that to the faggot who treats 1girls like still life
>>109256978>>109256948
>didn't make either collageits ogre
>>109256996make your own, we are adults
No Gemini watermark because I used Google Flow
>>109256648Mark Rothko sponsored OP
>>109256932You're unwell.
>>109258829
is there a trick to getting consistent style with anima? no matter what artist tags i include it never ends up looking like the artist in question and it also always looks like scribbly line dogshit. coming from wai btw.
>>109259139Training a lora is the easiest way to get consistent style in my experience.
>>109255655Did you set the expiry time really short or is litterbox just broken?
>>109259139embrace the random>>109259550it was an hour yeah. i forgot to set it longer then nobody said anything so i didnt bother to up again.
>>109259604Could you reupload please? I asked before I went to bed because I knew we were gonna hit the post limit.
>>109259633slightly differenterhttps://litter.catbox.moe/xo7tsz.png
>>109256932fuckin saved
>>109259713ty love you
>>109256648Running ComfyUI locally here:. So what's the big difference between using an edit model like Qwen-edit and an inpainting-capable model? I was interested in playing around with edit models for the heck of it after relearning how to inpaint using Anima but even fp8 Qwen edit models are like 20 GB+. Even when I asked Kimi (yes go ahead and laugh) for recommendations it kept insisting I download the bf16 version that would have been like 40 GB. My rig is more than Pitbull running it but I'm questioning whether or not it's necessary for me to download all of that and potentially waste 20 to 40 GB of storage? Is Qwen edit really that good? What I need a model like that to do things like pic rel or to do more advanced things like redesigning my bedroom or something? (My parents used chat GPT's image added capabilities in order to help them redesign their kitchen and backyard patio so I'm wondering if something like this would A) be possible to do locally and B) whether or not you need fatass models like thesehttps://huggingface.co/Comfy-Org/Qwen-Image-Edit_ComfyUI/tree/main/split_files/diffusion_models
>>109259765Use flux klein 9b for edits instead, it's smaller than qwenAlso use fp8 instead of bf16, there is like no difference
>>109259765>A) be possible to do locallyyea but they are shit, they will change the style or the face of the cahracter> B) whether or not you need fatass models like theseYou can already do this locally, but most local models are still pretty bad at it. They tend to change the art style or mess up the character's face instead of actually editing what you asked for. If you want to compare models yourself, you can use Arena.ai. It lets you try both local and cloud models for free, and you do not even need a GPU as long as the content is SFW. That is how I realized Qwen Edit is probably the best one for image editing right now, although that thing is fucking heavy.
>>109259874nta and I don't really have a use for edit models, but can't you just set a latent noise mask to prevent it from changing other parts of the image? Not quite as turnkey, but it's a reasonable compromise for running a model locally.
>Qwen Edit is fucking heavyuse gguf
Is there a model that makes everything I want?
$25just kidding. it can barely generate fingers.
>>109260169damn now I remember when luddites swore up and down AI would NEVER be able to gen fingers
>>109260236kawaii...
>>109260236...catbox...?
>>109260280it's nai
>>109260280>>109260300erm it's anima actuallyhttps://litter.catbox.moe/v6iw8bil225ez9vb.pnghttps://litter.catbox.moe/8tbgggsa73em1klk.safetensors
>prompt anima for "creepy">get thisoh no haha please don't come closer ahaha how terrifying I'm scared stiff
>>109260326i don't get a workflow when i drag that image into comfy and also i have no idea what that safetensors file is for. am i silly?
>>109260392I'm using swarmui, you can drop the image into https://spell.novelai.dev to see the metadataAnd the safetensors file is the lora I used
>>109260176I legitimately think about this a lot. It's kind of insane and concerning how many absurdly shorted-sighted predictions luddites made about AI.
>>109259061
>>109260326>(loli:2)>animait sure is.
>>109260326ty for the lora
I just bought a 5070 ti 16GB but my CPU is a Ryzen 5 5500X3D. Is it a big bottleneck for video generation with WAN 2.1/2.2? Or does it not matter that much?
I always have a tough time reliably genning hebes without freakishly long legs. Any pointers?
>>109261136style
>>109261136proportions are strongly affected by aspect ratio. If you want shorties, make your outputs more square
>>109260326If I load this metadata, do I also get the LoRA automatically?
uma?
>>109261136I usually say something about accurate anatomy, short stature, typical features of a etc.
>>109261854I forgot to also say to this post >>109261222very nice.
>>109261773Blind dog
>>109261806come to Brazil!
>>109262161So the metadata doesn't come?
>>109261773>>109263386You download it separate and put it in the /lora/ folder The .safetensor file is the lora
>>109263512And why isn't that in the OP? is the idea to leave things half explained so people have to come here and ask??>How to generate anime:image : https://rentry.org/comfyui_guide_1girl
>>109263512why even bother
>>109263561Fuck off Ani
>>109259723I like this Slayers AIgold.
>>109256932Wtf i love anime now
>>109256648I didn't use this in foreverWhats the current model meta?Do we have models that produce non blatant AI images yet?
>>109264855Prompt? Catbox? Model?
>>109264949Simple. No. https://civitaiarchive.com/models/523485?modelVersionId=3084641
>>109264529tweeeeeeeeeeenty fiiiiiiiiiiiiiiiiive doooooooooooooooooooooolll haaaaaaairs
Who took 'er jabs?, this general use'ta be good, these post count use'ta be high, this general use'ta have quality, now they took 'er jabsTHEY TOOK 'ER JABS! THEY TOOK 'ER JAAAABS
Bump.
need oguri
