IL RPG Ver:These LoRAs are designed for my ComfyUI workflows only, Please take a look at the details on my Patreon <a target="_blank" rel="ugc" href="https://www.patreon.com/chambersu">Chamber | creating Game, Art | Patreon</a>, Download the generated images above, then drag them to ComfyUI to load the workflow behind them (the images generated by ComfyUI will attach their workflow data inside the image files).~ You can also try to use them in another way, but I can't promise their efficacy in other cases/projects.~ (I will provide support to Patreon followers in 24 hours if needed)If you are interested in or want to download more...<ul><li>more my personal LoRAs, merged checkpoints for game design</li></ul><ul><li>The raw generated image with generation info,</li><li>LoRA training preset files,</li><li>raw pictures, data training set, and tag preparation,</li><li>prompt structures</li><li>and how to model these images into 3d models,</li><li>...</li></ul>please take a look at my Patreon <a target="_blank" rel="ugc" href="https://www.patreon.com/chambersu">Chamber | 3d Asset, Dev Tool, Game | Patreon</a>, I will provide support to followers there, and answer whatever I know in 24 hours~ Patreon Subscriptions can help me update further and create more~There are detailed posts and assets there, and I am willing to help if needed~###IL Merged Ver: I prefer to use the LoRAs array(with BG3/Odin-MMORPG/Anime Mod/MUJI...), which are more predictable/controllable/easier to train for different styles, <a target="_blank" rel="ugc" href="https://civarchive.com/models/694655/ilf1xl-bg3-turn-multi-view-turnaround-model-sheet-character-design">IL/F1/XL BG3 Turn, Multi-View, Turnaround, Model Sheet, Character Design - IL | Illustrious LoRA | Civitai</a> <a target="_blank" rel="ugc" href="https://civarchive.com/models/1055288/il-anime-model-turn-multi-view-turnaround-model-sheet-character-design">IL Anime Model Turn, Multi-View, Turnaround, Model Sheet, Character Design - v1.0 | Illustrious LoRA | Civitai</a>But a merged one can simplify the turn LoRAs arrays as one LoRA, then we get simplified workflows, easier to understand/use, (but worse outputs in most cases I tested...) Also simplified all unnecessary nodes like the complex Detail Daemon Samplers(which are used to decrease details to clean anime designs),expanded and organized the nodes for workflow previewing easier~###XL Ver:*Every like, comment, and post will support me to go further!!Can find project files(.blend) on my Patreon <a target="_blank" rel="ugc" href="https://www.patreon.com/chambersu">Chamber | creating Game, Art | Patreon</a>. Subscriptions can help me update further and create more~Plan, future, coming soon...- [ ] more data from 3d models rendering test(about 100(will post if any improvement - [ ] more steps test (around 5000 (will post if any improvement- [ ] better tags and prompt test (including wildcards(will post if any improvement - [ ] charturn (also others‘ turntable) ti for sdxl/sd1.5 - [ ] charturn (also others‘ turntable)remix ver for both sd1.5 and sdxl, based on about 60 picked-generated images to try to make the output prettier(will post if any improvement, relate to the current version based on 20-50 3d model render images aimed at multi-view correction - [ ] charturn age/body shape/height/suit outline/face control test(will post if any improvement - [ ] my LoRA/models will post in public, and my train/generate tool/workflow will post on my Patreon onlyMerged all turntable images (~80 3D models render images) as training data to create this LoRA, seems to work better on clothing, as for bodyturn/faceturn/mechturn, they still need the special Helper LoRA. Tried XL CKPT/LoRA plan to improve the hand/palm/finger, hair/helmet/mask, and clothing, for controllable elements/styles, quantity, and consistency, they did have some improvements, you can feel from these samples above, The main improvement I feel is a better understanding of the prompts, like some kind of clothes and characters, better to control the outputs by prompts details,but after training and testing, I think SDXL has some shorting at this stage(3/22/2024):<ul><li>the main one, there are not enough awesome checkpoints XL and LoRAs XL to make results as good as SD 1.5, both for design tools like style offset, and their beauty level as base/beautify models</li><li>The open pose controller XL is not as good as SD 1.5 version,</li><li>Slower and harder to train/generate (4-5min for each 1536*1024 upscale to 1920*1024 then turn on ADetailer with 4 faces, actually with hands improvements may generate better but too slow to add more than 10 palms in ADetailer), while lighting versions work faster but their results don't look good enough for me now.</li></ul>I am not familiar with CKPT XL, and it seems too expensive to train now, so let's support other great creators and wait for some better results, For more tools LoRA XL, I will do some research and try later~Good news, RTX 4080 Laptop 12Gb VRAM is enough to train XL LoRA after some settings and updating kohya_ss~!!! If you want similar results, Please repeat most of the generate/sample settings from the raw pictures examples above, including ControlNet(OpenPose) and also others' LoRA, different extra LoRA/TI have their special efforts, add or delete them to test what suits your targets~Here is one of the LoRAs I trained for my new workflow, AI-powered designs, and textures(camera-based projection) tools. for SD XL<a target="_blank" rel="ugc" href="https://civarchive.com/models/362573/clothturn-helper-xl?modelVersionId=405133">Clothturn Helper XL - v1.0 | Stable Diffusion LoRA | Civitai</a><a target="_blank" rel="ugc" href="https://civarchive.com/models/362559/charturn-helper-merged-xl">Charturn Helper Merged XL - v1.0 | Stable Diffusion LoRA | Civitai</a>for SD 1.5:-<a target="_blank" rel="ugc" href="https://civarchive.com/models/361229/charturn-helper-merged">Charturn Helper Merged - v4.0 | Stable Diffusion LoRA | Civitai</a> - <a target="_blank" rel="ugc" href="https://civarchive.com/models/359478">Bodyturn Helper - v3.0 | Stable Diffusion LoRA | Civitai</a> - <a target="_blank" rel="ugc" href="https://civarchive.com/models/359567?modelVersionId=401878">Faceturn Helper - v2.0 | Stable Diffusion LoRA | Civitai</a> - <a target="_blank" rel="ugc" href="https://civarchive.com/models/359595?modelVersionId=401915">Hairturn Helper - v1.0 | Stable Diffusion LoRA | Civitai</a> - <a target="_blank" rel="ugc" href="https://civarchive.com/models/359624?modelVersionId=401943">Clothturn Helper - v1.0 | Stable Diffusion LoRA | Civitai</a> - <a target="_blank" rel="ugc" href="https://civarchive.com/models/359646/mechturn-armorturn-helper?modelVersionId=401961">Mechturn / Armorturn Helper - v1.0 | Stable Diffusion LoRA | Civitai</a> <a target="_blank" rel="ugc" href="https://civarchive.com/models/359659/sword-gen">Sword Gen - v1.0B | Stable Diffusion LoRA | Civitai</a>They work best with my character Library on Blender in the future, including 3d models' mesh shape edit guide, auto texturing/shading, weight painting, and so on...~ Mostly design for asset creation to develop my own game, which may help other independent developers like me.Personally Recommend Checkpoint(only one suitable for SDXL now after testing): <a target="_blank" rel="ugc" href="https://civarchive.com/models/356201/abyssorange-xl-else">AbyssOrange XL Else - v1.0 | Stable Diffusion Checkpoint | Civitai</a>Also training data here, about 30 percent of them rendered from artwork I created and posted before <a target="_blank" rel="ugc" href="https://sketchfab.com/chambersu1996/models">3D models by ChamberSu1996 - Sketchfab</a>, the rest are bought and collected from the internet: <a target="_blank" rel="ugc" href="http://TrainDataPack.zip">TrainDataPack.zip</a> Every LoRA was trained from a relatively tiny collection including 20~ 50 pictures for this early and test version, but it still helps a lot now.And training setting for kohya_ss for every LoRA I create, only uses different data, image size, and epoch, the rest are the same(Turn on XL option): <a target="_blank" rel="ugc" href="https://1drv.ms/u/s!Am0lZ75OjQ1mmNYzEGtWF_sl3XbOww?e=G0hjqy">ToonHairTurn_v1.json </a> (recommend trying your own better data and setting, for I am not familiar with training now, just try to make every step/feature workable this time and keep settings as simple/default as possible, will learn further about reasonable data, tags, and train settings later, then train better by only using my own artwork render images to ensure it will be open to use, and follow my styles/tastes)Add dynamic prompt and wildcard from Civital to help get unlimited different results,  using Gemini and Chat GPT to generate prompt wildcards for testing work well too~if any questions or anything here interests you, Feel free to contact me by leaving a comment or mailing to <a target="_blank" rel="ugc" href="mailto:chambersu1996@gmail.com">chambersu1996@gmail.com</a> Also Discord Channel here~ <a target="_blank" rel="ugc" href="https://discord.gg/SFJFcctfgD">https://discord.gg/SFJFcctfgD</a>Every like, comment, post, and Patreon <a target="_blank" rel="ugc" href="https://www.patreon.com/chambersu">Chamber | creating Game, Art | Patreon</a> Subscriptions can help to get the budget to update further, create more, and post them~

IL/XL Charturn Merged, Multi-View, Turnaround, Model Sheet, Character Design

320 images, 1536*1024, from bg3/odin/ani mod, dim 32 alpha 16, 1440 steps, batch size 2, 4 epochs, Prodigy,

Files with SHA256: cec49cb41813897d73682b59ec41e0e42acfde426cfe6b7bab9509163421796e