<h2 id="this-comfyui-workflow-was-mainly-created-to-do-one-thing-in-sdxl:-&quot;multiple-characters-loras&quot;-that-embeds-every-process-of-how-the-image-is-generated.">This ComfyUI workflow was mainly created to do one thing in SDXL: "Multiple characters Loras" that embeds every process of how the image is generated.</h2>It's like semi-inpaint that starts from the beginning of the generative process, and is compatible with "Tile Upscale (Ultimate SD Upscale)".<h2 id="it-also-works-well-without-character-lora.">It also works well without character lora.</h2><h3 id="it-works-well-with-amd-gpu-should-work-with-an-nvidia-gpu-too-if-the-custom-nodes-work-as-intended.">It works well with AMD GPU, should work with an Nvidia GPU too, if the custom nodes work as intended.</h3><hr /><h1 id="featuring">📌Featuring</h1><ul><li>Illustrious 2.0 Focused Workflow (Supports SDXL models)</li><li>Regional Prompts With Regional Lora (Up to 3 characters)</li><li>All-in-one Workflow (KSampler/Ultimate SD Upscaler/Detailer/Color Correction/Upscale/Inpaint)</li><li>Optional ControlNets (OpenPose, DepthAnything, Canny, SoftEdge, Tile)</li><li>Separated Base/Detailer Prompts (less upscaler/detailer hallucinate)</li><li>Flexible Workflow Starts (Starts your generation from any part of the process)</li><li>Semi-Auto Mask (starts workflow right away without manual masking involvement)</li></ul><h3 id="suitable-for-users-who-can-navigate-the-workflow-spaghetti.">Suitable for users who can navigate the workflow spaghetti.</h3>However, I've arranged the node groups in order from left to right and up and down. Categorized separately by Header and Textboxes. It should make it easier to follow through and input text/number accordingly. With Hotkeys to quickly navigate the workflowThe Guide, how to use, and that sort of stuff are inside the workflow. Ask me if there are any confusing words in the guide that may need improvement in the comments.Please avoid flooding me with personal chat as best as you can. I encourage posting your question in the comments.<h1></h1><hr /><h1 id="forewarning">⚠️Forewarning</h1><h3 id="backup-your-local-comfyui-environment.">Backup your Local ComfyUI Environment.</h3>When installing/updating new custom nodes, things can break from forcing to install certain version of site-packages, which may be incompatible with your settings.<h3 id="not-suitable-for-impatient-users.-the-workflow-is-very-delayed.">Not suitable for impatient users. The workflow is very delayed.</h3>Depending on how many group toggles are enabled.<h3 id="and-if-using-&quot;regional-promptlora&quot;-it-will-generate-very-slowly.">And if using "Regional Prompt/Lora", it will generate very slowly.</h3>2-4 times slower (depending on how many characters/character loras enabled).<h2 id="there's-a-basic-auto-masking-but-complicated-poses-still-require-manual-masking.">There's a basic auto masking, but complicated poses still require manual masking.</h2>Similar to "Inpainting", it requires tracing the character from the image to give proper facial expression/pose and prevent character features from bleeding.<hr /><h1 id="download-section">💾Download Section</h1><h2 id="controlnet-(models-greater-controlnet)"> 🎛️ControlNet (models &gt; controlnet)</h2><ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/thibaud/controlnet-openpose-sdxl-1.0/blob/main/OpenPoseXL2.safetensors">OpenPose XL v2</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/TencentARC/t2i-adapter-depth-midas-sdxl-1.0/resolve/main/diffusion_pytorch_model.fp16.safetensors">Depth Anything</a> - Rename file to "t2i-adapter-depth-midas-sdxl-1.0".</li><li><a target="_blank" rel="ugc" href="https://huggingface.co/TencentARC/t2i-adapter-canny-sdxl-1.0/resolve/main/diffusion_pytorch_model.fp16.safetensors">Canny</a> - Rename file to "t2i-adapter-canny-sdxl-1.0".</li><li><a target="_blank" rel="ugc" href="https://huggingface.co/Eugeoter/noob-sdxl-controlnet-softedge_hed/resolve/main/noob-sdxl-controlnet-softedge_hed.safetensors">Soft Edge</a></li><li><a target="_blank" rel="ugc" href="https://huggingface.co/TTPlanet/TTPLanet_SDXL_Controlnet_Tile_Realistic/blob/main/TTPLANET_Controlnet_Tile_realistic_v2_fp16.safetensors">TTPLanet Tile</a></li></ul><h2 id="detailer-(models-greater-ultralytics)">🔍Detailer (models &gt; ultralytics)</h2><h3 id="bbox-(ultralytics-greater-bbox)">📦BBOX (ultralytics &gt; bbox)</h3><ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Bingsu/adetailer/blob/main/face_yolov8n_v2.pt">Face yolov8n v2</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/178518?modelVersionId=582143">Eyeful v2 Paired</a> or <a target="_blank" rel="ugc" href="https://civarchive.com/models/178518?modelVersionId=582139">Eyeful v2 Individual </a>(Can use either Eye Detector models)</li><li><a target="_blank" rel="ugc" href="https://huggingface.co/Bingsu/adetailer/blob/main/person_yolov8m-seg.pt">Person Yolo8m SEGS</a></li></ul><h3 id="segment-(ultralytics-greater-segm)">📐SEGMENT (ultralytics &gt; segm)</h3><ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/iahhnim/adetailer_collection/blob/main/PitHandDetailer-v2-Test-v9c.pt">PitHandDetailer-v2-Test-v9c</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/1306938/adetailer-2d-mouth-detection-yolosegmentation">2d Mouth</a></li></ul><h2 id="sams-detector-(models-greater-sams)">🧠Sams Detector (models &gt; sams)</h2><ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/datasets/Gourieff/ReActor/blob/main/models/sams/sam_vit_b_01ec64.pth">sam_vit_b_01ec64</a></li></ul><h2 id="upscale-model">💎Upscale Model</h2><ul><li><a target="_blank" rel="ugc" href="https://openmodeldb.info/models/2x-AnimeSharpV4-Fast-RCAN-PU">AnimeSharpV4_Fast_RCAN_PU</a> </li></ul><hr /><h3 id="known-issue">⚠️Known Issue</h3><ul><li>[If using Load Image node] You need to set your image "Aspect Ratio" to match your "Load Image". Still trying to fix the dependency cycles error when implementing "Auto Get Image Size".</li><li>VERY LAGGY/DELAYED when generating due to the sheer amount of nodes and checks on the frontend. Can be remedied by keeping the browser size small.</li><li>At the last part of every prompt. You have to manually add a comma "," at the end.</li><li>Overburn/Overbaked Image could still happen if using 3 characters Loras, the countermeasure is to lower CFG by 1-3 for KSampler or use Sampler: "Euler_a_cfg_pp" with CFG: 1.3-1.5.</li><li>Somehow Sampler: "Euler_a_cfg_pp" in Detailer causes artifacting in the corner, hence I use the standard Sampler/Scheduler combo for this part</li><li>[Detailer "Tile SEGS" and Detailer Fix's "Character Bound" Mode] It is somehow forcing the generation outside of the detected area. It still works, but takes unnecessary long generative time.</li></ul><hr /><h1 id="credits">Credits</h1><h3 id="nerdy-rodent-(video-link)"><a target="_blank" rel="ugc" href="https://www.youtube.com/@NerdyRodent">Nerdy Rodent</a> (<a target="_blank" rel="ugc" href="https://www.youtube.com/watch?v=dASMuSCCNSU">Video Link</a>)</h3>His video sparks my interest in creating this madness AIO of workflow using his method.<h2 id="not-so-simple-(or-is-it)-illustrious-workflow-by"><a target="_blank" rel="ugc" href="https://civarchive.com/models/741620/not-so-simple-or-is-it-illustrious-workflow">Not so simple (or is it?) Illustrious Workflow</a> by @Gladas</h2>took some good sampler/scheduler combo, nodes, and settings from there<h3 id="regional-conditioning-with-attentioncouple(lora-supported)-by"><a target="_blank" rel="ugc" href="https://civarchive.com/models/1086525/regional-conditioning-with-attentioncouplelora-supported?modelVersionId=2218825">Regional Conditioning with AttentionCouple(LoRA supported)</a> by @Nyraen07</h3>Took some inspiration from there too. Also works similarly to mine but with different custom nodes (uses "A8R8 Attention Coupling" + built-in "Set Cond Props" nodes, my workflow only uses ComfyUI built-in "Set Cond Props + Set Default Cond Props")<hr /><h2 id="end-note">END NOTE</h2>I took an interest in generating stuff that I'm... interested in, along with gaining some useful insight, and sharing images/videos. This workflow is part of my learning and experimenting node and workflows.<h3 id="if-you-find-this-workflow-on-another-site-besides-civitai-the-uploader-is-responsible-for-providing-the-support.">If you find this workflow on another site besides Civitai, the uploader is responsible for providing the support.</h3>I have no way of knowing what gets changed by the uploader.May you find my personal workflow useful and suits your needs.

Silly All-in-one, Multiple Characters With Regional Loras

Test Release<ul><li>Adds an example prompt. Can generate everything from start to finish.</li><li>Flexible generative process. Continue where you left off with Load Image.</li><li>Adds experimental Inpaint using Detailer and BBOX person detector. Manually fixing the bad anatomy part of the image is possible.</li></ul>There are still a few known issues that aren't generative breaking bugs; just need to be a bit mindful when using them.

Files with SHA256: b94aefddda7e6a401abdacc46be88cbb8976d7f36b7901e5566821e01d909c36