{"prompt": {"37": {"inputs": {"lora_name": "Illustrious\\detailed hand focus style illustriousXL v1.1.safetensors", "strength_model": 0.7, "model": ["42", 0]}, "class_type": "LoraLoaderModelOnly", "_meta": {"title": "LoraLoaderModelOnly"}}, "38": {"inputs": {"text": "bad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines", "clip": ["53", 0]}, "class_type": "CLIPTextEncode", "_meta": {"title": "CLIP Text Encode (Negative)"}}, "39": {"inputs": {"lora_name": "Illustrious\\detailed foot focus style illustriousXL v1.safetensors", "strength_model": 0.7, "model": ["37", 0]}, "class_type": "LoraLoaderModelOnly", "_meta": {"title": "LoraLoaderModelOnly"}}, "40": {"inputs": {"lora_name": "Illustrious\\illustriousXLv01_stabilizer_v1.198.safetensors", "strength_model": 0.6, "model": ["39", 0]}, "class_type": "LoraLoaderModelOnly", "_meta": {"title": "LoraLoaderModelOnly"}}, "41": {"inputs": {"lora_name": "Illustrious\\Add_Details_v1.2.safetensors", "strength_model": 0.8, "model": ["40", 0]}, "class_type": "LoraLoaderModelOnly", "_meta": {"title": "LoraLoaderModelOnly"}}, "42": {"inputs": {"ckpt_name": "waiNSFWIllustrious_v140.safetensors"}, "class_type": "CheckpointLoaderSimple", "_meta": {"title": "Load Checkpoint"}}, "43": {"inputs": {"width": 1024, "height": 1024, "batch_size": 1}, "class_type": "EmptyLatentImage", "_meta": {"title": "Empty Latent Image"}}, "51": {"inputs": {"seed": 257243673939605, "steps": 18, "cfg": 6.8, "sampler_name": "euler_ancestral", "scheduler": "karras", "denoise": 0.35, "model": ["41", 0], "positive": ["52", 0], "negative": ["38", 0], "latent_image": ["54", 0]}, "class_type": "KSampler", "_meta": {"title": "KSampler (Refiner) 257243673939605"}}, "52": {"inputs": {"text": "son goku,super saiyan blue,energy,aura,space,stars,galaxy", "clip": ["53", 0]}, "class_type": "CLIPTextEncode", "_meta": {"title": "CLIP Text Encode (Positive)"}}, "53": {"inputs": {"stop_at_clip_layer": -2, "clip": ["57", 1]}, "class_type": "CLIPSetLastLayer", "_meta": {"title": "CLIP Set Last Layer"}}, "54": {"inputs": {"scale_method": "area", "scale_factor": 1.0, "use_tiled_vae": true, "samples": ["61", 0], "vae": ["57", 2], "upscale_model_opt": ["56", 0]}, "class_type": "LatentPixelScale", "_meta": {"title": "Latent Scale (on Pixel Space)"}}, "56": {"inputs": {"model_name": "4x-AnimeSharp.pth"}, "class_type": "UpscaleModelLoader", "_meta": {"title": "Load Upscale Model"}}, "57": {"inputs": {"ckpt_name": "waiNSFWIllustrious_v150.safetensors"}, "class_type": "CheckpointLoaderSimple", "_meta": {"title": "Load Checkpoint"}}, "58": {"inputs": {"seed": 311375798830678, "strength": 0.01, "image": ["73", 1]}, "class_type": "ImageAddNoise", "_meta": {"title": "ImageAddNoise"}}, "59": {"inputs": {"filename_prefix": "STUDIO1911A2 Machine Gun Masterpiece by Freyja Pixel", "images": ["58", 0]}, "class_type": "SaveImage", "_meta": {"title": "Save Image"}}, "61": {"inputs": {"seed": 321736340638903, "steps": 32, "cfg": 7.2, "sampler_name": "euler_ancestral", "scheduler": "normal", "denoise": 1.0, "model": ["57", 0], "positive": ["52", 0], "negative": ["38", 0], "latent_image": ["43", 0]}, "class_type": "KSampler", "_meta": {"title": "KSampler (First Pass) 321736340638903"}}, "72": {"inputs": {"scale_method": "lanczos", "scale_factor": 1.0, "use_tiled_vae": true, "samples": ["51", 0], "vae": ["57", 2], "upscale_model_opt": ["74", 0]}, "class_type": "LatentPixelScale", "_meta": {"title": "Latent Scale (on Pixel Space)"}}, "73": {"inputs": {"scale_method": "area", "scale_factor": 1.0, "use_tiled_vae": true, "samples": ["72", 0], "vae": ["57", 2], "upscale_model_opt": ["56", 0]}, "class_type": "LatentPixelScale", "_meta": {"title": "Latent Scale (on Pixel Space)"}}, "74": {"inputs": {"model_name": "RealESRGAN_x4plus_anime_6B.pth"}, "class_type": "UpscaleModelLoader", "_meta": {"title": "Load Upscale Model"}}}, "workflow": {"id": "e3dc2d81-5357-4f29-a7d9-bb1dd63765c7", "revision": 0, "last_node_id": 74, "last_link_id": 97, "nodes": [{"id": 40, "type": "LoraLoaderModelOnly", "pos": [5135.11536293114, 1451.8291669267237], "size": [324.9131223761374, 82.74650922824856], "flags": {}, "order": 25, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 67}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [68]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.67", "Node name for S&R": "LoraLoaderModelOnly", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.2.2"}}, "widgets_values": ["Illustrious\\illustriousXLv01_stabilizer_v1.198.safetensors", 0.6]}, {"id": 42, "type": "CheckpointLoaderSimple", "pos": [4833.459567132996, 1007.7573453529355], "size": [299.99504371627063, 106.50022410121454], "flags": {}, "order": 0, "mode": 0, "inputs": [], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [69]}, {"name": "CLIP", "type": "CLIP", "links": []}, {"name": "VAE", "type": "VAE", "links": null}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.73", "Node name for S&R": "CheckpointLoaderSimple", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": ["waiNSFWIllustrious_v140.safetensors"]}, {"id": 37, "type": "LoraLoaderModelOnly", "pos": [5136.244918532845, 1315.152373688168], "size": [326.00711059885225, 85.20041765647107], "flags": {}, "order": 20, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 64}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [66]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.67", "Node name for S&R": "LoraLoaderModelOnly", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.2.2"}}, "widgets_values": ["Illustrious\\detailed hand focus style illustriousXL v1.1.safetensors", 0.7]}, {"id": 39, "type": "LoraLoaderModelOnly", "pos": [5476.143648635178, 1315.2688945079497], "size": [301.20821393281585, 82.80010441411775], "flags": {}, "order": 23, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 66}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [67]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.67", "Node name for S&R": "LoraLoaderModelOnly", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.2.2"}}, "widgets_values": ["Illustrious\\detailed foot focus style illustriousXL v1.safetensors", 0.7]}, {"id": 46, "type": "MarkdownNote", "pos": [4548.984023226235, 1154.2735442598544], "size": [582.1497400095363, 109.96641282878102], "flags": {}, "order": 1, "mode": 0, "inputs": [], "outputs": [], "title": "Checkpoint Loads \ud83d\udda5\ufe0f", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.1", "input_ue_unconnectable": {}}}, "widgets_values": [" : \u0317\u0300\u204d Left is the First /Primary with a single LoRA loader and latent Hi-Res Upscale pass : \u0317\u0300\u204d Right is the refiner and LoRAs : \u0317\u0300\u204d\n\n : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d : \u0317\u0300\u204d\n\nI tested 15-15, 15-14, 14-14, and 14-15! I liked 15-14 best, and 14-15 leaned more anime/comic style \u2661\n\nVersion 15's VAE seemed superior in gen time, so I used that for all the wiring!"], "color": "#432", "bgcolor": "#653"}, {"id": 41, "type": "LoraLoaderModelOnly", "pos": [5475.35193207743, 1452.2865293160812], "size": [305.60831431988163, 88.27514185556436], "flags": {}, "order": 27, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 68}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [73]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.67", "Node name for S&R": "LoraLoaderModelOnly", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.2.2"}}, "widgets_values": ["Illustrious\\Add_Details_v1.2.safetensors", 0.8]}, {"id": 47, "type": "MarkdownNote", "pos": [5137.176058863044, 1155.9346572472423], "size": [926.9176634350233, 108.13711841847436], "flags": {}, "order": 2, "mode": 0, "inputs": [], "outputs": [], "title": "LoRAs and Loaders for second pass \ud83c\udfa8", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.1", "input_ue_unconnectable": {}}}, "widgets_values": ["The rgthree Power Lora Loader is above, looped into the refiner pass! I included some of my testing character and style LoRAs and strengths! You can \"push\" the initial gen by adding a low-strength lora to (near the Ksampler) generally between .2-.4 and add a similar or different LoRA (I have only tested using the same Lora) to match a character or style in the Power LoRA Loader (generally at a lower than creator recommended strength like .4-.8).\n\n-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-\n\nMy tuned and tested refinement LoRAs phase is below! Be careful with these settings! It's not the size, but the motion of the ocean! I've also found that running just the Stabilizer and Add Detailer without or with bypassed the hands and feet LoRAs seem to produce stable anatomical results!"], "color": "#432", "bgcolor": "#653"}, {"id": 48, "type": "MarkdownNote", "pos": [4547.9581116696245, 1637.6143982062733], "size": [584.0191362738829, 88.60113028859337], "flags": {}, "order": 3, "mode": 0, "inputs": [], "outputs": [], "title": "KSampler and Prompt Order \ud83d\udccb", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.1", "input_ue_unconnectable": {}}}, "widgets_values": ["<3 <3 <3 Left is First Pass/Main K-Sampler <3 <3 <3 <3 <3 <3 <3 <3 <3 <3 <3 <3 Right is Second Pass/Refiner <3 <3 <3\n\n-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-\n\n~#Left is Positive Prompt#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#~#Right is Negative Prompt#~"], "color": "#432", "bgcolor": "#653"}, {"id": 68, "type": "MarkdownNote", "pos": [4546.562307282017, 1762.2066636407646], "size": [583.0996068541963, 88], "flags": {}, "order": 4, "mode": 0, "inputs": [], "outputs": [], "title": "AUTOCOMPLETE WARNING \u26a0\ufe0f", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.2", "input_ue_unconnectable": {}}}, "widgets_values": ["I noted that when I was using the \"AutoComplete\" ComfyUI set up, that the default settings were adjusting my commas and input syntax within the prompt boxes. This was causing significant and noticable generation errors. I would reccomend checking your ComfyUI menu \"Settings\" to \"AutocompletePlus\" and disabling the \"AutoFormatting\" entirely. "], "color": "#432", "bgcolor": "#653"}, {"id": 53, "type": "CLIPSetLastLayer", "pos": [4259.811484496944, 847.4275606453095], "size": [280.824352959567, 58.74650922824861], "flags": {}, "order": 18, "mode": 0, "inputs": [{"name": "clip", "type": "CLIP", "link": 78}], "outputs": [{"name": "CLIP", "type": "CLIP", "links": [70]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.71", "Node name for S&R": "CLIPSetLastLayer", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": [-2]}, {"id": 38, "type": "CLIPTextEncode", "pos": [4846.86294641102, 1916.2635496815842], "size": [285.2749843560564, 360.2764926832941], "flags": {}, "order": 21, "mode": 0, "inputs": [{"name": "clip", "type": "CLIP", "link": 65}], "outputs": [{"name": "CONDITIONING", "type": "CONDITIONING", "links": [75, 87]}], "title": "CLIP Text Encode (Negative)", "properties": {"cnr_id": "comfy-core", "ver": "0.3.71", "Node name for S&R": "CLIPTextEncode", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": ["bad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines"]}, {"id": 55, "type": "LoraLoaderModelOnly", "pos": [4544.996471371185, 840.785901543012], "size": [270, 82], "flags": {}, "order": 17, "mode": 4, "inputs": [{"name": "model", "type": "MODEL", "link": 82}], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [85]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.75", "Node name for S&R": "LoraLoaderModelOnly", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": ["Illustrious\\Unchou_Kan-u-04.safetensors", 0.6]}, {"id": 49, "type": "Power Lora Loader (rgthree)", "pos": [5138.863725582507, 826.313302929664], "size": [324.82119140625, 286], "flags": {}, "order": 19, "mode": 4, "inputs": [{"dir": 3, "name": "model", "type": "MODEL", "link": 69}, {"dir": 3, "name": "clip", "type": "CLIP", "link": 70}], "outputs": [{"dir": 4, "name": "MODEL", "shape": 3, "type": "MODEL", "links": [64]}, {"dir": 4, "name": "CLIP", "shape": 3, "type": "CLIP", "links": [65, 77]}], "properties": {"cnr_id": "rgthree-comfy", "ver": "1.0.2511270846", "Show Strengths": "Single Strength", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": [{}, {"type": "PowerLoraLoaderHeaderWidget"}, {"on": false, "lora": "Illustrious\\Unchou_Kan-u-04.safetensors", "strength": 0.7, "strengthTwo": null}, {"on": false, "lora": "Illustrious\\JanePorterXL_Character-12.safetensors", "strength": 0.8, "strengthTwo": null}, {"on": false, "lora": "Illustrious\\Mount_LadyYu_Takeyama_My_Hero_Academia.safetensors", "strength": 1, "strengthTwo": null}, {"on": false, "lora": "Illustrious\\Egirl-000010.safetensors", "strength": 0, "strengthTwo": null}, {"on": false, "lora": "Illustrious\\JessicaRabbitXL_character-12-IL.safetensors", "strength": 0.8, "strengthTwo": null}, {"on": false, "lora": "Illustrious\\Marin_Kitagawa_S1Arisa_Izayoi_cosplay_My_Dress-Up_Darling.safetensors", "strength": 1, "strengthTwo": null}, {"on": false, "lora": "Illustrious\\ALYX_style_IL.safetensors", "strength": 0.8, "strengthTwo": null}, {}, ""]}, {"id": 67, "type": "MarkdownNote", "pos": [5142.362719135386, 2605.9514541743356], "size": [957.4460883025804, 129.95082395532927], "flags": {}, "order": 5, "mode": 0, "inputs": [], "outputs": [], "title": "Latent Bounty Hunter Cyborg \ud83d\udc96", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.2", "input_ue_unconnectable": {}}}, "widgets_values": ["I started with literally with no positive prompt and only my personal negative prompt, which I never altered.\n\nHere is the final prompt after engineering and working through some artifacts. Oddly enough, adding \"black cat\" in at the end enhanced the character, and I tried adding it because I saw a shadow of a black cat in a low angle full body shot from earlier in my gens, with my standard fixed seeds.\n\nPos (clean without black cat artifact, it would be the final tag after \"night sky\"):\nmasterpiece,best quality,amazing quality,\ncowboy shot,dutch angle,depth of field,full body,\n1girl,cyborg,\nred eyes,mechanical eyes,looking at viewer,smile,(naughty face:0.9),\nwhite hair,long hair,\npale skin,cybernetics,covered nipples,\nblack leather jacket,blue crop top,black gloves,black leather pants,\nbullet earrings,choker,\nstanding,legs apart,holding handguns,\ncity background,cityscape,cyberpunk,neon lights,night sky\n\nNeg:\nbad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines"], "color": "#432", "bgcolor": "#653"}, {"id": 44, "type": "MarkdownNote", "pos": [3798.301908755628, 1450.709082312696], "size": [740.6113751863916, 1895.0084582300458], "flags": {}, "order": 6, "mode": 0, "inputs": [], "outputs": [], "title": "Freyja Pixel's \ud83d\udc96 STUDIO1911A2 Bounty Hunter Masterpiece 15-14", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.1", "input_ue_unconnectable": {}}}, "widgets_values": ["README\n\n /|_____________________,/| - _ \u2661\n /_________masterpiece () -_ -\u2661 \n \\ ___- _____________/ \n / / /__/__/ /bounty/_\n / /___-/ /hunter/_o----------------------------* BUZZ\n /_______-/\n\n \nFine-tuned for Anime Portraits! (\u25d0\u03c9\u25d0) ...kawaii~!\n\n\nWhat is it? Hoes does it work? Is it Anime AI magic?\n\n\u2022\u1d17\u2022 ... Yes. And now with MOAR anatomy, eyeshine, and balance!!!\n\n\n\u2661 Impatient n00b\ud83d\ude0d TLDR \u2661 \n---------------------------\n\n(\u0e07\u0398 \ufe39\u0398)\u0e07 me angi... i h8 read... gib waifus...\n\n\u2661 How to use \u2661\n1) Install ComfyUI, and get the ComfyUI Manager\n2) Save a snapshot in your ComfyUI Manager\n3) Download the Checkpoints, LoRAs, and Custom Nodes listed below (MANDATORY), and place them into the ComfyUI \"checkpoints\" and \"loras\" folders.\n4) Craft your prompt using the prompt guide. \"Tags\" only is preferred.\n5) Fire a shot! This runs on fixed seeds, so if you want to do a batch, you need to change your prompt every gen, or go random and risk abominations and morphs.\n6) Analyze the image for errors or artifacts. Pay close attention to the eyes, clothing, and limbs. If you see something weird, like 7 toes or a oddly shiny and colored object you didn't ask for, then you need to adjust your prompt for clarity of input and proper tag use.\n7) Fire another gen, and hope that it gets better! Once you get a good base image, fire at an increased \"batch_size\", preferring 2 or 4. You're likely to get weird morphs in at least one of these.\n8) Celebrate your imagination and creativity by looping 5-7 times until you have the perfect image.\n9) If none of those worked, try switching out KSampler1 seed with some of the recommended seeds in the JSON, then try numbers 5-8 again.\n\nNow go get those waifus!\n\n\u2661 Bonus Tip \u2661\n\n\u2661 I've FIXED the seeds with known \"good gen\" seeds from my testing.\nGo random at your own peril! Or enjoyment, who knows! It's random!\n\u2661 This machine runs on TAGS! It's best to use danbooru wiki or the Wai-Illustrious SAA Character select to make sure you have PROPER tags.\n\n\n~+~+~+~+~+~+~+~+~+~+~+~+~+~+~+~+~\n\n\nFor those of you who still care... You're always my favorites ( \u02f6\u02d8 \u00b3\u02d8(\u02f5 \u0361\u00b0 \u035c\u0296 \u0361\u00b0\u02f5) \u2661\n\nTHIS BABY (ting-ting!) is the NEW AND IMPROVED two pass, hand-tuned, Wai-Illustrious-based workflow! This version is the public release, the STUDIO1911A2 \"Master Bounty Hunter\" model!\n\nThis iteration is an attempt at a \"zero-shot gen\" of ONE 2048x2048 anime-style image with accurate anatomy and body posture! It's especially effective with single character portraits! I was able to get about 90% accuracy with this version, and the improvements in this one include: more stability through latent hi-res upscaling, fixed gen seeds, latent seed bank with associated character traits (eye color, hair color, etc), standardized prompts, a post-gen noise filter, a prompting guide, multiple example prompts, and a more polished README and instructions! As a final step, I made some variant workflows to modify the output resolution and gen times for you to have a choice of your preferred image resolution!\n\nThis version uses Wai-Illustrious version 15, with a single LoRA booster for the first pass, and Wai-Illustrious version 14 for the second pass refiner.\nEvery KSampler parameter has been tuned to Anime Style Screenshots. The second phase has an rgthree Power LoRA loaded with CLIP support so you can patch in more of your visions and characters!\n\nThere are three subtypes of this workflow! Each one has it's own modification and benefits!\n\nThe \"A2 Bounty Hunter Masterpiece\" is a high quality image generation workflow designed to create 2048x2048 images. My average gen times were between 60-100 seconds.\n\nThe \"A2 Big Game Hunter\" is an evolution of the Masterpiece, allowing for UHD outputs at both 4k and 8k image resolution. My average gen times were between 200 - 400 seconds depending upon hi-res output settings.\n\nThe \"A2 Machine Gun Gacha\" is a turbo version designed for quick and clean 1024x1024 image generation. My average gen times were between 20-40 seconds.\n\nEach subtype also includes these adjustments:\n\n\u2661 The two \"LoRA sights\" have to be properly synched up for styles and characters, normally with a lower value on the first pass and a higher value on the second pass.\n\n\u2661 If you're running into issues (or just want to do something wild), try randomizing the seeds, rolling ONLY the negative prompt with NO positive prompting attached, changing the checkpoint's from 15-14 to 15-15, 14-14, OR use different checkpoints and LoRAs entirely! I would suggest only changing ONE of those things at a time!\n\n\u2661 I've also completely vetted and verified my refiner LoRAs! The \"Detailed hands\" and \"detailed feet\" work well in tandem, and the Stabilizer and Add Detail slider all add a lot of polish! You can run it without the feet and hands, but it will reduce the overall skin detail. It produces consistent anatomy with just the Stabilizer and Add Detail Slider!\n\n\u2661 I like to keep the full 4 refiner LoRAS on for the overall detail it provides, since I found the \"Skin Detail\", and I find many other style or anatomy detail LoRAs to be a little off, uncanny valley, or lacking. Wai-Illustrious is also known for it's polish and detail, so not much is needed in my experience.\n\n\u2661 Adding in style or characters LoRAs can help with anatomy and composition as well (as long as the LoRAs have good instructions, training images, and prompts), creating enough variation on subject and scene, while not increasing the risk of severely morphing the subjects into ungodly abominations. (\u251b\u25c9\u0414\u25c9)\u251b\u5f61\u253b\u2501\u253b THERE'S TOO MANY &@%$#ING TOES!!!\n\n\u2661 If you're having problems with eyes, hands, and feet while using LoRAs, try re-engineering the prompt based on my examples and guide, rather than cutting and pasting a bunch of LoRA tags straight into the positive prompt box.\nTrust me... People do NOT clean their LoRA prompts. The output images are MUCH cleaner if you have a properly engineered and structured prompt. I've seen things most people only see in nightmares trying to get this information...\nI hope you appreciate it... \u0ca0_\u0ca0\n\nBe careful, user. I've managed to get most of the morphing out, except an occasional missing toe (it's always the damn toes)... In general, The more complicated characters and objects you add to your prompts, the more morphs you may get, so be warned! Greedy boys get weird freaky hands and feet! (or worse)\n\nComfyUI Danbooru and e621 database tag helpers:\nhttps://civitai.com/models/950325/danboorue621-autocomplete-tag-lists-incl-aliases-krita-ai-support\n\nhttps://github.com/newtextdoc1111/ComfyUI-Autocomplete-Plus/\n\nWai-Illustrious tag resources:\n(online character and tag finder) https://huggingface.co/spaces/flagrantia/character_select_saa\n\n(SAA Character Select)\nhttps://github.com/mirabarukaso/character_select_stand_alone_app\n\n\nIt uses ComfyUI (download and install https://www.comfy.org/download)\nComfyUI Manager (https://github.com/Comfy-Org/ComfyUI-Manager)\nand rgthree's Custom nodes (https://www.runcomfy.com/comfyui-nodes/rgthree-comfy).\n\nIt uses Wai-Illustrious v15.0 for the primary pass.\nhttps://civitai.com/models/827184/wai-illustrious-sdxl.\n\nIt uses Wai-Illustrious v14.0 for refiner pass with specific detailing and stabilization LoRAs.\nhttps://civitai.com/models/827184?modelVersionId=1761560\n\nDetailed hands: \nhttps://civitai.com/models/200255?modelVersionId=2212079, \n\nDetailed feet: \nhttps://civitai.com/models/200251?modelVersionId=1464471\n\nIllustrious XL Stabilizer:\nhttps://civitai.com/models/971952?modelVersionId=2055853\n\nDetail Slider:\nhttps://civitai.com/models/1333749/add-detail-slider?modelVersionId=1506032) on the second pass!\n\n\nI've meticulously hand-tuned and tested each and every thing in this workflow, and I've double and triple verified my work. I wouldn't recommend changing much outside of the prompts, unless you want to get freaky and experimental...?\nThat was rhetorical.\nI don't even wanna know. \u0ca0_\u0ca0\n\n\nSAVE your work FIRST, and THEN get creative and chaotic!\nFeel free to experiment in prompting with characters, body postures, or other things! It seems stable with most characters, especially if you add that specific character or style LoRA to the first gen, and some more support to the Power LoRA Loader!\nIf you're having consistent issues, it's probably an error in the prompt rather than the workflow settings!\n\n\nWith that in mind, here are the most successful positive and negative prompts and parameters that I tested:\n\n\n(parentheses may be helpful:1.1) ... \u2022\u1d17\u2022 ? You can do this by highlighting the tag and pressing \"Ctrl+\u2191\" to increase or reinforce or pressing \"Ctrl+\u2193\" to diminish or decrease.\n\nDefault Settings:\n\nKSampler1:\nSeed: 321736340638903\nSteps: 32\ncfg: 7.2\neuler_ancestral\nnormal\n1.00\n\nKSampler2:\nSeed: 257243673939605\nSteps: 18\ncfe: 6.8\neuler_ancestral\nkarras\n0.35\n\nGACHA-SYTLE PROMPTS, for more detail, please read the prompting guide.\n\nPositive (substitute any character for mai's tag):\nmasterpiece,best quality,amazing quality,shiranui mai\n\nNegative:\nbad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines\n\nPerspective Prompts:\nfrom above,from below,from behind,dutch angle,cowboy shot,front view\n\nPose and Body Prompts:\nstanding,legs apart,feet up,standing on one leg,side split,arms up,hands up,arms at sides,soles,toes spread,middrif,kicking,high kick,combat stance,holding weapon\n\nFun Modifiers:\nabsurdres,happy,joyful,blush,sweat,slimy skin,sad,crying,angry,blood,bruise,bikini,lingerie,high heels,revealing clothes,see-through clothes,wet\n\n\nFOR THE LOVE OF EVERYTHING, PLEASE NOTE:\n\n\nNEGATIVE PROMPTS ARE MORE POWERFUL THAN POSITIVE PROMPTS IN EVERY ILLUSTRIOUS MODEL! YOU CAN LITERALLY JUST PROMPT WITH MY DEFAULT NEG AND NO POSITIVE PROMPT! I MADE SHOWCASE WORTHY IMAGES WITH JUST NEGATIVE PROMPTS!\n\nREAD THE WHITEPAPER, YOU FILTHY DEGENERATE! \u2022\u1d17< \n\n(Illustrious whitepaper: https://arxiv.org/abs/2409.19946)\n\n\u2661 Disclaimer \u2661\n\nResults may vary! Don't blame me for skill issues. I take no responsibility or accountability for anything made with this AI tool.\n\nGood luck, cuties, and as always, enjoooy~! (^ 3 ^) \u2661 \u2661 \u2661"], "color": "#432", "bgcolor": "#653"}, {"id": 66, "type": "MarkdownNote", "pos": [5142.374950251171, 2764.513010916983], "size": [958.1952840021604, 88], "flags": {}, "order": 7, "mode": 0, "inputs": [], "outputs": [], "title": "Miku-chan~ \ud83e\ude75", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.1", "input_ue_unconnectable": {}}}, "widgets_values": ["Positive: masterpiece,best quality,amazing quality, straight-on,depth of field,soft lighting, 1girl, hatsune miku, white pupils, power elements, microphone, vibrant blue color palette, abstract,abstract background, dreamlike atmosphere, delicate linework, wind-swept hair, energy\n\nNegative: bad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines"], "color": "#432", "bgcolor": "#653"}, {"id": 69, "type": "MarkdownNote", "pos": [5143.322198425205, 2883.5489834186137], "size": [961.2825172841385, 230.772370636887], "flags": {}, "order": 8, "mode": 0, "inputs": [], "outputs": [], "title": "Mai \u2764\ufe0f Sparring with Chun-Li \ud83d\udc99!", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.2", "input_ue_unconnectable": {}}}, "widgets_values": ["Positive:\nmasterpiece,best quality,amazing quality,general,\n\naction shot,from above,atmospheric perspective,cinematic angle,depth of field,soft lighting,haze,\n\n2girls,\n\nshiranui mai,brown eyes,looking away,smile,\nbrown hair,high ponytail,\nthick thighs,\nstanding,combat stance,\n\nchun-li,brown eyes,looking at another,angry,\nblack hair,double bun,\nmuscular,\njumping,kicking,\n\noutdoors,city,alleyway,grafitti,dumpster,trash can\n\n\n\nNegative:\n\nbad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines\n\n\nNegative:\n\nbad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines"], "color": "#432", "bgcolor": "#653"}, {"id": 71, "type": "MarkdownNote", "pos": [5145.349143544074, 3140.7267151372916], "size": [958.996642744898, 437.41681189983547], "flags": {}, "order": 9, "mode": 0, "inputs": [], "outputs": [], "title": "Kitagawa Marin Photoshoot! \ud83d\udcf8\ud83d\udc83", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.2", "input_ue_unconnectable": {}}}, "widgets_values": ["Positive 1:\n\nmasterpiece,best quality,amazing quality,general,\n\nfrom above,dutch angle,cowboy shot,cinematic angle,depth of field,soft lighting,\n\n1girl,\n\nkitagawa marin,red eyes,looking at viewer,smile,blush,\nblonde hair,long hair,\nschool uniform,\nselfie,standing,reaching towards viewer,legs apart,\n\nindoors,bedroom,otaku room,bed,black sheets,black pillows,tan walls,posters,black lamp,window\n\n\nPositive 2:\n\nmasterpiece,best quality,amazing quality,general,\n\nfrom above,dutch angle,cowboy shot,cinematic angle,depth of field,soft lighting,\n\n1girl,\n\nkitagawa marin,red eyes,looking at viewer,smile,blush,\nblonde hair,long hair,\nblack bikini,\nselfie,standing,reaching towards viewer,legs apart,\n\nindoors,bedroom,otaku room,bed,black sheets,black pillows,tan walls,posters,black lamp,window\n\nPositive 3:\n\nmasterpiece,best quality,amazing quality,general,\n\nfrom above,dutch angle,cowboy shot,cinematic angle,depth of field,soft lighting,\n\n1girl,\n\nkitagawa marin,red eyes,looking at viewer,excited,blush,\nblonde hair,long hair,\nwhite wedding dress,lace trim,bridal veil,\nselfie,standing,reaching towards viewer,legs apart,\n\nindoors,bedroom,otaku room,bed,black sheets,black pillows,tan walls,posters,black lamp,window\n\nNegative:\n\nbad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines"], "color": "#432", "bgcolor": "#653"}, {"id": 45, "type": "MarkdownNote", "pos": [4083.0913168273332, 950.6311769363805], "size": [448.09013625752823, 314.88608437228834], "flags": {}, "order": 10, "mode": 0, "inputs": [], "outputs": [], "title": "CLIP, Latent Image, and Seeds \ud83d\udcf8", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.1", "input_ue_unconnectable": {}}}, "widgets_values": ["CLIP is -2 for Wai-Illustrious models and LoRAS\n\n=-=-=-=-=-=-=-=-=-=-=-\n\n1024x1024 is stable! Anything else may morph!\n\nBatch_size tests is 1, 2, 4, or 8! High bastches may lead to more distortion. To reduce abominations shoot slowly and refine the prompt, rather than massively batch images.\n\n=-=-=-=-=-=-=-=-=-=-=-\nFirst Pass (Left KSampler) Seeds\n\nWhite Hair Red Eyes: 321736340638903\n\nBlack Hair Red Eyes: 321736340638904\n\nBlonde Hair Blue Eyes: 321736340638905\n\nBlack Hair Red Eyes Wings Red Magic Aura/Crown Male Protagonist: 135714383167589\n\nRed Hair Yellow Eyes: 374096964340620\n\n"], "color": "#432", "bgcolor": "#653"}, {"id": 61, "type": "KSampler", "pos": [4551.296327811228, 1312.4707672694947], "size": [297.491796875, 279.1247309696677], "flags": {}, "order": 24, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 85}, {"name": "positive", "type": "CONDITIONING", "link": 86}, {"name": "negative", "type": "CONDITIONING", "link": 87}, {"name": "latent_image", "type": "LATENT", "link": 88}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [79]}], "title": "KSampler (First Pass) 321736340638903", "properties": {"cnr_id": "comfy-core", "ver": "0.3.71", "Node name for S&R": "KSampler", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": [321736340638903, "fixed", 32, 7.2, "euler_ancestral", "normal", 1]}, {"id": 43, "type": "EmptyLatentImage", "pos": [4323.007017592539, 1310.7551283142877], "size": [212.18210294759365, 107.4547352983958], "flags": {}, "order": 11, "mode": 0, "inputs": [], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [88]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.71", "Node name for S&R": "EmptyLatentImage", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": [1024, 1024, 1]}, {"id": 64, "type": "MarkdownNote", "pos": [5146.80001756235, 1576.0267184839781], "size": [952.7420111455967, 1003.840712184392], "flags": {}, "order": 12, "mode": 0, "inputs": [], "outputs": [], "title": "Wai-Illustrious Prompting Guide - Double Click the text to see the full prompting structure", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.1", "input_ue_unconnectable": {}}}, "widgets_values": ["I've been using Wai-Illustrious since I began my public work with generative AI in January 2025. I've used most of the versions from 8 - 15, and I am currently testing multiple versions to create additional specialized ComfyUI workflows. Here's some of the assembled information I have learned about prompting this particular checkpoint, combination KSamplers, and workflow.\n\n\n1. This thing works best with pure danbooru tags. Natural language can work, but it seems to add more morphing, weird problems, and artifacts. Use natural language sparingly and sparsely for when a tag does not quite fit your vision.\n\n\n2. Negative prompts are exponentially stronger than positive prompts. For example, you can generate a complete and mostly accurate image with just the default or slightly improved Wai-Illustrious creator negative prompt without any positive prompting. As such, use lighter and tighter negative prompts rather than flooding the model with instruction noise during generation by feeding a 500 character negative prompt.\n\n\n3. I try to keep my prompts as short and concise as possible, using white space and line breaks for human readability and separation of different sections of the prompt. This is what the creator of Wai Illustrious does as well, though they seem to prefer just the use of white space as form of separation. Less characters is better, since its a much more clear and much less ambiguous input. Simple words make for simple tokens which reduce ambiguity during generations. For example, imagine a dude emailing \"MAKE ME TIFA!\" for an artist commission is 17 different sentences using mostly the same words, each emphasizing different \"best\" parts of Tifa without really specific details about their vision of the scene, composition, lighting, styling, etc. This dude then starts doubly repeating what they don't want to see or the \"worst\" parts of Tifa across 34 sentences in a completely separate email, and then ending that entire email chain with paragraph with \"masterpiece,best quality,high quality, no weird stuff\".... That's not going to be a good commission for the artist to create or iterate for improvements. That's essentially what a person is doing when they flood the CLIP and generation model with excessive natural language positive and negative prompts. Compare that previous example to a commission email that simply reads \"masterpiece anime style picture of Tifa Lockhart from above while pursing her lips. No bad images, sketches, comic style, floating text blocks, or bad hands.\" That's a much clearer request, input, and prompt for that artist to create a better image.\n\n\n4. Structure your prompts! This helps you find your best prompts in the moment as well as archive them for later, and it helps the model know what you want in an orderly fashion. Most AI systems, including ComfyUI, Illustrious, and SDXL, have \"prompt weights\" which favor things closer the to beginning of the prompt. You need to tell the AI things in a standardized format to help it understand what you want to see as well. Moving things around in your prompt, from lower to higher in the order or vice versa, helps alter or define the image! Use parentheses sparingly but effectively to highlight the most important part of the prompt or to coach the model where it is having problems. My standard structure is listed below with a \"zero-shot\" prompt example, or a prompt that got what I wanted the first try, as well as an example of engineering out a tag spelling and order error. This was not a perfect prompt, because such things do not exist, however it was very quality and resulted in multiple showcase or publication quality images with no morphs or anatomy problems when run on a 4 image batch. That's as close to \"zero-shot\" as I have gotten with these checkpoints. I've created this specific prompt engineering structure using official guides, released images and workflows, community resources, multiple AI virtual employees, danbooru wiki, and the Wai-Illustrious SAA Character Selector.\n\n\n5. It is better to run one gen at a time, analyze, adjust your prompt, and get a stellar result than prompt 100 times in a long overnight batch using the same prompt and hope the robot will get it right. Robots follow YOUR instructions, they will not \"get it right\" one time by chance. You will likely get more machine garbage and morphs than you can physically sort using the largest and fastest batch method rather than my recommended single shot iteration strategy.\n\n\n6. If you find an image you like, save the seeds, prompts, and parameters somewhere safe to experiment with later!\n\n\n7. There are many, many tools and variations. Some examples are LLM-assisted prompting, auto-tag converters, wildcards, dynamic prompting, embeddings, multiple koma/comic panel style images, multiple characters, anthro/furry content, and much more. I have only begun to scratch the surface here and in my testing. Community resources are often your best bet to learn and challenge yourself! Try some CivitAI bounties! That's how I was inspired to make this version of the 1911!\n\n\nStandardized Positive Prompt Structure\n\nDEFAULT PROMPTS FIRST to help coach the model into a great artistic space.\n\nFollow that with construction elements in this order perspective, angle, depth of field/focus, lighting, number of subjects, and general style.\n\nSpecific characters and franchises or series.\n\n1girl,2boys, etc. tags to help focus the model on the desired number of subjects.\n\nSubject 1 description. Eyes, view/gaze, expression, hair, make-up, physical bodily description, clothing and shoes, jewelry, pose and posture, sex acts.\n\nSubject 2 description.\n\nBackground description, indoor or outdoor, room, objects, end on white space\nI USED THE BLONDE SEED FOR SAMUS FOR BEST RESULTS!\n\nSample ORIGINAL ERROR positive prompt which still produced favorable results:\n\nmasterpiece,best quality,amazing quality,general,\n\nfrom below,cinematic angle,depth of field,soft lighting,\n\n1girl,\n\nsamus aran,blue eyes,looking away,blonde hair,ponytail,\n\ncombat stance,holding handgun,\n\ncave interior,stalactites,green glow,monster,metroid\n\n\nWhile making this guide, I attempted correcting this prompt to the following by injecting some \"metroid\" into the end of the prompt to hope that the AI would add something interesting or creative. Instead, it morphed Samus's chest to have her orange armor and glowing purple button. Not what I intended. I corrected it below with \"alien\" as a replacement tag! I ended up keeping these as examples for you to run yourself, to see which works better for you and your environment.\n\nSample CORRECTED Positive prompt with 4 batch perfect run (on my machine/environment):\n\nmasterpiece,best quality,amazing quality,general,\n\nfrom below,cinematic angle,depth of field,soft lighting,\n\n1girl,\n\nsamus aran,blue eyes,looking away,blonde hair,ponytail,\n\ncombat stance,holding handgun,\n\ncave interior,stalactites,green glow,monster,alien\n\n\n\n\nSample negative prompts:\n\nbad quality,worst quality,worst detail,sketch,censor,patreon,watermark\n\nbad quality,worst quality,worst detail,sketch,censor,patreon,watermark,motion lines,emphasis lines,speed lines\n\nbad quality,worst quality,worst detail,sketch,censor,patreon,watermark,bad anatomy,bad hands,bad feet,text,motion lines,emphasis lines,speed lines\n\nOR\n\nbad quality,worst quality,worst detail,sketch,censor,bad anatomy,bad hands,bad feet,patreon,watermark,text,motion lines,emphasis lines,speed lines"], "color": "#432", "bgcolor": "#653"}, {"id": 51, "type": "KSampler", "pos": [4842.480028046694, 1313.2039679016257], "size": [290.38520270836534, 280.5812991399544], "flags": {}, "order": 28, "mode": 0, "inputs": [{"name": "model", "type": "MODEL", "link": 73}, {"name": "positive", "type": "CONDITIONING", "link": 74}, {"name": "negative", "type": "CONDITIONING", "link": 75}, {"name": "latent_image", "type": "LATENT", "link": 76}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [90]}], "title": "KSampler (Refiner) 257243673939605", "properties": {"cnr_id": "comfy-core", "ver": "0.3.67", "Node name for S&R": "KSampler", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.2.2"}}, "widgets_values": [257243673939605, "fixed", 18, 6.8, "euler_ancestral", "karras", 0.35]}, {"id": 54, "type": "LatentPixelScale", "pos": [4545.043025672721, 962.7152301704981], "size": [276.0610310086057, 153.86325345215494], "flags": {}, "order": 26, "mode": 0, "inputs": [{"name": "samples", "type": "LATENT", "link": 79}, {"name": "vae", "type": "VAE", "link": 80}, {"name": "upscale_model_opt", "shape": 7, "type": "UPSCALE_MODEL", "link": 81}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [76]}, {"name": "IMAGE", "type": "IMAGE", "links": []}], "properties": {"cnr_id": "comfyui-impact-pack", "ver": "8.23.0", "Node name for S&R": "LatentPixelScale", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.2.2"}}, "widgets_values": ["area", 1, true]}, {"id": 57, "type": "CheckpointLoaderSimple", "pos": [4544.092702899628, 701.0575137714912], "size": [277.1173577747977, 104.96207382327907], "flags": {}, "order": 13, "mode": 0, "inputs": [], "outputs": [{"name": "MODEL", "type": "MODEL", "links": [82]}, {"name": "CLIP", "type": "CLIP", "links": [78]}, {"name": "VAE", "type": "VAE", "links": [80, 93, 94]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.71", "Node name for S&R": "CheckpointLoaderSimple", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": ["waiNSFWIllustrious_v150.safetensors"]}, {"id": 72, "type": "LatentPixelScale", "pos": [5796.850706368363, 1325.1148948384662], "size": [276.0610310086057, 153.86325345215494], "flags": {}, "order": 29, "mode": 0, "inputs": [{"name": "samples", "type": "LATENT", "link": 90}, {"name": "vae", "type": "VAE", "link": 93}, {"name": "upscale_model_opt", "shape": 7, "type": "UPSCALE_MODEL", "link": 95}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": [91]}, {"name": "IMAGE", "type": "IMAGE", "links": []}], "properties": {"cnr_id": "comfyui-impact-pack", "ver": "8.23.0", "Node name for S&R": "LatentPixelScale", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.2.2"}}, "widgets_values": ["lanczos", 1, true]}, {"id": 56, "type": "UpscaleModelLoader", "pos": [4257.551085712881, 744.7867109233347], "size": [276.01199179891955, 62.64613875373726], "flags": {}, "order": 14, "mode": 0, "inputs": [], "outputs": [{"name": "UPSCALE_MODEL", "type": "UPSCALE_MODEL", "links": [81, 96]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.43", "Node name for S&R": "UpscaleModelLoader", "ue_properties": {"version": "7.0.1", "widget_ue_connectable": {}}}, "widgets_values": ["4x-AnimeSharp.pth"]}, {"id": 74, "type": "UpscaleModelLoader", "pos": [6122.84647652635, 1529.6084876427992], "size": [270, 58], "flags": {}, "order": 15, "mode": 0, "inputs": [], "outputs": [{"name": "UPSCALE_MODEL", "type": "UPSCALE_MODEL", "links": [95]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.76", "Node name for S&R": "UpscaleModelLoader", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.2"}}, "widgets_values": ["RealESRGAN_x4plus_anime_6B.pth"]}, {"id": 73, "type": "LatentPixelScale", "pos": [6117.468352288572, 1324.0447227766035], "size": [276.0610310086057, 153.86325345215494], "flags": {}, "order": 30, "mode": 0, "inputs": [{"name": "samples", "type": "LATENT", "link": 91}, {"name": "vae", "type": "VAE", "link": 94}, {"name": "upscale_model_opt", "shape": 7, "type": "UPSCALE_MODEL", "link": 96}], "outputs": [{"name": "LATENT", "type": "LATENT", "links": []}, {"name": "IMAGE", "type": "IMAGE", "links": [97]}], "properties": {"cnr_id": "comfyui-impact-pack", "ver": "8.23.0", "Node name for S&R": "LatentPixelScale", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.2.2"}}, "widgets_values": ["area", 1, true]}, {"id": 60, "type": "MarkdownNote", "pos": [6086.032792346463, 1157.1409674931367], "size": [621.5911657517272, 107.35772800264783], "flags": {}, "order": 16, "mode": 0, "inputs": [], "outputs": [], "title": "Quick Refine and Noise Buffer \u2728", "properties": {"ue_properties": {"widget_ue_connectable": {}, "version": "7.5.1", "input_ue_unconnectable": {}}}, "widgets_values": ["Quick refine uses upscale models to tidy up your image! It adds about 5-10 seconds to compute. You can change this to strict latent upscaling by increasing the \"scale_factor\". 1.00 = none, 1.25 = 25 percent increase, 2.00 = 100 percent increase (from 1024 to 2048), and so forth. I would be careful you don't OOM your GPU!\n\nNoise Buffer\n\nDisable with CTRL+B\n\n0.01 or 0.02 for strength seems great!"], "color": "#432", "bgcolor": "#653"}, {"id": 58, "type": "ImageAddNoise", "pos": [6415.95695027078, 1343.3355356558104], "size": [276.0113028859332, 106], "flags": {}, "order": 31, "mode": 0, "inputs": [{"name": "image", "type": "IMAGE", "link": 97}], "outputs": [{"name": "IMAGE", "type": "IMAGE", "links": [84]}], "properties": {"cnr_id": "comfy-core", "ver": "0.3.75", "Node name for S&R": "ImageAddNoise", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": [311375798830678, "randomize", 0.01]}, {"id": 59, "type": "SaveImage", "pos": [6716.697002087594, 1345.0886479739436], "size": [762.6816306346636, 716.4261081619645], "flags": {}, "order": 32, "mode": 0, "inputs": [{"name": "images", "type": "IMAGE", "link": 84}], "outputs": [], "properties": {"cnr_id": "comfy-core", "ver": "0.3.71", "Node name for S&R": "SaveImage", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": ["STUDIO1911A2 Machine Gun Masterpiece by Freyja Pixel"]}, {"id": 52, "type": "CLIPTextEncode", "pos": [4548.588871420533, 1914.4644553260662], "size": [282.74921317847657, 360.5165503419086], "flags": {}, "order": 22, "mode": 0, "inputs": [{"name": "clip", "type": "CLIP", "link": 77}], "outputs": [{"name": "CONDITIONING", "type": "CONDITIONING", "links": [74, 86]}], "title": "CLIP Text Encode (Positive)", "properties": {"cnr_id": "comfy-core", "ver": "0.3.71", "Node name for S&R": "CLIPTextEncode", "ue_properties": {"widget_ue_connectable": {}, "input_ue_unconnectable": {}, "version": "7.5.1"}}, "widgets_values": ["son goku,super saiyan blue,energy,aura,space,stars,galaxy"]}], "links": [[64, 49, 0, 37, 0, "MODEL"], [65, 49, 1, 38, 0, "CLIP"], [66, 37, 0, 39, 0, "MODEL"], [67, 39, 0, 40, 0, "MODEL"], [68, 40, 0, 41, 0, "MODEL"], [69, 42, 0, 49, 0, "MODEL"], [70, 53, 0, 49, 1, "CLIP"], [73, 41, 0, 51, 0, "MODEL"], [74, 52, 0, 51, 1, "CONDITIONING"], [75, 38, 0, 51, 2, "CONDITIONING"], [76, 54, 0, 51, 3, "LATENT"], [77, 49, 1, 52, 0, "CLIP"], [78, 57, 1, 53, 0, "CLIP"], [79, 61, 0, 54, 0, "LATENT"], [80, 57, 2, 54, 1, "VAE"], [81, 56, 0, 54, 2, "UPSCALE_MODEL"], [82, 57, 0, 55, 0, "MODEL"], [84, 58, 0, 59, 0, "IMAGE"], [85, 55, 0, 61, 0, "MODEL"], [86, 52, 0, 61, 1, "CONDITIONING"], [87, 38, 0, 61, 2, "CONDITIONING"], [88, 43, 0, 61, 3, "LATENT"], [90, 51, 0, 72, 0, "LATENT"], [91, 72, 0, 73, 0, "LATENT"], [93, 57, 2, 72, 1, "VAE"], [94, 57, 2, 73, 1, "VAE"], [95, 74, 0, 72, 2, "UPSCALE_MODEL"], [96, 56, 0, 73, 2, "UPSCALE_MODEL"], [97, 73, 1, 58, 0, "IMAGE"]], "groups": [], "config": {}, "extra": {"ue_links": [], "ds": {"scale": 2.3939204936919185, "offset": [-4445.508450125455, -1834.9890710987904]}, "links_added_by_ue": [], "frontendVersion": "1.32.10", "VHS_latentpreview": false, "VHS_latentpreviewrate": 0, "VHS_MetadataImage": true, "VHS_KeepIntermediate": true, "workflowRendererVersion": "LG"}, "version": 0.4, "widget_idx_map": {"51": {"seed": 0, "sampler_name": 4, "scheduler": 5}, "58": {"seed": 0}, "61": {"seed": 0, "sampler_name": 4, "scheduler": 5}}, "seed_widgets": {"51": 0, "58": 0, "61": 0}}}

Freyja Pixel's Anime Studio