YiffyMix

Use e621 tags (no underscore), Artist tag very effective in YiffyMix.Species/Artists grid list (update v51) & Furry LoRAs (update SDXL)/samples/wildcardsRecommend Artist List for new user & comfyUI workflowYiffyMix v5x & v6x SDXL settingYiffyMix v5x doesn't need PonyXL score tags.Don't use SD1.5 vae in SDXL, you will get fried image.steps=12~24, cfg=4, sampler="Euler A SGMUniform","DDPM","Euler A Turbo"YiffyMix v4x V-pred Setting# YiffyMix v4x A1111 SD-WebUI or Forge V-pred Setting1.Download config file ".yaml" paste next to the model.2.Raname ".yaml" same as model name. ( check the ".yaml" has [parameterization: "v"] line )3.Restart SD-WebUI ( if config fails to load, the model just generate noise )# YiffyMix v4x ComfyUI or StableSwarmUI V-pred SettingAdd node "ModelSamplingDiscrete", Sampling = "v_prediction"workflow demo# ComfyUI V-pred Setting (old version)1.Copy config ".yaml" to [ComfyUI\models\configs] and refresh ComfyUI2."Load Checkpoint (With Config)" [Right Click\AddNode\advanced\model_merging]workflow:[Load Checkpoint (With Config)]-[KSampler]-[VAE Decode]-[Save Image]# v-pred mode troubleshootIf you use new version webui-forge and failed detect v-pred model.Try update webui-forge (click update.bat).When loading v-pred model, you will see this line in cmd console:left over keys: dict_keys(['v_pred'])# v4.x some time will get little fried issue image in this caseUse "Dynamic Prompts with wildcards prompts" and "batch size > 1"Just generate again with same prompts with parament, it will return normal.Example SettingSetting SDXL-lightningSteps = 12~24Sampler = "DDPM", "Euler A SGMUniform", "Euler SGMUniform"CFG scale = 4~6Negative embeddings SDXL = ac-neg1, ac-neg2 (You don't really need this)Postive LoRA SDXL = SeaArt Quality Tags LoRAStop at CLIP layers = 2Setting SD 1.5 + vpredSteps = 30~40Sampler = "DPM++ 2M Karras", "DPM++ SDE Karras", "DDIM", "UniPC"CFG scale = 6~8Negative embeddings = deformity_v6, bwu [SD-WebUI\embeddings]Stop at CLIP layers = 1Eta noise seed delta = 0SD VAE = kl-f8-anime2, Furception [SD-WebUI\models\VAE]Hires. fixHires steps = Steps * Denoising strengthDenoising strength = 0.25Hires upscaler = 4x-UltraMix_Smooth [SD-WebUI\models\ESRGAN]Img2img First Pass ( Upscale enhance )Scale to = 1.5Denoising strength = 0.4~0.45Extra ( seed ) = trueVariation seed = -1 ( use same seed )Variation strength = 0.9Img2img Second Pass ( Detail enhance )Scale to = 1Denoising strength = 0.3~0.35Extra ( seed ) = trueVariation seed = -1 ( use same seed )Variation strength = 0.95ControlNetControlNet = softedge_hed, control_v11p_sd15_softedgeControlNet SDXL = softedge_hed, controlnet-sd-xl-1.0-softedge-dexinedControlNet Weight = 0.35~0.5ControlNet Pixel Perfect = trueRecommend SD WebUIreForgeForgeA1111LoRA Trainingimgs count = 15~50total steps = epoch * imgs count * folder loop = 3000~4500network_dim = 64network_alpha = 128learning_rate = 0.0002unet_lr = 0.0001text_encoder_lr = 0.00005lr_scheduler = "cosine_with_restarts"mixed_precision = "bf16"optimizer_type = "Adafactor"optimizer_args = [ "scale_parameter=False", "relative_step=False", "warmup_init=False", ]Version Infov1.x [2D,512~768] old model, unstable, low resolutionv2.x [2D,512~896] e621 dataset + 30~40% SD5 datasetv3.0 [2D,3D,512~1088] large dataset then v2.x version, can do 2D, 3Dv3.1 [2D,3D,Real,512~1088]more realistic, but loss some concept (e621 tag count lower 1000)v3.2 [2D,3D,Real,512~1088]unstable version, use SNR version FluffyRock, more noise detailv3.3 [2D,3D,Real,512~1088]stable version, more detail, more sensitive to promptsv3.4 [2D,3D,Real,512~1088] ※include Fluffy Rock Quality Tags-LoRAstable version, more detail, clear result, reduce some noise (like bush,pattern)v3.5 [2D,3D,Real,512~1088]unstable version, contrast & detailed enhancev3.6 [2D,3D,Real,512~1088]stable version, more sensitive to e621 tagsv3.7 [2D,3D,Realistic,res:512~1088,VAE Baked]stable version, reduce noise overactivity, fix blurry eyes and finger error, fix failed background depthv4.0 [2D,3D,Real,512~1088,VAE Baked]v-pred version, remix from EasyFluffaccuracy anatomy and style with fewer prompt,little dim blue average color with smooth noise, weakly negative prompt issuev4.1 [2D,3D,Real,512~1088,VAE Baked]v-pred version, color and low contrast fix, reduced realistic noise.v4.2 [2D,3D,Real,512~1088,VAE Baked]v-pred version, more contrast and detail, yelow issue.if you feel too fried use Rescale CFG =0.35v4.3 [2D,3D,Real,512~1088,VAE Baked]v-pred version, clear artist style, fix most yellow/brown issuethe version doesn't need CFG Rescalev4.4 [2D,3D,Real,512~1088,VAE Baked]v-pred version, no more yellow/brown issue.the version doesn't need CFG Rescalev5.0 [2D,3D,Real,896~1536]SDXL-lightning version, base on Compassmix XL.v5.1 [2D,3D,Real,896~1536]SDXL-lightning version, more e621 dataset, lower human face, better NSFW stuff.v5.2 [2D,3D,Realistic,res:896~1536]SDXL-lightning version, increase average quality. Little unstable than v51 but more creative.v6.0 [2D,3D,Realistic,res:896~1536]NoobXL version.Basic StyleNegative Prompt SD1.5unusual anatomy, mutilated, malformed, watermark,amputee, mosaic censorship, sketch, monochromeNegative Prompt SDXLmalformed, worst quality, bad quality,signature, text, url3D Artwork StylePrompt:blender (software), ray tracing, 3d, unreal enginePhotorealistic StylePrompt:[:photorealistic, analog style, realistic, photorealism:0.5]Negative prompt:[:bwu:0.5]Recommend Refiner:IndigoFurryMix v110-Realistic Switch at 0.5Prompt SDXL:by Mandy Disher, by Wim Wenders, by Robert Rauschenberg, [by Fossa666::0.65],ultra realistic, photorealism, photograph,Negative prompt SDXL:sketch, manga, vector, line art, toonySome SD tricks using a1111 [a:b:value] script prompt:1.High details increase[:some style or artist:0.15]Let first 10%~20% steps free, then style will create on high detail base.2.Avoid character pose affected by artist or background[:artist or background prompt:0.2]Skipping first 20% steps can avoid this.3."A artist character" + "B artist style"[A artist:B artist:0.35]Most base concept is done in first 30%~50% steps, last steps only change style.