HuanCai||Color Fantasy (Auxiliary Fine-Tuning Model)

Welcome to useColorful Lora ModelThis modelsupports art stylemicro-adjustmentand is compatible&nbsp;&nbsp;2D, 2.5D, realistic models; can be used independently as character Lora-------------------------------------------------------------------------------------------------------------------------------Recommendations for Use:Recommended to use in combination with anime or realistic models compatible with 2.5D modelsWhen combined with other Lora suggestions, weight of this model should be 0.8-1; other Lora weight 0.6-0.8If using the ADetailer plugin, make sure to enable it when exporting imagesModel adaptation to base model conditions and sampler compatibility chart<img src="https://liblibai-online.liblib.cloud/web/image/bba723ac80733437f9232c90ffe6bb26f1fb1696fb1fea1e80ce99397c28724e.png?x-oss-process=image/resize,w_1290,m_lfit/format,webp" alt="web/image/bba723ac80733437f9232c90ffe6bb26f1fb1696fb1fea1e80ce99397c28724e.png" data-href="https://liblibai-online.liblib.cloud/web/image/bba723ac80733437f9232c90ffe6bb26f1fb1696fb1fea1e80ce99397c28724e.png?x-oss-process=image/resize,w_1290,m_lfit/format,webp" style="">As shown in the chart, it is not recommended to use the DPM++ 2M SDE samplerThe cover image is based on MiaoKa_reality【Fusion】drawn-------------------------------------------------------------------------------------------------------------------------------Recommended starting prompts: best quality, masterpiece, realistic, HDR, UHD, 8K, masterpiece, high quality, exquisite wallpaper, 16K, animation, illustration, (white_skin),Suggested image parametersSampling method (Sampler):Euler a: 25-30 stepsDPM++ SDE Karras: 20-30 stepsDPM++ 2M Karras: 20-40 stepsPrompt word guide factor (CFG Scale): 7Character resolution: 512x768, 768x768, 512x1024Background and landscape resolution: 768x512, 1024x512, 1536x512High-definition repair: R-ESRGAN 4x, high-end iteration steps 15-30, redraw range 0.3-0.5Clip Skip: 2Negative prompt words:Recommended to use EasyNegative in the embedded embeddings model, negative_hand-negTwo sets of additional negative prompt words:Set 1: (worst quality, low quality: 2), monochrome, zombie, (interlocked fingers: 1.2),Set 2: NSFW, logo, text, blurry, low quality, bad anatomy, lowres, worst quality, watermark, bad proportions, username, EasyNegative,-------------------------------------------------------------------------------------------------------------------------------Model feedback AI technology exchange QQ group: 377209248This model does not include any real-life image sources or training code from the internet. The model is only for research and interest exchange

Files with SHA256: 892dec0abf00741e371932006601bac4ad4515b5cf5cbeb6852eba4fbbe8ba4d