![746132387628525820](https://image.tensorartassets.com/cdn-cgi/image/anim=false,plain=false,w=640,q=85/workflow_template_showcase/746132387628525820/874da606-71e6-b1a7-45a1-6c09949c0f87.png)
[P0624 06] “Eris to anything” (v3.1)
IPAdapter models are very powerful for image-to-image conditioning, enabling the easy transfer of the subject or style of reference images onto new creations. You need two images for referencing, one for posing and the other for copy its style; this is a basic (below 100 nodes), but complete version of the “style transfer workflow” for local comfyui. It includes: · Functional two control tabs for jumping between three IPAdapter boxes without connect or disconnect anything and enable or disable watermark boxes.To show the potential of this workflow I added three presets (pre-loaded) on it, so you can generate for free samples to see the potential of this workflow in action; in the SAMPLE example you’ll fast generate something into a base “Retro Goth illustration (prompt style)”.Instructions to generate a sample image: 1. Go to [] 2. Upload any image for reference. 3. Upload a ControlNet image [R 2:3][512x768] (You can get some free ControlNet images here [https://tensor.art/articles/732001331050630805]. 4. Enable one style and disable the rest (linear, style transfer or composition). 5. Select a seed number or pick it randomly. 6. Introduce your nick or social link. 7. Click on generateCredits Model A: Dark Fantasy V1 Model B: Dark Age LoRA [1]: Classic Oil Painting Style LoRA [2]: OC Eris Etolia LoRA [3]: Castlevania Style v1.0b⚠️ Models and LoRA used in the SAMPLE version generation are not included.Useful Documentation · Standard (medium strength): This model achieves medium-intensity modifications, suited for a broad array of applications, balancing subtle and noticeable alterations. · CLIP-ViT-H-14-laion2B-s32B-b79K.safetensors: This model variant is a part of the CLIP (Contrastive Language–Image Pre-training) family, specifically designed to understand and interpret visual content in relation to textual information. · Linear: Applies weight evenly from the start to the end of the processing pipeline, ensuring a consistent influence throughout without modulation in intensity. · Style Transfer: This option focuses on adapting the style from the reference image to the target image. · Composition: It aims to merge the compositional elements of the reference and target images, emphasizing the arrangement and interaction of elements within the space.
![PictureT](https://image.tensorartassets.com/cdn-cgi/image/anim=false,plain=false,w=24,q=85/users/common/avatar/619049867326525670/691d5c15-cd03-d607-6158-90fa30456e73.png)
PictureT
![Daily Challenge](https://image.tensorartassets.com/cdn-cgi/image/anim=true,plain=false,w=14,q=85/badges/e13137b7-b048-4d79-8c98-e2bb580be0ac.jpeg)