(This template is used for Workflow Contest)

What this workflow does

👉 [Please add here]

Simple Workflow using text to image, ipadpter and image to image tha creates very interesting results :)

works with SD1.5 and SDXL models.

How to use this workflow

👉 [Please add here]

the values that have most impact on the result are the weaight, noise and weight type of the ipadapter,

the "start_at" and "end_at"valus in the apply ipadater node determine at what point during the redner the ipapater influence starts, and or ends. do for example "start_at"  0.250 means the ipadapter 2kicks"in adter 25% of the render steps are done, and "end_at" 0.600 would mean the influence of the ipadter on the rendering stops after 60% of the render steps are done.

the denoise strength on the img2img

and ofcourse the prompt ( positive prompt has a wildcard picker, so you can even have a little random in there too if you like)

:D

oh. and you have to enable the populate checkbox on the wildcard prompt, otherwise it always uses the same prompt node, for some reason thats always off by default

Tips about this workflow

just load some image and enter a prompt, queueprompt and see what happens. :D

for exampe if you load an image of a mouse into the ipadpter, a picture of a hambuger in the img2img and then use the prompt "realistic photograph of a cute grinning little monster" you get a little grinning hamburger monster :D

🎥 Video demo link (optional)

👉 [Please add here]