hello, recently been getting into stable diffusion. I already have the basics down such as text2img, and some img2img. However, for the life of me i cannot seem to get neither the text2img nor the img2img to work correctly for pregnancy. What loras do yall use? what models as well? and how can one effectively make someone pregnant with img2img? thanks!

I've used RealsticVision as a base model which worked pretty good for clothed pregnancy.

You can use controlnet with OpenPose for better posing. Also the mode "reference only" worked quite well to get alternative images.

Make sure your prompts contain model specifc keywords. For realistic vision those positive and negative prompts you need to add are written on the civitai page.

(August 5, 2023, 4:48 pm)peterwein I've used RealsticVision as a base model which worked pretty good for clothed pregnancy.

You can use controlnet with OpenPose for better posing. Also the mode "reference only" worked quite well to get alternative images.

Make sure your prompts contain model specifc keywords. For realistic vision those positive and negative prompts you need to add are written on the civitai page.

thanks for the help man. I understood mostly everything but would like some clarification on your second sentence, "You can use controlnet with OpenPose for better posing. Also the mode "reference only" worked quite well to get alternative images. where can i find these options?

controlnet is an extension for the automatic1111 UI that you are probably using. Openpose and referrence only are models within of controlnet. Just have a look at YouTube