ClashSAN

You will need 8 gb to train/make this I believe, it is a similar requirement to textual inversion. I don't know how it plays out if you're using smaller images,...

I've been wondering about this, the resulting image would have a load of metadata, but at least these variant images would be more reproducible than inpainting or color-sketch images.

>inpainting - only one color of the brush, no zoom. It is not possible to cancel everything and return to the original image. Color sketch for inpainting is already a...

The commit adding this: https://github.com/AUTOMATIC1111/stable-diffusion-webui/commit/c26732fbee2a57e621ac22bf70decf7496daa4cd Auto's comment: https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/1695

>Leaving it as is causes a crash when DDIM sampler is selected. https://github.com/AUTOMATIC1111/stable-diffusion-webui/commit/c26732fbee2a57e621ac22bf70decf7496daa4cd In line 115 "composition via AND is not supported for DDIM/PLMS samplers" >negative prompts are the solution...

Please decide which is more complete, thank you https://github.com/AUTOMATIC1111/stable-diffusion-webui/pull/3755 already open

**EDIT: my bad this is `Masked content: Fill - new color and opacity options`** nice it will fullfill https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/1783 and https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/3337 It was previously like that @ntrpnr https://github.com/AUTOMATIC1111/stable-diffusion-webui/issues/1362 >It used...

The tool would be much more powerful if it could be used in inpainting, as you can just paint a character, switch to inpaint brush and mask that section alone,...

try running without medvram. What is your GPU?

we have similar issue open here: #914