Open TruthSearchers opened 1 year ago
I like this idea. I don't quite understand all the specifics being asked for, but here's my idea: we can specify a base image that the bot considers its own selfie. Then, it can run img2img on that image with various parameters either built-in or in a user request, and give the result.
Alternatively, we could specify a folder full of images meant to represent the bot, and then when asked for a selfie it picks one of them and sends it. OR it could then do img2img on it. Or both, both could be options. 😄
Yeah img2img with a prompt let's suppose user asked for i want selfie with a car so it generate that character image with a car
Plus if user asked generate an image of Elon Musk (text to image ) , it generate image (just like now )
So both combined
Interesting. Something to keep in mind. What do folks think of using the bot's discord avatar as an image source? That way this feature could potentially generate images for other users as well.
If it's possible it will be cool
K, leaving this on the backlog for now but will keep it open for when I look at image gen more.
Interesting. Something to keep in mind. What do folks think of using the bot's discord avatar as an image source? That way this feature could potentially generate images for other users as well.
@chrisrude I like it!
Traditional text2img can work as well, especially if we can define a prompt template the bot should use for its own "selfies".
Interesting. Something to keep in mind. What do folks think of using the bot's discord avatar as an image source? That way this feature could potentially generate images for other users as well.
You could also have it "see" itself as a character that is trained in Lora (or something similar). For example, if you asked it to make an image of itself, it would put
Hey there is it possible to add a argument --image (image given locally) Like when we say generate a selfie or pic of yourself It generate it's multiple variant of image given locally through argument based on prompt on chat (selfie) plus also generate other images (i mean based on prompt just like now) Both combined
just like there is extension sd_api injection type in ooobaboga
It will be amazing