31st January 2023 Update:
djzAirlockV15 released!19th January 2023 Update:
djzAirlockV21-512-inpainting released!
you will need the 2-1-512-inpainting.yaml config, so grab that too!
2.1-768 model: Example merging Rick-Roll, JunglePunk & Airlock:
credit: RickInversion
These "strong style" Models are intended to be merged with each other and any model for Stable Diffusion 2.1
I recommend merging with 0.5 (50/50 blend) then using prompt weighting to control the Aesthetic gradient.
example merged model prompt with automatic1111:
(Airlock:1.2) (yourmodeltoken:0.8)
if you drop the "djz" and the "V21" what remains is the token you need to call up the concept in the model. All examples shown were the Raw Token, no other words. Tokens are case sensitive and in almost all models it will match the filename.
It is possible to merge these models with each other using a different value. It is possible to pair models and then merge those resulting models. In this way we can blend abstract concepts together and then weight the tokens to achieve the result we may wish to create. Of course to eliminate all those tokens, you can simply train a new custom model from the outputs, which means you are back to a single token.
A video explanation will follow, but for now the above explanation should do. We are focused on getting as many style/aesthetic models into artist hands to enhance the creativity already at their finger tips.
Art Freedom for all!!
[all original artwork used for training with full permission from Drift Johnson]
Description
ensure you take the correct config!
demo photos were random pictures i took with a camera.
you can do better easily!
FAQ
Comments (17)
this should just be an embedding
nope.
If you want to eat Steak with a spoon that's up to you, but we all use a fork with our Knives. Spoons are useful too just not for this
You gonna explain how to Inpaint with a textual inversion now, or just leave us all hanging?
@driftjohnson i don't know if you're joking or not, but you can use embeddings when inpainting, and it's a lot more useful and variable than this
@Aili Thanks, we use many Textual Inversions for both positive and negative embeddings, with multiple Hypernetworks, multiple Lora's and merged Checkpoints. Thanks for asking ;)
We don't advocate for one approach, we advocate for them all.... At this point we do not need to demonstrate the power of merged models or the strength and variability of "strong style models" thanks again for your continued support & enjoy the inpainting version !
also yes we like to make jokes - aha!
@Aili yes.
lmao at how this aged. get wrecked
@driftjohnson skill issue
@r3l4x whats your point?
@r3l4x i guess some people didn't understand how to use this resource, yes. That's ok, its all part of the journey.
@driftjohnson lmao. it was clear from the beginning. its funny as fuck to see the cope.
@r3l4x It's ok if you have no point. I don't judge people that failed to realize what these resources are for. I tried to explain it, everyone else gets it. Beyond endless demonstrations proving this, you are just trolling ;) You clearly didn't look for the LORA Version.... but that's ok.
@r3l4x stop coping. if you don't like this resource go find one that you do like. damn.
@KekwaeOdinson Im happy you recognize you were wrong from the beginning. Next time I hope you dont assume so much.
@driftjohnson Im happy you recognize you were wrong from the beginning. Next time I hope you dont assume so much.
@r3l4x I do not, but i do recognise you coming back two months later to make silly comments. The resource works fine exactly as it should, and LORA was not a thing when this came out. Get over it man, you just didn't understand what it is or what it's for. You can just not use it ;)