--English
CoffeeBreak is a checkpoint merge model.
This is just a merge of the following two checkpoints.
CoffeeNSFW https://civarchive.com/models/43784/coffeensfw
BreakDomain https://civarchive.com/models/50520/breakdomain
I figured the CoffeeNSFW character art and the BreakDomain background art would suffice for most things. It was exactly that.
I thought it might not be necessary to publish it because it was too simple, but I have posted many images created with this model, so I thought it would be better to publish it.
I will delete it if it seems that the reputation is bad after seeing everyone's reaction.
Baked VAE: kl-f8-anime2
Ultimate SD scale and 4x-AnimeSharp are used for the sample.
--日本語
CoffeeBreakはチェックポイントのマージモデルです。
これは次の2つのチェックポイントをマージしているだけのものです。
CoffeeNSFW https://civarchive.com/models/43784/coffeensfw
BreakDomain https://civarchive.com/models/50520/breakdomain
私はCoffeNSFWのキャラクターイラストと、BreakDomainの背景イラストがあれば大抵の事が足りると考えました。実際にその通りだったと思います。
シンプルすぎて公開しなくても良いかもと思いましたが、私がこのモデルで作成した画像を沢山投稿しているので公開した方が良いと考えました。
皆さんの反応を見て評判が悪いようでしたら削除します。
VAE : kl-f8-anime2 を焼いてます。
サンプルにはUltimate SD scale、および4x-AnimeSharpを使用しています。
Description
FAQ
Comments (8)
lol, i was so excited to try this, waiting for the early access was a nightmare, i'm gonna have fun using it. thanks! i'll leave a review later :D
Thanks for waiting! Looking forward to your review.
相同的参数，相同的embedding，相同的lora，生成出来的图和你的示例图总是有30%的差异，不知道是什么原因
I can't find out what the cause is based on that information alone. Is it the same for SEED?
I apologize if you didn't seem to ask for help.
maybe because xformers enabled or not
You're being robbed, this is a top tier model here. I kept seeing those cool retro anime lora examples and I had to know the source. I've been at it all day! Thank you!
Could you add your model to https://tensor.art/ plis.
Some of the images exhibit a retro anime aesthetic, with sort of toned down washed away paint color composition and a blurred cel anime feel, while others adopt a crystal-clear 2020's digital gacha design and art style. However, upon closer examination, they don't appear to have additional LoRA or prompts to establish the atmosphere. How are these effects achieved? Can someone provide an explanation? The only logical inference is that it is ingrained in each character's LoRA. Is there a method to recreate this?