v2:REBIRTH

theres no new one besides v1.1, the one uploaded newer than that is 1.1-inpainting which isnt for generating images but for fixing images in inpainting

You asked this before. Again it’s small as it can be

Too many models published here, to remember all interactions. No problem.

just do some inpainting afterwards

@cromoose I know what CLIP is, but what is CLIP 2? Is it just openCLIP? Which openCLIP?

@haodocowsfly clip skip 2...

I don't understand it either. Everywhere they write "Use CLIP skip: 2", but no one writes where it should be used. In the Easy Diffusion v2.5 settings, I did not find these settings anywhere. It should be written in a text prompt???

@3dzver207
What: https://www.youtube.com/watch?v=Ff9NkfyV2_0
How: https://www.youtube.com/watch?v=UKHpGHepq4A

https://www.youtube.com/watch?v=IkMIoRCfCgE

/Settings/Stable Diffusion, bottom of page

@Relick There are no such settings in ED 2.5

@nfjdsngjf696 There is nothing in these videos about the settings in Easy Diffusion 2.5 

?? this model is 2 weeks old. Its new

@s6yx oh wtf my mistake, my bad... i was looking at the v1.0 didn't see the 1.1 was just released recently

@honfoetrash793 if it's good why new version? If it's not broken don't fix it
jk of course

@Rowlinson it's just that it's so good that recently i even started merging other models and using Rev as a base and it game me some solid results, it's basically one of the top models for me, but yeah lol v1.1 was big fix for a lot of the stuff that felt weird at 1.0 for me, if ever there's a new version or whatever it'd be probably fucking gold.
thx o7

@honfoe  I ask him on reddit he said this is not yet finish, it was release early, so yup there will be more.

its the current model available (1.1). Just the naming scheme i use when testing

This is a fairly good tutorial.
https://stable-diffusion-art.com/how-to-use-vae/
(They have many other tutorials there as well..)
But basically put VAE here: stable-diffusion-webui/models/VAE
And then:
To use a VAE in AUTOMATIC1111 GUI, go to the Settings tab and find a section called SD VAE (Use Ctrl+F if you cannot find it). In the dropdown menu, select the VAE file you want to use.

https://huggingface.co/stabilityai/sd-vae-ft-mse-original/tree/main

Many seem to be using the ki-f8-anime2 VAE
https://huggingface.co/hakurei/waifu-diffusion-v1-4/blob/main/vae/kl-f8-anime2.ckpt

pickletensor

@s6yx will there be a pruned version ?

@Sakura96 ?? this is already the lowest it can be

try removing --no-half and --disable-nan-check

@s6yx unfortunately I just get an error that says
'NansException: A tensor with all NaNs was produced in Unet. This could be either because there's not enough precision to represent the picture, or because your video card does not support half type. Try setting the "Upcast cross attention layer to float32" option in Settings > Stable Diffusion or using the --no-half commandline argument to fix this. Use --disable-nan-check commandline argument to disable this check.'

which would just put me back at square 1. I appreciate the suggestion though

1. bad-artist-anime is a embedding which seems to be taken down
2. I use this embeddding no need to link it to me since i use it on most of my generations

@s6yx 

1) Unfortuntate :(

2) I added the link for other peeps, since they might not know where to get it :)

BTW, thanks for the reply, and also thanks for the great model!

https://huggingface.co/nick-x-hacker/bad-artist/tree/main bad-artist and bad-artist-anime are both here. well, that's where i got them anyway lol

second this, the model looks great but its a merge, and the instructions for the VAE are not clear, lazy.

suggestions havent changed from 1.0 version

@s6yx what VAE to use? You did not write it anywhere

@civitai_bot2 yes i did, check 1.0 model tag and youll see it there

try Dreambooth: https://github.com/XavierXiao/Dreambooth-Stable-Diffusion
here's a tutorial:
https://www.youtube.com/watch?v=9Nu5tUl2zQw

@kyactive thanks，Let me try

youtube can help you with that

corona

@1234popo405 If you want to help me, I will thank you, maybe some people will not, but you can't say that China is all like this

@woncho thanks,I found some tutorials and am trying

Your comment was the best reason I got to download this model, thank you.

ok

go to model v1.0 description and you will see...

install it and try it out

@s6yx It's tough to get non humans right. But i'm using it for sure for my human and humanoid NPCs. No better model for that that is for sure

Ah, I was just about to ask! Do I need those VAEs for good image generation, or are you simply informing that you used them in the training process, and this checkpoint doesn't need to be paired with any VAE?

@TernaryM01 the former

I think you may have been overly irritated when you did this lol, I don't think blessed2.vae.pt is a valid URL. Hoping that it saves you clicks- it's NoCrypt/blessed_vae · Hugging Face

@mindframe lol oh you could tell. thanks for headsup, fixed in the main info section

Sorry, got a quick chuckle from this...Youre an absolute legend fam, dont bother with the VAE orcs.
Again, you got all my admiration and respect for these great models. (made a meme on civitai discord memes topic inspired from your comment)

@s6yx also seems like orangemix.vae.pt link is invalid too

@UselessMeaning thank you, fixed

you did not read the model description. I liked the VAE.....