AI [AI Art] - Show Us Your AI Skill *NO TEENS*

4.80 star(s) 6 Votes

Schlongborn

Member
May 4, 2019
456
1,635
I wanted to reproduce the pictures from an earlier post I did in this thread from 5th January, and it proved to be almost impossible! So I just had to produce a bunch of similar images using the same prompt instead:
01644-1126304191-beautiful gorgeous full body 8k unity daz3d render, busty thicc ebony female,...png 01653-1636403152-beautiful gorgeous full body 8k unity daz3d render, busty thicc ebony female,...png

There were actually a bunch of changes to the webui project so that the model hashes changed, and also changes to how the samplers work. In the newest versions there are now settings to switch back to the old behaviour which helps, but still I just couldn't figure out what I was missing anymore.

I did find that reproduceability is not as easy that I thought in general, I tried to reproduce a picture Sepheyer posted here, but I just couldn't get the same result as him:
01821-1729759877-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png 01823-1729759877-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png

It looks very similar, but something must have still been different.

So, instead of trying to figure out why it didn't work, I decided my time is better spend on trying something else. So I trained a LORA with the pictures of that Daz character I already used as example in my first post here. Basicly I trained with the pictures of this little series:

And then I tried applying that LORA to Sepheyer's prompt:
01835-3592413857-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png 01839-3941786890-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png 01841-1378958709-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png 01842-1832049134-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png
01845-2200728019-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png 01856-258638541-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bok...png 01869-1884413959-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png 01872-1095882939-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bo...png

And I was actually really surprised how easy training a LORA was and how relatively well it worked. In particular the hair and the clothing was picked up (only the shirt, the skirt was barely present in the training data, still came out looking like I expected it though), even tough the training data wasn't really that comprehensive. Though I think the glasses were not a good thing to try and use for a first attempt at training a LORA. They caused ugly effects when applying the LORA with too much weight, and also it was kind of difficult to then make it create an image without glasses.

I also applied the LORA to pictures of the training data, with inpainting. It kind of worked as a way to recreate the same character, but with a certain style. More realistic, or more anime like, some before after pics:
020.png 01127-947022458-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bok...png
002.png 01131-7515208-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bokeh...png
01114-2132200728-busty naked dripping wet blonde girl with thick thighs standing at the beach,...png 01182-527197445-style ultrarealistic photography, 4k, 8k, 16k, 200mm lens, depth of field, bok...png
 
Last edited:

Mr-Fox

Well-Known Member
Jan 24, 2020
1,401
3,808
I wanted to reproduce the pictures from an earlier post I did in this thread from 5th January, and it proved to be almost impossible! So I just had to produce a bunch of similar images using the same prompt instead:
View attachment 2470090 View attachment 2470092

There were actually a bunch of changes to the webui project so that the model hashes changed, and also changes to how the samplers work. In the newest versions there are now settings to switch back to the old behaviour which helps, but still I just couldn't figure out what I was missing anymore.

I did find that reproduceability is not as easy that I thought in general, I tried to reproduce a picture Sepheyer posted here, but I just couldn't get the same result as him:
View attachment 2470097 View attachment 2470099

It looks very similar, but something must have still been different.

So, instead of trying to figure out why it didn't work, I decided my time is better spend on trying something else. So I trained a LORA with the pictures of that Daz character I already used as example in my first post here. Basicly I trained with the pictures of this little series:

And then I tried applying that LORA to Sepheyer's prompt:
View attachment 2470106 View attachment 2470107 View attachment 2470110 View attachment 2470111
View attachment 2470112 View attachment 2470113 View attachment 2470115 View attachment 2470117

And I was actually really surprised how easy training a LORA was and how relatively well it worked. In particular the hair and the clothing was picked up (only the shirt, the skirt was barely present in the training data, still came out looking like I expected it though), even tough the training data wasn't really that comprehensive. Though I think the glasses were not a good thing to try and use for a first attempt at training a LORA. They caused ugly effects when applying the LORA with too much weight, and also it was kind of difficult to then make it create an image without glasses.

I also applied the LORA to pictures of the training data, with inpainting. It kind of worked as a way to recreate the same character, but with a certain style. More realistic, or more anime like, some before after pics:
View attachment 2470131 View attachment 2470132
View attachment 2470134 View attachment 2470136
View attachment 2470138 View attachment 2470139
Awsome job. If you wouldn't mind. Could you make a post in the learning thread about making lora's?
I understand that you just tried it, but you did learn a lot I bet. Many people me included are guaranteed to be curious about trying it ourselves.
 

Sepheyer

Well-Known Member
Dec 21, 2020
1,611
3,863
This might be useful, or completely not. See how the girl on the left has grain in her face and hair? These are actually artefacts of the face restoration algorithm. There are two major algos for "face restore". I assume you are using WebUI by A1111, and they use CodeFormer as the built-in option. Lots of times it works well and seamlessly, tho sometimes it has these artefacts that one discards assuming this is grain and a pleasant addition to the image.

So, just so the folks know there is an alternative algo, GFPGAN. And here is how these algos compare: link.

A1111 has an option to use GFPGAN - this can get rid of that "staircase / grain" artefact in the woman's hair.

This is what GodOfPandas's image looks like with artifacts removed by GFPGAN:

00052rs.png

"How to" here: link.
 
Last edited:
4.80 star(s) 6 Votes