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

[hu lover]

ok next question is....
How do you "train a Lora"?

I just googled it for you. Here is a tutorial explanation below:

How to train Lora models

1. Train a Lora model. Step 1: Collect training images. Step 2: Upload images to Google Drive. Step 3: Create captions. Running the LoRA trainer. Image path. Other settings. Start auto-captioning. Revising the captions. Step 4: LoRA training. Source model. Folders. Parameters. ...
2. Using the LoRA.
3. Remarks.
4. Reference.

Here is a full tutorial that goes more in depth.

You must be registered to see the links

 

[felldude]

Text to image with 2 custom Lora's I've been working on. (768x768)
Then a upscale with details in the prompt.

Regarding lora's I train locally on my own machine (With the latest BF16 I can even train XL SD)
with a tool called

You must be registered to see the links

 

[PandaRepublic]

ok next question is....
How do you "train a Lora"?

Here is a good training guide on everything about training LoRA's

You must be registered to see the links

 

[Sepheyer]

ok next question is....
How do you "train a Lora"?

There's an AI prompt learning thread here as well: link.

 

[felldude]

Still trying to determine how much skin detail is to much for your brain with the Pixar style

 

[Rumbrandt]

A cowgirl at the bar

 

[FoolsBeGone]

In case you couldn't tell I love ocean sunset scenes lol

 

[thelastof9]

A cowgirl at the bar

View attachment 2914756

What a nice setting and pretty shirt.... I mean hat, very nice hat

 

[FoolsBeGone]

Ok I am going to post a picture and include below it in a spoiler all the prompt info, what I wrote and what was supplied by the system

Eowyn before the battle

Spoiler: Prompts

You don't have permission to view the spoiler content. Log in or register now.

 

[FoolsBeGone]

Ok one more time with the Ring-wraith/Nazgul

 

[FoolsBeGone]

These are the incarnations that led to the final Nazgul picture. Each had something wrong with it...the first two were missing the rider, the third was missing the moon and I didn't like the way the pants looked, the fourth one you could see that the rider was white from just above the face mask which ringwraiths are not. They are black as the night because they have no actual substance so that one was also a no go.

 

[Sharinel]

These are the incarnations that led to the final Nazgul picture. Each had something wrong with it...the first two were missing the rider

They had just taken off their capes, so were invisible. Someone has never done a project breakdown to management, always keep it positive

 

[FoolsBeGone]

Epic Eowyn

View attachment deacd34b9a5dac6f406844aa3f3f5a73.webp View attachment e938bc55bbbe2dabbfb9c69ea30ba7e0.webp

 

[FoolsBeGone]

and now for Gandalf- the program took his sword(Glamdring)and his staff and made them into one item but other than that it looks pretty good

 

[Artiour]

<lorahoto real:0.800000>
Model: Brookers Stlye XL

lora with XL model!? I didn't know they created LoRAs for XL yet
I need to go out there and update myself

 

[FoolsBeGone]

I couldn't get the faces to come out right but I suspect that is an issue with the Lora(?)/Checkpoint.

 

[FoolsBeGone]

Gandalf the White

 

[FoolsBeGone]

Spoiler: Prompts

You don't have permission to view the spoiler content. Log in or register now.

 

[FoolsBeGone]

 

[felldude]

lora with XL model!? I didn't know they created LoRAs for XL yet
I need to go out there and update myself

Yeah they had a contest and everything. The checkpoints and lora's are getting better but we still can't train the dual text encoder for shit.