Preface

After my last computer crash I have to sort my pictures again. I found a few photos of a cat that I rescued from the street years ago. Since I wanted to see how I could use cat photos for a LoRA model anyway, I used these photos to create this model.

What the Lora Does

The LoRA is able to create a photo realistic red cat.

Trigger Words

The trigger words for this LoRA are

red cat

Training Images

The training image have a resolution of 1024 x 1536 pixel. The related aspect ratio is 3:2. I downscaled them, so that one side should have a length of 1024 pixel.

I used in total 19 of my best images.

Training settings was

• 19 Images

• 38 buckets

• 2 repeats

• Batch size 2

• 250 epochs

• 2500 learning steps

Quality of The Created Images

My first tests shows that the quality of the created images are depending on the model. Some models are better with respect to right anatomy than others.

Final Words

Have fun. Be inspired!

Donation

I am happy about each Buzz!

I am also happy to accept donations in one of 250+ crypto currencies:

https://nowpayments.io/donation/zentrocdot

Leave an comment when doing so that I know in what you are interested in.