Hello everyone. This is my first public Lora, so don't be too harsh :D
First of all, the Lora is not perfect. But I've also seen much worse than my Lora. It's best to try it out for yourself and then decide whether you like the lora or not. :)
_______________________________________________________________________________________________
The Lora was trained with a diverse set of 388 photos. All of them are ******* photos without post-processing, shot by myself with a “Canon EOS 1100D”.
the set contains:
Portraits
Landscapes
Family photos
Party photos
A few car photos
Animal photos
_______________________________________________________________________________________________ Recommended Settings:
Distilled CFG Scale:
between 2-3.5 (depending on your own taste)
Sampling method and Schedule type:
I personally use Euler/Simple.
You have to decide for yourself if other combinations deliver better results.
Steps:
30-50
Resolution:
I personally use 768x1024 / 1024x768. The Lora is trained in the same aspect ratio. You will have to test if other aspect ratios work.
Lora Weight:
I like the results between 0.7-1 the most. But again, it all depends on your own taste.
Trigger:
1100d Also works without trigger, but I like the result more with trigger.
Prompt:
Start with “1100d” and then simply enter your prompt. I tested it with short tags and with long natural language. Both gave me good results.
_______________________________________________________________________________________________
I would be very grateful if you could leave me your feedback here and post pictures with your results.