🌞 Requests
You can request a commission on my Ko-Fi
🗃 Info & Tips
Models are trained on Pony Realism but should work with other Pony models
The model datasets are from 3DGI, so their style leans towards that. Use a weight of 0.6 - 0.7; the higher the weight, the more 3D the style will be.
Experiment with samplers, as some provide more realism than others.
Read each model Trigger words on the Model Details.
Recommended the use of positive prompt of: realistic, photo, real hair, detailed skin.
Recommended the use of negative prompt of: illustration, cartoon, anime, comic, 3d, cgi.
🏋️♀️ Training your own character
Want to train your own character? Read the following article for tips and my training preset
📜 Index
More to come....
⚡️ Buzz for the Best Images ⚡️
Description
FAQ
Comments (7)
Wow this looks like the best Alyx Lora yet..... I donated and can't wait to try it out
For whatever reason it's making me subscribe to civit or w/e to download
This is the first time I've seen this
Is this intentional? It's weird to paywall stuff that is AI generated.
@ZyloO what does that mean in this context? Are you still working on the LoRA? what changes are you planning?
"stuff that is AI generated" 😃😂 this "stuffs" are created by humans! some of them require a lot of effort and time to create. Consumers who only want to consume without giving anything in return are terrible.
Buzz. The system has been up on Civitai for quite a while already...
@PetroUA It must take you a lot of effort and time because you've only ever released 2 LoRAs. I can shit out a character LoRA in 30 minutes, and most of that time is dataset preparation.
And to be clear, dataset preparation means finding images that other people have made.