Type | |
Stats | 46 192 39 |
Reviews | (5) |
Published | Oct 6, 2024 |
Base Model | |
Training | Steps: 1,100 Epochs: 15 |
Usage Tips | Clip Skip: 2 |
Trigger Words | Drestoria Style |
Hash | AutoV2 2121C1C762 |
I wanted a slightly softer looking LoRA, however this turned out to do slight semi-realism sometimes probably because I trained a base LoRA and then merged it slightly with others, to get the intended soft-look, I wanted for version 1. So... stuff was attempted. Version 2 and 3 was trained on my ExoCore XL/Pony merge model on civit.
Side note, I suggest anything above or at 4 CFG, when using this LoRA. This also works best with my merge model, which makes sense the base was trained on it. Also feel free to try anything with this, just know some LoRAs and models may not work with this, that well, tested that out myself.
Don't forget to publish images made by this if you want. I like seeing people's images made with my LoRAs and stuff.
The base trigger word is the "Drestoria-Style_V1" one, you don't need "semi-realism, nose", unless it's trying to produce semi-realism.
For version 2 the base trigger is "Drestoria-Style_V2" good luck on generating images with this, I hope to see your creations made with this LoRA.
Version 3 has the trigger "Drestoria Style" doesn't really require it but, the trigger word sort of acts like a fix for the eyes and nose, so it's still recommended.
Certain prompting will cause a overbaked look, so be wary as the LoRA is hard to control.
⚠️ IF YOU'RE GOING TO STEAL MY MODELS, MAKE IT SO THE DOWNLOAD IS ENABLED. ⚠️