So, after some more tests, i figured out Lianima style is a little too strong to work correctly with some style LoRAs. And instead of publishing a full V2 for what is basically just a patch, here is a LoRA in its original usage: publishing just enough weight for correction without releasing and making people download a full V2 ^^;

Usage: Just shove this LoRA at strength 1.0~1.2 on top of Lianima before applying a style LoRA (a little stronger too if needed) to get a result closer to the intended look. Main goal is to "flatten" a bit the initial model and restore just a little bit some modified keys in the llm adapter. This should not Liadora from the model knowledge in any case 😉

Legal

The CircleStone Model is licensed by CircleStone Labs LLC under the CircleStone Non-Commercial License. Copyright CircleStone Labs LLC. IN NO EVENT SHALL CIRCLESTONE LABS LLC BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH USE OF THIS MODEL.