Use the Pixel_pony keyword to activate.

My Lora was designed to simplify the process of creating monochrome pixel art for minecraft, but my dataset was not perfect, so this version can be considered unsuccessful.

I keep it as a reminder to myself that there is something to grow towards.

The training was based on the "Super_PixelArt_XL_M_V1" model

My mistake was to trust the automatic detection of the dataset contents, so it's hard for me to say how predictable the model will be.