This is a VERY niche model that I trained with Dreambooth from SDv1.5 a long while ago. It is trained on images from the music video for the song Mariana Trench by the band Bright Eyes. The video has a unique watercolor hand-drawn style. Trigger word is "BEWatercolor". Generalizes pretty well. (I purchased a couple of these actual drawings from a promo the band was doing, I have them framed, I just like the look/style of it)