A lora trained on many oil paintings, the style is modern impressionism so loose brush marks and thick paint. At high levels it will simplify dramatically at lower levels it will allow more detail, not trigger but "impressionz" was in every caption. Visible brush marks and paint textures help as well. Due to being in the UK I can no longer upload here so my models are here: https://huggingface.co/treeshark
Description
Ver 1 was quite abstract and simplified Ver 2 was a little more resolved and quieter in color, V 3 is somewhere between. It is better at fighting off the tendency in Flux for faces to revert to photos. No trigger but prompting , "facilz, oil painting on canvas, thick paint, visible brush marks and sketch" will make the lora react strongly as they were in every caption.