Trained V2 on images produced with v1
Added a keyword
Testing an amalgam of styles. Likely too broad and will need to be split