1.0 model, this one was trained on roughly 100 headshots. I am already in the works on including more body shots into the mix and finding clear headshots for a 2.0