Downblouse Video

This is the sixteenth epoch of a series with a combined video and image dataset. Seems to give best output at guidance 8. This was the last epoch that produced good sample images during training; after this, coherency began to degrade.