During the first iteration of the workshop at ECCV 2018, we organized a challenge on a large scale dataset of over 300K high quality fashion images. The dataset consists of multiple angles of each product against a standardized background, worn by a model under consistent lighting conditions, along with text descriptions of the products. The competition was on generating images conditioned on the text descriptions. The quality of the generated images was evaluated both by a pre-trained network (inception score) and by human evaluators.