A PyTorch implementation of state of the art video captioning models from 2015-2019 on MSVD and MSRVTT datasets.