Cider Metrics for image captioning

I’ve figured out what the issue is. It seems that CiDEr can’t calculate for a single dataset; it requires a set of datasets to work with