In table 8 of the paper you describe that the embedding size for YAGO is set to 300.
When I want to reproduce these results:
Do I have to set also the hyperparameters initsize and hidsize to 200, or only set the parameter embsize?
parser.add_argument('--initsize', type=int, default=200, help='size of initial representation dimension')
parser.add_argument('--embsize', type=int, default=200, help='size of output embeddings')
parser.add_argument('--hidsize', type=int, default= 200, help='size of representation dimension in the core function')