Evaluation of ConvE model is poor

[chehefei]

Hi. When I run the ConvE model (on the FB15K dataset), the evaluation is poor as follows:

$ pykg2vec-train -exp True -mn ConvE -ds fb15k -hpf ./custom_hp.yaml

--# of entities, # of relations: 14951, 1345 --mr, filtered mr : 7280.2817, 7178.6094 --mrr, filtered mrr : 0.0005, 0.0006 --hits1 : 0.0000 --filtered hits1 : 0.0000 --hits3 : 0.0000 --filtered hits3 : 0.0000 --hits5 : 0.0000 --filtered hits5 : 0.0000 --hits10 : 0.0000 --filtered hits10 : 0.0000

The hyperparameter configuration file ./custom_hp.yaml has been modified. I can't find the benchmark data description of ConvE in the documentation. I want to know what's wrong with it, thank you