You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, authors,
When I reproduced the code, I can't get the same results as the ones shown in paper. I used those parameters provided in README, but Hits@10 is (under main_literals-FB15k-DistMult) about 0.75, can't get 0.82. And increasing the epochs seems contribute little in the final results. So I want to ask if I missed something in it. Or what is the best setup of parameters to get the results in paper. Thank you very much.
Hi, authors, When I reproduced the code, I can't get the same results as the ones shown in paper. I used those parameters provided in README, but Hits@10 is (under main_literals-FB15k-DistMult) about 0.75, can't get 0.82. And increasing the epochs seems contribute little in the final results. So I want to ask if I missed something in it. Or what is the best setup of parameters to get the results in paper. Thank you very much.
Same question, just get Hits @10 about 0.75 (under main_literals-FB15k-DistMult)
No description provided.
The text was updated successfully, but these errors were encountered: