question
Opened this issue · 4 comments
HengYuh commented
Dear Xiou:
I'm sorry to bother you again, but I reproduce the result in CompoundE on wn18rr,
My code as follows:
but I tried two methods with wn18rr in Rotate and Rotate3d, the results were much lower than expected, I wonder if there are any special methods or wrong with my codes.
b1ackwater3 commented
bro,did you successfully reproduce the results of the CompounE paper on the WN18RR & FB15k-237 dataset?
HengYuh commented
NO，I guess it may because the normaliztion trick
…---Original---
From: "long ***@***.***>
Date: Mon, May 27, 2024 15:31 PM
To: ***@***.***>;
Cc: ***@***.******@***.***>;
Subject: Re: [hughxiouge/CompoundE] question (Issue #2)
bro,did you successfully reproduce the results of the CompounE paper on the WN18RR & FB15k-237 dataset?
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: ***@***.***>
b1ackwater3 commented
Could you release the source code for reproducing FB15k-237? I would
greatly appreciate it.
HengYu ***@***.***> 于2024年5月27日周一 18:42写道：
… It is OK for FB15K237，while wn18rr a little far away，you don not have to
stick to this，I guess it might because some tircks，such as head
normalization，and the method present in paper may be wrong. You can refer
to CompoundE3D more-except for YAGO3-10
---Original---
From: "long ***@***.***>
Date: Mon, May 27, 2024 15:31 PM
To: ***@***.***>;
Cc: ***@***.******@***.***>;
Subject: Re: [hughxiouge/CompoundE] question (Issue #2)
bro,did you successfully reproduce the results of the CompounE paper on
the WN18RR & FB15k-237 dataset?
—
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID:
***@***.***>
—
Reply to this email directly, view it on GitHub
<#2 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/A57WRZJI5J2OO5COPSDDBOTZEMEZTAVCNFSM6AAAAABFYL4GFCVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCMZTGE4TQOBXGI>
.
You are receiving this because you commented.Message ID:
***@***.***>
b1ackwater3 commented
Here is my source code,idk what is wrong.
