search

INK-USC / RE-Net

Recurrent Event Network: Autoregressive Structure Inference over Temporal Knowledge Graphs (EMNLP 2020)
http://inklab.usc.edu/renet/
435 stars 95 forks source link

Probably a typo in the paper #37

Open ZifengDing opened 3 years ago

ZifengDing commented 3 years ago

Hello,

I was checking your paper. In table 4, hits3 of ICEWS14 on RE-Net seems not very reasonable (it might be higher?). And could you please also provide the hits1 performance?

Best, Zifeng

fangyiming commented 3 years ago

Hello,

I was checking your paper. In table 4, hits3 of ICEWS14 on RE-Net seems not very reasonable (it might be higher?). And could you please also provide the hits1 performance?

Best, Zifeng

HELLO,I'm checking too.Can you run his code? Or could you please send your code to me? Thank you, my email is1043630181@qq.com

ZifengDing commented 3 years ago

Hello, I was checking your paper. In table 4, hits3 of ICEWS14 on RE-Net seems not very reasonable (it might be higher?). And could you please also provide the hits1 performance? Best, Zifeng

HELLO,I'm checking too.Can you run his code? Or could you please send your code to me? Thank you, my email is1043630181@qq.com

Hello, I ran it and got 26.495 as Hits@3 for ICEWS14 raw result. The code provided by author is runable, just a bit slow.

fangyiming commented 3 years ago

Thank you for your reply, I run the code a lot of problems, can you send your code and configuration information to me, what I use is python3.7, torch1.1, cuda11.0, thank you very much

------------------ 原始邮件 ------------------ 发件人: "INK-USC/RE-Net" <notifications@github.com>; 发送时间: 2020年12月15日(星期二) 上午7:41 收件人: "INK-USC/RE-Net"<RE-Net@noreply.github.com>; 抄送: "房意明"<1043630181@qq.com>;"Comment"<comment@noreply.github.com>; 主题: Re: [INK-USC/RE-Net] Probably a typo in the paper (#37)

Hello, I was checking your paper. In table 4, hits3 of ICEWS14 on RE-Net seems not very reasonable (it might be higher?). And could you please also provide the hits1 performance? Best, Zifeng

HELLO,I'm checking too.Can you run his code? Or could you please send your code to me? Thank you, my email is1043630181@qq.com

Hello, I ran it and got 26.495 as Hits@3 for ICEWS14 raw result. The code provided by author is runable, just a bit slow.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

woojeongjin commented 3 years ago

Hi! Thanks for letting me know. The result on the paper is Hit@1. It should be 26.56.

ZhengJialin1000 commented 2 years ago

Thank you for your reply, I run the code a lot of problems, can you send your code and configuration information to me, what I use is python3.7, torch1.1, cuda11.0, thank you very much … ------------------ 原始邮件 ------------------ 发件人: "INK-USC/RE-Net" <notifications@github.com>; 发送时间: 2020年12月15日(星期二) 上午7:41 收件人: "INK-USC/RE-Net"<RE-Net@noreply.github.com>; 抄送: "房意明"<1043630181@qq.com>;"Comment"<comment@noreply.github.com>; 主题: Re: [INK-USC/RE-Net] Probably a typo in the paper (#37) Hello, I was checking your paper. In table 4, hits3 of ICEWS14 on RE-Net seems not very reasonable (it might be higher?). And could you please also provide the hits1 performance? Best, Zifeng HELLO,I'm checking too.Can you run his code? Or could you please send your code to me? Thank you, my email is1043630181@qq.com Hello, I ran it and got 26.495 as Hits@3 for ICEWS14 raw result. The code provided by author is runable, just a bit slow. — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

HELLO,I'm checking too.Could you run his code now? I met the environment problem cause dgl updated.My device is the same to you.Could you send a version of yours.Thank you, my email is 1655791750@qq.com