How to run "train.py " faster and eficientlier?

INK-USC / RE-Net

Recurrent Event Network: Autoregressive Structure Inference over Temporal Knowledge Graphs (EMNLP 2020)

http://inklab.usc.edu/renet/

436 stars 95 forks source link

Open MrLiuCC opened 2 years ago

MrLiuCC commented 2 years ago

It takes too much time for one epoch. How to solve this problem?

woojeongjin commented 2 years ago

Hi it might be an environment issue. What versions of DGL, torch, etc did you use? Have you tried the versions in the readme? Thanks!

MrLiuCC commented 2 years ago

Hi it might be an environment issue. What versions of DGL, torch, etc did you use? Have you tried the versions in the readme? Thanks!

Thanks! I will try again