Skip to content

Code of the NVIDIA winning solution to the 2nd OGB-LSC at the NeurIPS 2022 challenge with dataset PCQM4Mv2

License

Notifications You must be signed in to change notification settings

CValse/NVIDIA-PCQM4Mv2

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

83 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

NVIDIA-PCQM4Mv2

Code of the NVIDIA winning solution to the 2nd OGB-LSC at the NeurIPS 2022 challenge with dataset PCQM4Mv2

In order to reproduce our solution one has to execute code in directories. Each directory has instructions in its README.md file.

  • Create training folds by running the code in the data directory.
  • Train models by running code in these directories (order is not important): cnn, modelcular_transformer, pd_dgn, TransformerM. Each of these downloads a variant of the competition dataset.
  • Run the ensembling notebook in ensemble directory.

The PDF document describes our solution.It is now available at https://arxiv.org/abs/2211.11035

About

Code of the NVIDIA winning solution to the 2nd OGB-LSC at the NeurIPS 2022 challenge with dataset PCQM4Mv2

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Jupyter Notebook 77.1%
  • Python 22.5%
  • Other 0.4%