Skip to content

MatGL-based neural network potential that computes excited state energies and forces

License

Notifications You must be signed in to change notification settings

pfnet-research/ex_matgl

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Excited state neural network potential using M3GNet model

Ex_MatGL is a MatGL-based neural network potential that computes excited state energies and forces.

Installation

  1. Install Pytorch. This package is tested on

    • CUDA==11.7
    • Python=3.9.17
    • torch==2.0.0
    python -m pip install torch==2.0.0+cu117 --index-url https://download.pytorch.org/whl/cu117
    
  2. Install DGL.

    python -m pip install dgl==1.0.1+cu117 -f https://data.dgl.ai/wheels/cu117/repo.html
    python -m pip install dglgo -f https://data.dgl.ai/wheels-test/repo.html
    
  3. Install Matgl

    Clone matgl repository and install latest version.1

    git clone [email protected]:materialsvirtuallab/matgl.git
    python -m pip install ./matgl
    
  4. Install this package

    python -m pip install .
    

Usage

Sample code is here. Please read instructuion.

References

Footnotes

  1. If you install matgl 0.8.5 version via pip and try to run sample code using GPU, you see the following error.TypeError: can't convert cuda:0 device type tensor to numpy. Use Tensor.cpu() to copy the tensor to host memory first.

About

MatGL-based neural network potential that computes excited state energies and forces

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages