Skip to content

alejands/DeepMETv1

 
 

Repository files navigation

DeepMETv1

DeepMETv1 is a fully-connected neural network (FCNN) for MET reconstruction in CMS data. This branch is focused on training on Run3 conditions.

The original repository by Yongbing Feng was used to train DeepMETv1 on Run2 conditions. A copy of this original code is in the branch Run2.


Install the necessary packages with MiniConda. You can use the provided env_deepmetv1.yml file.

conda env create -y -f env_deepmetv1.yml

Activate the environment

conda activate METTraining

Prepare the HDF5 training files (to see log info, add option -v)

python convertNanoToHDF5.py inputfile.root outputfile.h5

To see more options

python convertNanoToHDF5.py --help

Run the training

python train_ptmiss_mine.py -i input.txt

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%