GSoC: Phase 1 Evaluation



This is a short update on my progress till the Phase 1 Evaluations in GSoC 2020. The following goals were set for the 1st month according to my proposal, of which I was able to complete the following.

Header MethodPyTorch

  • Add Initial Structure for MethodPyTorch.h header file.

Class MethodPyTorch Implement the basic outline of MethodPyTorch.cxx Class.

  • Implement the DeclareOptions method
  • Implement ProcessOptions method
  • Implement SetupTorchModel method
  • Implement Validation Sample Splits
  • Implement train methods. (Could not complete)
    • Implement simple basic structure for training
    • Add ability to load code from users
    • Convert numpy based dataloaders in TMVA to PyTorch dataloaders
    • Use default optimizer (optional optimizer SGD)
    • Add savebest model
    • Add scheduler
  • Implement helper function

The train method in the MethodPyTorch requires a few more preliminary options which will facilitate loading user code into the TMVA Interface, but it’s not implemented in PyMVA yet. Lorenzo (my mentor) is working on a solution for that exact problem and as soon as that is fixed, I should be able to implement the train method.

Hopefully, I’ll get this done in the next couple of weeks!🤞