Skip to content

Latest commit

 

History

History
16 lines (9 loc) · 694 Bytes

README.md

File metadata and controls

16 lines (9 loc) · 694 Bytes
Raw

Lightning JTML (LitJTML)

This repo is for moving our JTML neural network code to PyTorch Lightning and using WandB logging.

Setup:

  1. Create the conda environment lit-jtml-env from the environment.yml using the command conda create env create -f environment.yml.
  2. Activate the conda env with conda activate lit-jtml-env.

Use:

  1. Be in the LitJTML directory (use the cd command to change the directory to the blah/blah/LitJTML/ directory).
  2. To fit (train) a model, call python scripts/fit.py my_config where my_config is the name of the config.py file in the config/ directory.
    • The config file should specify the model, data, and other parameters.