setup.md

Setup help:

Make sure you have git installed. Open terminal, go to wherever you want your local repository, and type:

git init
git pull git@github.com:EmoryMLIP/DynamicBlocks.git

For running on a linux server, you may want to change the Git default editor:

git config --global core.editor "vim"

All the code should follow python3, so make sure that you use that throughout. (I get weird errors about the import statements if I try to run anything using python 2.) You may wish to set up a virtual environment (if not installed, install virtualenv). I called mine torchEnv.

python3 -m virtualenv torchEnv  

(If this line gives you trouble, see *Troubleshooting.)

To activate the virtual environment:

source torchEnv/bin/activate

I did a pip freeze to get all the requirements, removed the hazardous dependencies, and stored them in requirements.txt To load them into your virtual environment, run:

pip3 install -r requirements.txt 

Now run the default network (RK4 scheme using a Double Symmetric Layer on the CIFAR-10 dataset):

python3 RKNet.py 

NOTE: default behavior assumes that when running on cpu, the user is debugging. As such, functions will overwrite many parameters to train much smaller models and prevent the cpu from crashing.

When running outside of an IDE:

To run python functions in any subfolders, add the root folder to the path via:

source startup.sh

Then, run the functions from the root folder; for example, run the rk4 test as:

python3 modules/testRK4Block.py 

---

*Troubleshooting setup

Check the default python for creating virtual environments

echo $VIRTUALENV_PYTHON

You can set this to python3:

export VIRTUALENV_PYTHON=/usr/bin/python3

Also, virtualenv -p python3 torchEnv can be an alternative to python3 -m virtualenv torchEnv .