These are tutorials held at DESY for HEP (and Astroparticle) physicists
To get the tutorials:
git clone https://github.com/dkgithub/DESY_ML_PyTorch.git
Connect to your machine with a tunnel:
mkdir tmp
ssh -i A_DESY.pem -S./tmp -L 1080:127.0.0.1:8888 user@machine
To get the right environment
- on AWS (Amazon)
cd DESY_ML_PyTorch
source startROOT.sh
We need some extra software
pip install sklearn
jupyter lab --no-browser
Look for the token.
Access the jupyter server on your laptop: http://localhost:1080/ Copy the token as password.
If you are working on a machine with different people jupyter will look for a free port number. If this happens you need to connect by the above (1080:127.0.0.1:8888<--change to something free, e.g. 1234)
jupyter lab --no-browser --port 1234
Navigate to the files
Pick the "conda_pytorch_p27" environment for all except the create toy example below.
CTRL-C CTRL-C to shutdown jupyter
The createToy.ipynb example needs a different setup!! Pick the "python2" environment!!