Repository for the DNN for HH->WWGG Snowmass analysis.

To run jupyter from within ingrid, do the two steps:

on your local, ssh -NL port_number:localhost:port_number -i .ssh/ssh_key ingrid1-gwceci
on the cluster, jupyter notebook --no-browser --port port_number

Give a port number of your choice (one that you know is not being used at that moment); 1234, 1235 etc.

(ingrid1-gwceci is the name in my ssh config to connect to ingrid.)

What I run, for example, is sh -NL 1234:localhost:1234 -i .ssh/id_rsa.ceci ingrid1-gwceci

Once you trained your model and saved it with tensorflow, use tf2onnx.

To install: pip install -U tf2onnx

Once installed : python -m tf2onnx.convert --saved-model tensorflow-model-path --output model.onnx will convert your tf model to onnx.