I made it so that we have better logging done in python code. We
previously relied on the scala experiment code to save logs and model
parameters, but with that (likely) going away, I added some
functionality for it in python.
I updated the parameters for the decomposable attention example
experiment to match the original model as closely as DeepQA currently
allows.
I removed some old, unnecessary stuff from .gitignore, and added one
line for a models directory.
I made it so that we have better logging done in python code. We previously relied on the scala experiment code to save logs and model parameters, but with that (likely) going away, I added some functionality for it in python.
I updated the parameters for the decomposable attention example experiment to match the original model as closely as DeepQA currently allows.
I removed some old, unnecessary stuff from .gitignore, and added one line for a models directory.