Description
PyData Chicago 2016
Slides: https://mheilman.github.io/pydata_chicago_2016/#/
Grid search over hyperparameters is an important but computationally expensive process in machine learning, particularly for deep learning and tree ensembles. In this talk, I will describe how one can use joblib's recently added custom backend functionality to do distributed grid search on Amazon EC2 for a TensorFlow deep text classifier that follows the scikit-learn estimator API.