automl_leaderboard(): Returns the AutoML Leaderboard from the AutoML run when the model was created.
automl_update_model(): Makes it easy to swap out models from the AutoML Leaderboard.
h2o.init()to skip on CRAN per H2O recommendation. The original issue was related to
libxgboost4j_gpu*.sobeing left on the user system when the user’s temp library. This is being corrected by H2O.
This is the initial release with key functions:
automl_reg(): Uses H2O AutoML as a backend for forecasting.
load_h2o_model(): Saving and loading H2O models and Workflow/Parsnip objects