Adding xgb_model paramter allows the continuation of model training. Model has to be saved by calling `model.get_booster().save_model(path)`
Julian Niedermeier