Poisson Loss Function for Gradient Boosting

Question

I have applied gradient boosting regression using sklearn.ensemble.GradientBoostingRegressor for the prediction of some count data. As there are 4 loss functions(loss: {‘ls’