Deep learning in H2O4GPU

Question

Good day! I had initially built a deep learning estimator on the h2o library in R, and I was thinking of using the h2o4gpu library instead to build it in order to expedite the process. Does anyone know if there is an existing function in h2o4gpu that is equivalent to the h2o.DeepLearningEstimator function in h2o? If not, is there a way for my h2o function to incorporate utilizing the GPU? Thank you!

I used the following function in R:

model_one = h2o.deeplearning(x = independent_variables,
                             training_frame = results_h2o,
                             autoencoder = TRUE,
                             #reproducible = TRUE,
                             seed = -1,
                             hidden = c(2000,1000,500,250,125,50),
                             epochs = 30,
                             activation = "Tanh"#
)

Answer 1

H2O deeplearning does not utilize GPU. However, there is the parameter fast_mode you can use to speed up your training.