Implementing Essential Deep Learning Modules - Week 9

Building up on our work from the last week on optimizing our ANN framework, we went forward with implementing EvaluateWithGradient() function for the FFN and RNN classes as well.

Though we had done the same with the aim of reducing code duplication in mind initially, we realized that with the above function implemented, we were able to obtain atleast a 30% speedup in the case of simple FFN networks!

For the case of RNN class, the speedup was slightly lower at 22% ~ 25%, primarily because of the heavier gradient computation routines being used. Nevertheless, we also applied the above function inside our GAN::EvaluateWithGradient() function, so a certain amount of speedup is expected there as well!

I also received my Phase II evaluations this week, and I'm glad that Marcus is satisfied with the effort that we have put in. I will continue to build up on my work on RBMs and hopefully, we can merge them as well before this month ends.

Mirupafshim