search

modern-fortran / neural-fortran

A parallel framework for deep learning
MIT License
395 stars 82 forks source link

Mv optimizer from the network level to the layer level #184

Open jvdp1 opened 3 months ago

jvdp1 commented 3 months ago

As discussed, here is a draft in which I suggst to moved the optimizer from the network level to the layer level.

This is just a draft with an implementation for the dense layer only.

Here are the wall clock times using my dataset (with 2 hidden dense layers):

v0.17.0

Current PR