2
$\begingroup$

I am new to deep learning. I am working on training an SSD model on a set of small objects. I am using Adam gradient descent for optimization and a large input (800x800), but I seem to only get an improvement of 0.010 after every 20 or so epochs(350 steps).

What can I do or look for to speed up convergence on this model?

$\endgroup$
1
  • 1
    $\begingroup$ please put your architecture. Your input is very large, you may end up with very large dense layers after convolutional layers. $\endgroup$ Dec 4 '18 at 21:02
2
$\begingroup$

Implement the below mentioned techniques and check

  1. Add Batch Normalization

  2. Increase the learning rate

  3. Standard/Normalize the inputs if you have not done it already

$\endgroup$
0

Not the answer you're looking for? Browse other questions tagged or ask your own question.