2
$\begingroup$

It's taking me over 4 days to train a deep learning network with just 10000 images of 224px x 224px x 3 channels size, with batch size 25. The machine has 32GB RAM, a Core i7 CPU, and a GTX 960 GPU. I'm further using matconvnet and cudnn. Instead of training from scratch, I am using the ready made vggface model.

My question is: what could be the reason for the long training time and how could I fix this?

$\endgroup$
1
$\begingroup$

There are a number of reasons training might be taking a long time, but the first reason that comes to mind is that you haven't set an appropriate learning rate. If your learning rate is too high, instead of descending towards a minimum, your gradient path will bounce around uncontrollably/erratically, and this could happen indefinitely. If your learning rate is too low (I've seen this happen less frequently in practice), your model will take a longer time to reach the minimum, but it will eventually arrive. I'm not familiar with the particular libraries you've referenced, so unfortunately I can't offer library-specific details.

There is a pretty good general answer (with references) on Stack Overflow about setting a good learning rate in neural networks. See the link below.

https://stackoverflow.com/questions/11414374/neural-network-learning-rate-and-batch-weight-update

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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