Gradient clipping with custom feed-forward net

16 views (last 30 days)
Everytime I am training my custom feed-forward net with 2 inputs and one output( timeseries) with the train(net,....) function:
after ~10 training epochs the value of the gradient reaches the prestet value and the training stops.
Changing the networks architecture is not an option in my case.
Is there a way to implement "gradient clipping" with a feed-forward net?
Or is there any other workaround for the "exploding gradient"?

Accepted Answer

Vineet Joshi
Vineet Joshi on 1 Sep 2021
Hi
The following documentation link will provide you suitable details regarding dealing with exploding gradients in MATLAB.
Hope this helps.
Thanks
  1 Comment
Artem Lenskiy
Artem Lenskiy on 4 Dec 2022 at 12:52
The answer you provided is not for a custom loop. See this example https://au.mathworks.com/help/deeplearning/ug/train-network-using-custom-training-loop.html there is the following line
[loss,gradients,state] = dlfeval(@modelLoss,net,X,T);
The question is how to apply clipping to gradients. Is there are standard Matlab function can do this for me or should I implement it myself.

Sign in to comment.

More Answers (1)

Artem Lenskiy
Artem Lenskiy on 4 Dec 2022 at 13:29
Please check this link that illustrates several examples on how to implement training options that you would usually define via trainingOptions() and use with trainNetwork() but for customs loops. Here is an L2 clipping example given in the link above
function gradients = thresholdL2Norm(gradients,gradientThreshold)
gradientNorm = sqrt(sum(gradients(:).^2));
if gradientNorm > gradientThreshold
gradients = gradients * (gradientThreshold / gradientNorm);
end
end
You might also find this link useful https://au.mathworks.com/help/deeplearning/ug/detect-vanishing-gradients-in-deep-neural-networks.html that discuss detection of vanishing gradients in deep neural networks.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!