Reproducibility convolutional neural network training with gpu

Diego Alvarez Estevez

2018 9 20

1 个回答

回答已采纳

9 次查看（30 天）

2 个投票

Hello,

I am training a CNN using my local GPU (to speed up training) for classification problems and would like to try different parameterizations. To avoid the variability effects due to different data and/or weights initialization I am resetting the random seeds each time before training:

% Initialize random seed (thus same dataset on same architecture would lead
% to predictable result)
rng(0);
%parallel.gpu.rng(0, 'CombRecursive');
randStream = parallel.gpu.RandStream('CombRecursive', 'Seed', 0);
parallel.gpu.RandStream.setGlobalStream(randStream);
% Train the CNN network
net = trainNetwork(TR.data,TR.reference,layers,options);

The problem is that when using GPU I am getting different results on each execution, even if initializing the GPU random seed to the same value. Strange thing is if I use CPU instead, then I do get the reproducible results. I am doing something wrong with GPU random seed initialization? Is there a know problem for this situation or something I am missing?

Thanks beforehand.

PS: I am using Matlab R2017b

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Follow Question

采纳的回答

Joss Knight 2018-9-20

1 个投票

Use of the GPU has non-deterministic behaviour. You cannot guarantee identical results when training your network, because it depends on the whims of floating point precision and parallel computations of the form (a + b) + c ~= a + (b + c).

Most of our GPU algorithms are in fact deterministic but a few are not, for instance, backward convolution.