zgrkpnr

@leeskyed The issue you refer talks about keeping `convnet_base.trainable` True. My problem is about using data augmentation while recording pretrained modles outputs as Numpy array first. If I am wrong,...

That would be awesome. I am using MathNet distributions and I copied the related source code into my project just to try `Samples(double[] values)` as `Samples(Span values)`. I used `stackalloc`...