Torch. manualₛeed (3407) is all you need: On the influence of random seeds in deep learning architectures for computer vision

In this paper I investigate the effect of random seed selection on the when using popular deep learning architectures for computer vision. I a large amount of seeds (up to 10⁴) on CIFAR 10 and I also scan fewer on Imagenet using pre-trained models to investigate large scale datasets. conclusions are that even if the variance is not very large, it is easy to find an outlier that performs much better or much worse the average.