ParentFull threadsneed_chucker·Deep Learning, as in, "efficiently trained neural networks utilizing gradient descent" really isn'tView on HN