PyTorch Archives

Diagram showing how batch normalization makes neural networks faster, with the formula mu_B, sigma_B squared, gamma, and beta applied to layer inputs

by Sanksshep Mahendra Updated May 29, 2026 at 11:18 am

AI

Batch Normalization Makes Neural Networks Faster

Why it matters: See how batch normalization speeds up neural network training, what its formula means, and how to add it in PyTorch and Keras the right way.

Diagram showing cross entropy loss in machine learning with the formula, predicted probability, and true label for a binary classification example.

by Sanksshep Mahendra Updated May 31, 2026 at 8:29 pm

AI

Cross Entropy Loss in Machine Learning

Why it matters: Cross entropy loss explained: binary cross entropy loss formula, categorical cross entropy, focal loss, label smoothing, PyTorch code, and production tips.

Diagram of the softmax activation function turning neural network logits into probabilities

by Sanksshep Mahendra Updated June 5, 2026 at 2:11 pm

AI

Softmax Activation Function

Why it matters: Master the softmax activation function: math, gradient, temperature scaling, transformer attention, PyTorch code, calibration risks, and modern alternatives.