skip to content
mathematics · 2 posts

Why Do Transformers Divide Attention Scores by √d?

nlp ·transformers ·mathematics

Why Adam Works: From Gradient Descent to Adaptive Optimization

deep-learning ·optimization ·mathematics