search

udacity / deep-reinforcement-learning

Repo for the Deep Reinforcement Learning Nanodegree program
https://www.udacity.com/course/deep-reinforcement-learning-nanodegree--nd893
MIT License
4.85k stars 2.34k forks source link

calculate fan-in correctly #15

Closed dantp-ai closed 2 years ago

dantp-ai commented 5 years ago

Fan-in is defined to be the maximum number of inputs to a layer. The weight matrix is transposed. This means that the number of inputs are equal to the second component in the shape not the first one (0-indexing).