MLS Course 2 Week2: Advanced optimization
From the screen shot with earlier timestamp: Adam would have as many learning rates as the number of weights in the Neural network + one learning rate for b.
In the later screenshot of the same slide: it shows that the learning rate initiated was only 1 single value.(?) Expected to see multiple intiialized values.
I understand that Adam indepth discussion is reserved for advanced Deep learning courses. Trying to understand keras implementation discussed in here.
Thanks