Using logistics function as activation in each RNN cell

[prateekagrawal06]

Hi Professor,

The Assignment 4 bonus part tells that we need to use logistic as the activation function for each cell in the hidden layer. Does this mean that we need to change the tanh activation to sigmoid in the BASICRnnCell implementation of tensorflow?

[aronwc]

That would be one way to do it.

On Apr 10, 2017, at 6:51 PM, prateekagrawal06 notifications@github.com wrote:

BASICRnnCell

[prateekagrawal06]

Thanks, professor