Skip to content

shashidharakc/recurrent-network-raw-lowlevel-api-tensorflow

Repository files navigation

recurrent-network-raw-lowlevel-api-tensorflow

Purpose of this repo is to code RNN, LSTM and GRU, network using low level tensorflow API's, To get better understanding.

Implementations are done based on the below equations :

LSTM:

Forward Computation : 
Ft = sigmoid(Whf*Ht^-1 + Wxf*Xt)
It = sigmoid(Whi*Ht^-1 + Wxi*Xt)
Ct_Dash = tanh(Whc*Ht^-1 + Wxc*Xt)
Ct = Ft*Ct^-1 + It*Ct_Dash
Ot = sigmoid(Who*Ht^-1 + Wxo*Xt)
Ht = Ot*tanh(Ct)

GRU:

Forward Computation : 
Zt = sigmoid(Whz*Ht^-1 + Wxz*Xt)
Rt = sigmoid(Whr*Ht^-1 + Wxr*Xt)
Ht_Dash = tanh((Rt*Whh*Ht^-1) + Wxh*Xt)
Ht = (1-Zt)*Ht^-1+Zt*Ht_Dash

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages