simpleRnn.py

class SimpleRNN:
  def __init__():
    self.W_sh = np.zeros(shape=[rnn_size, rnn_size])
    self.W_xh = np.zeros(shape=[word_vecotr_size, rnn_size])
    self.s = np.zeros(shape=[rnn_size,1]
    
  def step(self, x):
    # update the state
    self.s = np.tanh(np.dot(self.W_sh, self.s) + np.dot(self.W_xh, x))
    # compute the output vector
    y = softmax(self.s)
    return y

rnn = SimpleRnn()
y = rnn.step(x)