def __init__(self, vocab_size, wordvec_size, hidden_size): V, D, H = vocab_size, wordvec_size, hidden_size rn = np.random.randn embed_W = (rn(V, D) / 100).astype('f') lstm_Wx = (rn(D, 4 * H) / np.sqrt(D)).astype('f') lstm_Wh = (rn(H, 4 * H) / np.sqrt(H)).astype('f') lstm_b = np.zeros(4 * H).astype('f') self.embed = TimeEmbedding(embed_W) self.lstm = TimeLSTM(lstm_Wx, lstm_Wh, lstm_b, stateful=False) self.params = self.embed.params + self.lstm.params self.grads = self.embed.grads + self.lstm.grads self.hs = None
def __init__(self, vocab_size, wordvec_size, hidden_size): V, D, H = vocab_size, wordvec_size, hidden_size rn = np.random.randn embed_W = (rn(V, D) / 100).astype('f') lstm_Wx = (rn(D, 4 * H) / np.sqrt(D)).astype('f') lstm_Wh = (rn(H, 4 * H) / np.sqrt(H)).astype('f') lstm_b = np.zeros(4 * H).astype('f') affine_W = (rn(H, V) / np.sqrt(H)).astype('f') affine_b = np.zeros(V).astype('f') self.embed = TimeEmbedding(embed_W) self.lstm = TimeLSTM(lstm_Wx, lstm_Wh, lstm_b, stateful=True) self.affine = TimeAffine(affine_W, affine_b) self.params, self.grads = [], [] for layer in (self.embed, self.lstm, self.affine): self.params += layer.params self.grads += layer.grads