search

zhaodongh / Encoding-Word-Order-in-Complex-valued-Embedding

The code of Encoding Word Order in Complex-valued Embedding
42 stars 3 forks source link

Which version of tensoflow do you use? #2

Open wujsAct opened 4 years ago

wujsAct commented 4 years ago

I revise the code to run with tf15. And when I run Complex_vanilla model in folder BiLSTM, the program is stuck in and can not continue to run: def feed_neural_work(self): self.cell = URNNCell(num_units = self.max_input_left, num_in = self.embedding_size) outputs = tf.nn.dynamic_rnn(self.cell,self.embedded_chars_q,dtype=tf.float32)

I also revise the code urnn_cell.py: class URNNCell(tf.nn.rnn_cell.RNNCell): """The most basic URNN cell. Args: num_units (int): The number of units in the LSTM cell, hidden layer size. num_in: Input vector size, input layer size. """ def init(self, num_units, num_in, reuse=None): super(URNNCell, self).init()

wujsAct commented 4 years ago

@zhaodongh Complex_vanilla model in folder BiLSTM is stuck, have you ever encounter this problem?

zhaodongh commented 4 years ago

1.2.1