What are the TensorFlow dynamic_rnn options for this simple dataset?

Question

I want to train an RNN language model using TensorFlow.

My training data is a sequence of 5 tokens represented by integers such as

 x = [0, 1, 2, 3, 4]

I want the deployed RNN length to be 4 and the batch size to 2. (I selected these values to require padding.)

Each token has an attachment of length 3 so that

 0 -> [0, 0 ,0]
 1 -> [10, 10, 10]
 2 -> [20, 20, 20]
 3 -> [30, 30, 30]
 4 -> [40, 40, 40]

What should I pass as parameters tf.nn.dynamic_rnn?

" TensorFlow tf.nn.dynamic_rnn ?" . . , TensorFlow, . ( , , : 1.) , , , , , .

0

W.P. McNeill 21 . '16 1:35

1

Eugene Brevdo · Answer 1 · 2017-03-24T20:08:01+0000

rnn- , , /. , tf.nn.embedding_lookup. 3- x x embedding_depth ( 3)

embedding_lookup batch_size x time.