How to pass `None` Batch Size to tensorflow dynamic rnn? - tensorflow

Question

I am trying to build a CNN+LSTM+CTC model for word recognition.
Initially I have an image, I am converting the features extracted using CNN on word image and building a sequence of features which I am passing as sequential data to RNN.
Follwing is how I am converting the features to sequential data:
[[a1,b1,c1],[a2,b2,c2],[a3,b3,c3]] -> [[a1,a2,a3],[b1,b2,b3],[c1,c2,c3]]
Where a,b,c are 3 features extracted using CNN.
Presently I can pass a constant batch_size to the model common.BATCH_SIZE, but what I want is to be able to pass variable batch_size to the model.
How can this be done ?
inputs = tf.placeholder(tf.float32, [common.BATCH_SIZE, common.OUTPUT_SHAPE[1], common.OUTPUT_SHAPE[0], 1])
# Here we use sparse_placeholder that will generate a
# SparseTensor required by ctc_loss op.
targets = tf.sparse_placeholder(tf.int32)
# 1d array of size [batch_size]
seq_len = tf.placeholder(tf.int32, [common.BATCH_SIZE])
model = tf.layers.conv2d(inputs, 64, (3,3),strides=(1, 1), padding='same', name='c1')
model = tf.layers.max_pooling2d(model, (3,3), strides=(2,2), padding='same', name='m1')
model = tf.layers.conv2d(model, 128,(3,3), strides=(1, 1), padding='same', name='c2')
model = tf.layers.max_pooling2d(model, (3,3),strides=(2,2), padding='same', name='m2')
model = tf.transpose(model, [3,0,1,2])
shape = model.get_shape().as_list()
model = tf.reshape(model, [shape[0],-1,shape[2]*shape[3]])
cell = tf.nn.rnn_cell.LSTMCell(common.num_hidden, state_is_tuple=True)
cell = tf.nn.rnn_cell.DropoutWrapper(cell, input_keep_prob=0.5, output_keep_prob=0.5)
stack = tf.nn.rnn_cell.MultiRNNCell([cell]*common.num_layers, state_is_tuple=True)
outputs, _ = tf.nn.dynamic_rnn(cell, model, seq_len, dtype=tf.float32,time_major=True)
UPDATE:
batch_size = tf.placeholder(tf.int32, None, name='batch_size')
inputs = tf.placeholder(tf.float32, [batch_size, common.OUTPUT_SHAPE[1], common.OUTPUT_SHAPE[0], 1])
# Here we use sparse_placeholder that will generate a
# SparseTensor required by ctc_loss op.
targets = tf.sparse_placeholder(tf.int32)
# 1d array of size [batch_size]
seq_len = tf.placeholder(tf.int32, [batch_size])
model = tf.layers.conv2d(inputs, 64, (3,3),strides=(1, 1), padding='same', name='c1')
model = tf.layers.max_pooling2d(model, (3,3), strides=(2,2), padding='same', name='m1')
model = tf.layers.conv2d(model, 128,(3,3), strides=(1, 1), padding='same', name='c2')
model = tf.layers.max_pooling2d(model, (3,3),strides=(2,2), padding='same', name='m2')
model = tf.transpose(model, [3,0,1,2])
shape = model.get_shape().as_list()
model = tf.reshape(model, [shape[0],-1,shape[2]*shape[3]])
cell = tf.nn.rnn_cell.LSTMCell(common.num_hidden, state_is_tuple=True)
cell = tf.nn.rnn_cell.DropoutWrapper(cell, input_keep_prob=0.5, output_keep_prob=0.5)
stack = tf.nn.rnn_cell.MultiRNNCell([cell]*common.num_layers, state_is_tuple=True)
outputs, _ = tf.nn.dynamic_rnn(cell, model, seq_len, dtype=tf.float32,time_major=True)
I am getting error as below:
Traceback (most recent call last):
File "lstm_and_ctc_ocr_train.py", line 203, in <module>
train()
File "lstm_and_ctc_ocr_train.py", line 77, in train
logits, inputs, targets, seq_len, batch_size = model.get_train_model()
File "/home/himanshu/learning-tf/tf/code/tensorflow_lstm_ctc_ocr/model.py", line 20, in get_train_model
inputs = tf.placeholder(tf.float32, [batch_size, common.OUTPUT_SHAPE[1], common.OUTPUT_SHAPE[0], 1])
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/array_ops.py", line 1530, in placeholder
return gen_array_ops._placeholder(dtype=dtype, shape=shape, name=name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/gen_array_ops.py", line 1954, in _placeholder
name=name)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/op_def_library.py", line 705, in apply_op
attr_value.shape.CopyFrom(_MakeShape(value, key))
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/op_def_library.py", line 198, in _MakeShape
return tensor_shape.as_shape(v).as_proto()
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/tensor_shape.py", line 798, in as_shape
return TensorShape(shape)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/tensor_shape.py", line 434, in __init__
self._dims = [as_dimension(d) for d in dims_iter]
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/tensor_shape.py", line 376, in as_dimension
return Dimension(value)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/tensor_shape.py", line 32, in __init__
self._value = int(value)
TypeError: int() argument must be a string or a number, not 'Tensor'

Answer 1

You should be able to pass batch_size to dynamic RNN as a placeholder. In my experience, the only headache you might experience is if you don't specify its shape in advance, so you should pass [] to make things work, something like this:
batchsize = tf.placeholder(tf.int32, [], name='batchsize')
and then feed its value during sess.run() in the usual way. This works well for me while training on a large batch size, but then generating with batch of 1.
But strictly speaking, you don't even need to specify batch size for dynamic_rnn specifically, do you? You need it if you use MultiRNNCell to get zero state, but I don't see you doing that in your code...
*** UPDATE:
As discussed in the comment, your problem seems to have nothing to do with dynamic_rnn and more to do with the fact that you are using a placeholder inputs to specify the shape of another placeholder seq_len. Here's the code that reproduces the same error:
import tensorflow as tf
a = tf.placeholder(tf.int32, None, name='a')
b = tf.placeholder(tf.int32, [a, 5], name='b')
c = b * 5
with tf.Session() as sess:
C = sess.run(c, feed_dict={a:1, b:[[1,2,3,4,5]]})
And here's the error:
TypeError: int() argument must be a string, a bytes-like object or a number, not 'Tensor'
Before getting to the headache of dynamic_rnn, I would recommend finding a way around this, either by changing your code or by asking a separate question on how to fudge it with placeholders.