Queue popup from python iterator

Question

I would like to create a queue that is populated from an iterator. However, the following MWE always runs the same value:

import tensorflow as tf
import numpy as np

# data
imgs = [np.random.randn(i,i) for i in [2,3,4,5]]

# iterate through data infinitly
def data_iterator():
    while True:
        for img in imgs:
            yield img

it = data_iterator()

# create queue for data
q = tf.FIFOQueue(capacity=5, dtypes=[tf.float64])

# feed next element from iterator
enqueue_op = q.enqueue(list(next(it)))

# setup queue runner
numberOfThreads = 1 
qr = tf.train.QueueRunner(q, [enqueue_op] * numberOfThreads)
tf.train.add_queue_runner(qr) 

# dequeue
dequeue_op  = q.dequeue() 
dequeue_op = tf.Print(dequeue_op, data=[dequeue_op], message="dequeue()")

# We start the session as usual ...
with tf.Session() as sess:
    coord = tf.train.Coordinator()
    threads = tf.train.start_queue_runners(coord=coord)

    for i in range(10):
        data = sess.run(dequeue_op)
        print(data)
.
    coord.request_stop()
    coord.join(threads)

Should i use feed_dict? If so, how do I use it in combination with QueueRunner?

Answer 1

At startup

enqueue_op = q.enqueue(list(next(it)))

tensorflow will execute the list (next (it)) exactly once. After that, it will save this first list and add it to q every time it starts enqueue_op. To avoid this, you should use a placeholder. Feed placeholders are not compatible with tf.train.QueueRunner. Use this instead:

import tensorflow as tf
import numpy as np
import threading

# data
imgs = [np.random.randn(i,i) for i in [2,3,4,5]]

# iterate through data infinitly
def data_iterator():
    while True:
        for img in imgs:
            yield img

it = data_iterator()

# create queue for data
q = tf.FIFOQueue(capacity=5, dtypes=[tf.float64])

# feed next element from iterator

img_p = tf.placeholder(tf.float64, [None, None])
enqueue_op = q.enqueue(img_p)

dequeue_op  = q.dequeue()


with tf.Session() as sess:
    coord = tf.train.Coordinator()

    def enqueue_thread():
        with coord.stop_on_exception():
            while not coord.should_stop():
                sess.run(enqueue_op, feed_dict={img_p: list(next(it))})

    numberOfThreads = 1
    for i in range(numberOfThreads):
      threading.Thread(target=enqueue_thread).start()



    for i in range(3):
        data = sess.run(dequeue_op)
        print(data)