How To Correctly Implement Lazy Loading In Tensorflow?

The following code (while trying to replicate code structure in https://danijar.com/structuring-your-tensorflow-models/ ) import tensorflow as tf class Model: def __init__(s

Solution 1:

The problem is that the call to sess.run(tf.global_variables_initializer()) happens before the variables are created, in the first call to model.output on the following line.

To fix the problem, you must somehow access model.output before calling sess.run(tf.global_variables_initializer()). For example, the following code works:

import tensorflow as tf

classModel:

    def__init__(self, x):
        self.x = x
        self._output = None    @propertydefoutput(self):
        # NOTE: You must use `if self._output is None` when `self._output` can# be a tensor, because `if self._output` on a tensor object will raise# an exception.if self._output isNone:
            weight = tf.Variable(tf.constant(4.0))
            bias = tf.Variable(tf.constant(2.0))
            self._output = tf.multiply(self.x, weight) + bias
        return self._output

defmain():
    x = tf.placeholder(tf.float32)
    model = Model(x)

    # The variables are created on this line.
    output_t = model.output

    with tf.Session() as sess:
        sess.run(tf.global_variables_initializer())
        output = sess.run(output_t, {x: 4.0})
        print(output)

if __name__ == '__main__':
    main()