tf.squeeze to remove dimensions of [1]

If I had a tensor of dimension [1, 2, 3], tf.squeze() would turn it into [2, 3]. tf.squeeze  |  TensorFlow Core r1.14  |  TensorFlow.

It removes any dimensions of shape 1 it finds. If I provide an axis parameter, it will only look at that dimension.


fixum – a fixed pay.
Elided - definition of elided by The Free Dictionary tr.v. e·lid·ed, e·lid·ing, e·lides

    a. To omit or slur over (a syllable, for example) in pronunciation.
    b. To strike out (something written).
    a. To eliminate or leave out of consideration.
    b. To cut short; abridge.

First seen as “[elided 10 identical lines from previous traceback]” in Tensorflow/Python

Learning Tensorflow

For a more systematic understanding: * jtoy/awesome-tensorflow: TensorFlow - A curated list of dedicated resources http://tensorflow.org is a list of nice stuff, not necessarily meant to be read in the same order * astorfi/TensorFlow-Roadmap: Organized & Useful Resources about Deep Learning with TensorFlow has a more “more GIFs to the god of GIFs” feel to it, but is meant to be read in a sequential order, and the resources look very nice.