Closed MBSMGW closed 3 years ago
Well, i found a pattern in the numbers (The numbers, Mason.. What do they mean?!) from the first (layer 3), every 5th number is the value of what that spot +3+%4 should be so
0 3 1 0 2 1 3 2 4 7 5 4 6 5 7 6 8 11
Is it just the skip connections causing this?
@MBSMGW hey there is no logic in the naming. The number related to the activations is given by Keras and is a counter that is incremented every time an activation layer is defined. For the Conv Layers, we give them names based on where they are but nobody has ever checked in details.
Describe the bug When using the tcn_full_summary() to output the model arrangement, all of the numberings are either out of order or random. If the numbers were all ascending maybe i wouldn't have batted an eyelid, but i just can not find anything to explain why the numbers are jumbled.
See for example below, the names of the activation layers, moving from input to output: activation_3 (Activation)3 activation (Activation)0 1 2 7 4 5 6 11 8 9 10 15 12 13 14 19 16 17 18 23 20 21 22 27 24 25 26 31 28 29 30 35 32 33 34
Paste a snippet
Returns the following:
Dependencies Running in Jupyter, Python 3.8, Tensorflow 2.4.1, keras-tcn 3.4.0