Info#

The demo script Train.py can used to demo the following functions:

*   Demonstrates that the Compile Model can process an Output

*   Demonstrates who to export the Tensorflow Model (Issue)

*   Demonstrates how to visualise slices of the Convolutional Layers

*   Saves model summary and Image to the Model folder

Controllers#

By changing the following lines to True will dictate the what function is processed.

test_batch = False
get_graph = False
show_activation = False
show_cmd_output = True
  • test_batch - this will compile the model, generate some random random data and then predict the output.
  • get_graph - this will when it works how to export the Tensorflow graph
  • show_activation - this function displays using Matplotlib the visulisation of 10 slices from one Covnet Layer
  • show_cmd_output - this will save a summary of the file to txt file and export the diagram of the model to a folder. Instructions are contained in the file of how to run this.

Layer Slices#

Slice1 Slice2 Slice3 Slice4 Slice5 Slice6 Slice7 Slice8 Slice9 Slice10