# custom model

Tensorflow insights - part 6: Custom model - Inception V3

The VGG block boils down to only a sub-network that contains a sequence of convolutional layers and a max-pooling layer. Each layer is just connected right after another layer in a consecutive manner, which is exactly in the same way as all the networks that we used before part 4. For that reason, you might not have gained the full advantage of using the Tensorflow custom layer/model. In this post, we will get familiar with the idea of parallel paths and implement the Inception module which is used by the variants of the Inception network. To be practical, we will then show you how to implement the Inception-v3 network architecture. Throughout this post, you will see a lot more of the power of the Tensorflow custom layer/model.


Tensorflow insights - part 5: Custom model - VGG - continue

In the last part, we have shown how to use the custom model to implement the VGG network. However, one problem that remained is we cannot use model.summary() to see the output shape of each layer. In addition, we also cannot get the shape of filters. Although we know how the VGG is constructed, overcoming this problem will help the end-users - who only use our checkpoint files to investigate the model. In particular, it is very important for us to get the output shape of each layer/block when using the file test.py.


Tensorflow insights - part 4: Custom model - VGG

In this post, we will use the Tensorflow custom model to efficiently implement the VGG architecture so we can easily experiment with many variants of the network. The network architecture is deeper and will help us to increase the final performance.