PyTorch Models and CNN Theory Part 2

PYTORCH CUSTOM MODELS

How are you feeling? Ready for some more Python? Last time around we read through the basic theory behind how convolutional neural networks work. We saw that CNNs extract key features from our visual data (images). Now we’re ready to get going and see how everything falls into place as we apply the theory in real models.
In workshop 2 of this series, we introduced PyTorch, and showed you how to create datasets, data loaders, and bind it all together into a training and validation loop. In that article we were using pre-trained models which we fine-tuned with our own dataset to try and get better results. This time around we are delving into depth about how we can build a basic CNN network ourselves!