Machine learning models can be optimised in several different ways.
One such type of optimisations is focused on reducing the number of parameters and operations involved by removing the connections between neural layers.
This pruning API is built on top of Keras making it easy for developers to apply it to any Keras training program.
Weight pruning has several advantages, the most immediate being disk compression. A 90% sparse model can be compressed from 12MB to nearly 2MB. Visit the website to get a complete picture of how it works.
TO DO: Use TensorFlow Model Optimization Toolkit to optimize neural network
We curate authentic and awesome content. Sign up for our newsletter for your daily dose of inspiration !