We provide a notebook implementation of Conditional Neural Processes (CNPs) that can be run in the browser. The notebook includes an overview of the different building blocks of the model and the code to run it. Any further details can be found in our paper.
The easiest way to run the code is to download the .ipynb file and upload it
to Colab. Colaboratory is a free Jupyter
notebook environment provided by Google that requires no setup and runs entirely
in the cloud. The hosted runtime already includes the following dependencies:
- Numpy
- Tensorflow
- Matplotlib
which are all we need to run CNPs.
Alternative you can open the .ipynb file using
Jupyter notebook. If you do this you will
also have to set up a local kernel that includes Tensorflow.
... you might like our follow up work Neural Processes, where we introduce a latent variable version of CNPs that allows us to generate different samples given the same context.
If you like our work and end up using CNPs for your reseach give us a shout-out:
Garnelo M, Rosenbaum D, Maddison CJ, Ramalho T, Saxton D, Shanahan M, Teh YW, Rezende DJ, Eslami SM. Conditional Neural Processes. In International Conference on Machine Learning 2018.
Any feedback is much appreciated! Drop us a line at garnelo@google.com
This is not an official Google product.