The tutorial will be based on three notebooks available at this link. They will also be uploaded s Google colab notebooks. Participants will either be able to open the copy of the notebook, run a cell to install PennyLane and do everything locally, or they can copy the colab notebooks and work online.
The three hours will consist of 3 modules, and in each module participants will go through one of the notebooks for about 25 minutes, and then there will be 20 minutes for a task (which participants can solve in the very same instance). The third block will also have some theory slides.
---------------------------------------------------------
Part I: Classical machine learning with automatic differentiation
Notebook 1-classical-ml-with-automatic-differentiation
Learning objectives:
- Be able to explain the concept of automatic differentiation
- Be able to train a simple linear model using automatic differentiation
Part II: Differentiable quantum computing
Notebook 2-differentiable-quantum-computations
Learning objectives:
- Be able to implement a variational quantum circuit in PennyLane
- Compute the gradient of a variational quantum circuit
- Train a variational quantum circuit like a machine learning model
Part III: Quantum gradients on remote devices
Slides and Notebook 3-quantum-gradients-with-braket
Learning objectives:
- Be able to name and explain three different ways to compute quantum gradients
- Understand why parameter-shift rules are needed for hardware
- Be able to compute a quantum gradient on a remote backend
