z-quantum-optimizers

What is it?

z-quantum-optimizers is a module with basic optimizers to be used in workflows with Orquestra – a platform for performing computations on quantum computers developed by Zapata Computing.

Currently this library includes the following optimizers:

Grid search
Scipy optimizers
CMS-ES (Covariance Matrix Adaptation Evolution Strategy)

It also supports the optimization of variational circuits.

Usage

Workflow

In order to use z-quantum-optimizers in your workflow, you need to add it as an import in your Orquestra workflow:

imports:
- name: z-quantum-optimizers
  type: git
  parameters:
    repository: "git@github.com:zapatacomputing/z-quantum-optimizers.git"
    branch: "main"

and then add it in the imports argument of your step:

- name: my-step
  config:
    runtime:
      language: python3
      imports: [z-quantum-optimizers]

Once that is done you can:

use any z-quantum-optimizers function by specifying its name and path as follows:

- name: optimize-circuit
  config:
    runtime:
      language: python3
      imports: [z-quantum-optimizers]
      parameters:
        file: z-quantum-optimizers/steps/optimizers.py
        function: optimize_variational_circuit

use tasks which import zquantum.optimizers in the python code (see below).

Python

Here's an example of how to use methods from z-quantum-optimizers in a python task:

from zquantum.optimizers import ScipyOptimizer
optimizer = ScipyOptimizer(method='L-BFGS-B')

or use optimizer-specs parameter to make our code work with other backends too:

from zquantum.core.utils import create_object
optimizer_specs = {{inputs.parameters.optimizer-specs}}
optimizer = create_object(optimizer_specs)

Even though it's intended to be used with Orquestra, z-quantum-optimizers can be also used as a standalone Python module. This can be done by running pip install . from the src/ directory.

Development and contribution

You can find the development guidelines in the z-quantum-core repository.

Running tests

In order to run tests please run pytest . from the main directory.

In order for the tests related to the daemon optimizer to work you need to first specify FLASK_APP environmental variable: export FLASK_APP=/path/to/z-quantum-optimizer/src/python/orquestra/optimizers/daemon-optimizer/proxy/rest.py.