Investigating Quantum Approximate Optimization Algorithms under Bang-bang Protocols

This repository contains a python implementation of

Investigating Quantum Approximate Optimization Algorithms under Bang-bang Protocols

Daniel Liang, Li Li and Stefan Leichenauer, Phys. Rev. Research 2, 033402 (2020)

Paper link: https://link.aps.org/doi/10.1103/PhysRevResearch.2.033402

Requirements

pip install -r requirements.txt

Search an optimal bang-bang protocol using stochastic descent

NOTE: This demo code runs on a single machine. For the results presented in the paper, we run the stochastic descent search on a distributed system.

python -m bangbang_qaoa.search_bangbang_qaoa_protocols \
  --num_literals=10 \
  --num_clauses=10 \
  --initialization_method=random \
  --num_chunks=100 \
  --total_time=3 \
  --num_samples=1

Depends on the initial random seed, the result will look like below:

Optimal protocol:
[0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 1, 1, 1, 1, 0, 1, 1, 0, 0,
 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0]
Protocol eval: 0.931884
Number of epoch: 74

The QAOA p value of this protocol is approximately 2, which is consistent with Figure 3(e) in the paper.