genQC · Generative Quantum Circuits

Code repository for generating quantum circuits with diffusion models.

📰 News

• 🔥 [2025-06-02] Paper release: Synthesis of discrete-continuous quantum circuits with multimodal diffusion models.
• 🔥 [2025-06-01] Discrete-continuous circuits with multimodal diffusion - model released on Hugging Face: huggingface.co/collections/Floki00.

The codebase

The code contained within this repo allows the sampling of pre-trained diffusion models and includes our pipeline to fine-tune and train models from scratch. Pre-trained weights can be found on [Hugging Face] and can be downloaded automatically via our code (see minimal example). For the text CLIP model weights we use the OpenCLIP library, which will download (and cache) the CLIP model on first usage of our pipeline. In case you prefer reading a documentation, rather than notebooks or code, see the project page under [Documentation].

This repo inlcudes:

1. genQC/ a full release of our used diffusion pipeline.
2. src/examples/ examples and tutorials to show how to use the library.
3. src/ the source notebooks for nbdev.

Examples

Minimal example

A minimal example to compile the 4-qubit Quantum Fourier transform (QFT) unitary, using parameterized circuits.

``` python import torch from genQC.pipeline.multimodaldiffusionpipeline import MultimodalDiffusionPipelineParametrizedCompilation from genQC.inference.sampling import generatecompilationtensors, decodetensorstobackend from genQC.utils.miscutils import infertorchdevice, setseed from genQC.platform.tokenizer.circuitstokenizer import CircuitTokenizer from genQC.benchmark.benchcompilation import SpecialUnitaries from genQC.platform.simulation import Simulator, CircuitBackendType

device = infertorchdevice() set_seed(0)

pipeline = MultimodalDiffusionPipelineParametrizedCompilation.frompretrained( repoid="Floki00/cirditmultimodalcompile3to5qubit", device=device)

pipeline.scheduler.settimesteps(40) pipeline.schedulerw.set_timesteps(40)

pipeline.gh, pipeline.gw = 0.3, 0.1 pipeline.lambdah, pipeline.lambdaw = 1.0, 0.35

U = SpecialUnitaries.QFT(num_qubits=4).to(torch.complex64)

outtensor, params = generatecompilationtensors(pipeline, prompt="Compile 4 qubits using: ['h', 'cx', 'ccx', 'swap', 'rx', 'ry', 'rz', 'cp']", U=U, samples=8, systemsize=5, numofqubits=4, max_gates=32) ```

``` python vocabulary = {g:i+1 for i, g in enumerate(pipeline.gate_pool)} tokenizer = CircuitTokenizer(vocabulary) simulator = Simulator(CircuitBackendType.QISKIT)

qclist, _ = decodetensorstobackend(simulator, tokenizer, outtensor, params) qclist[0].draw("mpl") ```

Further examples

More detailed examples and tutorial notebooks are provided on the project page [tutorials] or in the directory src/examples/.

Installation

The installation of genQC is done via pip within a few minutes, depending on your downloading speed.

Method 1: pip install

To install genQC just run:

sh pip install genQC

Note, this will install missing requirements automatically. You may want to install some of them manually beforehand, e.g. torch for specific cuda support, see https://pytorch.org/get-started/locally/.

Requirements: genQC depends on python (min. version 3.12) and the libraries: torch, numpy, matplotlib, scipy, omegaconf, qiskit, tqdm, joblib, open_clip_torch, ipywidgets, pylatexenc, safetensors, tensordict and huggingface_hub. All can be installed with pip install. In src/RELEASES.md [doc] and the GitHub release descriptions, specific tested-on versions are listed.

Method 2: clone the repository

To use the latest GitHub code, you can clone the repository by running:

sh git clone https://github.com/FlorianFuerrutter/genQC.git cd genQC

The library genQC is built using jupyter notebooks and nbdev. To install the library use in the clone directory:

sh pip install -e .

Test installation

You can run the provided src/examples/Quantum circuit synthesis with diffusion models/0_hello_circuit [doc] [notebook] example to test your installation. On a computer with a moderate GPU this inference example notebook should run under half a minute.

License

The code and weights in this repository are licensed under the Apache License 2.0.

BibTeX

We kindly ask you to cite our paper if any of the previous material was useful for your work.

Quantum circuit synthesis with diffusion models

latex @article{furrutter2024quantum, title={Quantum circuit synthesis with diffusion models}, author={F{\"u}rrutter, Florian and Mu{\~n}oz-Gil, Gorka and Briegel, Hans J}, journal={Nature Machine Intelligence}, doi = {https://doi.org/10.1038/s42256-024-00831-9}, vol = {6}, pages = {515-–524}, pages={1--10}, year={2024}, publisher={Nature Publishing Group UK London} }