Search papers, labs, and topics across Lattice.
This paper introduces a framework-agnostic quantum neural network (QNN) architecture designed to mitigate vendor lock-in within quantum machine learning (QML). The architecture employs a unified computational graph and hardware abstraction layer (HAL) to provide transparent access to diverse quantum computing backends, including IBM Quantum, Amazon Braket, and Azure Quantum. Benchmarking on standard classification tasks shows training time parity (within 8% overhead) and identical classification accuracy compared to native framework implementations, demonstrating the viability of the approach.
Escape the QML vendor lock-in: this framework-agnostic QNN lets you train once and deploy on any major quantum hardware platform with minimal overhead.
Quantum machine learning (QML) stands at the intersection of quantum computing and artificial intelligence, offering the potential to solve problems that remain intractable for classical methods. However, the current landscape of QML software frameworks suffers from severe fragmentation: models developed in TensorFlow Quantum cannot execute on PennyLane backends, circuits authored in Qiskit Machine Learning cannot be deployed to Amazon Braket hardware, and researchers who invest in one ecosystem face prohibitive switching costs when migrating to another. This vendor lock-in impedes reproducibility, limits hardware access, and slows the pace of scientific discovery. In this paper, we present a framework-agnostic quantum neural network (QNN) architecture that abstracts away vendor-specific interfaces through a unified computational graph, a hardware abstraction layer (HAL), and a multi-framework export pipeline. The core architecture supports simultaneous integration with TensorFlow, PyTorch, and JAX as classical co-processors, while the HAL provides transparent access to IBM Quantum, Amazon Braket, Azure Quantum, IonQ, and Rigetti backends through a single application programming interface (API). We introduce three pluggable data encoding strategies (amplitude, angle, and instantaneous quantum polynomial encoding) that are compatible with all supported backends. An export module leveraging Open Neural Network Exchange (ONNX) metadata enables lossless circuit translation across Qiskit, Cirq, PennyLane, and Braket representations. We benchmark our framework on the Iris, Wine, and MNIST-4 classification tasks, demonstrating training time parity (within 8\% overhead) compared to native framework implementations, while achieving identical classification accuracy.
