The machine learning framework has the ability to construct quantum datasets, prototype hybrid quantum and classic machine learning models, support quantum circuit simulators and train both discriminative and generative quantum models.
According to a Google AI blog, TensorFlow Quantum is able to create quantum models with standard Keras functions and by providing quantum circuit simulators and quantum computing primitives compatible with existing TensorFlow APIs.
The release of TensorFlow Quantum comes after Microsoft’s launch of Azure Quantum and the recent news that Honeywell is developing a quantum computer with a quantum volume of at least 64 which will be available in the next three months.
In an abstract for a paper, authored by members of Google’s X unit, The Institute for Quantum Computing at the University of Waterloo, NASA’s Quantum AI Lab, Volkswagen and Google Research, submitted to the preprint repository arXiv, the authors explain what they believe TensorFlow Quantum can achieve, saying:
“We hope this framework provides the necessary tools for the quantum computing and machine learning research communities to explore models of both natural and artificial quantum systems, and ultimately discover new quantum algorithms which could potentially yield a quantum advantage.”
The paper provides details on the TensorFlow Quantum software stack that combines the open source quantum circuit library Cirq with the TensorFlow machine learning platform.
TensorFlow Quantum’s ability to simulate properties will hopefully lead to advances in the fields of life sciences, decryption, chemical or material development and optimization.