QHack 2024: Quantum Machine Learning Challenge Course

QHack is the annual quantum machine learning hackathon organized by Xanadu, the team behind PennyLane, in collaboration with the Quantum Open Source Foundation. Each year the event includes a set of pre-event learning challenges designed to bring participants up to speed on quantum ML concepts and PennyLane usage before the main hackathon begins. These challenges function as a structured self-study course, and all materials including recorded lectures, problem sets, and community solutions are made available on YouTube and the QOSF GitHub after the event ends.

The 2024 learning challenges covered the landscape of quantum machine learning with hands-on PennyLane exercises. Topics included variational quantum circuits as function approximators, the construction and training of hybrid quantum-classical models, quantum kernels and their relationship to classical kernel methods in machine learning, and the design of quantum circuits for specific learning tasks. The challenge problems increase in difficulty and are designed to push intermediate learners into genuinely research-adjacent territory by the later rounds.

The QOSF community dimension of QHack is a meaningful part of its value. Solution discussions, team collaborations, and community writeups from each year’s event accumulate into a body of worked examples that goes well beyond what any single instructor could produce. For someone learning quantum ML at the intermediate level, working through the QHack 2024 challenges alongside community solutions and then comparing approaches is a particularly effective learning path.

What you’ll learn

• Variational quantum circuits: design, parameterization, and training for learning tasks
• Hybrid quantum-classical models: where to place the quantum component and how to train end-to-end
• Quantum kernels: constructing them from circuits and using them with classical support vector machines
• PennyLane workflows at an intermediate level: custom devices, noise models, and optimization strategies
• Contest-style problem solving: reading specifications carefully and implementing solutions under constraints
• Community engagement: reviewing and understanding multiple valid approaches to the same problem

Who is this for?

• Intermediate PennyLane users who want structured challenge problems beyond tutorials
• Quantum ML researchers and students looking for a community-grounded learning path
• Anyone who learns well through problem sets and wants to compare their solutions to community answers
• Developers who want exposure to the kinds of problems the quantum ML research community cares about