PennyLane v0.12 has arrived! 🤩
- New mixed-state simulator
- Minimize the number of measurement settings for HW
- Access the state vector from circuits
- Quantum noisy channels
- Preview the next-generation QNode
And more!
2) We’re excited to introduce `default.mixed` 🔊, a mixed state implementation of our default qubit simulator, ideal for simulating and optimizing circuits that include noise and quantum channels.
Use it alongside our new noise channels.
3) Our new `qml.grouping` module makes it easy to group simultaneously measurable observables.
Use the new `optimize_measurements` function to reduce the number of measurements needed on quantum hardware during optimization.
4) QNodes keep getting better!
You can now return the quantum state from supported devices via the convenient `qml.state()` return function. It even supports differentiation when using backprop!
5) The next-generation QNode is now in preview and available for testing! Try out the new ‘tape-mode’ 📼 for in-QNode classical processing, less restrictions, and significant speed improvements.
This new mode is experimental; feedback is welcomed!
6) 🔌 QNode caching has been introduced, allowing PennyLane to keep track of the results of previous device executions and reuse those results in subsequent calls.
Useful when rapid prototyping to reduce simulator overhead!
7) The Strawberry Fields plugin has a new device: , allowing simulated photonic circuits to be trained faster via
@TensorFlow
and backpropagation.
Full autodifferentiation, now available in the Fock basis ❤️
8) The ecosystem of connected quantum platforms keeps expanding!
In addition to
@qiskit
,
@rigetti
Forest,
@GoogleAI
Cirq,
@Honeywell
+ more, you can now include neutral-atom devices from
@pasqalio
directly in your QML models.