Sample Flux Layer

This example demonstrates the creation of a sample Flux layer.

Example

This example shows test code that uses the Flux.gradient() call to calculate the gradients of the Circuit object.

using Flux
using Qaintessent
using Qaintellect
using LinearAlgebra
using IterTools: ncycle

# construct parametrized circuit
rx = RxGate(2π*rand())
ry = RyGate(2π*rand())

cgc = [
    circuit_gate(1, rx),
    circuit_gate(1, ry),
]

# using Pauli-Z matrix as observable
meas = [MeasurementOperator([1 0; 0 -1], (1,))]

c = Circuit{1}(cgc, meas)

# We set the initial input as the $\lvert 0 \rangle$ state, and the target expectation value of the measurement as $-1$.
ψ = ComplexF64[1, 0]
e = -1

# create loss function: note that circuit `c` is applied to `x`
loss(x, y) = Flux.mse(c(x), y)

# gather parameters from Circuit
paras = Flux.params(c)
@show(paras)

# define optimizer
opt = ADAM(0.5)

# set up data for training; using `ncycle()` to repeatedly feed the input quantum state into the training algorithm
data = ncycle([(ψ, e)], 128)

# define evaluation function
evalcb() = @show(loss(ψ, e))

# example: compute gradients
grads = gradient(() -> loss(ψ, e), paras)
grads[ry.θ]

Flux.train!(loss, paras, data, opt, cb=Flux.throttle(evalcb, 0.01))

# verify output
@show apply(c, ψ)