Tests with PennyLane and JAX.
jaxopt_run.py
: simple sine-function learning of a VQC withpennylane
,jax.jit
andjaxopt.GradientDescent
. The.run()
method is called on the optimizer.jaxopt_optimization_loop.py
: same asjaxopt.py
, but breaks down the optmization loop, rather than calling therun()
method on the optimizer object.optax_full_set.py
: same asjaxopt_optimization_loop.py
, but usesoptax.adam
as optimizer and updates with the full dataset.optax_batch.py
: same asoptax_full_set.py
but stochastic: uses a data iterator to batchX
andy
.optax_classification.py
: classification of the iris dataset withoptax
. Optimization part is the same asoptax_batch.py
, but the data now is preprocessed and converted from NumPy to JAX.optax_kta.py
: optimizes the kernel alignment of a quantum kernel wrt the (binary) iris dataset.