Due to increasing popularity and nice functionality, one should think about offering JAX support for CBX. Curretnly CBX support numpy and torch as array/tensor backends. In its current OOP style CBXPy does not directly support JAX as abackend. In particular, different slicing operations would need to be abstracted to allow a shared numpy and JAX interface.
Due to increasing popularity and nice functionality, one should think about offering JAX support for CBX. Curretnly CBX support numpy and torch as array/tensor backends. In its current OOP style CBXPy does not directly support JAX as abackend. In particular, different slicing operations would need to be abstracted to allow a shared numpy and JAX interface.