`QCNN` backend issues
Opened this issue · 0 comments
BrunoLiegiBastonLiegi commented
The QCNN
doesn't allow to pass a backend, but always makes use of the GlobalBackend
Line 12 in 0d86ec5
In the tests, indeed, only the GlobalBackend
is tested which, by chance, was working (I'm guessing the numpy one was used).
Some quick test I ran showed that the TensorflowBackend
might have problems with it, and probably other backends as well.