Martin Ganahl

e.g. `eigs`

improve testing of symmetric backend

2

testing of the symmetric backend is currently done independently of the other backends. We should pull all these tests together. A good way would be to build a custom test...

also add it to tn.expm (including tests)

repr for Tensor

8

We need to add a useful `__repr__` to `Tensor`

Tensor initialization with dtype and pytorch backend breaks

8

`Tensor` initialization methods for pytorch currently break for this type of initialization: ```python t = tn.random_uniform((2,2), dtype=np.float64, backend='pytorch') ``` (works for JAX and tensorflow) @alewis can you take a look?