Fix JAX eigh

[abhinavd]

This is a bug coming from a commit on JAX, namely 012c5bd. The eigenvalues and eigenvectors are wrapped in a container object that needs to be unpacked. Otherwise, it causes an error down the line, like the following:

Custom VJP fwd rule jaxeigh_fwd for function adaware_eigh must produce a pair (list or tuple of length two) where the first element represents the primal output (equal to the output of the custom_vjp-decorated function adaware_eigh) and the second element represents residuals (i.e. values stored from the forward pass for use on the backward pass), but instead the fwd rule output's first element had container/pytree structure:
(float32[321, complex64[32, 32])
while the custom_vjp-decorated function adaware_eigh had output container/pytree structure:
EighResult (eigenvalues=float32[32], eigenvectors=complex64[32,32]) .

[refraction-ray]

One problem of the change proposed here is that it is not compatible with older versions of Jax, where the result is just a tuple?
What about e, v = jnp.linalg.eigh(A); return e, v, will this change be compatible with both tuple and namedtuple?

[abhinavd]

Ah, yes, that would work! Shall I update it?

[codecov]

Codecov Report

Attention: Patch coverage is 57.14286% with 3 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
tensorcircuit/backends/jax_ops.py	57.14%	3 Missing ⚠️

Files with missing lines	Coverage Δ
tensorcircuit/backends/jax_ops.py	82.72% <57.14%> (-1.09%)	⬇️

[refraction-ray]

LGTM, thank you for the contribution!