Modify Fock backend, supporting MCMC measurement

[Jooyuza]

Fock backend modified and new feature added:

Parameter is_prob of forward ,forward_fockhas been changed to:
is_prob (None or bool, optional): Choices of return type. Set None to return entire unitary matrix,
Set True to return probabilities for Fock basis states;
Set False to return amplitude dictionary of final state.
Default: None
Returns:
Union[torch.Tensor, Dict]:
Tensor of unitary matrix(is_prob = None), Dictionary of final state(is_prob = False),
or Dictionary of probabilty distribution(is_prob = True),
of the photonic quantum circuit after applying the operators.

Fock backend now supports：

1. mcmc sampling;
2. batch-friendly features: when QumodeCircuit is encoded with batched data, forward function will have a batched return and measure function will generate batched outcome.

Brief notebook demo:"\docs\test_fockbackend_is_prob.ipynb"

[sansiro77]

Fock后端的measure根据forward输出是unitary、dict、tensor，分成三个

[sansiro77]

我对docstring和代码做了些调整，可以再检查一下，再尝试支持处理带batch的初态，然后测试的ipynb别忘了删

[sansiro77]

Good job