Image2DModel and Image3DModel docs rendered without text

[LucaMarconato]

This seems to have happened in the latest modification to the docs. See more details here: https://scverse.zulipchat.com/#narrow/channel/443514-spatialdata-dev/topic/docs/near/493413340.

models.mkd doesn't show any obvious reasons for the different results in the docs.

# Models

The elements (building-blocks) that constitute `SpatialData`.

```{eval-rst}
.. currentmodule:: spatialdata.models

.. autoclass:: Image2DModel
   :members:
   :undoc-members:
.. autoclass:: Image3DModel
   :members:
   :undoc-members:
.. autoclass:: Labels2DModel
   :members:
   :undoc-members:
.. autoclass:: Labels3DModel
   :members:
   :undoc-members:
.. autoclass:: ShapesModel
   :members:
   :undoc-members:
.. autoclass:: PointsModel
   :members:
   :undoc-members:
.. autoclass:: TableModel
   :members:
   :undoc-members:

[quentinblampey]

Hi @LucaMarconato, I think the docstrings for parse are not displaying because autoclass:: may not be looking at the dosctrings of the parent method (i.e., RasterSchema.parse). I'm not an expert in sphinx but I can check if there is an attribute to autoclass:: that does that

[LucaMarconato]

I now noticed that Labels2DModel and Labels3DModel were overriding .parse() from the base class; this explains why the docs were correct for those. Thanks for the fix!