REF: Inherit base model class from ABC to enforece native abstraction

Inherit base model class from `ABC` to enforece native abstraction.

Fixes:
```
Abstract methods are allowed in classes whose metaclass is 'ABCMeta'
```

raised locally by the IDE.

Author: jhlegarreta

src/nifreeze/model/base.py

@@ -22,7 +22,7 @@
 #
 """Base infrastructure for nifreeze's models."""
 
-from abc import abstractmethod
+from abc import ABC, ABCMeta, abstractmethod
 from warnings import warn
 
 import numpy as np
@@ -76,7 +76,7 @@ def init(model: str | None = None, **kwargs):
         raise NotImplementedError(f"Unsupported model <{model}>.")
 
 
-class BaseModel:
+class BaseModel(ABC):
     """
     Defines the interface and default methods.
 
@@ -87,6 +87,8 @@ class BaseModel:
 
     """
 
+    __metaclass__ = ABCMeta
+
     __slots__ = ("_dataset", "_locked_fit")
 
     def __init__(self, dataset, **kwargs):
@@ -99,7 +101,7 @@ def __init__(self, dataset, **kwargs):
             warn(mask_absence_warn_msg, stacklevel=2)
 
     @abstractmethod
-    def fit_predict(self, index: int | None = None, **kwargs) -> np.ndarray:
+    def fit_predict(self, index: int | None = None, **kwargs):
         """
         Fit and predict the indicated index of the dataset (abstract signature).
 
@@ -115,7 +117,7 @@ def fit_predict(self, index: int | None = None, **kwargs) -> np.ndarray:
             If ``None``, no prediction will be executed.
 
         """
-        raise NotImplementedError("Cannot call fit_predict() on a BaseModel instance.")
+        return
 
 
 class TrivialModel(BaseModel):