refactor: let user pass base function and not implement load method b…

…ecause it is always the same

src/piven/models/mlp_regressor.py

@@ -51,7 +51,7 @@ def check_model_params(
 
 
 # Make build function for the model wrapper
-def piven_model(
+def piven_mlp_model(
     input_dim, dense_units, dropout_rate, lambda_, bias_init_low, bias_init_high, lr
 ):
     model = build_keras_piven(
@@ -74,7 +74,7 @@ class PivenMlpModel(PivenBaseModel):
     def build(self, preprocess: Union[None, Pipeline, TransformerMixin] = None):
         # All build params are passed to init and should be checked here
         check_model_params(**self.params)
-        model = PivenKerasRegressor(build_fn=piven_model, **self.params)
+        model = PivenKerasRegressor(build_fn=piven_mlp_model, **self.params)
         if preprocess is None:
             pipeline = Pipeline([("model", model)])
         else:
@@ -84,11 +84,3 @@ def build(self, preprocess: Union[None, Pipeline, TransformerMixin] = None):
             regressor=pipeline, transformer=StandardScaler()
         )
         return self
-
-    @classmethod
-    def load(cls, path: str):
-        model_config = PivenMlpModel.load_model_config(path)
-        model = PivenMlpModel.load_model_from_disk(piven_model, path)
-        run = cls(**model_config)
-        run.model = model
-        return run

tests/test_mlp_model.py

@@ -3,7 +3,7 @@
 import pytest
 import tensorflow as tf
 import numpy as np
-from piven.models import PivenMlpModel
+from piven.models import PivenMlpModel, piven_mlp_model
 from sklearn.pipeline import Pipeline
 from sklearn.preprocessing import StandardScaler, OneHotEncoder
 from sklearn.compose import ColumnTransformer
@@ -104,7 +104,7 @@ def test_experiment_io(self, mock_data, experiment):
         with tempfile.TemporaryDirectory() as tmpdir:
             experiment.save(tmpdir)
             assert (Path(tmpdir) / "experiment_params.json").is_file()
-            _ = PivenMlpModel.load(path=tmpdir)
+            _ = PivenMlpModel.load(path=tmpdir, build_fn=piven_mlp_model)
 
     def test_experiment_scoring(self, mock_data, experiment):
         x_train, x_valid, y_train, y_valid = mock_data