chore: add weights clipping

docs/advanced_examples/LogisticRegressionTraining.ipynb

@@ -372,7 +372,7 @@
     "\n",
     "batch_size = model_concrete_partial.batch_size\n",
     "\n",
-    "classes = np.unique(y2_train)\n",
+    "classes = np.unique(y_train)\n",
     "\n",
     "# Go through the training batches\n",
     "accuracy_scores = []\n",

src/concrete/ml/sklearn/_fhe_training_utils.py

@@ -1,6 +1,6 @@
 """Utility functions for FHE training."""
 
-from typing import Tuple
+from typing import List, Optional, Tuple
 
 import numpy
 import torch
@@ -29,7 +29,12 @@ class LogisticRegressionTraining(torch.nn.Module):
     The forward function iterates the SGD over a given certain number of times.
     """
 
-    def __init__(self, iterations: int = 1, learning_rate: float = 1.0, fit_bias: bool = True):
+    def __init__(
+        self,
+        iterations: int = 1,
+        learning_rate: float = 1.0,
+        fit_bias: bool = True,
+    ):
         """Instantiate the model.
 
         Args:

src/concrete/ml/sklearn/linear_model.py

@@ -253,7 +253,7 @@ def __init__(
 
             if self.parameters_range is None:
                 raise ValueError(
-                    "Setting 'parameter_range' is mandatory if FHE training is enabled "
+                    "Setting 'parameters_range' is mandatory if FHE training is enabled "
                     f"({fit_encrypted=}). Got {parameters_range=}"
                 )
 
@@ -358,6 +358,7 @@ def _get_training_quantized_module(
         # Enable the underlying FHE circuit to be composed with itself
         # This feature is used in order to be able to iterate in the clear n times without having
         # to encrypt/decrypt the weight/bias values between each loop
+        # configuration = Configuration(composable=True, detect_overflow_in_simulation=False)
         configuration = Configuration(composable=True)
 
         # Compile the model using the compile set
@@ -446,7 +447,6 @@ def _fit_encrypted(
                     f" was: {self.classes_}"
                 )
 
-
         n_samples, n_features = X.shape
         weight_shape = (1, n_features, 1)
         bias_shape = (1, 1, 1)