PyGAD 3.0.0 parallel processing

There was an issue when using the `parallel_processing` parameter with Keras and PyTorch. As Keras/PyTorch are not thread-safe, the `predict()` method gives incorrect and weird results when more than 1 thread is used. ahmedfgad/GeneticAlgorithmPython#145 #5 ahmedfgad/KerasGA#6. Thanks to this StackOverflow answer https://stackoverflow.com/a/75606666/5426539.

README.md

@@ -13,9 +13,10 @@ pip install pygad
 
 # Donation
 
-You can donate via [Open Collective](https://opencollective.com/pygad): [opencollective.com/pygad](https://opencollective.com/pygad). 
-
-To donate using PayPal, use either this link: [paypal.me/ahmedfgad](https://paypal.me/ahmedfgad) or the e-mail address ahmed.f.gad@gmail.com.
+- [Credit/Debit Card](https://donate.stripe.com/eVa5kO866elKgM0144): https://donate.stripe.com/eVa5kO866elKgM0144
+- [Open Collective](https://opencollective.com/pygad): [opencollective.com/pygad](https://opencollective.com/pygad)
+- PayPal: Use either this link: [paypal.me/ahmedfgad](https://paypal.me/ahmedfgad) or the e-mail address ahmed.f.gad@gmail.com
+- Interac e-Transfer: Use e-mail address ahmed.f.gad@gmail.com
 
 # Installation
 
@@ -72,7 +73,7 @@ import numpy
 function_inputs = [4,-2,3.5,5,-11,-4.7]
 desired_output = 44
 
-def fitness_func(solution, solution_idx):
+def fitness_func(ga_instance, solution, solution_idx):
     output = numpy.sum(solution*function_inputs)
     fitness = 1.0 / (numpy.abs(output - desired_output) + 0.000001)
     return fitness
@@ -153,7 +154,7 @@ import torch
 import torchga
 import pygad
 
-def fitness_func(solution, sol_idx):
+def fitness_func(ga_instance, solution, sol_idx):
     global data_inputs, data_outputs, torch_ga, model, loss_function
 
     model_weights_dict = torchga.model_weights_as_dict(model=model,
@@ -240,7 +241,7 @@ import torch
 import torchga
 import pygad
 
-def fitness_func(solution, sol_idx):
+def fitness_func(ga_instance, solution, sol_idx):
     global data_inputs, data_outputs, torch_ga, model, loss_function
 
     model_weights_dict = torchga.model_weights_as_dict(model=model,

__init__.py

@@ -1,3 +1,3 @@
 from .torchga import *
 
-__version__ = "1.1.0"
+__version__ = "1.2.0"

pytorch_pygad_XOR_classification.py

@@ -2,7 +2,7 @@
 import pygad.torchga
 import pygad
 
-def fitness_func(solution, sol_idx):
+def fitness_func(ga_instanse, solution, sol_idx):
     global data_inputs, data_outputs, torch_ga, model, loss_function
 
     predictions = pygad.torchga.predict(model=model, 
@@ -57,6 +57,7 @@ def callback_generation(ga_instance):
                        num_parents_mating=num_parents_mating, 
                        initial_population=initial_population,
                        fitness_func=fitness_func,
+                       parallel_processing=3,
                        on_generation=callback_generation)
 
 # Start the genetic algorithm evolution.

pytorch_pygad_image_classification_CNN.py

@@ -3,7 +3,7 @@
 import pygad
 import numpy
 
-def fitness_func(solution, sol_idx):
+def fitness_func(ga_instanse, solution, sol_idx):
     global data_inputs, data_outputs, torch_ga, model, loss_function
 
     predictions = pygad.torchga.predict(model=model, 

pytorch_pygad_image_classification_Dense.py

@@ -3,7 +3,7 @@
 import pygad
 import numpy
 
-def fitness_func(solution, sol_idx):
+def fitness_func(ga_instanse, solution, sol_idx):
     global data_inputs, data_outputs, torch_ga, model, loss_function
 
     predictions = pygad.torchga.predict(model=model, 

pytorch_pygad_regression.py

@@ -2,7 +2,7 @@
 import pygad.torchga
 import pygad
 
-def fitness_func(solution, sol_idx):
+def fitness_func(ga_instanse, solution, sol_idx):
     global data_inputs, data_outputs, torch_ga, model, loss_function
 
     predictions = pygad.torchga.predict(model=model, 

torchga.py

@@ -20,7 +20,7 @@ def model_weights_as_dict(model, weights_vector):
 
     start = 0
     for key in weights_dict:
-        # Calling detach() to remove the computational graph from the layer.
+        # Calling detach() to remove the computational graph from the layer. 
         # cpu() is called for making shore the data is moved from GPU to cpu
         # numpy() is called for converting the tensor into a NumPy array.
         w_matrix = weights_dict[key].cpu().detach().numpy()
@@ -41,9 +41,10 @@ def predict(model, solution, data):
                                                weights_vector=solution)
 
     # Use the current solution as the model parameters.
-    model.load_state_dict(model_weights_dict)
+    _model = copy.deepcopy(model)
+    _model.load_state_dict(model_weights_dict)
 
-    predictions = model(data)
+    predictions = _model(data)
 
     return predictions