Methods to save/load the network/model/weights

conx/network.py

@@ -26,10 +26,12 @@
 from functools import reduce
 import signal
 import numbers
+import pickle
 import base64
 import html
 import copy
 import io
+import os
 import PIL
 
 import numpy as np
@@ -1600,30 +1602,77 @@ def describe_connection_to(self, layer1, layer2):
         ## FIXME: how to show merged layer weights?
         return retval
 
-    def save(self, filename=None):
+    def save(self, foldername=None):
         """
-        Save the weights to a file.
+        Save the network to a folder.
         """
-        if filename is None:
-            filename = "%s.wts" % self.name
-        with open(filename, "wb") as fp:
-            for layer in self.model.layers:
-                for weight in layer.get_weights():
-                    np.save(fp, weight)
+        if foldername is None:
+            foldername = "%s.conx" % self.name
+        if not os.path.isdir(foldername):
+            os.makedirs(foldername)
+        if self.model:
+            self.save_model(foldername)
+            self.save_weights(foldername)
+        self.model, tmp_model = None, self.model
+        with open("%s/network.pickle" % foldername, "wb") as fp:
+            pickle.dump(self, fp)
+        self.model = tmp_model
+
+    @classmethod
+    def load(cls, foldername):
+        """
+        Load the network from a folder.
+        """
+        with open("%s/network.pickle" % foldername, "rb") as fp:
+            net = pickle.load(fp)
+        net.load_model(foldername)
+        net.load_weights(foldername)
+        return net
+
+    def save_weights(self, foldername=None):
+        """
+        Save the model weights to a folder.
+        """
+        if self.model:
+            if foldername is None:
+                foldername = "%s.conx" % self.name
+            if not os.path.isdir(foldername):
+                os.makedirs(foldername)
+            self.model.save_weights("%s/weights.h5" % foldername)
+        else:
+            raise Exception("need to compile network first")
+
+    def save_model(self, foldername=None):
+        """
+        Save the model to a folder.
+        """
+        if self.model:
+            if foldername is None:
+                foldername = "%s.conx" % self.name
+            if not os.path.isdir(foldername):
+                os.makedirs(foldername)
+            self.model.save("%s/model.h5" % foldername)
+        else:
+            raise Exception("need to compile network first")
+
+    def load_weights(self, foldername=None):
+        """
+        Load the model weights from a folder.
+        """
+        if self.model:
+            if foldername is None:
+                foldername = "%s.conx" % self.name
+            if os.path.isfile("%s/model.h5" % foldername):
+                self.model.load_weights("%s/weights.h5" % foldername)
 
-    def load(self, filename=None):
+    def load_model(self, foldername=None):
         """
-        Load the weights from a file.
+        Load and set the model from a folder.
         """
-        if filename is None:
-            filename = "%s.wts" % self.name
-        with open(filename, "rb") as fp:
-            for layer in self.model.layers:
-                weights = layer.get_weights()
-                new_weights = []
-                for w in range(len(weights)):
-                    new_weights.append(np.load(fp))
-                layer.set_weights(new_weights)
+        if foldername is None:
+            foldername = "%s.conx" % self.name
+        if os.path.isfile("%s/model.h5" % foldername):
+            self.model = keras.models.load_model("%s/model.h5" % foldername)
 
     def get_inputs_length(self):
         """