fix: hotfix for a cyclic import in the keras-encoder, also added some…

… minor documentation

README.md

@@ -83,6 +83,31 @@ unproductive for an input resolution of 32x32 pixels:
 
 Keep in mind that the Graph is reverse-engineerd from the PyTorch JIT-compiler, therefore no looping-logic is
 allowed within the forward pass of the model.
+Also, the use of the functional-library for stateful-operations like pooling or convolutional layers is discouraged,
+since it can cause malformed graphs during the reverse engineering of the compute graph.
+
+#### Tensorflow / Keras
+
+Similar to the PyTorch-import, models can be imported from TensorFlow as well.
+
+The code below will create a graph of VGG16 and visualize it using GraphViz and color all layers that are predicted to be
+unproductive for an input resolution of 32x32 pixels. This is analog to the PyTorch-variant depicted above:
+
+```python
+
+from keras.applications.vgg19 import VGG19
+from rfa_toolbox import create_graph_from_tensorflow_model
+
+model = VGG19(include_top=True, weights=None)
+graph: EnrichedNetworkNode = create_graph_from_tensorflow_model(model)
+visualize_architecture(graph, "VGG16", input_res=32).view()
+
+```
+
+This will create the following visualization:
+![vgg16.PNG](https://github.com/MLRichter/receptive_field_analysis_toolbox/blob/main/images/vgg16.PNG?raw=true)
+
+Currently, only the use of the keras.Model object is supported.
 
 #### Custom
 

rfa_toolbox/__init__.py

@@ -15,7 +15,7 @@ def create_graph_from_pytorch_model(*args, **kwargs):
     from rfa_toolbox.encodings.tensorflow_keras.ingest_architecture import (
         create_graph_from_model as create_graph_from_tensorflow_model,
     )
-except ImportError:
+except ValueError:
 
     def create_graph_from_tensorflow_model(*args, **kwargs):
         raise ImportError("This function is not available, tensorflow not installed")

rfa_toolbox/encodings/tensorflow_keras/ingest_architecture.py

@@ -3,7 +3,6 @@
 
 from tensorflow.keras.models import Model
 
-from rfa_toolbox import visualize_architecture
 from rfa_toolbox.encodings.tensorflow_keras.layer_handlers import (
     AnyHandler,
     DenseHandler,
@@ -138,12 +137,3 @@ def create_graph_from_model(model: Model) -> EnrichedNetworkNode:
     """
     model_dict = keras_model_to_dict(model)
     return model_dict_to_enriched_graph(model_dict)
-
-
-if __name__ == "__main__":
-    from tensorflow.keras.applications.inception_resnet_v2 import InceptionResNetV2
-
-    graph: EnrichedNetworkNode = create_graph_from_model(
-        InceptionResNetV2(weights=None)
-    )
-    visualize_architecture(graph, "InceptionResNetV2", input_res=32).view()

rfa_toolbox/vis_script.py

@@ -170,10 +170,14 @@ def vgg19_perf2() -> EnrichedNetworkNode:
 
 
 if __name__ == "__main__":
-    m = vgg19_perf2
+    # m = vgg19_perf2
 
-    dot = visualize_architecture(m(), "vgg19_perf", input_res=16).view()
+    # dot = visualize_architecture(m(), "vgg19_perf", input_res=16).view()
 
     from keras.applications.vgg19 import VGG19
 
-    VGG19(weights=None, input_shape=(224, 224, 3), include_top=True).to_json()
+    from rfa_toolbox import create_graph_from_tensorflow_model
+
+    model = VGG19(include_top=True, weights=None)
+    graph: EnrichedNetworkNode = create_graph_from_tensorflow_model(model)
+    visualize_architecture(graph, "VGG16", input_res=32).view()