Update for flatbuffers python wrappers

wrappers/python/model_microservice.py

@@ -134,31 +134,31 @@ def __init__(self,user_model):
         super(SeldonFlatbuffersServer, self).__init__()
         self.user_model = user_model
 
-    async def handle_stream(self, stream, address):
+    @gen.coroutine
+    def handle_stream(self, stream, address):
         while True:
             try:
-                data = await stream.read_bytes(4)
+                data = yield stream.read_bytes(4)
                 obj = struct.unpack('<i',data)
                 len_msg = obj[0]
-                data = await stream.read_bytes(len_msg)
+                data = yield stream.read_bytes(len_msg)
                 try:
                     features,names = SeldonRPCToNumpyArray(data)
                     predictions = np.array(predict(self.user_model,features,names))
                     if len(predictions.shape)>1:
-                        print(predictions.shape)
                         class_names = get_class_names(self.user_model, predictions.shape[1])
                     else:
                         class_names = []
                     outData = NumpyArrayToSeldonRPC(predictions,class_names)
-                    await stream.write(outData)
+                    yield stream.write(outData)
                 except StreamClosedError:
                     print("Stream closed during processing:",address)
                     break
                 except Exception:
                     tb = traceback.format_exc()
                     print("Caught exception during processing:",address,tb)
                     outData = CreateErrorMsg(tb)
-                    await stream.write(outData)
+                    yield stream.write(outData)
                     stream.close()
                     break;
             except StreamClosedError:

wrappers/python/seldon_flatbuffers.py

@@ -38,7 +38,6 @@ def SeldonRPCToNumpyArray(data):
                 shape = []
                 for i in range(tensor.ShapeLength()):
                     shape.append(tensor.Shape(i))
-                    print(shape)
                 values = tensor.ValuesAsNumpy()
                 values = values.reshape(shape)
                 return (values,names)
@@ -130,7 +129,7 @@ def CreateNumpyVector(builder, x):
     if not isinstance(x, np.ndarray):
         raise TypeError("non-numpy-ndarray passed to CreateNumpyVector")
 
-    if x.data.ndim > 1:
+    if x.ndim > 1:
         raise TypeError("multidimensional-ndarray passed to CreateNumpyVector")
 
     builder.StartVector(x.itemsize, x.size, x.dtype.alignment)
@@ -139,7 +138,7 @@ def CreateNumpyVector(builder, x):
     if x.dtype.str[0] == "<":
         x_lend = x
     else:
-        x_lend = x.byteswap()
+        x_lend = x.byteswap(inplace=False)
 
     # Calculate total length
     l = UOffsetTFlags.py_type(x_lend.itemsize * x_lend.size)

wrappers/s2i/python/build_python2.sh

@@ -0,0 +1 @@
+make build

wrappers/s2i/python/test_fbs/.s2i/environment

@@ -0,0 +1,4 @@
+MODEL_NAME=Test
+API_TYPE=FBS
+SERVICE_TYPE=MODEL
+PERSISTENCE=0

wrappers/s2i/python/test_fbs/READNE.txt

@@ -0,0 +1 @@
+A simple dummy model to test wrapping.

wrappers/s2i/python/test_fbs/Test.py

@@ -0,0 +1,23 @@
+class Test(object):
+    """
+    Model template. You can load your model parameters in __init__ from a location accessible at runtime
+    """
+
+    def __init__(self):
+        """
+        Add any initialization parameters. These will be passed at runtime from the graph definition parameters defined in your seldondeployment kubernetes resource manifest.
+        """
+        print("Initializing")
+
+    def predict(self,X,features_names):
+        """
+        Return a prediction.
+
+        Parameters
+        ----------
+        X : array-like
+        feature_names : array of feature names (optional)
+        """
+        print("Predict called - will run identity function")
+        print(X)
+        return X

wrappers/s2i/python/test_fbs/contract.json

@@ -0,0 +1,39 @@
+{
+    "features":[
+	{
+	    "name":"sepal_length",
+	    "dtype":"FLOAT",
+	    "ftype":"continuous",
+	    "range":[4,8]
+	},
+	{
+	    "name":"sepal_width",
+	    "dtype":"FLOAT",
+	    "ftype":"continuous",
+	    "range":[2,5]
+	},
+	{
+	    "name":"petal_length",
+	    "dtype":"FLOAT",
+	    "ftype":"continuous",
+	    "range":[1,10]
+	},
+	{
+	    "name":"petal_width",
+	    "dtype":"FLOAT",
+	    "ftype":"continuous",
+	    "range":[0,3]
+	}
+    ],
+    "targets":[
+	{
+	    "name":"class",
+	    "dtype":"FLOAT",
+	    "ftype":"continuous",
+	    "range":[0,1],
+	    "repeat":3
+	}
+    ]
+}
+
+

wrappers/s2i/python/test_fbs/requirements.txt

@@ -0,0 +1 @@
+numpy

wrappers/s2i/python/test_fbs/run_test_python2

@@ -0,0 +1,6 @@
+
+s2i build . seldonio/seldon-core-s2i-python2:0.2-SNAPSHOT test-wrapper-py2:0.1
+docker run --name "test_predictor" --rm -d -p 5000:5000 test-wrapper-py2:0.1
+sleep 2
+python ../../../testing/tester.py contract.json 0.0.0.0 5000 -p --fbs
+docker rm -f test_predictor

wrappers/s2i/python/test_fbs/run_test_python3

@@ -0,0 +1,6 @@
+
+s2i build . seldonio/seldon-core-s2i-python3:0.2-SNAPSHOT test-wrapper-py3:0.1
+docker run --name "test_predictor" --rm -d -p 5000:5000 test-wrapper-py3:0.1
+sleep 2
+python ../../../testing/tester.py contract.json 0.0.0.0 5000 -p --fbs
+docker rm -f test_predictor

wrappers/testing/tester_flatbuffers.py

@@ -74,7 +74,6 @@ def SeldonRPCToNumpyArray(data):
             shape = []
             for i in range(tensor.ShapeLength()):
                 shape.append(tensor.Shape(i))
-                print(shape)
             values = tensor.ValuesAsNumpy()
             values = values.reshape(shape)
             return (values,names)