Upgrade version

Author: xadupre

CHANGELOGS.rst

@@ -4,6 +4,7 @@ Change Logs
 0.3.0
 +++++
 
+* :pr:`93`: fixes evaluator type in ``compare_onnx_execution``
 * :pr:`92`: avoids recursion errors in profiling
 * :pr:`87`: adds command line to replace contant by ConstantOfShape
 * :pr:`79`: first draft to export to GraphBuilder

_doc/index.rst

@@ -187,5 +187,7 @@ to know onnx for that. See :ref:`l-numpy-api-onnx`.
 Older versions
 ++++++++++++++
 
+* `0.3.0 <../v0.3.0/index.html>`_
+* `0.2.0 <../v0.2.0/index.html>`_
 * `0.1.3 <../v0.1.3/index.html>`_
 * `0.1.2 <../v0.1.2/index.html>`_

_unittests/ut_npx/test_npx.py

@@ -208,7 +208,7 @@ def local1(
             return x
 
         def local2(
-            x: TensorType[ElemType.floats, "T"]
+            x: TensorType[ElemType.floats, "T"],
         ) -> TensorType[ElemType.floats, "T"]:
             return x
 

onnx_array_api/__init__.py

@@ -2,5 +2,5 @@
 APIs to create ONNX Graphs.
 """
 
-__version__ = "0.2.0"
+__version__ = "0.3.0"
 __author__ = "Xavier Dupré"

onnx_array_api/npx/npx_functions_test.py

@@ -22,21 +22,21 @@
 
 @npxapi_function
 def _min_max(
-    x: TensorType[ElemType.numerics, "T"]
+    x: TensorType[ElemType.numerics, "T"],
 ) -> TupleType[TensorType[ElemType.numerics, "T"], TensorType[ElemType.numerics, "T"]]:
     return tuple_var(var(x, op="ReduceMin"), var(x, op="ReduceMax"))
 
 
 @npxapi_inline
 def _min_max_inline(
-    x: TensorType[ElemType.numerics, "T"]
+    x: TensorType[ElemType.numerics, "T"],
 ) -> TupleType[TensorType[ElemType.numerics, "T"], TensorType[ElemType.numerics, "T"]]:
     return tuple_var(var(x, op="ReduceMin"), var(x, op="ReduceMax"))
 
 
 @npxapi_function
 def absolute(
-    x: TensorType[ElemType.numerics, "T"]
+    x: TensorType[ElemType.numerics, "T"],
 ) -> TensorType[ElemType.numerics, "T"]:
     "See :func:`numpy.absolute`."
     return var(x, op="Abs")
@@ -90,7 +90,7 @@ def log1p(x: TensorType[ElemType.floats, "T"]) -> TensorType[ElemType.floats, "T
 
 @npxapi_function
 def negative(
-    x: TensorType[ElemType.numerics, "T"]
+    x: TensorType[ElemType.numerics, "T"],
 ) -> TensorType[ElemType.numerics, "T"]:
     "See :func:`numpy.negative`."
     return var(x, op="Neg")

onnx_array_api/validation/tools.py

@@ -20,7 +20,7 @@
 
 
 def randomize_proto(
-    onx: Union[ModelProto, GraphProto, FunctionProto, NodeProto, TensorProto]
+    onx: Union[ModelProto, GraphProto, FunctionProto, NodeProto, TensorProto],
 ) -> Union[ModelProto, GraphProto, FunctionProto, NodeProto, TensorProto]:
     """
     Randomizes float initializers or constant nodes.