BroadcastTo関連のコードを移動し、pads関連のエクスポート時のコードがおかしかったのを修正

sony · Apr 25, 2018 · 1f54698 · 1f54698
1 parent c46d264
commit 1f54698
Show file tree

Hide file tree

Showing 6 changed files with 384 additions and 319 deletions.
diff --git a/build-tools/code_generator/function_types.yaml b/build-tools/code_generator/function_types.yaml
@@ -232,6 +232,9 @@ Transpose:
 Broadcast:
   float: [float]
   half: [Half]
+BroadcastTo:
+  float: [float]
+  half: [Half]
 OneHot:
   float: [int, float]
   half: [int, Half]

diff --git a/build-tools/code_generator/functions.yaml b/build-tools/code_generator/functions.yaml
@@ -1827,6 +1827,24 @@ Array Manipulation:
     outputs:
       y:
         doc: Broadcasted N-D array
+  BroadcastTo:
+    snake_name: broadcast_to
+    doc: |2
+
+      Broadcasting ND-array to the specified buffer.
+    inputs:
+      x:
+        doc: N-D array
+      y:
+        doc: N-D array
+    arguments:
+      axis:
+        doc: Target axis to start broadcasting. If this is not set, broadcast will try to fit y to x starting from the last dimension
+        type: int64
+        default: -1
+    outputs:
+      z:
+        doc: Broadcasted N-D array
   OneHot:
     snake_name: one_hot
     doc: |2

diff --git a/python/src/nnabla/utils/converter/functions.yaml b/python/src/nnabla/utils/converter/functions.yaml
@@ -1827,6 +1827,24 @@ Array Manipulation:
     outputs:
       y:
         doc: Broadcasted N-D array
+  BroadcastTo:
+    snake_name: broadcast_to
+    doc: |2
+
+      Broadcasting ND-array to the specified buffer.
+    inputs:
+      x:
+        doc: N-D array
+      y:
+        doc: N-D array
+    arguments:
+      axis:
+        doc: Target axis to start broadcasting. If this is not set, broadcast will try to fit y to x starting from the last dimension
+        type: int64
+        default: -1
+    outputs:
+      z:
+        doc: Broadcasted N-D array
   OneHot:
     snake_name: one_hot
     doc: |2

diff --git a/python/src/nnabla/utils/converter/onnx/exporter.py b/python/src/nnabla/utils/converter/onnx/exporter.py
@@ -113,7 +113,7 @@ def convert_to_nodes(func, variables, input_types, output_types, broadcast_targe
         # Copy kernel, stride, and pads values
         k = onnx.helper.make_attribute("kernel_shape", mpp.kernel.dim)
         s = onnx.helper.make_attribute("strides", mpp.stride.dim)
-        p = onnx.helper.make_attribute("pads", mpp.pad.dim*2)
+        p = onnx.helper.make_attribute("pads", mpp.pad.dim[:]*2)
         n.attribute.extend([k, s, p])
         nl.append(n)
     elif func.type == "Convolution":
@@ -138,7 +138,7 @@ def convert_to_nodes(func, variables, input_types, output_types, broadcast_targe
                                        weight_shape.dim[weight_base:])
         d = onnx.helper.make_attribute("dilations", cp.dilation.dim)
         s = onnx.helper.make_attribute("strides", cp.stride.dim)
-        p = onnx.helper.make_attribute("pads", cp.pad.dim*2)
+        p = onnx.helper.make_attribute("pads", cp.pad.dim[:]*2)
         g = onnx.helper.make_attribute("group", cp.group)
         n.attribute.extend([k, d, s, p, g])
         nl.append(n)
@@ -170,7 +170,7 @@ def convert_to_nodes(func, variables, input_types, output_types, broadcast_targe
         # Copy kernel, stride, and pads values
         k = onnx.helper.make_attribute("kernel_shape", app.kernel.dim)
         s = onnx.helper.make_attribute("strides", app.stride.dim)
-        p = onnx.helper.make_attribute("pads", app.pad.dim*2)
+        p = onnx.helper.make_attribute("pads", app.pad.dim[:]*2)
         n.attribute.extend([k, s, p])
         nl.append(n)
     elif func.type == "BatchNormalization":