Clarify behavior in ConvTranspose (onnx#2343)

* Fix the wrong behavior in ConvTranspose * Address comments
usteiner9 · Sep 21, 2019 · c5af774 · c5af774
1 parent a20ba2f
commit c5af774
Show file tree

Hide file tree

Showing 3 changed files with 12 additions and 4 deletions.
diff --git a/docs/Changelog.md b/docs/Changelog.md
@@ -10762,7 +10762,7 @@ This version of the operator has been available since version 11 of the default
 <dt><tt>kernel_shape</tt> : list of ints</dt>
 <dd>The shape of the convolution kernel. If not present, should be inferred from input W.</dd>
 <dt><tt>output_padding</tt> : list of ints</dt>
-<dd>The zero-padding added to one side of the output. This is also called adjs/adjustment in some frameworks.</dd>
+<dd>Additional elements added to the side with higher coordinate indices in the output. Each padding value in "output_padding" must be less than the corresponding stride/dilation dimension. By default, this attribute is a zero vector. Note that this attribute doesn't directly affect the computed output values. It only controls the selection of the computed values, so changing this attribute only adds or removes output elements. If "output_shape" is explicitly provided, "output_padding" does not contribute additional size to "output_shape" but participates in the computation of the needed padding amount. This is also called adjs or adjustment in some frameworks.</dd>
 <dt><tt>output_shape</tt> : list of ints</dt>
 <dd>The shape of the output can be explicitly set which will cause pads values to be auto generated. If output_shape is specified pads values are ignored. See doc for details for equations to generate pads</dd>
 <dt><tt>pads</tt> : list of ints</dt>

diff --git a/docs/Operators.md b/docs/Operators.md
@@ -2948,7 +2948,7 @@ Other versions of this operator: <a href="Changelog.md#ConvTranspose-1">ConvTran
 <dt><tt>kernel_shape</tt> : list of ints</dt>
 <dd>The shape of the convolution kernel. If not present, should be inferred from input W.</dd>
 <dt><tt>output_padding</tt> : list of ints</dt>
-<dd>The zero-padding added to one side of the output. This is also called adjs/adjustment in some frameworks.</dd>
+<dd>Additional elements added to the side with higher coordinate indices in the output. Each padding value in "output_padding" must be less than the corresponding stride/dilation dimension. By default, this attribute is a zero vector. Note that this attribute doesn't directly affect the computed output values. It only controls the selection of the computed values, so changing this attribute only adds or removes output elements. If "output_shape" is explicitly provided, "output_padding" does not contribute additional size to "output_shape" but participates in the computation of the needed padding amount. This is also called adjs or adjustment in some frameworks.</dd>
 <dt><tt>output_shape</tt> : list of ints</dt>
 <dd>The shape of the output can be explicitly set which will cause pads values to be auto generated. If output_shape is specified pads values are ignored. See doc for details for equations to generate pads</dd>
 <dt><tt>pads</tt> : list of ints</dt>

diff --git a/onnx/defs/nn/defs.cc b/onnx/defs/nn/defs.cc
@@ -1299,8 +1299,16 @@ output_shape can also be explicitly specified in which case pads values are auto
         OPTIONAL);
     schema.Attr(
         "output_padding",
-        "The zero-padding added to one side of the output."
-        " This is also called adjs/adjustment in some frameworks.",
+        "Additional elements added to the side with higher coordinate indices in the output. "
+        "Each padding value in \"output_padding\" must be less than the corresponding stride/dilation dimension. "
+        "By default, this attribute is a zero vector. "
+        "Note that this attribute doesn't directly affect the computed output values. "
+        "It only controls the selection of the computed values, "
+        "so changing this attribute only adds or removes output elements. "
+        "If \"output_shape\" is explicitly provided, "
+        "\"output_padding\" does not contribute additional size to \"output_shape\" but "
+        "participates in the computation of the needed padding amount. "
+        "This is also called adjs or adjustment in some frameworks.",
         AttributeProto::INTS,
         OPTIONAL);
     schema.Attr(