Update several links (#846)

Author: guspan-tanadi

docs/SpecCodingConventions.md

@@ -151,6 +151,6 @@ Example:
 ### Dictionary Members
 
 * Dictionary members are referenced using dotted property syntax. e.g. _options.padding_
-   * Note that this is contrary to Web IDL + Infra; formally, a JavaScript object has been mapped to a Web IDL [dictionary](https://webidl.spec.whatwg.org/#idl-dictionaries) and then processed into an Infra [map](ordered) by the time a spec is using it. So formally the syntax _options["padding"]_ should be used.
+   * Note that this is contrary to Web IDL + Infra; formally, a JavaScript object has been mapped to a Web IDL [dictionary](https://webidl.spec.whatwg.org/#idl-dictionaries) and then processed into an Infra [map](https://infra.spec.whatwg.org/#ordered-map) by the time a spec is using it. So formally the syntax _options["padding"]_ should be used.
 * Dictionary members should be linked to, both in algorithms and in other text. e.g. `|options|.{{MLOptionsDict/member}}` (in the steps for an algorithm) or `{{MLOptionsDict/member}}` (outside an algorithm).
 * Dictionary members should be given definitions somewhere in the text. This is usually done with a `<dl dfn-type=dict-member dfn-for=...>` for the dictionary as a whole, containing a `<dfn>` for each member.

mltensor-explainer.md

@@ -2,7 +2,7 @@
 
 ## Authors
 
-- [Austin Sullivan]([email protected]) (Google)
+- [Austin Sullivan](mailto:[email protected]) (Google)
 
 ## Participate
 
@@ -419,4 +419,4 @@ partial interface MLContext {
 partial interface ML {
   Promise<MLContext> createContext(GPUDevice device);
 };
-```
+```

op_compatibility/conv2d.md

@@ -2,9 +2,9 @@ Tracking issue: https://github.com/webmachinelearning/webnn/issues/28
 
 |  WebNN | NNAPI | MPS | DirectML | BNNS | DNNL (MKL-DNN) | ONNX | Paddle-Lite |
 |------|---------|------|------------|--------|---------------------|---|---|
-|  **conv2d** | [ANEURALNETWORKS_CONV_2D](https://developer.android.com/ndk/reference/group/neural-networks#group___neural_networks_1ggaabbe492c60331b13038e39d4207940e0a34a73b5eaf458b67db5eda71557d1d01)  | [MPSCNNConvolution](https://developer.apple.com/documentation/metalperformanceshaders/mpscnnconvolution) | [DML_CONVOLUTION_OPERATOR_DESC](https://docs.microsoft.com/en-us/windows/win32/api/directml/ns-directml-dml_convolution_operator_desc) when **DimensionCount** = 2, **GroupCount** = 1, **Mode** = DML_CONVOLUTION_MODE_CROSS_CORRELATION, **Direction** = DML_CONVOLUTION_DIRECTION_FORWARD | [BNNSFilterCreateConvolutionLayer](https://developer.apple.com/documentation/accelerate/1642537-bnnsfiltercreateconvolutionlayer?language=objc) | [convolution_forward](https://intel.github.io/mkl-dnn/structdnnl_1_1convolution__forward.html) | [Conv](https://github.com/onnx/onnx/blob/master/docs/Operators.md#Conv) | [conv2d](https://www.paddlepaddle.org.cn/documentation/docs/en/api/layers/conv2d.html) |
+|  **conv2d** | [ANEURALNETWORKS_CONV_2D](https://developer.android.com/ndk/reference/group/neural-networks#group___neural_networks_1ggaabbe492c60331b13038e39d4207940e0a34a73b5eaf458b67db5eda71557d1d01)  | [MPSCNNConvolution](https://developer.apple.com/documentation/metalperformanceshaders/mpscnnconvolution) | [DML_CONVOLUTION_OPERATOR_DESC](https://docs.microsoft.com/en-us/windows/win32/api/directml/ns-directml-dml_convolution_operator_desc) when **DimensionCount** = 2, **GroupCount** = 1, **Mode** = DML_CONVOLUTION_MODE_CROSS_CORRELATION, **Direction** = DML_CONVOLUTION_DIRECTION_FORWARD | [BNNSFilterCreateConvolutionLayer](https://developer.apple.com/documentation/accelerate/1642537-bnnsfiltercreateconvolutionlayer?language=objc) | [convolution_forward](https://uxlfoundation.github.io/oneDNN/struct_dnnl_convolution_forward.html) | [Conv](https://github.com/onnx/onnx/blob/master/docs/Operators.md#Conv) | [conv2d](https://www.paddlepaddle.org.cn/documentation/docs/en/api/paddle/nn/functional/conv2d_en.html) |
 | **x:** 4-dim tensor with logical shape [N, H, W, in_channels] | **input[0]:** 4-dim tensor of shape [N, H, W, in_channels] (default) or [N, in_channels, H, W] with **input[10]** (explicit padding) = True | **sourceImage** of MPSCNNKernel.encodeToCommandBuffer: 4-dim tensor in [MPSImage](https://developer.apple.com/documentation/metalperformanceshaders/mpsimage?language=objc)<br> described by MPSImageDescriptor {numberOfImages, featureChannels, width, height} | **InputTensor** of DML_CONVOLUTION_OPERATOR_DESC: 4-dim tensor in buffer resource of shape [N, in_channels, H, W] | **in_desc**: 4-dim tensor of shape [N, in_channels, H, W] | **src_desc** of convolution_forward::desc: 4-dim tensor of shape [N, in_channels, H, W] | **X**: 4-dim tensor of shape [N, in_channels, H, W] | **input**: 4-D tensor with shape [N, C, H, W] (default) or [N, H, W, C] with **data_format** = "NHWC" |
 | **filter:** 4-dim tensor with logical shape [H, W, in_channels, out_channels] | **inputs[1]:** 4-dim tensor of shape [out_channels, H, W, in_channels] | **weights** of MPSCNNConvolutionDataSource:<br>4-dim tensor of shape [out_channels, H, W, in_channels] | **FilterTensor** of DML_CONVOLUTION_OPERATOR_DESC: 4-dim tensor of buffer resource of shape [out_channels, in_channels, H, W] | **weigths** of BNNSConvolutionLayerParameters: 4-dim tensor of shape [out_channels, in_channels, H, W] | **weights_desc** of convolution_forward::desc: 4-dim tensor of shape [out_channels, in_channels, H, W] | **W**: 4-dim tensor of shape [out_channels, in_channels, H, W] with **group** = 1 | **filter**: 4-D tensor (MCHW) of shape [depth_out, depth_in, filter_height, filter_width] |
 | **padding:** array of ints. Padding for the beginning and ending along each spatial axis. | (explicit padding) **input[3:6]:** int, padding to beginning_W, ending_W, beginning_H, ending_H | **offset** of MPSCNNKernel: in MPSOffset {x, y, z}, int, offset (-padding) for beginning_W and beginning_H.<br>**Open:** padding for ending? | **StartPadding** of DML_CONVOLUTION_OPERATOR_DESC: unsigned int, padding to the start of the corresponding axis [beginning_H, beginning_W] <br>**EndPadding** of DML_CONVOLUTION_OPERATOR_DESC: unsigned int, padding to the end of corresponding axis [ending_H, ending_W] | **x_padding** of BNNSConvolutionLayerParameters: unsigned int, beginning and ending padding of width axis <br> **y_padding** of BNNSConvolutionLayerParameters:  unsigned int, beginning and ending padding of height axis <br> **Open**: only support same beginning and ending paddings along each spatial axis. | **padding_l** of  of convolution_forward::desc: int64, [beginning_H, beginning_W] <br> **padding_r** of convolution_forward::desc: int64, [ending_H, ending_W] | **pads**: list of ints, padding for the beginning and ending along each spatial axis, [beginning_H, beginning_W, ending_H, ending_W] with **auto_pad** = NOTSET | **padding**: list of ints, [pad_height_top, pad_height_bottom, pad_width_left, pad_width_right] |
 |  **strides:** array of ints that have length 1, 2 or 4. The stride of the sliding window for each dimension of input. | **input[7:8]:** int, stride_W and stride_H <br> **Open:** support strides of N and C? | **strideInPixelsX** and **strideInPixelsY** of MPSCNNConvolutionDescriptor: unsigned int, stride_W and stride_H <br> **Open:** support strides of N and C? | **Strides** of DML_CONVOLUTION_OPERATOR_DESC: unsigned int, [stride_H, stride_W] | **x_stride** and **y_stride** of BNNSConvolutionLayerParameters: unsigned int, stride_W and stride_H <br> **Open:** support strides of N and C?  | **strides** of convolution_forward::desc: int64, stride_W, stride_H <br> **Open:** support strides of N and C?  | **strides**: list of ints, stride along each spatial axis | **stride**: tuple of ints,  (stride_height, stride_width) |
-|  **dilations:** array of ints of length 2: [dil_height, dil_width] The dilation factor for each spatial dimension of the input. | **input[11:12]:** int, dilation_width and dilation_height | **dilationRateX** and  **dilationRateY** of MPSCNNConvolutionDescriptor: unsigned int, dilation_width and dilation_height | **Dilations** of DML_CONVOLUTION_OPERATOR_DESC: unsigned int, [dilation_H, dilation_W] | **Open**: not supported | **dilates** of convolution_forward::desc: int64, [dilation_width, dilation_height] | **dilations**: list of ints, dilation value along each spatial axis of the filter | **dilation**: tuple of ints, (dilation_height, dilation_width) |
+|  **dilations:** array of ints of length 2: [dil_height, dil_width] The dilation factor for each spatial dimension of the input. | **input[11:12]:** int, dilation_width and dilation_height | **dilationRateX** and  **dilationRateY** of MPSCNNConvolutionDescriptor: unsigned int, dilation_width and dilation_height | **Dilations** of DML_CONVOLUTION_OPERATOR_DESC: unsigned int, [dilation_H, dilation_W] | **Open**: not supported | **dilates** of convolution_forward::desc: int64, [dilation_width, dilation_height] | **dilations**: list of ints, dilation value along each spatial axis of the filter | **dilation**: tuple of ints, (dilation_height, dilation_width) |