/// Copyright (c) 2021 ARM Limited and Contributors. All rights reserved.
///
/// SPDX-License-Identifier: MIT
///

namespace armnn
{
/**
@page operator_list Arm NN Operators

@tableofcontents

@section S5_1_operator_list Arm NN Operators

Arm NN supports operators that are listed in below table.

Arm NN supports a wide list of data-types.
The main data-types that the Machine Learning functions support are the following:
  <ul>
    <li><b>BFLOAT16:</b> 16-bit non-standard brain floating point
    <li><b>QASYMMU8:</b> 8-bit unsigned asymmetric quantized
    <li><b>QASYMMS8:</b> 8-bit signed asymmetric quantized
    <li><b>QUANTIZEDSYMM8PERAXIS:</b> 8-bit signed symmetric quantized
    <li><b>QSYMMS8:</b> 8-bit signed symmetric quantized
    <li><b>QSYMMS16:</b> 16-bit signed symmetric quantized
    <li><b>FLOAT32:</b> 32-bit single precision floating point
    <li><b>FLOAT16:</b> 16-bit half precision floating point
    <li><b>SIGNED32:</b> 32-bit signed integer
    <li><b>BOOLEAN:</b> 8-bit unsigned char
    <li><b>All:</b> Agnostic to any specific data type
  </ul>

Arm NN supports the following data layouts (fast changing dimension from right to left):
  <ul>
    <li><b>NHWC:</b> Layout where channels are in the fastest changing dimension
    <li><b>NCHW:</b> Layout where width is in the fastest changing dimension
    <li><b>All:</b> Agnostic to any specific data layout
  </ul>
where N = batches, C = channels, H = height, W = width

<table>
<caption id="multi_row"></caption>
<tr>
  <th>Operator
  <th>Description
  <th>Equivalent Android NNAPI Operator
  <th>Backends
  <th>Data Layouts
  <th>Data Types
<tr>
  <td rowspan="3">AbsLayer
  <td rowspan="3"> Layer to perform absolute operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_ABS
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
       <tr><th>
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>SIGNED32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
       <tr><th>
       <tr><td>FLOAT16
       <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ActivationLayer
  <td rowspan="3" style="width:200px;"> Layer to simulate an activation layer with the specified activation function.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_ABS
       <li>ANEURALNETWORKS_ELU
       <li>ANEURALNETWORKS_HARD_SWISH
       <li>ANEURALNETWORKS_LOGISTIC
       <li>ANEURALNETWORKS_PRELU
       <li>ANEURALNETWORKS_RELU
       <li>ANEURALNETWORKS_RELU1
       <li>ANEURALNETWORKS_RELU6
       <li>ANEURALNETWORKS_SQRT
       <li>ANEURALNETWORKS_TANH
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
    <ul>
      <li>All
    </ul>
  <td>
    <table>
      <tr><th>
      <tr><td>QASYMMU8
      <tr><td>QASYMMS8
      <tr><td>QSYMMS16
      <tr><td>FLOAT16
      <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
      <tr><th>
      <tr><td>QASYMMU8
      <tr><td>QASYMMS8
      <tr><td>QSYMMS16
      <tr><td>FLOAT16
      <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">AdditionLayer
  <td rowspan="3" style="width:200px;"> Layer to add 2 tensors.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_ADD
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ArgMinMaxLayer
  <td rowspan="3" style="width:200px;"> Layer to calculate the index of the minimum or maximum values in a tensor
  based on an axis.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_ARGMAX
       <li>ANEURALNETWORKS_ARGMIN
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
       <tr><td>SIGNED64
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">BatchNormalizationLayer
  <td rowspan="3" style="width:200px;"> Layer to perform batch normalization.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT32
    <tr><td>FLOAT16
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT32
    <tr><td>FLOAT16
    </table>
<tr>
  <td rowspan="3">BatchToSpaceNdLayer
  <td rowspan="3" style="width:200px;"> Layer to perform a batch to space transformation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_BATCH_TO_SPACE_ND
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">CastLayer
  <td rowspan="3" style="width:200px;"> Layer to cast a tensor to a type.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_CAST
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QSYMMS8
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMS8
    <tr><td>QASYMMU8
    <tr><td>FLOAT16
    <tr><td>SIGNED32
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMS8
    <tr><td>QASYMMU8
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ChannelShuffleLayer
  <td rowspan="3" style="width:200px;"> Layer to reorganize the channels of a tensor.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_CHANNEL_SHUFFLE
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QSYMMS8
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMS8
    <tr><td>QASYMMU8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMS8
    <tr><td>QASYMMU8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ComparisonLayer
  <td rowspan="3" style="width:200px;"> Layer to compare 2 tensors.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_EQUAL
       <li>ANEURALNETWORKS_GREATER
       <li>ANEURALNETWORKS_GREATER_EQUAL
       <li>ANEURALNETWORKS_LESS
       <li>ANEURALNETWORKS_LESS_EQUAL
       <li>ANEURALNETWORKS_NOT_EQUAL
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>BOOLEAN
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
   <td>CpuAcc
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>All
      </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">ConcatLayer
  <td rowspan="3" style="width:200px;"> Layer to concatenate tensors along a given axis.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_CONCATENATION
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ConstantLayer
  <td rowspan="3" style="width:200px;"> Layer to provide a constant tensor.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">ConvertBf16ToFp32Layer
  <td rowspan="3" style="width:200px;"> Layer to convert BFloat16 tensor to Float32 tensor.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>BFLOAT16
         <tr><td>FLOAT32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
         <tr><th>
         <tr><td>BFLOAT16
         <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
         <tr><th>
         <tr><td>BFLOAT16
         <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ConvertFp16ToFp32Layer
  <td rowspan="3" style="width:200px;"> Layer to convert Float16 tensor to Float32 tensor.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>FLOAT16
         <tr><td>FLOAT32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
         <tr><th>
         <tr><td>FLOAT16
         <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
         <tr><th>
         <tr><td>FLOAT16
         <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ConvertFp32ToBf16Layer
  <td rowspan="3" style="width:200px;"> Layer to convert Float32 tensor to BFloat16 tensor.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>BFLOAT16
         <tr><td>FLOAT32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
         <tr><th>
         <tr><td>BFLOAT16
         <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
         <tr><th>
         <tr><td>BFLOAT16
         <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ConvertFp32ToFp16Layer
  <td rowspan="3" style="width:200px;"> Layer to convert Float32 tensor to Float16 tensor.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>FLOAT16
         <tr><td>FLOAT32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
         <tr><th>
         <tr><td>FLOAT16
         <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
         <tr><th>
         <tr><td>FLOAT16
         <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">Convolution2dLayer
  <td rowspan="3" style="width:200px;"> Layer to compute a convolution operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_CONV_2D
       <li>ANEURALNETWORKS_GROUPED_CONV_2D
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QUANTIZEDSYMM8PERAXIS
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QUANTIZEDSYMM8PERAXIS
    </table>
<tr>
  <td rowspan="3">Convolution3dLayer
  <td rowspan="3" style="width:200px;"> Layer to compute a 3D convolution operation.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>NDHWC
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>N/A
      </ul>
  <td>
      <ul>
       <li>N/A
      </ul>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>N/A
      </ul>
  <td>
      <ul>
       <li>N/A
      </ul>
<tr>
  <td rowspan="1">DebugLayer
  <td rowspan="1" style="width:200px;"> Layer to print out inter layer tensor information.
  <td rowspan="1">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td rowspan="3">DepthToSpaceLayer
  <td rowspan="3" style="width:200px;"> Layer to perform Depth to Space transformation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_DEPTH_TO_SPACE
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">DepthwiseConvolution2dLayer
  <td rowspan="3" style="width:200px;"> Layer to compute a deconvolution or transpose convolution.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_DEPTHWISE_CONV_2D
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QUANTIZEDSYMM8PERAXIS
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QUANTIZEDSYMM8PERAXIS
    </table>
<tr>
  <td rowspan="3">DequantizeLayer
  <td rowspan="3" style="width:200px;"> Layer to dequantize the values in a tensor.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_DEQUANTIZE
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QUANTIZEDSYMM8PERAXIS
    <tr><td>QSYMMS8
    <tr><td>QSYMMS16
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QUANTIZEDSYMM8PERAXIS
    <tr><td>QSYMMS8
    <tr><td>QSYMMS16
    </table>
<tr>
  <td rowspan="2">DetectionPostProcessLayer
  <td rowspan="2" style="width:200px;"> Layer to generate the detection output based on center size encoded boxes, class prediction and anchors by doing non maximum suppression (NMS).
  <td rowspan="2">
      <ul>
       <li>ANEURALNETWORKS_DETECTION_POSTPROCESSING
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">DivisionLayer
  <td rowspan="3" style="width:200px;"> Layer to divide 2 tensors.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_DIV
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ElementwiseBaseLayer
  <td rowspan="3" style="width:200px;"> Layer to perform Add - Div - Max - Min - Mul operations.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_ADD
       <li>ANEURALNETWORKS_DIV
       <li>ANEURALNETWORKS_MAXIMUM
       <li>ANEURALNETWORKS_MINIMUM
       <li>ANEURALNETWORKS_MUL
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ElementwiseUnaryLayer
  <td rowspan="3" style="width:200px;"> Layer to perform Rsqrt - Exp - Neg - Log - Abs - Sin - Sqrt operations.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_ABS
       <li>ANEURALNETWORKS_EXP
       <li>ANEURALNETWORKS_LOG
       <li>ANEURALNETWORKS_NEG
       <li>ANEURALNETWORKS_RSQRT
       <li>ANEURALNETWORKS_SIN
       <li>ANEURALNETWORKS_SQRT
      </ul>
  <td>CpuRef
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
    </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="1">FakeQuantizationLayer
  <td rowspan="1" style="width:200px;"> Layer to quantize float values and dequantize afterwards. The current implementation does not dequantize the values.
  <td rowspan="1">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>FLOAT32
      </table>
<tr>
  <td rowspan="3">FillLayer
  <td rowspan="3" style="width:200px;"> Layer to set the values of a tensor with a given value.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_FILL
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">FloorLayer
  <td rowspan="3" style="width:200px;"> Layer to round the value to the lowest whole number.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_FLOOR
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT32
    <tr><td>FLOAT16
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT32
    <tr><td>FLOAT16
    </table>
<tr>
  <td rowspan="3">FullyConnectedLayer
  <td rowspan="3" style="width:200px;"> Layer to perform a fully connected / dense operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_FULLY_CONNECTED
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    </table>
<tr>
  <td rowspan="3">GatherLayer
  <td rowspan="3" style="width:200px;"> Layer to perform the gather operation along the chosen axis.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_GATHER
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">GatherNdLayer
  <td rowspan="3" style="width:200px;"> Layer to perform the gatherNd operation.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>TBD
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>TBD
    </table>
<tr>
  <td rowspan="1">InputLayer
  <td rowspan="1" style="width:200px;"> Special layer used to provide input data to the computational network.
  <td rowspan="1">
      <ul>
       <li>N/A
      </ul>
   <td>All
   <td>
         <ul>
          <li>All
         </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">InstanceNormalizationLayer
  <td rowspan="3" style="width:200px;"> Layer to perform an instance normalization on a given axis.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_INSTANCE_NORMALIZATION
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">L2NormalizationLayer
  <td rowspan="3" style="width:200px;"> Layer to perform an L2 normalization on a given axis.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_L2_NORMALIZATION
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">LogSoftmaxLayer
  <td rowspan="3" style="width:200px;"> Layer to perform the log softmax activations given logits.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">LogicalBinaryLayer
  <td rowspan="3" style="width:200px;"> Layer to perform Logical AND - Logical NOT - Logical OR operations.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_LOGICAL_AND
       <li>ANEURALNETWORKS_LOGICAL_NOT
       <li>ANEURALNETWORKS_LOGICAL_OR
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>BOOLEAN
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>BOOLEAN
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>BOOLEAN
    </table>
<tr>
  <td rowspan="3">LstmLayer
  <td rowspan="3" style="width:200px;"> Layer to perform a single time step in a Long Short-Term Memory (LSTM) operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_LSTM
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">MapLayer
  <td rowspan="3" style="width:200px;"> Layer to perform map operation on tensor.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>All
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">MaximumLayer
  <td rowspan="3" style="width:200px;"> Layer to perform an elementwise maximum of two tensors.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    </table>
<tr>
  <td rowspan="3">MeanLayer
  <td rowspan="3" style="width:200px;"> Layer to perform reduce mean operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_MEAN
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">MemCopyLayer
  <td rowspan="3" style="width:200px;"> Layer to perform memory copy operation.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>BOOLEAN
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">MemImportLayer
  <td rowspan="3" style="width:200px;"> Layer to perform memory import operation.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>All
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">MergeLayer
  <td rowspan="3" style="width:200px;"> Layer to concatenate tensors along a given axis.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_CONCATENATION
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">MinimumLayer
  <td rowspan="3" style="width:200px;"> Layer to perform an elementwise minimum of two tensors.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_MINIMUM
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    </table>
<tr>
  <td rowspan="3">MultiplicationLayer
  <td rowspan="3" style="width:200px;"> Layer to perform an elementwise multiplication of two tensors.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_MUL
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    </table>
<tr>
  <td rowspan="3">NormalizationLayer
  <td rowspan="3" style="width:200px;"> Layer to compute normalization operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_LOCAL_RESPONSE_NORMALIZATION
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT32
    <tr><td>FLOAT16
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT32
    <tr><td>FLOAT16
    </table>
<tr>
  <td rowspan="1">OutputLayer
  <td rowspan="1" style="width:200px;"> A special layer providing access to a user supplied buffer into which the output of a network can be written.
  <td rowspan="1">
      <ul>
       <li>N/A
      </ul>
   <td>All
   <td>
         <ul>
          <li>All
         </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">PadLayer
  <td rowspan="3" style="width:200px;"> Layer to pad a tensor.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_PAD
       <li>ANEURALNETWORKS_PAD_V2
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">PermuteLayer
  <td rowspan="3" style="width:200px;"> Layer to transpose an ND tensor.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_TRANSPOSE
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">Pooling2dLayer
  <td rowspan="3" style="width:200px;"> Layer to perform 2D pooling with the specified pooling operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_AVERAGE_POOL_2D
       <li>ANEURALNETWORKS_L2_POOL_2D
       <li>ANEURALNETWORKS_MAX_POOL_2D
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">Pooling3dLayer
  <td rowspan="3" style="width:200px;"> Layer to perform 3D pooling with the specified pooling operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_AVERAGE_POOL_3D
       <li>ANEURALNETWORKS_L2_POOL_3D
       <li>ANEURALNETWORKS_MAX_POOL_3D
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>NDHWC
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NA
      </ul>
  <td>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NDHWC
      </ul>
<tr>
  <td rowspan="1">PreCompiledLayer
  <td rowspan="1" style="width:200px;"> Opaque layer provided by a backend which provides an executable representation of a subgraph from the original network.
  <td rowspan="1">
      <ul>  
       <li>N/A
      </ul>
   <td>N/A
   <td>N/A
   <td>N/A
<tr>
  <td rowspan="3">PreluLayer
  <td rowspan="3" style="width:200px;"> Layer to compute the activation layer with the PRELU activation function.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_PRELU
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">QLstmLayer
  <td rowspan="3" style="width:200px;"> Layer to perform quantized LSTM (Long Short-Term Memory) operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_QUANTIZED_LSTM
       <li>ANEURALNETWORKS_QUANTIZED_16BIT_LSTM
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>All
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMS8
    <tr><td>QASYMMU8
    <tr><td>SIGNED32
    <tr><td>QSYMMS16
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMS8
    <tr><td>QASYMMU8
    <tr><td>SIGNED32
    <tr><td>QSYMMS16
    </table>
<tr>
  <td rowspan="3">QuantizeLayer
  <td rowspan="3" style="width:200px;"> Layer to perform quantization operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_QUANTIZE
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QASYMM16
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QASYMM16
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">QuantizedLstmLayer
  <td rowspan="3" style="width:200px;"> Layer to perform quantized LSTM (Long Short-Term Memory) operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_QUANTIZED_LSTM
       <li>ANEURALNETWORKS_QUANTIZED_16BIT_LSTM
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>All
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>SIGNED32
    <tr><td>QASYMMU8
    <tr><td>QSYMMS16
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>SIGNED32
    <tr><td>QASYMMU8
    <tr><td>QSYMMS16
    </table>
<tr>
  <td rowspan="3">RankLayer
  <td rowspan="3" style="width:200px;"> Layer to perform a rank operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_RANK
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>All
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">ReduceLayer
  <td rowspan="3" style="width:200px;"> Layer to perform reduce with the following operations - ARG_IDX_MAX: Index of the max value - ARG_IDX_MIN: Index of the min value - MEAN_SUM:    Mean of sum - PROD:        Product - SUM_SQUARE:  Sum of squares - SUM:         Sum - MIN:         Min - MAX:         Max
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_REDUCE_MAX
       <li>ANEURALNETWORKS_REDUCE_MIN
       <li>ANEURALNETWORKS_REDUCE_SUM
       <li>ANEURALNETWORKS_REDUCE_PROD
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    </table>
<tr>
  <td rowspan="3">ReshapeLayer
  <td rowspan="3" style="width:200px;"> Layer to reshape a tensor.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_RESHAPE
       <li>ANEURALNETWORKS_SQUEEZE
       <li>ANEURALNETWORKS_EXPAND_DIMS
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
       <tr><td>BOOLEAN
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">ResizeLayer
  <td rowspan="3" style="width:200px;"> Layer to perform resize of a tensor using one of the interpolation methods: - Bilinear -  Nearest Neighbor.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_RESIZE_BILINEAR
       <li>ANEURALNETWORKS_RESIZE_NEAREST_NEIGHBOR
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">RsqrtLayer
  <td rowspan="3" style="width:200px;"> Layer to perform Rsqrt operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_RSQRT
      </ul>
  <td>CpuRef
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
    </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>SIGNED32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">ShapeLayer
  <td rowspan="3" style="width:200px;"> Layer to return the shape of the input tensor.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
  <td>CpuRef
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">SliceLayer
  <td rowspan="3" style="width:200px;"> Layer to perform tensor slicing.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_SLICE
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">SoftmaxLayer
  <td rowspan="3" style="width:200px;"> Layer to perform softmax, log-softmax operation over the specified axis.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_LOG_SOFTMAX
       <li>ANEURALNETWORKS_SOFTMAX
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">SpaceToBatchNdLayer
  <td rowspan="3" style="width:200px;"> Layer to divide spatial dimensions of the tensor into a grid of blocks and interleaves these blocks with the batch dimension.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_SPACE_TO_BATCH_ND
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">SpaceToDepthLayer
  <td rowspan="3" style="width:200px;"> Layer to rearrange blocks of spatial data into depth.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_SPACE_TO_DEPTH
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">SplitterLayer
  <td rowspan="3" style="width:200px;"> Layer to split a tensor along a given axis.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_SPLIT
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">StackLayer
  <td rowspan="3" style="width:200px;"> Layer to stack tensors along an axis.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="1">StandInLayer
  <td rowspan="1" style="width:200px;"> A layer to represent "unknown" or "unsupported" operations in the input graph. It has a configurable number of input and output slots and an optional name.
  <td rowspan="1">
      <ul>
       <li>N/A
      </ul>
   <td>N/A
   <td>N/A
   <td>N/A
<tr>
  <td rowspan="3">StridedSliceLayer
  <td rowspan="3" style="width:200px;"> Layer to extract a strided slice of a tensor.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_STRIDED_SLICE
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">SubtractionLayer
  <td rowspan="3" style="width:200px;"> Layer to perform an elementwise subtract of 2 tensors.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_SUB
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
       <tr><td>SIGNED32
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QSYMMS16
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    </table>
<tr>
  <td rowspan="3">TransposeConvolution2dLayer
  <td rowspan="3" style="width:200px;"> Layer to perform 2D transpose convolution (deconvolution) operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_TRANSPOSE_CONV_2D
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QUANTIZEDSYMM8PERAXIS
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>SIGNED32
    <tr><td>FLOAT16
    <tr><td>FLOAT32
    <tr><td>QASYMMU8
    <tr><td>QASYMMS8
    <tr><td>QUANTIZEDSYMM8PERAXIS
    </table>
<tr>
  <td rowspan="3">TransposeLayer
  <td rowspan="3" style="width:200px;"> Layer to transpose a tensor.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_TRANSPOSE
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
       <tr><th>
       <tr><td>BFLOAT16
       <tr><td>FLOAT16
       <tr><td>FLOAT32
       <tr><td>QASYMMS8
       <tr><td>QASYMMU8
       <tr><td>QSYMMS16
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>All
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td rowspan="3">UnidirectionalSquenceLstmLayer
  <td rowspan="3" style="width:200px;"> Layer to perform unidirectional sequence LSTM operation.
  <td rowspan="3">
      <ul>
       <li>ANEURALNETWORKS_UNIDIRECTIONAL_SEQUENCE_LSTM
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>Input Types
         <tr><td>FLOAT32
      </table>
      <table>
         <tr><th>Weight Types
         <tr><td>FLOAT32
         <tr><td>QASYMMS8
      </table>
   <td>CpuAcc
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>Input Types
         <tr><td>FLOAT32
      </table>
      <table>
         <tr><th>Weight Types
         <tr><td>FLOAT32
      </table>
   <td>GpuAcc
    <td>
        <ul>
         <li>All
        </ul>
    <td>
     <table>
        <tr><th>Input Types
        <tr><td>FLOAT32
     </table>
     <table>
        <tr><th>Weight Types
        <tr><td>FLOAT32
     </table>
<tr>
  <td rowspan="3">UnmapLayer
  <td rowspan="3" style="width:200px;"> Layer to perform unmap operation on tensor.
  <td rowspan="3">
      <ul>
       <li>N/A
      </ul>
   <td>CpuRef
     <td>
         <ul>
          <li>All
         </ul>
     <td>
      <table>
         <tr><th>
         <tr><td>All
      </table>
<tr>
  <td>CpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
<tr>
  <td>GpuAcc
  <td>
      <ul>
       <li>NHWC
       <li>NCHW
      </ul>
  <td>
    <table>
    <tr><th>
    <tr><td>All
    </table>
</table>

*/
} // namespace