ArmNN  NotReleased
src/armnnOnnxParser/OnnxSupport.md
Go to the documentation of this file.
1 # ONNX operators that the Arm NN SDK supports
2 
3 This reference guide provides a list of ONNX operators the Arm NN SDK currently supports.
4 
5 The Arm NN SDK ONNX parser currently only supports fp32 operators.
6 
7 ## Fully supported
8 
9 **Add**
10 
11 See the ONNX [Add documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#Add) for more information
12 
13 **AveragePool**
14 
15 See the ONNX [AveragePool documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#AveragePool) for more information.
16 
17 **Constant**
18 
19 See the ONNX [Constant documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#Constant) for more information.
20 
21 **GlobalAveragePool**
22 
23 See the ONNX [GlobalAveragePool documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#GlobalAveragePool) for more information.
24 
25 **MaxPool**
26 
27 See the ONNX [max_pool documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#MaxPool) for more information.
28 
29 **Relu**
30 
31 See the ONNX [Relu documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#Relu) for more information.
32 
33 **Reshape**
34 
35 See the ONNX [Reshape documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#Reshape) for more information.
36 
37 ## Partially supported
38 
39 **Conv**
40 
41 The parser only supports 2D convolutions with a dilation rate of [1, 1] and group = 1 or group = #Nb_of_channel (depthwise convolution)
42 See the ONNX [Conv documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#Conv) for more information.
43 
44 **BatchNormalization**
45 
46 The parser does not support training mode. See the ONNX [BatchNormalization documentation](https://github.com/onnx/onnx/blob/master/docs/Operators.md#BatchNormalization) for more information.
47 
48 **MatMul**
49 
50 The parser only supports constant weights in a fully connected layer.
51 
52 ## Tested networks
53 
54 Arm tested these operators with the following ONNX fp32 neural networks:
55 
56 * Simple MNIST. See the ONNX [MNIST documentation](https://github.com/onnx/models/tree/master/mnist) for more information.
57 
58 * Mobilenet_v2. See the ONNX [MobileNet documentation](https://github.com/onnx/models/tree/master/models/image_classification/mobilenet) for more information.
59 
60 More machine learning operators will be supported in future releases.