ONNX operators currently supported should cover more opset
Required prerequisites
-
Make sure you've read the documentation. Your issue may be addressed there. -
Search the issue tracker and discussions to verify that this hasn't already been reported. +1 or comment there if it has.
What commit version of aidge do you use
-
aidge_core
: 0.2.2 -
aidge_onnx
: 0.2.2
Problem description
For now, it is quite common for an operator import to work for a single opset: the one of the model that required to import the operator into Aidge first.
Suggestion
Add support for each opset of Operators currently supported
-
Add -
AveragePool -
BatchNormalization -
Concat -
Constant -
Conv -
Dropout -
Div -
Erf -
Gather -
Gemm -
GlobalAveragePool -
Identity -
LeakyRelu -
LSTM -
MatMul -
MaxPool -
Mul -
Pow -
ReduceMean -
Relu -
Reshape -
Shape -
Sigmoid -
Slice -
Softmax -
Split -
Sqrt -
Sub -
Transpose
Edited by Maxence Naud