When I use torch.nn.LayerNorm in my model and perform a conversion to ONNX model representation, I observe that the (layer_norm) mapping is missing and it’s represented as a number of smaller ops performing the math for layer norm. I also checked ONNX operators.md list and there is no LayerNorm type listed there. Does this mean there is no support for layernorm type?