Super new to ONNX, I have a fundamental question. I would like to speed up my model’s inference and training. The slow part of the model is this portion:
ONNX supports a 1d convolution with dilation, tanh and sigmoid. Is it possible to export this part of the model for inference and/or training?