backend := gorgonnx.NewGraph()
model := onnx.NewModel(backend)
err = model.UnmarshalBinary(binaryData)
This is the way to initialize model but to make multiple predictions do I need to do these steps every time or is there a way to skip the UnmarshalBinary step as my model is of big size and this takes more than 5ms and doing this for every prediction is not possible and if I don't do this the model gives increasing output for each input
backend := gorgonnx.NewGraph() model := onnx.NewModel(backend) err = model.UnmarshalBinary(binaryData)
This is the way to initialize model but to make multiple predictions do I need to do these steps every time or is there a way to skip the UnmarshalBinary step as my model is of big size and this takes more than 5ms and doing this for every prediction is not possible and if I don't do this the model gives increasing output for each input