[Pytorch Android] How can I predict with two inputs?

Mobile
1

Tensor outputTensor_head = mModule.forward((IValue.from(inputTensor),IValue.from(inputTensor))).toTensor();

I just made my pytorch android model, that takes two tensors.
But In java, How can define two inputs?? (with array?)

error: ‘)’ expected
Tensor outputTensor_head = mModule.forward((IValue.from(inputTensor),IValue.from(inputTensor))).toTensor();

Above error is came out.
Help me…T.T

2

Is this a double post from here with a new syntax error or what’s the difference between these posts?

3

That was python code (converting pytorch model to pytorch android model) qustion.
Now I just got some idea that was bug of pytorch mobile, the swap opearation or concat operation
can not be able to made a pytorch mobile model.
After adding some action(like CNN or ReLU) at python model’s forward, the conversion is passed.

But I still remain qustion, How Can I predict with converted model with two tensor is JAVA?
The example code only contains one tensor… So I added this question.