1model = Model() 2i4 = Int32Scalar("b4", 1) 3i5 = Int32Scalar("b5", 1) 4i6 = Int32Scalar("b6", 1) 5i7 = Int32Scalar("b7", 1) 6i8 = Int32Scalar("b8", 0) 7i2 = Input("op2", "TENSOR_FLOAT32", "{1, 8, 8, 3}") # input 0 8i3 = Output("op3", "TENSOR_FLOAT32", "{1, 8, 8, 3}") # output 0 9i0 = Parameter("op0", "TENSOR_FLOAT32", "{1, 1, 1, 3}", [-0.966213, -0.467474, -0.82203]) # parameters 10i1 = Parameter("op1", "TENSOR_FLOAT32", "{3}", [0, 0, 0]) # parameters 11model = model.DepthWiseConv(i2, i0, i1, i4, i5, i6, i7, i8).To(i3) 12