I am having a look at neural networks in Arraymancer. After training the model,
how do I test it at a particular point. See ??? in the last line of code below:

`let (N, D_in, H, D_out) = (10, 2, 3, 1)

# Create the autograd context that will hold the computational graph
let ctx = newContext Tensor[float32]

# Create random Tensors to hold inputs and outputs, and wrap them in Variables.
let
#   x = ctx.variable(randomTensor[float32](N, D_in, 1'f32))
  xx = [0.000783'f32, 0.153779, 0.560532, 0.865013, 0.276724, 0.895919,
0.704462, 0.886472, 0.929641, 0.469290, 0.350208, 0.941637, 0.096535, 0.457211,
0.346164, 0.970019, 0.114938, 0.769819, 0.341565, 0.684224
].toTensor.reshape(10,2)
  x = ctx.variable( xx )
  y = [3.000602'f32, 5.738535, 4.316186, 6.618693, 6.045585, 4.771489, 3.230002,
4.798756, 3.455619, 4.285201].toTensor.reshape(10,1)

# ##################################################################
# Define the model

network TwoLayersNet:
  layers:
    fc1: Linear(D_in, H)
    fc2: Linear(H, D_out)
  forward x:
    # x.fc1.relu.fc2
    x.fc1.sigmoid.fc2

let
  model = ctx.init(TwoLayersNet)
  optim = model.optimizer(SGD, learning_rate = 1e-4'f32)

# ##################################################################
# Training

for t in 0 ..< 500:
  let
    y_pred = model.forward(x)
    loss = y_pred.mse_loss(y)

  # echo &"Epoch {t}: loss {loss.value[0]}"

  loss.backprop()
  optim.update()

# ##################################################################
# Test at (0.25, 0.5)
let res = [0.2'f32, 0.5 ].toTensor.reshape(2,1)
???`