Function test_Conv1D

numpy_ml/tests/test_nn.py:1267–1345

(N=15)

def test_Conv1D(N=15):
    from numpy_ml.neural_nets.layers import Conv1D
    from numpy_ml.neural_nets.activations import Tanh, ReLU, Sigmoid, Affine

    N = np.inf if N is None else N

    np.random.seed(12345)

    acts = [
        (Tanh(), nn.Tanh(), "Tanh"),
        (Sigmoid(), nn.Sigmoid(), "Sigmoid"),
        (ReLU(), nn.ReLU(), "ReLU"),
        (Affine(), TorchLinearActivation(), "Affine"),
    ]

    i = 1
    while i < N + 1:
        n_ex = np.random.randint(1, 10)
        l_in = np.random.randint(1, 10)
        n_in, n_out = np.random.randint(1, 3), np.random.randint(1, 3)
        f_width = min(l_in, np.random.randint(1, 5))
        p, s = np.random.randint(0, 5), np.random.randint(1, 3)
        d = np.random.randint(0, 5)

        fc = f_width * (d + 1) - d
        l_out = int(1 + (l_in + 2 * p - fc) / s)

        if l_out <= 0:
            continue

        X = random_tensor((n_ex, l_in, n_in), standardize=True)

        # randomly select an activation function
        act_fn, torch_fn, act_fn_name = acts[np.random.randint(0, len(acts))]

        # initialize Conv2D layer
        L1 = Conv1D(
            out_ch=n_out,
            kernel_width=f_width,
            act_fn=act_fn,
            pad=p,
            stride=s,
            dilation=d,
        )

        # forward prop
        y_pred = L1.forward(X)

        # backprop
        dLdy = np.ones_like(y_pred)
        dLdX = L1.backward(dLdy)

        # get gold standard gradients
        gold_mod = TorchConv1DLayer(
            n_in, n_out, torch_fn, L1.parameters, L1.hyperparameters
        )
        golds = gold_mod.extract_grads(X)