public override torch.Tensor forward(torch.Tensor features)
{
// TODO: try whitening-like techniques
// take <s> token (equiv. to [CLS])
using var x = features[torch.TensorIndex.Colon, torch.TensorIndex.Single(0), torch.TensorIndex.Colon];
return(Classifier.forward(x));
}
public override TorchTensor forward(TorchTensor input)
{
using (var f = features.forward(input))
using (var avg = avgPool.forward(f))
using (var x = avg.view(new long[] { avg.shape[0], 256 * 2 * 2 }))
return(classifier.forward(x));
}
static void Main(string[] args)
{
Sequential xornet = new Sequential(
new Linear(2, 100),
new ReLU(),
new Linear(100, 1));
double[,] x = { { 0, 0 }, { 0, 1 }, { 1, 0 }, { 1, 1 } };
double[,] y = { { 0 }, { 1 }, { 1 }, { 0 } };
NDArray a_np = np.array(x);
NDArray b_np = np.array(y);
Tensor input = new Tensor(a_np);
Tensor label = new Tensor(b_np);
int epoch = 1000;
SGD optim = new SGD(xornet.parameters(), 0.05);
MSELoss mse = new MSELoss();
for (int i = 1; i <= epoch; i++)
{
Tensor output = xornet.forward(input);
Tensor loss = mse.forward(output, label);
optim.zero_grad();
loss.backward();
optim.step();
Console.WriteLine("[+] Epoch: " + i + " Loss: " + loss);
}
Tensor z = new Tensor(new NDArray(x));
Tensor outputs = xornet.forward(z);
Console.WriteLine("Result: " + outputs.data.flatten().ToString());
Console.ReadLine();
}
public override torch.Tensor forward(torch.Tensor x)
{
using var x1 = x.permute(1, 2, 0);
using var conv = Conv.forward(x1);
return(conv.permute(2, 0, 1));
}
public override torch.Tensor forward(torch.Tensor x, Dictionary <string, object> param)
{
using var layerOutput = FullConnects.forward(x);
using var layerOuptutIntermediate = layerOutput.add_(x);
return(FinalLayerNorm.forward(layerOutput));
}
