private static float TrainModel()
{
float cumulative_train_loss = 0;
foreach (var(data, label) in train_dataloader)
{
NDArray loss_result = null;
using (var ag = Autograd.Record())
{
var output = net.Call(data);
loss_result = loss.Call(output, label);
loss_result.Backward();
}
trainer.Step(batch_size);
cumulative_train_loss += nd.Sum(loss_result).AsScalar <float>();
}
return(cumulative_train_loss);
}
public override NDArrayOrSymbol HybridForward(NDArrayOrSymbol x, params NDArrayOrSymbol[] args)
{
var @out = output.Call(x, args);
if (use_shortcut)
{
if (x.IsNDArray)
{
@out = nd.ElemwiseAdd(@out, x.NdX);
}
else
{
@out = sym.ElemwiseAdd(@out, x.SymX);
}
}
return(@out);
}
public override NDArrayOrSymbol HybridForward(NDArrayOrSymbol x, params NDArrayOrSymbol[] args)
{
var residual = x;
x = body.Call(x, args);
if (ds != null)
{
residual = ds.Call(residual, args);
}
if (x.IsNDArray)
{
x = nd.Activation(x.NdX + residual.NdX, ActivationType.Relu);
}
else
{
x = sym.Activation(x.SymX + residual.SymX, ActivationType.Relu);
}
return(x);
}
public override NDArrayOrSymbol HybridForward(NDArrayOrSymbol x, params NDArrayOrSymbol[] args)
{
var matches = _matchers.Call(x);
return(this.ComposeMatches(matches.NdX));
}