public void ExecuteTest()
{
float max_err = 0;
foreach (int batch in new int[] { 1, 2 })
{
foreach (int channels in new int[] { 1, 2, 3, 4, 5, 6, 7, 8 })
{
foreach (int stride in new int[] { 2, 3, 4 })
{
foreach (int inwidth in new int[] { 5, 7, 11 })
{
foreach (int inheight in new int[] { 5, 7, 11 })
{
foreach (int indepth in new int[] { 5, 7, 11 })
{
int outwidth = inwidth / stride, outheight = inheight / stride, outdepth = indepth / stride;
float[] yval = (new float[outwidth * outheight * outdepth * channels * batch]).Select((_, idx) => idx * 1e-3f).ToArray();
Map3D y = new Map3D(channels, outwidth, outheight, outdepth, batch, yval);
Map3D x = Reference(y, inwidth, inheight, indepth, stride);
OverflowCheckedTensor x_tensor = new OverflowCheckedTensor(Shape.Map3D(channels, inwidth, inheight, indepth, batch));
OverflowCheckedTensor y_tensor = new OverflowCheckedTensor(Shape.Map3D(channels, outwidth, outheight, outdepth, batch), yval);
AverageUnpooling ope = new AverageUnpooling(inwidth, inheight, indepth, channels, stride, batch);
ope.Execute(y_tensor, x_tensor);
float[] x_expect = x.ToArray();
float[] x_actual = x_tensor.State;
CollectionAssert.AreEqual(yval, y_tensor.State);
AssertError.Tolerance(x_expect, x_actual, 1e-7f, 1e-5f, ref max_err, $"mismatch value {channels},{stride},{inwidth},{inheight},{indepth},{batch}");
Console.WriteLine($"pass: {channels},{stride},{inwidth},{inheight},{indepth},{batch}");
}
}
}
}
}
}
Console.WriteLine($"maxerr:{max_err}");
}
public void SpeedTest()
{
int outwidth = 512, outheight = 512, channels = 32, stride = 2;
int inwidth = outwidth / stride, inheight = outheight / stride;
OverflowCheckedTensor x_tensor = new OverflowCheckedTensor(Shape.Map2D(channels, inwidth, inheight));
OverflowCheckedTensor y_tensor = new OverflowCheckedTensor(Shape.Map2D(channels, outwidth, outheight));
AverageUnpooling ope = new AverageUnpooling(outwidth, outheight, channels, stride);
Stopwatch sw = new Stopwatch();
sw.Start();
ope.Execute(x_tensor, y_tensor);
ope.Execute(x_tensor, y_tensor);
ope.Execute(x_tensor, y_tensor);
ope.Execute(x_tensor, y_tensor);
sw.Stop();
Console.WriteLine($"{sw.ElapsedMilliseconds / 4} msec");
}
public bool CreateLayer(int nCount, ELayerType type, ActivationSettings activationSettings)
{
Layer.Utility.Layer layer;
switch (type)
{
case ELayerType.Invalid:
throw new ArgumentException("Invalid \"type\" argument.");
case ELayerType.AveragePooling:
layer = new AveragePooling(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.AverageUnpooling:
layer = new AverageUnpooling(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.Convolutional:
layer = new Convolutional(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.Deconvolutional:
layer = new Deconvolutional(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.Dropout:
layer = new Dropout(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.FullyConnected:
layer = new FullyConnected(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.GatedRecurrent:
layer = new GatedRecurrent(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.LSTM:
layer = new LSTM(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.MaxPooling:
layer = new MaxPooling(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.MaxUnpooling:
layer = new MaxUnpooling(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
case ELayerType.Recurrent:
layer = new Recurrent(nCount, Layers.Count, activationSettings);
Layers.Add(layer);
return(true);
default:
throw new ArgumentException("Invalid \"type\" argument.");
}
}
