andr3/gotch
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

WIP(fix BatchAccuracyForLogits)

Browse Source
This commit is contained in:
sugarme 2020-07-11 07:10:50 +10:00
parent c265ba007e
commit 290421a30f
2 changed files with 88 additions and 11 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
example/cifar/main.go
View File

@ -118,6 +118,8 @@ func main() {
var lossVal float64
startTime := time.Now()
var bestAccuracy float64
for epoch := 0; epoch < 150; epoch++ {
// opt.SetLR(learningRate(epoch))
optConfig := nn.NewSGDConfig(0.9, 0.0, 5e-4, true)
@ -169,9 +171,14 @@ func main() {
loss.MustDrop()
}
// testAcc := ts.BatchAccuracyForLogits(net, ds.TestImages, ds.TestLabels, vs.Device(), 512)
// fmt.Printf("Epoch:\t %v\t Loss: \t %.3f \tAcc: %10.2f%%\n", epoch, lossVal, testAcc*100.0)
fmt.Printf("Epoch: %10.0d\tLoss:%10.3f\n", epoch, lossVal)
vs.Freeze()
testAcc := batchAccuracyForLogits(net, ds.TestImages, ds.TestLabels, vs.Device(), 512)
vs.Unfreeze()
fmt.Printf("Epoch:\t %v\t Loss: \t %.3f \tAcc: %10.2f%%\n", epoch, lossVal, testAcc*100.0)
// fmt.Printf("Epoch: %10.0d\tLoss:%10.3f\n", epoch, lossVal)
if testAcc > bestAccuracy {
bestAccuracy = testAcc
}
iter.Drop()
/*
@ -188,7 +195,39 @@ func main() {
* */
}
testAcc := ts.BatchAccuracyForLogits(net, ds.TestImages, ds.TestLabels, vs.Device(), 512)
fmt.Printf("Accuracy: %10.2f%%\n", testAcc*100.0)
// testAcc := ts.BatchAccuracyForLogits(net, ds.TestImages, ds.TestLabels, vs.Device(), 512)
fmt.Printf("Best Accuracy: %10.2f%%\n", bestAccuracy*100.0)
fmt.Printf("Taken time:\t%.2f mins\n", time.Since(startTime).Minutes())
}
func batchAccuracyForLogits(m ts.ModuleT, xs, ys ts.Tensor, d gotch.Device, batchSize int) (retVal float64) {
var (
sumAccuracy float64 = 0.0
sampleCount float64 = 0.0
)
iter2 := ts.MustNewIter2(xs, ys, int64(batchSize))
for {
item, ok := iter2.Next()
if !ok {
break
}
size := float64(item.Data.MustSize()[0])
bImages := item.Data.MustTo(d, true)
bLabels := item.Label.MustTo(d, true)
logits := m.ForwardT(bImages, false)
acc := logits.AccuracyForLogits(bLabels)
sumAccuracy += acc.Values()[0] * size
sampleCount += size
bImages.MustDrop()
bLabels.MustDrop()
acc.MustDrop()
}
return sumAccuracy / sampleCount
}

50
example/mnist/cnn.go
View File

@ -125,6 +125,8 @@ func runCNN1() {
logits := net.ForwardT(bImages, true)
loss := logits.CrossEntropyForLogits(bLabels)
// loss = loss.MustSetRequiresGrad(true)
opt.BackwardStep(loss)
epocLoss = loss.MustShallowClone()
@ -138,10 +140,12 @@ func runCNN1() {
// loss.MustDrop()
}
// testAccuracy := ts.BatchAccuracyForLogitsIdx(net, testImages, testLabels, vs.Device(), 1024)
// fmt.Printf("Epoch: %v\t Loss: %.2f \t Test accuracy: %.2f%%\n", epoch, epocLoss.Values()[0], testAccuracy*100)
vs.Freeze()
testAccuracy := batchAccuracyForLogits(net, testImages, testLabels, vs.Device(), 1024)
vs.Unfreeze()
fmt.Printf("Epoch: %v\t Loss: %.2f \t Test accuracy: %.2f%%\n", epoch, epocLoss.Values()[0], testAccuracy*100.0)
fmt.Printf("Epoch:\t %v\tLoss: \t %.2f\n", epoch, epocLoss.Values()[0])
// fmt.Printf("Epoch:\t %v\tLoss: \t %.2f\n", epoch, epocLoss.Values()[0])
epocLoss.MustDrop()
imagesTs.MustDrop()
labelsTs.MustDrop()
@ -197,13 +201,47 @@ func runCNN2() {
loss.MustDrop()
}
fmt.Printf("Epoch:\t %v\tLoss: \t %.2f\n", epoch, lossVal)
// fmt.Printf("Epoch:\t %v\tLoss: \t %.2f\n", epoch, lossVal)
// testAcc := ts.BatchAccuracyForLogits(net, ds.TestImages, ds.TestLabels, vs.Device(), batchCNN)
// fmt.Printf("Epoch:\t %v\tLoss: \t %.2f\t Accuracy: %.2f\n", epoch, lossVal, testAcc*100)
vs.Freeze()
testAcc := batchAccuracyForLogits(net, ds.TestImages, ds.TestLabels, vs.Device(), batchCNN)
vs.Unfreeze()
fmt.Printf("Epoch:\t %v\tLoss: \t %.2f\t Accuracy: %.2f\n", epoch, lossVal, testAcc*100.0)
}
testAcc := ts.BatchAccuracyForLogits(net, ds.TestImages, ds.TestLabels, vs.Device(), batchCNN)
fmt.Printf("Loss: \t %.2f\t Accuracy: %.2f\n", lossVal, testAcc*100)
fmt.Printf("Taken time:\t%.2f mins\n", time.Since(startTime).Minutes())
}
func batchAccuracyForLogits(m ts.ModuleT, xs, ys ts.Tensor, d gotch.Device, batchSize int) (retVal float64) {
var (
sumAccuracy float64 = 0.0
sampleCount float64 = 0.0
)
iter2 := ts.MustNewIter2(xs, ys, int64(batchSize))
for {
item, ok := iter2.Next()
if !ok {
break
}
size := float64(item.Data.MustSize()[0])
bImages := item.Data.MustTo(d, true)
bLabels := item.Label.MustTo(d, true)
logits := m.ForwardT(bImages, false)
acc := logits.AccuracyForLogits(bLabels)
sumAccuracy += acc.Values()[0] * size
sampleCount += size
bImages.MustDrop()
bLabels.MustDrop()
acc.MustDrop()
}
return sumAccuracy / sampleCount
}