parent
ece6a17b11
commit
d607bec6f9
|
@ -291,7 +291,7 @@ public class RegressionEvalTest extends BaseNd4jTest {
|
||||||
for (Metric m : Metric.values()) {
|
for (Metric m : Metric.values()) {
|
||||||
double d1 = e4d.scoreForMetric(m);
|
double d1 = e4d.scoreForMetric(m);
|
||||||
double d2 = e2d.scoreForMetric(m);
|
double d2 = e2d.scoreForMetric(m);
|
||||||
assertEquals(m.toString(), d2, d1, 1e-6);
|
assertEquals(m.toString(), d2, d1, 1e-5);
|
||||||
}
|
}
|
||||||
}
|
}
|
||||||
|
|
||||||
|
@ -385,7 +385,7 @@ public class RegressionEvalTest extends BaseNd4jTest {
|
||||||
for(Metric m : Metric.values()){
|
for(Metric m : Metric.values()){
|
||||||
double d1 = e4d_m1.scoreForMetric(m);
|
double d1 = e4d_m1.scoreForMetric(m);
|
||||||
double d2 = e2d_m1.scoreForMetric(m);
|
double d2 = e2d_m1.scoreForMetric(m);
|
||||||
assertEquals(m.toString(), d2, d1, 1e-6);
|
assertEquals(m.toString(), d2, d1, 1e-5);
|
||||||
}
|
}
|
||||||
|
|
||||||
//Check per-output masking:
|
//Check per-output masking:
|
||||||
|
|
|
@ -551,15 +551,13 @@ public class SpecialTests extends BaseNd4jTest {
|
||||||
int[] inputShape = new int[]{1, 2, 2, 1};
|
int[] inputShape = new int[]{1, 2, 2, 1};
|
||||||
|
|
||||||
int M = 2;
|
int M = 2;
|
||||||
int[] blockShape = new int[]{M, 1};
|
|
||||||
int[] paddingShape = new int[]{M, 2};
|
|
||||||
|
|
||||||
INDArray input = Nd4j.randn(inputShape).castTo(DataType.DOUBLE);
|
INDArray input = Nd4j.randn(inputShape).castTo(DataType.DOUBLE);
|
||||||
INDArray blocks = Nd4j.create(new float[]{2, 2}, blockShape).castTo(DataType.INT);
|
INDArray blocks = Nd4j.createFromArray(2, 2);
|
||||||
INDArray padding = Nd4j.create(new float[]{0, 0, 0, 0}, paddingShape).castTo(DataType.INT);
|
INDArray padding = Nd4j.createFromArray(0, 0, 0, 0).reshape(2,2);
|
||||||
|
|
||||||
INDArray expOut = Nd4j.create(DataType.DOUBLE, miniBatch, 1, 1, 1);
|
INDArray expOut = Nd4j.create(DataType.DOUBLE, miniBatch, 1, 1, 1);
|
||||||
val op = DynamicCustomOp.builder("space_to_batch")
|
val op = DynamicCustomOp.builder("space_to_batch_nd")
|
||||||
.addInputs(input, blocks, padding)
|
.addInputs(input, blocks, padding)
|
||||||
.addOutputs(expOut).build();
|
.addOutputs(expOut).build();
|
||||||
Nd4j.getExecutioner().execAndReturn(op);
|
Nd4j.getExecutioner().execAndReturn(op);
|
||||||
|
@ -573,15 +571,13 @@ public class SpecialTests extends BaseNd4jTest {
|
||||||
int[] inputShape = new int[]{miniBatch, 1, 1, 1};
|
int[] inputShape = new int[]{miniBatch, 1, 1, 1};
|
||||||
|
|
||||||
int M = 2;
|
int M = 2;
|
||||||
int[] blockShape = new int[]{M, 1};
|
|
||||||
int[] cropShape = new int[]{M, 2};
|
|
||||||
|
|
||||||
INDArray input = Nd4j.randn(inputShape).castTo(DataType.DOUBLE);
|
INDArray input = Nd4j.randn(inputShape).castTo(DataType.DOUBLE);
|
||||||
INDArray blocks = Nd4j.create(new float[]{2, 2}, blockShape).castTo(DataType.INT);
|
INDArray blocks = Nd4j.createFromArray(2, 2);
|
||||||
INDArray crops = Nd4j.create(new float[]{0, 0, 0, 0}, cropShape).castTo(DataType.INT);
|
INDArray crops = Nd4j.createFromArray(0, 0, 0, 0).reshape(2,2);
|
||||||
|
|
||||||
INDArray expOut = Nd4j.create(DataType.DOUBLE, 1, 2, 2, 1);
|
INDArray expOut = Nd4j.create(DataType.DOUBLE, 1, 2, 2, 1);
|
||||||
DynamicCustomOp op = DynamicCustomOp.builder("batch_to_space")
|
DynamicCustomOp op = DynamicCustomOp.builder("batch_to_space_nd")
|
||||||
.addInputs(input, blocks, crops)
|
.addInputs(input, blocks, crops)
|
||||||
.addOutputs(expOut).build();
|
.addOutputs(expOut).build();
|
||||||
Nd4j.getExecutioner().execAndReturn(op);
|
Nd4j.getExecutioner().execAndReturn(op);
|
||||||
|
|
Loading…
Reference in New Issue