use of org.nd4j.linalg.api.ops.impl.broadcast.BroadcastDivOp in project deeplearning4j by deeplearning4j.
the class L2NormalizeVertex method doBackward.
@Override
public Pair<Gradient, INDArray[]> doBackward(boolean tbptt) {
if (!canDoBackward())
throw new IllegalStateException("Cannot do backward pass: errors not set (L2NormalizeVertex " + vertexName + " idx " + vertexIndex + ")");
INDArray x = inputs[0];
int[] dimensions = getDimensions(x);
INDArray norm = x.norm2(dimensions);
INDArray norm3 = Transforms.pow(norm, 3.0, true);
// in case of div/0
Transforms.max(norm, eps, false);
Transforms.max(norm3, eps, false);
INDArray dLdx;
if (x.rank() == 2) {
// 2D case
dLdx = epsilon.divColumnVector(norm);
INDArray xDivNorm3 = x.divColumnVector(norm3);
dLdx.subi(xDivNorm3.muliColumnVector(epsilon.mul(x).sum(1)));
} else {
//RNN and CNN case - Broadcast along dimension 0
INDArray dx = epsilon.mul(x).sum(dimensions);
//x / |x|_2^3 * sum_k (dLda*x)
INDArray xDivNorm3 = Nd4j.createUninitialized(x.shape(), x.ordering());
Nd4j.getExecutioner().exec(new BroadcastDivOp(x, norm3, xDivNorm3, 0));
Nd4j.getExecutioner().exec(new BroadcastMulOp(xDivNorm3, dx, xDivNorm3, 0));
//1/|x|_2 * dLda - above
dLdx = Nd4j.createUninitialized(epsilon.shape(), epsilon.ordering());
Nd4j.getExecutioner().exec(new BroadcastDivOp(epsilon, norm, dLdx, 0));
dLdx.subi(xDivNorm3);
}
return new Pair<>(null, new INDArray[] { dLdx });
}
use of org.nd4j.linalg.api.ops.impl.broadcast.BroadcastDivOp in project deeplearning4j by deeplearning4j.
the class MaskedReductionUtil method maskedPoolingEpsilonTimeSeries.
public static INDArray maskedPoolingEpsilonTimeSeries(PoolingType poolingType, INDArray input, INDArray mask, INDArray epsilon2d, int pnorm) {
if (input.rank() != 3) {
throw new IllegalArgumentException("Expect rank 3 input activation array: got " + input.rank());
}
if (mask.rank() != 2) {
throw new IllegalArgumentException("Expect rank 2 array for mask: got " + mask.rank());
}
if (epsilon2d.rank() != 2) {
throw new IllegalArgumentException("Expected rank 2 array for errors: got " + epsilon2d.rank());
}
switch(poolingType) {
case MAX:
//TODO This is ugly - replace it with something better... Need something like a Broadcast CAS op
INDArray negInfMask = Transforms.not(mask);
BooleanIndexing.replaceWhere(negInfMask, Double.NEGATIVE_INFINITY, Conditions.equals(1.0));
INDArray withInf = Nd4j.createUninitialized(input.shape());
Nd4j.getExecutioner().exec(new BroadcastAddOp(input, negInfMask, withInf, 0, 2));
//At this point: all the masked out steps have value -inf, hence can't be the output of the MAX op
INDArray isMax = Nd4j.getExecutioner().execAndReturn(new IsMax(withInf, 2));
return Nd4j.getExecutioner().execAndReturn(new BroadcastMulOp(isMax, epsilon2d, isMax, 0, 1));
case AVG:
case SUM:
//if out = sum(in,dims) then dL/dIn = dL/dOut -> duplicate to each step and mask
//if out = avg(in,dims) then dL/dIn = 1/N * dL/dOut
//With masking: N differs for different time series
INDArray out = Nd4j.createUninitialized(input.shape(), 'f');
//Broadcast copy op, then divide and mask to 0 as appropriate
Nd4j.getExecutioner().exec(new BroadcastCopyOp(out, epsilon2d, out, 0, 1));
Nd4j.getExecutioner().exec(new BroadcastMulOp(out, mask, out, 0, 2));
if (poolingType == PoolingType.SUM) {
return out;
}
//[minibatchSize,tsLength] -> [minibatchSize,1]
INDArray nEachTimeSeries = mask.sum(1);
Nd4j.getExecutioner().exec(new BroadcastDivOp(out, nEachTimeSeries, out, 0));
return out;
case PNORM:
//Similar to average and sum pooling: there's no N term here, so we can just set the masked values to 0
INDArray masked2 = Nd4j.createUninitialized(input.shape());
Nd4j.getExecutioner().exec(new BroadcastMulOp(input, mask, masked2, 0, 2));
INDArray abs = Transforms.abs(masked2, true);
Transforms.pow(abs, pnorm, false);
INDArray pNorm = Transforms.pow(abs.sum(2), 1.0 / pnorm);
INDArray numerator;
if (pnorm == 2) {
numerator = input.dup();
} else {
INDArray absp2 = Transforms.pow(Transforms.abs(input, true), pnorm - 2, false);
numerator = input.mul(absp2);
}
INDArray denom = Transforms.pow(pNorm, pnorm - 1, false);
denom.rdivi(epsilon2d);
Nd4j.getExecutioner().execAndReturn(new BroadcastMulOp(numerator, denom, numerator, 0, 1));
//Apply mask
Nd4j.getExecutioner().exec(new BroadcastMulOp(numerator, mask, numerator, 0, 2));
return numerator;
case NONE:
throw new UnsupportedOperationException("NONE pooling type not supported");
default:
throw new UnsupportedOperationException("Unknown or not supported pooling type: " + poolingType);
}
}
Aggregations