Losses
Helpers & Utilities · AiDotNet.Tensors.NN.Losses
No summary documentation available yet.
Methods
| Method | Summary |
|---|---|
Tensor<T> CosineEmbeddingLoss(Tensor<T> x1, Tensor<T> x2, Tensor<int> y, double margin, LossReduction reduction) |
|
Tensor<T> GaussianNllLoss(Tensor<T> input, Tensor<T> target, Tensor<T> variance, bool full, double eps, LossReduction reduction) |
|
Tensor<T> HingeEmbeddingLoss(Tensor<T> input, Tensor<int> y, double margin, LossReduction reduction) |
|
Tensor<T> KLDiv(Tensor<T> input, Tensor<T> target, bool logTarget, LossReduction reduction) |
|
Tensor<T> MarginRankingLoss(Tensor<T> input1, Tensor<T> input2, Tensor<T> y, double margin, LossReduction reduction) |
|
Tensor<T> MultiLabelMarginLoss(Tensor<T> input, Tensor<int> target, LossReduction reduction) |
|
Tensor<T> MultiLabelSoftMarginLoss(Tensor<T> input, Tensor<T> target, Tensor<T> weight, LossReduction reduction) |
|
Tensor<T> MultiMarginLoss(Tensor<T> input, Tensor<int> target, int p, double margin, Tensor<T> weight, LossReduction reduction) |
|
Tensor<T> PoissonNllLoss(Tensor<T> input, Tensor<T> target, bool logInput, bool full, double eps, LossReduction reduction) |
|
Tensor<T> SmoothL1Loss(Tensor<T> input, Tensor<T> target, double beta, LossReduction reduction) |
|
Tensor<T> TripletMarginLoss(Tensor<T> anchor, Tensor<T> positive, Tensor<T> negative, double margin, double p, double eps, LossReduction reduction) |
|
Tensor<T> TripletMarginWithDistanceLoss(Tensor<T> anchor, Tensor<T> positive, Tensor<T> negative, Func<Tensor<T>, Tensor<T>, Tensor<T>> distanceFn, double margin, bool swap, LossReduction reduction) |