predictSoftly

open override fun predictSoftly(dataset: Dataset, batchSize: Int, callbacks: List<Callback>): Array<FloatArray>

Generates output predictions for the input samples. Each prediction is a vector of probabilities instead of specific class in predict method.

Return

Array of labels. All labels are vectors that represents the probability distributions of a list of potential outcomes. The length is equal to the Number of samples on the dataset.

Parameters

dataset

Data to predict on.

batchSize

Number of samples per batch of computation.

callbacks

Callbacks to be used during prediction phase.

open override fun predictSoftly(inputData: FloatArray, predictionTensorName: String): FloatArray

Predicts vector of probabilities instead of specific class in predict method.

Return

Vector that represents the probability distributions of a list of potential outcomes

Parameters

inputData

The single example with unknown vector of probabilities.

predictionTensorName

The name of prediction tensor. It could be changed, if you need to get alternative outputs from model graph.