Adaptive Sampled Softmax with Kernel Based Sampling

AdaptiveSampledSoftmaxwithKernelBasedSamplingGuyBlanc1SteffenRendle2Abstractmizationalgorithm,e.g.,stochasticgradientdescent,needstocomputethegradientswithrespecttotheloss.WhentheSoftmaxisthemostco...