Sampled Softmax with Random Fourier Features
arXiv:1907.10747
Abstract
The computational cost of training with softmax cross entropy loss grows linearly with the number of classes. For the settings where a large number of classes are involved, a common method to speed up training is to sample a subset of classes and utilize an estimate of the loss gradient based on these classes, known as the sampled softmax method. However, the sampled softmax provides a biased estimate of the gradient unless the samples are drawn from the exact softmax distribution, which is again expensive to compute. Therefore, a widely employed practical approach involves sampling from a simpler distribution in the hope of approximating the exact softmax distribution. In this paper, we develop the first theoretical understanding of the role that different sampling distributions play in determining the quality of sampled softmax. Motivated by our analysis and the work on kernel-based sampling, we propose the Random Fourier Softmax (RF-softmax) method that utilizes the powerful Random Fourier Features to enable more efficient and accurate sampling from an approximate softmax distribution. We show that RF-softmax leads to low bias in estimation in terms of both the full softmax distribution and the full softmax gradient. Furthermore, the cost of RF-softmax scales only logarithmically with the number of classes.
arXiv commentIn NeurIPS 2019
References in corpus (11)
- Distributed Representations of Words and Phrases and their Compositionality
- Orthogonal Random Features
- An Exploration of Softmax Alternatives Belonging to the Spherical Loss Family
- Compact Random Feature Maps
- Optimal Rates for Random Fourier Features
- Deep Networks With Large Output Spaces
- Stochastic Negative Mining for Learning with Large Output Spaces
- Unbiased scalable softmax optimization
- Probabilistic Structured Predictors
- Fast Amortized Inference and Learning in Log-linear Models with Randomly Perturbed Nearest Neighbor Search
- DS-MLR: Exploiting Double Separability for Scaling up Distributed Multinomial Logistic Regression