Ensemble-Compression: A New Method for Parallel Training of Deep Neural Networks
arXiv:1606.00575
Abstract
Parallelization framework has become a necessity to speed up the training of deep neural networks (DNN) recently. Such framework typically employs the Model Average approach, denoted as MA-DNN, in which parallel workers conduct respective training based on their own local data while the parameters of local models are periodically communicated and averaged to obtain a global model which serves as the new start of local models. However, since DNN is a highly non-convex model, averaging parameters cannot ensure that such global model can perform better than those local models. To tackle this problem, we introduce a new parallel training framework called Ensemble-Compression, denoted as EC-DNN. In this framework, we propose to aggregate the local models by ensemble, i.e., averaging the outputs of local models instead of the parameters. As most of prevalent loss functions are convex to the output of DNN, the performance of ensemble-based global model is guaranteed to be at least as good as the average performance of local models. However, a big challenge lies in the explosion of model size since each round of ensemble can give rise to multiple times size increment. Thus, we carry out model compression after each ensemble, specialized by a distillation based method in this paper, to reduce the size of the global model to be the same as the local ones. Our experimental results demonstrate the prominent advantage of EC-DNN over MA-DNN in terms of both accuracy and speedup.
ECML 2017
References in corpus (9)
- Distilling the Knowledge in a Neural Network
- Caffe: Convolutional Architecture for Fast Feature Embedding
- FitNets: Hints for Thin Deep Nets
- Exploiting Linear Structure Within Convolutional Networks for Efficient Evaluation
- Going Deeper with Convolutions
- Compressing Deep Convolutional Networks using Vector Quantization
- Revisiting Distributed Synchronous SGD
- Compressing Neural Networks with the Hashing Trick
- Parallel training of DNNs with Natural Gradient and Parameter Averaging