[1503.02531] Distilling the Knowledge in a Neural Network

Abstract:A very simple way to improve the performance of almost any machine learning algorithm is to train many different models on the same data and then to average their predictions. Unfortunately, making predictions using a whole ensemble of models is cumbersome and may be too computationally expensive to allow deployment to a large number of users, especially if the individual models are large neural nets. Caruana and his collaborators have shown that it is possible to compress the knowledge in an ensemble into a single model which is much easier to deploy and we develop this approach further using a different compression technique. We achieve some surprising results on MNIST and we show that we can significantly improve the acoustic model of a heavily used commercial system by distilling the knowledge in an ensemble of models into a single model. We also introduce a new type of ensemble composed of one or more full models and many specialist models which learn to distinguish fine-grained classes that the full models confuse. Unlike a mixture of experts, these specialist models can be trained rapidly and in parallel.

Distilling the Knowledge in a Neural Network Geoffrey Hinton∗ † Oriol Vinyals† Jeff Dean Google Inc. Google Inc. Google Inc. arXiv:1503.02531v1 [stat.ML] 9 Mar 2015 Mountain View Mountain View Mountain View geoffhinton@google.com…

related reading

• Distill — Latest articles about machine learningdistill.pub
• [1511.03643] Unifying distillation and privileged informationarxiv.org
• Distillation Walkthroughvladfeinberg.com
• [1706.04599] On Calibration of Modern Neural Networksarxiv.org
• Understanding deep learning requires rethinking generalizationarxiv.org
• [1701.06538] Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layerarxiv.org
• Zero-Shot Knowledge Distillation in Deep Networksarxiv.org
• [2605.01172] A Theory of Generalization in Deep Learningarxiv.org
• [2201.02177] Grokking: Generalization Beyond Overfitting on Small Algorithmic Datasetsarxiv.org
• arxiv.org/pdf/1805.08522arxiv.org
• Efficient LLM inferencefinbarrtimbers.substack.com
• [2605.12715] Scaling Laws for Mixture Pretraining Under Data Constraintsarxiv.org