Accelerating the Low-Rank Decomposed Models
This addresses the issue of model compression for AI practitioners, but appears incremental as it builds on existing low-rank decomposition methods.
The paper tackles the problem of low-rank decomposition in AI models causing increased depth and latency, and presents a modified technique to achieve high accuracy, low memory consumption, and faster training and inference.
Tensor decomposition is a mathematically supported technique for data compression. It consists of applying some kind of a Low Rank Decomposition technique on the tensors or matrices in order to reduce the redundancy of the data. However, it is not a popular technique for compressing the AI models duo to the high number of new layers added to the architecture after decomposition. Although the number of parameters could shrink significantly, it could result in the model be more than twice deeper which could add some latency to the training or inference. In this paper, we present a comprehensive study about how to modify low rank decomposition technique in AI models so that we could benefit from both high accuracy and low memory consumption as well as speeding up the training and inference