Building an Efficiency Pipeline: Commutativity and Cumulativeness of Efficiency Operators for Transformers
This provides practical guidelines for applying efficiency methods in real-world NLP applications, though it is incremental as it builds on existing techniques.
The paper investigates the feasibility of combining multiple efficiency methods (like pruning and quantization) into a pipeline for NLP models, finding that these operators are commutative (order doesn't matter) and cumulative (combined effects can be estimated from individual ones).
There exists a wide variety of efficiency methods for natural language processing (NLP) tasks, such as pruning, distillation, dynamic inference, quantization, etc. We can consider an efficiency method as an operator applied on a model. Naturally, we may construct a pipeline of multiple efficiency methods, i.e., to apply multiple operators on the model sequentially. In this paper, we study the plausibility of this idea, and more importantly, the commutativity and cumulativeness of efficiency operators. We make two interesting observations: (1) Efficiency operators are commutative -- the order of efficiency methods within the pipeline has little impact on the final results; (2) Efficiency operators are also cumulative -- the final results of combining several efficiency methods can be estimated by combining the results of individual methods. These observations deepen our understanding of efficiency operators and provide useful guidelines for their real-world applications.