Sanity Checks for Lottery Tickets: Does Your Winning Ticket Really Win the Jackpot?
This work addresses methodological inconsistencies in a foundational area of neural network pruning, which is incremental but important for researchers in machine learning efficiency.
The authors tackled inconsistencies in identifying 'winning tickets' in the lottery ticket hypothesis by proposing a more rigorous definition and conducting extensive experiments across major DNN architectures. They found that key training hyperparameters and architecture characteristics are highly correlated with the identification of winning tickets, leading to a guideline for parameter settings.
There have been long-standing controversies and inconsistencies over the experiment setup and criteria for identifying the "winning ticket" in literature. To reconcile such, we revisit the definition of lottery ticket hypothesis, with comprehensive and more rigorous conditions. Under our new definition, we show concrete evidence to clarify whether the winning ticket exists across the major DNN architectures and/or applications. Through extensive experiments, we perform quantitative analysis on the correlations between winning tickets and various experimental factors, and empirically study the patterns of our observations. We find that the key training hyperparameters, such as learning rate and training epochs, as well as the architecture characteristics such as capacities and residual connections, are all highly correlated with whether and when the winning tickets can be identified. Based on our analysis, we summarize a guideline for parameter settings in regards of specific architecture characteristics, which we hope to catalyze the research progress on the topic of lottery ticket hypothesis. Our codes are publicly available at: https://github.com/boone891214/sanity-check-LTH.