MILO: Model-Agnostic Subset Selection Framework for Efficient Model Training and Tuning
This addresses the efficiency bottleneck for researchers and practitioners in machine learning by enabling faster model development and tuning, though it is incremental as it builds on existing subset selection methods.
The paper tackles the problem of computationally intensive training and hyperparameter tuning of deep networks on large datasets by proposing MILO, a model-agnostic subset selection framework that decouples subset selection from model training, resulting in 3x to 10x faster model training and 20x to 75x faster hyperparameter tuning without performance loss.
Training deep networks and tuning hyperparameters on large datasets is computationally intensive. One of the primary research directions for efficient training is to reduce training costs by selecting well-generalizable subsets of training data. Compared to simple adaptive random subset selection baselines, existing intelligent subset selection approaches are not competitive due to the time-consuming subset selection step, which involves computing model-dependent gradients and feature embeddings and applies greedy maximization of submodular objectives. Our key insight is that removing the reliance on downstream model parameters enables subset selection as a pre-processing step and enables one to train multiple models at no additional cost. In this work, we propose MILO, a model-agnostic subset selection framework that decouples the subset selection from model training while enabling superior model convergence and performance by using an easy-to-hard curriculum. Our empirical results indicate that MILO can train models $3\times - 10 \times$ faster and tune hyperparameters $20\times - 75 \times$ faster than full-dataset training or tuning without compromising performance.