Amos: An Adam-style Optimizer with Adaptive Weight Decay towards Model-Oriented Scale
This work addresses the problem of efficient and memory-friendly optimization for deep neural networks, particularly in pre-training large language models, representing an incremental improvement over existing optimizers like AdamW.
The paper introduces Amos, an Adam-style optimizer with adaptive weight decay, which accelerates convergence for training BERT and T5 models, achieving better validation loss in ≤70% of training steps and time while using ≤51% memory for slot variables.
We present Amos, a stochastic gradient-based optimizer designed for training deep neural networks. It can be viewed as an Adam optimizer with theoretically supported, adaptive learning-rate decay and weight decay. A key insight behind Amos is that it leverages model-specific information to determine the initial learning-rate and decaying schedules. When used for pre-training BERT variants and T5, Amos consistently converges faster than the state-of-the-art settings of AdamW, achieving better validation loss within <=70% training steps and time, while requiring <=51% memory for slot variables. Our code is open-sourced at: https://github.com/google-research/jestimator