A DNN Optimizer that Improves over AdaBelief by Suppression of the Adaptive Stepsize Range

arxiv(2023)

引用 0|浏览15
暂无评分
摘要
We make contributions towards improving adaptive-optimizer performance. Our improvements are based on suppression of the range of adaptive stepsizes in the AdaBelief optimizer. Firstly, we show that the particular placement of the parameter epsilon within the update expressions of AdaBelief reduces the range of the adaptive stepsizes, making AdaBelief closer to SGD with momentum. Secondly, we extend AdaBelief by further suppressing the range of the adaptive stepsizes. To achieve the above goal, we perform mutual layerwise vector projections between the gradient g_t and its first momentum m_t before using them to estimate the second momentum. The new optimization method is referred to as Aida. Thirdly, extensive experimental results show that Aida outperforms nine optimizers when training transformers and LSTMs for NLP, and VGG and ResNet for image classification over CIAF10 and CIFAR100 while matching the best performance of the nine methods when training WGAN-GP models for image generation tasks. Furthermore, Aida produces higher validation accuracies than AdaBelief for training ResNet18 over ImageNet. Code is available at this URL
更多
查看译文
关键词
dnn optimizer,adabelief,range
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要