اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدHow Do Adam and Training Strategies Help BNNs Optimization?
72
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
The best performing Binary Neural Networks (BNNs) are usually attained using Adam optimization and its multi-step training variants. However, to the best of our knowledge, few studies explore the fundamental reasons why Adam is superior to other optimizers like SGD for BNN optimization or provide analytical explanations that support specific training strategies. To address this, in this paper we first investigate the trajectories of gradients and weights in BNNs during the training process. We show the regularization effect of second-order momentum in Adam is crucial to revitalize the weights that are dead due to the activation saturation in BNNs. We find that Adam, through its adaptive learning rate strategy, is better equipped to handle the rugged loss surface of BNNs and reaches a better optimum with higher generalization ability. Furthermore, we inspect the intriguing role of the real-valued weights in binary networks, and reveal the effect of weight decay on the stability and sluggishness of BNN optimization. Through extensive experiments and analysis, we derive a simple training scheme, building on existing Adam-based optimization, which achieves 70.5% top-1 accuracy on the ImageNet dataset using the same architecture as the state-of-the-art ReActNet while achieving 1.1% higher accuracy. Code and models are available at https://github.com/liuzechun/AdamBNN.
قيم البحث
اقرأ أيضاً
In this paper, we introduce a new perspective on training deep neural networks capable of state-of-the-art performance without the need for the expensive over-parameterization by proposing the concept of In-Time Over-Parameterization (ITOP) in sparse
training. By starting from a random sparse network and continuously exploring sparse connectivities during training, we can perform an Over-Parameterization in the space-time manifold, closing the gap in the expressibility between sparse training and dense training. We further use ITOP to understand the underlying mechanism of Dynamic Sparse Training (DST) and indicate that the benefits of DST come from its ability to consider across time all possible parameters when searching for the optimal sparse connectivity. As long as there are sufficient parameters that have been reliably explored during training, DST can outperform the dense neural network by a large margin. We present a series of experiments to support our conjecture and achieve the state-of-the-art sparse training performance with ResNet-50 on ImageNet. More impressively, our method achieves dominant performance over the overparameterization-based sparse methods at extreme sparsity levels. When trained on CIFAR-100, our method can match the performance of the dense model even at an extreme sparsity (98%). Code can be found https://github.com/Shiweiliuiiiiiii/In-Time-Over-Parameterization.
Binary Neural Networks (BNNs) show promising progress in reducing computational and memory costs but suffer from substantial accuracy degradation compared to their real-valued counterparts on large-scale datasets, e.g., ImageNet. Previous work mainly
focused on reducing quantization errors of weights and activations, whereby a series of approximation methods and sophisticated training tricks have been proposed. In this work, we make several observations that challenge conventional wisdom. We revisit some commonly used techniques, such as scaling factors and custom gradients, and show that these methods are not crucial in training well-performing BNNs. On the contrary, we suggest several design principles for BNNs based on the insights learned and demonstrate that highly accurate BNNs can be trained from scratch with a simple training strategy. We propose a new BNN architecture BinaryDenseNet, which significantly surpasses all existing 1-bit CNNs on ImageNet without tricks. In our experiments, BinaryDenseNet achieves 18.6% and 7.6% relative improvement over the well-known XNOR-Network and the current state-of-the-art Bi-Real Net in terms of top-1 accuracy on ImageNet, respectively.
Gaining a better understanding of how and what machine learning systems learn is important to increase confidence in their decisions and catalyze further research. In this paper, we analyze the predictions made by a specific type of recurrent neural
network, mixture density RNNs (MD-RNNs). These networks learn to model predictions as a combination of multiple Gaussian distributions, making them particularly interesting for problems where a sequence of inputs may lead to several distinct future possibilities. An example is learning internal models of an environment, where different events may or may not occur, but where the average over different events is not meaningful. By analyzing the predictions made by trained MD-RNNs, we find that their different Gaussian components have two complementary roles: 1) Separately modeling different stochastic events and 2) Separately modeling scenarios governed by different rules. These findings increase our understanding of what is learned by predictive MD-RNNs, and open up new research directions for further understanding how we can benefit from their self-organizing model decomposition.
Mixup is a popular data augmentation technique based on taking convex combinations of pairs of examples and their labels. This simple technique has been shown to substantially improve both the robustness and the generalization of the trained model. H
owever, it is not well-understood why such improvement occurs. In this paper, we provide theoretical analysis to demonstrate how using Mixup in training helps model robustness and generalization. For robustness, we show that minimizing the Mixup loss corresponds to approximately minimizing an upper bound of the adversarial loss. This explains why models obtained by Mixup training exhibits robustness to several kinds of adversarial attacks such as Fast Gradient Sign Method (FGSM). For generalization, we prove that Mixup augmentation corresponds to a specific type of data-adaptive regularization which reduces overfitting. Our analysis provides new insights and a framework to understand Mixup.
Recent advances in one-shot semi-supervised learning have lowered the barrier for deep learning of new applications. However, the state-of-the-art for semi-supervised learning is slow to train and the performance is sensitive to the choices of the la
beled data and hyper-parameter values. In this paper, we present a one-shot semi-supervised learning method that trains up to an order of magnitude faster and is more robust than state-of-the-art methods. Specifically, we show that by combining semi-supervised learning with a one-stage, single network version of self-training, our FROST methodology trains faster and is more robust to choices for the labeled samples and changes in hyper-parameters. Our experiments demonstrate FROSTs capability to perform well when the composition of the unlabeled data is unknown; that is when the unlabeled data contain unequal numbers of each class and can contain out-of-distribution examples that dont belong to any of the training classes. High performance, speed of training, and insensitivity to hyper-parameters make FROST the most practical method for one-shot semi-supervised training. Our code is available at https://github.com/HelenaELiu/FROST.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
