ترغب بنشر مسار تعليمي؟ اضغط هنا

A Closer Look at Codistillation for Distributed Training

120   0   0.0 ( 0 )
 نشر من قبل Shagun Sodhani
 تاريخ النشر 2020
  مجال البحث الهندسة المعلوماتية
والبحث باللغة English




اسأل ChatGPT حول البحث

Codistillation has been proposed as a mechanism to share knowledge among concurrently trained models by encouraging them to represent the same function through an auxiliary loss. This contrasts with the more commonly used fully-synchronous data-parallel stochastic gradient descent methods, where different model replicas average their gradients (or parameters) at every iteration and thus maintain identical parameters. We investigate codistillation in a distributed training setup, complementing previous work which focused on extremely large batch sizes. Surprisingly, we find that even at moderate batch sizes, models trained with codistillation can perform as well as models trained with synchronous data-parallel methods, despite using a much weaker synchronization mechanism. These findings hold across a range of batch sizes and learning rate schedules, as well as different kinds of models and datasets. Obtaining this level of accuracy, however, requires properly accounting for the regularization effect of codistillation, which we highlight through several empirical observations. Overall, this work contributes to a better understanding of codistillation and how to best take advantage of it in a distributed computing environment.



قيم البحث

اقرأ أيضاً

We consider distributed optimization under communication constraints for training deep learning models. We propose a new algorithm, whose parameter updates rely on two forces: a regular gradient step, and a corrective direction dictated by the curren tly best-performing worker (leader). Our method differs from the parameter-averaging scheme EASGD in a number of ways: (i) our objective formulation does not change the location of stationary points compared to the original optimization problem; (ii) we avoid convergence decelerations caused by pulling local workers descending to different local minima to each other (i.e. to the average of their parameters); (iii) our update by design breaks the curse of symmetry (the phenomenon of being trapped in poorly generalizing sub-optimal solutions in symmetric non-convex landscapes); and (iv) our approach is more communication efficient since it broadcasts only parameters of the leader rather than all workers. We provide theoretical analysis of the batch version of the proposed algorithm, which we call Leader Gradient Descent (LGD), and its stochastic variant (LSGD). Finally, we implement an asynchronous version of our algorithm and extend it to the multi-leader setting, where we form groups of workers, each represented by its own local leader (the best performer in a group), and update each worker with a corrective direction comprised of two attractive forces: one to the local, and one to the global leader (the best performer among all workers). The multi-leader setting is well-aligned with current hardware architecture, where local workers forming a group lie within a single computational node and different groups correspond to different nodes. For training convolutional neural networks, we empirically demonstrate that our approach compares favorably to state-of-the-art baselines.
Current methods for training robust networks lead to a drop in test accuracy, which has led prior works to posit that a robustness-accuracy tradeoff may be inevitable in deep learning. We take a closer look at this phenomenon and first show that real image datasets are actually separated. With this property in mind, we then prove that robustness and accuracy should both be achievable for benchmark datasets through locally Lipschitz functions, and hence, there should be no inherent tradeoff between robustness and accuracy. Through extensive experiments with robustness methods, we argue that the gap between theory and practice arises from two limitations of current methods: either they fail to impose local Lipschitzness or they are insufficiently generalized. We explore combining dropout with robust training methods and obtain better generalization. We conclude that achieving robustness and accuracy in practice may require using methods that impose local Lipschitzness and augmenting them with deep learning generalization techniques. Code available at https://github.com/yangarbiter/robust-local-lipschitz
We study how the behavior of deep policy gradient algorithms reflects the conceptual framework motivating their development. To this end, we propose a fine-grained analysis of state-of-the-art methods based on key elements of this framework: gradient estimation, value prediction, and optimization landscapes. Our results show that the behavior of deep policy gradient algorithms often deviates from what their motivating framework would predict: the surrogate objective does not match the true reward landscape, learned value estimators fail to fit the true value function, and gradient estimates poorly correlate with the true gradient. The mismatch between predicted and empirical behavior we uncover highlights our poor understanding of current methods, and indicates the need to move beyond current benchmark-centric evaluation methods.
Being able to segment unseen classes not observed during training is an important technical challenge in deep learning, because of its potential to reduce the expensive annotation required for semantic segmentation. Prior zero-label semantic segmenta tion works approach this task by learning visual-semantic embeddings or generative models. However, they are prone to overfitting on the seen classes because there is no training signal for them. In this paper, we study the challenging generalized zero-label semantic segmentation task where the model has to segment both seen and unseen classes at test time. We assume that pixels of unseen classes could be present in the training images but without being annotated. Our idea is to capture the latent information on unseen classes by supervising the model with self-produced pseudo-labels for unlabeled pixels. We propose a consistency regularizer to filter out noisy pseudo-labels by taking the intersections of the pseudo-labels generated from different augmentations of the same image. Our framework generates pseudo-labels and then retrain the model with human-annotated and pseudo-labelled data. This procedure is repeated for several iterations. As a result, our approach achieves the new state-of-the-art on PascalVOC12 and COCO-stuff datasets in the challenging generalized zero-label semantic segmentation setting, surpassing other existing methods addressing this task with more complex strategies.
Modern machine learning algorithms are increasingly computationally demanding, requiring specialized hardware and distributed computation to achieve high performance in a reasonable time frame. Many hyperparameter search algorithms have been proposed for improving the efficiency of model selection, however their adaptation to the distributed compute environment is often ad-hoc. We propose Tune, a unified framework for model selection and training that provides a narrow-waist interface between training scripts and search algorithms. We show that this interface meets the requirements for a broad range of hyperparameter search algorithms, allows straightforward scaling of search to large clusters, and simplifies algorithm implementation. We demonstrate the implementation of several state-of-the-art hyperparameter search algorithms in Tune. Tune is available at http://ray.readthedocs.io/en/latest/tune.html.

الأسئلة المقترحة

التعليقات
جاري جلب التعليقات جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
mircosoft-partner

هل ترغب بارسال اشعارات عن اخر التحديثات في شمرا-اكاديميا