Large-batch training has been essential in leveraging large-scale datasets and models in deep learning. While it is computationally beneficial to use large batch sizes, it often requires a specially designed learning rate (LR) schedule to achieve a comparable level of performance as in smaller batch training. Especially, when the number of training epochs is constrained, the use of a large LR and a warmup strategy is critical in the final performance of large-batch training due to the reduced number of updating steps. In this work, we propose an automated LR scheduling algorithm which is effective for neural network training with a large batch size under the given epoch budget. In specific, the whole schedule consists of two phases: adaptive warmup and predefined decay, where the LR is increased until the training loss no longer decreases and decreased to zero until the end of training. Here, whether the training loss has reached the minimum value is robustly checked with Gaussian process smoothing in an online manner with a low computational burden. Coupled with adaptive stochastic optimizers such as AdamP and LAMB, the proposed scheduler successfully adjusts the LRs without cumbersome hyperparameter tuning and achieves comparable or better performances than tuned baselines on various image classification benchmarks and architectures with a wide range of batch sizes.
University of North Carolina at Charlotte（北卡罗来纳大学夏洛特分校）；Iowa State University（爱荷华州立大学）；Oak Ridge National Laboratory（橡树岭国家实验室）
Today high-performance computing (HPC) platforms are still dominated by batch jobs. Accordingly, effective batch job scheduling is crucial to obtain high system efficiency. Existing HPC batch job schedulers typically leverage heuristic priority functions to prioritize and schedule jobs. But, once configured and deployed by the experts, such priority functions can hardly adapt to the changes of job loads, optimization goals, or system settings, potentially leading to degraded system efficiency when changes occur. To address this fundamental issue, we present RLScheduler, an automated HPC batch …
National University of Singapore（新加坡国立大学）；ByteDance（字节跳动）
Efficient GPU resource scheduling is essential to maximize resource utilization and save training costs for the increasing amount of deep learning workloads in shared GPU clusters. Existing GPU schedulers largely rely on static policies to leverage the performance characteristics of deep learning jobs. However, they can hardly reach optimal efficiency due to the lack of elasticity. To address the problem, we propose ONES, an ONline Evolutionary Scheduler for elastic batch size orchestration. ONES automatically manages the elasticity of each job based on the training batch size, so as to maximi…
Scheduling the batch size to increase is an effective strategy to control gradient noise when training deep neural networks. Current approaches implement scheduling heuristics that neglect structure within the optimization procedure, limiting their flexibility to the training dynamics and capacity to discern the impact of their adaptations on generalization. We introduce Arbiter as a new hyperparameter optimization algorithm to perform batch size adaptations for learnable scheduling heuristics using gradients from a meta-objective function, which overcomes previous heuristic constraints by enf…
Automated Batch Distillation Process Simulation for a Large Hybrid Dataset for Deep Anomaly Detection
Fraunhofer Institute for Industrial Mathematics (ITWM)（弗劳恩霍夫工业数学研究所）；Laboratory of Engineering Thermodynamics (LTD)（工程热力学实验室）；RPTU Kaiserslautern（科隆大学（卡塞尔拉特恩分校））
Data augmentation have been intensively used in training deep neural network to improve the generalization, whether in original space (e.g., image space) or representation space. Although being successful, the connection between the synthesized data and the original data is largely ignored in training, without considering the distribution information that the synthesized samples are surrounding the original sample in training. Hence, the behavior of the network is not optimized for this. However, that behavior is crucially important for generalization, even in the adversarial setting, for the …
BERT has recently attracted a lot of attention in natural language understanding (NLU) and achieved state-of-the-art results in various NLU tasks. However, its success requires large deep neural networks and huge amount of data, which result in long training time and impede development progress. Using stochastic gradient methods with large mini-batch has been advocated as an efficient tool to reduce the training time. Along this line of research, LAMB is a prominent example that reduces the training time of BERT from 3 days to 76 minutes on a TPUv3 Pod. In this paper, we propose an accelerated…
Mini-batch sub-sampling in neural network training is unavoidable, due to growing data demands, memory-limited computational resources such as graphical processing units (GPUs), and the dynamics of on-line learning. In this study we specifically distinguish between static mini-batch sub-sampled loss functions, where mini-batches are intermittently fixed during training, resulting in smooth but biased loss functions; and the dynamic sub-sampling equivalent, where new mini-batches are sampled at every loss evaluation, trading bias for variance in sampling induced discontinuities. These render au…
FedQueue: Queue-Aware Federated Learning for Cross-Facility HPC Training
Computer Science Division, Argonne National Laboratory, Lemont, IL, USA（阿贡国家实验室计算机科学部）；Learning Division, Argonne National Laboratory, Lemont, IL, USA（阿贡国家实验室学习部）
Automated writing evaluation (AWE) has been shown to be an effective mechanism for quickly providing feedback to students. It has already seen wide adoption in enterprise-scale applications and is starting to be adopted in large-scale contexts. Training an AWE model has historically required a single batch of several hundred writing examples and human scores for each of them. This requirement limits large-scale adoption of AWE since human-scoring essays is costly. Here we evaluate algorithms for ensuring that AWE models are consistently trained using the most informative essays. Our results sh…
Moog Inc. has automated the evaluation of copper (Cu) alloy grain size using a deep-learning convolutional neural network (CNN). The proof-of-concept automated image acquisition and batch-wise image processing offers the potential for significantly reduced labor, improved accuracy of grain evaluation, and decreased overall turnaround times for approving Cu alloy bar stock for use in flight critical aircraft hardware. A classification accuracy of 91.1% on individual sub-images of the Cu alloy coupons was achieved. Process development included minimizing the variation in acquired image color, br…
