arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2012.09272cs.LG
收藏

深度神经网络大批量分布式训练的数据优化

Data optimization for large batch distributed training of deep neural networks

Oak Ridge National Lab(橡树岭国家实验室)

Shubhankar Gahlot, Junqi Yin, Mallikarjun Shankar

摘要:

随着数据和模型的增长,深度学习(DL)中的分布式训练是常见的做法。当前深度神经网络的分布式训练实践在大规模操作时面临通信瓶颈的挑战,并且随着全局批量大小的增加,模型准确性恶化。目前的解决方案集中在提高消息交换效率以及实现技术来调整训练过程中的批量大小和模型。训练精度的损失通常是因为损失函数陷入局部最小值。我们观察到损失景观最小化是由模型和训练数据塑造的,并提出了一种数据优化方法,该方法利用机器学习隐式地平滑损失景观,从而减少局部最小值。我们的方法过滤掉了对特征学习不太重要的数据点,使我们能够在更大的批量上加快模型的训练,以提高准确性。

英文摘要:

Distributed training in deep learning (DL) is common practice as data and models grow. The current practice for distributed training of deep neural networks faces the challenges of communication bottlenecks when operating at scale, and model accuracy deterioration with an increase in global batch size. Present solutions focus on improving message exchange efficiency as well as implementing techniques to tweak batch sizes and models in the training process. The loss of training accuracy typically happens because the loss function gets trapped in a local minima. We observe that the loss landscape minimization is shaped by both the model and training data and propose a data optimization approach that utilizes machine learning to implicitly smooth out the loss landscape resulting in fewer local minima. Our approach filters out data points which are less important to feature learning, enabling us to speed up the training of models on larger batch sizes to improved accuracy.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑