论文标题
比较DL框架的抽象成本
Comparing the costs of abstraction for DL frameworks
论文作者
论文摘要
用于实施，培训和测试深度学习（DL）模型的高级抽象比比皆是。此类框架主要是通过抽象任意神经体系结构的实现细节来发挥作用，从而使研究人员和工程师能够专注于设计。原则上，这样的框架可能是“零成本的抽象”。实际上，它们会产生翻译和间接开销。我们研究了DL模型的工程生命周期中的哪些点，支付了最高成本以及是否可以缓解它们。我们使用Pytorch，Libtorch，Torchscript和Cudnn在代表性数据集上训练，测试和评估代表性的DL模型，从而比较准确性，执行时间和内存效率。
High level abstractions for implementing, training, and testing Deep Learning (DL) models abound. Such frameworks function primarily by abstracting away the implementation details of arbitrary neural architectures, thereby enabling researchers and engineers to focus on design. In principle, such frameworks could be "zero-cost abstractions"; in practice, they incur translation and indirection overheads. We study at which points exactly in the engineering life-cycle of a DL model the highest costs are paid and whether they can be mitigated. We train, test, and evaluate a representative DL model using PyTorch, LibTorch, TorchScript, and cuDNN on representative datasets, comparing accuracy, execution time and memory efficiency.
