Benchmark of DNN Model Search at Deployment Time

TL;DR

This paper introduces a set of automated model search strategies for deploying pre-trained deep learning models, reducing manual effort and improving selection accuracy across diverse applications.

Contribution

It proposes multiple similarity and non-similarity-based model search methods and evaluates their effectiveness in various deployment scenarios.

Findings

Asymmetric similarity-based measurement outperforms symmetric methods in most workloads.

The proposed approaches reduce manual effort in model selection.

Evaluation covers diverse tasks like image recognition and NLP.

Abstract

Deep learning has become the most popular direction in machine learning and artificial intelligence. However, the preparation of training data, as well as model training, are often time-consuming and become the bottleneck of the end-to-end machine learning lifecycle. Reusing models for inferring a dataset can avoid the costs of retraining. However, when there are multiple candidate models, it is challenging to discover the right model for reuse. Although there exist a number of model sharing platforms such as ModelDB, TensorFlow Hub, PyTorch Hub, and DLHub, most of these systems require model uploaders to manually specify the details of each model and model downloaders to screen keyword search results for selecting a model. We are lacking a highly productive model search tool that selects models for deployment without the need for any manual inspection and/or labeled data from the target domain. This paper proposes multiple model search strategies including various similarity-based approaches and non-similarity-based approaches. We design, implement, and evaluate these approaches on multiple model inference scenarios, including activity recognition, image recognition, text classification, natural language processing, and entity matching. The experimental evaluation showed that our proposed asymmetric similarity-based measurement, adaptivity, outperformed symmetric similarity-based measurements and non-similarity-based measurements in most of the workloads.