Software engineers develop, fine-tune, and deploy deep learning (DL) models using a variety of development frameworks and runtime environments. DL model converters move models between frameworks and to runtime environments. Conversion errors compromise model quality and disrupt deployment. However, the failure characteristics of DL model converters are unknown, adding risk when using DL interoperability technologies. This paper analyzes failures in DL model converters. We survey software engineers about DL interoperability tools, use cases, and pain points (N=92). Then, we characterize failures in model converters associated with the main interoperability tool, ONNX (N=200 issues in PyTorch and TensorFlow). Finally, we formulate and test two hypotheses about structural causes for the failures we studied. We find that the node conversion stage of a model converter accounts for ∼75% of the defects and 33% of reported failure are related to semantically incorrect models. The cause of semantically incorrect models is elusive, but models with behaviour inconsistencies share operator sequences. Our results motivate future research on making DL interoperability software simpler to maintain, extend, and validate. Research into behavioural tolerances and architectural coverage metrics would be fruitful.

Context: Many engineering organizations are reimplementing and extending deep neural networks from the research community. We describe this process as deep learning model reengineering. Deep learning model reengineering — reusing, replicating, adapting, and enhancing state-of-the-art deep learning approaches — is challenging for reasons including under-documented reference models, changing requirements, and the cost of implementation and testing. Objective: Prior work has characterized the challenges of deep learning model development, but as yet we know little about the deep learning model reengineering process and its common challenges. Prior work has examined DL systems from a “product” view, examining defects from projects regardless of the engineers’ purpose. Our study is focused on reengineering activities from a “process” view, and focuses on engineers specifically engaged in the reengineering process. Method: Our goal is to understand the characteristics and challenges of deep learning model reengineering. We conducted a mixed-methods case study of this phenomenon, focusing on the context of computer vision. Our results draw from two data sources: defects reported in open-source reeengineering projects, and interviews conducted with practitioners and the leaders of a reengineering team. From the defect data source, we analyzed 348 defects from 27 open-source deep learning projects. Meanwhile, our reengineering team replicated 7 deep learning models over two years; we interviewed 2 open-source contributors, 4 practitioners, and 6 reengineering team leaders to understand their experiences. Results: Our results describe how deep learning-based computer vision techniques are reengineered, quantitatively analyze the distribution of defects in this process, and qualitatively discuss challenges and practices. We found that most defects (58%) are reported by re-users, and that reproducibility-related defects tend to be discovered during training (68% of them are). Our analysis shows that most environment defects (88%) are interface defects, and most environment defects (46%) are caused by API defects. We found that training defects have diverse symptoms and root causes. We identified four main challenges in the DL reengineering process: model operationalization, performance debugging, portability of DL operations, and customized data pipeline. Integrating our quantitative and qualitative data, we propose a novel reengineering workflow. Conclusions: Our findings inform several conclusion, including: standardizing model reengineering practices, developing validation tools to support model reengineering, automated support beyond manual model reengineering, and measuring additional unknown aspects of model reengineering.

The development and training of deep learning models have become increasingly costly and complex. Consequently, software engineers are adopting pre-trained models (PTMs) for their downstream applications. The dynamics of the PTM supply chain remain largely unexplored, signaling a clear need for structured datasets that document not only the metadata but also the subsequent applications of these models. Without such data, the MSR community cannot comprehensively understand the impact of PTM adoption and reuse. This paper presents the PeaTMOSS dataset, which comprises metadata for 281,638 PTMs and detailed snapshots for all PTMs with over 50 monthly downloads (14,296 PTMs), along with 28,575 open-source software repositories from GitHub that utilize these models. Additionally, the dataset includes 44,337 mappings from 15,129 downstream GitHub repositories to the 2,530 PTMs they use. To enhance the dataset's comprehensiveness, we developed prompts for a large language model to automatically extract model metadata, including the model's training datasets, parameters, and evaluation metrics. Our analysis of this dataset provides the first summary statistics for the PTM supply chain, showing the trend of PTM development and common shortcomings of PTM package documentation. Our example application reveals inconsistencies in software licenses across PTMs and their dependent projects. PeaTMOSS lays the foundation for future research, offering rich opportunities to investigate the PTM supply chain. We outline mining opportunities on PTMs, their downstream usage, and cross-cutting questions. Our artifact is available at https://github.com/PurdueDualityLab/PeaTMOSS-Artifact. Our dataset is available at https://transfer.rcac.purdue.edu/file-manager?origin_id=ff978999-16c2-4b50-ac7a-947ffdc3eb1d&origin_path=%2F.

Background: Software Package Registries (SPRs) are an integral part of the software supply chain. These collaborative platforms unite contributors, users, and code for streamlined package management. Prior work has characterized the SPRs associated with traditional software, such as NPM (JavaScript) and PyPI (Python). Pre-Trained Model (PTM) Registries are an emerging class of SPR of increasing importance, because they support the deep learning supply chain. A growing body of empirical research has examined PTM registries from various angles, such as vulnerabilities, reuse processes, and evolution. However, no synthesis provides a systematic understanding of current knowledge. Furthermore, much of the existing research includes non-quantified qualitative observations. Aims: First, we aim to provide a systematic knowledge synthesis. Second, we quantify qualitative claims. Methods: We conducted a systematic literature review (SLR). We then observed that some of the claims are qualitative, lacking quantitative evidence. We identify quantifiable metrics associated with those claims, and measure in order to substantiate these claims. Results: We identify 12 claims about PTM reuse on the HuggingFace platform, 4 of which lack quantitative support. We tested 3 of these claims through a quantitative analysis, and directly compare the fourth with traditional software. Our most notable findings are: (1) PTMs have a significantly higher turnover rate than traditional software, indicating more rapid evolution; and (2) There is a strong correlation between documentation quality and PTM popularity. Conclusions: Our findings validate several qualitative research claims with concrete metrics, confirming prior research. Our measures motivate further research on the dynamics of PTM reuse.

This study explores the performance of machine learning algorithms on the classification of fossil teeth in the Family Bovidae. Isolated bovid teeth are typically the most common fossils found in southern Africa and they often constitute the basis for paleoenvironmental reconstructions. Taxonomic identification of fossil bovid teeth, however, is often imprecise and subjective. Using modern teeth with known taxons, machine learning algorithms can be trained to classify fossils. Previous work by Brophy et al. [Quantitative morphological analysis of bovid teeth and implications for paleoenvironmental reconstruction of plovers lake, Gauteng Province, South Africa, J. Archaeol. Sci. 41 (2014), pp. 376–388] uses elliptical Fourier analysis of the form (size and shape) of the outline of the occlusal surface of each tooth as features in a linear discriminant analysis (LDA) framework. This manuscript expands on that previous work by exploring how different machine learning approaches classify the teeth and testing which technique is best for classification. In addition to LDA, four other machine learning techniques were considered (neural networks, nuclear penalized multinomial regression,random forests, and support vector machines) with support vector machines and random forests performing the best in terms of log loss and classification rate.

As innovation in deep learning continues, many engineers are incorporating Pre-Trained Models (PTMs) as components in computer systems. Some PTMs are foundation models, and others are fine-tuned variations adapted to different needs. When these PTMs are named well, it facilitates model discovery and reuse. However, prior research has shown that model names are not always well chosen and can sometimes be inaccurate and misleading. The naming practices for PTM packages have not been systematically studied, which hampers engineers’ ability to efficiently search for and reliably reuse these models. In this paper, we conduct the first empirical investigation of PTM naming practices in the Hugging Face PTM registry. We begin by reporting on a survey of 108 Hugging Face users, highlighting differences from traditional software package naming and presenting findings on PTM naming practices. The survey results indicate a mismatch between engineers’ preferences and current practices in PTM naming. We then introduce DARA, the first automated D NN AR chitecture A ssessment technique designed to detect PTM naming inconsistencies. Our results demonstrate that architectural information alone is sufficient to detect these inconsistencies, achieving an accuracy of 94% in identifying model types and promising performance (over 70%) in other architectural metadata as well. We also highlight potential use cases for automated naming tools, such as model validation, PTM metadata generation and verification, and plagiarism detection. Our study provides a foundation for automating naming inconsistency detection. Finally, we envision future work focusing on automated tools for standardizing package naming, improving model selection and reuse, and strengthening the security of the PTM supply chain. “The main idea is to treat a program as a piece of literature, addressed to human beings rather than to a computer” —D. Knuth