The State of Documentation Practices of Third-party Machine Learning Models and Datasets
This highlights a critical transparency gap for users relying on model stores, though it is incremental as it builds on existing documentation standards.
The study assessed documentation practices for third-party machine learning models and datasets on Hugging Face, finding that only 39.62% of models and 28.48% of datasets have documentation, with inconsistencies in ethics and transparency-related content.
Model stores offer third-party ML models and datasets for easy project integration, minimizing coding efforts. One might hope to find detailed specifications of these models and datasets in the documentation, leveraging documentation standards such as model and dataset cards. In this study, we use statistical analysis and hybrid card sorting to assess the state of the practice of documenting model cards and dataset cards in one of the largest model stores in use today--Hugging Face (HF). Our findings show that only 21,902 models (39.62\%) and 1,925 datasets (28.48\%) have documentation. Furthermore, we observe inconsistency in ethics and transparency-related documentation for ML models and datasets.