DataPerf: Benchmarks for Data-Centric AI Development

• Main Authors: Mazumder, Mark, Banbury, Colby, Yao, Xiaozhe, Karlaš, Bojan, Rojas, William Gaviria, Diamos, Sudnya, Diamos, Greg, He, Lynn, Parrish, Alicia, Kirk, Hannah Rose, Quaye, Jessica, Rastogi, Charvi, Kiela, Douwe, Jurado, David, Kanter, David, Mosquera, Rafael, Ciro, Juan, Aroyo, Lora, Acun, Bilge, Chen, Lingjiao, Raje, Mehul Smriti, Bartolo, Max, Eyuboglu, Sabri, Ghorbani, Amirata, Goodman, Emmett, Inel, Oana, Kane, Tariq, Kirkpatrick, Christine R, Kuo, Tzu-Sheng, Mueller, Jonas, Thrush, Tristan, Vanschoren, Joaquin, Warren, Margaret, Williams, Adina, Yeung, Serena, Ardalani, Newsha, Paritosh, Praveen, Bat-Leah, Lilith, Zhang, Ce, Zou, James, Wu, Carole-Jean, Coleman, Cody, Ng, Andrew, Mattson, Peter, Reddi, Vijay Janapa
• Format: Journal Article
• Language: English
• Published: 20.07.2022
• Subjects: Computer Science - Learning
• DOI: 10.48550/arxiv.2207.10062

Machine learning research has long focused on models rather than datasets, and prominent datasets are used for common ML tasks without regard to the breadth, difficulty, and faithfulness of the underlying problems. Neglecting the fundamental importance of data has given rise to inaccuracy, bias, and fragility in real-world applications, and research is hindered by saturation across existing dataset benchmarks. In response, we present DataPerf, a community-led benchmark suite for evaluating ML datasets and data-centric algorithms. We aim to foster innovation in data-centric AI through competition, comparability, and reproducibility. We enable the ML community to iterate on datasets, instead of just architectures, and we provide an open, online platform with multiple rounds of challenges to support this iterative development. The first iteration of DataPerf contains five benchmarks covering a wide spectrum of data-centric techniques, tasks, and modalities in vision, speech, acquisition, debugging, and diffusion prompting, and we support hosting new contributed benchmarks from the community. The benchmarks, online evaluation platform, and baseline implementations are open source, and the MLCommons Association will maintain DataPerf to ensure long-term benefits to academia and industry.