{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/pre-training-differentially-private-models","title":"Pre-training Differentially Private Models with Limited Public Data","arxiv_id":"2402.18752","date":"2024-02-28","proceeding":null,"authors":["Zhiqi Bu","Xinwei Zhang","Mingyi Hong","Sheng Zha","George Karypis"],"abstract":"The superior performance of large foundation models relies on the use of massive amounts of high-quality data, which often contain sensitive, private and copyrighted material that requires formal protection. While differential privacy (DP) is a prominent method to gauge the degree of security provided to the models, its application is commonly limited to the model fine-tuning stage, due to the performance degradation when applying DP during the pre-training stage. Consequently, DP is yet not capable of protecting a substantial portion of the data used during the initial pre-training process. In this work, we first provide a theoretical understanding of the efficacy of DP training by analyzing the per-iteration loss improvement. We make a key observation that DP optimizers' performance degradation can be significantly mitigated by the use of limited public data, which leads to a novel DP continual pre-training strategy. Empirically, using only 10\\% of public data, our strategy can achieve DP accuracy of 41.5\\% on ImageNet-21k (with $\\epsilon=8$), as well as non-DP accuracy of 55.7\\% and and 60.0\\% on downstream tasks Places365 and iNaturalist-2021, respectively, on par with state-of-the-art standard pre-training and substantially outperforming existing DP pre-trained models. Our DP pre-trained models are released in fastDP library (https://github.com/awslabs/fast-differential-privacy/releases/tag/v2.1)","url_abs":"https://arxiv.org/abs/2402.18752v2","url_pdf":"https://arxiv.org/pdf/2402.18752v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"pre-training-differentially-private-models","repo_url":"https://github.com/awslabs/fast-differential-privacy","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"pre-training-differentially-private-models","repo_url":"https://github.com/awslabs/fast-differential-privacy/tree/main/fastDP","is_official":1,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"tag","task_name":"TAG"}],"methods":[{"method_slug":null,"method_name":"Library"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2402.18752","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2402.18752"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/awslabs/fast-differential-privacy/tree/main/fastDP","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/awslabs/fast-differential-privacy","reach":null}],"summary":{"unverified":2},"by_repo_kind":{"official":{"samples":1,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":1,"samples":[{"code_sha256_prefix":"1fbb306860e852f7","entry":"main","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"1fbb306860e852f7"}},{"code_sha256_prefix":"cd53567f8d66c13e","entry":"main","repo":"awslabs/fast-differential-privacy","repo_kind":"official","path":"examples/image_classification/CIFAR_TIMM.py","file_url":"https://github.com/awslabs/fast-differential-privacy/blob/HEAD/examples/image_classification/CIFAR_TIMM.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"cd53567f8d66c13e"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}