{"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/secure-distributed-training-at-scale","title":"Secure Distributed Training at Scale","arxiv_id":"2106.11257","date":"2021-06-21","proceeding":null,"authors":["Eduard Gorbunov","Alexander Borzunov","Michael Diskin","Max Ryabinin"],"abstract":"Many areas of deep learning benefit from using increasingly larger neural networks trained on public data, as is the case for pre-trained models for NLP and computer vision. Training such models requires a lot of computational resources (e.g., HPC clusters) that are not available to small research groups and independent researchers. One way to address it is for several smaller groups to pool their computational resources together and train a model that benefits all participants. Unfortunately, in this case, any participant can jeopardize the entire training run by sending incorrect updates, deliberately or by mistake. Training in presence of such peers requires specialized distributed training algorithms with Byzantine tolerance. These algorithms often sacrifice efficiency by introducing redundant communication or passing all updates through a trusted server, making it infeasible to apply them to large-scale deep learning, where models can have billions of parameters. In this work, we propose a novel protocol for secure (Byzantine-tolerant) decentralized training that emphasizes communication efficiency.","url_abs":"https://arxiv.org/abs/2106.11257v4","url_pdf":"https://arxiv.org/pdf/2106.11257v4.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":"secure-distributed-training-at-scale","repo_url":"https://github.com/yandex-research/btard","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"secure-distributed-training-at-scale","repo_url":"https://github.com/iclr-paper/btard","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok"}},{"paper_slug":"secure-distributed-training-at-scale","repo_url":"https://github.com/learning-at-home/hivemind","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"distributed-optimization","task_name":"Distributed Optimization"},{"task_slug":"image-classification","task_name":"Image Classification"},{"task_slug":"language-modelling","task_name":"Language Modelling"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2106.11257","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2106.11257"}},"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/learning-at-home/hivemind","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/iclr-paper/btard","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/yandex-research/btard","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"unverified":9},"by_repo_kind":{"official":{"samples":9,"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":0,"samples":[{"code_sha256_prefix":"92df667d07531d60","entry":"centered_clip","repo":"yandex-research/btard","repo_kind":"official","path":"resnet/centered_clip.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/resnet/centered_clip.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":"92df667d07531d60"}},{"code_sha256_prefix":"b6beffa06be72594","entry":"create_instances_from_document","repo":"yandex-research/btard","repo_kind":"official","path":"albert/experiments/tokenize_wikitext103.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/albert/experiments/tokenize_wikitext103.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":"b6beffa06be72594"}},{"code_sha256_prefix":"a904de8f2aa42d6f","entry":"evaluate_accuracy","repo":"yandex-research/btard","repo_kind":"official","path":"resnet/training_utils.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/resnet/training_utils.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":"a904de8f2aa42d6f"}},{"code_sha256_prefix":"1960851ee3b8f1ba","entry":"get_model","repo":"yandex-research/btard","repo_kind":"official","path":"albert/experiments/run_trainer.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/albert/experiments/run_trainer.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":"1960851ee3b8f1ba"}},{"code_sha256_prefix":"96f98edcffa1a140","entry":"restore_from_parts","repo":"yandex-research/btard","repo_kind":"official","path":"resnet/centered_clip.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/resnet/centered_clip.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":"96f98edcffa1a140"}},{"code_sha256_prefix":"df3533d0b1a9d882","entry":"sample_tensors","repo":"yandex-research/btard","repo_kind":"official","path":"albert/hivemind/benchmarks/benchmark_averaging.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/albert/hivemind/benchmarks/benchmark_averaging.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":"df3533d0b1a9d882"}},{"code_sha256_prefix":"52afa7d9072a594d","entry":"split_into_parts","repo":"yandex-research/btard","repo_kind":"official","path":"resnet/centered_clip.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/resnet/centered_clip.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":"52afa7d9072a594d"}},{"code_sha256_prefix":"d2b88e48b87e9450","entry":"tokenize_function","repo":"yandex-research/btard","repo_kind":"official","path":"albert/experiments/tokenize_wikitext103.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/albert/experiments/tokenize_wikitext103.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":"d2b88e48b87e9450"}},{"code_sha256_prefix":"74f2dafa5b436318","entry":"verify_equal_parameters","repo":"yandex-research/btard","repo_kind":"official","path":"resnet/training_utils.py","file_url":"https://github.com/yandex-research/btard/blob/HEAD/resnet/training_utils.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":"74f2dafa5b436318"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}