{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:7MB2SIZ7R54TSCUA24OKJEWGDF","short_pith_number":"pith:7MB2SIZ7","schema_version":"1.0","canonical_sha256":"fb03a9233f8f79390a80d71ca492c61942b2062b6edb04a76a1869fd7e745aff","source":{"kind":"arxiv","id":"2011.02828","version":1},"attestation_state":"computed","paper":{"title":"Local SGD: Unified Theory and New Efficient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Eduard Gorbunov, Filip Hanzely, Peter Richt\\'arik","submitted_at":"2020-11-03T13:02:50Z","abstract_excerpt":"We present a unified framework for analyzing local SGD methods in the convex and strongly convex regimes for distributed/federated training of supervised machine learning models. We recover several known methods as a special case of our general framework, including Local-SGD/FedAvg, SCAFFOLD, and several variants of SGD not originally designed for federated learning. Our framework covers both the identical and heterogeneous data settings, supports both random and deterministic number of local steps, and can work with a wide array of local stochastic gradient estimators, including shifted estim"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2011.02828","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-11-03T13:02:50Z","cross_cats_sorted":[],"title_canon_sha256":"ba1a33c108e5e12ac54e22583900553c045af8315dfa9db4faf71c35daa1cacd","abstract_canon_sha256":"76ae2bc85e8b9c16fdad7deb16e87161d7ac906e82b736e52708da19b69d94b7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:49:26.157727Z","signature_b64":"xYbh3cwh/HXB6c+3Jm104vPtZAtinTkU8Ndll+E4Y5TFrMnEDO7tenGgfMNFZw5GsMUgICakZ4vYHLMBmKWJCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fb03a9233f8f79390a80d71ca492c61942b2062b6edb04a76a1869fd7e745aff","last_reissued_at":"2026-07-05T01:49:26.157297Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:49:26.157297Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Local SGD: Unified Theory and New Efficient Methods","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Eduard Gorbunov, Filip Hanzely, Peter Richt\\'arik","submitted_at":"2020-11-03T13:02:50Z","abstract_excerpt":"We present a unified framework for analyzing local SGD methods in the convex and strongly convex regimes for distributed/federated training of supervised machine learning models. We recover several known methods as a special case of our general framework, including Local-SGD/FedAvg, SCAFFOLD, and several variants of SGD not originally designed for federated learning. Our framework covers both the identical and heterogeneous data settings, supports both random and deterministic number of local steps, and can work with a wide array of local stochastic gradient estimators, including shifted estim"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2011.02828","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2011.02828/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2011.02828","created_at":"2026-07-05T01:49:26.157371+00:00"},{"alias_kind":"arxiv_version","alias_value":"2011.02828v1","created_at":"2026-07-05T01:49:26.157371+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2011.02828","created_at":"2026-07-05T01:49:26.157371+00:00"},{"alias_kind":"pith_short_12","alias_value":"7MB2SIZ7R54T","created_at":"2026-07-05T01:49:26.157371+00:00"},{"alias_kind":"pith_short_16","alias_value":"7MB2SIZ7R54TSCUA","created_at":"2026-07-05T01:49:26.157371+00:00"},{"alias_kind":"pith_short_8","alias_value":"7MB2SIZ7","created_at":"2026-07-05T01:49:26.157371+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.13434","citing_title":"Rescaled Asynchronous SGD: Optimal Distributed Optimization under Data and System Heterogeneity","ref_index":49,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF","json":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF.json","graph_json":"https://pith.science/api/pith-number/7MB2SIZ7R54TSCUA24OKJEWGDF/graph.json","events_json":"https://pith.science/api/pith-number/7MB2SIZ7R54TSCUA24OKJEWGDF/events.json","paper":"https://pith.science/paper/7MB2SIZ7"},"agent_actions":{"view_html":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF","download_json":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF.json","view_paper":"https://pith.science/paper/7MB2SIZ7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2011.02828&json=true","fetch_graph":"https://pith.science/api/pith-number/7MB2SIZ7R54TSCUA24OKJEWGDF/graph.json","fetch_events":"https://pith.science/api/pith-number/7MB2SIZ7R54TSCUA24OKJEWGDF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF/action/storage_attestation","attest_author":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF/action/author_attestation","sign_citation":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF/action/citation_signature","submit_replication":"https://pith.science/pith/7MB2SIZ7R54TSCUA24OKJEWGDF/action/replication_record"}},"created_at":"2026-07-05T01:49:26.157371+00:00","updated_at":"2026-07-05T01:49:26.157371+00:00"}