{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:EU4543U6F2MZNYM25AHRDP3W3O","short_pith_number":"pith:EU4543U6","schema_version":"1.0","canonical_sha256":"2539de6e9e2e9996e19ae80f11bf76dba26b6fc1dbba839de926088f225433a3","source":{"kind":"arxiv","id":"2106.04759","version":2},"attestation_state":"computed","paper":{"title":"Communication-efficient SGD: From Local SGD to One-Shot Averaging","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.OC"],"primary_cat":"cs.DC","authors_text":"Alex Olshevsky, Artin Spiridonoff, Ioannis Ch. Paschalidis","submitted_at":"2021-06-09T01:10:34Z","abstract_excerpt":"We consider speeding up stochastic gradient descent (SGD) by parallelizing it across multiple workers. We assume the same data set is shared among $N$ workers, who can take SGD steps and coordinate with a central server. While it is possible to obtain a linear reduction in the variance by averaging all the stochastic gradients at every step, this requires a lot of communication between the workers and the server, which can dramatically reduce the gains from parallelism. The Local SGD method, proposed and analyzed in the earlier literature, suggests machines should make many local steps between"},"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":"2106.04759","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2021-06-09T01:10:34Z","cross_cats_sorted":["cs.LG","math.OC"],"title_canon_sha256":"041e747c07d788a5d6270c53b6cf666f6620d745c17e105e683326126ae663cd","abstract_canon_sha256":"8b61a7782b131b4fe63450c6dd68cf128499ff582f531e4015e757d600253b73"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:26:17.870742Z","signature_b64":"cH3EEXBKB1R8qdzG9vopPSGG2TQ45D35IpIIDeylI7ujXZojGTUcKYDojVoaep+bxn78FXjyhjPdiim7yCFVDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2539de6e9e2e9996e19ae80f11bf76dba26b6fc1dbba839de926088f225433a3","last_reissued_at":"2026-07-05T03:26:17.870184Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:26:17.870184Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Communication-efficient SGD: From Local SGD to One-Shot Averaging","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","math.OC"],"primary_cat":"cs.DC","authors_text":"Alex Olshevsky, Artin Spiridonoff, Ioannis Ch. Paschalidis","submitted_at":"2021-06-09T01:10:34Z","abstract_excerpt":"We consider speeding up stochastic gradient descent (SGD) by parallelizing it across multiple workers. We assume the same data set is shared among $N$ workers, who can take SGD steps and coordinate with a central server. While it is possible to obtain a linear reduction in the variance by averaging all the stochastic gradients at every step, this requires a lot of communication between the workers and the server, which can dramatically reduce the gains from parallelism. The Local SGD method, proposed and analyzed in the earlier literature, suggests machines should make many local steps between"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2106.04759","kind":"arxiv","version":2},"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/2106.04759/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":"2106.04759","created_at":"2026-07-05T03:26:17.870245+00:00"},{"alias_kind":"arxiv_version","alias_value":"2106.04759v2","created_at":"2026-07-05T03:26:17.870245+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2106.04759","created_at":"2026-07-05T03:26:17.870245+00:00"},{"alias_kind":"pith_short_12","alias_value":"EU4543U6F2MZ","created_at":"2026-07-05T03:26:17.870245+00:00"},{"alias_kind":"pith_short_16","alias_value":"EU4543U6F2MZNYM2","created_at":"2026-07-05T03:26:17.870245+00:00"},{"alias_kind":"pith_short_8","alias_value":"EU4543U6","created_at":"2026-07-05T03:26:17.870245+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O","json":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O.json","graph_json":"https://pith.science/api/pith-number/EU4543U6F2MZNYM25AHRDP3W3O/graph.json","events_json":"https://pith.science/api/pith-number/EU4543U6F2MZNYM25AHRDP3W3O/events.json","paper":"https://pith.science/paper/EU4543U6"},"agent_actions":{"view_html":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O","download_json":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O.json","view_paper":"https://pith.science/paper/EU4543U6","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2106.04759&json=true","fetch_graph":"https://pith.science/api/pith-number/EU4543U6F2MZNYM25AHRDP3W3O/graph.json","fetch_events":"https://pith.science/api/pith-number/EU4543U6F2MZNYM25AHRDP3W3O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O/action/storage_attestation","attest_author":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O/action/author_attestation","sign_citation":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O/action/citation_signature","submit_replication":"https://pith.science/pith/EU4543U6F2MZNYM25AHRDP3W3O/action/replication_record"}},"created_at":"2026-07-05T03:26:17.870245+00:00","updated_at":"2026-07-05T03:26:17.870245+00:00"}