{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:5EWHP2DI3MQ673UTH6FAWYKSUA","short_pith_number":"pith:5EWHP2DI","schema_version":"1.0","canonical_sha256":"e92c77e868db21efee933f8a0b6152a0394dad0f08fb81a2d7f878bdc2b9e2e2","source":{"kind":"arxiv","id":"1907.09356","version":3},"attestation_state":"computed","paper":{"title":"Decentralized Deep Learning with Arbitrary Communication Compression","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC","cs.DS","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Anastasia Koloskova, Martin Jaggi, Sebastian U. Stich, Tao Lin","submitted_at":"2019-07-22T14:53:02Z","abstract_excerpt":"Decentralized training of deep learning models is a key element for enabling data privacy and on-device learning over networks, as well as for efficient scaling to large compute clusters. As current approaches suffer from limited bandwidth of the network, we propose the use of communication compression in the decentralized training context. We show that Choco-SGD $-$ recently introduced and analyzed for strongly-convex objectives only $-$ converges under arbitrary high compression ratio on general non-convex functions at the rate $O\\bigl(1/\\sqrt{nT}\\bigr)$ where $T$ denotes the number of itera"},"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":"1907.09356","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-07-22T14:53:02Z","cross_cats_sorted":["cs.DC","cs.DS","math.OC","stat.ML"],"title_canon_sha256":"c7fa254731ea332d4967287af026e66c34a876c3c07ce65ac0c4ecffc0202bba","abstract_canon_sha256":"434a946d50fa0e7c829e341240ccafed7beb8ea084db90dc6611922476d0991f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T01:50:57.029172Z","signature_b64":"fFbdr7WPPFxEdjW3omgGUX+nxuJXbGehNQ+W5E+ThWRO/lWpqd+zYU41YLviNx6cIBcbjyMcmChAyKF/Vjx5CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e92c77e868db21efee933f8a0b6152a0394dad0f08fb81a2d7f878bdc2b9e2e2","last_reissued_at":"2026-07-05T01:50:57.028744Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T01:50:57.028744Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Decentralized Deep Learning with Arbitrary Communication Compression","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC","cs.DS","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Anastasia Koloskova, Martin Jaggi, Sebastian U. Stich, Tao Lin","submitted_at":"2019-07-22T14:53:02Z","abstract_excerpt":"Decentralized training of deep learning models is a key element for enabling data privacy and on-device learning over networks, as well as for efficient scaling to large compute clusters. As current approaches suffer from limited bandwidth of the network, we propose the use of communication compression in the decentralized training context. We show that Choco-SGD $-$ recently introduced and analyzed for strongly-convex objectives only $-$ converges under arbitrary high compression ratio on general non-convex functions at the rate $O\\bigl(1/\\sqrt{nT}\\bigr)$ where $T$ denotes the number of itera"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1907.09356","kind":"arxiv","version":3},"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/1907.09356/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":"1907.09356","created_at":"2026-07-05T01:50:57.028798+00:00"},{"alias_kind":"arxiv_version","alias_value":"1907.09356v3","created_at":"2026-07-05T01:50:57.028798+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1907.09356","created_at":"2026-07-05T01:50:57.028798+00:00"},{"alias_kind":"pith_short_12","alias_value":"5EWHP2DI3MQ6","created_at":"2026-07-05T01:50:57.028798+00:00"},{"alias_kind":"pith_short_16","alias_value":"5EWHP2DI3MQ673UT","created_at":"2026-07-05T01:50:57.028798+00:00"},{"alias_kind":"pith_short_8","alias_value":"5EWHP2DI","created_at":"2026-07-05T01:50:57.028798+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2607.01755","citing_title":"Decentralized Stochastic Subgradient-type Methods with Communication Compression for Nonsmooth Nonconvex Optimization","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2606.11339","citing_title":"Quantized Stochastic Primal-Dual Methods for Distributed Optimization under Relaxed Global Geometry","ref_index":9,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA","json":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA.json","graph_json":"https://pith.science/api/pith-number/5EWHP2DI3MQ673UTH6FAWYKSUA/graph.json","events_json":"https://pith.science/api/pith-number/5EWHP2DI3MQ673UTH6FAWYKSUA/events.json","paper":"https://pith.science/paper/5EWHP2DI"},"agent_actions":{"view_html":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA","download_json":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA.json","view_paper":"https://pith.science/paper/5EWHP2DI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1907.09356&json=true","fetch_graph":"https://pith.science/api/pith-number/5EWHP2DI3MQ673UTH6FAWYKSUA/graph.json","fetch_events":"https://pith.science/api/pith-number/5EWHP2DI3MQ673UTH6FAWYKSUA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA/action/storage_attestation","attest_author":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA/action/author_attestation","sign_citation":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA/action/citation_signature","submit_replication":"https://pith.science/pith/5EWHP2DI3MQ673UTH6FAWYKSUA/action/replication_record"}},"created_at":"2026-07-05T01:50:57.028798+00:00","updated_at":"2026-07-05T01:50:57.028798+00:00"}