{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:4NTHSFGUBZHTXAA3CAQLNQSA55","short_pith_number":"pith:4NTHSFGU","schema_version":"1.0","canonical_sha256":"e3667914d40e4f3b801b1020b6c240ef57d196dbec523a23137983debfd88e7c","source":{"kind":"arxiv","id":"1904.11943","version":2},"attestation_state":"computed","paper":{"title":"SWALP : Stochastic Weight Averaging in Low-Precision Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Andrew Gordon Wilson, Christopher De Sa, Guandao Yang, Junwen Bai, Polina Kirichenko, Tianyi Zhang","submitted_at":"2019-04-26T17:22:06Z","abstract_excerpt":"Low precision operations can provide scalability, memory savings, portability, and energy efficiency. This paper proposes SWALP, an approach to low precision training that averages low-precision SGD iterates with a modified learning rate schedule. SWALP is easy to implement and can match the performance of full-precision SGD even with all numbers quantized down to 8 bits, including the gradient accumulators. Additionally, we show that SWALP converges arbitrarily close to the optimal solution for quadratic objectives, and to a noise ball asymptotically smaller than low precision SGD in strongly"},"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":"1904.11943","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-04-26T17:22:06Z","cross_cats_sorted":["cs.AI","stat.ML"],"title_canon_sha256":"bc419e328c5e7a2094be434afb0f966453a1f5959caa654111491dd2736fff7a","abstract_canon_sha256":"8465fec6fdcdaca7663eb5a775d57a79ab8f2de381fac7ed06227dbf8f72d16c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:45:51.697805Z","signature_b64":"zgVcKU5aYgpNzOCaCADRskSSjqgdW2yYwfMpAVaK2qujald5fcIQJw+Jq1DjoNeVfow9WQ4GfwzN8fh8GbwuAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e3667914d40e4f3b801b1020b6c240ef57d196dbec523a23137983debfd88e7c","last_reissued_at":"2026-05-17T23:45:51.697428Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:45:51.697428Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SWALP : Stochastic Weight Averaging in Low-Precision Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","stat.ML"],"primary_cat":"cs.LG","authors_text":"Andrew Gordon Wilson, Christopher De Sa, Guandao Yang, Junwen Bai, Polina Kirichenko, Tianyi Zhang","submitted_at":"2019-04-26T17:22:06Z","abstract_excerpt":"Low precision operations can provide scalability, memory savings, portability, and energy efficiency. This paper proposes SWALP, an approach to low precision training that averages low-precision SGD iterates with a modified learning rate schedule. SWALP is easy to implement and can match the performance of full-precision SGD even with all numbers quantized down to 8 bits, including the gradient accumulators. Additionally, we show that SWALP converges arbitrarily close to the optimal solution for quadratic objectives, and to a noise ball asymptotically smaller than low precision SGD in strongly"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1904.11943","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":""},"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":"1904.11943","created_at":"2026-05-17T23:45:51.697482+00:00"},{"alias_kind":"arxiv_version","alias_value":"1904.11943v2","created_at":"2026-05-17T23:45:51.697482+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1904.11943","created_at":"2026-05-17T23:45:51.697482+00:00"},{"alias_kind":"pith_short_12","alias_value":"4NTHSFGUBZHT","created_at":"2026-05-18T12:33:10.108867+00:00"},{"alias_kind":"pith_short_16","alias_value":"4NTHSFGUBZHTXAA3","created_at":"2026-05-18T12:33:10.108867+00:00"},{"alias_kind":"pith_short_8","alias_value":"4NTHSFGU","created_at":"2026-05-18T12:33:10.108867+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/4NTHSFGUBZHTXAA3CAQLNQSA55","json":"https://pith.science/pith/4NTHSFGUBZHTXAA3CAQLNQSA55.json","graph_json":"https://pith.science/api/pith-number/4NTHSFGUBZHTXAA3CAQLNQSA55/graph.json","events_json":"https://pith.science/api/pith-number/4NTHSFGUBZHTXAA3CAQLNQSA55/events.json","paper":"https://pith.science/paper/4NTHSFGU"},"agent_actions":{"view_html":"https://pith.science/pith/4NTHSFGUBZHTXAA3CAQLNQSA55","download_json":"https://pith.science/pith/4NTHSFGUBZHTXAA3CAQLNQSA55.json","view_paper":"https://pith.science/paper/4NTHSFGU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1904.11943&json=true","fetch_graph":"https://pith.science/api/pith-number/4NTHSFGUBZHTXAA3CAQLNQSA55/graph.json","fetch_events":"https://pith.science/api/pith-number/4NTHSFGUBZHTXAA3CAQLNQSA55/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/4NTHSFGUBZHTXAA3CAQLNQSA55/action/timestamp_anchor","attest_storage":"https://pith.science/pith/4NTHSFGUBZHTXAA3CAQLNQSA55/action/storage_attestation","attest_author":"https://pith.science/pith/4NTHSFGUBZHTXAA3CAQLNQSA55/action/author_attestation","sign_citation":"https://pith.science/pith/4NTHSFGUBZHTXAA3CAQLNQSA55/action/citation_signature","submit_replication":"https://pith.science/pith/4NTHSFGUBZHTXAA3CAQLNQSA55/action/replication_record"}},"created_at":"2026-05-17T23:45:51.697482+00:00","updated_at":"2026-05-17T23:45:51.697482+00:00"}