{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2017:J2LUZ564FUOB74RUDD6Q7Q3WSE","short_pith_number":"pith:J2LUZ564","schema_version":"1.0","canonical_sha256":"4e974cf7dc2d1c1ff23418fd0fc3769117ee8ca36cd3f9d9c35c3d6506abae22","source":{"kind":"arxiv","id":"1712.04432","version":4},"attestation_state":"computed","paper":{"title":"Integrated Model, Batch and Domain Parallelism in Training Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Amir Gholami, Ariful Azad, Aydin Buluc, Kurt Keutzer, Peter Jin","submitted_at":"2017-12-12T18:42:07Z","abstract_excerpt":"We propose a new integrated method of exploiting model, batch and domain parallelism for the training of deep neural networks (DNNs) on large distributed-memory computers using minibatch stochastic gradient descent (SGD). Our goal is to find an efficient parallelization strategy for a fixed batch size using $P$ processes. Our method is inspired by the communication-avoiding algorithms in numerical linear algebra. We see $P$ processes as logically divided into a $P_r \\times P_c$ grid where the $P_r$ dimension is implicitly responsible for model/domain parallelism and the $P_c$ dimension is impl"},"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":"1712.04432","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2017-12-12T18:42:07Z","cross_cats_sorted":["stat.ML"],"title_canon_sha256":"b8baf00ad4866c3851c7191b5f08cbd95b50df7aaca2b26c630040a1594676ae","abstract_canon_sha256":"9b75ed351ac0dd0d3a76c433d64caab5c6b6bdc12b3a0eb714049ac559a45bb1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T00:15:52.143447Z","signature_b64":"D9S7vv5P0duu3vbMG6t891YL5iTmEepvU8D/mqVjiHU1vWn1ur+44KWBWcvt/M+wCMFZGBr6LrQdiOPktl3tCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4e974cf7dc2d1c1ff23418fd0fc3769117ee8ca36cd3f9d9c35c3d6506abae22","last_reissued_at":"2026-05-18T00:15:52.142853Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T00:15:52.142853Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Integrated Model, Batch and Domain Parallelism in Training Neural Networks","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["stat.ML"],"primary_cat":"cs.LG","authors_text":"Amir Gholami, Ariful Azad, Aydin Buluc, Kurt Keutzer, Peter Jin","submitted_at":"2017-12-12T18:42:07Z","abstract_excerpt":"We propose a new integrated method of exploiting model, batch and domain parallelism for the training of deep neural networks (DNNs) on large distributed-memory computers using minibatch stochastic gradient descent (SGD). Our goal is to find an efficient parallelization strategy for a fixed batch size using $P$ processes. Our method is inspired by the communication-avoiding algorithms in numerical linear algebra. We see $P$ processes as logically divided into a $P_r \\times P_c$ grid where the $P_r$ dimension is implicitly responsible for model/domain parallelism and the $P_c$ dimension is impl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1712.04432","kind":"arxiv","version":4},"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":"1712.04432","created_at":"2026-05-18T00:15:52.142946+00:00"},{"alias_kind":"arxiv_version","alias_value":"1712.04432v4","created_at":"2026-05-18T00:15:52.142946+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1712.04432","created_at":"2026-05-18T00:15:52.142946+00:00"},{"alias_kind":"pith_short_12","alias_value":"J2LUZ564FUOB","created_at":"2026-05-18T12:31:21.493067+00:00"},{"alias_kind":"pith_short_16","alias_value":"J2LUZ564FUOB74RU","created_at":"2026-05-18T12:31:21.493067+00:00"},{"alias_kind":"pith_short_8","alias_value":"J2LUZ564","created_at":"2026-05-18T12:31:21.493067+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.02473","citing_title":"Analyzing Reverse Address Translation Overheads in Multi-GPU Scale-Up Pods","ref_index":41,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE","json":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE.json","graph_json":"https://pith.science/api/pith-number/J2LUZ564FUOB74RUDD6Q7Q3WSE/graph.json","events_json":"https://pith.science/api/pith-number/J2LUZ564FUOB74RUDD6Q7Q3WSE/events.json","paper":"https://pith.science/paper/J2LUZ564"},"agent_actions":{"view_html":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE","download_json":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE.json","view_paper":"https://pith.science/paper/J2LUZ564","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1712.04432&json=true","fetch_graph":"https://pith.science/api/pith-number/J2LUZ564FUOB74RUDD6Q7Q3WSE/graph.json","fetch_events":"https://pith.science/api/pith-number/J2LUZ564FUOB74RUDD6Q7Q3WSE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE/action/storage_attestation","attest_author":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE/action/author_attestation","sign_citation":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE/action/citation_signature","submit_replication":"https://pith.science/pith/J2LUZ564FUOB74RUDD6Q7Q3WSE/action/replication_record"}},"created_at":"2026-05-18T00:15:52.142946+00:00","updated_at":"2026-05-18T00:15:52.142946+00:00"}