{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2018:AH5S4WE2T4KISLKPOBTLFWQMX3","short_pith_number":"pith:AH5S4WE2","schema_version":"1.0","canonical_sha256":"01fb2e589a9f14892d4f7066b2da0cbedbddaa333d2b71aa702d1be61b7854be","source":{"kind":"arxiv","id":"1811.06992","version":2},"attestation_state":"computed","paper":{"title":"Image Classification at Supercomputer Scale","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.DC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chris Ying, Dehao Chen, Sameer Kumar, Tao Wang, Youlong Cheng","submitted_at":"2018-11-16T19:01:40Z","abstract_excerpt":"Deep learning is extremely computationally intensive, and hardware vendors have responded by building faster accelerators in large clusters. Training deep learning models at petaFLOPS scale requires overcoming both algorithmic and systems software challenges. In this paper, we discuss three systems-related optimizations: (1) distributed batch normalization to control per-replica batch sizes, (2) input pipeline optimizations to sustain model throughput, and (3) 2-D torus all-reduce to speed up gradient summation. We combine these optimizations to train ResNet-50 on ImageNet to 76.3% accuracy in"},"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":"1811.06992","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2018-11-16T19:01:40Z","cross_cats_sorted":["cs.DC","stat.ML"],"title_canon_sha256":"c879f910810daf3a6556b22a8dc037f789043349e85fd5208e01139c26e12f16","abstract_canon_sha256":"4dc7164dc5cc885869b040779ccc8c1e3d21fba4c5a9aed25473a7831f9a8cd1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:59:22.045615Z","signature_b64":"aQx+1FpcOLeaBVeDQyHVdhr3QlnU2mU6krroFSgzzExy2EcnpdPSXXpeSGISNlT99k2CMwn/asFHwsz+tbwQDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"01fb2e589a9f14892d4f7066b2da0cbedbddaa333d2b71aa702d1be61b7854be","last_reissued_at":"2026-05-17T23:59:22.045226Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:59:22.045226Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Image Classification at Supercomputer Scale","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.DC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chris Ying, Dehao Chen, Sameer Kumar, Tao Wang, Youlong Cheng","submitted_at":"2018-11-16T19:01:40Z","abstract_excerpt":"Deep learning is extremely computationally intensive, and hardware vendors have responded by building faster accelerators in large clusters. Training deep learning models at petaFLOPS scale requires overcoming both algorithmic and systems software challenges. In this paper, we discuss three systems-related optimizations: (1) distributed batch normalization to control per-replica batch sizes, (2) input pipeline optimizations to sustain model throughput, and (3) 2-D torus all-reduce to speed up gradient summation. We combine these optimizations to train ResNet-50 on ImageNet to 76.3% accuracy in"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1811.06992","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":"1811.06992","created_at":"2026-05-17T23:59:22.045290+00:00"},{"alias_kind":"arxiv_version","alias_value":"1811.06992v2","created_at":"2026-05-17T23:59:22.045290+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1811.06992","created_at":"2026-05-17T23:59:22.045290+00:00"},{"alias_kind":"pith_short_12","alias_value":"AH5S4WE2T4KI","created_at":"2026-05-18T12:32:13.499390+00:00"},{"alias_kind":"pith_short_16","alias_value":"AH5S4WE2T4KISLKP","created_at":"2026-05-18T12:32:13.499390+00:00"},{"alias_kind":"pith_short_8","alias_value":"AH5S4WE2","created_at":"2026-05-18T12:32:13.499390+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"1904.00962","citing_title":"Large Batch Optimization for Deep Learning: Training BERT in 76 minutes","ref_index":17,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3","json":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3.json","graph_json":"https://pith.science/api/pith-number/AH5S4WE2T4KISLKPOBTLFWQMX3/graph.json","events_json":"https://pith.science/api/pith-number/AH5S4WE2T4KISLKPOBTLFWQMX3/events.json","paper":"https://pith.science/paper/AH5S4WE2"},"agent_actions":{"view_html":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3","download_json":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3.json","view_paper":"https://pith.science/paper/AH5S4WE2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1811.06992&json=true","fetch_graph":"https://pith.science/api/pith-number/AH5S4WE2T4KISLKPOBTLFWQMX3/graph.json","fetch_events":"https://pith.science/api/pith-number/AH5S4WE2T4KISLKPOBTLFWQMX3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3/action/storage_attestation","attest_author":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3/action/author_attestation","sign_citation":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3/action/citation_signature","submit_replication":"https://pith.science/pith/AH5S4WE2T4KISLKPOBTLFWQMX3/action/replication_record"}},"created_at":"2026-05-17T23:59:22.045290+00:00","updated_at":"2026-05-17T23:59:22.045290+00:00"}