{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2018:6VARSBRX6E3HDBTBZP3WZOXG6B","short_pith_number":"pith:6VARSBRX","canonical_record":{"source":{"id":"1811.05233","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-11-13T11:52:04Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"d6cd81252e89069e2d685a8db0e14d63ceaabc34f2ea6bef568baa2887ba52e9","abstract_canon_sha256":"64a9eeb5c319b65bc3c68598ef29104918cd63cfecf9a846b0e74e3fb43d040f"},"schema_version":"1.0"},"canonical_sha256":"f541190637f136718661cbf76cbae6f0739e46cf4ee80f156d42ce5b21a0d6c1","source":{"kind":"arxiv","id":"1811.05233","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1811.05233","created_at":"2026-05-17T23:52:04Z"},{"alias_kind":"arxiv_version","alias_value":"1811.05233v2","created_at":"2026-05-17T23:52:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1811.05233","created_at":"2026-05-17T23:52:04Z"},{"alias_kind":"pith_short_12","alias_value":"6VARSBRX6E3H","created_at":"2026-05-18T12:32:11Z"},{"alias_kind":"pith_short_16","alias_value":"6VARSBRX6E3HDBTB","created_at":"2026-05-18T12:32:11Z"},{"alias_kind":"pith_short_8","alias_value":"6VARSBRX","created_at":"2026-05-18T12:32:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2018:6VARSBRX6E3HDBTBZP3WZOXG6B","target":"record","payload":{"canonical_record":{"source":{"id":"1811.05233","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-11-13T11:52:04Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"d6cd81252e89069e2d685a8db0e14d63ceaabc34f2ea6bef568baa2887ba52e9","abstract_canon_sha256":"64a9eeb5c319b65bc3c68598ef29104918cd63cfecf9a846b0e74e3fb43d040f"},"schema_version":"1.0"},"canonical_sha256":"f541190637f136718661cbf76cbae6f0739e46cf4ee80f156d42ce5b21a0d6c1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:52:04.414826Z","signature_b64":"qVacuSf+RVHKpILiCUhTrivj8gHRe+kvKEhIUqCjrP17nIyjT+tp6SDaft9zouuDg08rqHx2OQZDLUPgpd41Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f541190637f136718661cbf76cbae6f0739e46cf4ee80f156d42ce5b21a0d6c1","last_reissued_at":"2026-05-17T23:52:04.414293Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:52:04.414293Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1811.05233","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:52:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"4UlzK2aC8EaNznvwtckg315aG5r+vw2hO5/hWLpg9aPz+Z4aDv86Aw3Sq4OBDwc3NVvF6BK7KUB7EoQg1eRBDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T16:56:10.808338Z"},"content_sha256":"d528e1ba5834414ee6d18f4c78823cd282c9077a0187ee41feb77ef7bee41d56","schema_version":"1.0","event_id":"sha256:d528e1ba5834414ee6d18f4c78823cd282c9077a0187ee41feb77ef7bee41d56"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2018:6VARSBRX6E3HDBTBZP3WZOXG6B","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Massively Distributed SGD: ImageNet/ResNet-50 Training in a Flash","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Hiroaki Mikami, Hisahiro Suganuma, Pongsakorn U-Chupala, Yoshiki Tanaka, Yuichi Kageyama","submitted_at":"2018-11-13T11:52:04Z","abstract_excerpt":"Scaling the distributed deep learning to a massive GPU cluster level is challenging due to the instability of the large mini-batch training and the overhead of the gradient synchronization. We address the instability of the large mini-batch training with batch-size control and label smoothing. We address the overhead of the gradient synchronization with 2D-Torus all-reduce. Specifically, 2D-Torus all-reduce arranges GPUs in a logical 2D grid and performs a series of collective operation in different orientations. These two techniques are implemented with Neural Network Libraries (NNL). We have"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1811.05233","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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:52:04Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"74Nix4ZvmdIl22t9gFudXuARekowrMQI/Ym1e3WW6AYj70XFhc7Weu/N1rm+m8n0jSXSTkaFb8DUHVqr3iGmDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-22T16:56:10.809030Z"},"content_sha256":"bf674f1c6f85e53272e0803f8d38a281cc9a76d30868ce294014f73e9b5cb689","schema_version":"1.0","event_id":"sha256:bf674f1c6f85e53272e0803f8d38a281cc9a76d30868ce294014f73e9b5cb689"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6VARSBRX6E3HDBTBZP3WZOXG6B/bundle.json","state_url":"https://pith.science/pith/6VARSBRX6E3HDBTBZP3WZOXG6B/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6VARSBRX6E3HDBTBZP3WZOXG6B/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-05-22T16:56:10Z","links":{"resolver":"https://pith.science/pith/6VARSBRX6E3HDBTBZP3WZOXG6B","bundle":"https://pith.science/pith/6VARSBRX6E3HDBTBZP3WZOXG6B/bundle.json","state":"https://pith.science/pith/6VARSBRX6E3HDBTBZP3WZOXG6B/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6VARSBRX6E3HDBTBZP3WZOXG6B/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2018:6VARSBRX6E3HDBTBZP3WZOXG6B","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"64a9eeb5c319b65bc3c68598ef29104918cd63cfecf9a846b0e74e3fb43d040f","cross_cats_sorted":["cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-11-13T11:52:04Z","title_canon_sha256":"d6cd81252e89069e2d685a8db0e14d63ceaabc34f2ea6bef568baa2887ba52e9"},"schema_version":"1.0","source":{"id":"1811.05233","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1811.05233","created_at":"2026-05-17T23:52:04Z"},{"alias_kind":"arxiv_version","alias_value":"1811.05233v2","created_at":"2026-05-17T23:52:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1811.05233","created_at":"2026-05-17T23:52:04Z"},{"alias_kind":"pith_short_12","alias_value":"6VARSBRX6E3H","created_at":"2026-05-18T12:32:11Z"},{"alias_kind":"pith_short_16","alias_value":"6VARSBRX6E3HDBTB","created_at":"2026-05-18T12:32:11Z"},{"alias_kind":"pith_short_8","alias_value":"6VARSBRX","created_at":"2026-05-18T12:32:11Z"}],"graph_snapshots":[{"event_id":"sha256:bf674f1c6f85e53272e0803f8d38a281cc9a76d30868ce294014f73e9b5cb689","target":"graph","created_at":"2026-05-17T23:52:04Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"paper":{"abstract_excerpt":"Scaling the distributed deep learning to a massive GPU cluster level is challenging due to the instability of the large mini-batch training and the overhead of the gradient synchronization. We address the instability of the large mini-batch training with batch-size control and label smoothing. We address the overhead of the gradient synchronization with 2D-Torus all-reduce. Specifically, 2D-Torus all-reduce arranges GPUs in a logical 2D grid and performs a series of collective operation in different orientations. These two techniques are implemented with Neural Network Libraries (NNL). We have","authors_text":"Hiroaki Mikami, Hisahiro Suganuma, Pongsakorn U-Chupala, Yoshiki Tanaka, Yuichi Kageyama","cross_cats":["cs.CV"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-11-13T11:52:04Z","title":"Massively Distributed SGD: ImageNet/ResNet-50 Training in a Flash"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1811.05233","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:d528e1ba5834414ee6d18f4c78823cd282c9077a0187ee41feb77ef7bee41d56","target":"record","created_at":"2026-05-17T23:52:04Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"64a9eeb5c319b65bc3c68598ef29104918cd63cfecf9a846b0e74e3fb43d040f","cross_cats_sorted":["cs.CV"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2018-11-13T11:52:04Z","title_canon_sha256":"d6cd81252e89069e2d685a8db0e14d63ceaabc34f2ea6bef568baa2887ba52e9"},"schema_version":"1.0","source":{"id":"1811.05233","kind":"arxiv","version":2}},"canonical_sha256":"f541190637f136718661cbf76cbae6f0739e46cf4ee80f156d42ce5b21a0d6c1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f541190637f136718661cbf76cbae6f0739e46cf4ee80f156d42ce5b21a0d6c1","first_computed_at":"2026-05-17T23:52:04.414293Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:52:04.414293Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"qVacuSf+RVHKpILiCUhTrivj8gHRe+kvKEhIUqCjrP17nIyjT+tp6SDaft9zouuDg08rqHx2OQZDLUPgpd41Cw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:52:04.414826Z","signed_message":"canonical_sha256_bytes"},"source_id":"1811.05233","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d528e1ba5834414ee6d18f4c78823cd282c9077a0187ee41feb77ef7bee41d56","sha256:bf674f1c6f85e53272e0803f8d38a281cc9a76d30868ce294014f73e9b5cb689"],"state_sha256":"7009375f9ecb7b586f0389db7d979591133fb9ad9fc1814e7780c0278eae74a5"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"SKkOZDIhfeTHBpoTjULwOK/PN/MLDP4NrnncDwYOn30RAaqW36pl9bwjE3kr1ZmPprqYm2LvLQQO4pHMfC6kCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-22T16:56:10.812776Z","bundle_sha256":"483ecee5e8104ea9a1d4288107ba7172b294e21d0a501bc78a8caa3b76131164"}}