{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:VWMY65AJTT2XBISMIBUWAGDILZ","short_pith_number":"pith:VWMY65AJ","canonical_record":{"source":{"id":"1903.06681","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2019-03-15T17:25:01Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"d3973d41ac957390f0312381da69eb5ea8dcb0a49004a97e6886df015dea04a1","abstract_canon_sha256":"c08544c994a8ffce91a76f0d3851f2559677fb0646a9adb2c227296fe437adf1"},"schema_version":"1.0"},"canonical_sha256":"ad998f74099cf570a24c40696018685e42b4d93f85226988a52d860556ae1711","source":{"kind":"arxiv","id":"1903.06681","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1903.06681","created_at":"2026-05-17T23:51:10Z"},{"alias_kind":"arxiv_version","alias_value":"1903.06681v1","created_at":"2026-05-17T23:51:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1903.06681","created_at":"2026-05-17T23:51:10Z"},{"alias_kind":"pith_short_12","alias_value":"VWMY65AJTT2X","created_at":"2026-05-18T12:33:30Z"},{"alias_kind":"pith_short_16","alias_value":"VWMY65AJTT2XBISM","created_at":"2026-05-18T12:33:30Z"},{"alias_kind":"pith_short_8","alias_value":"VWMY65AJ","created_at":"2026-05-18T12:33:30Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:VWMY65AJTT2XBISMIBUWAGDILZ","target":"record","payload":{"canonical_record":{"source":{"id":"1903.06681","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2019-03-15T17:25:01Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"d3973d41ac957390f0312381da69eb5ea8dcb0a49004a97e6886df015dea04a1","abstract_canon_sha256":"c08544c994a8ffce91a76f0d3851f2559677fb0646a9adb2c227296fe437adf1"},"schema_version":"1.0"},"canonical_sha256":"ad998f74099cf570a24c40696018685e42b4d93f85226988a52d860556ae1711","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:51:10.088973Z","signature_b64":"7fCl8k+YZ24niyleKv7gNCBF5kaiDFhIqu8l0ZmsHWgFotmmx3WbROBM84fOftyNP/9yY9pREpbxNhwqcGfbCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ad998f74099cf570a24c40696018685e42b4d93f85226988a52d860556ae1711","last_reissued_at":"2026-05-17T23:51:10.088356Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:51:10.088356Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1903.06681","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-05-17T23:51:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hT3Neh921+PKc8sJsb2MSbZydWi2BGjw81AmEiBH0vtFTKjIQhSg94yCh1yDH9JWVSe2Eb2liBtmakJC0vWoCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T13:23:55.050553Z"},"content_sha256":"e431c8356304ac7c15807db514cb10a99ce0eacd3726963b30b4e1bb6d2e2257","schema_version":"1.0","event_id":"sha256:e431c8356304ac7c15807db514cb10a99ce0eacd3726963b30b4e1bb6d2e2257"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:VWMY65AJTT2XBISMIBUWAGDILZ","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Improving Strong-Scaling of CNN Training by Exploiting Finer-Grained Parallelism","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.DC","authors_text":"Brian Van Essen, Marc Snir, Naoya Maruyama, Nikoli Dryden, Tim Moon, Tom Benson","submitted_at":"2019-03-15T17:25:01Z","abstract_excerpt":"Scaling CNN training is necessary to keep up with growing datasets and reduce training time. We also see an emerging need to handle datasets with very large samples, where memory requirements for training are large. Existing training frameworks use a data-parallel approach that partitions samples within a mini-batch, but limits to scaling the mini-batch size and memory consumption makes this untenable for large samples. We describe and implement new approaches to convolution, which parallelize using spatial decomposition or a combination of sample and spatial decomposition. This introduces man"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1903.06681","kind":"arxiv","version":1},"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":"2026-05-17T23:51:10Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"61F0GH9Iq8P3fhn/+4Q95rJDRH+gEfVACnpXrHpZjbhPi8oCI8oCb3/xRgw77TgU6lfP+bvm/5xqtbrkyfUOAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-25T13:23:55.051209Z"},"content_sha256":"c380f797e6eb594e29bb86248a3d6ad3ced28da5052dabbdb0b04c30c9e8b1ca","schema_version":"1.0","event_id":"sha256:c380f797e6eb594e29bb86248a3d6ad3ced28da5052dabbdb0b04c30c9e8b1ca"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VWMY65AJTT2XBISMIBUWAGDILZ/bundle.json","state_url":"https://pith.science/pith/VWMY65AJTT2XBISMIBUWAGDILZ/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VWMY65AJTT2XBISMIBUWAGDILZ/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":"2026-05-25T13:23:55Z","links":{"resolver":"https://pith.science/pith/VWMY65AJTT2XBISMIBUWAGDILZ","bundle":"https://pith.science/pith/VWMY65AJTT2XBISMIBUWAGDILZ/bundle.json","state":"https://pith.science/pith/VWMY65AJTT2XBISMIBUWAGDILZ/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VWMY65AJTT2XBISMIBUWAGDILZ/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:VWMY65AJTT2XBISMIBUWAGDILZ","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":"c08544c994a8ffce91a76f0d3851f2559677fb0646a9adb2c227296fe437adf1","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2019-03-15T17:25:01Z","title_canon_sha256":"d3973d41ac957390f0312381da69eb5ea8dcb0a49004a97e6886df015dea04a1"},"schema_version":"1.0","source":{"id":"1903.06681","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1903.06681","created_at":"2026-05-17T23:51:10Z"},{"alias_kind":"arxiv_version","alias_value":"1903.06681v1","created_at":"2026-05-17T23:51:10Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1903.06681","created_at":"2026-05-17T23:51:10Z"},{"alias_kind":"pith_short_12","alias_value":"VWMY65AJTT2X","created_at":"2026-05-18T12:33:30Z"},{"alias_kind":"pith_short_16","alias_value":"VWMY65AJTT2XBISM","created_at":"2026-05-18T12:33:30Z"},{"alias_kind":"pith_short_8","alias_value":"VWMY65AJ","created_at":"2026-05-18T12:33:30Z"}],"graph_snapshots":[{"event_id":"sha256:c380f797e6eb594e29bb86248a3d6ad3ced28da5052dabbdb0b04c30c9e8b1ca","target":"graph","created_at":"2026-05-17T23:51:10Z","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 CNN training is necessary to keep up with growing datasets and reduce training time. We also see an emerging need to handle datasets with very large samples, where memory requirements for training are large. Existing training frameworks use a data-parallel approach that partitions samples within a mini-batch, but limits to scaling the mini-batch size and memory consumption makes this untenable for large samples. We describe and implement new approaches to convolution, which parallelize using spatial decomposition or a combination of sample and spatial decomposition. This introduces man","authors_text":"Brian Van Essen, Marc Snir, Naoya Maruyama, Nikoli Dryden, Tim Moon, Tom Benson","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2019-03-15T17:25:01Z","title":"Improving Strong-Scaling of CNN Training by Exploiting Finer-Grained Parallelism"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1903.06681","kind":"arxiv","version":1},"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:e431c8356304ac7c15807db514cb10a99ce0eacd3726963b30b4e1bb6d2e2257","target":"record","created_at":"2026-05-17T23:51:10Z","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":"c08544c994a8ffce91a76f0d3851f2559677fb0646a9adb2c227296fe437adf1","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2019-03-15T17:25:01Z","title_canon_sha256":"d3973d41ac957390f0312381da69eb5ea8dcb0a49004a97e6886df015dea04a1"},"schema_version":"1.0","source":{"id":"1903.06681","kind":"arxiv","version":1}},"canonical_sha256":"ad998f74099cf570a24c40696018685e42b4d93f85226988a52d860556ae1711","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ad998f74099cf570a24c40696018685e42b4d93f85226988a52d860556ae1711","first_computed_at":"2026-05-17T23:51:10.088356Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:51:10.088356Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"7fCl8k+YZ24niyleKv7gNCBF5kaiDFhIqu8l0ZmsHWgFotmmx3WbROBM84fOftyNP/9yY9pREpbxNhwqcGfbCg==","signature_status":"signed_v1","signed_at":"2026-05-17T23:51:10.088973Z","signed_message":"canonical_sha256_bytes"},"source_id":"1903.06681","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e431c8356304ac7c15807db514cb10a99ce0eacd3726963b30b4e1bb6d2e2257","sha256:c380f797e6eb594e29bb86248a3d6ad3ced28da5052dabbdb0b04c30c9e8b1ca"],"state_sha256":"2402c7170b93c10ebcd61cafacf200aaec064ac77e8d9375e1e4cecef751c7a2"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5BJ9Kd7OkfQoS0/nnsoFA99jVs4rRNKWNbpyMNV1dhWWot0Bva4fgR5p55g8ZsSOyby2BSwkbSIoYUy1PqARCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-25T13:23:55.056011Z","bundle_sha256":"f62e44631d457cfea2a4d6813de38e2c3e75ad7eaffd8e34e10000a920aa9443"}}