{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:IEJ5JAJP7IBVZXH3BDKJTRL3YS","short_pith_number":"pith:IEJ5JAJP","canonical_record":{"source":{"id":"2104.05343","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-04-12T10:47:16Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"c7585625f45fa4b2c2522f8933ff2190e0fd565e1c6d815be3e2ff89fa3847d5","abstract_canon_sha256":"061a8c6daa988df2196f556e7f35ac09265d167a5351871211a1f50c83620d0b"},"schema_version":"1.0"},"canonical_sha256":"4113d4812ffa035cdcfb08d499c57bc4a821be690f6a3a00f53a26a6e80f82c3","source":{"kind":"arxiv","id":"2104.05343","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2104.05343","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"arxiv_version","alias_value":"2104.05343v1","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.05343","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"pith_short_12","alias_value":"IEJ5JAJP7IBV","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"pith_short_16","alias_value":"IEJ5JAJP7IBVZXH3","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"pith_short_8","alias_value":"IEJ5JAJP","created_at":"2026-07-05T02:31:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:IEJ5JAJP7IBVZXH3BDKJTRL3YS","target":"record","payload":{"canonical_record":{"source":{"id":"2104.05343","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-04-12T10:47:16Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"c7585625f45fa4b2c2522f8933ff2190e0fd565e1c6d815be3e2ff89fa3847d5","abstract_canon_sha256":"061a8c6daa988df2196f556e7f35ac09265d167a5351871211a1f50c83620d0b"},"schema_version":"1.0"},"canonical_sha256":"4113d4812ffa035cdcfb08d499c57bc4a821be690f6a3a00f53a26a6e80f82c3","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:31:11.425287Z","signature_b64":"Wb/I7t0oSn7Y/lkOMrpe8Ep2oiBh5HuT8ySI+wwEpR8bLOb2ud4vfDHyuxAYMBkXy3QY+AGroylClA/tzmNfCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4113d4812ffa035cdcfb08d499c57bc4a821be690f6a3a00f53a26a6e80f82c3","last_reissued_at":"2026-07-05T02:31:11.424801Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:31:11.424801Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2104.05343","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":"2026-07-05T02:31:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KcBgbcmtlJKW3aG2vyjuKcqlX7885D1naSvmP/Nu9cddhcahY3dVCAqVldSrlU+7umrbWY3FnmO1J0+/0+uuDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T07:36:11.647908Z"},"content_sha256":"0c5828b932badd95c98722b9394e68c5f778c20b5548eb2073010a49758b2de1","schema_version":"1.0","event_id":"sha256:0c5828b932badd95c98722b9394e68c5f778c20b5548eb2073010a49758b2de1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:IEJ5JAJP7IBVZXH3BDKJTRL3YS","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"An Efficient 2D Method for Training Super-Large Deep Learning Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Chaoyu Gong, Qifan Xu, Shenggui Li, Yang You","submitted_at":"2021-04-12T10:47:16Z","abstract_excerpt":"Huge neural network models have shown unprecedented performance in real-world applications. However, due to memory constraints, model parallelism must be utilized to host large models that would otherwise not fit into the memory of a single device. Previous methods like Megatron partition the parameters of the entire model among multiple devices, while each device has to accommodate the redundant activations in forward and backward pass. In this work, we propose Optimus, a highly efficient and scalable 2D-partition paradigm of model parallelism that would facilitate the training of infinitely "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.05343","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2104.05343/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"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":"2026-07-05T02:31:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"48msX0ln4CU9c0d1sg+pAWW14s9Ki6AZ0p/ykqycEh3XDgoLH49TulsDoEbgW7U1dMI87BFINMxdwYrA4oKVAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-01T07:36:11.648406Z"},"content_sha256":"4d4e0d59fddc8c97de99c3b1304bc56dace37684898ec1f9e792b794765a5003","schema_version":"1.0","event_id":"sha256:4d4e0d59fddc8c97de99c3b1304bc56dace37684898ec1f9e792b794765a5003"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/IEJ5JAJP7IBVZXH3BDKJTRL3YS/bundle.json","state_url":"https://pith.science/pith/IEJ5JAJP7IBVZXH3BDKJTRL3YS/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/IEJ5JAJP7IBVZXH3BDKJTRL3YS/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":"2026-08-01T07:36:11Z","links":{"resolver":"https://pith.science/pith/IEJ5JAJP7IBVZXH3BDKJTRL3YS","bundle":"https://pith.science/pith/IEJ5JAJP7IBVZXH3BDKJTRL3YS/bundle.json","state":"https://pith.science/pith/IEJ5JAJP7IBVZXH3BDKJTRL3YS/state.json","well_known_bundle":"https://pith.science/.well-known/pith/IEJ5JAJP7IBVZXH3BDKJTRL3YS/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:IEJ5JAJP7IBVZXH3BDKJTRL3YS","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":"061a8c6daa988df2196f556e7f35ac09265d167a5351871211a1f50c83620d0b","cross_cats_sorted":["cs.DC"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-04-12T10:47:16Z","title_canon_sha256":"c7585625f45fa4b2c2522f8933ff2190e0fd565e1c6d815be3e2ff89fa3847d5"},"schema_version":"1.0","source":{"id":"2104.05343","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2104.05343","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"arxiv_version","alias_value":"2104.05343v1","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2104.05343","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"pith_short_12","alias_value":"IEJ5JAJP7IBV","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"pith_short_16","alias_value":"IEJ5JAJP7IBVZXH3","created_at":"2026-07-05T02:31:11Z"},{"alias_kind":"pith_short_8","alias_value":"IEJ5JAJP","created_at":"2026-07-05T02:31:11Z"}],"graph_snapshots":[{"event_id":"sha256:4d4e0d59fddc8c97de99c3b1304bc56dace37684898ec1f9e792b794765a5003","target":"graph","created_at":"2026-07-05T02:31:11Z","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"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2104.05343/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Huge neural network models have shown unprecedented performance in real-world applications. However, due to memory constraints, model parallelism must be utilized to host large models that would otherwise not fit into the memory of a single device. Previous methods like Megatron partition the parameters of the entire model among multiple devices, while each device has to accommodate the redundant activations in forward and backward pass. In this work, we propose Optimus, a highly efficient and scalable 2D-partition paradigm of model parallelism that would facilitate the training of infinitely ","authors_text":"Chaoyu Gong, Qifan Xu, Shenggui Li, Yang You","cross_cats":["cs.DC"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-04-12T10:47:16Z","title":"An Efficient 2D Method for Training Super-Large Deep Learning Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2104.05343","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":"sha256:0c5828b932badd95c98722b9394e68c5f778c20b5548eb2073010a49758b2de1","target":"record","created_at":"2026-07-05T02:31:11Z","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":"061a8c6daa988df2196f556e7f35ac09265d167a5351871211a1f50c83620d0b","cross_cats_sorted":["cs.DC"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2021-04-12T10:47:16Z","title_canon_sha256":"c7585625f45fa4b2c2522f8933ff2190e0fd565e1c6d815be3e2ff89fa3847d5"},"schema_version":"1.0","source":{"id":"2104.05343","kind":"arxiv","version":1}},"canonical_sha256":"4113d4812ffa035cdcfb08d499c57bc4a821be690f6a3a00f53a26a6e80f82c3","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"4113d4812ffa035cdcfb08d499c57bc4a821be690f6a3a00f53a26a6e80f82c3","first_computed_at":"2026-07-05T02:31:11.424801Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T02:31:11.424801Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Wb/I7t0oSn7Y/lkOMrpe8Ep2oiBh5HuT8ySI+wwEpR8bLOb2ud4vfDHyuxAYMBkXy3QY+AGroylClA/tzmNfCA==","signature_status":"signed_v1","signed_at":"2026-07-05T02:31:11.425287Z","signed_message":"canonical_sha256_bytes"},"source_id":"2104.05343","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0c5828b932badd95c98722b9394e68c5f778c20b5548eb2073010a49758b2de1","sha256:4d4e0d59fddc8c97de99c3b1304bc56dace37684898ec1f9e792b794765a5003"],"state_sha256":"6df5a4a738137148c442963407f07b9328a9b3e28a571fae7ba050ee3b61cebb"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LEDavQBXmYiLfsq8kSfDsQivrICIxGXdmVWKCvC8A9U0wPUfJgF7O9AhnwFZIwlFavYrjocYke30aUAIKhFNCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-01T07:36:11.653138Z","bundle_sha256":"e14385058e57bea59b75415568fc80d282d6a40e3e7e792a67075e2e96efc338"}}