{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2021:HQBKPCGKW52HNMYVOTYGQACC6N","short_pith_number":"pith:HQBKPCGK","canonical_record":{"source":{"id":"2110.03888","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T04:24:51Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"d85d2047e6c6353110efd485453e8302144b18d626b2d59c3ecd39d82c593674","abstract_canon_sha256":"e18c0223102adc9d3a24af047e4f496b9cd104f775a93fc35951dc818e4b3aef"},"schema_version":"1.0"},"canonical_sha256":"3c02a788cab77476b31574f0680042f35f6831c5bc3344b81387d5f0e7f8772b","source":{"kind":"arxiv","id":"2110.03888","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2110.03888","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"arxiv_version","alias_value":"2110.03888v3","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.03888","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"pith_short_12","alias_value":"HQBKPCGKW52H","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"pith_short_16","alias_value":"HQBKPCGKW52HNMYV","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"pith_short_8","alias_value":"HQBKPCGK","created_at":"2026-07-05T03:25:17Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2021:HQBKPCGKW52HNMYVOTYGQACC6N","target":"record","payload":{"canonical_record":{"source":{"id":"2110.03888","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T04:24:51Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"d85d2047e6c6353110efd485453e8302144b18d626b2d59c3ecd39d82c593674","abstract_canon_sha256":"e18c0223102adc9d3a24af047e4f496b9cd104f775a93fc35951dc818e4b3aef"},"schema_version":"1.0"},"canonical_sha256":"3c02a788cab77476b31574f0680042f35f6831c5bc3344b81387d5f0e7f8772b","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:25:17.733924Z","signature_b64":"Yj3cSZwwgl1ASQHK2u030lQC6gQuvFZtbU60cYxOZ1qVu8e6Udo0H9Ep6cZBmYcFJIU7k4As2WXH/GfY+p46Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3c02a788cab77476b31574f0680042f35f6831c5bc3344b81387d5f0e7f8772b","last_reissued_at":"2026-07-05T03:25:17.733456Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:25:17.733456Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2110.03888","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T03:25:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gzhcSvyZ3rS5Xw25rblYi8Ffy8UWaZQ46QeB3512CRbeI384+XeyCnYD1sCUUPrxYDEB1xWY1KKc4cD/ltSBCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T10:54:08.352864Z"},"content_sha256":"d64e6e6eb06c7b1835c0a7d1cbc44f1a4f4aff6732fb765446081e7bec5d7f27","schema_version":"1.0","event_id":"sha256:d64e6e6eb06c7b1835c0a7d1cbc44f1a4f4aff6732fb765446081e7bec5d7f27"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2021:HQBKPCGKW52HNMYVOTYGQACC6N","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Ang Wang, An Yang, Chang Zhou, Hongxia Yang, Jie Zhang, Jingren Zhou, Jinze Bai, Junyang Lin, Le Jiang, Wei Lin, Xianyan Jia, Yong Li","submitted_at":"2021-10-08T04:24:51Z","abstract_excerpt":"Recent expeditious developments in deep learning algorithms, distributed training, and even hardware design for large models have enabled training extreme-scale models, say GPT-3 and Switch Transformer possessing hundreds of billions or even trillions of parameters. However, under limited resources, extreme-scale model training that requires enormous amounts of computes and memory footprint suffers from frustratingly low efficiency in model convergence. In this paper, we propose a simple training strategy called \"Pseudo-to-Real\" for high-memory-footprint-required large models. Pseudo-to-Real i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.03888","kind":"arxiv","version":3},"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/2110.03888/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":"2026-07-05T03:25:17Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"qygDto8b22o78T564ClgKo1mUn1o0Vb7ibHd6RgObgStbk3gEXbKG7U4eCVVWRBLlaBShW1E0h4dMOn71mrTDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-19T10:54:08.353262Z"},"content_sha256":"0699bd0aa8e7b6bcb276886ca63a9199bb34be1dc97be87b1b03ad9bfc59e72d","schema_version":"1.0","event_id":"sha256:0699bd0aa8e7b6bcb276886ca63a9199bb34be1dc97be87b1b03ad9bfc59e72d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/HQBKPCGKW52HNMYVOTYGQACC6N/bundle.json","state_url":"https://pith.science/pith/HQBKPCGKW52HNMYVOTYGQACC6N/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/HQBKPCGKW52HNMYVOTYGQACC6N/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":"2026-08-19T10:54:08Z","links":{"resolver":"https://pith.science/pith/HQBKPCGKW52HNMYVOTYGQACC6N","bundle":"https://pith.science/pith/HQBKPCGKW52HNMYVOTYGQACC6N/bundle.json","state":"https://pith.science/pith/HQBKPCGKW52HNMYVOTYGQACC6N/state.json","well_known_bundle":"https://pith.science/.well-known/pith/HQBKPCGKW52HNMYVOTYGQACC6N/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2021:HQBKPCGKW52HNMYVOTYGQACC6N","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":"e18c0223102adc9d3a24af047e4f496b9cd104f775a93fc35951dc818e4b3aef","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T04:24:51Z","title_canon_sha256":"d85d2047e6c6353110efd485453e8302144b18d626b2d59c3ecd39d82c593674"},"schema_version":"1.0","source":{"id":"2110.03888","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2110.03888","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"arxiv_version","alias_value":"2110.03888v3","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2110.03888","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"pith_short_12","alias_value":"HQBKPCGKW52H","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"pith_short_16","alias_value":"HQBKPCGKW52HNMYV","created_at":"2026-07-05T03:25:17Z"},{"alias_kind":"pith_short_8","alias_value":"HQBKPCGK","created_at":"2026-07-05T03:25:17Z"}],"graph_snapshots":[{"event_id":"sha256:0699bd0aa8e7b6bcb276886ca63a9199bb34be1dc97be87b1b03ad9bfc59e72d","target":"graph","created_at":"2026-07-05T03:25:17Z","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/2110.03888/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent expeditious developments in deep learning algorithms, distributed training, and even hardware design for large models have enabled training extreme-scale models, say GPT-3 and Switch Transformer possessing hundreds of billions or even trillions of parameters. However, under limited resources, extreme-scale model training that requires enormous amounts of computes and memory footprint suffers from frustratingly low efficiency in model convergence. In this paper, we propose a simple training strategy called \"Pseudo-to-Real\" for high-memory-footprint-required large models. Pseudo-to-Real i","authors_text":"Ang Wang, An Yang, Chang Zhou, Hongxia Yang, Jie Zhang, Jingren Zhou, Jinze Bai, Junyang Lin, Le Jiang, Wei Lin, Xianyan Jia, Yong Li","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T04:24:51Z","title":"M6-10T: A Sharing-Delinking Paradigm for Efficient Multi-Trillion Parameter Pretraining"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2110.03888","kind":"arxiv","version":3},"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:d64e6e6eb06c7b1835c0a7d1cbc44f1a4f4aff6732fb765446081e7bec5d7f27","target":"record","created_at":"2026-07-05T03:25:17Z","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":"e18c0223102adc9d3a24af047e4f496b9cd104f775a93fc35951dc818e4b3aef","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2021-10-08T04:24:51Z","title_canon_sha256":"d85d2047e6c6353110efd485453e8302144b18d626b2d59c3ecd39d82c593674"},"schema_version":"1.0","source":{"id":"2110.03888","kind":"arxiv","version":3}},"canonical_sha256":"3c02a788cab77476b31574f0680042f35f6831c5bc3344b81387d5f0e7f8772b","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3c02a788cab77476b31574f0680042f35f6831c5bc3344b81387d5f0e7f8772b","first_computed_at":"2026-07-05T03:25:17.733456Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T03:25:17.733456Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Yj3cSZwwgl1ASQHK2u030lQC6gQuvFZtbU60cYxOZ1qVu8e6Udo0H9Ep6cZBmYcFJIU7k4As2WXH/GfY+p46Bg==","signature_status":"signed_v1","signed_at":"2026-07-05T03:25:17.733924Z","signed_message":"canonical_sha256_bytes"},"source_id":"2110.03888","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d64e6e6eb06c7b1835c0a7d1cbc44f1a4f4aff6732fb765446081e7bec5d7f27","sha256:0699bd0aa8e7b6bcb276886ca63a9199bb34be1dc97be87b1b03ad9bfc59e72d"],"state_sha256":"008be5f938d364fd442b9bb78554ef7fd689ab8dc795fbe89da6857e09779a7e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5UxrO6yXf0wCojSy+uSsueyoSvVzb195iqm3KcqrYEKYoqb/4z7u3AfZgm16O8mAXWd04yjSoj3o6teL1SEKAQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-19T10:54:08.355646Z","bundle_sha256":"29d82a1b04a2e6231145f33b57c6f2d61a9e19f51311ed7abae2a8f7a415faca"}}