{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:LW4HVKOQV4AAG4DENZPHIIHXUH","short_pith_number":"pith:LW4HVKOQ","schema_version":"1.0","canonical_sha256":"5db87aa9d0af000370646e5e7420f7a1eb1320ea52b4667813d56b320071f756","source":{"kind":"arxiv","id":"2404.11068","version":1},"attestation_state":"computed","paper":{"title":"ScaleFold: Reducing AlphaFold Initial Training Time to 10 Hours","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.DC","q-bio.QM"],"primary_cat":"cs.LG","authors_text":"Arkadiusz Nowaczynski, Feiwen Zhu, Jie Xin, Jun Yang, Michael Andersch, Michal Marcinkiewicz, Rundong Li, Sukru Burc Eryilmaz, Yifei Song","submitted_at":"2024-04-17T04:55:33Z","abstract_excerpt":"AlphaFold2 has been hailed as a breakthrough in protein folding. It can rapidly predict protein structures with lab-grade accuracy. However, its implementation does not include the necessary training code. OpenFold is the first trainable public reimplementation of AlphaFold. AlphaFold training procedure is prohibitively time-consuming, and gets diminishing benefits from scaling to more compute resources. In this work, we conducted a comprehensive analysis on the AlphaFold training procedure based on Openfold, identified that inefficient communications and overhead-dominated computations were t"},"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":"2404.11068","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2024-04-17T04:55:33Z","cross_cats_sorted":["cs.AI","cs.DC","q-bio.QM"],"title_canon_sha256":"1bac553e4a1590b07e4812eace29f421cba490c9adb5613cad1cbcbb063fbe14","abstract_canon_sha256":"3c4fb94bde755c00d87b57e98c7362f4017ea26f9f31fafbf1fcc245d16240d8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:09:02.928460Z","signature_b64":"YJqB8g4G3PiGzRuY7cgDvIKAYsly/y1cBu9QHJntC6eucBw4wjdadCMcdxi1zbJYnTvHfPX34i9s2II2+CEkAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5db87aa9d0af000370646e5e7420f7a1eb1320ea52b4667813d56b320071f756","last_reissued_at":"2026-07-05T08:09:02.927966Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:09:02.927966Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ScaleFold: Reducing AlphaFold Initial Training Time to 10 Hours","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.DC","q-bio.QM"],"primary_cat":"cs.LG","authors_text":"Arkadiusz Nowaczynski, Feiwen Zhu, Jie Xin, Jun Yang, Michael Andersch, Michal Marcinkiewicz, Rundong Li, Sukru Burc Eryilmaz, Yifei Song","submitted_at":"2024-04-17T04:55:33Z","abstract_excerpt":"AlphaFold2 has been hailed as a breakthrough in protein folding. It can rapidly predict protein structures with lab-grade accuracy. However, its implementation does not include the necessary training code. OpenFold is the first trainable public reimplementation of AlphaFold. AlphaFold training procedure is prohibitively time-consuming, and gets diminishing benefits from scaling to more compute resources. In this work, we conducted a comprehensive analysis on the AlphaFold training procedure based on Openfold, identified that inefficient communications and overhead-dominated computations were t"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.11068","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/2404.11068/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2404.11068","created_at":"2026-07-05T08:09:02.928037+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.11068v1","created_at":"2026-07-05T08:09:02.928037+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.11068","created_at":"2026-07-05T08:09:02.928037+00:00"},{"alias_kind":"pith_short_12","alias_value":"LW4HVKOQV4AA","created_at":"2026-07-05T08:09:02.928037+00:00"},{"alias_kind":"pith_short_16","alias_value":"LW4HVKOQV4AAG4DE","created_at":"2026-07-05T08:09:02.928037+00:00"},{"alias_kind":"pith_short_8","alias_value":"LW4HVKOQ","created_at":"2026-07-05T08:09:02.928037+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.05439","citing_title":"Design-CP: Context Parallelism for Design of Protein Nanoparticles","ref_index":93,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH","json":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH.json","graph_json":"https://pith.science/api/pith-number/LW4HVKOQV4AAG4DENZPHIIHXUH/graph.json","events_json":"https://pith.science/api/pith-number/LW4HVKOQV4AAG4DENZPHIIHXUH/events.json","paper":"https://pith.science/paper/LW4HVKOQ"},"agent_actions":{"view_html":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH","download_json":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH.json","view_paper":"https://pith.science/paper/LW4HVKOQ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.11068&json=true","fetch_graph":"https://pith.science/api/pith-number/LW4HVKOQV4AAG4DENZPHIIHXUH/graph.json","fetch_events":"https://pith.science/api/pith-number/LW4HVKOQV4AAG4DENZPHIIHXUH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH/action/storage_attestation","attest_author":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH/action/author_attestation","sign_citation":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH/action/citation_signature","submit_replication":"https://pith.science/pith/LW4HVKOQV4AAG4DENZPHIIHXUH/action/replication_record"}},"created_at":"2026-07-05T08:09:02.928037+00:00","updated_at":"2026-07-05T08:09:02.928037+00:00"}