{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:25X6ZO74XWYHAYEJ3JIVQ64YAW","short_pith_number":"pith:25X6ZO74","canonical_record":{"source":{"id":"2412.21124","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-30T17:55:28Z","cross_cats_sorted":["math.OC","stat.ML"],"title_canon_sha256":"4621507e37f9afdde61b9056ef6d21e3ee21dc1b1114f231be4e2dd78449267d","abstract_canon_sha256":"bdf52e51bb146d2928ec0090580003154b70ee952f7288db6b946d96add21b30"},"schema_version":"1.0"},"canonical_sha256":"d76fecbbfcbdb0706089da51587b9805bc5637dc626aa9aa23d57813cbbd99c0","source":{"kind":"arxiv","id":"2412.21124","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.21124","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"arxiv_version","alias_value":"2412.21124v2","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.21124","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"pith_short_12","alias_value":"25X6ZO74XWYH","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"pith_short_16","alias_value":"25X6ZO74XWYHAYEJ","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"pith_short_8","alias_value":"25X6ZO74","created_at":"2026-07-05T10:31:52Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:25X6ZO74XWYHAYEJ3JIVQ64YAW","target":"record","payload":{"canonical_record":{"source":{"id":"2412.21124","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-30T17:55:28Z","cross_cats_sorted":["math.OC","stat.ML"],"title_canon_sha256":"4621507e37f9afdde61b9056ef6d21e3ee21dc1b1114f231be4e2dd78449267d","abstract_canon_sha256":"bdf52e51bb146d2928ec0090580003154b70ee952f7288db6b946d96add21b30"},"schema_version":"1.0"},"canonical_sha256":"d76fecbbfcbdb0706089da51587b9805bc5637dc626aa9aa23d57813cbbd99c0","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:31:52.360443Z","signature_b64":"e7Q4coUBZheoMNEgjvDQMBKTrw0Tg7sMN0eTRaWBjhvJ1HwqiW4ZszZed65OAtkbtGH+0wlr+vCxnkV7SGujDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d76fecbbfcbdb0706089da51587b9805bc5637dc626aa9aa23d57813cbbd99c0","last_reissued_at":"2026-07-05T10:31:52.359930Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:31:52.359930Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2412.21124","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":"2026-07-05T10:31:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eXIUOzYvpldnDlFj+MgizZzkUhoi0v7mqUjYgI8yDKlKDaOhTdm+/cmEeNBnBWpgjFcrMBfBVihmSpVm8P0PDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T13:35:59.793788Z"},"content_sha256":"315ebca5a26375dbffac4a18aeb67f9654fb8958a0e487eade4d9b2e5e631099","schema_version":"1.0","event_id":"sha256:315ebca5a26375dbffac4a18aeb67f9654fb8958a0e487eade4d9b2e5e631099"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:25X6ZO74XWYHAYEJ3JIVQ64YAW","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Adaptive Batch Size Schedules for Distributed Training of Language Models with Data and Model Parallelism","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Chenwei Xu, Han Liu, Mladen Kolar, Tim Tsz-Kit Lau, Weijian Li","submitted_at":"2024-12-30T17:55:28Z","abstract_excerpt":"An appropriate choice of batch sizes in large-scale model training is crucial, yet it involves an intrinsic yet inevitable dilemma: large-batch training improves training efficiency in terms of memory utilization, while generalization performance often deteriorates due to small amounts of gradient noise. Despite this dilemma, the common practice of choosing batch sizes in language model training often prioritizes training efficiency -- employing either constant large sizes with data parallelism or implementing batch size warmup schedules. However, such batch size schedule designs remain heuris"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.21124","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":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2412.21124/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":"2026-07-05T10:31:52Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"AhHFWGvYjPiH67tSsl2UFs9aYg57Z7Hka82kzlfw//ysjoSKxP3an/dafhUWcq7+gqMLScPaVDtXipp6RO+sDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T13:35:59.794512Z"},"content_sha256":"873d9df92cf0b9f109760fd5e4725e060332d908fbe1e4bf2ef35d6cf69700e3","schema_version":"1.0","event_id":"sha256:873d9df92cf0b9f109760fd5e4725e060332d908fbe1e4bf2ef35d6cf69700e3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/25X6ZO74XWYHAYEJ3JIVQ64YAW/bundle.json","state_url":"https://pith.science/pith/25X6ZO74XWYHAYEJ3JIVQ64YAW/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/25X6ZO74XWYHAYEJ3JIVQ64YAW/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":"2026-08-14T13:35:59Z","links":{"resolver":"https://pith.science/pith/25X6ZO74XWYHAYEJ3JIVQ64YAW","bundle":"https://pith.science/pith/25X6ZO74XWYHAYEJ3JIVQ64YAW/bundle.json","state":"https://pith.science/pith/25X6ZO74XWYHAYEJ3JIVQ64YAW/state.json","well_known_bundle":"https://pith.science/.well-known/pith/25X6ZO74XWYHAYEJ3JIVQ64YAW/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:25X6ZO74XWYHAYEJ3JIVQ64YAW","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":"bdf52e51bb146d2928ec0090580003154b70ee952f7288db6b946d96add21b30","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-30T17:55:28Z","title_canon_sha256":"4621507e37f9afdde61b9056ef6d21e3ee21dc1b1114f231be4e2dd78449267d"},"schema_version":"1.0","source":{"id":"2412.21124","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2412.21124","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"arxiv_version","alias_value":"2412.21124v2","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.21124","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"pith_short_12","alias_value":"25X6ZO74XWYH","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"pith_short_16","alias_value":"25X6ZO74XWYHAYEJ","created_at":"2026-07-05T10:31:52Z"},{"alias_kind":"pith_short_8","alias_value":"25X6ZO74","created_at":"2026-07-05T10:31:52Z"}],"graph_snapshots":[{"event_id":"sha256:873d9df92cf0b9f109760fd5e4725e060332d908fbe1e4bf2ef35d6cf69700e3","target":"graph","created_at":"2026-07-05T10:31:52Z","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/2412.21124/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"An appropriate choice of batch sizes in large-scale model training is crucial, yet it involves an intrinsic yet inevitable dilemma: large-batch training improves training efficiency in terms of memory utilization, while generalization performance often deteriorates due to small amounts of gradient noise. Despite this dilemma, the common practice of choosing batch sizes in language model training often prioritizes training efficiency -- employing either constant large sizes with data parallelism or implementing batch size warmup schedules. However, such batch size schedule designs remain heuris","authors_text":"Chenwei Xu, Han Liu, Mladen Kolar, Tim Tsz-Kit Lau, Weijian Li","cross_cats":["math.OC","stat.ML"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-30T17:55:28Z","title":"Adaptive Batch Size Schedules for Distributed Training of Language Models with Data and Model Parallelism"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.21124","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":"sha256:315ebca5a26375dbffac4a18aeb67f9654fb8958a0e487eade4d9b2e5e631099","target":"record","created_at":"2026-07-05T10:31:52Z","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":"bdf52e51bb146d2928ec0090580003154b70ee952f7288db6b946d96add21b30","cross_cats_sorted":["math.OC","stat.ML"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-12-30T17:55:28Z","title_canon_sha256":"4621507e37f9afdde61b9056ef6d21e3ee21dc1b1114f231be4e2dd78449267d"},"schema_version":"1.0","source":{"id":"2412.21124","kind":"arxiv","version":2}},"canonical_sha256":"d76fecbbfcbdb0706089da51587b9805bc5637dc626aa9aa23d57813cbbd99c0","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d76fecbbfcbdb0706089da51587b9805bc5637dc626aa9aa23d57813cbbd99c0","first_computed_at":"2026-07-05T10:31:52.359930Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:31:52.359930Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"e7Q4coUBZheoMNEgjvDQMBKTrw0Tg7sMN0eTRaWBjhvJ1HwqiW4ZszZed65OAtkbtGH+0wlr+vCxnkV7SGujDw==","signature_status":"signed_v1","signed_at":"2026-07-05T10:31:52.360443Z","signed_message":"canonical_sha256_bytes"},"source_id":"2412.21124","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:315ebca5a26375dbffac4a18aeb67f9654fb8958a0e487eade4d9b2e5e631099","sha256:873d9df92cf0b9f109760fd5e4725e060332d908fbe1e4bf2ef35d6cf69700e3"],"state_sha256":"becdc522eaeb41f36efd70ac0b0c1ec64ed369ef55da5252dcb75d4591c8b818"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"fktUL7ENd2L+8aOXYND+Ux8j7X98sHKQ1V8ofiPT+yLv5AuKa53i32HzxIQRQaW4nuoGHISJjOb6xFo2Bpz3AQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T13:35:59.800309Z","bundle_sha256":"3f6c955474fbc284fe1063b9c788daeb6014a0d275a9f3ea2f01549e2d99f8fa"}}