{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:S6WUQ2UUUKVUWNQEWYMMMEU56C","short_pith_number":"pith:S6WUQ2UU","schema_version":"1.0","canonical_sha256":"97ad486a94a2ab4b3604b618c6129df08ce837c485099fe8f8a9092f821c75ba","source":{"kind":"arxiv","id":"2401.02088","version":1},"attestation_state":"computed","paper":{"title":"Re-evaluating the Memory-balanced Pipeline Parallelism: BPipe","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.DC"],"primary_cat":"cs.LG","authors_text":"Chao Wang, Chi Ma, Lei Yu, Mincong Huang, Peng Zhang, Yineng Zhang","submitted_at":"2024-01-04T06:23:22Z","abstract_excerpt":"Pipeline parallelism is an essential technique in the training of large-scale Transformer models. However, it suffers from imbalanced memory consumption, leading to insufficient memory utilization. The BPipe technique was proposed to address this issue and has proven effective in the GPT-3 model. Nevertheless, our experiments have not yielded similar benefits for LLaMA training. Additionally, BPipe only yields negligible benefits for GPT-3 training when applying flash attention. We analyze the underlying causes of the divergent performance of BPipe on GPT-3 and LLaMA. Furthermore, we introduce"},"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":"2401.02088","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-01-04T06:23:22Z","cross_cats_sorted":["cs.CL","cs.DC"],"title_canon_sha256":"ac183a68c4b4ca47b7df2930538e003f25ca09b3608fe48de5564e58257f6b37","abstract_canon_sha256":"f0f408fe17e576d376818d8b423a9e411524e1840d829e4d2dd131da994d8aa4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:30:10.861789Z","signature_b64":"HWEYdWtK2ccJJ+vLXIqV2FgcPmdilPZcjFpyIjK1Q0nQaPciCdHpKcS/5Hq0kGs09RvPCWYc+BCPiBa7XRX1CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"97ad486a94a2ab4b3604b618c6129df08ce837c485099fe8f8a9092f821c75ba","last_reissued_at":"2026-07-05T07:30:10.861267Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:30:10.861267Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Re-evaluating the Memory-balanced Pipeline Parallelism: BPipe","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL","cs.DC"],"primary_cat":"cs.LG","authors_text":"Chao Wang, Chi Ma, Lei Yu, Mincong Huang, Peng Zhang, Yineng Zhang","submitted_at":"2024-01-04T06:23:22Z","abstract_excerpt":"Pipeline parallelism is an essential technique in the training of large-scale Transformer models. However, it suffers from imbalanced memory consumption, leading to insufficient memory utilization. The BPipe technique was proposed to address this issue and has proven effective in the GPT-3 model. Nevertheless, our experiments have not yielded similar benefits for LLaMA training. Additionally, BPipe only yields negligible benefits for GPT-3 training when applying flash attention. We analyze the underlying causes of the divergent performance of BPipe on GPT-3 and LLaMA. Furthermore, we introduce"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.02088","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/2401.02088/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":"2401.02088","created_at":"2026-07-05T07:30:10.861332+00:00"},{"alias_kind":"arxiv_version","alias_value":"2401.02088v1","created_at":"2026-07-05T07:30:10.861332+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.02088","created_at":"2026-07-05T07:30:10.861332+00:00"},{"alias_kind":"pith_short_12","alias_value":"S6WUQ2UUUKVU","created_at":"2026-07-05T07:30:10.861332+00:00"},{"alias_kind":"pith_short_16","alias_value":"S6WUQ2UUUKVUWNQE","created_at":"2026-07-05T07:30:10.861332+00:00"},{"alias_kind":"pith_short_8","alias_value":"S6WUQ2UU","created_at":"2026-07-05T07:30:10.861332+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2512.16776","citing_title":"Kling-Omni Technical Report","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C","json":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C.json","graph_json":"https://pith.science/api/pith-number/S6WUQ2UUUKVUWNQEWYMMMEU56C/graph.json","events_json":"https://pith.science/api/pith-number/S6WUQ2UUUKVUWNQEWYMMMEU56C/events.json","paper":"https://pith.science/paper/S6WUQ2UU"},"agent_actions":{"view_html":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C","download_json":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C.json","view_paper":"https://pith.science/paper/S6WUQ2UU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2401.02088&json=true","fetch_graph":"https://pith.science/api/pith-number/S6WUQ2UUUKVUWNQEWYMMMEU56C/graph.json","fetch_events":"https://pith.science/api/pith-number/S6WUQ2UUUKVUWNQEWYMMMEU56C/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C/action/timestamp_anchor","attest_storage":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C/action/storage_attestation","attest_author":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C/action/author_attestation","sign_citation":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C/action/citation_signature","submit_replication":"https://pith.science/pith/S6WUQ2UUUKVUWNQEWYMMMEU56C/action/replication_record"}},"created_at":"2026-07-05T07:30:10.861332+00:00","updated_at":"2026-07-05T07:30:10.861332+00:00"}