{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:ULXCJOW2J7UWS3CCWCVUCKDDKA","short_pith_number":"pith:ULXCJOW2","schema_version":"1.0","canonical_sha256":"a2ee24bada4fe9696c42b0ab41286350348b6c9d1f9b8a153828d1abda93bab2","source":{"kind":"arxiv","id":"2605.17821","version":1},"attestation_state":"computed","paper":{"title":"TierCheck: Tiered Checkpointing for Fault Tolerance in Large Language Model Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Feng Jiang, Lichen Pan, Nannan Zhao, Patrick P. C. Lee, Shujie Han, Xiaonan Zhao, Xiao Zhang, Zhijie Huang","submitted_at":"2026-05-18T03:48:32Z","abstract_excerpt":"Large Language Model (LLM) training is frequently interrupted by a heterogeneous spectrum of failures, from common GPU crashes to catastrophic cluster-wide outages. Existing checkpointing systems rely on monolithic, single-tier storage backend, forcing a trade-off between state-saving overhead and recovery speed. We propose TierCheck, a cluster-aware tiered checkpointing system that aligns storage placement with failure heterogeneity. TierCheck adopts a three-tier design that maintains lightweight differential checkpoints in local and peer memory for fast localized recovery, while asynchronous"},"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":"2605.17821","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2026-05-18T03:48:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"13878dc0b22a52c2150ee2a233d12231182f864959efa0456fed760d5d42ced4","abstract_canon_sha256":"8bbafb845cafdf28755ea30d50431ea73cadd9dc64d3ed70bbeb2c448b98b23b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T00:05:00.049934Z","signature_b64":"rTVuEUspuNwnAIBNHEPi8Z3QFCV1m2CTuzzyfBN7vrG/sj72Ci7bkXuGg1Gq0wCkU4tU1iXkXtGh1xlgH8zwBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a2ee24bada4fe9696c42b0ab41286350348b6c9d1f9b8a153828d1abda93bab2","last_reissued_at":"2026-05-20T00:05:00.049075Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T00:05:00.049075Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TierCheck: Tiered Checkpointing for Fault Tolerance in Large Language Model Training","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.DC","authors_text":"Feng Jiang, Lichen Pan, Nannan Zhao, Patrick P. C. Lee, Shujie Han, Xiaonan Zhao, Xiao Zhang, Zhijie Huang","submitted_at":"2026-05-18T03:48:32Z","abstract_excerpt":"Large Language Model (LLM) training is frequently interrupted by a heterogeneous spectrum of failures, from common GPU crashes to catastrophic cluster-wide outages. Existing checkpointing systems rely on monolithic, single-tier storage backend, forcing a trade-off between state-saving overhead and recovery speed. We propose TierCheck, a cluster-aware tiered checkpointing system that aligns storage placement with failure heterogeneity. TierCheck adopts a three-tier design that maintains lightweight differential checkpoints in local and peer memory for fast localized recovery, while asynchronous"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.17821","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/2605.17821/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":"2605.17821","created_at":"2026-05-20T00:05:00.049228+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.17821v1","created_at":"2026-05-20T00:05:00.049228+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.17821","created_at":"2026-05-20T00:05:00.049228+00:00"},{"alias_kind":"pith_short_12","alias_value":"ULXCJOW2J7UW","created_at":"2026-05-20T00:05:00.049228+00:00"},{"alias_kind":"pith_short_16","alias_value":"ULXCJOW2J7UWS3CC","created_at":"2026-05-20T00:05:00.049228+00:00"},{"alias_kind":"pith_short_8","alias_value":"ULXCJOW2","created_at":"2026-05-20T00:05:00.049228+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA","json":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA.json","graph_json":"https://pith.science/api/pith-number/ULXCJOW2J7UWS3CCWCVUCKDDKA/graph.json","events_json":"https://pith.science/api/pith-number/ULXCJOW2J7UWS3CCWCVUCKDDKA/events.json","paper":"https://pith.science/paper/ULXCJOW2"},"agent_actions":{"view_html":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA","download_json":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA.json","view_paper":"https://pith.science/paper/ULXCJOW2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.17821&json=true","fetch_graph":"https://pith.science/api/pith-number/ULXCJOW2J7UWS3CCWCVUCKDDKA/graph.json","fetch_events":"https://pith.science/api/pith-number/ULXCJOW2J7UWS3CCWCVUCKDDKA/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA/action/storage_attestation","attest_author":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA/action/author_attestation","sign_citation":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA/action/citation_signature","submit_replication":"https://pith.science/pith/ULXCJOW2J7UWS3CCWCVUCKDDKA/action/replication_record"}},"created_at":"2026-05-20T00:05:00.049228+00:00","updated_at":"2026-05-20T00:05:00.049228+00:00"}