{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:VL3FEATYQWIG74IAV3UTJQKVE7","short_pith_number":"pith:VL3FEATY","schema_version":"1.0","canonical_sha256":"aaf652027885906ff100aee934c15527f16cb47b277c6d33eebcbab6dd644b15","source":{"kind":"arxiv","id":"2306.11800","version":3},"attestation_state":"computed","paper":{"title":"Inshrinkerator: Compressing Deep Learning Training Checkpoints via Dynamic Quantization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexey Tumanov, Amey Agrawal, Kexin Rong, Sameer Reddy, Satwik Bhattamishra, Venkata Prabhakara Sarath Nookala, Vidushi Vashishth","submitted_at":"2023-06-20T18:00:31Z","abstract_excerpt":"With the increase in the scale of Deep Learning (DL) training workloads in terms of compute resources and time consumption, the likelihood of encountering in-training failures rises substantially, leading to lost work and resource wastage. Such failures are typically offset by a checkpointing mechanism, which comes at the cost of storage and network bandwidth overhead. State-of-the-art approaches involve lossy model compression mechanisms, which induce a tradeoff between the resulting model quality (accuracy) and compression ratio. Delta compression is then used to further reduce the overhead "},"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":"2306.11800","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-06-20T18:00:31Z","cross_cats_sorted":[],"title_canon_sha256":"43f88bb9e62cd1943fbbfc490abd627476a4111f7934e37911165e71a1ef1767","abstract_canon_sha256":"a56fd5ed2e16c107e1779016b905866c3b4108abf82c86a9e3313fb8735a12ba"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:19:23.500844Z","signature_b64":"7jTHS1H3+ZuwWpek3DchDXEpFWQxtcvr3zlAoCnl9/b2Cyx/7L9WfS++9R+oTagquoPySKu9WUQJEbljNeMEBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aaf652027885906ff100aee934c15527f16cb47b277c6d33eebcbab6dd644b15","last_reissued_at":"2026-07-05T09:19:23.500350Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:19:23.500350Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Inshrinkerator: Compressing Deep Learning Training Checkpoints via Dynamic Quantization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Alexey Tumanov, Amey Agrawal, Kexin Rong, Sameer Reddy, Satwik Bhattamishra, Venkata Prabhakara Sarath Nookala, Vidushi Vashishth","submitted_at":"2023-06-20T18:00:31Z","abstract_excerpt":"With the increase in the scale of Deep Learning (DL) training workloads in terms of compute resources and time consumption, the likelihood of encountering in-training failures rises substantially, leading to lost work and resource wastage. Such failures are typically offset by a checkpointing mechanism, which comes at the cost of storage and network bandwidth overhead. State-of-the-art approaches involve lossy model compression mechanisms, which induce a tradeoff between the resulting model quality (accuracy) and compression ratio. Delta compression is then used to further reduce the overhead "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.11800","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/2306.11800/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":"2306.11800","created_at":"2026-07-05T09:19:23.500409+00:00"},{"alias_kind":"arxiv_version","alias_value":"2306.11800v3","created_at":"2026-07-05T09:19:23.500409+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.11800","created_at":"2026-07-05T09:19:23.500409+00:00"},{"alias_kind":"pith_short_12","alias_value":"VL3FEATYQWIG","created_at":"2026-07-05T09:19:23.500409+00:00"},{"alias_kind":"pith_short_16","alias_value":"VL3FEATYQWIG74IA","created_at":"2026-07-05T09:19:23.500409+00:00"},{"alias_kind":"pith_short_8","alias_value":"VL3FEATY","created_at":"2026-07-05T09:19:23.500409+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.12000","citing_title":"An Efficient Compression of Deep Neural Network Checkpoints Based on Prediction and Context Modeling","ref_index":9,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7","json":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7.json","graph_json":"https://pith.science/api/pith-number/VL3FEATYQWIG74IAV3UTJQKVE7/graph.json","events_json":"https://pith.science/api/pith-number/VL3FEATYQWIG74IAV3UTJQKVE7/events.json","paper":"https://pith.science/paper/VL3FEATY"},"agent_actions":{"view_html":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7","download_json":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7.json","view_paper":"https://pith.science/paper/VL3FEATY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2306.11800&json=true","fetch_graph":"https://pith.science/api/pith-number/VL3FEATYQWIG74IAV3UTJQKVE7/graph.json","fetch_events":"https://pith.science/api/pith-number/VL3FEATYQWIG74IAV3UTJQKVE7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7/action/storage_attestation","attest_author":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7/action/author_attestation","sign_citation":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7/action/citation_signature","submit_replication":"https://pith.science/pith/VL3FEATYQWIG74IAV3UTJQKVE7/action/replication_record"}},"created_at":"2026-07-05T09:19:23.500409+00:00","updated_at":"2026-07-05T09:19:23.500409+00:00"}