{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:KYYWFCTZUQNFW2Q57KOSVEI7WL","short_pith_number":"pith:KYYWFCTZ","schema_version":"1.0","canonical_sha256":"5631628a79a41a5b6a1dfa9d2a911fb2de9835f4e8e37c1508e7023196635e3d","source":{"kind":"arxiv","id":"2003.05649","version":1},"attestation_state":"computed","paper":{"title":"Machine Learning on Volatile Instances","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Carlee Joe-Wong, Gauri Joshi, Jianyu Wang, Xiaoxi Zhang","submitted_at":"2020-03-12T07:47:34Z","abstract_excerpt":"Due to the massive size of the neural network models and training datasets used in machine learning today, it is imperative to distribute stochastic gradient descent (SGD) by splitting up tasks such as gradient evaluation across multiple worker nodes. However, running distributed SGD can be prohibitively expensive because it may require specialized computing resources such as GPUs for extended periods of time. We propose cost-effective strategies to exploit volatile cloud instances that are cheaper than standard instances, but may be interrupted by higher priority workloads. To the best of our"},"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":"2003.05649","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-03-12T07:47:34Z","cross_cats_sorted":["cs.DC","stat.ML"],"title_canon_sha256":"ccd1318eaa150299bd6803ced1414359f44632d15c204f35ed89d428c382e41d","abstract_canon_sha256":"4a76d0777b4d2d8b569d5749008d54d7473c97af6d9c72431adc45a9d459e046"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:47:32.713950Z","signature_b64":"TfbTqR5ASJbTWpdZAArrxs8/C6cLUMsryM3sbQJG3OqaVBHWvaAeeI39o+SW4Fi4b0GHjqGCo10vke4UuiZxDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5631628a79a41a5b6a1dfa9d2a911fb2de9835f4e8e37c1508e7023196635e3d","last_reissued_at":"2026-07-05T00:47:32.713330Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:47:32.713330Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Machine Learning on Volatile Instances","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Carlee Joe-Wong, Gauri Joshi, Jianyu Wang, Xiaoxi Zhang","submitted_at":"2020-03-12T07:47:34Z","abstract_excerpt":"Due to the massive size of the neural network models and training datasets used in machine learning today, it is imperative to distribute stochastic gradient descent (SGD) by splitting up tasks such as gradient evaluation across multiple worker nodes. However, running distributed SGD can be prohibitively expensive because it may require specialized computing resources such as GPUs for extended periods of time. We propose cost-effective strategies to exploit volatile cloud instances that are cheaper than standard instances, but may be interrupted by higher priority workloads. To the best of our"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2003.05649","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/2003.05649/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":"2003.05649","created_at":"2026-07-05T00:47:32.713405+00:00"},{"alias_kind":"arxiv_version","alias_value":"2003.05649v1","created_at":"2026-07-05T00:47:32.713405+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2003.05649","created_at":"2026-07-05T00:47:32.713405+00:00"},{"alias_kind":"pith_short_12","alias_value":"KYYWFCTZUQNF","created_at":"2026-07-05T00:47:32.713405+00:00"},{"alias_kind":"pith_short_16","alias_value":"KYYWFCTZUQNFW2Q5","created_at":"2026-07-05T00:47:32.713405+00:00"},{"alias_kind":"pith_short_8","alias_value":"KYYWFCTZ","created_at":"2026-07-05T00:47:32.713405+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/KYYWFCTZUQNFW2Q57KOSVEI7WL","json":"https://pith.science/pith/KYYWFCTZUQNFW2Q57KOSVEI7WL.json","graph_json":"https://pith.science/api/pith-number/KYYWFCTZUQNFW2Q57KOSVEI7WL/graph.json","events_json":"https://pith.science/api/pith-number/KYYWFCTZUQNFW2Q57KOSVEI7WL/events.json","paper":"https://pith.science/paper/KYYWFCTZ"},"agent_actions":{"view_html":"https://pith.science/pith/KYYWFCTZUQNFW2Q57KOSVEI7WL","download_json":"https://pith.science/pith/KYYWFCTZUQNFW2Q57KOSVEI7WL.json","view_paper":"https://pith.science/paper/KYYWFCTZ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2003.05649&json=true","fetch_graph":"https://pith.science/api/pith-number/KYYWFCTZUQNFW2Q57KOSVEI7WL/graph.json","fetch_events":"https://pith.science/api/pith-number/KYYWFCTZUQNFW2Q57KOSVEI7WL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KYYWFCTZUQNFW2Q57KOSVEI7WL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KYYWFCTZUQNFW2Q57KOSVEI7WL/action/storage_attestation","attest_author":"https://pith.science/pith/KYYWFCTZUQNFW2Q57KOSVEI7WL/action/author_attestation","sign_citation":"https://pith.science/pith/KYYWFCTZUQNFW2Q57KOSVEI7WL/action/citation_signature","submit_replication":"https://pith.science/pith/KYYWFCTZUQNFW2Q57KOSVEI7WL/action/replication_record"}},"created_at":"2026-07-05T00:47:32.713405+00:00","updated_at":"2026-07-05T00:47:32.713405+00:00"}