{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:2RMOVIVNFZIVG4WD42BFY2PYS7","short_pith_number":"pith:2RMOVIVN","schema_version":"1.0","canonical_sha256":"d458eaa2ad2e515372c3e6825c69f897fc508ab5ebbe7a8527cce908514e0f23","source":{"kind":"arxiv","id":"2312.08361","version":1},"attestation_state":"computed","paper":{"title":"Distributed Inference and Fine-tuning of Large Language Models Over The Internet","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Alexander Borzunov, Artem Chumachenko, Colin Raffel, Dmitry Baranchuk, Max Ryabinin, Pavel Samygin, Tim Dettmers, Younes Belkada","submitted_at":"2023-12-13T18:52:49Z","abstract_excerpt":"Large language models (LLMs) are useful in many NLP tasks and become more capable with size, with the best open-source models having over 50 billion parameters. However, using these 50B+ models requires high-end hardware, making them inaccessible to most researchers. In this work, we investigate methods for cost-efficient inference and fine-tuning of LLMs, comparing local and distributed strategies. We observe that a large enough model (50B+) can run efficiently even on geodistributed devices in a consumer-grade network. This could allow running LLM efficiently by pooling together idle compute"},"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":"2312.08361","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-12-13T18:52:49Z","cross_cats_sorted":["cs.DC"],"title_canon_sha256":"9eb74edb681f1e393fe9208eacc38ac0ba24330110b9b4c4ecbe6ccda9ade3c8","abstract_canon_sha256":"c73e0c4f6da2c4fc74da4fce2b930a109058869a6ba378cc475684a290b1a732"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:23:52.391324Z","signature_b64":"TneUbOV6rAtoWNOL+uJIUTyTrlwAH8owU2LMTuRi26UvkT7/caFnGgo7B8W2nrndZ24/DhhZ7/FTTznb6dNEAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d458eaa2ad2e515372c3e6825c69f897fc508ab5ebbe7a8527cce908514e0f23","last_reissued_at":"2026-07-05T07:23:52.390876Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:23:52.390876Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Distributed Inference and Fine-tuning of Large Language Models Over The Internet","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC"],"primary_cat":"cs.LG","authors_text":"Alexander Borzunov, Artem Chumachenko, Colin Raffel, Dmitry Baranchuk, Max Ryabinin, Pavel Samygin, Tim Dettmers, Younes Belkada","submitted_at":"2023-12-13T18:52:49Z","abstract_excerpt":"Large language models (LLMs) are useful in many NLP tasks and become more capable with size, with the best open-source models having over 50 billion parameters. However, using these 50B+ models requires high-end hardware, making them inaccessible to most researchers. In this work, we investigate methods for cost-efficient inference and fine-tuning of LLMs, comparing local and distributed strategies. We observe that a large enough model (50B+) can run efficiently even on geodistributed devices in a consumer-grade network. This could allow running LLM efficiently by pooling together idle compute"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.08361","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/2312.08361/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":"2312.08361","created_at":"2026-07-05T07:23:52.390942+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.08361v1","created_at":"2026-07-05T07:23:52.390942+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.08361","created_at":"2026-07-05T07:23:52.390942+00:00"},{"alias_kind":"pith_short_12","alias_value":"2RMOVIVNFZIV","created_at":"2026-07-05T07:23:52.390942+00:00"},{"alias_kind":"pith_short_16","alias_value":"2RMOVIVNFZIVG4WD","created_at":"2026-07-05T07:23:52.390942+00:00"},{"alias_kind":"pith_short_8","alias_value":"2RMOVIVN","created_at":"2026-07-05T07:23:52.390942+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/2RMOVIVNFZIVG4WD42BFY2PYS7","json":"https://pith.science/pith/2RMOVIVNFZIVG4WD42BFY2PYS7.json","graph_json":"https://pith.science/api/pith-number/2RMOVIVNFZIVG4WD42BFY2PYS7/graph.json","events_json":"https://pith.science/api/pith-number/2RMOVIVNFZIVG4WD42BFY2PYS7/events.json","paper":"https://pith.science/paper/2RMOVIVN"},"agent_actions":{"view_html":"https://pith.science/pith/2RMOVIVNFZIVG4WD42BFY2PYS7","download_json":"https://pith.science/pith/2RMOVIVNFZIVG4WD42BFY2PYS7.json","view_paper":"https://pith.science/paper/2RMOVIVN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.08361&json=true","fetch_graph":"https://pith.science/api/pith-number/2RMOVIVNFZIVG4WD42BFY2PYS7/graph.json","fetch_events":"https://pith.science/api/pith-number/2RMOVIVNFZIVG4WD42BFY2PYS7/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2RMOVIVNFZIVG4WD42BFY2PYS7/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2RMOVIVNFZIVG4WD42BFY2PYS7/action/storage_attestation","attest_author":"https://pith.science/pith/2RMOVIVNFZIVG4WD42BFY2PYS7/action/author_attestation","sign_citation":"https://pith.science/pith/2RMOVIVNFZIVG4WD42BFY2PYS7/action/citation_signature","submit_replication":"https://pith.science/pith/2RMOVIVNFZIVG4WD42BFY2PYS7/action/replication_record"}},"created_at":"2026-07-05T07:23:52.390942+00:00","updated_at":"2026-07-05T07:23:52.390942+00:00"}