{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:HNXIJQYP7RXNCOM7OTLMPLTA3G","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"b8991906585dbaf8b70cdbde3e3b4eca76b4e8fb464afeb1c50bfbeb07f20e80","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2023-10-28T00:33:37Z","title_canon_sha256":"c27bf50003626de9bb384186a1f061512b45c11b89137022581f2f9b085fd67b"},"schema_version":"1.0","source":{"id":"2310.18547","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2310.18547","created_at":"2026-07-05T07:06:09Z"},{"alias_kind":"arxiv_version","alias_value":"2310.18547v1","created_at":"2026-07-05T07:06:09Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.18547","created_at":"2026-07-05T07:06:09Z"},{"alias_kind":"pith_short_12","alias_value":"HNXIJQYP7RXN","created_at":"2026-07-05T07:06:09Z"},{"alias_kind":"pith_short_16","alias_value":"HNXIJQYP7RXNCOM7","created_at":"2026-07-05T07:06:09Z"},{"alias_kind":"pith_short_8","alias_value":"HNXIJQYP","created_at":"2026-07-05T07:06:09Z"}],"graph_snapshots":[{"event_id":"sha256:85b3ae8cb1c88acec54f9fac5124265e3f3fd92cd75d1b2e014bdad9391eed98","target":"graph","created_at":"2026-07-05T07:06:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2310.18547/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Low-rank adaptation (LoRA) has become an important and popular method to adapt pre-trained models to specific domains. We present Punica, a system to serve multiple LoRA models in a shared GPU cluster. Punica contains a new CUDA kernel design that allows batching of GPU operations for different LoRA models. This allows a GPU to hold only a single copy of the underlying pre-trained model when serving multiple, different LoRA models, significantly enhancing GPU efficiency in terms of both memory and computation. Our scheduler consolidates multi-tenant LoRA serving workloads in a shared GPU clust","authors_text":"(2) Duke University), Arvind Krishnamurthy (1) ((1) University of Washington, Danyang Zhuo (2), Lequn Chen (1), Luis Ceze (1), Yongji Wu (2), Zihao Ye (1)","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2023-10-28T00:33:37Z","title":"Punica: Multi-Tenant LoRA Serving"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.18547","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:e155d9756d9bc616f319bcfd2e012b4bbba09af00434efdd37d2f28c4b9996ac","target":"record","created_at":"2026-07-05T07:06:09Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"b8991906585dbaf8b70cdbde3e3b4eca76b4e8fb464afeb1c50bfbeb07f20e80","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.DC","submitted_at":"2023-10-28T00:33:37Z","title_canon_sha256":"c27bf50003626de9bb384186a1f061512b45c11b89137022581f2f9b085fd67b"},"schema_version":"1.0","source":{"id":"2310.18547","kind":"arxiv","version":1}},"canonical_sha256":"3b6e84c30ffc6ed1399f74d6c7ae60d9b4ea8acbad3ba8e327eb38db4127f34f","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3b6e84c30ffc6ed1399f74d6c7ae60d9b4ea8acbad3ba8e327eb38db4127f34f","first_computed_at":"2026-07-05T07:06:09.949192Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:06:09.949192Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Uw3GsDuUChM6u108QXNw+t1n/jdxFZJkn5TVDyAUTnBPljZsG4pAy/JY32uXXBj4H0Kfh6Sd15O5/uUQHUrkBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T07:06:09.949748Z","signed_message":"canonical_sha256_bytes"},"source_id":"2310.18547","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:e155d9756d9bc616f319bcfd2e012b4bbba09af00434efdd37d2f28c4b9996ac","sha256:85b3ae8cb1c88acec54f9fac5124265e3f3fd92cd75d1b2e014bdad9391eed98"],"state_sha256":"f6ac9e8dc5a2bcaeb38f31370bc8d44690018f0872572315f4936764e2297222"}