{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:SNKAEL6QO77P4ELEIZMMAU5PLG","short_pith_number":"pith:SNKAEL6Q","schema_version":"1.0","canonical_sha256":"9354022fd077fefe11644658c053af598c82817cdd7684e5abbbab2d5c5e7e3e","source":{"kind":"arxiv","id":"2501.06663","version":2},"attestation_state":"computed","paper":{"title":"Ultra Memory-Efficient On-FPGA Training of Transformers via Tensor-Compressed Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AR","cs.CL"],"primary_cat":"cs.LG","authors_text":"Cong Hao, Hai Li, Ian Young, Jiayi Tian, Jinming Lu, Xiangwei Wang, Zheng Zhang","submitted_at":"2025-01-11T23:29:51Z","abstract_excerpt":"Transformer models have achieved state-of-the-art performance across a wide range of machine learning tasks. There is growing interest in training transformers on resource-constrained edge devices due to considerations such as privacy, domain adaptation, and on-device scientific machine learning. However, the significant computational and memory demands required for transformer training often exceed the capabilities of an edge device. Leveraging low-rank tensor compression, this paper presents the first on-FPGA accelerator for end-to-end transformer training. On the algorithm side, we present "},"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":"2501.06663","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-11T23:29:51Z","cross_cats_sorted":["cs.AR","cs.CL"],"title_canon_sha256":"ffbc6827cce21509100ee2240ceb2931ced38b5a85ca8747c10797deed4ff9bb","abstract_canon_sha256":"70bada34e6df6b7da7f434381c45e6062f5c784fe4bbe9b2f0df9a3a0b91faad"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:49:32.426542Z","signature_b64":"sBq9L+jU2woE0bDk/CrTnPxsck1wSSLYNYdBYjx/rS8NGPDj/v7M3dAcvFE3DFKOl6/QVPcQPemlaoWD9BD4Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9354022fd077fefe11644658c053af598c82817cdd7684e5abbbab2d5c5e7e3e","last_reissued_at":"2026-07-05T11:49:32.426080Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:49:32.426080Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Ultra Memory-Efficient On-FPGA Training of Transformers via Tensor-Compressed Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AR","cs.CL"],"primary_cat":"cs.LG","authors_text":"Cong Hao, Hai Li, Ian Young, Jiayi Tian, Jinming Lu, Xiangwei Wang, Zheng Zhang","submitted_at":"2025-01-11T23:29:51Z","abstract_excerpt":"Transformer models have achieved state-of-the-art performance across a wide range of machine learning tasks. There is growing interest in training transformers on resource-constrained edge devices due to considerations such as privacy, domain adaptation, and on-device scientific machine learning. However, the significant computational and memory demands required for transformer training often exceed the capabilities of an edge device. Leveraging low-rank tensor compression, this paper presents the first on-FPGA accelerator for end-to-end transformer training. On the algorithm side, we present "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.06663","kind":"arxiv","version":2},"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/2501.06663/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":"2501.06663","created_at":"2026-07-05T11:49:32.426135+00:00"},{"alias_kind":"arxiv_version","alias_value":"2501.06663v2","created_at":"2026-07-05T11:49:32.426135+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.06663","created_at":"2026-07-05T11:49:32.426135+00:00"},{"alias_kind":"pith_short_12","alias_value":"SNKAEL6QO77P","created_at":"2026-07-05T11:49:32.426135+00:00"},{"alias_kind":"pith_short_16","alias_value":"SNKAEL6QO77P4ELE","created_at":"2026-07-05T11:49:32.426135+00:00"},{"alias_kind":"pith_short_8","alias_value":"SNKAEL6Q","created_at":"2026-07-05T11:49:32.426135+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/SNKAEL6QO77P4ELEIZMMAU5PLG","json":"https://pith.science/pith/SNKAEL6QO77P4ELEIZMMAU5PLG.json","graph_json":"https://pith.science/api/pith-number/SNKAEL6QO77P4ELEIZMMAU5PLG/graph.json","events_json":"https://pith.science/api/pith-number/SNKAEL6QO77P4ELEIZMMAU5PLG/events.json","paper":"https://pith.science/paper/SNKAEL6Q"},"agent_actions":{"view_html":"https://pith.science/pith/SNKAEL6QO77P4ELEIZMMAU5PLG","download_json":"https://pith.science/pith/SNKAEL6QO77P4ELEIZMMAU5PLG.json","view_paper":"https://pith.science/paper/SNKAEL6Q","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2501.06663&json=true","fetch_graph":"https://pith.science/api/pith-number/SNKAEL6QO77P4ELEIZMMAU5PLG/graph.json","fetch_events":"https://pith.science/api/pith-number/SNKAEL6QO77P4ELEIZMMAU5PLG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SNKAEL6QO77P4ELEIZMMAU5PLG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SNKAEL6QO77P4ELEIZMMAU5PLG/action/storage_attestation","attest_author":"https://pith.science/pith/SNKAEL6QO77P4ELEIZMMAU5PLG/action/author_attestation","sign_citation":"https://pith.science/pith/SNKAEL6QO77P4ELEIZMMAU5PLG/action/citation_signature","submit_replication":"https://pith.science/pith/SNKAEL6QO77P4ELEIZMMAU5PLG/action/replication_record"}},"created_at":"2026-07-05T11:49:32.426135+00:00","updated_at":"2026-07-05T11:49:32.426135+00:00"}