{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:OYGGK2WSTTK6K2Q4WINSBOJX7Y","short_pith_number":"pith:OYGGK2WS","schema_version":"1.0","canonical_sha256":"760c656ad29cd5e56a1cb21b20b937fe21cee831a3c21366f6e9a8c471a65b6d","source":{"kind":"arxiv","id":"2505.23420","version":1},"attestation_state":"computed","paper":{"title":"The Warmup Dilemma: How Learning Rate Strategies Impact Speech-to-Text Model Convergence","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alessio Brutti, Luisa Bentivogli, Marco Gaido, Marco Matassoni, Matteo Negri, Mauro Cettolo, Mohamed Nabih, Roberto Gretter, Sara Papi","submitted_at":"2025-05-29T13:10:57Z","abstract_excerpt":"Training large-scale models presents challenges not only in terms of resource requirements but also in terms of their convergence. For this reason, the learning rate (LR) is often decreased when the size of a model is increased. Such a simple solution is not enough in the case of speech-to-text (S2T) trainings, where evolved and more complex variants of the Transformer architecture -- e.g., Conformer or Branchformer -- are used in light of their better performance. As a workaround, OWSM designed a double linear warmup of the LR, increasing it to a very small value in the first phase before upd"},"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":"2505.23420","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-29T13:10:57Z","cross_cats_sorted":[],"title_canon_sha256":"4fd24b2b5687d499911d6db0bee9b391b389cd312148943ef1722a2e2132acc7","abstract_canon_sha256":"6c9400e75be998582421932c37a7bef1476c46ee9457c8c3cc93979cf1a32db8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:12:01.482802Z","signature_b64":"h/x3R0pyvZRkjBW+T3s2+po2dRwDv6WKhkk3z0RjKGXPXS+5tH3eVHhgC8JhpldJ5apnhHGX6mk0zOLDuBPwCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"760c656ad29cd5e56a1cb21b20b937fe21cee831a3c21366f6e9a8c471a65b6d","last_reissued_at":"2026-07-05T11:12:01.482327Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:12:01.482327Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"The Warmup Dilemma: How Learning Rate Strategies Impact Speech-to-Text Model Convergence","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alessio Brutti, Luisa Bentivogli, Marco Gaido, Marco Matassoni, Matteo Negri, Mauro Cettolo, Mohamed Nabih, Roberto Gretter, Sara Papi","submitted_at":"2025-05-29T13:10:57Z","abstract_excerpt":"Training large-scale models presents challenges not only in terms of resource requirements but also in terms of their convergence. For this reason, the learning rate (LR) is often decreased when the size of a model is increased. Such a simple solution is not enough in the case of speech-to-text (S2T) trainings, where evolved and more complex variants of the Transformer architecture -- e.g., Conformer or Branchformer -- are used in light of their better performance. As a workaround, OWSM designed a double linear warmup of the LR, increasing it to a very small value in the first phase before upd"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.23420","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/2505.23420/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":"2505.23420","created_at":"2026-07-05T11:12:01.482389+00:00"},{"alias_kind":"arxiv_version","alias_value":"2505.23420v1","created_at":"2026-07-05T11:12:01.482389+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.23420","created_at":"2026-07-05T11:12:01.482389+00:00"},{"alias_kind":"pith_short_12","alias_value":"OYGGK2WSTTK6","created_at":"2026-07-05T11:12:01.482389+00:00"},{"alias_kind":"pith_short_16","alias_value":"OYGGK2WSTTK6K2Q4","created_at":"2026-07-05T11:12:01.482389+00:00"},{"alias_kind":"pith_short_8","alias_value":"OYGGK2WS","created_at":"2026-07-05T11:12:01.482389+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/OYGGK2WSTTK6K2Q4WINSBOJX7Y","json":"https://pith.science/pith/OYGGK2WSTTK6K2Q4WINSBOJX7Y.json","graph_json":"https://pith.science/api/pith-number/OYGGK2WSTTK6K2Q4WINSBOJX7Y/graph.json","events_json":"https://pith.science/api/pith-number/OYGGK2WSTTK6K2Q4WINSBOJX7Y/events.json","paper":"https://pith.science/paper/OYGGK2WS"},"agent_actions":{"view_html":"https://pith.science/pith/OYGGK2WSTTK6K2Q4WINSBOJX7Y","download_json":"https://pith.science/pith/OYGGK2WSTTK6K2Q4WINSBOJX7Y.json","view_paper":"https://pith.science/paper/OYGGK2WS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2505.23420&json=true","fetch_graph":"https://pith.science/api/pith-number/OYGGK2WSTTK6K2Q4WINSBOJX7Y/graph.json","fetch_events":"https://pith.science/api/pith-number/OYGGK2WSTTK6K2Q4WINSBOJX7Y/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OYGGK2WSTTK6K2Q4WINSBOJX7Y/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OYGGK2WSTTK6K2Q4WINSBOJX7Y/action/storage_attestation","attest_author":"https://pith.science/pith/OYGGK2WSTTK6K2Q4WINSBOJX7Y/action/author_attestation","sign_citation":"https://pith.science/pith/OYGGK2WSTTK6K2Q4WINSBOJX7Y/action/citation_signature","submit_replication":"https://pith.science/pith/OYGGK2WSTTK6K2Q4WINSBOJX7Y/action/replication_record"}},"created_at":"2026-07-05T11:12:01.482389+00:00","updated_at":"2026-07-05T11:12:01.482389+00:00"}