{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:YSNBEHWPVDARQMELMELXVBKYVG","short_pith_number":"pith:YSNBEHWP","schema_version":"1.0","canonical_sha256":"c49a121ecfa8c118308b61177a8558a9829c3c828f96f01d39abf381c20e384c","source":{"kind":"arxiv","id":"1912.12384","version":1},"attestation_state":"computed","paper":{"title":"Improved Multi-Stage Training of Online Attention-based Encoder-Decoder Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD","eess.SP","stat.ML"],"primary_cat":"eess.AS","authors_text":"Abhinav Garg, Ankur Kumar, Chanwoo Kim, Dhananjaya Gowda, Kwangyoun Kim, Mehul Kumar","submitted_at":"2019-12-28T02:29:33Z","abstract_excerpt":"In this paper, we propose a refined multi-stage multi-task training strategy to improve the performance of online attention-based encoder-decoder (AED) models. A three-stage training based on three levels of architectural granularity namely, character encoder, byte pair encoding (BPE) based encoder, and attention decoder, is proposed. Also, multi-task learning based on two-levels of linguistic granularity namely, character and BPE, is used. We explore different pre-training strategies for the encoders including transfer learning from a bidirectional encoder. Our encoder-decoder models with onl"},"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":"1912.12384","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"eess.AS","submitted_at":"2019-12-28T02:29:33Z","cross_cats_sorted":["cs.LG","cs.SD","eess.SP","stat.ML"],"title_canon_sha256":"203f205b96e4dd1be12191d3e1caecc98f03b1be0613e5315947bcd15d3e2d23","abstract_canon_sha256":"891e64196b422dda604237f6317496c6e25c39e10d08dfb1f88f00775493ec38"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:28:59.476487Z","signature_b64":"BM7VN7ILn8+XVNW1/OBfk6beVpnvrWe3IIZxxwB8OdReKPDiDsMrD7LtGS79mP2IQrovQB/RfdwElk3i6002BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c49a121ecfa8c118308b61177a8558a9829c3c828f96f01d39abf381c20e384c","last_reissued_at":"2026-07-05T00:28:59.476063Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:28:59.476063Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Improved Multi-Stage Training of Online Attention-based Encoder-Decoder Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.SD","eess.SP","stat.ML"],"primary_cat":"eess.AS","authors_text":"Abhinav Garg, Ankur Kumar, Chanwoo Kim, Dhananjaya Gowda, Kwangyoun Kim, Mehul Kumar","submitted_at":"2019-12-28T02:29:33Z","abstract_excerpt":"In this paper, we propose a refined multi-stage multi-task training strategy to improve the performance of online attention-based encoder-decoder (AED) models. A three-stage training based on three levels of architectural granularity namely, character encoder, byte pair encoding (BPE) based encoder, and attention decoder, is proposed. Also, multi-task learning based on two-levels of linguistic granularity namely, character and BPE, is used. We explore different pre-training strategies for the encoders including transfer learning from a bidirectional encoder. Our encoder-decoder models with onl"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1912.12384","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/1912.12384/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":"1912.12384","created_at":"2026-07-05T00:28:59.476124+00:00"},{"alias_kind":"arxiv_version","alias_value":"1912.12384v1","created_at":"2026-07-05T00:28:59.476124+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1912.12384","created_at":"2026-07-05T00:28:59.476124+00:00"},{"alias_kind":"pith_short_12","alias_value":"YSNBEHWPVDAR","created_at":"2026-07-05T00:28:59.476124+00:00"},{"alias_kind":"pith_short_16","alias_value":"YSNBEHWPVDARQMEL","created_at":"2026-07-05T00:28:59.476124+00:00"},{"alias_kind":"pith_short_8","alias_value":"YSNBEHWP","created_at":"2026-07-05T00:28:59.476124+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/YSNBEHWPVDARQMELMELXVBKYVG","json":"https://pith.science/pith/YSNBEHWPVDARQMELMELXVBKYVG.json","graph_json":"https://pith.science/api/pith-number/YSNBEHWPVDARQMELMELXVBKYVG/graph.json","events_json":"https://pith.science/api/pith-number/YSNBEHWPVDARQMELMELXVBKYVG/events.json","paper":"https://pith.science/paper/YSNBEHWP"},"agent_actions":{"view_html":"https://pith.science/pith/YSNBEHWPVDARQMELMELXVBKYVG","download_json":"https://pith.science/pith/YSNBEHWPVDARQMELMELXVBKYVG.json","view_paper":"https://pith.science/paper/YSNBEHWP","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1912.12384&json=true","fetch_graph":"https://pith.science/api/pith-number/YSNBEHWPVDARQMELMELXVBKYVG/graph.json","fetch_events":"https://pith.science/api/pith-number/YSNBEHWPVDARQMELMELXVBKYVG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YSNBEHWPVDARQMELMELXVBKYVG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YSNBEHWPVDARQMELMELXVBKYVG/action/storage_attestation","attest_author":"https://pith.science/pith/YSNBEHWPVDARQMELMELXVBKYVG/action/author_attestation","sign_citation":"https://pith.science/pith/YSNBEHWPVDARQMELMELXVBKYVG/action/citation_signature","submit_replication":"https://pith.science/pith/YSNBEHWPVDARQMELMELXVBKYVG/action/replication_record"}},"created_at":"2026-07-05T00:28:59.476124+00:00","updated_at":"2026-07-05T00:28:59.476124+00:00"}