{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:ARAEYOPYGHAQP4FWUYMKVFWGP3","short_pith_number":"pith:ARAEYOPY","schema_version":"1.0","canonical_sha256":"04404c39f831c107f0b6a618aa96c67ec54abd3dc39143d7bd02a65d4855959e","source":{"kind":"arxiv","id":"2506.10299","version":1},"attestation_state":"computed","paper":{"title":"Scheduled Interleaved Speech-Text Training for Speech-to-Speech Translation with LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Emiru Tsunoo, Hassan Shahmohammadi, Hayato Futami, Shinji Watanabe, Siddhant Arora, Yosuke Kashiwagi, Yuki Ito","submitted_at":"2025-06-12T02:24:44Z","abstract_excerpt":"Speech-to-speech translation (S2ST) has been advanced with large language models (LLMs), which are fine-tuned on discrete speech units. In such approaches, modality adaptation from text to speech has been an issue. LLMs are trained on text-only data, which presents challenges to adapt them to speech modality with limited speech-to-speech data. To address the training difficulty, we propose scheduled interleaved speech--text training in this study. We use interleaved speech--text units instead of speech units during training, where aligned text tokens are interleaved at the word level. We gradu"},"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":"2506.10299","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2025-06-12T02:24:44Z","cross_cats_sorted":["cs.SD","eess.AS"],"title_canon_sha256":"8b8aba6c2082a11d8ef657be6cf319f757d145de72b4805d768f24cf6ec76314","abstract_canon_sha256":"b3943bf4d00bfa70030b0a2345da19be67c118c649791fcbcb932ab7b7ef1f9e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:20:21.107016Z","signature_b64":"z5MjN0dP49IpaaGvDjh7NzBDE2zxkvyYp3uUdi/c4kiiO0dY9EQwOHsR/De4oeC3i1TpN6mE+1coUN1vbl2iDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"04404c39f831c107f0b6a618aa96c67ec54abd3dc39143d7bd02a65d4855959e","last_reissued_at":"2026-07-05T11:20:21.106424Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:20:21.106424Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Scheduled Interleaved Speech-Text Training for Speech-to-Speech Translation with LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.SD","eess.AS"],"primary_cat":"cs.CL","authors_text":"Emiru Tsunoo, Hassan Shahmohammadi, Hayato Futami, Shinji Watanabe, Siddhant Arora, Yosuke Kashiwagi, Yuki Ito","submitted_at":"2025-06-12T02:24:44Z","abstract_excerpt":"Speech-to-speech translation (S2ST) has been advanced with large language models (LLMs), which are fine-tuned on discrete speech units. In such approaches, modality adaptation from text to speech has been an issue. LLMs are trained on text-only data, which presents challenges to adapt them to speech modality with limited speech-to-speech data. To address the training difficulty, we propose scheduled interleaved speech--text training in this study. We use interleaved speech--text units instead of speech units during training, where aligned text tokens are interleaved at the word level. We gradu"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.10299","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/2506.10299/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":"2506.10299","created_at":"2026-07-05T11:20:21.106505+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.10299v1","created_at":"2026-07-05T11:20:21.106505+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.10299","created_at":"2026-07-05T11:20:21.106505+00:00"},{"alias_kind":"pith_short_12","alias_value":"ARAEYOPYGHAQ","created_at":"2026-07-05T11:20:21.106505+00:00"},{"alias_kind":"pith_short_16","alias_value":"ARAEYOPYGHAQP4FW","created_at":"2026-07-05T11:20:21.106505+00:00"},{"alias_kind":"pith_short_8","alias_value":"ARAEYOPY","created_at":"2026-07-05T11:20:21.106505+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2604.16287","citing_title":"NaijaS2ST: A Multi-Accent Benchmark for Speech-to-Speech Translation in Low-Resource Nigerian Languages","ref_index":4,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3","json":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3.json","graph_json":"https://pith.science/api/pith-number/ARAEYOPYGHAQP4FWUYMKVFWGP3/graph.json","events_json":"https://pith.science/api/pith-number/ARAEYOPYGHAQP4FWUYMKVFWGP3/events.json","paper":"https://pith.science/paper/ARAEYOPY"},"agent_actions":{"view_html":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3","download_json":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3.json","view_paper":"https://pith.science/paper/ARAEYOPY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.10299&json=true","fetch_graph":"https://pith.science/api/pith-number/ARAEYOPYGHAQP4FWUYMKVFWGP3/graph.json","fetch_events":"https://pith.science/api/pith-number/ARAEYOPYGHAQP4FWUYMKVFWGP3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3/action/storage_attestation","attest_author":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3/action/author_attestation","sign_citation":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3/action/citation_signature","submit_replication":"https://pith.science/pith/ARAEYOPYGHAQP4FWUYMKVFWGP3/action/replication_record"}},"created_at":"2026-07-05T11:20:21.106505+00:00","updated_at":"2026-07-05T11:20:21.106505+00:00"}