{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:IJCIFOTYXQ3IJVRAPL7ANO6G2S","short_pith_number":"pith:IJCIFOTY","schema_version":"1.0","canonical_sha256":"424482ba78bc3684d6207afe06bbc6d4b2e4aa4035f4d9296a9a45f7e30b46d8","source":{"kind":"arxiv","id":"2606.09234","version":1},"attestation_state":"computed","paper":{"title":"End-to-End Training for Discrete Token LLM based TTS System","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SD","authors_text":"Changfeng Gao, Jun Yuan, ShiDong Shang, Ye Bai, Yong Ren, Zhao You","submitted_at":"2026-06-08T09:07:23Z","abstract_excerpt":"Recent state-of-the-art (SOTA) text-to-speech (TTS) systems typically adopt a cascaded pipeline consisting of a speech tokenizer, an autoregressive large language model (LLM), and a diffusion based flow-matching (FM) model, with these components trained independently. In this paper, we propose a fully end-to-end (E2E) optimization framework that unifies the training of the speech tokenizer, LLM, FM model, and an additional reward model (RM). Specifically, we first jointly optimize the tokenizer using multi-task objectives derived from reconstruction for FM, next-token prediction for LLM, and m"},"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":"2606.09234","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.SD","submitted_at":"2026-06-08T09:07:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"a8f1682541077c877446a1da9ea89211ab62b644f3e361ace89c332c5abafa92","abstract_canon_sha256":"3910ee088baec1cf69b0376606804658676851b239957d2d746c870ce4ef791d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T02:08:09.064843Z","signature_b64":"1/MkaIZM4ZegqR2jchk9DaWaezV7tqv8/Egk09LRoWL3CN1EJZMQ7WRJYhuokb9FDaLFncB8m84qO0VBCbEJDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"424482ba78bc3684d6207afe06bbc6d4b2e4aa4035f4d9296a9a45f7e30b46d8","last_reissued_at":"2026-06-09T02:08:09.063783Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T02:08:09.063783Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"End-to-End Training for Discrete Token LLM based TTS System","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.SD","authors_text":"Changfeng Gao, Jun Yuan, ShiDong Shang, Ye Bai, Yong Ren, Zhao You","submitted_at":"2026-06-08T09:07:23Z","abstract_excerpt":"Recent state-of-the-art (SOTA) text-to-speech (TTS) systems typically adopt a cascaded pipeline consisting of a speech tokenizer, an autoregressive large language model (LLM), and a diffusion based flow-matching (FM) model, with these components trained independently. In this paper, we propose a fully end-to-end (E2E) optimization framework that unifies the training of the speech tokenizer, LLM, FM model, and an additional reward model (RM). Specifically, we first jointly optimize the tokenizer using multi-task objectives derived from reconstruction for FM, next-token prediction for LLM, and m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.09234","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/2606.09234/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":"2606.09234","created_at":"2026-06-09T02:08:09.063932+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.09234v1","created_at":"2026-06-09T02:08:09.063932+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.09234","created_at":"2026-06-09T02:08:09.063932+00:00"},{"alias_kind":"pith_short_12","alias_value":"IJCIFOTYXQ3I","created_at":"2026-06-09T02:08:09.063932+00:00"},{"alias_kind":"pith_short_16","alias_value":"IJCIFOTYXQ3IJVRA","created_at":"2026-06-09T02:08:09.063932+00:00"},{"alias_kind":"pith_short_8","alias_value":"IJCIFOTY","created_at":"2026-06-09T02:08:09.063932+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/IJCIFOTYXQ3IJVRAPL7ANO6G2S","json":"https://pith.science/pith/IJCIFOTYXQ3IJVRAPL7ANO6G2S.json","graph_json":"https://pith.science/api/pith-number/IJCIFOTYXQ3IJVRAPL7ANO6G2S/graph.json","events_json":"https://pith.science/api/pith-number/IJCIFOTYXQ3IJVRAPL7ANO6G2S/events.json","paper":"https://pith.science/paper/IJCIFOTY"},"agent_actions":{"view_html":"https://pith.science/pith/IJCIFOTYXQ3IJVRAPL7ANO6G2S","download_json":"https://pith.science/pith/IJCIFOTYXQ3IJVRAPL7ANO6G2S.json","view_paper":"https://pith.science/paper/IJCIFOTY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.09234&json=true","fetch_graph":"https://pith.science/api/pith-number/IJCIFOTYXQ3IJVRAPL7ANO6G2S/graph.json","fetch_events":"https://pith.science/api/pith-number/IJCIFOTYXQ3IJVRAPL7ANO6G2S/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IJCIFOTYXQ3IJVRAPL7ANO6G2S/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IJCIFOTYXQ3IJVRAPL7ANO6G2S/action/storage_attestation","attest_author":"https://pith.science/pith/IJCIFOTYXQ3IJVRAPL7ANO6G2S/action/author_attestation","sign_citation":"https://pith.science/pith/IJCIFOTYXQ3IJVRAPL7ANO6G2S/action/citation_signature","submit_replication":"https://pith.science/pith/IJCIFOTYXQ3IJVRAPL7ANO6G2S/action/replication_record"}},"created_at":"2026-06-09T02:08:09.063932+00:00","updated_at":"2026-06-09T02:08:09.063932+00:00"}