{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:VEX4F6XOQ4TMTPBDJKMN6JVKYY","short_pith_number":"pith:VEX4F6XO","schema_version":"1.0","canonical_sha256":"a92fc2faee8726c9bc234a98df26aac62e701717c199752a5d67b752e551f8a8","source":{"kind":"arxiv","id":"2412.21037","version":2},"attestation_state":"computed","paper":{"title":"TangoFlux: Super Fast and Faithful Text to Audio Generation with Flow Matching and Clap-Ranked Preference Optimization","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","eess.AS"],"primary_cat":"cs.SD","authors_text":"Ambuj Mehrish, Amir Ali Bagherzadeh, Bryan Catanzaro, Chia-Yu Hung, Chuan Li, Navonil Majumder, Rafael Valle, Soujanya Poria, Zhifeng Kong","submitted_at":"2024-12-30T16:02:44Z","abstract_excerpt":"We introduce TangoFlux, an efficient Text-to-Audio (TTA) generative model with 515M parameters, capable of generating up to 30 seconds of 44.1kHz audio in just 3.7 seconds on a single A40 GPU. A key challenge in aligning TTA models lies in the difficulty of creating preference pairs, as TTA lacks structured mechanisms like verifiable rewards or gold-standard answers available for Large Language Models (LLMs). To address this, we propose CLAP-Ranked Preference Optimization (CRPO), a novel framework that iteratively generates and optimizes preference data to enhance TTA alignment. We demonstrate"},"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":"2412.21037","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.SD","submitted_at":"2024-12-30T16:02:44Z","cross_cats_sorted":["cs.AI","cs.CL","eess.AS"],"title_canon_sha256":"1691dfe3d2f6b49c91fb50fc97607a5cb514ce1c96c5c857a0c123c95e146f41","abstract_canon_sha256":"076cbb7258f3996519cadf5bfdfcbccfb76f5efba3fa66b420861933dc78705a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:46:59.610614Z","signature_b64":"faJIT79qyT9TPOnmsnD1DI11hzSPBlG4i4aQCvNVxTyIbi1CeJChrlL0hbSgXpem7zN9S8GXdPK/MKiciSIUAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a92fc2faee8726c9bc234a98df26aac62e701717c199752a5d67b752e551f8a8","last_reissued_at":"2026-07-05T10:46:59.610115Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:46:59.610115Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"TangoFlux: Super Fast and Faithful Text to Audio Generation with Flow Matching and Clap-Ranked Preference Optimization","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.CL","eess.AS"],"primary_cat":"cs.SD","authors_text":"Ambuj Mehrish, Amir Ali Bagherzadeh, Bryan Catanzaro, Chia-Yu Hung, Chuan Li, Navonil Majumder, Rafael Valle, Soujanya Poria, Zhifeng Kong","submitted_at":"2024-12-30T16:02:44Z","abstract_excerpt":"We introduce TangoFlux, an efficient Text-to-Audio (TTA) generative model with 515M parameters, capable of generating up to 30 seconds of 44.1kHz audio in just 3.7 seconds on a single A40 GPU. A key challenge in aligning TTA models lies in the difficulty of creating preference pairs, as TTA lacks structured mechanisms like verifiable rewards or gold-standard answers available for Large Language Models (LLMs). To address this, we propose CLAP-Ranked Preference Optimization (CRPO), a novel framework that iteratively generates and optimizes preference data to enhance TTA alignment. We demonstrate"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.21037","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/2412.21037/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":"2412.21037","created_at":"2026-07-05T10:46:59.610176+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.21037v2","created_at":"2026-07-05T10:46:59.610176+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.21037","created_at":"2026-07-05T10:46:59.610176+00:00"},{"alias_kind":"pith_short_12","alias_value":"VEX4F6XOQ4TM","created_at":"2026-07-05T10:46:59.610176+00:00"},{"alias_kind":"pith_short_16","alias_value":"VEX4F6XOQ4TMTPBD","created_at":"2026-07-05T10:46:59.610176+00:00"},{"alias_kind":"pith_short_8","alias_value":"VEX4F6XO","created_at":"2026-07-05T10:46:59.610176+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":18,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.05196","citing_title":"Unified Audio Intelligence Without Regressing on Text Intelligence","ref_index":119,"is_internal_anchor":true},{"citing_arxiv_id":"2606.23080","citing_title":"AudioCALM: Continuous Autoregressive Language Modeling for Universal Audio Generation","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2606.23064","citing_title":"STAR-VAE: Structured Topology-Aware Regularization for Audio Reconstruction and Generation","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20101","citing_title":"RFM-Editing 2: Text-Guided Audio Editing with Rectified Flow Matching and Coarse-to-Fine Diffusion Transformers","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20101","citing_title":"RFM-Editing 2: Text-Guided Audio Editing with Rectified Flow Matching and Coarse-to-Fine Diffusion Transformers","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2606.12555","citing_title":"AudioX-Turbo: A Unified Framework for Efficient Anything-to-Audio Generation","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.30965","citing_title":"ImmersiveTTS: Environment-Aware Text-to-Speech with Multimodal Diffusion Transformer and Domain-Specific Representation Alignment","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31026","citing_title":"OTCache: Optimal Transport for Geometry-Aware Caching in Diffusion Models","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27838","citing_title":"Dasheng AudioGen: A Unified Model for Generating Coherent Audio Scenes from Text","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.31530","citing_title":"UNISON: A Unified Sound Generation and Editing Framework via Deep LLM Fusion","ref_index":82,"is_internal_anchor":false},{"citing_arxiv_id":"2505.13777","citing_title":"Sat2Sound: A Unified Framework for Zero-Shot Soundscape Mapping","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11866","citing_title":"AuDirector: A Self-Reflective Closed-Loop Framework for Immersive Audio Storytelling","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2509.14003","citing_title":"RFM-Editing: Rectified Flow Matching for Text-guided Audio Editing","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2509.23727","citing_title":"AudioMoG: Guiding Audio Generation with Mixture-of-Guidance","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2604.01929","citing_title":"Woosh: A Sound Effects Foundation Model","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11866","citing_title":"AuDirector: A Self-Reflective Closed-Loop Framework for Immersive Audio Storytelling","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00329","citing_title":"Fast Text-to-Audio Generation with One-Step Sampling via Energy-Scoring and Auxiliary Contextual Representation Distillation","ref_index":15,"is_internal_anchor":false},{"citing_arxiv_id":"2604.15086","citing_title":"ControlFoley: Unified and Controllable Video-to-Audio Generation with Cross-Modal Conflict Handling","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY","json":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY.json","graph_json":"https://pith.science/api/pith-number/VEX4F6XOQ4TMTPBDJKMN6JVKYY/graph.json","events_json":"https://pith.science/api/pith-number/VEX4F6XOQ4TMTPBDJKMN6JVKYY/events.json","paper":"https://pith.science/paper/VEX4F6XO"},"agent_actions":{"view_html":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY","download_json":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY.json","view_paper":"https://pith.science/paper/VEX4F6XO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.21037&json=true","fetch_graph":"https://pith.science/api/pith-number/VEX4F6XOQ4TMTPBDJKMN6JVKYY/graph.json","fetch_events":"https://pith.science/api/pith-number/VEX4F6XOQ4TMTPBDJKMN6JVKYY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY/action/storage_attestation","attest_author":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY/action/author_attestation","sign_citation":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY/action/citation_signature","submit_replication":"https://pith.science/pith/VEX4F6XOQ4TMTPBDJKMN6JVKYY/action/replication_record"}},"created_at":"2026-07-05T10:46:59.610176+00:00","updated_at":"2026-07-05T10:46:59.610176+00:00"}