TangoFlux: Super Fast and Faithful Text to Audio Generation with Flow Matching and Clap-Ranked Preference Optimization

2024-12-30

AI Summary

To address the scarcity of preference data, misalignment between text and audio, and low inference efficiency in text-to-audio (TTA) generation, this paper proposes TangoFlux—a novel TTA model—and the CLAP-Ranked Preference Optimization (CRPO) framework. CRPO eliminates the need for manual annotations by leveraging CLAP’s multimodal embeddings to automatically rank preferences and iteratively synthesize high-quality audio-text pairs. The core generator employs Flow Matching, enhancing modeling fidelity and sampling efficiency. Evaluated on high-fidelity 44.1 kHz audio generation, the 515M-parameter TangoFlux model generates 30-second audio in just 3.7 seconds on a single A40 GPU, achieving state-of-the-art performance in both objective metrics (e.g., CLAPScore, FAD) and human evaluations. All code and models are publicly released.

Technology Category

• Natural Language Processing: Generation
• Machine Learning: Learning Preferences or Rankings
• Computer Vision: Diffusion Models for Vision

Application Category

• Search and Retrieval-Augmented AI: Web search models and ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications

Abstract

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 that the audio preference dataset generated using CRPO outperforms existing alternatives. With this framework, TangoFlux achieves state-of-the-art performance across both objective and subjective benchmarks. We open source all code and models to support further research in TTA generation.

Problem

• Text-to-Speech
• Efficiency
• Quality

Innovation

• TangoFlux Model
• CLAP Ranking Preference Optimization (CRPO)
• High-quality Audio Generation