# NotaGen: Advancing Musicality in Symbolic Music Generation with Large Language Model Training Paradigms

arXiv: https://arxiv.org/abs/2502.18008 · HTML: https://mlsanity.com/paper/2502.18008/

## Key facts

- Cited by 30 papers in the map (data through 14 Oct 2026).
- Builds on 23 of its 42 references that are in the map.
- First posted on arXiv 25 Feb 2025.

## Builds on (most cited first, 23 of 23)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer](https://mlsanity.com/paper/1910.10683/) (2019)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [GLM: General Language Model Pretraining with Autoregressive Blank Infilling](https://mlsanity.com/paper/2103.10360/) (2021)
- [RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedback](https://mlsanity.com/paper/2309.00267/) (2023)
- [Smaug: Fixing Failure Modes of Preference Optimisation with DPO-Positive](https://mlsanity.com/paper/2402.13228/) (2024)
- [LakhNES: Improving multi-instrumental music generation with cross-domain pre-training](https://mlsanity.com/paper/1907.04868/) (2019)
- [Music transcription modelling and composition using deep learning](https://mlsanity.com/paper/1604.08723/) (2016)
- [Symphony Generation with Permutation Invariant Language Model](https://mlsanity.com/paper/2205.05448/) (2022)
- [MuPT: A Generative Symbolic Music Pretrained Transformer](https://mlsanity.com/paper/2404.06393/) (2024)
- [MusicRL: Aligning Music Generation to Human Preferences](https://mlsanity.com/paper/2402.04229/) (2024)
- [TunesFormer: Forming Irish Tunes with Control Codes by Bar Patching](https://mlsanity.com/paper/2301.02884/) (2023)
- [RL-Duet: Online Music Accompaniment Generation Using Deep Reinforcement Learning](https://mlsanity.com/paper/2002.03082/) (2020)
- [Frechet Music Distance: A Metric For Generative Symbolic Music Evaluation](https://mlsanity.com/paper/2412.07948/) (2024)
- [CLaMP: Contrastive Language-Music Pre-training for Cross-Modal Symbolic Music Information Retrieval](https://mlsanity.com/paper/2304.11029/) (2023)
- [MelodyT5: A Unified Score-to-Score Transformer for Symbolic Music Processing](https://mlsanity.com/paper/2407.02277/) (2024)
- [CLaMP 2: Multimodal Music Information Retrieval Across 101 Languages Using Large Language Models](https://mlsanity.com/paper/2410.13267/) (2024)
- [MelodyGLM: Multi-task Pre-training for Symbolic Melody Generation](https://mlsanity.com/paper/2309.10738/) (2023)
- [Score Transformer: Generating Musical Score from Note-level Representation](https://mlsanity.com/paper/2112.00355/) (2021)
- [Exploring Tokenization Methods for Multitrack Sheet Music Generation](https://mlsanity.com/paper/2410.17584/) (2024)

## Built on by (most cited first, 30 of 30)

- [DiffRhythm 2: Efficient and High Fidelity Song Generation via Block Flow Matching](https://mlsanity.com/paper/2510.22950/) (2025)
- [Scaling Self-Supervised Representation Learning for Symbolic Piano Performance](https://mlsanity.com/paper/2506.23869/) (2025)
- [AutoMV: An Automatic Multi-Agent System for Music Video Generation](https://mlsanity.com/paper/2512.12196/) (2025)
- [Towards an AI Musician: Synthesizing Sheet Music Problems for Musical Reasoning](https://mlsanity.com/paper/2509.04059/) (2025)
- [ABC-Eval: Benchmarking Large Language Models on Symbolic Music Understanding and Instruction Following](https://mlsanity.com/paper/2509.23350/) (2025)
- [Via Score to Performance: Efficient Human-Controllable Long Song Generation with Bar-Level Symbolic Notation](https://mlsanity.com/paper/2508.01394/) (2025)
- [Composer Vector: Style-steering Symbolic Music Generation in a Latent Space](https://mlsanity.com/paper/2604.03333/) (2026)
- [BEAT: Tokenizing and Generating Symbolic Music by Uniform Temporal Steps](https://mlsanity.com/paper/2604.19532/) (2026)
- [Musical Score Understanding Benchmark: Evaluating Large Language Models' Comprehension of Complete Musical Scores](https://mlsanity.com/paper/2511.20697/) (2025)
- [PhraseVAE and PhraseLDM: Latent Diffusion for Full-Song Multitrack Symbolic Music Generation](https://mlsanity.com/paper/2512.11348/) (2025)
- [SymphonyGen: 3D Hierarchical Orchestral Generation with Controllable Harmony Skeleton](https://mlsanity.com/paper/2604.25498/) (2026)
- [From Generality to Mastery: Composer-Style Symbolic Music Generation via Large-Scale Pre-training](https://mlsanity.com/paper/2506.17497/) (2025)
- [CoComposer: LLM Multi-agent Collaborative Music Composition](https://mlsanity.com/paper/2509.00132/) (2025)
- [ProGress: Structured Music Generation via Graph Diffusion and Hierarchical Music Analysis](https://mlsanity.com/paper/2510.10249/) (2025)
- [MIDI-LLM: Improving Text-to-MIDI Music Generation via Adapting Large Language Models](https://mlsanity.com/paper/2511.03942/) (2025)
- [How Far Can Pretrained LLMs Go in Symbolic Music? Controlled Comparisons of Supervised and Preference-based Adaptation](https://mlsanity.com/paper/2601.22764/) (2026)
- [BMdataset: A Musicologically Curated LilyPond Dataset](https://mlsanity.com/paper/2604.10628/) (2026)
- [Optical Music Recognition for Real-World Manuscripts with Synthetic Data](https://mlsanity.com/paper/2606.09479/) (2026)
- [Large Language Models' Internal Perception of Symbolic Music](https://mlsanity.com/paper/2507.12808/) (2025)
- [Amadeus: Autoregressive Model with Bidirectional Attribute Modelling for Symbolic Music](https://mlsanity.com/paper/2508.20665/) (2025)
- [Aligning Generative Music AI with Human Preferences: Methods and Challenges](https://mlsanity.com/paper/2511.15038/) (2025)
- [Whole-Piece Training for Symbolic Music Language Models via Full-Horizon Compressed Recurrence](https://mlsanity.com/paper/2602.19816/) (2026)
- [Can LLMs understand LilyPond? A benchmark for symbolic music generation and understanding](https://mlsanity.com/paper/2606.08722/) (2026)
- [Libretto: Giving LLM Agents a Sense of Musical Structure](https://mlsanity.com/paper/2606.22708/) (2026)
- [UOT-IR: Structured Routing of High-Polyphony Symbolic Music into Fixed-Budget Representations](https://mlsanity.com/paper/2608.00576/) (2026)
- [DDSynth-RL: Audio Synthesizer Inversion via Discrete Diffusion with Reinforcement Learning](https://mlsanity.com/paper/2608.03032/) (2026)
- [TUTTI: Toward generalizable audio-to-score transcription via fully synthesized data](https://mlsanity.com/paper/2609.00640/) (2026)
- [Which Constraints Are Missing? Ask the Verifier: Graded Rewards for Constraint-Following Music Generation](https://mlsanity.com/paper/2609.23665/) (2026)
- [Getting Motif-ated: Controllable AI Compositions from Injected Motif Prompts](https://mlsanity.com/paper/2609.32788/) (2026)
- [Harmonic Representations in Symbolic Music Transformers are Localized and Controllable](https://mlsanity.com/paper/2610.15957/) (2026)

## Abstract (arXiv)

We introduce NotaGen, a symbolic music generation model aiming to explore the potential of producing high-quality classical sheet music. Inspired by the success of Large Language Models (LLMs), NotaGen adopts pre-training, fine-tuning, and reinforcement learning paradigms (henceforth referred to as the LLM training paradigms). It is pre-trained on 1.6M pieces of music in ABC notation, and then fine-tuned on approximately 9K high-quality classical compositions conditioned on "period-composer-instrumentation" prompts. For reinforcement learning, we propose the CLaMP-DPO method, which further enhances generation quality and controllability without requiring human annotations or predefined rewards. Our experiments demonstrate the efficacy of CLaMP-DPO in symbolic music generation models with different architectures and encoding schemes. Furthermore, subjective A/B tests show that NotaGen outperforms baseline models against human compositions, greatly advancing musical aesthetics in symbolic music generation.