File: README.md Copy Chapter

ACE-Step

A Step Towards Music Generation Foundation Model

Table of Contents Copy Chapter

• ✨ Features
• 📦 Installation
• 🚀 Usage
• 📱 User Interface Guide
• 🔨 Train

📝 Abstract Copy Chapter

We introduce ACE-Step, a novel open-source foundation model for music generation that overcomes key limitations of existing approaches and achieves state-of-the-art performance through a holistic architectural design. Current methods face inherent trade-offs between generation speed, musical coherence, and controllability. For instance, LLM-based models (e.g., Yue, SongGen) excel at lyric alignment but suffer from slow inference and structural artifacts. Diffusion models (e.g., DiffRhythm), on the other hand, enable faster synthesis but often lack long-range structural coherence.

ACE-Step bridges this gap by integrating diffusion-based generation with Sana’s Deep Compression AutoEncoder (DCAE) and a lightweight linear transformer. It further leverages MERT and m-hubert to align semantic representations (REPA) during training, enabling rapid convergence. As a result, our model synthesizes up to 4 minutes of music in just 20 seconds on an A100 GPU—15× faster than LLM-based baselines—while achieving superior musical coherence and lyric alignment across melody, harmony, and rhythm metrics. Moreover, ACE-Step preserves fine-grained acoustic details, enabling advanced control mechanisms such as voice cloning, lyric editing, remixing, and track generation (e.g., lyric2vocal, singing2accompaniment).

Rather than building yet another end-to-end text-to-music pipeline, our vision is to establish a foundation model for music AI: a fast, general-purpose, efficient yet flexible architecture that makes it easy to train sub-tasks on top of it. This paves the way for developing powerful tools that seamlessly integrate into the creative workflows of music artists, producers, and content creators. In short, we aim to build the Stable Diffusion moment for music.

📢 News and Updates Copy Chapter

• 🎉 2026.01.28: Released ACE-Step v1.5 - Our latest and most advanced model is now available!

• 📃 2025.06.02: Released ACE-Step Technical Report (PDF).

• 🎮 2025.05.14: Add Stable Audio Open Small sampler pingpong. Use SDE to achieve better music consistency and quality, including lyric alignment and style alignment. Use a better method to re-implement Audio2Audio

• 🎤 2025.05.12: Release RapMachine and fix lora training issues

  • See ZH_RAP_LORA.md for details. Audio Examples: https://ace-step.github.io/#RapMachine
  • See TRAIN_INSTRUCTION.md for detailed training instructions.

• 🔥 2025.05.10: Memory Optimization Update

  • Reduced Max VRAM to 8GB, making it more compatible with consumer devices
  • Recommended launch options:
    bash Copy

    acestep --torch_compile true --cpu_offload true --overlapped_decode true

    Windows need to install triton:
    text Copy

    pip install triton-windows

• 📢 2025.05.09: Graidio Demo support Audio2Audio. ComfyUI: Ace_Step_4x_a2a.json

• 🚀 2025.05.08: ComfyUI_ACE-Step node is now available! Explore the power of ACE-Step within ComfyUI. 🎉

• 🚀 2025.05.06: Open source demo code and model

✨ Features Copy Chapter

🎯 Baseline Quality Copy Section

🌈 Diverse Styles & Genres

• 🎸 Supports all mainstream music styles with various description formats including short tags, descriptive text, or use-case scenarios
• 🎷 Capable of generating music across different genres with appropriate instrumentation and style

🌍 Multiple Languages

• 🗣️ Supports 19 languages with top 10 well-performing languages including:
  • 🇺🇸 English, 🇨🇳 Chinese, 🇷🇺 Russian, 🇪🇸 Spanish, 🇯🇵 Japanese, 🇩🇪 German, 🇫🇷 French, 🇵🇹 Portuguese, 🇮🇹 Italian, 🇰🇷 Korean
• ⚠️ Due to data imbalance, less common languages may underperform

🎻 Instrumental Styles

• 🎹 Supports various instrumental music generation across different genres and styles
• 🎺 Capable of producing realistic instrumental tracks with appropriate timbre and expression for each instrument
• 🎼 Can generate complex arrangements with multiple instruments while maintaining musical coherence

🎤 Vocal Techniques

• 🎙️ Capable of rendering various vocal styles and techniques with good quality
• 🗣️ Supports different vocal expressions including various singing techniques and styles

🎛️ Controllability Copy Section

🔄 Variations Generation

• ⚙️ Implemented using training-free, inference-time optimization techniques
• 🌊 Flow-matching model generates initial noise, then uses trigFlow's noise formula to add additional Gaussian noise
• 🎚️ Adjustable mixing ratio between original initial noise and new Gaussian noise to control variation degree

🎨 Repainting

• 🖌️ Implemented by adding noise to the target audio input and applying mask constraints during the ODE process
• 🔍 When input conditions change from the original generation, only specific aspects can be modified while preserving the rest
• 🔀 Can be combined with Variations Generation techniques to create localized variations in style, lyrics, or vocals

✏️ Lyric Editing

• 💡 Innovatively applies flow-edit technology to enable localized lyric modifications while preserving melody, vocals, and accompaniment
• 🔄 Works with both generated content and uploaded audio, greatly enhancing creative possibilities
• ℹ️ Current limitation: can only modify small segments of lyrics at once to avoid distortion, but multiple edits can be applied sequentially

🚀 Applications Copy Section

🎤 Lyric2Vocal (LoRA)

• 🔊 Based on a LoRA fine-tuned on pure vocal data, allowing direct generation of vocal samples from lyrics
• 🛠️ Offers numerous practical applications such as vocal demos, guide tracks, songwriting assistance, and vocal arrangement experimentation
• ⏱️ Provides a quick way to test how lyrics might sound when sung, helping songwriters iterate faster

📝 Text2Samples (LoRA)

• 🎛️ Similar to Lyric2Vocal, but fine-tuned on pure instrumental and sample data
• 🎵 Capable of generating conceptual music production samples from text descriptions
• 🧰 Useful for quickly creating instrument loops, sound effects, and musical elements for production

🔮 Coming Soon Copy Section

🎤 RapMachine

• 🔥 Fine-tuned on pure rap data to create an AI system specialized in rap generation
• 🏆 Expected capabilities include AI rap battles and narrative expression through rap
• 📚 Rap has exceptional storytelling and expressive capabilities, offering extraordinary application potential

🎛️ StemGen

• 🎚️ A controlnet-lora trained on multi-track data to generate individual instrument stems
• 🎯 Takes a reference track and specified instrument (or instrument reference audio) as input
• 🎹 Outputs an instrument stem that complements the reference track, such as creating a piano accompaniment for a flute melody or adding jazz drums to a lead guitar

🎤 Singing2Accompaniment

• 🔄 The reverse process of StemGen, generating a mixed master track from a single vocal track
• 🎵 Takes a vocal track and specified style as input to produce a complete vocal accompaniment
• 🎸 Creates full instrumental backing that complements the input vocals, making it easy to add professional-sounding accompaniment to any vocal recording

📋 Roadmap Copy Chapter

• Release training code 🔥
• Release LoRA training code 🔥
• Release RapMachine LoRA 🎤
• Release evaluation performance and technical report 📄
• Train and Release ACE-Step V1.5
• Release ControlNet training code 🔥
• Release Singing2Accompaniment ControlNet 🎮

🖥️ Hardware Performance Copy Chapter

We have evaluated ACE-Step across different hardware setups, yielding the following throughput results:

We use RTF (Real-Time Factor) to measure the performance of ACE-Step. Higher values indicate faster generation speed. 27.27x means to generate 1 minute of music, it takes 2.2 seconds (60/27.27). The performance is measured on a single GPU with batch size 1 and 27 steps.

📦 Installation Copy Chapter

1. Clone the Repository Copy Section

First, clone the ACE-Step repository to your local machine and navigate into the project directory:

git clone https://github.com/ace-step/ACE-Step.git
cd ACE-Step

2. Prerequisites Copy Section

Ensure you have the following installed:

• Python: Version 3.10 or later is recommended. You can download it from python.org.
• Conda or venv: For creating a virtual environment (Conda is recommended).

3. Set Up a Virtual Environment Copy Section

It is highly recommended to use a virtual environment to manage project dependencies and avoid conflicts. Choose one of the following methods:

Option A: Using Conda

1. Create the environment named ace_step with Python 3.10:

   bash Copy

   conda create -n ace_step python=3.10 -y

2. Activate the environment:

   bash Copy

   conda activate ace_step

Option B: Using venv

1. Navigate to the cloned ACE-Step directory.

2. Create the virtual environment (commonly named venv):

   bash Copy

   python -m venv venv

3. Activate the environment:

   • On Windows (cmd.exe):
     bash Copy

     venv\Scripts\activate.bat

   • On Windows (PowerShell):
     powershell Copy

     .\venv\Scripts\Activate.ps1

     (If you encounter execution policy errors, you might need to run Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope Process first)
   • On Linux / macOS (bash/zsh):
     bash Copy

     source venv/bin/activate

4. Install Dependencies Copy Section

Once your virtual environment is activated:
a. (Windows Only) If you are on Windows and plan to use an NVIDIA GPU, install PyTorch with CUDA support first:

pip3 install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu126

(Adjust cu126 if you have a different CUDA version. For other PyTorch installation options, refer to the official PyTorch website).

b. Install ACE-Step and its core dependencies:

pip install -e .

The ACE-Step application is now installed. The GUI works on Windows, macOS, and Linux. For instructions on how to run it, please see the Usage section.

🚀 Usage Copy Chapter

🔍 Basic Usage Copy Section

acestep --port 7865

⚙️ Advanced Usage Copy Section

acestep --checkpoint_path /path/to/checkpoint --port 7865 --device_id 0 --share true --bf16 true

• If --checkpoint_path is set and models exist at the path, load from checkpoint_path.
• If --checkpoint_path is set but models do not exist at the path, auto download models to checkpoint_path.
• If --checkpoint_path is not set, auto download models to the default path ~/.cache/ace-step/checkpoints.

If you are using macOS, please use --bf16 false to avoid errors.

🔍 API Usage

If you intend to integrate ACE-Step as a library into your own Python projects, you can install the latest version directly from GitHub using the following pip command.

Direct Installation via pip:

1. Ensure Git is installed: This method requires Git to be installed on your system and accessible in your system's PATH.
2. Execute the installation command:
   bash Copy

   pip install git+https://github.com/ace-step/ACE-Step.git

   It's recommended to use this command within a virtual environment to avoid conflicts with other packages.

🛠️ Command Line Arguments

• --checkpoint_path: Path to the model checkpoint (default: downloads automatically)
• --server_name: IP address or hostname for the Gradio server to bind to (default: '127.0.0.1'). Use '0.0.0.0' to make it accessible from other devices on the network.
• --port: Port to run the Gradio server on (default: 7865)
• --device_id: GPU device ID to use (default: 0)
• --share: Enable Gradio sharing link (default: False)
• --bf16: Use bfloat16 precision for faster inference (default: True)
• --torch_compile: Use torch.compile() to optimize the model, speeding up inference (default: False).
  • Windows need to install triton:
    text Copy

    pip install triton-windows

• --cpu_offload: Offload model weights to CPU to save GPU memory (default: False)
• --overlapped_decode: Use overlapped decoding to speed up inference (default: False)

📱 User Interface Guide Copy Chapter

The ACE-Step interface provides several tabs for different music generation and editing tasks:

📝 Text2Music Tab Copy Section

1. 📋 Input Fields:

   • 🏷️ Tags: Enter descriptive tags, genres, or scene descriptions separated by commas
   • 📜 Lyrics: Enter lyrics with structure tags like [verse], [chorus], and [bridge]
   • ⏱️ Audio Duration: Set the desired duration of the generated audio (-1 for random)

2. ⚙️ Settings:

   • 🔧 Basic Settings: Adjust inference steps, guidance scale, and seeds
   • 🔬 Advanced Settings: Fine-tune scheduler type, CFG type, ERG settings, and more

3. 🚀 Generation: Click "Generate" to create music based on your inputs

🔄 Retake Tab Copy Section

• 🎲 Regenerate music with slight variations using different seeds
• 🎚️ Adjust variance to control how much the retake differs from the original

🎨 Repainting Tab Copy Section

• 🖌️ Selectively regenerate specific sections of the music
• ⏱️ Specify start and end times for the section to repaint
• 🔍 Choose the source audio (text2music output, last repaint, or upload)

✏️ Edit Tab Copy Section

• 🔄 Modify existing music by changing tags or lyrics
• 🎛️ Choose between "only_lyrics" mode (preserves melody) or "remix" mode (changes melody)
• 🎚️ Adjust edit parameters to control how much of the original is preserved

📏 Extend Tab Copy Section

• ➕ Add music to the beginning or end of an existing piece
• 📐 Specify left and right extension lengths
• 🔍 Choose the source audio to extend

📂 Examples Copy Chapter

The examples/input_params directory contains sample input parameters that can be used as references for generating music.

🏗️ Architecture Copy Chapter

🔨 Train Copy Chapter

See TRAIN_INSTRUCTION.md for detailed training instructions.

📜 License & Disclaimer Copy Chapter

This project is licensed under Apache License 2.0

ACE-Step enables original music generation across diverse genres, with applications in creative production, education, and entertainment. While designed to support positive and artistic use cases, we acknowledge potential risks such as unintentional copyright infringement due to stylistic similarity, inappropriate blending of cultural elements, and misuse for generating harmful content. To ensure responsible use, we encourage users to verify the originality of generated works, clearly disclose AI involvement, and obtain appropriate permissions when adapting protected styles or materials. By using ACE-Step, you agree to uphold these principles and respect artistic integrity, cultural diversity, and legal compliance. The authors are not responsible for any misuse of the model, including but not limited to copyright violations, cultural insensitivity, or the generation of harmful content.

🔔 Important Notice
The only official website for the ACE-Step project is our GitHub Pages site.
We do not operate any other websites.
🚫 Fake domains include but are not limited to:
ac**p.com, a**p.org, a***c.org
⚠️ Please be cautious. Do not visit, trust, or make payments on any of those sites.

🙏 Acknowledgements Copy Chapter

This project is co-led by ACE Studio and StepFun.

📖 Citation Copy Chapter

If you find this project useful for your research, please consider citing:

@misc{gong2025acestep,
	title={ACE-Step: A Step Towards Music Generation Foundation Model},
	author={Junmin Gong, Wenxiao Zhao, Sen Wang, Shengyuan Xu, Jing Guo}, 
	howpublished={\url{https://github.com/ace-step/ACE-Step}},
	year={2025},
	note={GitHub repository}
}

File: docker-compose.yaml Copy Chapter

services:
&name ace-step:
build:
context: https://github.com/ace-step/ACE-Step.git
dockerfile: Dockerfile
container_name: *name
hostname: *name
stop_grace_period: 2s
ports:
- "7865:7865"
volumes:
- ./checkpoints:/app/checkpoints
- ./outputs:/app/outputs
- ./logs:/app/logs
environment:
- ACE_OUTPUT_DIR=/app/outputs
# command: python app.py --server_name 0.0.0.0 --port 7865 --share False --bf16 True --torch_compile True --device-id 0
healthcheck:
test: ["CMD", "curl", "-f", "http://localhost:7865/"]
interval: 60s
timeout: 10s
retries: 30
start_period: 3s
restart: unless-stopped
runtime: nvidia
deploy:
resources:
reservations:
devices:
- driver: nvidia
count: all
capabilities: ["compute", "utility", "graphics", "video"]

File: requirements.txt Copy Chapter

datasets==3.4.1
diffusers>=0.33.0
gradio
librosa==0.11.0
loguru==0.7.3
matplotlib==3.10.1
numpy
pypinyin==0.53.0
pytorch_lightning==2.5.1
soundfile==0.13.1
torch
torchaudio
torchvision
tqdm
transformers==4.50.0
py3langid==0.3.0
hangul-romanize==0.1.0
num2words==0.5.14
spacy==3.8.4
accelerate==1.6.0
cutlet
fugashi[unidic-lite]
click
peft
tensorboard
tensorboardX

File: TRAIN_INSTRUCTION.md Copy Chapter

Training Instruction

1. Data Preparation Copy Chapter

Required File Format Copy Section

For each audio sample, you need exactly 3 files in the data directory:

1. filename.mp3 - The audio file
2. filename_prompt.txt - Audio characteristics (comma-separated tags)
3. filename_lyrics.txt - Song lyrics (optional, but recommended)

Example Data Structure Copy Section

data/
├── test_track_001.mp3
├── test_track_001_prompt.txt
└── test_track_001_lyrics.txt

File Content Format Copy Section

`*_prompt.txt` - Audio Tags

Simple comma-separated audio characteristics describing the sound, instruments, genre, mood, etc.

Example:

melodic techno, male vocal, electronic, emotional, minor key, 124 bpm, synthesizer, driving, atmospheric

Guidelines for creating prompt tags:

• Include genre (e.g., "rap", "pop", "rock", "electronic")
• Include vocal type (e.g., "male vocal", "female vocal", "spoken word")
• Include instruments actually heard (e.g., "guitar", "piano", "synthesizer", "drums")
• Include mood/energy (e.g., "energetic", "calm", "aggressive", "melancholic")
• Include tempo if known (e.g., "120 bpm", "fast tempo", "slow tempo")
• Include key if known (e.g., "major key", "minor key", "C major")

`*_lyrics.txt` - Song Lyrics

Standard song lyrics with verse/chorus structure.

Example:

[Verse]
Lately I've been wondering
Why do I do this to myself
I should be over it

[Chorus]  
It makes me want to cry
If you knew what you meant to me
I wonder if you'd come back

⚠️ Important Notes Copy Section

• File naming is strict: Must follow filename.mp3, filename_prompt.txt, filename_lyrics.txt pattern
• JSON files are NOT supported - the converter only reads the simple text files above
• Complex multi-variant descriptions are NOT used - only the simple comma-separated prompt format works

2. Convert to Huggingface Dataset Format Copy Chapter

Run the following command to convert your data to the training format:

python convert2hf_dataset.py --data_dir "./data" --repeat_count 2000 --output_name "zh_lora_dataset"

Parameters:

• --data_dir: Path to your data directory containing the MP3, prompt, and lyrics files
• --repeat_count: Number of times to repeat your data (use higher values for small datasets)
• --output_name: Name of the output dataset directory

What the Converter Creates Copy Section

The converter processes your files and creates a Huggingface dataset with these features:

Dataset Features:
{
    'keys': string,              # filename (e.g., "test_track_001")
    'filename': string,          # path to MP3 file  
    'tags': list[string],        # parsed prompt tags as array
    'speaker_emb_path': string,  # (empty, not used)
    'norm_lyrics': string,       # full lyrics text
    'recaption': dict            # (empty, not used)
}

Example processed sample:

{
    'keys': 'test_track_001',
    'filename': 'data/test_track_001.mp3',
    'tags': ['melodic techno', 'male vocal', 'electronic', 'emotional', 'minor key', '124 bpm', 'synthesizer', 'driving', 'atmospheric'],
    'speaker_emb_path': '',
    'norm_lyrics': '[Verse]\nLately I\'ve been wondering\nWhy do I do this to myself...',
    'recaption': {}
}

3. Configure Lora Parameters Copy Chapter

Refer to config/zh_rap_lora_config.json for configuring Lora parameters.

If your VRAM is not sufficient, you can reduce the r and lora_alpha parameters in the configuration file. Such as:

{
	"r": 16,
	"lora_alpha": 32,
	"target_modules": [
		"linear_q",
		"linear_k",
		"linear_v",
		"to_q",
		"to_k",
		"to_v",
		"to_out.0"
	]
}

4. Run Training Copy Chapter

Run python trainer.py with the following parameter introduction:

Trainer Parameter Interpretation

1. General Settings Copy Chapter

1. --num_nodes: This parameter specifies the number of nodes for the training process. It is an integer value, and the default is set to 1. In scenarios where distributed training across multiple nodes is applicable, this parameter determines how many nodes will be utilized. For example, if you have a cluster of machines and want to distribute the training load, you can increase this number. However, for most single-machine or basic training setups, the default value of 1 is sufficient.
2. --shift: It is a floating-point parameter with a default value of 3.0. Although its specific function depends on the implementation details of the model, it is likely used for some internal calculations related to the model, such as adjusting certain biases or offsets in the neural network architecture during the training process.

2. Training Hyperparameters Copy Chapter

1. --learning_rate: This is a crucial hyperparameter for the training process. It is a floating-point value with a default of 1e-4 (0.0001). The learning rate determines the step size at each iteration while updating the model's weights. A smaller learning rate will make the training process more stable but may require more training steps to converge. On the other hand, a larger learning rate can lead to faster convergence but might cause the model to overshoot the optimal solution and result in unstable training or even divergence.
2. --num_workers: This parameter defines the number of worker processes that will be used for data loading. It is an integer with a default value of 8. Having multiple workers can significantly speed up the data loading process, especially when dealing with large datasets. However, it also consumes additional system resources, so you may need to adjust this value based on the available resources of your machine (e.g., CPU cores and memory).
3. --epochs: It represents the number of times the entire training dataset will be passed through the model. It is an integer, and the default value is set to -1. When set to -1, the training will continue until another stopping condition (such as reaching the maximum number of steps) is met. If you set a positive integer value, the training will stop after that number of epochs.
4. --max_steps: This parameter specifies the maximum number of training steps. It is an integer with a default value of 2000000. Once the model has completed this number of training steps, the training process will stop, regardless of whether the model has fully converged or not. This is useful for setting a limit on the training duration in terms of the number of steps.
5. --every_n_train_steps: It is an integer parameter with a default of 2000. It determines how often certain operations (such as saving checkpoints, logging training progress, etc.) will be performed during the training. For example, with a value of 2000, these operations will occur every 2000 training steps.

3. Dataset and Experiment Settings Copy Chapter

1. --dataset_path: This is a string parameter that indicates the path to the dataset in the Huggingface dataset format. The default value is "./zh_lora_dataset". You need to ensure that the dataset at this path is correctly formatted and contains the necessary data for training.
2. --exp_name: It is a string parameter used to name the experiment. The default value is "chinese_rap_lora". This name can be used to distinguish different training experiments, and it is often used in logging and saving checkpoints to organize and identify the results of different runs.

4. Training Precision and Gradient Settings Copy Chapter

1. --precision: This parameter specifies the precision of the training. It is a string with a default value of "32", which usually means 32-bit floating-point precision. Higher precision can lead to more accurate training but may also consume more memory and computational resources. You can adjust this value depending on your hardware capabilities and the requirements of your model.
2. --accumulate_grad_batches: It is an integer parameter with a default value of 1. It determines how many batches of data will be used to accumulate gradients before performing an optimization step. For example, if you set it to 4, the gradients from 4 consecutive batches will be accumulated, and then the model's weights will be updated. This can be useful in scenarios where you want to simulate larger batch sizes when your available memory does not allow for actual large batch training.
3. --gradient_clip_val: This is a floating-point parameter with a default value of 0.5. It is used to clip the gradients during the backpropagation process. Clipping the gradients helps prevent the issue of gradient explosion, where the gradients become extremely large and cause the model to become unstable. By setting a clip value, the gradients will be adjusted to be within a certain range.
4. --gradient_clip_algorithm: It is a string parameter with a default value of "norm". This parameter specifies the algorithm used for gradient clipping. The "norm" algorithm is one common method, but there may be other algorithms available depending on the implementation of the training framework.

5. Checkpoint and Logging Settings Copy Chapter

1. --devices: This is an integer parameter with a default value of 1. It specifies the number of devices (such as GPUs) that will be used for training. If you have multiple GPUs available and want to use them for parallel training, you can increase this number accordingly.
2. --logger_dir: It is a string parameter with a default value of "./exps/logs/". This parameter indicates the directory where the training logs will be saved. The logs can be useful for monitoring the training progress, analyzing the performance of the model during training, and debugging any issues that may arise.
3. --ckpt_path: It is a string parameter with a default value of None. If you want to resume training from a previously saved checkpoint, you can specify the path to the checkpoint file using this parameter. If set to None, the training will start from scratch.
4. --checkpoint_dir: This is a string parameter with a default value of None. It specifies the directory where the checkpoints of the model will be saved during the training process. If set to None, checkpoints may not be saved or may be saved in a default location depending on the training framework.

6. Validation and Reloading Settings Copy Chapter

1. --reload_dataloaders_every_n_epochs: It is an integer parameter with a default value of 1. It determines how often the data loaders will be reloaded during the training process. Reloading the data loaders can be useful when you want to ensure that the data is shuffled or processed differently for each epoch, especially when dealing with datasets that may change or have some specific requirements.
2. --every_plot_step: It is an integer parameter with a default value of 2000. It specifies how often some visualizations or plots (such as loss curves, accuracy plots, etc.) will be generated during the training process. For example, with a value of 2000, the plots will be updated every 2000 training steps.
3. --val_check_interval: This is an integer parameter with a default value of None. It determines how often the validation process will be performed during the training. If set to a positive integer, the model will be evaluated on the validation dataset every specified number of steps. If set to None, no regular validation checks will be performed.
4. --lora_config_path: It is a string parameter with a default value of "config/zh_rap_lora_config.json". This parameter specifies the path to the configuration file for the Lora (Low-Rank Adaptation) module. The Lora configuration file contains settings related to the Lora module, such as the rank of the low-rank matrices, the learning rate for the Lora parameters, etc.

File: ZH_RAP_LORA.md Copy Chapter

🎤 RapMachine Release

We meticulously curated and trained this model on Chinese rap/hip-hop datasets, with rigorous data cleaning and recaptioning. The results include:

• Improved pronunciation for Chinese lyrics
• Enhanced adherence to hip-hop and electronic music styles
• Greater diversity in hip-hop vocal performances

**How to Use** Copy Section

1. Generate higher-quality Chinese songs (⚠️ It's not just for Chinese songs. You can also use it in other ways. )

2. Create better hip-hop tracks

3. Blend it with other genres to:

   • Produce music with richer vocal details
   • Experiment with underground or street culture flavors

4. Fine-tune outputs using the following dimensions:

   • vocal_timbre: Describes the inherent qualities of the voice.

     • Examples: Bright, dark, warm, cold, breathy, nasal, gritty, smooth, husky, metallic, whispery, resonant, airy, smoky, sultry, light, clear, high-pitched, raspy, powerful, ethereal, flute-like, hollow, velvety, shrill, hoarse, mellow, thin, thick, reedy, silvery, twangy.

   • techniques:

     • Examples:
       • Rap styles: mumble rap, chopper rap, melodic rap, lyrical rap, trap flow, double-time rap
       • Vocal effects: auto-tune, reverb, delay, distortion
       • Delivery styles: whispered, shouted, spoken word, narration, singing
       • Other vocalizations: ad-libs, call-and-response, harmonized

Community Note Copy Chapter

We’ve revamped and expanded the LoRA training guide with finer details. This release is a proof of concept—showcasing the potential of ACE-Step.

While a Chinese rap LoRA might seem niche for non-Chinese communities, we consistently demonstrate through such projects that ACE-step - as a music generation foundation model - holds boundless potential. It doesn't just improve pronunciation in one language, but spawns new styles.

The universal human appreciation of music is a precious asset. Like abstract LEGO blocks, these elements will eventually combine in more organic ways. May our open-source contributions propel the evolution of musical history forward.

Enjoy it, customize it, and create something entirely new.

We can’t wait to hear what you’ll build!

File: acestep/models/config.json Copy Chapter

{
"_class_name": "Transformer2DModel",
"_diffusers_version": "0.27.2",
"in_channels": 8,
"num_layers": 24,
"inner_dim": 2560,
"attention_head_dim": 128,
"num_attention_heads": 20,
"mlp_ratio": 2.5,
"out_channels": 8,
"max_position": 32768,
"rope_theta": 1000000.0,
"speaker_embedding_dim": 512,
"text_embedding_dim": 768,
"ssl_encoder_depths": [8, 8],
"ssl_names": ["mert", "m-hubert"],
"ssl_latent_dims": [1024, 768],
"patch_size": [16, 1],
"max_height": 16,
"max_width": 32768,
"lyric_encoder_vocab_size": 6693,
"lyric_hidden_size": 1024
}

File: assets/Ace_Step_4x_a2a.json Copy Chapter

{
"id": "88ac5dad-efd7-40bb-84fe-fbaefdee1fa9",
"revision": 0,
"last_node_id": 63,
"last_link_id": 173,
"nodes": [
{
"id": 45,
"type": "ModelSamplingSD3",
"pos": [
542.9647827148438,
-90.26543426513672
],
"size": [
270,
58
],
"flags": {},
"order": 3,
"mode": 0,
"inputs": [
{
"name": "model",
"type": "MODEL",
"link": 111
}
],
"outputs": [
{
"name": "MODEL",
"type": "MODEL",
"links": [
112,
116,
123,
132
]
}
],
"properties": {
"Node name for S&R": "ModelSamplingSD3"
},
"widgets_values": [
4.000000000000001
]
},
{
"id": 18,
"type": "VAEDecodeAudio",
"pos": [
886.7603149414062,
-89.60011291503906
],
"size": [
150.93612670898438,
46
],
"flags": {},
"order": 9,
"mode": 0,
"inputs": [
{
"name": "samples",
"type": "LATENT",
"link": 101
},
{
"name": "vae",
"type": "VAE",
"link": 83
}
],
"outputs": [
{
"name": "AUDIO",
"type": "AUDIO",
"links": [
26
]
}
],
"properties": {
"Node name for S&R": "VAEDecodeAudio"
},
"widgets_values": []
},
{
"id": 47,
"type": "VAEDecodeAudio",
"pos": [
876.7603759765625,
180.3998260498047
],
"size": [
150.93612670898438,
46
],
"flags": {},
"order": 12,
"mode": 0,
"inputs": [
{
"name": "samples",
"type": "LATENT",
"link": 117
},
{
"name": "vae",
"type": "VAE",
"link": 118
}
],
"outputs": [
{
"name": "AUDIO",
"type": "AUDIO",
"links": [
120
]
}
],
"properties": {
"Node name for S&R": "VAEDecodeAudio"
},
"widgets_values": []
},
{
"id": 54,
"type": "VAEDecodeAudio",
"pos": [
886.7603149414062,
530.3997192382812
],
"size": [
150.93612670898438,
46
],
"flags": {},
"order": 15,
"mode": 0,
"inputs": [
{
"name": "samples",
"type": "LATENT",
"link": 147
},
{
"name": "vae",
"type": "VAE",
"link": 128
}
],
"outputs": [
{
"name": "AUDIO",
"type": "AUDIO",
"links": [
130
]
}
],
"properties": {
"Node name for S&R": "VAEDecodeAudio"
},
"widgets_values": []
},
{
"id": 40,
"type": "CheckpointLoaderSimple",
"pos": [
83.06548309326172,
-53.6221923828125
],
"size": [
375,
98
],
"flags": {},
"order": 0,
"mode": 0,
"inputs": [],
"outputs": [
{
"name": "MODEL",
"type": "MODEL",
"links": [
111
]
},
{
"name": "CLIP",
"type": "CLIP",
"links": [
80
]
},
{
"name": "VAE",
"type": "VAE",
"links": [
83,
118,
128,
137,
149
]
}
],
"properties": {
"Node name for S&R": "CheckpointLoaderSimple"
},
"widgets_values": [
"ace_step_v1_3.5b.safetensors"
]
},
{
"id": 57,
"type": "VAEDecodeAudio",
"pos": [
886.7603149414062,
800.3997802734375
],
"size": [
150.93612670898438,
46
],
"flags": {},
"order": 18,
"mode": 0,
"inputs": [
{
"name": "samples",
"type": "LATENT",
"link": 136
},
{
"name": "vae",
"type": "VAE",
"link": 137
}
],
"outputs": [
{
"name": "AUDIO",
"type": "AUDIO",
"links": [
168
]
}
],
"properties": {
"Node name for S&R": "VAEDecodeAudio"
},
"widgets_values": []
},
{
"id": 62,
"type": "Reroute",
"pos": [
80,
60
],
"size": [
75,
26
],
"flags": {},
"order": 7,
"mode": 0,
"inputs": [
{
"name": "",
"type": "",
"link": 173
}
],
"outputs": [
{
"name": "",
"type": "LATENT",
"links": [
172
]
}
],
"properties": {
"showOutputText": false,
"horizontal": false
}
},
{
"id": 60,
"type": "VAEEncodeAudio",
"pos": [
-122.92820739746094,
7.071800231933594
],
"size": [
150.13333129882812,
46
],
"flags": {},
"order": 5,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 150
},
{
"name": "vae",
"type": "VAE",
"link": 149
}
],
"outputs": [
{
"name": "LATENT",
"type": "LATENT",
"links": [
173
]
}
],
"properties": {
"Node name for S&R": "VAEEncodeAudio"
},
"color": "#223",
"bgcolor": "#335"
},
{
"id": 17,
"type": "EmptyAceStepLatentAudio",
"pos": [
-245.60769653320312,
112.92820739746094
],
"size": [
270,
82
],
"flags": {},
"order": 1,
"mode": 0,
"inputs": [],
"outputs": [
{
"name": "LATENT",
"type": "LATENT",
"links": []
}
],
"properties": {
"Node name for S&R": "EmptyAceStepLatentAudio"
},
"widgets_values": [
69,
1
],
"color": "#223",
"bgcolor": "#335"
},
{
"id": 44,
"type": "ConditioningZeroOut",
"pos": [
291.3594665527344,
486.2049865722656
],
"size": [
197.712890625,
26
],
"flags": {},
"order": 6,
"mode": 0,
"inputs": [
{
"name": "conditioning",
"type": "CONDITIONING",
"link": 108
}
],
"outputs": [
{
"name": "CONDITIONING",
"type": "CONDITIONING",
"links": [
109,
114,
125,
134
]
}
],
"properties": {
"Node name for S&R": "ConditioningZeroOut"
},
"widgets_values": []
},
{
"id": 46,
"type": "KSampler",
"pos": [
526.7609252929688,
330.3997802734375
],
"size": [
315,
262
],
"flags": {},
"order": 10,
"mode": 0,
"inputs": [
{
"name": "model",
"type": "MODEL",
"link": 116
},
{
"name": "positive",
"type": "CONDITIONING",
"link": 115
},
{
"name": "negative",
"type": "CONDITIONING",
"link": 114
},
{
"name": "latent_image",
"type": "LATENT",
"link": 169
}
],
"outputs": [
{
"name": "LATENT",
"type": "LATENT",
"slot_index": 0,
"links": [
117,
170
]
}
],
"properties": {
"Node name for S&R": "KSampler"
},
"widgets_values": [
701612340335405,
"increment",
30,
6,
"euler",
"sgm_uniform",
0.20000000000000004
]
},
{
"id": 53,
"type": "KSampler",
"pos": [
526.7609252929688,
640.3997802734375
],
"size": [
315,
262
],
"flags": {},
"order": 13,
"mode": 0,
"inputs": [
{
"name": "model",
"type": "MODEL",
"link": 123
},
{
"name": "positive",
"type": "CONDITIONING",
"link": 124
},
{
"name": "negative",
"type": "CONDITIONING",
"link": 125
},
{
"name": "latent_image",
"type": "LATENT",
"link": 170
}
],
"outputs": [
{
"name": "LATENT",
"type": "LATENT",
"slot_index": 0,
"links": [
147,
171
]
}
],
"properties": {
"Node name for S&R": "KSampler"
},
"widgets_values": [
801612340336968,
"increment",
30,
6,
"euler",
"sgm_uniform",
0.20000000000000004
]
},
{
"id": 56,
"type": "KSampler",
"pos": [
526.7609252929688,
960.3997802734375
],
"size": [
315,
262
],
"flags": {},
"order": 16,
"mode": 0,
"inputs": [
{
"name": "model",
"type": "MODEL",
"link": 132
},
{
"name": "positive",
"type": "CONDITIONING",
"link": 133
},
{
"name": "negative",
"type": "CONDITIONING",
"link": 134
},
{
"name": "latent_image",
"type": "LATENT",
"link": 171
}
],
"outputs": [
{
"name": "LATENT",
"type": "LATENT",
"slot_index": 0,
"links": [
136
]
}
],
"properties": {
"Node name for S&R": "KSampler"
},
"widgets_values": [
205612340336968,
"increment",
30,
6,
"euler",
"sgm_uniform",
0.20000000000000004
]
},
{
"id": 3,
"type": "KSampler",
"pos": [
524.7819213867188,
24.392868041992188
],
"size": [
315,
262
],
"flags": {},
"order": 8,
"mode": 0,
"inputs": [
{
"name": "model",
"type": "MODEL",
"link": 112
},
{
"name": "positive",
"type": "CONDITIONING",
"link": 110
},
{
"name": "negative",
"type": "CONDITIONING",
"link": 109
},
{
"name": "latent_image",
"type": "LATENT",
"link": 172
}
],
"outputs": [
{
"name": "LATENT",
"type": "LATENT",
"slot_index": 0,
"links": [
101,
169
]
}
],
"properties": {
"Node name for S&R": "KSampler"
},
"widgets_values": [
654498564630224,
"increment",
30,
6,
"euler",
"sgm_uniform",
0.4000000000000001
]
},
{
"id": 59,
"type": "LoadAudio",
"pos": [
-547.00732421875,
44.58967590332031
],
"size": [
274,
136
],
"flags": {},
"order": 2,
"mode": 0,
"inputs": [],
"outputs": [
{
"name": "AUDIO",
"type": "AUDIO",
"links": [
150
]
}
],
"properties": {
"Node name for S&R": "LoadAudio"
},
"widgets_values": [
"The_Gael_Synth.mp3",
null,
""
],
"color": "#432",
"bgcolor": "#653"
},
{
"id": 14,
"type": "TextEncodeAceStepAudio",
"pos": [
81.86527252197266,
132.01605224609375
],
"size": [
410.834716796875,
305.39215087890625
],
"flags": {},
"order": 4,
"mode": 0,
"inputs": [
{
"name": "clip",
"type": "CLIP",
"link": 80
}
],
"outputs": [
{
"name": "CONDITIONING",
"type": "CONDITIONING",
"links": [
108,
110,
115,
124,
133
]
}
],
"properties": {
"Node name for S&R": "TextEncodeAceStepAudio"
},
"widgets_values": [
"techno, eurodance, melodic, dark synth, dark soundscape",
""
],
"color": "#232",
"bgcolor": "#353"
},
{
"id": 19,
"type": "SaveAudio",
"pos": [
-662.6826782226562,
248.9967041015625
],
"size": [
725.9335327148438,
112
],
"flags": {},
"order": 11,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 26
}
],
"outputs": [],
"properties": {},
"widgets_values": [
"audio/asdf"
]
},
{
"id": 61,
"type": "SaveAudio",
"pos": [
-649.760498046875,
716.68798828125
],
"size": [
705.8839721679688,
112
],
"flags": {},
"order": 19,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 168
}
],
"outputs": [],
"properties": {},
"widgets_values": [
"audio/asdf"
]
},
{
"id": 49,
"type": "SaveAudio",
"pos": [
-654.6276245117188,
408.31817626953125
],
"size": [
715.1640014648438,
112
],
"flags": {},
"order": 14,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 120
}
],
"outputs": [],
"properties": {},
"widgets_values": [
"audio/asdf"
]
},
{
"id": 55,
"type": "SaveAudio",
"pos": [
-659.52783203125,
558.103515625
],
"size": [
723.4810180664062,
112
],
"flags": {},
"order": 17,
"mode": 0,
"inputs": [
{
"name": "audio",
"type": "AUDIO",
"link": 130
}
],
"outputs": [],
"properties": {},
"widgets_values": [
"audio/asdf"
]
}
],
"links": [
[
26,
18,
0,
19,
0,
"AUDIO"
],
[
80,
40,
1,
14,
0,
"CLIP"
],
[
83,
40,
2,
18,
1,
"VAE"
],
[
101,
3,
0,
18,
0,
"LATENT"
],
[
108,
14,
0,
44,
0,
"CONDITIONING"
],
[
109,
44,
0,
3,
2,
"CONDITIONING"
],
[
110,
14,
0,
3,
1,
"CONDITIONING"
],
[
111,
40,
0,
45,
0,
"MODEL"
],
[
112,
45,
0,
3,
0,
"MODEL"
],
[
114,
44,
0,
46,
2,
"CONDITIONING"
],
[
115,
14,
0,
46,
1,
"CONDITIONING"
],
[
116,
45,
0,
46,
0,
"MODEL"
],
[
117,
46,
0,
47,
0,
"LATENT"
],
[
118,
40,
2,
47,
1,
"VAE"
],
[
120,
47,
0,
49,
0,
"AUDIO"
],
[
123,
45,
0,
53,
0,
"MODEL"
],
[
124,
14,
0,
53,
1,
"CONDITIONING"
],
[
125,
44,
0,
53,
2,
"CONDITIONING"
],
[
128,
40,
2,
54,
1,
"VAE"
],
[
130,
54,
0,
55,
0,
"AUDIO"
],
[
132,
45,
0,
56,
0,
"MODEL"
],
[
133,
14,
0,
56,
1,
"CONDITIONING"
],
[
134,
44,
0,
56,
2,
"CONDITIONING"
],
[
136,
56,
0,
57,
0,
"LATENT"
],
[
137,
40,
2,
57,
1,
"VAE"
],
[
147,
53,
0,
54,
0,
"LATENT"
],
[
149,
40,
2,
60,
1,
"VAE"
],
[
150,
59,
0,
60,
0,
"AUDIO"
],
[
168,
57,
0,
61,
0,
"AUDIO"
],
[
169,
3,
0,
46,
3,
"LATENT"
],
[
170,
46,
0,
53,
3,
"LATENT"
],
[
171,
53,
0,
56,
3,
"LATENT"
],
[
172,
62,
0,
3,
3,
"LATENT"
],
[
173,
60,
0,
62,
0,
""
]
],
"groups": [],
"config": {},
"extra": {
"frontendVersion": "1.18.9",
"VHS_latentpreview": false,
"VHS_latentpreviewrate": 0
},
"version": 0.4
}

File: config/zh_rap_lora_config.json Copy Chapter

{
"r": 256,
"lora_alpha": 32,
"target_modules": [
"speaker_embedder",
"linear_q",
"linear_k",
"linear_v",
"to_q",
"to_k",
"to_v",
"to_out.0"
],
"use_rslora": true
}

File: data/test_track_001_lyrics.txt Copy Chapter

[Intro]
"System booting... 语言 模型 loading..."

[Verse 1]
硅谷 那个 coder 调试 neural network
北京 的 极客 训练 A I 写 report
不同 架构 的 chip 不同 算法 的 war
屏幕上 跑的 全是 machine learning (learning)

[Bridge]
多少年 我们 chase 摩尔 定律 的 trend (yeah)
这两年 换他们 study 中文 N L P
Convolution L S T M
好烧脑 的 backprop 好暴力 的 big data

[Verse 2]
Python 强 say加加 刚 Python 调用 C++ 的 A P I
say加加 嫌 Python 太 slow Python 笑 C++ 太 hardcore
L L V M 默默 generate 中间 code
到底 interpreter 还是 compiler 屌？

[Verse 3]
P M 和 engineer
白板 画满 flow chart 服务器 闪着 red light
P M 说 add feature engineer 说 no way
需求 变更 code 重构
不知 是 P M 太 fly 还是 deadline 太 high

[Chorus]
全世界 都在 train neural network
Transformer 的 paper 越来越 难 go through
全世界 都在 tune 超参数
我们 写的 bug 让 G P U 都 say no

[Verse 4]
柏林 hackathon demo blockchain contract
上海 的 dev 用 federated learning 破 data wall
各种 语言 的 error 各种 框架 的 doc
terminal 里 滚的 全是 dependency 冲突

[Bridge]
曾以为 English 才是 coding 的 language (yeah)
直到见 G P T 用 文言文 generate 正则 expression
Gradient explode
好硬核 的 prompt 好头秃 的 debug road

[Verse 5]
有个 bug 叫 quantum
测试 环境 run perfect 上线 立即就 crash
查 log 看 monitor 发现是 thread 不同步
改 sync 加 lock 慢 deadlock 更难办
量子 computer 也解不开 这 chaos chain

[Verse 6]
你说 996 我说 007
你说 福报 我说 burnout
Product 要 agile Boss 要 KPI
Code 要 elegant deadline 是 tomorrow
不如 直接 script 自动 submit 离职信

[Outro]
"Warning: 内存 leak...core dumping..."
全世界 都在 train neural network (neural network)
Loss 还没 converge 天已经亮
全世界 都在 tune 超参数
我们 写的 code (让它) 让 world (reboot) 都 reboot 无效

File: data/test_track_001_prompt.txt Copy Chapter

articulate, spoken word, young adult, rap music, female, clear, energetic, warm

File: examples/default/input_params/output_20250426071706_0_input_params.json Copy Chapter

{
"prompt": "pop, rap, electronic, blues, hip-house, rhythm and blues",
"lyrics": "[verse]\n我走过深夜的街道\n冷风吹乱思念的漂亮外套\n你的微笑像星光很炫耀\n照亮了我孤独的每分每秒\n\n[chorus]\n愿你是风吹过我的脸\n带我飞过最远最遥远的山间\n愿你是风轻触我的梦\n停在心头不再飘散无迹无踪\n\n[verse]\n一起在喧哗避开世俗的骚动\n独自在天台探望月色的朦胧\n你说爱像音乐带点重节奏\n一拍一跳让我忘了心的温度多空洞\n\n[bridge]\n唱起对你的想念不隐藏\n像诗又像画写满藏不了的渴望\n你的影子挥不掉像风的倔强\n追着你飞扬穿越云海一样泛光\n\n[chorus]\n愿你是风吹过我的手\n暖暖的触碰像春日细雨温柔\n愿你是风盘绕我的身\n深情万万重不会有一天走远走\n\n[verse]\n深夜的钢琴弹起动人的旋律\n低音鼓砸进心底的每一次呼吸\n要是能将爱化作歌声传递\n你是否会听见我心里的真心实意",
"audio_duration": 170.63997916666668,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 3.191075086593628,
"diffusion": 17.459356784820557,
"latent2audio": 1.7095518112182617
},
"actual_seeds": [
3299954530
]
}

File: examples/default/input_params/output_20250426071812_0_input_params.json Copy Chapter

{
"prompt": "country rock, folk rock, southern rock, bluegrass, country pop",
"lyrics": "[verse]\nWoke up to the sunrise glow\nTook my heart and hit the road\nWheels hummin' the only tune I know\nStraight to where the wildflowers grow\n\n[verse]\nGot that old map all wrinkled and torn\nDestination unknown but I'm reborn\nWith a smile that the wind has worn\nChasin' dreams that can't be sworn\n\n[chorus]\nRidin' on a highway to sunshine\nGot my shades and my radio on fine\nLeave the shadows in the rearview rhyme\nHeart's racing as we chase the time\n\n[verse]\nMet a girl with a heart of gold\nTold stories that never get old\nHer laugh like a tale that's been told\nA melody so bold yet uncontrolled\n\n[bridge]\nClouds roll by like silent ghosts\nAs we drive along the coast\nWe toast to the days we love the most\nFreedom's song is what we post\n\n[chorus]\nRidin' on a highway to sunshine\nGot my shades and my radio on fine\nLeave the shadows in the rearview rhyme\nHeart's racing as we chase the time",
"audio_duration": 224.23997916666667,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 4.262240648269653,
"diffusion": 15.380569219589233,
"latent2audio": 2.3227272033691406
},
"actual_seeds": [
401640
]
}

File: examples/default/input_params/output_20250426072346_0_input_params.json Copy Chapter

[Binary file]

File: examples/default/input_params/output_20250426072508_0_input_params.json Copy Chapter

{
"prompt": "funk, pop, soul, rock, melodic, guitar, drums, bass, keyboard, percussion, 105 BPM, energetic, upbeat, groovy, vibrant, dynamic",
"lyrics": "[verse]\nNeon lights they flicker bright\nCity hums in dead of night\nRhythms pulse through concrete veins\nLost in echoes of refrains\n\n[verse]\nBassline groovin' in my chest\nHeartbeats match the city's zest\nElectric whispers fill the air\nSynthesized dreams everywhere\n\n[chorus]\nTurn it up and let it flow\nFeel the fire let it grow\nIn this rhythm we belong\nHear the night sing out our song\n\n[verse]\nGuitar strings they start to weep\nWake the soul from silent sleep\nEvery note a story told\nIn this night we’re bold and gold\n\n[bridge]\nVoices blend in harmony\nLost in pure cacophony\nTimeless echoes timeless cries\nSoulful shouts beneath the skies\n\n[verse]\nKeyboard dances on the keys\nMelodies on evening breeze\nCatch the tune and hold it tight\nIn this moment we take flight",
"audio_duration": 178.87997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.02882218360900879,
"diffusion": 16.91233205795288,
"latent2audio": 1.7794082164764404
},
"actual_seeds": [
401640
]
}

File: examples/default/input_params/output_20250426073829_0_input_params.json Copy Chapter

{
"prompt": "electronic rap",
"lyrics": "[verse]\nWaves on the bass, pulsing in the speakers,\nTurn the dial up, we chasing six-figure features,\nGrinding on the beats, codes in the creases,\nDigital hustler, midnight in sneakers.\n\n[chorus]\nElectro vibes, hearts beat with the hum,\nUrban legends ride, we ain't ever numb,\nCircuits sparking live, tapping on the drum,\nLiving on the edge, never succumb.\n\n[verse]\nSynthesizers blaze, city lights a glow,\nRhythm in the haze, moving with the flow,\nSwagger on stage, energy to blow,\nFrom the blocks to the booth, you already know.\n\n[bridge]\nNight's electric, streets full of dreams,\nBass hits collective, bursting at seams,\nHustle perspective, all in the schemes,\nRise and reflective, ain't no in-betweens.\n\n[verse]\nVibin' with the crew, sync in the wire,\nGot the dance moves, fire in the attire,\nRhythm and blues, soul's our supplier,\nRun the digital zoo, higher and higher.\n\n[chorus]\nElectro vibes, hearts beat with the hum,\nUrban legends ride, we ain't ever numb,\nCircuits sparking live, tapping on the drum,\nLiving on the edge, never succumb.",
"audio_duration": 221.42547916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.024875164031982422,
"diffusion": 20.566852569580078,
"latent2audio": 2.2281734943389893
},
"actual_seeds": [
401640
]
}

File: examples/default/input_params/output_20250426074037_0_input_params.json Copy Chapter

{
"prompt": "electronic, house, electro house, synthesizer, drums, bass, percussion, fast, energetic, uplifting, exciting",
"lyrics": "[verse]\n霓虹灯下我们追逐\n人群跃动像潮水满布\n热浪袭来吹散孤独\n跳进节奏不如停下脚步\n\n[pre-chorus]\n脚尖触电快点感受\n迎着风声释放自由\n心跳节拍配合节奏\n一切烦恼请靠边游\n\n[chorus]\n夏夜狂奔没有尽头\n星光闪烁舞池不朽\n尽情挥洒所有节奏\n无边热情把你包裹哦\n\n[verse]\n天空翻滚黑云入夜\n每颗星星像音乐律贴\n耳边回响那低音线\n环绕耳际如梦境般甜\n\n[pre-chorus]\n脚尖触电快点感受\n迎着风声释放自由\n心跳节拍配合节奏\n一切烦恼请靠边游\n\n[chorus]\n夏夜狂奔没有尽头\n星光闪烁舞池不朽\n尽情挥洒所有节奏\n无边热情把你包裹哦",
"audio_duration": 221.47997916666668,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.028400182723999023,
"diffusion": 13.195815324783325,
"latent2audio": 2.1679723262786865
},
"actual_seeds": [
3440445703
]
}

File: examples/default/input_params/output_20250426074214_0_input_params.json Copy Chapter

{
"prompt": "synth-pop, electronic, pop, synthesizer, drums, bass, piano, 128 BPM, energetic, uplifting, modern",
"lyrics": "[verse]\nWoke up in a city that's always alive\nNeon lights they shimmer they thrive\nElectric pulses beat they drive\nMy heart races just to survive\n\n[chorus]\nOh electric dreams they keep me high\nThrough the wires I soar and fly\nMidnight rhythms in the sky\nElectric dreams together we’ll defy\n\n[verse]\nLost in the labyrinth of screens\nVirtual love or so it seems\nIn the night the city gleams\nDigital faces haunted by memes\n\n[chorus]\nOh electric dreams they keep me high\nThrough the wires I soar and fly\nMidnight rhythms in the sky\nElectric dreams together we’ll defy\n\n[bridge]\nSilent whispers in my ear\nPixelated love serene and clear\nThrough the chaos find you near\nIn electric dreams no fear\n\n[verse]\nBound by circuits intertwined\nLove like ours is hard to find\nIn this world we’re truly blind\nBut electric dreams free the mind",
"audio_duration": 221.27997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.025463581085205078,
"diffusion": 15.243804454803467,
"latent2audio": 2.170398473739624
},
"actual_seeds": [
3400270027
]
}

File: examples/default/input_params/output_20250426074413_0_input_params.json Copy Chapter

{
"prompt": "Cuban music, salsa, son, Afro-Cuban, traditional Cuban",
"lyrics": "[verse]\nSun dips low the night ignites\nBassline hums with gleaming lights\nElectric guitar singing tales so fine\nIn the rhythm we all intertwine\n\n[verse]\nDrums beat steady calling out\nPercussion guides no room for doubt\nElectric pulse through every vein\nDance away every ounce of pain\n\n[chorus]\nFeel the rhythm feel the flow\nLet the music take control\nBassline deep electric hum\nIn this night we're never numb\n\n[bridge]\nStars above they start to glow\nEchoes of the night's soft glow\nElectric strings weave through the air\nIn this moment none compare\n\n[verse]\nHeartbeats sync with every tone\nLost in music never alone\nElectric tales of love and peace\nIn this groove we find release\n\n[chorus]\nFeel the rhythm feel the flow\nLet the music take control\nBassline deep electric hum\nIn this night we're never numb",
"audio_duration": 208.27997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.026132583618164062,
"diffusion": 15.139378070831299,
"latent2audio": 2.2071540355682373
},
"actual_seeds": [
3358899399
]
}

File: examples/default/input_params/output_20250426075107_0_input_params.json Copy Chapter

{
"prompt": "pop, piano, rap, dark, atmospheric",
"lyrics": "[verse]\n月光爬上窗 染白冷的床\n心跳的方向 带我入迷惘\n黑夜吞噬光 命运的纸张\n爱是血色霜 邪恶又芬芳\n\n[chorus]\n你是猎人的欲望 我是迷途的小羊\n深陷你眼眸的荒 唐突献出心脏\n我在夜里回荡 是谁给我希望\n黑暗风中飘荡 假装不再受伤\n\n[verse]\n心锁在门外 谁会解开关怀\n温柔的手拍 藏着冷酷杀害\n思绪如尘埃 撞击爱的霹雳\n灵魂的独白 为你沾满血迹\n\n[bridge]\n你是噩梦的歌唱 是灵魂的捆绑\n绝望中带着光 悬崖边的渴望\n心跳被你鼓掌 恶魔也痴痴想\n渐渐没了抵抗 古老诡计流淌\n\n[chorus]\n你是猎人的欲望 我是迷途的小羊\n深陷你眼眸的荒 唐突献出心脏\n我在夜里回荡 是谁给我希望\n黑暗风中飘荡 假装不再受伤\n\n[outro]\n爱如月黑无光 渗进梦的战场\n逃入无声的场 放手或心嚷嚷\n隐秘的极端 爱是极致风浪\n灵魂彻底交偿 你是终极虚妄",
"audio_duration": 146.91997916666668,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.03876018524169922,
"diffusion": 15.962624549865723,
"latent2audio": 1.4594337940216064
},
"actual_seeds": [
2065110378
]
}

File: examples/default/input_params/output_20250426075537_0_input_params.json Copy Chapter

{
"prompt": "surf music",
"lyrics": "[verse]\nSunshine on the boulevard the beach is calling loud\nWaves are dancing golden sand under a cotton cloud\nElectric heartbeat pounding fast the tide is on our side\nCatch a wave and feel alive we’ll take it for a ride\n\n[verse]\nPalm trees swaying left to right they know where we belong\nFeel the rhythm of the night it keeps us moving strong\nSea spray kisses salty air we’re flying with the breeze\nChampagne states of mind we ride we do just as we please\n\n[chorus]\nWe’re riding waves of life together hand in hand\nWith every beat we chase the beat it’s our own wonderland\nFeel the music take you higher as the shorelines blur\nThis is our world our endless summer as we live and learn\n\n[bridge]\nMoonlight paints the ocean blue reflections in our eyes\nStars align to light our path we’re surfing through the skies\nEvery moment like a song we sing it loud and clear\nEvery day’s a new adventure with you always near\n\n[verse]\nNeon lights and city sounds they blend with ocean views\nWe’re unstoppable tonight no way that we can lose\nDreams are written in the sand they sparkle in the sun\nTogether we’re a masterpiece our story’s just begun\n\n[chorus]\nWe’re riding waves of life together hand in hand\nWith every beat we chase the beat it’s our own wonderland\nFeel the music take you higher as the shorelines blur\nThis is our world our endless summer as we live and learn",
"audio_duration": 236.55997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.033666133880615234,
"diffusion": 16.291455507278442,
"latent2audio": 2.3726775646209717
},
"actual_seeds": [
508630535
]
}

File: examples/default/input_params/output_20250426075843_0_input_params.json Copy Chapter

{
"prompt": "alternative rock, pop, rock",
"lyrics": "[verse]\nBright lights flashing in the city sky\nRunning fast and we don't know why\nElectric nights got our hearts on fire\nChasing dreams we'll never tire\n\n[verse]\nGrit in our eyes wind in our hair\nBreaking rules we don't even care\nShouting loud above the crowd\nLiving life like we're unbowed\n\n[chorus]\nRunning wild in the night so free\nFeel the beat pumping endlessly\nHearts collide in the midnight air\nWe belong we don't have a care\n\n[verse]\nPiercing through like a lightning strike\nEvery moment feels like a hike\nDaring bold never backing down\nKings and queens without a crown\n\n[chorus]\nRunning wild in the night so free\nFeel the beat pumping endlessly\nHearts collide in the midnight air\nWe belong we don't have a care\n\n[bridge]\nClose your eyes let your spirit soar\nWe are the ones who wanted more\nBreaking chains of the mundane\nIn this world we'll make our claim",
"audio_duration": 202.19997916666668,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.02512216567993164,
"diffusion": 18.860822677612305,
"latent2audio": 2.0361969470977783
},
"actual_seeds": [
1255121549
]
}

File: examples/default/input_params/output_20250426080234_0_input_params.json Copy Chapter

{
"prompt": "rock, hip - hop, orchestral, bass, drums, electric guitar, piano, synthesizer, violin, viola, cello, fast, energetic, motivational, inspirational, empowering",
"lyrics": "### [Intro – Spoken] \n*"The streets whisper, their echoes never fade. \nEvery step I take leaves a mark—this ain't just a game."* \n\n### [Hook/Chorus] \nBorn in the chaos, I weather the storm, \nRising from ashes where warriors are born. \nChains couldn't hold me, the system’s a maze, \nI rewrite the rules, set the city ablaze! \n\n### [Verse 1] \nCold nights, empty pockets, dreams laced with fight, \nEvery loss made me sharper, cut deep like a knife. \nThey said I wouldn’t make it, now they watch in despair, \nFrom the curb to the throne, took the pain, made it rare. \nEvery siren’s a melody, every alley holds a tale, \nRose from the shadows, left my name on the trail. \nStreetlights flicker like warnings in the haze, \nBut I move like a phantom, unfazed by the blaze. \n\n### [Hook/Chorus] \nBorn in the chaos, I weather the storm, \nRising from ashes where warriors are born. \nChains couldn't hold me, the system’s a maze, \nI rewrite the rules, set the city ablaze! \n\n### [Verse 2] \nBarbed wire fences couldn't lock in my mind, \nEvery cage they designed, I left broken behind. \nThey want control, but I’m destined to roam, \nWhere the lost find their voice, where the heart sets the tone. \nSteel and concrete, where the lessons run deep, \nEvery crack in the pavement tells a story of heat. \nBut I rise, undefeated, like a king with no throne, \nWriting scripts in the struggle, my legacy’s stone. \n\n### [Bridge] \nFeel the rhythm of the underground roar, \nEvery wound tells a story of the battles before. \nBlood, sweat, and echoes fill the cold midnight, \nBut we move with the fire—unshaken, upright. \n\n### [Verse 3] \nNo regrets, no retreat, this game has no pause, \nEvery step that I take is a win for the lost. \nI took lessons from hustlers, wisdom from pain, \nNow the echoes of struggle carve power in my name. \nThey built walls, but I walk through the cracks, \nTurned dirt into gold, never looked back. \nThrough the struggle we rise, through the fire we claim, \nThis is more than just music—it's life in the frame. \n\n### [Hook/Chorus – Reprise] \nBorn in the chaos, I weather the storm, \nRising from ashes where warriors are born. \nChains couldn't hold me, the system’s a maze, \nI rewrite the rules, set the city ablaze! \n\n### [Outro – Spoken] \n*"The scars, the struggle, the grind—it’s all part of the rhythm. \nWe never break, we never fold. We rise."*",
"audio_duration": 153.95997916666667,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.04368758201599121,
"diffusion": 17.16369390487671,
"latent2audio": 1.5405471324920654
},
"actual_seeds": [
2659225017
]
}

File: examples/default/input_params/output_20250426080407_0_input_params.json Copy Chapter

{
"prompt": "tango finlandés, campanas, disco, dark pop, electro, guitarra clásica, corridos tumba",
"lyrics": "[inst]",
"audio_duration": 162.79997916666667,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.011058568954467773,
"diffusion": 9.924944400787354,
"latent2audio": 1.6034839153289795
},
"actual_seeds": [
780297686
]
}

File: examples/default/input_params/output_20250426080601_0_input_params.json Copy Chapter

{
"prompt": "Nightclubs, dance parties, workout playlists, radio broadcasts",
"lyrics": "Burning in motion, set me alight!\nEvery heartbeat turns into a fight!\nCaged in rhythm, chained in time!\nLove’s a battle— You're Mine! You're Mine!",
"audio_duration": 221.83997916666667,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.012485980987548828,
"diffusion": 14.345409154891968,
"latent2audio": 2.174558639526367
},
"actual_seeds": [
1318394052
]
}

File: examples/default/input_params/output_20250426081134_0_input_params.json Copy Chapter

{
"prompt": "melancholic, world, sad, medieval, soulful",
"lyrics": "[Verse]\nIn a world so grand he roams the skies alone\nHis heart a heavy stone a tale untold\nWhispers of his past echo through the night\nA lonely dragon searching for the light\n\n[Verse 2]\nOnce a mighty force now he drifts in pain\nHis scales once shimmered now they're dark with shame\nCast out by his kin in shadows he does hide\nA haunting sorrow burns deep inside\n\n[Chorus]\nRoaming endless fields with no friend in sight\nHis roar a mournful cry beneath the moon's pale light\nTears fall like stars as he flies on his way\nA lonely dragon yearning for the break of day\n\n[Bridge]\nThe world turns cold the nights grow long\nIn his heart he carries an ancient song\nOf battles fought and love long gone\nA legend now but his soul is torn\n\n[Verse 3]\nHoping for a day he'll find a kindred soul\nTo share his pain and make him whole\nTill then he drifts a shadow in the sky\nA lonely dragon with tears in his eye\n\n[Chorus]\nRoaming endless fields with no friend in sight\nHis roar a mournful cry beneath the moon's pale light\nTears fall like stars as he flies on his way\nA lonely dragon yearning for the break of day",
"audio_duration": 239.99997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.029100656509399414,
"diffusion": 22.503791570663452,
"latent2audio": 2.3603708744049072
},
"actual_seeds": [
2166832218
]
}

File: examples/default/input_params/output_20250426091716_0_input_params.json Copy Chapter

[Binary file]

File: examples/default/input_params/output_20250426092025_0_input_params.json Copy Chapter

{
"prompt": "dark, death rock, metal, hardcore, electric guitar, powerful, bass, drums, 110 bpm, G major",
"lyrics": "[Verse]\nMy lovers betray me\nThe snake in my garden is hissing\nIn the air is the sweetness of roses\nAnd under my skin\nThere's a thorn\n\n[Verse 2]\nI should have known\nThat God sends his angel in shadows\nWith blood in his veins\nI watch the enemy\nGivin' me the hand of my savior\n\n[Chorus]\nAnd I can't love again\nWith the echo of your name in my head\nWith the demons in my bed\nWith the memories\nYour ghost\nI see it\n'Cause it comes to haunt me\nJust to taunt me\nIt comes to haunt me\nJust to taunt me\n\n[Verse 3]\nWith sugar and spice\nIt's hard to ignore the nostalgia\nWith the men on their knees\nAt the gates of my heart\nHow they beg me\n\n[Verse 4]\nThey say\n"No one will ever love you\nThe way that I do\nNo one will ever touch you\nThe way that I do"\n\n[Chorus]\nAnd I can't love again\nWith the echo of your name in my head\nWith the demons in my bed\nWith the memories\nYour ghost\nI see it\n'Cause it comes to haunt me\nJust to taunt me\nIt comes to haunt me\nJust to taunt me",
"audio_duration": 174.27997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 3.8372838497161865,
"diffusion": 13.039669275283813,
"latent2audio": 1.7923030853271484
},
"actual_seeds": [
4064916393
]
}

File: examples/default/input_params/output_20250426093007_0_input_params.json Copy Chapter

{
"prompt": "aggressive, Heavy Riffs, Blast Beats, Satanic Black Metal",
"lyrics": "[verse]\nFloating through the galaxy on a midnight ride\nStars are dancing all around in cosmic tides\nFeel the pulse of space and time beneath our feet\nEvery beat a heartbeat in this endless suite\n\n[chorus]\nGalactic dreams under neon lights\nSailing through the velvet nights\nWe are echoes in a cosmic sea\nIn a universe where we are free\n\n[verse]\nPlanetary whispers in the sky tonight\nEvery constellation's got a secret sight\nDistant worlds and moons we have yet to see\nIn the void of space where we can just be\n\n[bridge]\nAsteroids and comets in a ballet they spin\nLost in the rhythm of where our dreams begin\nClose your eyes and let the synths take flight\nWe're voyagers on an electric night\n\n[verse]\nLet the piano keys unlock the stars above\nEvery chord a memory every note is love\nIn this synth symphony we find our grace\nDrifting forever in this boundless space\n\n[chorus]\nGalactic dreams under neon lights\nSailing through the velvet nights\nWe are echoes in a cosmic sea\nIn a universe where we are free",
"audio_duration": 181.99997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.025065898895263672,
"diffusion": 17.176705837249756,
"latent2audio": 1.8225171566009521
},
"actual_seeds": [
1132623236
]
}

File: examples/default/input_params/output_20250426093146_0_input_params.json Copy Chapter

{
"prompt": "r&b, soul, funk/soul",
"lyrics": "[verse]\nDancing through electric fires\nHeart is buzzing like live wires\nIn your arms I find desire\nFeel the beat as we get higher\n\n[chorus]\nElectric love in the night sky\nWe’re gonna soar baby you and I\nDrop the bass let the rhythm fly\nFeel the heat and don't ask why\n\n[verse]\nWhisper secrets that make me blush\nUnder the neon city hush\nYour touch gives me such a rush\nTurn it up we're feeling lush\n\n[chorus]\nElectric love in the night sky\nWe’re gonna soar baby you and I\nDrop the bass let the rhythm fly\nFeel the heat and don't ask why\n\n[bridge]\nThrough the lights and the smoky haze\nI see you in a thousand ways\nLove's a script and we’re the play\nTurn the page stay till we sway\n\n[chorus]\nElectric love in the night sky\nWe’re gonna soar baby you and I\nDrop the bass let the rhythm fly\nFeel the heat and don't ask why",
"audio_duration": 195.15997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": true,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.025553464889526367,
"diffusion": 18.250118494033813,
"latent2audio": 1.9400627613067627
},
"actual_seeds": [
2853131993
]
}

File: examples/zh_rap_lora/input_params/output_20250512101839_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512114703_0_input_params.json Copy Chapter

{
"lora_name_or_path": "/root/sag_train/data/ace_step_v1_chinese_rap_lora",
"task": "text2music",
"prompt": "Chorus Hook, Melodic Rap, Ambient Synth Pads, adult, rap, Very Fast, Storytelling, Chinese Rap, male, spoken word, bright, energetic, Melodic Flow, clear, clarity, 130 bpm",
"lyrics": "[Intro]\n舌 头 打 结 了... 快 念 快 念...\n\n[Verse 1]\n这 个 赌 鬼 蹲 在 柜 台 啃 着 苦 瓜 干 快 很 干\n赌 桌 堆 满 骨 牌 古 怪 股 票 和 五 块 钢 镚 儿 钢 镚\n他 甩 出 扑 克 牌 啪 啪 啪 拍 扁 螃 蟹 壳 哦 壳 扁\n又 摸 摸 麻 将 摸 出 幺 鸡 摸 出 发 财 摸 出 一 条 蛇 蛇 蛇\n庄 家 咳 嗽 咳 破 锣 嗓 子 喊 开 开 开 快 开 开\n赌 鬼 咕 嘟 咕 嘟 灌 咖 啡 灌 到 筷 子 戳 穿 碗 快 戳 穿\n空 气 里 飘 着 锅 巴 味 混 合 隔 夜 的 酸 奶 罐 哦 酸\n输 光 裤 带 还 想 翻 盘 翻 成 煎 饼 摊 老 板 快 翻 盘\n\n[Chorus]\n赌 鬼 赌 鬼 哦 赌 鬼 赌 鬼 快 很 快\n舌 头 打 结 着 念 这 段 哦 这 段 绕 口 令 牌\n若 念 错 一 字 就 罚 你 哦 罚 你 吞 十 斤 海 带\n赌 场 规 矩 就 是 绕 晕 你 哦 绕 晕 你 快 很 快\n\n[Verse 2]\n他 掏 出 铜 板 抠 出 口 袋 最 后 一 颗 快 很 颗\n庄 家 哗 啦 哗 啦 摇 骰 子 摇 出 三 点 又 三 点 哦 三 点\n赌 鬼 急 得 咬 牙 切 齿 咬 到 舌 头 打 蝴 蝶 结 快 打 结\n还 想 押 上 祖 传 的 拖 鞋 拖 把 铁 锅 和 半 包 盐 盐 盐\n突 然 警 笛 嘀 嘟 嘀 嘟 吓 得 他 钻 进 垃 圾 罐 哦 垃 圾\n警 察 咔 嚓 咔 嚓 拍 照 拍 到 他 头 顶 菠 菜 叶 快 拍 照\n最 后 赌 鬼 蹲 监 狱 天 天 背 这 首 绕 口 令 哦 背 不 完\n若 背 错 一 句 就 加 刑 十 年 再 加 十 年 快 加 刑\n\n[Outro]\n舌 头 打 结 了... 赌 鬼 哭 了 哦...\n这 首 歌... 绕 死 人 了 哦...",
"audio_duration": 186.59997916666666,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.7,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": false,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.03011012077331543,
"diffusion": 21.696259260177612,
"latent2audio": 1.7648537158966064
},
"actual_seeds": [
3776541388
],
"retake_seeds": [
4274500599
],
"retake_variance": 0.5,
"guidance_scale_text": 0,
"guidance_scale_lyric": 0,
"repaint_start": 0,
"repaint_end": 0,
"edit_n_min": 0.0,
"edit_n_max": 1.0,
"edit_n_avg": 1,
"src_audio_path": null,
"edit_target_prompt": null,
"edit_target_lyrics": null,
"audio2audio_enable": false,
"ref_audio_strength": 0.5,
"ref_audio_input": null,
"audio_path": "./outputs/output_20250512114703_0.wav"
}

File: examples/zh_rap_lora/input_params/output_20250512115409_0_input_params.json Copy Chapter

{
"lora_name_or_path": "/root/sag_train/data/ace_step_v1_chinese_rap_lora",
"task": "text2music",
"prompt": "electronic, hip-hop, rap, synthesizer, drums, vocals, fast, energetic, modern, uplifting, young adult, male, spoken word, singing, bright, energetic, clear, 140 bpm, female",
"lyrics": "[Verse 1]\n红鲤鱼绿鲤鱼，驴在河里追鲤鱼，\n驴追鲤鱼鱼躲驴，气得驴子直喘气。\n扁担长板凳宽，扁担绑在板凳边，\n扁担要绑板凳不让绑，扁担偏要绑上板凳面！\n\n[Chorus]\n绕口令，练嘴皮，\n说快说慢别迟疑，\n红鲤鱼驴扁担板凳，\n一口气念完算你赢！\n\n[Verse 2]\n四是四十是十，十四是十四四十是四十，\n谁说四十是十四，舌头打结别放肆。\n黑化肥会挥发，灰化肥也发黑，\n化肥混一起，黑灰不分嘴发废！\n\n[Chorus]\n绕口令，练嘴皮，\n说快说慢别迟疑，\n四十十四化肥灰，\n念错罚你唱十回！\n\n[Bridge]\n坡上立着一只鹅，坡下流着一条河，\n鹅要过河河渡鹅，河要渡鹅鹅笑河——\n到底谁更啰嗦？！\n\n[Outro]\n嘴皮子功夫别小瞧，\n绕口令rap我最飙，\n下次挑战准备好，\n舌头打结别求饶！",
"audio_duration": 123.2,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.7,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": false,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.026150941848754883,
"diffusion": 12.212433099746704,
"latent2audio": 1.1857895851135254
},
"actual_seeds": [
1415752189
],
"retake_seeds": [
685932970
],
"retake_variance": 0.5,
"guidance_scale_text": 0,
"guidance_scale_lyric": 0,
"repaint_start": 0,
"repaint_end": 0,
"edit_n_min": 0.0,
"edit_n_max": 1.0,
"edit_n_avg": 1,
"src_audio_path": null,
"edit_target_prompt": null,
"edit_target_lyrics": null,
"audio2audio_enable": false,
"ref_audio_strength": 0.5,
"ref_audio_input": null,
"audio_path": "./outputs/output_20250512115409_0.wav"
}

File: examples/zh_rap_lora/input_params/output_20250512120348_0_input_params.json Copy Chapter

{
"lora_name_or_path": "/root/sag_train/data/ace_step_v1_chinese_rap_lora",
"task": "text2music",
"prompt": "singing, bright, slightly nasal, energetic, spoken word, young adult, male, rap music",
"lyrics": "[Intro]\nYo, check it—speed demon, lyrical heat, uh!\nRatatat like a drum when the beat bumps, uh!\n\n[Verse 1]\nRapatapa tap tap, flash like a snap,\nRap tap tap, I don’t chat, I clap clap clap!\nFingers snap, flow don’t slack, rapataptaptap,\nSpit it fast, hit the gas, rap tap tap rap!\n\n[Pre-Chorus]\nBoom-bap, zoom past, leave ’em flat,\nRap taptaprapataptaptap—where ya at?\n\n[Chorus]\nRapatapa tap tap, yeah, I go brrrr,\nRap tap tap, make the crowd stir!\nRapataptaptap, no lag, just spit,\nRap taptaprapataptaptap—I’m lit!\n\n[Verse 2]\nTongue-twist, quick wrist, rapatapa boom,\nTap tap rap, leave ya stuck like glue-gum!\nNo slow-mo, turbo, rapataptaptap,\nRap tap rap, yeah, I clap clap clap!\n\n[Outro]\nRapatapa—TAP! Mic drop—that’s that.",
"audio_duration": 60,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": false,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.018491744995117188,
"diffusion": 8.084580898284912,
"latent2audio": 0.5694489479064941
},
"actual_seeds": [
226581098
],
"retake_seeds": [
1603201617
],
"retake_variance": 0.5,
"guidance_scale_text": 0,
"guidance_scale_lyric": 0,
"repaint_start": 0,
"repaint_end": 0,
"edit_n_min": 0.0,
"edit_n_max": 1.0,
"edit_n_avg": 1,
"src_audio_path": null,
"edit_target_prompt": null,
"edit_target_lyrics": null,
"audio2audio_enable": false,
"ref_audio_strength": 0.5,
"ref_audio_input": null,
"audio_path": "./outputs/output_20250512120348_0.wav"
}

File: examples/zh_rap_lora/input_params/output_20250512143242_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512145057_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512152217_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512153616_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512154907_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512161832_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512164224_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512171227_0_input_params.json Copy Chapter

{
"lora_name_or_path": "ACE-Step/ACE-Step-v1-chinese-rap-LoRA",
"task": "text2music",
"prompt": "Rap, Chinese Rap, J-Pop, Anime, kawaii pop, EDM, Aggressive, Intense, Crisp Snare, Super Fast, Clear",
"lyrics": "(Intro)\nLet's drift away...\n\n(Verse 1)\n现实是灰色的格子间，重复的工作，枯燥的报表 \n敲打着键盘，眼神却放空，意识早已挣脱了肉体的镣铐\n飘向窗外，飞过拥挤的街道，穿过云层，到达想象的群岛\n那里色彩斑斓，形状奇异，逻辑失效，一切都随心所欲地飘摇\n迷幻的鼓点，像心跳的变奏，忽快忽慢，难以预料\n抽象的采样，扭曲的人声，构建一个超现实的音景环绕\n我变成一只鸟，一条鱼，一束光，自由地变换形态和奔跑\n在这白日梦里，我无所不能，摆脱了所有现实的烦恼， feeling the afterglow\n\n(Chorus)\n意识漫游，逃离乏味的轨道 \n迷幻嘻哈的节拍，是白日梦的引导 \n抽象的世界，逻辑被重新构造\nMind wandering free, where reality starts to fade slow\n\n(Verse 2)\n会议室里老板在讲话，声音模糊，像隔着水听不清道\n我的思绪，早已潜入深海，与发光的水母一起舞蹈\n或者飞向外太空，在星云间穿梭，探索未知的星球和轨道\n现实的规则，在这里被打破，物理定律也失去效劳\n白日梦是我的避难所，是精神的氧气罩\n在乏味的现实里，为我注入一点色彩和奇妙\n虽然短暂，虽然虚幻，但它让我能够喘息，重新把能量找到\n然后回到现实，继续扮演那个，循规蹈矩的角色，把梦藏好， keep the dream aglow\n\n(Chorus)\n意识漫游，逃离乏味的轨道\n迷幻嘻哈的节拍，是白日梦的引导\n抽象的世界，逻辑被重新构造\nMind wandering free, where reality starts to fade slow\n",
"audio_duration": 153.7148,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": false,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.04823446273803711,
"diffusion": 13.158645629882812,
"latent2audio": 1.493880033493042
},
"actual_seeds": [
2945962357
],
"retake_seeds": [
2676242300
],
"retake_variance": 0.5,
"guidance_scale_text": 0.7,
"guidance_scale_lyric": 1.5,
"repaint_start": 0,
"repaint_end": 0,
"edit_n_min": 0.0,
"edit_n_max": 1.0,
"edit_n_avg": 1,
"src_audio_path": null,
"edit_target_prompt": null,
"edit_target_lyrics": null,
"audio2audio_enable": false,
"ref_audio_strength": 0.5,
"ref_audio_input": null,
"audio_path": "./outputs/output_20250512171227_0.wav"
}

File: examples/zh_rap_lora/input_params/output_20250512171809_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250512172941_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250513044511_0_input_params.json Copy Chapter

{
"lora_name_or_path": "/root/sag_train/data/ace_step_v1_chinese_rap_lora_100k",
"task": "text2music",
"prompt": "东北话, spoken word, male, Tempo - Fast, Elements - Chorus Hook, Subgenre-Satirical Hip Hop, Rap, Chinese-language music, energetic, slightly nasal, Instrument - Live Bass Guitar, adult, Vocals - Syncopated Flow, Genre - Hip-Hop, rapping, bright",
"lyrics": "[verse]\n挣着 憋屈的 工资 还得 装乐呵\n猫着 怂样儿 还搁 朋友圈 嘚瑟\n扛着 傻逼的 指标 没人 搭把手\n这儿 不是 托儿所 少整 那出儿 哭唧尿嚎\n\n俺们 就像 一条条 老板的 裤衩子\n陪着 笑脸 接他 每一回 突突\n哎呦 老板 今儿个 穿我呗\n他 撅个腚 眼角 瞟你 那熊样\n\n[chorus]\n他们 骂我 打工仔 太多人 没睡醒\n寻思 抠搜 老板 一天天 穷折腾\n不想 俺的 人生 烂在 这嘎达\n不想 俺的 将来 折在 这破棚\n\n老子 不想 上班 老子 是外星人\n你都 把俺 骂急眼了 俺还 这么淡定\n现实 才是 梦 啥时候 能醒啊\n那 糟践人的 答案 在西北风 里飘\n\n[verse]\n瞅见 二愣子 同事 给老板 舔腚沟子\n瞅见 浪蹄子 女同事 在老板 胯骨轴 扭搭\n瞅见 白瞎的 光阴 耗在 没亮儿的 道儿\n瞅见 公交车上 一帮 僵尸 吐酸水\n\n瞅见 俺的 命 定在 苦逼的 坑里\n瞅见 俺的 爱情 被轮了 成了 老处女\n瞅见 好事儿 全归 高富帅\n还有 那些 臭不要脸 扭腚的 货色\n\n[chorus]（重复）\n他们 骂我 打工仔 太多人 没睡醒...\n\n[bridge]\n加班 没补助 俺认了\n欠薪 揍员工 把俺 当牲口\n去你妈 的小姘头\n\n[verse]\n破逼 管理制度 净整 娱乐八卦\n撸管式 管理 也就 你自己 嗨\n出点儿 屁事儿 就往 下属 脑瓜子 扣\n挣俩 钢镚儿 立马 牛逼 不分 公母\n\n你挖个 大坑 把俺们 往里 踹\n说这 叫梦想 你当年 多能耐\n俺们 就当 听传销 洗脑课\n可怜 连骗人 你都 就会 这一套\n\n[outro]\n老子 不想 上班\n老子 不想 上班\n老子 不想 上班",
"audio_duration": 135.92,
"infer_step": 60,
"guidance_scale": 15,
"scheduler_type": "euler",
"cfg_type": "apg",
"omega_scale": 10,
"guidance_interval": 0.5,
"guidance_interval_decay": 0,
"min_guidance_scale": 3,
"use_erg_tag": true,
"use_erg_lyric": false,
"use_erg_diffusion": true,
"oss_steps": [],
"timecosts": {
"preprocess": 0.06204533576965332,
"diffusion": 35.75483560562134,
"latent2audio": 1.5193355083465576
},
"actual_seeds": [
4176354214
],
"retake_seeds": [
601086915
],
"retake_variance": 0.5,
"guidance_scale_text": 0,
"guidance_scale_lyric": 0,
"repaint_start": 0,
"repaint_end": 0,
"edit_n_min": 0.0,
"edit_n_max": 1.0,
"edit_n_avg": 1,
"src_audio_path": null,
"edit_target_prompt": null,
"edit_target_lyrics": null,
"audio2audio_enable": false,
"ref_audio_strength": 0.5,
"ref_audio_input": null,
"audio_path": "./outputs/output_20250513044511_0.wav"
}

File: examples/zh_rap_lora/input_params/output_20250513050200_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250513055451_0_input_params.json Copy Chapter

[Binary file]

File: examples/zh_rap_lora/input_params/output_20250513060150_0_input_params.json Copy Chapter

[Binary file]

File: zh_lora_dataset/dataset_info.json Copy Chapter

{
"citation": "",
"description": "",
"features": {
"keys": {
"dtype": "string",
"_type": "Value"
},
"filename": {
"dtype": "string",
"_type": "Value"
},
"tags": {
"feature": {
"dtype": "string",
"_type": "Value"
},
"_type": "Sequence"
},
"speaker_emb_path": {
"dtype": "string",
"_type": "Value"
},
"norm_lyrics": {
"dtype": "string",
"_type": "Value"
},
"recaption": {}
},
"homepage": "",
"license": ""
}

File: zh_lora_dataset/state.json Copy Chapter

{
"_data_files": [
{
"filename": "data-00000-of-00001.arrow"
}
],
"_fingerprint": "76c203d4bc1fdd7e",
"_format_columns": null,
"_format_kwargs": {},
"_format_type": null,
"_output_all_columns": false,
"_split": null
}