Instructions to use Nekochu/stable-audio-open-1.0-Music with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Stable Audio Tools
How to use Nekochu/stable-audio-open-1.0-Music with Stable Audio Tools:
import torch import torchaudio from einops import rearrange from stable_audio_tools import get_pretrained_model from stable_audio_tools.inference.generation import generate_diffusion_cond device = "cuda" if torch.cuda.is_available() else "cpu" # Download model model, model_config = get_pretrained_model("Nekochu/stable-audio-open-1.0-Music") sample_rate = model_config["sample_rate"] sample_size = model_config["sample_size"] model = model.to(device) # Set up text and timing conditioning conditioning = [{ "prompt": "128 BPM tech house drum loop", }] # Generate stereo audio output = generate_diffusion_cond( model, conditioning=conditioning, sample_size=sample_size, device=device ) # Rearrange audio batch to a single sequence output = rearrange(output, "b d n -> d (b n)") # Peak normalize, clip, convert to int16, and save to file output = output.to(torch.float32).div(torch.max(torch.abs(output))).clamp(-1, 1).mul(32767).to(torch.int16).cpu() torchaudio.save("output.wav", output, sample_rate) - Notebooks
- Google Colab
- Kaggle
You can use this model in stable-audio-tools, fine-tuned on my favorite song from my personal playlist.
Comparison Table
| Prompt | Base Model | Fine-Tuned |
|---|---|---|
| Feel-Good Vibes and Dramatic Atmosphere, alone hero, epic, get good yeah, better last night pop, follow follow, echoing, powerful vocal driving melancholic vocals dramatic Features rising tension, progressive electro house, far away, by Alan Walker, popular song tempo, girl, female synth, popular, titled: legend never die | ||
| Beautiful music progressive electro slap mood, upbeat, heavy bass, melancholic, hopeful; drums, vocals, dynamic shifts, building intensity, run far away, repetitive, let let go, think of us, titled popular lyrics: Mirror's Edge, popular lyrics say: "still still alive" | ||
| Chill soft wake up, slow down alt, night get lucky dance, relax music introspective 2017 2018 2019 2020 2021 2022, acoustic atmosphere uplifting dreams, dreamy indie pop, electric trap, percussion, higher reverb, really intensity melody, goodbye | ||
| Chill hip-hop beat, chillhop, lofi pop, favorite music |
Showcase Model Details
Test Settings:
- CFG: 7.0
- Steps: 100
- Seed: -1
Prompt have been chosen based on the top tagged words except last prompt which is used to compare effect on non-trained tags

- Prompt
- Chill soft wake up, slow down alt, night get lucky dance, relax music introspective 2017 2018 2019 2020 2021 2022, acoustic atmosphere uplifting dreams, dreamy indie pop, electric trap, percussion, higher reverb, really intensity melody, goodbye
- Negative Prompt
- blurry, cropped, ugly

- Prompt
- Chill hip-hop beat, chillhop, lofi pop, favorite music
- Negative Prompt
- blurry, cropped, ugly
Training
Dataset: 2-3 min music length
- All of my Liked music download and auto label so mostly copyright.
- Total number of samples: ~1383
"random_crop": truein dataset_config.json
Settings:
- Training epochs: 1
- Training steps: 1383
- Learning rate: 1e-05
- Downloads last month
- 272