zsxkib/flux-music 🔢📝❓✓ → ❓

▶️ 8.6K runs 📅 Sep 2024 ⚙️ Cog 0.9.24 🔗 GitHub 📄 Paper ⚖️ License

music-generation text-to-music

About

🎼FluxMusic Text-to-Music Generation with Rectified Flow Transformer🎶

Example Output

Prompt:

"A wild and rebellious rock anthem with psychedelic and bluesy undertones, this song is loud and energetic."

Output

{"wav":"https://replicate.delivery/pbxt/8WrfHSnuvs0kNqL4x9tx0ZrOITA7lryic6Vit0CaeCUN3ygTA/output.wav","melspectrogram":"https://replicate.delivery/pbxt/eDxBMtguqHWjU6zGezLBsab4Ck5dG7dG3M6IOwnmNqMO3ygTA/spectrogram.png"}

Performance Metrics

5.82s Prediction Time

173.68s Total Time

All Input Parameters

{
  "steps": 50,
  "prompt": "A wild and rebellious rock anthem with psychedelic and bluesy undertones, this song is loud and energetic.",
  "model_version": "base",
  "guidance_scale": 7,
  "negative_prompt": "low quality, gentle",
  "save_spectrogram": true
}

Input Parameters

seed Type: integer: Random seed. Leave blank to randomize the seed
steps Type: integerDefault: 50Range: 1 - 200: Number of sampling steps
prompt Type: stringDefault: The song is an epic blend of space-rock, rock, and post-rock genres.: Text prompt for music generation
model_version Default: base: Select the model version to use
guidance_scale Type: numberDefault: 7Range: 0 - 20: Classifier-free guidance scale
negative_prompt Type: stringDefault: low quality, gentle: Text prompt for negative guidance (unconditioned prompt)
save_spectrogram Type: booleanDefault: false: Whether to save the spectrogram image

Output Schema

wav Type: stringFormat: uri: Wav
melspectrogram Type: stringFormat: uri: Melspectrogram

Example Execution Logs

Using seed: 1535524279

Version Details

Version ID: eebfed4a1749bb1172f005f71fac5a1e0377502ec149c9d02b56ac1de3aa9f07
Version Created: September 26, 2024

Run on Replicate →