MusicGen music generation for Unity Sentis (Version 1.3.0-pre.3 Only*)

*Version 1.3 Sentis files will not work with version 1.4.0 or above

This is the Meta MusicGen model validated for Unity Sentis. This model creates up to 30 seconds of music in the style given by a text prompt. For example, you could use it to generate original music tracks for your games.

How to run this model

  • Put the sentis and json files in the Assets/StreamingAssets folder
  • Install com.unity.sentis version 1.3.0-pre.3 and com.unity.nuget.newtonsoft-json packages.
  • Put the c# script on the Main Camera
  • Put an AudioSource on the Main Camera
  • Change the prompt in the script file
  • Change the number of seconds in the script file
  • Press play and check for results in the console window

Info

This model is a bit slow at the moment. Please use our discussion forum to discuss this model or suggest improvements to the code.

Downloads last month
58
Inference Examples
Inference API (serverless) does not yet support unity-sentis models for this pipeline type.