Note
Sentis is now called Inference Engine. The documentation has moved to https://docs.unity3d.com/Packages/com.unity.ai.inference@latest. Refer to the new location for the latest updates and guidance. Make sure to update your bookmarks and references accordingly.
Run a model
After you create a worker, call Schedule to run the model.
worker.Schedule(inputTensor);
The first scheduling of a model within the Unity Editor might be slow as Sentis needs to compile code and shaders as well as allocating internal memory. Subsequent runs will be faster due to caching. Including a test run when starting the application is a good idea.
For an example, refer to the Run a model sample in the sample scripts.