Prebuilt Models
Deploying Prebuilt Models

You can deploy prebuilt models via Cerebrium using two methods.

  1. Doing simple one-click deploys from your dashboard
  2. Through our Python framework.

Deploying a prebuilt model through our framework is largely similar to deploying a custom model. The only difference is that you do not need to upload a model file. Instead, you specify which prebuilt model you want to deploy using the corresponding identifier.

Deploy a prebuilt Whisper Model
from cerebrium import deploy, model_type
endpoint = deploy((model_type.PREBUILT, "whisper-medium"), 'my-flow', "<API_KEY>")

You can not use prebuilt models in ensembles or with processing functions. They should be contained within single model flows only.

You can check out the available models through your dashboard or through our docs below