Deploying models
There are several ways to deploy a model in BaseTen, from simple one-liners with some model frameworks, pre-trained models via the UI, and complete freedom with custom models.

💻 Installing the python client

If you've trained your own model, you need the BaseTen package installed in your python environment. It's best to install BaseTen in the same environment that you use to train your models. The three steps for installing and configuring the BaseTen client are:
  1. 1.
    pip install baseten
  2. 2.
    baseten login It needs an API key. Generate one on the app's settings page, under "Authorization".
  3. 3.
    [Optional] baseten configure If you're self-hosting BaseTen, run this command to provide the client with the app's base URI. It will likely be similar to

🔮 ➡️ ☁️ Deploying an ML model

Currently Scikit-learn, Keras, and PyTorch models are supported out of the box, for everything else, there's Custom Models

🔄 Updating an ML model

To add a new version of an existing model, simply call baseten.deploy passing it the new model object and the same model_name as the existing model.
Last modified 2mo ago