PolarDB:Procedure to load an external model

With the MLOps capabilities provided by PolarDB for AI, you can also upload, deploy, and use externally trained models. A wide array of external models are supported, such as sklearn-compatible LightGBM and GBDT and algorithms created using deep learning frameworks like TensorFlow and PyTorch.

To load an external model, do the following steps:

• Upload a model

• Deploy a model

• Evaluate a model

• Perform model inference