Using a trained model for inference/future-testing

Flow Forecast Inference mode allows you to use a trained model to easily make predictions on new data . This is useful for model deployment, running additional ablation tests, and much more.

ReadTheDocs

Inference Mode Basic Notebook

Inference Mode Advanced Notebook/Abalation Tests

Steps

1. Loading the model: Before running the model you first will need to re-load the model into a PyTorchForecast class. This is very easy to do via our Google Cloud Provider integration or you can do it manually if you are currently using a different cloud provider.

2. After the model is loaded there a couple different methods that you can a

Summary