Skip to content

Bike Sharing Forecasting Zero-shot, Fine-tuning, and Performance Evaluation

TinyTimeMixers (TTMs) are compact pre-trained models for Multivariate Time-Series Forecasting, open-sourced by IBM Research. With less than 1 Million parameters, TTM introduces the notion of the first-ever "tiny" pre-trained models for Time-Series Forecasting. TTM outperforms several popular benchmarks demanding billions of parameters in zero-shot and few-shot forecasting and can easily be fine-tuned for multivariate forecasts.

In this lab, we cover zero-shot forecasting, as well as fine-tuning. This example makes use of the Kaggle bike sharing dataset which contains bikes rental demand with weather and seasonal information.

Prerequisites

This lab is a Jupyter notebook. Please follow the instructions in pre-work to run the lab.

Lab

Bike Sharing Forecasting Zero-shot, Fine-tuning, and Performance Evaluation notebook Bike Sharing Forecasting Zero-shot, Fine-tuning, and Performance Evaluation notebook

To run the notebook from your command line in Jupyter using the active virtual environment from the pre-work, run:

jupyter notebook notebooks/Bike_Sharing_Finetuning_with_Exogenous.ipynb

The path of the notebook file above is relative to the granite-workshop folder from the git clone in the pre-work.

Credits

This notebook is a modified version of the IBM Granite Community Bike Sharing Forecasting Zero-shot, Fine-tuning, and Performance Evaluation notebook. Refer to the IBM Granite Community for the official notebooks.