Tutorial

The linked Jupyter notebooks proivde an interactive introduction to the TACO Python library, including how to initialize tensors, define mode formats, and perform computations. There are three notebooks, which differ mainly in the final extended example: SpMV (useful for scientific computing), SDDMM (machine learning), and MTTKRP (data analytics).

These notebooks are hosted online and do not require any installation; i.e., they can be run without having TACO, Jupyter, or even Python installed locally. However, they may take a minute or two to build.

If, on the other hand, you would like to run the notebooks on your computer, please do the following:

1. Install Python 3.
2. Intall Jupyter.
3. Install NumPy and SciPy.
4. Install TACO. Follow the instructions to build TACO with the Python API.
5. Clone the repository with the tutorials.