I will explain automatic differentiation for the gifted amateur and how to use it in the PyTorch framework. You don't have to know the PyTorch framework but some experience with NumPy could help to follow the tutorial session. I would like to keep the session interactive but you will not need to run code on your laptops. Instead, I will present code snippets, discuss them and thus introduce some common pitfalls when using PyTorch. In the end I hope that you understand better what happens behind the curtain of PyTorch (and other autodiff libraries), helping you to more effectively debug your every day code.


    Access

    Download


    Export, share and cite



    Title :

    PyTorch and automatic differentiation


    Contributors:

    Conference:

    2021 ; Köln, Deutschland


    Publication date :

    2021-10-11


    Type of media :

    Conference paper


    Type of material :

    Electronic Resource


    Language :

    English




    Parallelized Forward Kinematics Using Product of Exponentials in PyTorch

    Vrigazov, Hristo / Yovchev, Kaloyan | TIBKAT | 2022


    Parallelized Forward Kinematics Using Product of Exponentials in PyTorch

    Vrigazov, Hristo / Yovchev, Kaloyan | Springer Verlag | 2022



    Automatic Differentiation in Robust Optimization

    Su, J. / Renaud, J. / American Institute of Aeronautics and Astronautics et al. | British Library Conference Proceedings | 1996