/PyTorch-Padawan

Exercises, Descriptions, and Visualizations to build intuitions and confidence in working with PyTorch for accelerated Scientific Computing

Primary LanguageJupyter Notebook

PyTorch Padawan

The grand goal of this repository is to offer an in-depth understanding of PyTorch for Scientific Computing. In particular, we will focus mainly on solving problems while also explaining the steps involved which will then offer good intuitions and confidence for working with tensors of arbitrary shapes.
Topic Description nbviewer colab github
einsum explanation and examples about processing tensors using einsum
tensor manipulation explanation and examples about general manipulations on the tensors
vectorization explanation and examples about how to write vectorized code in order to achieve memory & runtime efficiency
broadcasting explanation and examples for writing code without duplicating data to achieve memory & runtime efficiency in progress in progress in progress
multiple linear regression explanation and implementation of multiple linear regression with regularization & hyperparameter tuning in progress in progress in progress
linear algebra barebones implementation of common linear algebra operations with intuitive explanations planned planned planned
signal processing implementation interspersed with explanations of fundamental principles for processing signals planned planned planned
Acknowledgement

Special thanks 🙏 to my dear friend Kata Naszádi for introducing me the name and letting me to use it here.