/Tensor-library

Analogue to Math and ML libraries like numpy or pytorch

Primary LanguageJavaMIT LicenseMIT

Tensor-library

Analogue to Math and ML libraries like numpy or pytorch
From scratch in plain java

Dig deeper into the library

Features

  • Tensors
  • Operations: transposition, matrix multiplication, convolution etc.
  • Convenient creation of your own operations
  • Dynamic computation graphs
  • AutoGrad

Usage Tips

Disadvantage

Element-by-element execution of operations, without GPU acceleration of calculations.
I would like to add this feature, but learning OpenCL will take a lot of time.

Explanation why I won't continue

"The problem with programmers
is that when they make a car,
they are at the same time, reinvent
the wheel, steel mining methods and traffic rules"

But maybe someday...

References:

  • Weidman, S. (2019). Deep learning from scratch: Building with Python from first principles (First edition). O’Reilly Media, Inc.
  • Patterson, J., & Gibson, A. (2017). Deep learning: A practitioner’s approach (First edition). O’Reilly.
  • Евгений Разинков. (2021). Лекции по Deep Learning. https://www.youtube.com/playlist?list=PL6-BrcpR2C5QrLMaIOstSxZp4RfhveDSP
  • Raschka, S., Liu, Y., Mirjalili, V., & Dzhulgakov, D. (2022). Machine learning with PyTorch and Scikit-Learn: Develop machine learning and deep learning models with Python. Packt.
  • Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. The MIT press.
  • Rashid, T. (2016). Make Your own neural network. CreateSpace Independent Publishing Platform.

Licence

MIT License